d/freebsd-dev
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
9516bcdfb4
freebsd-dev/sys/compat/linux/linux_ioctl.c

3808 lines
94 KiB
C
Raw Normal View History

Start each of the license/copyright comments with /*-
2005-01-05 22:34:37 +00:00
/*-
Rationalize license text on Linuxolator files Many licenses on Linuxolator files contained small variations from the standard FreeBSD license text. To avoid license proliferation switch to the standard 2-clause FreeBSD license for those files where I have permission from each of the listed copyright holders. Additional files waiting on permission from others are listed in review D14210. Approved by: kan, marcel, sos, rdivacky MFC after: 1 week Sponsored by: The FreeBSD Foundation
2018-02-16 15:00:14 +00:00
* SPDX-License-Identifier: BSD-2-Clause-FreeBSD
sys/compat: further adoption of SPDX licensing ID tags. Mainly focus on files that use BSD 2-Clause license, however the tool I was using misidentified many licenses so this was mostly a manual - error prone - task. The Software Package Data Exchange (SPDX) group provides a specification to make it easier for automated tools to detect and summarize well known opensource licenses. We are gradually adopting the specification, noting that the tags are considered only advisory and do not, in any way, superceed or replace the license texts.
2017-11-27 15:13:23 +00:00
*
Convert files to UTF-8
2012-01-15 13:23:18 +00:00
* Copyright (c) 1994-1995 Søren Schmidt
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
Rationalize license text on Linuxolator files Many licenses on Linuxolator files contained small variations from the standard FreeBSD license text. To avoid license proliferation switch to the standard 2-clause FreeBSD license for those files where I have permission from each of the listed copyright holders. Additional files waiting on permission from others are listed in review D14210. Approved by: kan, marcel, sos, rdivacky MFC after: 1 week Sponsored by: The FreeBSD Foundation
2018-02-16 15:00:14 +00:00
* notice, this list of conditions and the following disclaimer.
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
*
Rationalize license text on Linuxolator files Many licenses on Linuxolator files contained small variations from the standard FreeBSD license text. To avoid license proliferation switch to the standard 2-clause FreeBSD license for those files where I have permission from each of the listed copyright holders. Additional files waiting on permission from others are listed in review D14210. Approved by: kan, marcel, sos, rdivacky MFC after: 1 week Sponsored by: The FreeBSD Foundation
2018-02-16 15:00:14 +00:00
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
*/
All opt_x.h includes go at the top of other includes.
2008-08-25 04:55:29 +00:00
#include "opt_compat.h"
Use __FBSDID().
2003-06-10 21:29:12 +00:00
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
#include <sys/param.h>
#include <sys/systm.h>
Completed function declarations and added prototypes. Removed some unnecessary #includes. Fixed warnings about nested externs.
1995-11-22 07:43:53 +00:00
#include <sys/sysproto.h>
Centralize compatability translation macros. Copy the CP, PTRIN, etc macros from freebsd32.h into a sys/abi_compat.h and replace existing definitation with includes where required. This eliminates duplicate code and allows Linux and FreeBSD compatability headers to be included in the same files. Input from: cem, jhb Obtained from: CheriBSD MFC after: 2 weeks Sponsored by: DARPA Differential Revision: https://reviews.freebsd.org/D24275
2020-04-14 20:30:48 +00:00
#ifdef COMPAT_LINUX32
#include <sys/abi_compat.h>
#endif
Update kernel inclusions of capability.h to use capsicum.h instead; some further refinement is required as some device drivers intended to be portable over FreeBSD versions rely on __FreeBSD_version to decide whether to include capability.h. MFC after: 3 weeks
2014-03-16 10:55:57 +00:00
#include <sys/capsicum.h>
- Handle mixer read ioctls correctly. They have the same group, number and argument size as their write counterparts and were handled as write ioctls. - Emulate some cdrom ioctls.
1999-04-29 04:37:57 +00:00
#include <sys/cdio.h>
Implement DVD-ROM ioctls. PR: 26955 Submitted by: Boris Nikolaus (email unknown)
2001-11-18 18:49:07 +00:00
#include <sys/dvdio.h>
Add #include <sys/conf.h> so we don't rely on <sys/disk.h> doing it.
2003-04-01 12:34:47 +00:00
#include <sys/conf.h>
Fix an XXX: and implement LINUX_BLKGETSIZE correctly.
2003-03-28 08:58:11 +00:00
#include <sys/disk.h>
Initiate deorbit burn sequence for <machine/console.h>. Replace all in-tree uses with necessary subset of <sys/{fb,kb,cons}io.h>. This is also the appropriate fix for exo-tree sources. Put warnings in <machine/console.h> to discourage use. November 15th 2000 the warnings will be converted to errors. January 15th 2001 the <machine/console.h> files will be removed.
2000-10-08 21:34:00 +00:00
#include <sys/consio.h>
Try to make Linux socket ioctls work. Up until now they've only *pretended* to work, but haven't really due to subtle differences in structs etc. This is still not perfect (some ioctls are still known not to work, while others haven't been tested at all), but it's enough to get Debian's ifconfig to produce relatively sane output. More work will be needed to get all ioctls (or at least a reasonable subset) working, and to support the Cisco Aironet config tool mentioned in the PR. PR: 26546 Submitted by: Doug Ambrisko <ambrisko@ambrisko.com>
2001-10-15 20:52:17 +00:00
#include <sys/ctype.h>
Don't #include <sys/fcntl.h> in <sys/file.h> if KERNEL is defined. Fixed everything that depended on getting fcntl.h stuff from the wrong place. Most things don't depend on file.h stuff at all.
1997-03-23 03:37:54 +00:00
#include <sys/fcntl.h>
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
#include <sys/file.h>
#include <sys/filedesc.h>
Don't include <sys/ioctl.h> in the kernel. Stage 3: include <sys/filio.h> instead of <sys/ioctl.h> in non-network non-tty files.
1997-03-24 11:37:53 +00:00
#include <sys/filio.h>
Add explicit includes for jail.h to the files that need them and remove the "hidden" one from vimage.h.
2009-06-17 15:01:01 +00:00
#include <sys/jail.h>
Initiate deorbit burn sequence for <machine/console.h>. Replace all in-tree uses with necessary subset of <sys/{fb,kb,cons}io.h>. This is also the appropriate fix for exo-tree sources. Put warnings in <machine/console.h> to discourage use. November 15th 2000 the warnings will be converted to errors. January 15th 2001 the <machine/console.h> files will be removed.
2000-10-08 21:34:00 +00:00
#include <sys/kbio.h>
- Protect the list of linux ioctl handlers with an sx lock. - Hold Giant while calling linux ioctl handlers for now as they aren't all known to be MPSAFE yet. - Mark linux_ioctl() MPSAFE.
2006-07-06 21:42:36 +00:00
#include <sys/kernel.h>
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
#include <sys/linker_set.h>
- Protect the list of linux ioctl handlers with an sx lock. - Hold Giant while calling linux ioctl handlers for now as they aren't all known to be MPSAFE yet. - Mark linux_ioctl() MPSAFE.
2006-07-06 21:42:36 +00:00
#include <sys/lock.h>
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
#include <sys/malloc.h>
Try to make Linux socket ioctls work. Up until now they've only *pretended* to work, but haven't really due to subtle differences in structs etc. This is still not perfect (some ioctls are still known not to work, while others haven't been tested at all), but it's enough to get Debian's ifconfig to produce relatively sane output. More work will be needed to get all ioctls (or at least a reasonable subset) working, and to support the Cisco Aironet config tool mentioned in the PR. PR: 26546 Submitted by: Doug Ambrisko <ambrisko@ambrisko.com>
2001-10-15 20:52:17 +00:00
#include <sys/proc.h>
Rewrite linux_ifconf() to be more like ifconf() in net/if.c so that we do not call uiomove() while IFNET_RLOCK() is held. This eliminates the witness warning: Calling uiomove() with the following non-sleepable locks held: exclusive sleep mutex ifnet r = 0 (0xc096dd60) locked @ /usr/src/sys/modules/linux/../../compat/linux/linux_ioctl.c:2170 MFC after: 2 days
2005-08-27 14:44:10 +00:00
#include <sys/sbuf.h>
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
#include <sys/socket.h>
Try to make Linux socket ioctls work. Up until now they've only *pretended* to work, but haven't really due to subtle differences in structs etc. This is still not perfect (some ioctls are still known not to work, while others haven't been tested at all), but it's enough to get Debian's ifconfig to produce relatively sane output. More work will be needed to get all ioctls (or at least a reasonable subset) working, and to support the Cisco Aironet config tool mentioned in the PR. PR: 26546 Submitted by: Doug Ambrisko <ambrisko@ambrisko.com>
2001-10-15 20:52:17 +00:00
#include <sys/sockio.h>
#include <sys/soundcard.h>
Add support for more linux ioctls. I've had this sitting in my tree for a long time and I can't seem to find who sent it to me in the first place, apologies to whoever is missing out on a Contributed by: line here. I belive it works as it should.
2004-06-14 07:26:23 +00:00
#include <sys/stdint.h>
- Protect the list of linux ioctl handlers with an sx lock. - Hold Giant while calling linux ioctl handlers for now as they aren't all known to be MPSAFE yet. - Mark linux_ioctl() MPSAFE.
2006-07-06 21:42:36 +00:00
#include <sys/sx.h>
Add FEATURE macros for v4l and v4l2 to the linuxulator. Suggested by: ae
2011-05-04 09:52:34 +00:00
#include <sys/sysctl.h>
Try to make Linux socket ioctls work. Up until now they've only *pretended* to work, but haven't really due to subtle differences in structs etc. This is still not perfect (some ioctls are still known not to work, while others haven't been tested at all), but it's enough to get Debian's ifconfig to produce relatively sane output. More work will be needed to get all ioctls (or at least a reasonable subset) working, and to support the Cisco Aironet config tool mentioned in the PR. PR: 26546 Submitted by: Doug Ambrisko <ambrisko@ambrisko.com>
2001-10-15 20:52:17 +00:00
#include <sys/tty.h>
#include <sys/uio.h>
This is v4l2 support for the linuxulator. This allows to access FreeBSD native devices which support the v4l2 API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd or multimedia/webcamd supplied drivers. Submitted by: nox MFC after: 1 month
2011-05-04 09:05:39 +00:00
#include <sys/types.h>
#include <sys/mman.h>
#include <sys/resourcevar.h>
Commit step 1 of the vimage project, (network stack) virtualization work done by Marko Zec (zec@). This is the first in a series of commits over the course of the next few weeks. Mark all uses of global variables to be virtualized with a V_ prefix. Use macros to map them back to their global names for now, so this is a NOP change only. We hope to have caught at least 85-90% of what is needed so we do not invalidate a lot of outstanding patches again. Obtained from: //depot/projects/vimage-commit2/... Reviewed by: brooks, des, ed, mav, julian, jamie, kris, rwatson, zec, ... (various people I forgot, different versions) md5 (with a bit of help) Sponsored by: NLnet Foundation, The FreeBSD Foundation X-MFC after: never V_Commit_Message_Reviewed_By: more people than the patch
2008-08-17 23:27:27 +00:00
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
#include <net/if.h>
The r48589 promised to remove implicit inclusion of if_var.h soon. Prepare to this event, adding if_var.h to files that do need it. Also, include all includes that now are included due to implicit pollution via if_var.h Sponsored by: Netflix Sponsored by: Nginx, Inc.
2013-10-26 17:58:36 +00:00
#include <net/if_var.h>
Add support for the SIOCGIFHWADDR ioctl, commonly used by license managers to obtain the host's ethernet address as a key. Note that this implementation takes the first hardware address for the first ethernet interface found, and disregards the interface name that may be passed in, as linux ethernet devices are all "ethX".
1997-06-02 06:31:49 +00:00
#include <net/if_dl.h>
#include <net/if_types.h>
Completed function declarations and added prototypes. Removed some unnecessary #includes. Fixed warnings about nested externs.
1995-11-22 07:43:53 +00:00
Add Evdev ioctl handler to the Linuxulator. PR: 218627 Submitted by: Jan Kokemüller Reported by: Jan Kokemüller MFC after: 1 week
2017-04-23 07:43:50 +00:00
#include <dev/evdev/input.h>
Bring USB fixes for linux(4). Intention of this commit is to let us take a full advantage of libusb(8) ported to Linux. This decreases a possibility of getting any collisions within ioctl() "command" space, especially with relation to LINUX_SNDCTL_SEQ... stuff. Basically, we provide commands, that will be mapped in the kernel to correct ones and forward those to the USB layer. Port enabling functionality brought with this patch is here: http://www.freebsd.org/cgi/query-pr.cgi?pr=146895 Bump __FreeBSD_version to catch, since which version installing a port makes sense. This patch should bring no regressions. So far, only i386 is tested. Tested by: thompsa@ Reviewed by: thompsa@ OKed by: netchild@
2010-05-24 07:04:00 +00:00
#include <dev/usb/usb_ioctl.h>
Match the LINUX32's style with existing style Submitted by: Jung-uk Kim <jkim@niksun.com> Use positive, not negative logic.
2005-01-14 04:44:56 +00:00
#ifdef COMPAT_LINUX32
Changes to MI Linux emulation code necessary to run 32-bit Linux binaries on AMD64, and the general case where the emulated platform has different size pointers than we use natively: - declare certain structure members as l_uintptr_t and use the new PTRIN and PTROUT macros to convert to and from native pointers. - declare some structures __packed on amd64 when the layout would differ from that used on i386. - include <machine/../linux32/linux.h> instead of <machine/../linux/linux.h> if compiling with COMPAT_LINUX32. This will need to be revisited before 32-bit and 64-bit Linux emulation support can coexist in the same kernel. - other small scattered changes. This should be a no-op on i386 and Alpha.
2004-08-16 07:28:16 +00:00
#include <machine/../linux32/linux.h>
#include <machine/../linux32/linux32_proto.h>
Match the LINUX32's style with existing style Submitted by: Jung-uk Kim <jkim@niksun.com> Use positive, not negative logic.
2005-01-14 04:44:56 +00:00
#else
#include <machine/../linux/linux.h>
#include <machine/../linux/linux_proto.h>
Changes to MI Linux emulation code necessary to run 32-bit Linux binaries on AMD64, and the general case where the emulated platform has different size pointers than we use natively: - declare certain structure members as l_uintptr_t and use the new PTRIN and PTROUT macros to convert to and from native pointers. - declare some structures __packed on amd64 when the layout would differ from that used on i386. - include <machine/../linux32/linux.h> instead of <machine/../linux/linux.h> if compiling with COMPAT_LINUX32. This will need to be revisited before 32-bit and 64-bit Linux emulation support can coexist in the same kernel. - other small scattered changes. This should be a no-op on i386 and Alpha.
2004-08-16 07:28:16 +00:00
#endif
The MI/MD split wasn't perfect and the MI files need hacks for the AlphaLinux compat bits. This will be better cleaned up soon. Agreed to what ever was necessary by: marcel
2000-11-01 19:48:35 +00:00
In order to reduce duplication between MD parts of the Linuxulator move bits that are MI out into the headers in compat/linux. For that remove bogus _packed attribute from struct l_sockaddr and use MI types for struct members. And continue to move into the linux_common module a code that is intended for both Linuxulator modules (both instruction set - 32 & 64 bit) or for external modules like linsysfs or linprocfs. To avoid header pollution introduce new sys/compat/linux_common.h header. Reviewed by: emaste MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D20137
2019-05-03 08:42:49 +00:00
#include <compat/linux/linux_common.h>
Update include directives.
2000-08-22 01:32:14 +00:00
#include <compat/linux/linux_ioctl.h>
#include <compat/linux/linux_mib.h>
To avoid excessive code duplication move MI definitions to the MI header file. As it is defined in Linux. Approved by: kib (mentor) MFC after: 1 month
2009-05-07 09:39:20 +00:00
#include <compat/linux/linux_socket.h>
Add Evdev ioctl handler to the Linuxulator. PR: 218627 Submitted by: Jan Kokemüller Reported by: Jan Kokemüller MFC after: 1 week
2017-04-23 07:43:50 +00:00
#include <compat/linux/linux_timer.h>
Update include directives.
2000-08-22 01:32:14 +00:00
#include <compat/linux/linux_util.h>
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
- Move videodev headers from compat/linux to contrib/v4l (cp from vendor and apply diff to compat/linux versions). - The cp implies an update of videodev2.h to the linux kernel 2.6.34.14 one. The update makes video in skype v4 work on FreeBSD. Tested by: Artyom Mirgorodskiy <artyom.mirgorodsky@gmail.com> (update of header only)
2013-07-06 19:59:06 +00:00
#include <contrib/v4l/videodev.h>
This is v4l support for the linuxulator. This allows to access FreeBSD native devices which support the v4l API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd driver. Not tested is firmware upload, framebuffer stuff and video tuner stuff due to lack of hardware. The clipping part (VIDIOCSWIN) needs a little bit of further work (partly in progress, but can not be tested due to lack of a suitable device). The submitter tested this sucessfully with Skype and flash apps on amd64 and i386 with the multimedia/pwcbsd driver. Submitted by: J.R. Oldroyd <fbsd@opal.com>
2009-12-04 21:06:54 +00:00
#include <compat/linux/linux_videodev_compat.h>
- Move videodev headers from compat/linux to contrib/v4l (cp from vendor and apply diff to compat/linux versions). - The cp implies an update of videodev2.h to the linux kernel 2.6.34.14 one. The update makes video in skype v4 work on FreeBSD. Tested by: Artyom Mirgorodskiy <artyom.mirgorodsky@gmail.com> (update of header only)
2013-07-06 19:59:06 +00:00
#include <contrib/v4l/videodev2.h>
This is v4l2 support for the linuxulator. This allows to access FreeBSD native devices which support the v4l2 API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd or multimedia/webcamd supplied drivers. Submitted by: nox MFC after: 1 month
2011-05-04 09:05:39 +00:00
#include <compat/linux/linux_videodev2_compat.h>
Overhaul CAM SG driver IOCTL interfaces. Make it really work for native FreeBSD programs. Before this it was broken for years due to different number of pointer dereferences in Linux and FreeBSD IOCTL paths, permanently returning errors to FreeBSD programs. This change breaks the driver FreeBSD IOCTL ABI, making it more strict, but since it was not working any way -- who bother. Add shims for 32-bit programs on 64-bit host, translating the argument of the SG_IO IOCTL for both FreeBSD and Linux ABIs. With this change I was able to run 32-bit Linux sg3_utils tools and simple 32 and 64-bit FreeBSD test tools on both 32 and 64-bit FreeBSD systems. MFC after: 1 month
2014-06-02 19:53:53 +00:00
#include <cam/scsi/scsi_sg.h>
Use the CTASSERT() macro instead of rolling my own, non-portable one using #error. Suggested by: jhb
2005-03-24 19:26:50 +00:00
CTASSERT(LINUX_IFNAMSIZ == IFNAMSIZ);
Compile errors are way more useful then panics later. Replace a KASSERT of LINUX_IFNAMSIZ == IFNAMSIZ with a preprocessor check and #error message. This will prevent nasty suprises if users change IFNAMSIZ without updating the linux code appropriatly.
2005-03-24 17:51:15 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
static linux_ioctl_function_t linux_ioctl_cdrom;
- Add support for IPC_64 extensions into shmctl(2), semctl(2) and msgctl(2); - add wrappers for mmap2(2) and ftruncate64(2) system calls; - don't spam console with printf's when VFAT_READDIR_BOTH ioctl(2) is invoked; - add support for SOUND_MIXER_READ_STEREODEVS ioctl(2); - make msgctl(IPC_STAT) and IPC_SET actually working by converting from BSD msqid_ds to Linux and vice versa; - properly return EINVAL if semget(2) is called with nsems being negative. Reviewed by: marcel Approved by: marcel Tested with: LSB runtime test
2002-10-11 11:43:09 +00:00
static linux_ioctl_function_t linux_ioctl_vfat;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
static linux_ioctl_function_t linux_ioctl_console;
Add support for more linux ioctls. I've had this sitting in my tree for a long time and I can't seem to find who sent it to me in the first place, apologies to whoever is missing out on a Contributed by: line here. I belive it works as it should.
2004-06-14 07:26:23 +00:00
static linux_ioctl_function_t linux_ioctl_hdio;
Fix some style bugs. The long line is in a chunk of code that's being rewritten, though. Submitted by: bde
2000-03-13 15:27:19 +00:00
static linux_ioctl_function_t linux_ioctl_disk;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
static linux_ioctl_function_t linux_ioctl_socket;
static linux_ioctl_function_t linux_ioctl_sound;
static linux_ioctl_function_t linux_ioctl_termio;
Add support for the "device private" ioctls soon to be used by the an driver. Also slightly change the name translation policy - only rename interfaces that have the IFF_BROADCAST flag set. This is not perfect, but is closer to how Linux names network interfaces.
2001-10-19 01:38:10 +00:00
static linux_ioctl_function_t linux_ioctl_private;
Add an ioctl handler for the DRM. This removes the need for the DRM_LINUX option, which has been a source of frustration for many users.
2003-04-24 23:36:35 +00:00
static linux_ioctl_function_t linux_ioctl_drm;
Add the CAM 'SG' peripheral device. This device implements a subset of the Linux SCSI SG passthrough device API. The intention is to allow for both running of Linux apps that want to talk to /dev/sg* nodes, and to facilitate porting of apps from Linux to FreeBSD. As such, both native and linuxolator entry points and definitions are provided. Caveats: - This does not support the procfs and sysfs nodes that the Linux SG driver provides. Some Linux apps may rely on these for operation, others may only use them for informational purposes. - More ioctls need to be implemented. - Linux uses a naming scheme of "sg[a-z]" for devices, while FreeBSD uses a scheme of "sg[0-9]". Devfs aliasis (symlinks) are automatically created to link the two together. However, tools like camcontrol only see the native names. - Some operations were originally designed to return byte counts or other data directly as the syscall return value. The linuxolator doesn't appear to support this well, so this driver just punts for these cases. Now that the driver is in place, others are welcome to add missing functionality. Thanks to Roman Divacky for pushing this work along.
2007-04-07 19:40:58 +00:00
static linux_ioctl_function_t linux_ioctl_sg;
This is v4l support for the linuxulator. This allows to access FreeBSD native devices which support the v4l API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd driver. Not tested is firmware upload, framebuffer stuff and video tuner stuff due to lack of hardware. The clipping part (VIDIOCSWIN) needs a little bit of further work (partly in progress, but can not be tested due to lack of a suitable device). The submitter tested this sucessfully with Skype and flash apps on amd64 and i386 with the multimedia/pwcbsd driver. Submitted by: J.R. Oldroyd <fbsd@opal.com>
2009-12-04 21:06:54 +00:00
static linux_ioctl_function_t linux_ioctl_v4l;
This is v4l2 support for the linuxulator. This allows to access FreeBSD native devices which support the v4l2 API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd or multimedia/webcamd supplied drivers. Submitted by: nox MFC after: 1 month
2011-05-04 09:05:39 +00:00
static linux_ioctl_function_t linux_ioctl_v4l2;
Deal with a few issues that cropped up following the recent changes to the code for translating socket and private ioctls: - Only perform socket ioctl translation if the file descriptor is a socket. - Treat socket ioctls on non-sockets specially, and for now assume that these are directed at a tap/vmnet device, so translate the ioctl numbers as appropriate (the way if_tap abuses some socket ioctls to pass non-ifreq data is utterly bogus, but this is how VMware on FreeBSD has always "worked"; I will deal with this later). - Add (untested) support for translating SIOCSIFADDR. - In all cases where we fail to translate an ioctl, return ENOIOCTL so that other handlers have a chance to do the translation. This should fix the "/dev/vmnet1: Invalid argument" errors that users of VMware were experiencing, though I have only verified this on RELENG_4. Submitted by: des (mostly) MFC after: 3 days
2001-11-19 15:43:50 +00:00
static linux_ioctl_function_t linux_ioctl_special;
Let us to use our libusb(3) in Linuxolator. With this change, Linux binaries can work with our libusb(3) when it's compiled against our header files on GNU/Linux system -- this solves the problem with differences between /dev layouts. With ported libusb(3), I am able to use my USB JTAG cable with Linux binaries that support it. Reviewed by: thompsa
2010-01-18 22:46:06 +00:00
static linux_ioctl_function_t linux_ioctl_fbsd_usb;
Add Evdev ioctl handler to the Linuxulator. PR: 218627 Submitted by: Jan Kokemüller Reported by: Jan Kokemüller MFC after: 1 week
2017-04-23 07:43:50 +00:00
static linux_ioctl_function_t linux_ioctl_evdev;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
static struct linux_ioctl_handler cdrom_handler =
{ linux_ioctl_cdrom, LINUX_IOCTL_CDROM_MIN, LINUX_IOCTL_CDROM_MAX };
- Add support for IPC_64 extensions into shmctl(2), semctl(2) and msgctl(2); - add wrappers for mmap2(2) and ftruncate64(2) system calls; - don't spam console with printf's when VFAT_READDIR_BOTH ioctl(2) is invoked; - add support for SOUND_MIXER_READ_STEREODEVS ioctl(2); - make msgctl(IPC_STAT) and IPC_SET actually working by converting from BSD msqid_ds to Linux and vice versa; - properly return EINVAL if semget(2) is called with nsems being negative. Reviewed by: marcel Approved by: marcel Tested with: LSB runtime test
2002-10-11 11:43:09 +00:00
static struct linux_ioctl_handler vfat_handler =
{ linux_ioctl_vfat, LINUX_IOCTL_VFAT_MIN, LINUX_IOCTL_VFAT_MAX };
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
static struct linux_ioctl_handler console_handler =
{ linux_ioctl_console, LINUX_IOCTL_CONSOLE_MIN, LINUX_IOCTL_CONSOLE_MAX };
Add support for more linux ioctls. I've had this sitting in my tree for a long time and I can't seem to find who sent it to me in the first place, apologies to whoever is missing out on a Contributed by: line here. I belive it works as it should.
2004-06-14 07:26:23 +00:00
static struct linux_ioctl_handler hdio_handler =
{ linux_ioctl_hdio, LINUX_IOCTL_HDIO_MIN, LINUX_IOCTL_HDIO_MAX };
Fix some style bugs. The long line is in a chunk of code that's being rewritten, though. Submitted by: bde
2000-03-13 15:27:19 +00:00
static struct linux_ioctl_handler disk_handler =
{ linux_ioctl_disk, LINUX_IOCTL_DISK_MIN, LINUX_IOCTL_DISK_MAX };
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
static struct linux_ioctl_handler socket_handler =
{ linux_ioctl_socket, LINUX_IOCTL_SOCKET_MIN, LINUX_IOCTL_SOCKET_MAX };
static struct linux_ioctl_handler sound_handler =
{ linux_ioctl_sound, LINUX_IOCTL_SOUND_MIN, LINUX_IOCTL_SOUND_MAX };
Add support for the "device private" ioctls soon to be used by the an driver. Also slightly change the name translation policy - only rename interfaces that have the IFF_BROADCAST flag set. This is not perfect, but is closer to how Linux names network interfaces.
2001-10-19 01:38:10 +00:00
static struct linux_ioctl_handler private_handler =
{ linux_ioctl_private, LINUX_IOCTL_PRIVATE_MIN, LINUX_IOCTL_PRIVATE_MAX };
Add an ioctl handler for the DRM. This removes the need for the DRM_LINUX option, which has been a source of frustration for many users.
2003-04-24 23:36:35 +00:00
static struct linux_ioctl_handler drm_handler =
{ linux_ioctl_drm, LINUX_IOCTL_DRM_MIN, LINUX_IOCTL_DRM_MAX };
Add the CAM 'SG' peripheral device. This device implements a subset of the Linux SCSI SG passthrough device API. The intention is to allow for both running of Linux apps that want to talk to /dev/sg* nodes, and to facilitate porting of apps from Linux to FreeBSD. As such, both native and linuxolator entry points and definitions are provided. Caveats: - This does not support the procfs and sysfs nodes that the Linux SG driver provides. Some Linux apps may rely on these for operation, others may only use them for informational purposes. - More ioctls need to be implemented. - Linux uses a naming scheme of "sg[a-z]" for devices, while FreeBSD uses a scheme of "sg[0-9]". Devfs aliasis (symlinks) are automatically created to link the two together. However, tools like camcontrol only see the native names. - Some operations were originally designed to return byte counts or other data directly as the syscall return value. The linuxolator doesn't appear to support this well, so this driver just punts for these cases. Now that the driver is in place, others are welcome to add missing functionality. Thanks to Roman Divacky for pushing this work along.
2007-04-07 19:40:58 +00:00
static struct linux_ioctl_handler sg_handler =
{ linux_ioctl_sg, LINUX_IOCTL_SG_MIN, LINUX_IOCTL_SG_MAX };
This is v4l support for the linuxulator. This allows to access FreeBSD native devices which support the v4l API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd driver. Not tested is firmware upload, framebuffer stuff and video tuner stuff due to lack of hardware. The clipping part (VIDIOCSWIN) needs a little bit of further work (partly in progress, but can not be tested due to lack of a suitable device). The submitter tested this sucessfully with Skype and flash apps on amd64 and i386 with the multimedia/pwcbsd driver. Submitted by: J.R. Oldroyd <fbsd@opal.com>
2009-12-04 21:06:54 +00:00
static struct linux_ioctl_handler video_handler =
{ linux_ioctl_v4l, LINUX_IOCTL_VIDEO_MIN, LINUX_IOCTL_VIDEO_MAX };
This is v4l2 support for the linuxulator. This allows to access FreeBSD native devices which support the v4l2 API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd or multimedia/webcamd supplied drivers. Submitted by: nox MFC after: 1 month
2011-05-04 09:05:39 +00:00
static struct linux_ioctl_handler video2_handler =
{ linux_ioctl_v4l2, LINUX_IOCTL_VIDEO2_MIN, LINUX_IOCTL_VIDEO2_MAX };
Let us to use our libusb(3) in Linuxolator. With this change, Linux binaries can work with our libusb(3) when it's compiled against our header files on GNU/Linux system -- this solves the problem with differences between /dev layouts. With ported libusb(3), I am able to use my USB JTAG cable with Linux binaries that support it. Reviewed by: thompsa
2010-01-18 22:46:06 +00:00
static struct linux_ioctl_handler fbsd_usb =
Bring USB fixes for linux(4). Intention of this commit is to let us take a full advantage of libusb(8) ported to Linux. This decreases a possibility of getting any collisions within ioctl() "command" space, especially with relation to LINUX_SNDCTL_SEQ... stuff. Basically, we provide commands, that will be mapped in the kernel to correct ones and forward those to the USB layer. Port enabling functionality brought with this patch is here: http://www.freebsd.org/cgi/query-pr.cgi?pr=146895 Bump __FreeBSD_version to catch, since which version installing a port makes sense. This patch should bring no regressions. So far, only i386 is tested. Tested by: thompsa@ Reviewed by: thompsa@ OKed by: netchild@
2010-05-24 07:04:00 +00:00
{ linux_ioctl_fbsd_usb, FBSD_LUSB_MIN, FBSD_LUSB_MAX };
Add Evdev ioctl handler to the Linuxulator. PR: 218627 Submitted by: Jan Kokemüller Reported by: Jan Kokemüller MFC after: 1 week
2017-04-23 07:43:50 +00:00
static struct linux_ioctl_handler evdev_handler =
{ linux_ioctl_evdev, LINUX_IOCTL_EVDEV_MIN, LINUX_IOCTL_EVDEV_MAX };
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
DATA_SET(linux_ioctl_handler_set, cdrom_handler);
- Add support for IPC_64 extensions into shmctl(2), semctl(2) and msgctl(2); - add wrappers for mmap2(2) and ftruncate64(2) system calls; - don't spam console with printf's when VFAT_READDIR_BOTH ioctl(2) is invoked; - add support for SOUND_MIXER_READ_STEREODEVS ioctl(2); - make msgctl(IPC_STAT) and IPC_SET actually working by converting from BSD msqid_ds to Linux and vice versa; - properly return EINVAL if semget(2) is called with nsems being negative. Reviewed by: marcel Approved by: marcel Tested with: LSB runtime test
2002-10-11 11:43:09 +00:00
DATA_SET(linux_ioctl_handler_set, vfat_handler);
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
DATA_SET(linux_ioctl_handler_set, console_handler);
Add support for more linux ioctls. I've had this sitting in my tree for a long time and I can't seem to find who sent it to me in the first place, apologies to whoever is missing out on a Contributed by: line here. I belive it works as it should.
2004-06-14 07:26:23 +00:00
DATA_SET(linux_ioctl_handler_set, hdio_handler);
Fix some style bugs. The long line is in a chunk of code that's being rewritten, though. Submitted by: bde
2000-03-13 15:27:19 +00:00
DATA_SET(linux_ioctl_handler_set, disk_handler);
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
DATA_SET(linux_ioctl_handler_set, socket_handler);
DATA_SET(linux_ioctl_handler_set, sound_handler);
Add support for the "device private" ioctls soon to be used by the an driver. Also slightly change the name translation policy - only rename interfaces that have the IFF_BROADCAST flag set. This is not perfect, but is closer to how Linux names network interfaces.
2001-10-19 01:38:10 +00:00
DATA_SET(linux_ioctl_handler_set, private_handler);
Add an ioctl handler for the DRM. This removes the need for the DRM_LINUX option, which has been a source of frustration for many users.
2003-04-24 23:36:35 +00:00
DATA_SET(linux_ioctl_handler_set, drm_handler);
Add the CAM 'SG' peripheral device. This device implements a subset of the Linux SCSI SG passthrough device API. The intention is to allow for both running of Linux apps that want to talk to /dev/sg* nodes, and to facilitate porting of apps from Linux to FreeBSD. As such, both native and linuxolator entry points and definitions are provided. Caveats: - This does not support the procfs and sysfs nodes that the Linux SG driver provides. Some Linux apps may rely on these for operation, others may only use them for informational purposes. - More ioctls need to be implemented. - Linux uses a naming scheme of "sg[a-z]" for devices, while FreeBSD uses a scheme of "sg[0-9]". Devfs aliasis (symlinks) are automatically created to link the two together. However, tools like camcontrol only see the native names. - Some operations were originally designed to return byte counts or other data directly as the syscall return value. The linuxolator doesn't appear to support this well, so this driver just punts for these cases. Now that the driver is in place, others are welcome to add missing functionality. Thanks to Roman Divacky for pushing this work along.
2007-04-07 19:40:58 +00:00
DATA_SET(linux_ioctl_handler_set, sg_handler);
This is v4l support for the linuxulator. This allows to access FreeBSD native devices which support the v4l API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd driver. Not tested is firmware upload, framebuffer stuff and video tuner stuff due to lack of hardware. The clipping part (VIDIOCSWIN) needs a little bit of further work (partly in progress, but can not be tested due to lack of a suitable device). The submitter tested this sucessfully with Skype and flash apps on amd64 and i386 with the multimedia/pwcbsd driver. Submitted by: J.R. Oldroyd <fbsd@opal.com>
2009-12-04 21:06:54 +00:00
DATA_SET(linux_ioctl_handler_set, video_handler);
This is v4l2 support for the linuxulator. This allows to access FreeBSD native devices which support the v4l2 API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd or multimedia/webcamd supplied drivers. Submitted by: nox MFC after: 1 month
2011-05-04 09:05:39 +00:00
DATA_SET(linux_ioctl_handler_set, video2_handler);
Let us to use our libusb(3) in Linuxolator. With this change, Linux binaries can work with our libusb(3) when it's compiled against our header files on GNU/Linux system -- this solves the problem with differences between /dev layouts. With ported libusb(3), I am able to use my USB JTAG cable with Linux binaries that support it. Reviewed by: thompsa
2010-01-18 22:46:06 +00:00
DATA_SET(linux_ioctl_handler_set, fbsd_usb);
Add Evdev ioctl handler to the Linuxulator. PR: 218627 Submitted by: Jan Kokemüller Reported by: Jan Kokemüller MFC after: 1 week
2017-04-23 07:43:50 +00:00
DATA_SET(linux_ioctl_handler_set, evdev_handler);
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
linux: fix ioctl performance for termios TCGETS et al are frequently issued by Linux binaries while the previous code avoidably ping-pongs a global sx lock and serializes on Giant. Note that even with the fix the common case will serialize on a per-tty lock.
2020-07-04 06:25:41 +00:00
/*
* Keep sorted by low.
*/
static struct linux_ioctl_handler linux_ioctls[] = {
{ .func = linux_ioctl_termio, .low = LINUX_IOCTL_TERMIO_MIN,
.high = LINUX_IOCTL_TERMIO_MAX },
};
On amd64 both Linux compat modules, linux.ko and linux64.ko, provide linux_ioctl_(un)register_handler that allows other driver modules to register ioctl handlers. The ioctl syscall implementation in each Linux compat module iterates over the list of handlers and forwards the call to the appropriate driver. Because the registration functions have the same name in each module it is not possible for a driver to support both 32 and 64 bit linux compatibility. Move the list of ioctl handlers to linux_common.ko so it is shared by both Linux modules and all drivers receive both 32 and 64 bit ioctl calls with one registration. These ioctl handlers normally forward the call to the FreeBSD ioctl handler which can handle both 32 and 64 bit. Keep the special COMPAT_LINUX32 ioctl handlers in linux.ko in a separate list for now and let the ioctl syscall iterate over that list first. Later, COMPAT_LINUX32 support can be added to the 64 bit ioctl handlers via a runtime check for ILP32 like is done for COMPAT_FREEBSD32 and then this separate list would disappear again. That is a much bigger effort however and this commit is meant to be MFCable. This enables linux64 support in x11/nvidia-driver*. PR: 206711 Reviewed by: kib MFC after: 3 days
2018-11-06 13:51:08 +00:00
#ifdef __i386__
static TAILQ_HEAD(, linux_ioctl_handler_element) linux_ioctl_handlers =
TAILQ_HEAD_INITIALIZER(linux_ioctl_handlers);
- Protect the list of linux ioctl handlers with an sx lock. - Hold Giant while calling linux ioctl handlers for now as they aren't all known to be MPSAFE yet. - Mark linux_ioctl() MPSAFE.
2006-07-06 21:42:36 +00:00
static struct sx linux_ioctl_sx;
Correct proper nouns in the Linuxulator - Capitalize Linux - Spell FreeBSD out in full - Address some style(9) on changed lines Sponsored by: Turing Robotic Industries Inc.
2018-02-22 02:24:17 +00:00
SX_SYSINIT(linux_ioctl, &linux_ioctl_sx, "Linux ioctl handlers");
On amd64 both Linux compat modules, linux.ko and linux64.ko, provide linux_ioctl_(un)register_handler that allows other driver modules to register ioctl handlers. The ioctl syscall implementation in each Linux compat module iterates over the list of handlers and forwards the call to the appropriate driver. Because the registration functions have the same name in each module it is not possible for a driver to support both 32 and 64 bit linux compatibility. Move the list of ioctl handlers to linux_common.ko so it is shared by both Linux modules and all drivers receive both 32 and 64 bit ioctl calls with one registration. These ioctl handlers normally forward the call to the FreeBSD ioctl handler which can handle both 32 and 64 bit. Keep the special COMPAT_LINUX32 ioctl handlers in linux.ko in a separate list for now and let the ioctl syscall iterate over that list first. Later, COMPAT_LINUX32 support can be added to the 64 bit ioctl handlers via a runtime check for ILP32 like is done for COMPAT_FREEBSD32 and then this separate list would disappear again. That is a much bigger effort however and this commit is meant to be MFCable. This enables linux64 support in x11/nvidia-driver*. PR: 206711 Reviewed by: kib MFC after: 3 days
2018-11-06 13:51:08 +00:00
#else
extern TAILQ_HEAD(, linux_ioctl_handler_element) linux_ioctl_handlers;
extern struct sx linux_ioctl_sx;
#endif
#ifdef COMPAT_LINUX32
static TAILQ_HEAD(, linux_ioctl_handler_element) linux32_ioctl_handlers =
TAILQ_HEAD_INITIALIZER(linux32_ioctl_handlers);
#endif
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
Add support for more linux ioctls. I've had this sitting in my tree for a long time and I can't seem to find who sent it to me in the first place, apologies to whoever is missing out on a Contributed by: line here. I belive it works as it should.
2004-06-14 07:26:23 +00:00
/*
* hdio related ioctls for VMWare support
*/
struct linux_hd_geometry {
u_int8_t heads;
u_int8_t sectors;
u_int16_t cylinders;
u_int32_t start;
};
struct linux_hd_big_geometry {
u_int8_t heads;
u_int8_t sectors;
u_int32_t cylinders;
u_int32_t start;
};
static int
linux_ioctl_hdio(struct thread *td, struct linux_ioctl_args *args)
{
struct file *fp;
int error;
u_int sectorsize, fwcylinders, fwheads, fwsectors;
off_t mediasize, bytespercyl;
Eliminate the overhead of gratuitous repeated reinitialization of cap_rights - Add macros to allow preinitialization of cap_rights_t. - Convert most commonly used code paths to use preinitialized cap_rights_t. A 3.6% speedup in fstat was measured with this change. Reported by: mjg Reviewed by: oshogbo Approved by: sbruno MFC after: 1 month
2018-05-09 18:47:24 +00:00
error = fget(td, args->fd, &cap_ioctl_rights, &fp);
Change the cap_rights_t type from uint64_t to a structure that we can extend in the future in a backward compatible (API and ABI) way. The cap_rights_t represents capability rights. We used to use one bit to represent one right, but we are running out of spare bits. Currently the new structure provides place for 114 rights (so 50 more than the previous cap_rights_t), but it is possible to grow the structure to hold at least 285 rights, although we can make it even larger if 285 rights won't be enough. The structure definition looks like this: struct cap_rights { uint64_t cr_rights[CAP_RIGHTS_VERSION + 2]; }; The initial CAP_RIGHTS_VERSION is 0. The top two bits in the first element of the cr_rights[] array contain total number of elements in the array - 2. This means if those two bits are equal to 0, we have 2 array elements. The top two bits in all remaining array elements should be 0. The next five bits in all array elements contain array index. Only one bit is used and bit position in this five-bits range defines array index. This means there can be at most five array elements in the future. To define new right the CAPRIGHT() macro must be used. The macro takes two arguments - an array index and a bit to set, eg. #define CAP_PDKILL CAPRIGHT(1, 0x0000000000000800ULL) We still support aliases that combine few rights, but the rights have to belong to the same array element, eg: #define CAP_LOOKUP CAPRIGHT(0, 0x0000000000000400ULL) #define CAP_FCHMOD CAPRIGHT(0, 0x0000000000002000ULL) #define CAP_FCHMODAT (CAP_FCHMOD | CAP_LOOKUP) There is new API to manage the new cap_rights_t structure: cap_rights_t *cap_rights_init(cap_rights_t *rights, ...); void cap_rights_set(cap_rights_t *rights, ...); void cap_rights_clear(cap_rights_t *rights, ...); bool cap_rights_is_set(const cap_rights_t *rights, ...); bool cap_rights_is_valid(const cap_rights_t *rights); void cap_rights_merge(cap_rights_t *dst, const cap_rights_t *src); void cap_rights_remove(cap_rights_t *dst, const cap_rights_t *src); bool cap_rights_contains(const cap_rights_t *big, const cap_rights_t *little); Capability rights to the cap_rights_init(), cap_rights_set(), cap_rights_clear() and cap_rights_is_set() functions are provided by separating them with commas, eg: cap_rights_t rights; cap_rights_init(&rights, CAP_READ, CAP_WRITE, CAP_FSTAT); There is no need to terminate the list of rights, as those functions are actually macros that take care of the termination, eg: #define cap_rights_set(rights, ...) \ __cap_rights_set((rights), __VA_ARGS__, 0ULL) void __cap_rights_set(cap_rights_t *rights, ...); Thanks to using one bit as an array index we can assert in those functions that there are no two rights belonging to different array elements provided together. For example this is illegal and will be detected, because CAP_LOOKUP belongs to element 0 and CAP_PDKILL to element 1: cap_rights_init(&rights, CAP_LOOKUP | CAP_PDKILL); Providing several rights that belongs to the same array's element this way is correct, but is not advised. It should only be used for aliases definition. This commit also breaks compatibility with some existing Capsicum system calls, but I see no other way to do that. This should be fine as Capsicum is still experimental and this change is not going to 9.x. Sponsored by: The FreeBSD Foundation
2013-09-05 00:09:56 +00:00
if (error != 0)
Add support for more linux ioctls. I've had this sitting in my tree for a long time and I can't seem to find who sent it to me in the first place, apologies to whoever is missing out on a Contributed by: line here. I belive it works as it should.
2004-06-14 07:26:23 +00:00
return (error);
switch (args->cmd & 0xffff) {
case LINUX_HDIO_GET_GEO:
case LINUX_HDIO_GET_GEO_BIG:
error = fo_ioctl(fp, DIOCGMEDIASIZE,
(caddr_t)&mediasize, td->td_ucred, td);
if (!error)
error = fo_ioctl(fp, DIOCGSECTORSIZE,
(caddr_t)&sectorsize, td->td_ucred, td);
if (!error)
error = fo_ioctl(fp, DIOCGFWHEADS,
(caddr_t)&fwheads, td->td_ucred, td);
if (!error)
error = fo_ioctl(fp, DIOCGFWSECTORS,
(caddr_t)&fwsectors, td->td_ucred, td);
/*
* XXX: DIOCGFIRSTOFFSET is not yet implemented, so
* so pretend that GEOM always says 0. This is NOT VALID
* for slices or partitions, only the per-disk raw devices.
*/
fdrop(fp, td);
if (error)
return (error);
/*
* 1. Calculate the number of bytes in a cylinder,
* given the firmware's notion of heads and sectors
* per cylinder.
* 2. Calculate the number of cylinders, given the total
* size of the media.
* All internal calculations should have 64-bit precision.
*/
bytespercyl = (off_t) sectorsize * fwheads * fwsectors;
fwcylinders = mediasize / bytespercyl;
Linuxulator depends on a fundamental kernel settings such as SMP. Many of them listed in opt_global.h which is not generated while building modules outside of a kernel and such modules never match real cofigured kernel. So, we should prevent our users from building obviously defective modules. Therefore, remove the root cause of the building of modules outside of a kernel - the possibility of building modules with DEBUG or KTR flags. And remove all of DEBUG printfs as it is incomplete and in threaded programms not informative, also a half of system call does not have DEBUG printf. For debuging Linux programms we have dtrace, ktr and ktrace ability. PR: 222861 Reviewed by: trasz MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D20178
2019-05-13 18:24:29 +00:00
Add support for more linux ioctls. I've had this sitting in my tree for a long time and I can't seem to find who sent it to me in the first place, apologies to whoever is missing out on a Contributed by: line here. I belive it works as it should.
2004-06-14 07:26:23 +00:00
if ((args->cmd & 0xffff) == LINUX_HDIO_GET_GEO) {
struct linux_hd_geometry hdg;
hdg.cylinders = fwcylinders;
hdg.heads = fwheads;
hdg.sectors = fwsectors;
hdg.start = 0;
error = copyout(&hdg, (void *)args->arg, sizeof(hdg));
} else if ((args->cmd & 0xffff) == LINUX_HDIO_GET_GEO_BIG) {
struct linux_hd_big_geometry hdbg;
linux_ioctl_hdio: fix kernel memory disclosure Stack-allocated struct linux_hd_big_geometry has undeclared padding copied to userland. admbugs: 765 Reported by: Vlad Tsyrklevich MFC after: 1 day Security: Kernel memory disclosure Sponsored by: The FreeBSD Foundation
2018-04-04 14:41:48 +00:00
memset(&hdbg, 0, sizeof(hdbg));
Add support for more linux ioctls. I've had this sitting in my tree for a long time and I can't seem to find who sent it to me in the first place, apologies to whoever is missing out on a Contributed by: line here. I belive it works as it should.
2004-06-14 07:26:23 +00:00
hdbg.cylinders = fwcylinders;
hdbg.heads = fwheads;
hdbg.sectors = fwsectors;
hdbg.start = 0;
error = copyout(&hdbg, (void *)args->arg, sizeof(hdbg));
}
return (error);
break;
default:
/* XXX */
linux_msg(td,
"ioctl fd=%d, cmd=0x%x ('%c',%d) is not implemented",
args->fd, (int)(args->cmd & 0xffff),
(int)(args->cmd & 0xff00) >> 8,
(int)(args->cmd & 0xff));
break;
}
fdrop(fp, td);
return (ENOIOCTL);
}
Implement Linux BLKGETSIZE ioctl, and open the door to implementing other BLK.* ioctls should the desire arize. Approved by: jkh (via dufault)
2000-03-09 15:14:14 +00:00
static int
KSE Milestone 2 Note ALL MODULES MUST BE RECOMPILED make the kernel aware that there are smaller units of scheduling than the process. (but only allow one thread per process at this time). This is functionally equivalent to teh previousl -current except that there is a thread associated with each process. Sorry john! (your next MFC will be a doosie!) Reviewed by: peter@freebsd.org, dillon@freebsd.org X-MFC after: ha ha ha ha
2001-09-12 08:38:13 +00:00
linux_ioctl_disk(struct thread *td, struct linux_ioctl_args *args)
Implement Linux BLKGETSIZE ioctl, and open the door to implementing other BLK.* ioctls should the desire arize. Approved by: jkh (via dufault)
2000-03-09 15:14:14 +00:00
{
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
struct file *fp;
Implement Linux BLKGETSIZE ioctl, and open the door to implementing other BLK.* ioctls should the desire arize. Approved by: jkh (via dufault)
2000-03-09 15:14:14 +00:00
int error;
Make linux(4) support the BLKPBSZGET ioctl. Oracle uses it. MFC after: 2 weeks Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D25694
2020-07-19 12:25:03 +00:00
u_int sectorsize, psectorsize;
Implement Linux BLKGETSIZE64 ioctl. MFC after: 2 weeks Sponsored by: The FreeBSD Foundation
2019-12-29 12:22:11 +00:00
uint64_t blksize64;
Make linux(4) support the BLKPBSZGET ioctl. Oracle uses it. MFC after: 2 weeks Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D25694
2020-07-19 12:25:03 +00:00
off_t mediasize, stripesize;
Implement Linux BLKGETSIZE ioctl, and open the door to implementing other BLK.* ioctls should the desire arize. Approved by: jkh (via dufault)
2000-03-09 15:14:14 +00:00
Eliminate the overhead of gratuitous repeated reinitialization of cap_rights - Add macros to allow preinitialization of cap_rights_t. - Convert most commonly used code paths to use preinitialized cap_rights_t. A 3.6% speedup in fstat was measured with this change. Reported by: mjg Reviewed by: oshogbo Approved by: sbruno MFC after: 1 month
2018-05-09 18:47:24 +00:00
error = fget(td, args->fd, &cap_ioctl_rights, &fp);
Change the cap_rights_t type from uint64_t to a structure that we can extend in the future in a backward compatible (API and ABI) way. The cap_rights_t represents capability rights. We used to use one bit to represent one right, but we are running out of spare bits. Currently the new structure provides place for 114 rights (so 50 more than the previous cap_rights_t), but it is possible to grow the structure to hold at least 285 rights, although we can make it even larger if 285 rights won't be enough. The structure definition looks like this: struct cap_rights { uint64_t cr_rights[CAP_RIGHTS_VERSION + 2]; }; The initial CAP_RIGHTS_VERSION is 0. The top two bits in the first element of the cr_rights[] array contain total number of elements in the array - 2. This means if those two bits are equal to 0, we have 2 array elements. The top two bits in all remaining array elements should be 0. The next five bits in all array elements contain array index. Only one bit is used and bit position in this five-bits range defines array index. This means there can be at most five array elements in the future. To define new right the CAPRIGHT() macro must be used. The macro takes two arguments - an array index and a bit to set, eg. #define CAP_PDKILL CAPRIGHT(1, 0x0000000000000800ULL) We still support aliases that combine few rights, but the rights have to belong to the same array element, eg: #define CAP_LOOKUP CAPRIGHT(0, 0x0000000000000400ULL) #define CAP_FCHMOD CAPRIGHT(0, 0x0000000000002000ULL) #define CAP_FCHMODAT (CAP_FCHMOD | CAP_LOOKUP) There is new API to manage the new cap_rights_t structure: cap_rights_t *cap_rights_init(cap_rights_t *rights, ...); void cap_rights_set(cap_rights_t *rights, ...); void cap_rights_clear(cap_rights_t *rights, ...); bool cap_rights_is_set(const cap_rights_t *rights, ...); bool cap_rights_is_valid(const cap_rights_t *rights); void cap_rights_merge(cap_rights_t *dst, const cap_rights_t *src); void cap_rights_remove(cap_rights_t *dst, const cap_rights_t *src); bool cap_rights_contains(const cap_rights_t *big, const cap_rights_t *little); Capability rights to the cap_rights_init(), cap_rights_set(), cap_rights_clear() and cap_rights_is_set() functions are provided by separating them with commas, eg: cap_rights_t rights; cap_rights_init(&rights, CAP_READ, CAP_WRITE, CAP_FSTAT); There is no need to terminate the list of rights, as those functions are actually macros that take care of the termination, eg: #define cap_rights_set(rights, ...) \ __cap_rights_set((rights), __VA_ARGS__, 0ULL) void __cap_rights_set(cap_rights_t *rights, ...); Thanks to using one bit as an array index we can assert in those functions that there are no two rights belonging to different array elements provided together. For example this is illegal and will be detected, because CAP_LOOKUP belongs to element 0 and CAP_PDKILL to element 1: cap_rights_init(&rights, CAP_LOOKUP | CAP_PDKILL); Providing several rights that belongs to the same array's element this way is correct, but is not advised. It should only be used for aliases definition. This commit also breaks compatibility with some existing Capsicum system calls, but I see no other way to do that. This should be fine as Capsicum is still experimental and this change is not going to 9.x. Sponsored by: The FreeBSD Foundation
2013-09-05 00:09:56 +00:00
if (error != 0)
Replace ffind_* with fget calls. Make fget MPsafe. Make fgetvp and fgetsock use the fget subsystem to reduce code bloat. Push giant down in fpathconf().
2002-01-14 00:13:45 +00:00
return (error);
Implement Linux BLKGETSIZE ioctl, and open the door to implementing other BLK.* ioctls should the desire arize. Approved by: jkh (via dufault)
2000-03-09 15:14:14 +00:00
switch (args->cmd & 0xffff) {
case LINUX_BLKGETSIZE:
Fix an XXX: and implement LINUX_BLKGETSIZE correctly.
2003-03-28 08:58:11 +00:00
error = fo_ioctl(fp, DIOCGSECTORSIZE,
(caddr_t)&sectorsize, td->td_ucred, td);
if (!error)
error = fo_ioctl(fp, DIOCGMEDIASIZE,
(caddr_t)&mediasize, td->td_ucred, td);
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
fdrop(fp, td);
Implement Linux BLKGETSIZE ioctl, and open the door to implementing other BLK.* ioctls should the desire arize. Approved by: jkh (via dufault)
2000-03-09 15:14:14 +00:00
if (error)
return (error);
Fix an XXX: and implement LINUX_BLKGETSIZE correctly.
2003-03-28 08:58:11 +00:00
sectorsize = mediasize / sectorsize;
/*
* XXX: How do we know we return the right size of integer ?
*/
return (copyout(&sectorsize, (void *)args->arg,
sizeof(sectorsize)));
Add support for more linux ioctls. I've had this sitting in my tree for a long time and I can't seem to find who sent it to me in the first place, apologies to whoever is missing out on a Contributed by: line here. I belive it works as it should.
2004-06-14 07:26:23 +00:00
break;
Implement Linux BLKGETSIZE64 ioctl. MFC after: 2 weeks Sponsored by: The FreeBSD Foundation
2019-12-29 12:22:11 +00:00
case LINUX_BLKGETSIZE64:
error = fo_ioctl(fp, DIOCGMEDIASIZE,
(caddr_t)&mediasize, td->td_ucred, td);
fdrop(fp, td);
if (error)
return (error);
blksize64 = mediasize;;
return (copyout(&blksize64, (void *)args->arg,
sizeof(blksize64)));
Implement BLKSSZGET ioctl for the Linuxulator. PR: 212700 Submitted by: Erik Cederstrand Reported by: Erik Cederstrand MFC after: 1 week
2016-09-17 08:10:01 +00:00
case LINUX_BLKSSZGET:
error = fo_ioctl(fp, DIOCGSECTORSIZE,
(caddr_t)&sectorsize, td->td_ucred, td);
fdrop(fp, td);
if (error)
return (error);
return (copyout(&sectorsize, (void *)args->arg,
sizeof(sectorsize)));
break;
Make linux(4) support the BLKPBSZGET ioctl. Oracle uses it. MFC after: 2 weeks Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D25694
2020-07-19 12:25:03 +00:00
case LINUX_BLKPBSZGET:
error = fo_ioctl(fp, DIOCGSTRIPESIZE,
(caddr_t)&stripesize, td->td_ucred, td);
if (error != 0) {
fdrop(fp, td);
return (error);
}
if (stripesize > 0 && stripesize <= 4096) {
psectorsize = stripesize;
} else {
error = fo_ioctl(fp, DIOCGSECTORSIZE,
(caddr_t)&sectorsize, td->td_ucred, td);
if (error != 0) {
fdrop(fp, td);
return (error);
}
psectorsize = sectorsize;
}
fdrop(fp, td);
return (copyout(&psectorsize, (void *)args->arg,
sizeof(psectorsize)));
Implement Linux BLKGETSIZE ioctl, and open the door to implementing other BLK.* ioctls should the desire arize. Approved by: jkh (via dufault)
2000-03-09 15:14:14 +00:00
}
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
fdrop(fp, td);
Implement Linux BLKGETSIZE ioctl, and open the door to implementing other BLK.* ioctls should the desire arize. Approved by: jkh (via dufault)
2000-03-09 15:14:14 +00:00
return (ENOIOCTL);
}
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
/*
* termio related ioctls
*/
Initial support for using linux X servers under emulation - to use an XFree86 server, users need to create the following links in their /compat/linux/dev directory (assuming kernel configured with 4 VTs). lrwxrwxrwx 1 root wheel 7 Aug 30 22:59 tty0 -> console lrwxrwxrwx 1 root wheel 5 Aug 30 22:45 tty1 -> ttyv0 lrwxrwxrwx 1 root wheel 5 Aug 30 22:45 tty2 -> ttyv1 lrwxrwxrwx 1 root wheel 5 Aug 30 22:45 tty3 -> ttyv2 lrwxrwxrwx 1 root wheel 5 Aug 30 22:45 tty4 -> ttyv3 VT switching is still not yet supported. Attempting to switch VT currently will cause Xserver bus error. Submitted by: Chain Lee <chain@110.net>
1998-08-31 06:55:02 +00:00
Add support for the old-style Linux termio (not termios) TCGETA etc. Also, LINUX_POSIX_VDISABLE is \0, FreeBSD's is 0xff. Convert between them. This enables some more programs to run, including the Livingston Portmaster utilities (PMtools). Submitted by: Robert Sanders <rsanders@mindspring.com>
1996-03-03 19:07:50 +00:00
struct linux_termio {
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
unsigned short c_iflag;
unsigned short c_oflag;
unsigned short c_cflag;
unsigned short c_lflag;
unsigned char c_line;
unsigned char c_cc[LINUX_NCC];
Add support for the old-style Linux termio (not termios) TCGETA etc. Also, LINUX_POSIX_VDISABLE is \0, FreeBSD's is 0xff. Convert between them. This enables some more programs to run, including the Livingston Portmaster utilities (PMtools). Submitted by: Robert Sanders <rsanders@mindspring.com>
1996-03-03 19:07:50 +00:00
};
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
struct linux_termios {
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
unsigned int c_iflag;
unsigned int c_oflag;
unsigned int c_cflag;
unsigned int c_lflag;
unsigned char c_line;
unsigned char c_cc[LINUX_NCCS];
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
};
struct linux_winsize {
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
unsigned short ws_row, ws_col;
unsigned short ws_xpixel, ws_ypixel;
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
};
Integrate the new MPSAFE TTY layer to the FreeBSD operating system. The last half year I've been working on a replacement TTY layer for the FreeBSD kernel. The new TTY layer was designed to improve the following: - Improved driver model: The old TTY layer has a driver model that is not abstract enough to make it friendly to use. A good example is the output path, where the device drivers directly access the output buffers. This means that an in-kernel PPP implementation must always convert network buffers into TTY buffers. If a PPP implementation would be built on top of the new TTY layer (still needs a hooks layer, though), it would allow the PPP implementation to directly hand the data to the TTY driver. - Improved hotplugging: With the old TTY layer, it isn't entirely safe to destroy TTY's from the system. This implementation has a two-step destructing design, where the driver first abandons the TTY. After all threads have left the TTY, the TTY layer calls a routine in the driver, which can be used to free resources (unit numbers, etc). The pts(4) driver also implements this feature, which means posix_openpt() will now return PTY's that are created on the fly. - Improved performance: One of the major improvements is the per-TTY mutex, which is expected to improve scalability when compared to the old Giant locking. Another change is the unbuffered copying to userspace, which is both used on TTY device nodes and PTY masters. Upgrading should be quite straightforward. Unlike previous versions, existing kernel configuration files do not need to be changed, except when they reference device drivers that are listed in UPDATING. Obtained from: //depot/projects/mpsafetty/... Approved by: philip (ex-mentor) Discussed: on the lists, at BSDCan, at the DevSummit Sponsored by: Snow B.V., the Netherlands dcons(4) fixed by: kan
2008-08-20 08:31:58 +00:00
struct speedtab {
int sp_speed; /* Speed. */
int sp_code; /* Code. */
};
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
static struct speedtab sptab[] = {
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
{ B0, LINUX_B0 }, { B50, LINUX_B50 },
{ B75, LINUX_B75 }, { B110, LINUX_B110 },
{ B134, LINUX_B134 }, { B150, LINUX_B150 },
{ B200, LINUX_B200 }, { B300, LINUX_B300 },
{ B600, LINUX_B600 }, { B1200, LINUX_B1200 },
{ B1800, LINUX_B1800 }, { B2400, LINUX_B2400 },
{ B4800, LINUX_B4800 }, { B9600, LINUX_B9600 },
{ B19200, LINUX_B19200 }, { B38400, LINUX_B38400 },
{ B57600, LINUX_B57600 }, { B115200, LINUX_B115200 },
{-1, -1 }
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
};
Implement rudumentry support for the linux TIOC[SG]ETSERIAL ioctl's. To complete this, some extra state has to be kept somewhere so that the B38400 flag in Linux can be correctly translated to/from either 38400, 57600 or 115200. Submitted by: Robert Sanders <rsanders@mindspring.com>
1996-03-10 22:30:53 +00:00
struct linux_serial_struct {
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
int type;
int line;
int port;
int irq;
int flags;
int xmit_fifo_size;
int custom_divisor;
int baud_base;
unsigned short close_delay;
char reserved_char[2];
int hub6;
unsigned short closing_wait;
unsigned short closing_wait2;
int reserved[4];
Implement rudumentry support for the linux TIOC[SG]ETSERIAL ioctl's. To complete this, some extra state has to be kept somewhere so that the B38400 flag in Linux can be correctly translated to/from either 38400, 57600 or 115200. Submitted by: Robert Sanders <rsanders@mindspring.com>
1996-03-10 22:30:53 +00:00
};
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
static int
linux_to_bsd_speed(int code, struct speedtab *table)
{
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
for ( ; table->sp_code != -1; table++)
if (table->sp_code == code)
return (table->sp_speed);
Linuxulator: apply style(9) to return Sponsored by: Turing Robotic Industries Inc.
2018-03-12 15:35:24 +00:00
return (-1);
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
}
static int
bsd_to_linux_speed(int speed, struct speedtab *table)
{
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
for ( ; table->sp_speed != -1; table++)
if (table->sp_speed == speed)
return (table->sp_code);
Linuxulator: apply style(9) to return Sponsored by: Turing Robotic Industries Inc.
2018-03-12 15:35:24 +00:00
return (-1);
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
}
static void
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
bsd_to_linux_termios(struct termios *bios, struct linux_termios *lios)
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
{
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
int i;
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
lios->c_iflag = 0;
if (bios->c_iflag & IGNBRK)
lios->c_iflag |= LINUX_IGNBRK;
if (bios->c_iflag & BRKINT)
lios->c_iflag |= LINUX_BRKINT;
if (bios->c_iflag & IGNPAR)
lios->c_iflag |= LINUX_IGNPAR;
if (bios->c_iflag & PARMRK)
lios->c_iflag |= LINUX_PARMRK;
if (bios->c_iflag & INPCK)
lios->c_iflag |= LINUX_INPCK;
if (bios->c_iflag & ISTRIP)
lios->c_iflag |= LINUX_ISTRIP;
if (bios->c_iflag & INLCR)
lios->c_iflag |= LINUX_INLCR;
if (bios->c_iflag & IGNCR)
lios->c_iflag |= LINUX_IGNCR;
if (bios->c_iflag & ICRNL)
lios->c_iflag |= LINUX_ICRNL;
if (bios->c_iflag & IXON)
lios->c_iflag |= LINUX_IXON;
if (bios->c_iflag & IXANY)
lios->c_iflag |= LINUX_IXANY;
if (bios->c_iflag & IXOFF)
lios->c_iflag |= LINUX_IXOFF;
if (bios->c_iflag & IMAXBEL)
lios->c_iflag |= LINUX_IMAXBEL;
lios->c_oflag = 0;
if (bios->c_oflag & OPOST)
lios->c_oflag |= LINUX_OPOST;
if (bios->c_oflag & ONLCR)
lios->c_oflag |= LINUX_ONLCR;
Integrate the new MPSAFE TTY layer to the FreeBSD operating system. The last half year I've been working on a replacement TTY layer for the FreeBSD kernel. The new TTY layer was designed to improve the following: - Improved driver model: The old TTY layer has a driver model that is not abstract enough to make it friendly to use. A good example is the output path, where the device drivers directly access the output buffers. This means that an in-kernel PPP implementation must always convert network buffers into TTY buffers. If a PPP implementation would be built on top of the new TTY layer (still needs a hooks layer, though), it would allow the PPP implementation to directly hand the data to the TTY driver. - Improved hotplugging: With the old TTY layer, it isn't entirely safe to destroy TTY's from the system. This implementation has a two-step destructing design, where the driver first abandons the TTY. After all threads have left the TTY, the TTY layer calls a routine in the driver, which can be used to free resources (unit numbers, etc). The pts(4) driver also implements this feature, which means posix_openpt() will now return PTY's that are created on the fly. - Improved performance: One of the major improvements is the per-TTY mutex, which is expected to improve scalability when compared to the old Giant locking. Another change is the unbuffered copying to userspace, which is both used on TTY device nodes and PTY masters. Upgrading should be quite straightforward. Unlike previous versions, existing kernel configuration files do not need to be changed, except when they reference device drivers that are listed in UPDATING. Obtained from: //depot/projects/mpsafetty/... Approved by: philip (ex-mentor) Discussed: on the lists, at BSDCan, at the DevSummit Sponsored by: Snow B.V., the Netherlands dcons(4) fixed by: kan
2008-08-20 08:31:58 +00:00
if (bios->c_oflag & TAB3)
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
lios->c_oflag |= LINUX_XTABS;
lios->c_cflag = bsd_to_linux_speed(bios->c_ispeed, sptab);
lios->c_cflag |= (bios->c_cflag & CSIZE) >> 4;
if (bios->c_cflag & CSTOPB)
lios->c_cflag |= LINUX_CSTOPB;
if (bios->c_cflag & CREAD)
lios->c_cflag |= LINUX_CREAD;
if (bios->c_cflag & PARENB)
lios->c_cflag |= LINUX_PARENB;
if (bios->c_cflag & PARODD)
lios->c_cflag |= LINUX_PARODD;
if (bios->c_cflag & HUPCL)
lios->c_cflag |= LINUX_HUPCL;
if (bios->c_cflag & CLOCAL)
lios->c_cflag |= LINUX_CLOCAL;
if (bios->c_cflag & CRTSCTS)
lios->c_cflag |= LINUX_CRTSCTS;
lios->c_lflag = 0;
if (bios->c_lflag & ISIG)
lios->c_lflag |= LINUX_ISIG;
if (bios->c_lflag & ICANON)
lios->c_lflag |= LINUX_ICANON;
if (bios->c_lflag & ECHO)
lios->c_lflag |= LINUX_ECHO;
if (bios->c_lflag & ECHOE)
lios->c_lflag |= LINUX_ECHOE;
if (bios->c_lflag & ECHOK)
lios->c_lflag |= LINUX_ECHOK;
if (bios->c_lflag & ECHONL)
lios->c_lflag |= LINUX_ECHONL;
if (bios->c_lflag & NOFLSH)
lios->c_lflag |= LINUX_NOFLSH;
if (bios->c_lflag & TOSTOP)
lios->c_lflag |= LINUX_TOSTOP;
if (bios->c_lflag & ECHOCTL)
lios->c_lflag |= LINUX_ECHOCTL;
if (bios->c_lflag & ECHOPRT)
lios->c_lflag |= LINUX_ECHOPRT;
if (bios->c_lflag & ECHOKE)
lios->c_lflag |= LINUX_ECHOKE;
if (bios->c_lflag & FLUSHO)
lios->c_lflag |= LINUX_FLUSHO;
if (bios->c_lflag & PENDIN)
lios->c_lflag |= LINUX_PENDIN;
if (bios->c_lflag & IEXTEN)
lios->c_lflag |= LINUX_IEXTEN;
for (i=0; i<LINUX_NCCS; i++)
lios->c_cc[i] = LINUX_POSIX_VDISABLE;
lios->c_cc[LINUX_VINTR] = bios->c_cc[VINTR];
lios->c_cc[LINUX_VQUIT] = bios->c_cc[VQUIT];
lios->c_cc[LINUX_VERASE] = bios->c_cc[VERASE];
lios->c_cc[LINUX_VKILL] = bios->c_cc[VKILL];
lios->c_cc[LINUX_VEOF] = bios->c_cc[VEOF];
lios->c_cc[LINUX_VEOL] = bios->c_cc[VEOL];
lios->c_cc[LINUX_VMIN] = bios->c_cc[VMIN];
lios->c_cc[LINUX_VTIME] = bios->c_cc[VTIME];
lios->c_cc[LINUX_VEOL2] = bios->c_cc[VEOL2];
lios->c_cc[LINUX_VSUSP] = bios->c_cc[VSUSP];
lios->c_cc[LINUX_VSTART] = bios->c_cc[VSTART];
lios->c_cc[LINUX_VSTOP] = bios->c_cc[VSTOP];
lios->c_cc[LINUX_VREPRINT] = bios->c_cc[VREPRINT];
lios->c_cc[LINUX_VDISCARD] = bios->c_cc[VDISCARD];
lios->c_cc[LINUX_VWERASE] = bios->c_cc[VWERASE];
lios->c_cc[LINUX_VLNEXT] = bios->c_cc[VLNEXT];
Add a hack to make ^T work for Linux binaries, enabled with 'compat.linux.preserve_vstatus=1' sysctl. MFC after: 2 weeks Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D21967
2019-12-16 20:15:44 +00:00
if (linux_preserve_vstatus)
lios->c_cc[LINUX_VSTATUS] = bios->c_cc[VSTATUS];
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
for (i=0; i<LINUX_NCCS; i++) {
Don't map LINUX_POSIX_VDISABLE to _POSIX_VDISABLE and vice versa for the VMIN and VTIME members of the c_cc array. These members are not special control characters. By not excluding these members we changed the noncanonical mode input processing when both members were 0 on entry (=LINUX_POSIX_VDISABLE) as we would remap them to 255 (=_POSIX_VDISABLE). See termios(4) case A for how that screws up your terminal I/O. PR: 23173 Originator: Bjarne Blichfeldt <bbl@dk.damgaard.com> Patch by: Boris Nikolaus <bn@dali.tellique.de> (original submission) Philipp Mergenthaler <philipp.mergenthaler@stud.uni-karlsruhe.de> Reminders by: Joseph Holland King <gte743n@cad.gatech.edu> MFC after: 5 days
2003-06-28 19:32:07 +00:00
if (i != LINUX_VMIN && i != LINUX_VTIME &&
lios->c_cc[i] == _POSIX_VDISABLE)
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
lios->c_cc[i] = LINUX_POSIX_VDISABLE;
}
lios->c_line = 0;
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
}
static void
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
linux_to_bsd_termios(struct linux_termios *lios, struct termios *bios)
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
{
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
int i;
bios->c_iflag = 0;
if (lios->c_iflag & LINUX_IGNBRK)
bios->c_iflag |= IGNBRK;
if (lios->c_iflag & LINUX_BRKINT)
bios->c_iflag |= BRKINT;
if (lios->c_iflag & LINUX_IGNPAR)
bios->c_iflag |= IGNPAR;
if (lios->c_iflag & LINUX_PARMRK)
bios->c_iflag |= PARMRK;
if (lios->c_iflag & LINUX_INPCK)
bios->c_iflag |= INPCK;
if (lios->c_iflag & LINUX_ISTRIP)
bios->c_iflag |= ISTRIP;
if (lios->c_iflag & LINUX_INLCR)
bios->c_iflag |= INLCR;
if (lios->c_iflag & LINUX_IGNCR)
bios->c_iflag |= IGNCR;
if (lios->c_iflag & LINUX_ICRNL)
bios->c_iflag |= ICRNL;
if (lios->c_iflag & LINUX_IXON)
bios->c_iflag |= IXON;
if (lios->c_iflag & LINUX_IXANY)
bios->c_iflag |= IXANY;
if (lios->c_iflag & LINUX_IXOFF)
bios->c_iflag |= IXOFF;
if (lios->c_iflag & LINUX_IMAXBEL)
bios->c_iflag |= IMAXBEL;
bios->c_oflag = 0;
if (lios->c_oflag & LINUX_OPOST)
bios->c_oflag |= OPOST;
if (lios->c_oflag & LINUX_ONLCR)
bios->c_oflag |= ONLCR;
if (lios->c_oflag & LINUX_XTABS)
Integrate the new MPSAFE TTY layer to the FreeBSD operating system. The last half year I've been working on a replacement TTY layer for the FreeBSD kernel. The new TTY layer was designed to improve the following: - Improved driver model: The old TTY layer has a driver model that is not abstract enough to make it friendly to use. A good example is the output path, where the device drivers directly access the output buffers. This means that an in-kernel PPP implementation must always convert network buffers into TTY buffers. If a PPP implementation would be built on top of the new TTY layer (still needs a hooks layer, though), it would allow the PPP implementation to directly hand the data to the TTY driver. - Improved hotplugging: With the old TTY layer, it isn't entirely safe to destroy TTY's from the system. This implementation has a two-step destructing design, where the driver first abandons the TTY. After all threads have left the TTY, the TTY layer calls a routine in the driver, which can be used to free resources (unit numbers, etc). The pts(4) driver also implements this feature, which means posix_openpt() will now return PTY's that are created on the fly. - Improved performance: One of the major improvements is the per-TTY mutex, which is expected to improve scalability when compared to the old Giant locking. Another change is the unbuffered copying to userspace, which is both used on TTY device nodes and PTY masters. Upgrading should be quite straightforward. Unlike previous versions, existing kernel configuration files do not need to be changed, except when they reference device drivers that are listed in UPDATING. Obtained from: //depot/projects/mpsafetty/... Approved by: philip (ex-mentor) Discussed: on the lists, at BSDCan, at the DevSummit Sponsored by: Snow B.V., the Netherlands dcons(4) fixed by: kan
2008-08-20 08:31:58 +00:00
bios->c_oflag |= TAB3;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
bios->c_cflag = (lios->c_cflag & LINUX_CSIZE) << 4;
if (lios->c_cflag & LINUX_CSTOPB)
bios->c_cflag |= CSTOPB;
if (lios->c_cflag & LINUX_CREAD)
bios->c_cflag |= CREAD;
if (lios->c_cflag & LINUX_PARENB)
bios->c_cflag |= PARENB;
if (lios->c_cflag & LINUX_PARODD)
bios->c_cflag |= PARODD;
if (lios->c_cflag & LINUX_HUPCL)
bios->c_cflag |= HUPCL;
if (lios->c_cflag & LINUX_CLOCAL)
bios->c_cflag |= CLOCAL;
if (lios->c_cflag & LINUX_CRTSCTS)
bios->c_cflag |= CRTSCTS;
bios->c_lflag = 0;
if (lios->c_lflag & LINUX_ISIG)
bios->c_lflag |= ISIG;
if (lios->c_lflag & LINUX_ICANON)
bios->c_lflag |= ICANON;
if (lios->c_lflag & LINUX_ECHO)
bios->c_lflag |= ECHO;
if (lios->c_lflag & LINUX_ECHOE)
bios->c_lflag |= ECHOE;
if (lios->c_lflag & LINUX_ECHOK)
bios->c_lflag |= ECHOK;
if (lios->c_lflag & LINUX_ECHONL)
bios->c_lflag |= ECHONL;
if (lios->c_lflag & LINUX_NOFLSH)
bios->c_lflag |= NOFLSH;
if (lios->c_lflag & LINUX_TOSTOP)
bios->c_lflag |= TOSTOP;
if (lios->c_lflag & LINUX_ECHOCTL)
bios->c_lflag |= ECHOCTL;
if (lios->c_lflag & LINUX_ECHOPRT)
bios->c_lflag |= ECHOPRT;
if (lios->c_lflag & LINUX_ECHOKE)
bios->c_lflag |= ECHOKE;
if (lios->c_lflag & LINUX_FLUSHO)
bios->c_lflag |= FLUSHO;
if (lios->c_lflag & LINUX_PENDIN)
bios->c_lflag |= PENDIN;
if (lios->c_lflag & LINUX_IEXTEN)
bios->c_lflag |= IEXTEN;
for (i=0; i<NCCS; i++)
bios->c_cc[i] = _POSIX_VDISABLE;
bios->c_cc[VINTR] = lios->c_cc[LINUX_VINTR];
bios->c_cc[VQUIT] = lios->c_cc[LINUX_VQUIT];
bios->c_cc[VERASE] = lios->c_cc[LINUX_VERASE];
bios->c_cc[VKILL] = lios->c_cc[LINUX_VKILL];
bios->c_cc[VEOF] = lios->c_cc[LINUX_VEOF];
bios->c_cc[VEOL] = lios->c_cc[LINUX_VEOL];
bios->c_cc[VMIN] = lios->c_cc[LINUX_VMIN];
bios->c_cc[VTIME] = lios->c_cc[LINUX_VTIME];
bios->c_cc[VEOL2] = lios->c_cc[LINUX_VEOL2];
bios->c_cc[VSUSP] = lios->c_cc[LINUX_VSUSP];
bios->c_cc[VSTART] = lios->c_cc[LINUX_VSTART];
bios->c_cc[VSTOP] = lios->c_cc[LINUX_VSTOP];
bios->c_cc[VREPRINT] = lios->c_cc[LINUX_VREPRINT];
bios->c_cc[VDISCARD] = lios->c_cc[LINUX_VDISCARD];
bios->c_cc[VWERASE] = lios->c_cc[LINUX_VWERASE];
bios->c_cc[VLNEXT] = lios->c_cc[LINUX_VLNEXT];
Add a hack to make ^T work for Linux binaries, enabled with 'compat.linux.preserve_vstatus=1' sysctl. MFC after: 2 weeks Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D21967
2019-12-16 20:15:44 +00:00
if (linux_preserve_vstatus)
bios->c_cc[VSTATUS] = lios->c_cc[LINUX_VSTATUS];
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
for (i=0; i<NCCS; i++) {
Don't map LINUX_POSIX_VDISABLE to _POSIX_VDISABLE and vice versa for the VMIN and VTIME members of the c_cc array. These members are not special control characters. By not excluding these members we changed the noncanonical mode input processing when both members were 0 on entry (=LINUX_POSIX_VDISABLE) as we would remap them to 255 (=_POSIX_VDISABLE). See termios(4) case A for how that screws up your terminal I/O. PR: 23173 Originator: Bjarne Blichfeldt <bbl@dk.damgaard.com> Patch by: Boris Nikolaus <bn@dali.tellique.de> (original submission) Philipp Mergenthaler <philipp.mergenthaler@stud.uni-karlsruhe.de> Reminders by: Joseph Holland King <gte743n@cad.gatech.edu> MFC after: 5 days
2003-06-28 19:32:07 +00:00
if (i != VMIN && i != VTIME &&
bios->c_cc[i] == LINUX_POSIX_VDISABLE)
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
bios->c_cc[i] = _POSIX_VDISABLE;
}
bios->c_ispeed = bios->c_ospeed =
linux_to_bsd_speed(lios->c_cflag & LINUX_CBAUD, sptab);
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
}
Add support for the old-style Linux termio (not termios) TCGETA etc. Also, LINUX_POSIX_VDISABLE is \0, FreeBSD's is 0xff. Convert between them. This enables some more programs to run, including the Livingston Portmaster utilities (PMtools). Submitted by: Robert Sanders <rsanders@mindspring.com>
1996-03-03 19:07:50 +00:00
static void
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
bsd_to_linux_termio(struct termios *bios, struct linux_termio *lio)
Add support for the old-style Linux termio (not termios) TCGETA etc. Also, LINUX_POSIX_VDISABLE is \0, FreeBSD's is 0xff. Convert between them. This enables some more programs to run, including the Livingston Portmaster utilities (PMtools). Submitted by: Robert Sanders <rsanders@mindspring.com>
1996-03-03 19:07:50 +00:00
{
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
struct linux_termios lios;
linuxulator: fix stack memory disclosure in linux_ioctl_termio admbugs: 765 Reported by: Vlad Tsyrklevich <vlad@tsyrklevich.net> Reviewed by: andrew MFC after: 1 day Security: Kernel stack memory disclosure Sponsored by: The FreeBSD Foundation
2019-01-21 16:21:03 +00:00
memset(lio, 0, sizeof(*lio));
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
bsd_to_linux_termios(bios, &lios);
lio->c_iflag = lios.c_iflag;
lio->c_oflag = lios.c_oflag;
lio->c_cflag = lios.c_cflag;
lio->c_lflag = lios.c_lflag;
lio->c_line = lios.c_line;
memcpy(lio->c_cc, lios.c_cc, LINUX_NCC);
Add support for the old-style Linux termio (not termios) TCGETA etc. Also, LINUX_POSIX_VDISABLE is \0, FreeBSD's is 0xff. Convert between them. This enables some more programs to run, including the Livingston Portmaster utilities (PMtools). Submitted by: Robert Sanders <rsanders@mindspring.com>
1996-03-03 19:07:50 +00:00
}
static void
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
linux_to_bsd_termio(struct linux_termio *lio, struct termios *bios)
{
struct linux_termios lios;
int i;
lios.c_iflag = lio->c_iflag;
lios.c_oflag = lio->c_oflag;
lios.c_cflag = lio->c_cflag;
lios.c_lflag = lio->c_lflag;
for (i=LINUX_NCC; i<LINUX_NCCS; i++)
lios.c_cc[i] = LINUX_POSIX_VDISABLE;
memcpy(lios.c_cc, lio->c_cc, LINUX_NCC);
linux_to_bsd_termios(&lios, bios);
}
static int
KSE Milestone 2 Note ALL MODULES MUST BE RECOMPILED make the kernel aware that there are smaller units of scheduling than the process. (but only allow one thread per process at this time). This is functionally equivalent to teh previousl -current except that there is a thread associated with each process. Sorry john! (your next MFC will be a doosie!) Reviewed by: peter@freebsd.org, dillon@freebsd.org X-MFC after: ha ha ha ha
2001-09-12 08:38:13 +00:00
linux_ioctl_termio(struct thread *td, struct linux_ioctl_args *args)
Add support for the old-style Linux termio (not termios) TCGETA etc. Also, LINUX_POSIX_VDISABLE is \0, FreeBSD's is 0xff. Convert between them. This enables some more programs to run, including the Livingston Portmaster utilities (PMtools). Submitted by: Robert Sanders <rsanders@mindspring.com>
1996-03-03 19:07:50 +00:00
{
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
struct termios bios;
struct linux_termios lios;
struct linux_termio lio;
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
struct file *fp;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
int error;
Eliminate the overhead of gratuitous repeated reinitialization of cap_rights - Add macros to allow preinitialization of cap_rights_t. - Convert most commonly used code paths to use preinitialized cap_rights_t. A 3.6% speedup in fstat was measured with this change. Reported by: mjg Reviewed by: oshogbo Approved by: sbruno MFC after: 1 month
2018-05-09 18:47:24 +00:00
error = fget(td, args->fd, &cap_ioctl_rights, &fp);
Change the cap_rights_t type from uint64_t to a structure that we can extend in the future in a backward compatible (API and ABI) way. The cap_rights_t represents capability rights. We used to use one bit to represent one right, but we are running out of spare bits. Currently the new structure provides place for 114 rights (so 50 more than the previous cap_rights_t), but it is possible to grow the structure to hold at least 285 rights, although we can make it even larger if 285 rights won't be enough. The structure definition looks like this: struct cap_rights { uint64_t cr_rights[CAP_RIGHTS_VERSION + 2]; }; The initial CAP_RIGHTS_VERSION is 0. The top two bits in the first element of the cr_rights[] array contain total number of elements in the array - 2. This means if those two bits are equal to 0, we have 2 array elements. The top two bits in all remaining array elements should be 0. The next five bits in all array elements contain array index. Only one bit is used and bit position in this five-bits range defines array index. This means there can be at most five array elements in the future. To define new right the CAPRIGHT() macro must be used. The macro takes two arguments - an array index and a bit to set, eg. #define CAP_PDKILL CAPRIGHT(1, 0x0000000000000800ULL) We still support aliases that combine few rights, but the rights have to belong to the same array element, eg: #define CAP_LOOKUP CAPRIGHT(0, 0x0000000000000400ULL) #define CAP_FCHMOD CAPRIGHT(0, 0x0000000000002000ULL) #define CAP_FCHMODAT (CAP_FCHMOD | CAP_LOOKUP) There is new API to manage the new cap_rights_t structure: cap_rights_t *cap_rights_init(cap_rights_t *rights, ...); void cap_rights_set(cap_rights_t *rights, ...); void cap_rights_clear(cap_rights_t *rights, ...); bool cap_rights_is_set(const cap_rights_t *rights, ...); bool cap_rights_is_valid(const cap_rights_t *rights); void cap_rights_merge(cap_rights_t *dst, const cap_rights_t *src); void cap_rights_remove(cap_rights_t *dst, const cap_rights_t *src); bool cap_rights_contains(const cap_rights_t *big, const cap_rights_t *little); Capability rights to the cap_rights_init(), cap_rights_set(), cap_rights_clear() and cap_rights_is_set() functions are provided by separating them with commas, eg: cap_rights_t rights; cap_rights_init(&rights, CAP_READ, CAP_WRITE, CAP_FSTAT); There is no need to terminate the list of rights, as those functions are actually macros that take care of the termination, eg: #define cap_rights_set(rights, ...) \ __cap_rights_set((rights), __VA_ARGS__, 0ULL) void __cap_rights_set(cap_rights_t *rights, ...); Thanks to using one bit as an array index we can assert in those functions that there are no two rights belonging to different array elements provided together. For example this is illegal and will be detected, because CAP_LOOKUP belongs to element 0 and CAP_PDKILL to element 1: cap_rights_init(&rights, CAP_LOOKUP | CAP_PDKILL); Providing several rights that belongs to the same array's element this way is correct, but is not advised. It should only be used for aliases definition. This commit also breaks compatibility with some existing Capsicum system calls, but I see no other way to do that. This should be fine as Capsicum is still experimental and this change is not going to 9.x. Sponsored by: The FreeBSD Foundation
2013-09-05 00:09:56 +00:00
if (error != 0)
Replace ffind_* with fget calls. Make fget MPsafe. Make fgetvp and fgetsock use the fget subsystem to reduce code bloat. Push giant down in fpathconf().
2002-01-14 00:13:45 +00:00
return (error);
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
switch (args->cmd & 0xffff) {
case LINUX_TCGETS:
In continuation of early fileop credential changes, modify fo_ioctl() to accept an 'active_cred' argument reflecting the credential of the thread initiating the ioctl operation. - Change fo_ioctl() to accept active_cred; change consumers of the fo_ioctl() interface to generally pass active_cred from td->td_ucred. - In fifofs, initialize filetmp.f_cred to ap->a_cred so that the invocations of soo_ioctl() are provided access to the calling f_cred. Pass ap->a_td->td_ucred as the active_cred, but note that this is required because we don't yet distinguish file_cred and active_cred in invoking VOP's. - Update kqueue_ioctl() for its new argument. - Update pipe_ioctl() for its new argument, pass active_cred rather than td_ucred to MAC for authorization. - Update soo_ioctl() for its new argument. - Update vn_ioctl() for its new argument, use active_cred rather than td->td_ucred to authorize VOP_IOCTL() and the associated VOP_GETATTR(). Obtained from: TrustedBSD Project Sponsored by: DARPA, NAI Labs
2002-08-17 02:36:16 +00:00
error = fo_ioctl(fp, TIOCGETA, (caddr_t)&bios, td->td_ucred,
td);
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
if (error)
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
bsd_to_linux_termios(&bios, &lios);
More caddr_t removal, in conjunction with copy{in,out}(9) this time. Also clean up some egregious casts and incorrect use of sizeof.
2003-03-03 09:14:26 +00:00
error = copyout(&lios, (void *)args->arg, sizeof(lios));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_TCSETS:
More caddr_t removal, in conjunction with copy{in,out}(9) this time. Also clean up some egregious casts and incorrect use of sizeof.
2003-03-03 09:14:26 +00:00
error = copyin((void *)args->arg, &lios, sizeof(lios));
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
if (error)
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
linux_to_bsd_termios(&lios, &bios);
In continuation of early fileop credential changes, modify fo_ioctl() to accept an 'active_cred' argument reflecting the credential of the thread initiating the ioctl operation. - Change fo_ioctl() to accept active_cred; change consumers of the fo_ioctl() interface to generally pass active_cred from td->td_ucred. - In fifofs, initialize filetmp.f_cred to ap->a_cred so that the invocations of soo_ioctl() are provided access to the calling f_cred. Pass ap->a_td->td_ucred as the active_cred, but note that this is required because we don't yet distinguish file_cred and active_cred in invoking VOP's. - Update kqueue_ioctl() for its new argument. - Update pipe_ioctl() for its new argument, pass active_cred rather than td_ucred to MAC for authorization. - Update soo_ioctl() for its new argument. - Update vn_ioctl() for its new argument, use active_cred rather than td->td_ucred to authorize VOP_IOCTL() and the associated VOP_GETATTR(). Obtained from: TrustedBSD Project Sponsored by: DARPA, NAI Labs
2002-08-17 02:36:16 +00:00
error = (fo_ioctl(fp, TIOCSETA, (caddr_t)&bios, td->td_ucred,
td));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_TCSETSW:
More caddr_t removal, in conjunction with copy{in,out}(9) this time. Also clean up some egregious casts and incorrect use of sizeof.
2003-03-03 09:14:26 +00:00
error = copyin((void *)args->arg, &lios, sizeof(lios));
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
if (error)
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
linux_to_bsd_termios(&lios, &bios);
In continuation of early fileop credential changes, modify fo_ioctl() to accept an 'active_cred' argument reflecting the credential of the thread initiating the ioctl operation. - Change fo_ioctl() to accept active_cred; change consumers of the fo_ioctl() interface to generally pass active_cred from td->td_ucred. - In fifofs, initialize filetmp.f_cred to ap->a_cred so that the invocations of soo_ioctl() are provided access to the calling f_cred. Pass ap->a_td->td_ucred as the active_cred, but note that this is required because we don't yet distinguish file_cred and active_cred in invoking VOP's. - Update kqueue_ioctl() for its new argument. - Update pipe_ioctl() for its new argument, pass active_cred rather than td_ucred to MAC for authorization. - Update soo_ioctl() for its new argument. - Update vn_ioctl() for its new argument, use active_cred rather than td->td_ucred to authorize VOP_IOCTL() and the associated VOP_GETATTR(). Obtained from: TrustedBSD Project Sponsored by: DARPA, NAI Labs
2002-08-17 02:36:16 +00:00
error = (fo_ioctl(fp, TIOCSETAW, (caddr_t)&bios, td->td_ucred,
td));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_TCSETSF:
More caddr_t removal, in conjunction with copy{in,out}(9) this time. Also clean up some egregious casts and incorrect use of sizeof.
2003-03-03 09:14:26 +00:00
error = copyin((void *)args->arg, &lios, sizeof(lios));
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
if (error)
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
linux_to_bsd_termios(&lios, &bios);
In continuation of early fileop credential changes, modify fo_ioctl() to accept an 'active_cred' argument reflecting the credential of the thread initiating the ioctl operation. - Change fo_ioctl() to accept active_cred; change consumers of the fo_ioctl() interface to generally pass active_cred from td->td_ucred. - In fifofs, initialize filetmp.f_cred to ap->a_cred so that the invocations of soo_ioctl() are provided access to the calling f_cred. Pass ap->a_td->td_ucred as the active_cred, but note that this is required because we don't yet distinguish file_cred and active_cred in invoking VOP's. - Update kqueue_ioctl() for its new argument. - Update pipe_ioctl() for its new argument, pass active_cred rather than td_ucred to MAC for authorization. - Update soo_ioctl() for its new argument. - Update vn_ioctl() for its new argument, use active_cred rather than td->td_ucred to authorize VOP_IOCTL() and the associated VOP_GETATTR(). Obtained from: TrustedBSD Project Sponsored by: DARPA, NAI Labs
2002-08-17 02:36:16 +00:00
error = (fo_ioctl(fp, TIOCSETAF, (caddr_t)&bios, td->td_ucred,
td));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_TCGETA:
In continuation of early fileop credential changes, modify fo_ioctl() to accept an 'active_cred' argument reflecting the credential of the thread initiating the ioctl operation. - Change fo_ioctl() to accept active_cred; change consumers of the fo_ioctl() interface to generally pass active_cred from td->td_ucred. - In fifofs, initialize filetmp.f_cred to ap->a_cred so that the invocations of soo_ioctl() are provided access to the calling f_cred. Pass ap->a_td->td_ucred as the active_cred, but note that this is required because we don't yet distinguish file_cred and active_cred in invoking VOP's. - Update kqueue_ioctl() for its new argument. - Update pipe_ioctl() for its new argument, pass active_cred rather than td_ucred to MAC for authorization. - Update soo_ioctl() for its new argument. - Update vn_ioctl() for its new argument, use active_cred rather than td->td_ucred to authorize VOP_IOCTL() and the associated VOP_GETATTR(). Obtained from: TrustedBSD Project Sponsored by: DARPA, NAI Labs
2002-08-17 02:36:16 +00:00
error = fo_ioctl(fp, TIOCGETA, (caddr_t)&bios, td->td_ucred,
td);
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
if (error)
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
bsd_to_linux_termio(&bios, &lio);
More caddr_t removal, in conjunction with copy{in,out}(9) this time. Also clean up some egregious casts and incorrect use of sizeof.
2003-03-03 09:14:26 +00:00
error = (copyout(&lio, (void *)args->arg, sizeof(lio)));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_TCSETA:
More caddr_t removal, in conjunction with copy{in,out}(9) this time. Also clean up some egregious casts and incorrect use of sizeof.
2003-03-03 09:14:26 +00:00
error = copyin((void *)args->arg, &lio, sizeof(lio));
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
if (error)
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
linux_to_bsd_termio(&lio, &bios);
In continuation of early fileop credential changes, modify fo_ioctl() to accept an 'active_cred' argument reflecting the credential of the thread initiating the ioctl operation. - Change fo_ioctl() to accept active_cred; change consumers of the fo_ioctl() interface to generally pass active_cred from td->td_ucred. - In fifofs, initialize filetmp.f_cred to ap->a_cred so that the invocations of soo_ioctl() are provided access to the calling f_cred. Pass ap->a_td->td_ucred as the active_cred, but note that this is required because we don't yet distinguish file_cred and active_cred in invoking VOP's. - Update kqueue_ioctl() for its new argument. - Update pipe_ioctl() for its new argument, pass active_cred rather than td_ucred to MAC for authorization. - Update soo_ioctl() for its new argument. - Update vn_ioctl() for its new argument, use active_cred rather than td->td_ucred to authorize VOP_IOCTL() and the associated VOP_GETATTR(). Obtained from: TrustedBSD Project Sponsored by: DARPA, NAI Labs
2002-08-17 02:36:16 +00:00
error = (fo_ioctl(fp, TIOCSETA, (caddr_t)&bios, td->td_ucred,
td));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_TCSETAW:
More caddr_t removal, in conjunction with copy{in,out}(9) this time. Also clean up some egregious casts and incorrect use of sizeof.
2003-03-03 09:14:26 +00:00
error = copyin((void *)args->arg, &lio, sizeof(lio));
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
if (error)
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
linux_to_bsd_termio(&lio, &bios);
In continuation of early fileop credential changes, modify fo_ioctl() to accept an 'active_cred' argument reflecting the credential of the thread initiating the ioctl operation. - Change fo_ioctl() to accept active_cred; change consumers of the fo_ioctl() interface to generally pass active_cred from td->td_ucred. - In fifofs, initialize filetmp.f_cred to ap->a_cred so that the invocations of soo_ioctl() are provided access to the calling f_cred. Pass ap->a_td->td_ucred as the active_cred, but note that this is required because we don't yet distinguish file_cred and active_cred in invoking VOP's. - Update kqueue_ioctl() for its new argument. - Update pipe_ioctl() for its new argument, pass active_cred rather than td_ucred to MAC for authorization. - Update soo_ioctl() for its new argument. - Update vn_ioctl() for its new argument, use active_cred rather than td->td_ucred to authorize VOP_IOCTL() and the associated VOP_GETATTR(). Obtained from: TrustedBSD Project Sponsored by: DARPA, NAI Labs
2002-08-17 02:36:16 +00:00
error = (fo_ioctl(fp, TIOCSETAW, (caddr_t)&bios, td->td_ucred,
td));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_TCSETAF:
More caddr_t removal, in conjunction with copy{in,out}(9) this time. Also clean up some egregious casts and incorrect use of sizeof.
2003-03-03 09:14:26 +00:00
error = copyin((void *)args->arg, &lio, sizeof(lio));
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
if (error)
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
linux_to_bsd_termio(&lio, &bios);
In continuation of early fileop credential changes, modify fo_ioctl() to accept an 'active_cred' argument reflecting the credential of the thread initiating the ioctl operation. - Change fo_ioctl() to accept active_cred; change consumers of the fo_ioctl() interface to generally pass active_cred from td->td_ucred. - In fifofs, initialize filetmp.f_cred to ap->a_cred so that the invocations of soo_ioctl() are provided access to the calling f_cred. Pass ap->a_td->td_ucred as the active_cred, but note that this is required because we don't yet distinguish file_cred and active_cred in invoking VOP's. - Update kqueue_ioctl() for its new argument. - Update pipe_ioctl() for its new argument, pass active_cred rather than td_ucred to MAC for authorization. - Update soo_ioctl() for its new argument. - Update vn_ioctl() for its new argument, use active_cred rather than td->td_ucred to authorize VOP_IOCTL() and the associated VOP_GETATTR(). Obtained from: TrustedBSD Project Sponsored by: DARPA, NAI Labs
2002-08-17 02:36:16 +00:00
error = (fo_ioctl(fp, TIOCSETAF, (caddr_t)&bios, td->td_ucred,
td));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
/* LINUX_TCSBRK */
case LINUX_TCXONC: {
switch (args->arg) {
case LINUX_TCOOFF:
args->cmd = TIOCSTOP;
break;
case LINUX_TCOON:
args->cmd = TIOCSTART;
break;
case LINUX_TCIOFF:
case LINUX_TCION: {
int c;
struct write_args wr;
In continuation of early fileop credential changes, modify fo_ioctl() to accept an 'active_cred' argument reflecting the credential of the thread initiating the ioctl operation. - Change fo_ioctl() to accept active_cred; change consumers of the fo_ioctl() interface to generally pass active_cred from td->td_ucred. - In fifofs, initialize filetmp.f_cred to ap->a_cred so that the invocations of soo_ioctl() are provided access to the calling f_cred. Pass ap->a_td->td_ucred as the active_cred, but note that this is required because we don't yet distinguish file_cred and active_cred in invoking VOP's. - Update kqueue_ioctl() for its new argument. - Update pipe_ioctl() for its new argument, pass active_cred rather than td_ucred to MAC for authorization. - Update soo_ioctl() for its new argument. - Update vn_ioctl() for its new argument, use active_cred rather than td->td_ucred to authorize VOP_IOCTL() and the associated VOP_GETATTR(). Obtained from: TrustedBSD Project Sponsored by: DARPA, NAI Labs
2002-08-17 02:36:16 +00:00
error = fo_ioctl(fp, TIOCGETA, (caddr_t)&bios,
td->td_ucred, td);
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
if (error)
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
fdrop(fp, td);
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
c = (args->arg == LINUX_TCIOFF) ? VSTOP : VSTART;
c = bios.c_cc[c];
if (c != _POSIX_VDISABLE) {
wr.fd = args->fd;
wr.buf = &c;
wr.nbyte = sizeof(c);
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_write(td, &wr));
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
} else
return (0);
}
default:
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
fdrop(fp, td);
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
return (EINVAL);
}
args->arg = 0;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = (sys_ioctl(td, (struct ioctl_args *)args));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
}
case LINUX_TCFLSH: {
In Linux, kernel parameters passed to ioctl are by value, while in FreeBSD they are passed by reference. Handle the difference within the linux_ioctl_termio on the LINUX_TCFLSH path. Submitted by: Jaroslav Drzik <jaro_AT_coop-voz_dot_sk>
2005-12-13 15:32:52 +00:00
int val;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
switch (args->arg) {
case LINUX_TCIFLUSH:
In Linux, kernel parameters passed to ioctl are by value, while in FreeBSD they are passed by reference. Handle the difference within the linux_ioctl_termio on the LINUX_TCFLSH path. Submitted by: Jaroslav Drzik <jaro_AT_coop-voz_dot_sk>
2005-12-13 15:32:52 +00:00
val = FREAD;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
break;
case LINUX_TCOFLUSH:
In Linux, kernel parameters passed to ioctl are by value, while in FreeBSD they are passed by reference. Handle the difference within the linux_ioctl_termio on the LINUX_TCFLSH path. Submitted by: Jaroslav Drzik <jaro_AT_coop-voz_dot_sk>
2005-12-13 15:32:52 +00:00
val = FWRITE;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
break;
case LINUX_TCIOFLUSH:
In Linux, kernel parameters passed to ioctl are by value, while in FreeBSD they are passed by reference. Handle the difference within the linux_ioctl_termio on the LINUX_TCFLSH path. Submitted by: Jaroslav Drzik <jaro_AT_coop-voz_dot_sk>
2005-12-13 15:32:52 +00:00
val = FREAD | FWRITE;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
break;
default:
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
fdrop(fp, td);
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
return (EINVAL);
}
In Linux, kernel parameters passed to ioctl are by value, while in FreeBSD they are passed by reference. Handle the difference within the linux_ioctl_termio on the LINUX_TCFLSH path. Submitted by: Jaroslav Drzik <jaro_AT_coop-voz_dot_sk>
2005-12-13 15:32:52 +00:00
error = (fo_ioctl(fp,TIOCFLUSH,(caddr_t)&val,td->td_ucred,td));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
}
case LINUX_TIOCEXCL:
args->cmd = TIOCEXCL;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = (sys_ioctl(td, (struct ioctl_args *)args));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_TIOCNXCL:
args->cmd = TIOCNXCL;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = (sys_ioctl(td, (struct ioctl_args *)args));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
Implement LINUX_TIOCSCTTY. PR: kern/42404
2002-09-05 02:51:25 +00:00
case LINUX_TIOCSCTTY:
args->cmd = TIOCSCTTY;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = (sys_ioctl(td, (struct ioctl_args *)args));
Implement LINUX_TIOCSCTTY. PR: kern/42404
2002-09-05 02:51:25 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_TIOCGPGRP:
args->cmd = TIOCGPGRP;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = (sys_ioctl(td, (struct ioctl_args *)args));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_TIOCSPGRP:
args->cmd = TIOCSPGRP;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = (sys_ioctl(td, (struct ioctl_args *)args));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
/* LINUX_TIOCOUTQ */
/* LINUX_TIOCSTI */
case LINUX_TIOCGWINSZ:
args->cmd = TIOCGWINSZ;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = (sys_ioctl(td, (struct ioctl_args *)args));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_TIOCSWINSZ:
args->cmd = TIOCSWINSZ;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = (sys_ioctl(td, (struct ioctl_args *)args));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_TIOCMGET:
args->cmd = TIOCMGET;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = (sys_ioctl(td, (struct ioctl_args *)args));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Add support for the old-style Linux termio (not termios) TCGETA etc. Also, LINUX_POSIX_VDISABLE is \0, FreeBSD's is 0xff. Convert between them. This enables some more programs to run, including the Livingston Portmaster utilities (PMtools). Submitted by: Robert Sanders <rsanders@mindspring.com>
1996-03-03 19:07:50 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_TIOCMBIS:
args->cmd = TIOCMBIS;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = (sys_ioctl(td, (struct ioctl_args *)args));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Add support for the old-style Linux termio (not termios) TCGETA etc. Also, LINUX_POSIX_VDISABLE is \0, FreeBSD's is 0xff. Convert between them. This enables some more programs to run, including the Livingston Portmaster utilities (PMtools). Submitted by: Robert Sanders <rsanders@mindspring.com>
1996-03-03 19:07:50 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_TIOCMBIC:
args->cmd = TIOCMBIC;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = (sys_ioctl(td, (struct ioctl_args *)args));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Add support for the old-style Linux termio (not termios) TCGETA etc. Also, LINUX_POSIX_VDISABLE is \0, FreeBSD's is 0xff. Convert between them. This enables some more programs to run, including the Livingston Portmaster utilities (PMtools). Submitted by: Robert Sanders <rsanders@mindspring.com>
1996-03-03 19:07:50 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_TIOCMSET:
args->cmd = TIOCMSET;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = (sys_ioctl(td, (struct ioctl_args *)args));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
/* TIOCGSOFTCAR */
/* TIOCSSOFTCAR */
case LINUX_FIONREAD: /* LINUX_TIOCINQ */
args->cmd = FIONREAD;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = (sys_ioctl(td, (struct ioctl_args *)args));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
/* LINUX_TIOCLINUX */
case LINUX_TIOCCONS:
args->cmd = TIOCCONS;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = (sys_ioctl(td, (struct ioctl_args *)args));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_TIOCGSERIAL: {
struct linux_serial_struct lss;
Fix kernel stack disclosures in the Linux and 4.3BSD compat layers. Submitted by: CTurt Security: SA-16:20 Security: SA-16:21
2016-05-31 16:56:30 +00:00
bzero(&lss, sizeof(lss));
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
lss.type = LINUX_PORT_16550A;
lss.flags = 0;
lss.close_delay = 0;
More caddr_t removal, in conjunction with copy{in,out}(9) this time. Also clean up some egregious casts and incorrect use of sizeof.
2003-03-03 09:14:26 +00:00
error = copyout(&lss, (void *)args->arg, sizeof(lss));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
}
case LINUX_TIOCSSERIAL: {
struct linux_serial_struct lss;
More caddr_t removal, in conjunction with copy{in,out}(9) this time. Also clean up some egregious casts and incorrect use of sizeof.
2003-03-03 09:14:26 +00:00
error = copyin((void *)args->arg, &lss, sizeof(lss));
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
if (error)
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
/* XXX - It really helps to have an implementation that
* does nothing. NOT!
*/
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
error = 0;
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
}
Add TIOCPKT and TIOCSPTLCK to the Linuxolator. We're very lucky, because the flags used by our TIOCPKT implementation are the same as flags used by Linux. We can safely enable TIOCPKT, assuming EXTPROC is not used. TIOCSPTLCK is used by unlockpt(). Because we don't need unlockpt() in our implementation, make this ioctl a no-op. Approved by: philip (mentor, implicit), rdivacky Obtained from: P4 (//depot/projects/mpsafetty/...)
2008-07-23 17:47:44 +00:00
case LINUX_TIOCPKT:
args->cmd = TIOCPKT;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = (sys_ioctl(td, (struct ioctl_args *)args));
Add TIOCPKT and TIOCSPTLCK to the Linuxolator. We're very lucky, because the flags used by our TIOCPKT implementation are the same as flags used by Linux. We can safely enable TIOCPKT, assuming EXTPROC is not used. TIOCSPTLCK is used by unlockpt(). Because we don't need unlockpt() in our implementation, make this ioctl a no-op. Approved by: philip (mentor, implicit), rdivacky Obtained from: P4 (//depot/projects/mpsafetty/...)
2008-07-23 17:47:44 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_FIONBIO:
args->cmd = FIONBIO;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = (sys_ioctl(td, (struct ioctl_args *)args));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_TIOCNOTTY:
args->cmd = TIOCNOTTY;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = (sys_ioctl(td, (struct ioctl_args *)args));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_TIOCSETD: {
int line;
switch (args->arg) {
case LINUX_N_TTY:
line = TTYDISC;
break;
case LINUX_N_SLIP:
line = SLIPDISC;
break;
case LINUX_N_PPP:
line = PPPDISC;
break;
default:
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
fdrop(fp, td);
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
return (EINVAL);
}
In continuation of early fileop credential changes, modify fo_ioctl() to accept an 'active_cred' argument reflecting the credential of the thread initiating the ioctl operation. - Change fo_ioctl() to accept active_cred; change consumers of the fo_ioctl() interface to generally pass active_cred from td->td_ucred. - In fifofs, initialize filetmp.f_cred to ap->a_cred so that the invocations of soo_ioctl() are provided access to the calling f_cred. Pass ap->a_td->td_ucred as the active_cred, but note that this is required because we don't yet distinguish file_cred and active_cred in invoking VOP's. - Update kqueue_ioctl() for its new argument. - Update pipe_ioctl() for its new argument, pass active_cred rather than td_ucred to MAC for authorization. - Update soo_ioctl() for its new argument. - Update vn_ioctl() for its new argument, use active_cred rather than td->td_ucred to authorize VOP_IOCTL() and the associated VOP_GETATTR(). Obtained from: TrustedBSD Project Sponsored by: DARPA, NAI Labs
2002-08-17 02:36:16 +00:00
error = (fo_ioctl(fp, TIOCSETD, (caddr_t)&line, td->td_ucred,
td));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
}
case LINUX_TIOCGETD: {
int linux_line;
int bsd_line = TTYDISC;
In continuation of early fileop credential changes, modify fo_ioctl() to accept an 'active_cred' argument reflecting the credential of the thread initiating the ioctl operation. - Change fo_ioctl() to accept active_cred; change consumers of the fo_ioctl() interface to generally pass active_cred from td->td_ucred. - In fifofs, initialize filetmp.f_cred to ap->a_cred so that the invocations of soo_ioctl() are provided access to the calling f_cred. Pass ap->a_td->td_ucred as the active_cred, but note that this is required because we don't yet distinguish file_cred and active_cred in invoking VOP's. - Update kqueue_ioctl() for its new argument. - Update pipe_ioctl() for its new argument, pass active_cred rather than td_ucred to MAC for authorization. - Update soo_ioctl() for its new argument. - Update vn_ioctl() for its new argument, use active_cred rather than td->td_ucred to authorize VOP_IOCTL() and the associated VOP_GETATTR(). Obtained from: TrustedBSD Project Sponsored by: DARPA, NAI Labs
2002-08-17 02:36:16 +00:00
error = fo_ioctl(fp, TIOCGETD, (caddr_t)&bsd_line,
td->td_ucred, td);
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
if (error)
Don't leak fp in case where fo_ioctl() returns an error. Reported by: C Turt <ecturt@gmail.com> MFC after: 1 week
2016-05-24 05:29:41 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
switch (bsd_line) {
case TTYDISC:
linux_line = LINUX_N_TTY;
break;
case SLIPDISC:
linux_line = LINUX_N_SLIP;
break;
case PPPDISC:
linux_line = LINUX_N_PPP;
break;
default:
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
fdrop(fp, td);
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
return (EINVAL);
}
More caddr_t removal, in conjunction with copy{in,out}(9) this time. Also clean up some egregious casts and incorrect use of sizeof.
2003-03-03 09:14:26 +00:00
error = (copyout(&linux_line, (void *)args->arg, sizeof(int)));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
}
/* LINUX_TCSBRKP */
/* LINUX_TIOCTTYGSTRUCT */
case LINUX_FIONCLEX:
args->cmd = FIONCLEX;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = (sys_ioctl(td, (struct ioctl_args *)args));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_FIOCLEX:
args->cmd = FIOCLEX;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = (sys_ioctl(td, (struct ioctl_args *)args));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_FIOASYNC:
args->cmd = FIOASYNC;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = (sys_ioctl(td, (struct ioctl_args *)args));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
/* LINUX_TIOCSERCONFIG */
/* LINUX_TIOCSERGWILD */
/* LINUX_TIOCSERSWILD */
/* LINUX_TIOCGLCKTRMIOS */
/* LINUX_TIOCSLCKTRMIOS */
Add BSD compatibility tty ioctls LINUX_TIOCSBRK and LINUX_TIOCCBRK. This addition appears to allow VMware 3 Workstation to operate with nmdm(4) as a virtual COM device. Tested by: Guido van Rooij
2004-02-19 12:38:12 +00:00
case LINUX_TIOCSBRK:
args->cmd = TIOCSBRK;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = (sys_ioctl(td, (struct ioctl_args *)args));
Add BSD compatibility tty ioctls LINUX_TIOCSBRK and LINUX_TIOCCBRK. This addition appears to allow VMware 3 Workstation to operate with nmdm(4) as a virtual COM device. Tested by: Guido van Rooij
2004-02-19 12:38:12 +00:00
break;
case LINUX_TIOCCBRK:
args->cmd = TIOCCBRK;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = (sys_ioctl(td, (struct ioctl_args *)args));
Add BSD compatibility tty ioctls LINUX_TIOCSBRK and LINUX_TIOCCBRK. This addition appears to allow VMware 3 Workstation to operate with nmdm(4) as a virtual COM device. Tested by: Guido van Rooij
2004-02-19 12:38:12 +00:00
break;
Linux compat bits needed to make linux programs use the new ptys : linux_ioctl.[ch] : Implement LINUX_TIOCGPTN, which returns the pty number linux_stats.c : - Return the magic number for devfs. - In various stats()-related functions, check that we're stating a file in /dev/pts, and if so, change the st_rdev field to match what linux expects to be there for a slave pty device. The glibc checks for this, and their openpty() fails if it is no correct.
2006-01-26 01:32:46 +00:00
case LINUX_TIOCGPTN: {
int nb;
Linuxolator whitespace cleanup A version of each of the MD files by necessity exists for each CPU architecture supported by the Linuxolator. Clean these up so that new architectures do not inherit whitespace issues. Clean up shared Linuxolator files while here. Sponsored by: Turing Robotic Industries Inc.
2018-02-05 17:29:12 +00:00
Linux compat bits needed to make linux programs use the new ptys : linux_ioctl.[ch] : Implement LINUX_TIOCGPTN, which returns the pty number linux_stats.c : - Return the magic number for devfs. - In various stats()-related functions, check that we're stating a file in /dev/pts, and if so, change the st_rdev field to match what linux expects to be there for a slave pty device. The glibc checks for this, and their openpty() fails if it is no correct.
2006-01-26 01:32:46 +00:00
error = fo_ioctl(fp, TIOCGPTN, (caddr_t)&nb, td->td_ucred, td);
if (!error)
error = copyout(&nb, (void *)args->arg,
sizeof(int));
break;
}
Add TIOCPKT and TIOCSPTLCK to the Linuxolator. We're very lucky, because the flags used by our TIOCPKT implementation are the same as flags used by Linux. We can safely enable TIOCPKT, assuming EXTPROC is not used. TIOCSPTLCK is used by unlockpt(). Because we don't need unlockpt() in our implementation, make this ioctl a no-op. Approved by: philip (mentor, implicit), rdivacky Obtained from: P4 (//depot/projects/mpsafetty/...)
2008-07-23 17:47:44 +00:00
case LINUX_TIOCSPTLCK:
/* Our unlockpt() does nothing. */
error = 0;
break;
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
default:
error = ENOIOCTL;
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
}
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
fdrop(fp, td);
return (error);
Add support for the old-style Linux termio (not termios) TCGETA etc. Also, LINUX_POSIX_VDISABLE is \0, FreeBSD's is 0xff. Convert between them. This enables some more programs to run, including the Livingston Portmaster utilities (PMtools). Submitted by: Robert Sanders <rsanders@mindspring.com>
1996-03-03 19:07:50 +00:00
}
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
/*
* CDROM related ioctls
*/
- Handle mixer read ioctls correctly. They have the same group, number and argument size as their write counterparts and were handled as write ioctls. - Emulate some cdrom ioctls.
1999-04-29 04:37:57 +00:00
struct linux_cdrom_msf
{
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
u_char cdmsf_min0;
u_char cdmsf_sec0;
u_char cdmsf_frame0;
u_char cdmsf_min1;
u_char cdmsf_sec1;
u_char cdmsf_frame1;
- Handle mixer read ioctls correctly. They have the same group, number and argument size as their write counterparts and were handled as write ioctls. - Emulate some cdrom ioctls.
1999-04-29 04:37:57 +00:00
};
struct linux_cdrom_tochdr
{
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
u_char cdth_trk0;
u_char cdth_trk1;
- Handle mixer read ioctls correctly. They have the same group, number and argument size as their write counterparts and were handled as write ioctls. - Emulate some cdrom ioctls.
1999-04-29 04:37:57 +00:00
};
union linux_cdrom_addr
{
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
struct {
u_char minute;
u_char second;
u_char frame;
} msf;
int lba;
- Handle mixer read ioctls correctly. They have the same group, number and argument size as their write counterparts and were handled as write ioctls. - Emulate some cdrom ioctls.
1999-04-29 04:37:57 +00:00
};
struct linux_cdrom_tocentry
{
Clean up whitespace, s/register //, refrain from strong urge to ANSIfy.
2003-03-02 15:56:49 +00:00
u_char cdte_track;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
u_char cdte_adr:4;
u_char cdte_ctrl:4;
Clean up whitespace, s/register //, refrain from strong urge to ANSIfy.
2003-03-02 15:56:49 +00:00
u_char cdte_format;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
union linux_cdrom_addr cdte_addr;
Clean up whitespace, s/register //, refrain from strong urge to ANSIfy.
2003-03-02 15:56:49 +00:00
u_char cdte_datamode;
- Handle mixer read ioctls correctly. They have the same group, number and argument size as their write counterparts and were handled as write ioctls. - Emulate some cdrom ioctls.
1999-04-29 04:37:57 +00:00
};
Implementation of the CDROMSUBCHNL ioctl.
1999-08-13 14:44:13 +00:00
struct linux_cdrom_subchnl
{
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
u_char cdsc_format;
u_char cdsc_audiostatus;
u_char cdsc_adr:4;
u_char cdsc_ctrl:4;
u_char cdsc_trk;
u_char cdsc_ind;
union linux_cdrom_addr cdsc_absaddr;
union linux_cdrom_addr cdsc_reladdr;
Implementation of the CDROMSUBCHNL ioctl.
1999-08-13 14:44:13 +00:00
};
Implement the CDROMREADAUDIO ioctl.
2002-10-19 21:11:43 +00:00
struct l_cdrom_read_audio {
union linux_cdrom_addr addr;
u_char addr_format;
l_int nframes;
u_char *buf;
};
Implement DVD-ROM ioctls. PR: 26955 Submitted by: Boris Nikolaus (email unknown)
2001-11-18 18:49:07 +00:00
struct l_dvd_layer {
u_char book_version:4;
u_char book_type:4;
u_char min_rate:4;
u_char disc_size:4;
u_char layer_type:4;
u_char track_path:1;
u_char nlayers:2;
u_char track_density:4;
u_char linear_density:4;
u_char bca:1;
u_int32_t start_sector;
u_int32_t end_sector;
u_int32_t end_sector_l0;
};
struct l_dvd_physical {
u_char type;
u_char layer_num;
struct l_dvd_layer layer[4];
};
struct l_dvd_copyright {
u_char type;
u_char layer_num;
u_char cpst;
u_char rmi;
};
struct l_dvd_disckey {
u_char type;
l_uint agid:2;
u_char value[2048];
};
struct l_dvd_bca {
u_char type;
l_int len;
u_char value[188];
};
struct l_dvd_manufact {
u_char type;
u_char layer_num;
l_int len;
u_char value[2048];
};
typedef union {
u_char type;
struct l_dvd_physical physical;
struct l_dvd_copyright copyright;
struct l_dvd_disckey disckey;
struct l_dvd_bca bca;
struct l_dvd_manufact manufact;
} l_dvd_struct;
typedef u_char l_dvd_key[5];
typedef u_char l_dvd_challenge[10];
struct l_dvd_lu_send_agid {
u_char type;
l_uint agid:2;
};
struct l_dvd_host_send_challenge {
u_char type;
l_uint agid:2;
l_dvd_challenge chal;
};
struct l_dvd_send_key {
u_char type;
l_uint agid:2;
l_dvd_key key;
};
struct l_dvd_lu_send_challenge {
u_char type;
l_uint agid:2;
l_dvd_challenge chal;
};
struct l_dvd_lu_send_title_key {
u_char type;
l_uint agid:2;
l_dvd_key title_key;
l_int lba;
l_uint cpm:1;
l_uint cp_sec:1;
l_uint cgms:2;
};
struct l_dvd_lu_send_asf {
u_char type;
l_uint agid:2;
l_uint asf:1;
};
struct l_dvd_host_send_rpcstate {
u_char type;
u_char pdrc;
};
struct l_dvd_lu_send_rpcstate {
u_char type:2;
u_char vra:3;
u_char ucca:3;
u_char region_mask;
u_char rpc_scheme;
};
typedef union {
u_char type;
struct l_dvd_lu_send_agid lsa;
struct l_dvd_host_send_challenge hsc;
struct l_dvd_send_key lsk;
struct l_dvd_lu_send_challenge lsc;
struct l_dvd_send_key hsk;
struct l_dvd_lu_send_title_key lstk;
struct l_dvd_lu_send_asf lsasf;
struct l_dvd_host_send_rpcstate hrpcs;
struct l_dvd_lu_send_rpcstate lrpcs;
} l_dvd_authinfo;
- Handle mixer read ioctls correctly. They have the same group, number and argument size as their write counterparts and were handled as write ioctls. - Emulate some cdrom ioctls.
1999-04-29 04:37:57 +00:00
static void
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
bsd_to_linux_msf_lba(u_char af, union msf_lba *bp, union linux_cdrom_addr *lp)
- Handle mixer read ioctls correctly. They have the same group, number and argument size as their write counterparts and were handled as write ioctls. - Emulate some cdrom ioctls.
1999-04-29 04:37:57 +00:00
{
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
if (af == CD_LBA_FORMAT)
lp->lba = bp->lba;
else {
lp->msf.minute = bp->msf.minute;
lp->msf.second = bp->msf.second;
lp->msf.frame = bp->msf.frame;
}
- Handle mixer read ioctls correctly. They have the same group, number and argument size as their write counterparts and were handled as write ioctls. - Emulate some cdrom ioctls.
1999-04-29 04:37:57 +00:00
}
static void
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
set_linux_cdrom_addr(union linux_cdrom_addr *addr, int format, int lba)
- Handle mixer read ioctls correctly. They have the same group, number and argument size as their write counterparts and were handled as write ioctls. - Emulate some cdrom ioctls.
1999-04-29 04:37:57 +00:00
{
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
if (format == LINUX_CDROM_MSF) {
addr->msf.frame = lba % 75;
lba /= 75;
lba += 2;
addr->msf.second = lba % 60;
addr->msf.minute = lba / 60;
} else
addr->lba = lba;
- Handle mixer read ioctls correctly. They have the same group, number and argument size as their write counterparts and were handled as write ioctls. - Emulate some cdrom ioctls.
1999-04-29 04:37:57 +00:00
}
Implement DVD-ROM ioctls. PR: 26955 Submitted by: Boris Nikolaus (email unknown)
2001-11-18 18:49:07 +00:00
static int
linux_to_bsd_dvd_struct(l_dvd_struct *lp, struct dvd_struct *bp)
{
bp->format = lp->type;
switch (bp->format) {
case DVD_STRUCT_PHYSICAL:
if (bp->layer_num >= 4)
return (EINVAL);
bp->layer_num = lp->physical.layer_num;
break;
case DVD_STRUCT_COPYRIGHT:
bp->layer_num = lp->copyright.layer_num;
break;
case DVD_STRUCT_DISCKEY:
bp->agid = lp->disckey.agid;
break;
case DVD_STRUCT_BCA:
case DVD_STRUCT_MANUFACT:
break;
default:
return (EINVAL);
}
return (0);
}
static int
bsd_to_linux_dvd_struct(struct dvd_struct *bp, l_dvd_struct *lp)
{
switch (bp->format) {
case DVD_STRUCT_PHYSICAL: {
struct dvd_layer *blp = (struct dvd_layer *)bp->data;
struct l_dvd_layer *llp = &lp->physical.layer[bp->layer_num];
memset(llp, 0, sizeof(*llp));
llp->book_version = blp->book_version;
llp->book_type = blp->book_type;
llp->min_rate = blp->max_rate;
llp->disc_size = blp->disc_size;
llp->layer_type = blp->layer_type;
llp->track_path = blp->track_path;
llp->nlayers = blp->nlayers;
llp->track_density = blp->track_density;
llp->linear_density = blp->linear_density;
llp->bca = blp->bca;
llp->start_sector = blp->start_sector;
llp->end_sector = blp->end_sector;
llp->end_sector_l0 = blp->end_sector_l0;
break;
}
case DVD_STRUCT_COPYRIGHT:
lp->copyright.cpst = bp->cpst;
lp->copyright.rmi = bp->rmi;
break;
case DVD_STRUCT_DISCKEY:
memcpy(lp->disckey.value, bp->data, sizeof(lp->disckey.value));
break;
case DVD_STRUCT_BCA:
lp->bca.len = bp->length;
memcpy(lp->bca.value, bp->data, sizeof(lp->bca.value));
break;
case DVD_STRUCT_MANUFACT:
lp->manufact.len = bp->length;
memcpy(lp->manufact.value, bp->data,
sizeof(lp->manufact.value));
Correct proper nouns in the Linuxulator - Capitalize Linux - Spell FreeBSD out in full - Address some style(9) on changed lines Sponsored by: Turing Robotic Industries Inc.
2018-02-22 02:24:17 +00:00
/* lp->manufact.layer_num is unused in Linux (redhat 7.0). */
Implement DVD-ROM ioctls. PR: 26955 Submitted by: Boris Nikolaus (email unknown)
2001-11-18 18:49:07 +00:00
break;
default:
return (EINVAL);
}
return (0);
}
static int
linux_to_bsd_dvd_authinfo(l_dvd_authinfo *lp, int *bcode,
struct dvd_authinfo *bp)
{
switch (lp->type) {
case LINUX_DVD_LU_SEND_AGID:
*bcode = DVDIOCREPORTKEY;
bp->format = DVD_REPORT_AGID;
bp->agid = lp->lsa.agid;
break;
case LINUX_DVD_HOST_SEND_CHALLENGE:
*bcode = DVDIOCSENDKEY;
bp->format = DVD_SEND_CHALLENGE;
bp->agid = lp->hsc.agid;
memcpy(bp->keychal, lp->hsc.chal, 10);
break;
case LINUX_DVD_LU_SEND_KEY1:
*bcode = DVDIOCREPORTKEY;
bp->format = DVD_REPORT_KEY1;
bp->agid = lp->lsk.agid;
break;
case LINUX_DVD_LU_SEND_CHALLENGE:
*bcode = DVDIOCREPORTKEY;
bp->format = DVD_REPORT_CHALLENGE;
bp->agid = lp->lsc.agid;
break;
case LINUX_DVD_HOST_SEND_KEY2:
*bcode = DVDIOCSENDKEY;
bp->format = DVD_SEND_KEY2;
bp->agid = lp->hsk.agid;
memcpy(bp->keychal, lp->hsk.key, 5);
break;
case LINUX_DVD_LU_SEND_TITLE_KEY:
*bcode = DVDIOCREPORTKEY;
bp->format = DVD_REPORT_TITLE_KEY;
bp->agid = lp->lstk.agid;
bp->lba = lp->lstk.lba;
break;
case LINUX_DVD_LU_SEND_ASF:
*bcode = DVDIOCREPORTKEY;
bp->format = DVD_REPORT_ASF;
bp->agid = lp->lsasf.agid;
break;
case LINUX_DVD_INVALIDATE_AGID:
*bcode = DVDIOCREPORTKEY;
bp->format = DVD_INVALIDATE_AGID;
bp->agid = lp->lsa.agid;
break;
case LINUX_DVD_LU_SEND_RPC_STATE:
*bcode = DVDIOCREPORTKEY;
bp->format = DVD_REPORT_RPC;
break;
case LINUX_DVD_HOST_SEND_RPC_STATE:
*bcode = DVDIOCSENDKEY;
bp->format = DVD_SEND_RPC;
bp->region = lp->hrpcs.pdrc;
break;
default:
return (EINVAL);
}
return (0);
}
static int
bsd_to_linux_dvd_authinfo(struct dvd_authinfo *bp, l_dvd_authinfo *lp)
{
switch (lp->type) {
case LINUX_DVD_LU_SEND_AGID:
lp->lsa.agid = bp->agid;
break;
case LINUX_DVD_HOST_SEND_CHALLENGE:
lp->type = LINUX_DVD_LU_SEND_KEY1;
break;
case LINUX_DVD_LU_SEND_KEY1:
memcpy(lp->lsk.key, bp->keychal, sizeof(lp->lsk.key));
break;
case LINUX_DVD_LU_SEND_CHALLENGE:
memcpy(lp->lsc.chal, bp->keychal, sizeof(lp->lsc.chal));
break;
case LINUX_DVD_HOST_SEND_KEY2:
lp->type = LINUX_DVD_AUTH_ESTABLISHED;
break;
case LINUX_DVD_LU_SEND_TITLE_KEY:
memcpy(lp->lstk.title_key, bp->keychal,
sizeof(lp->lstk.title_key));
lp->lstk.cpm = bp->cpm;
lp->lstk.cp_sec = bp->cp_sec;
lp->lstk.cgms = bp->cgms;
break;
case LINUX_DVD_LU_SEND_ASF:
lp->lsasf.asf = bp->asf;
break;
case LINUX_DVD_INVALIDATE_AGID:
break;
case LINUX_DVD_LU_SEND_RPC_STATE:
lp->lrpcs.type = bp->reg_type;
lp->lrpcs.vra = bp->vend_rsts;
lp->lrpcs.ucca = bp->user_rsts;
lp->lrpcs.region_mask = bp->region;
lp->lrpcs.rpc_scheme = bp->rpc_scheme;
break;
case LINUX_DVD_HOST_SEND_RPC_STATE:
break;
default:
return (EINVAL);
}
return (0);
}
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
static int
KSE Milestone 2 Note ALL MODULES MUST BE RECOMPILED make the kernel aware that there are smaller units of scheduling than the process. (but only allow one thread per process at this time). This is functionally equivalent to teh previousl -current except that there is a thread associated with each process. Sorry john! (your next MFC will be a doosie!) Reviewed by: peter@freebsd.org, dillon@freebsd.org X-MFC after: ha ha ha ha
2001-09-12 08:38:13 +00:00
linux_ioctl_cdrom(struct thread *td, struct linux_ioctl_args *args)
Implementation of the CDROMSUBCHNL ioctl.
1999-08-13 14:44:13 +00:00
{
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
struct file *fp;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
int error;
Eliminate the overhead of gratuitous repeated reinitialization of cap_rights - Add macros to allow preinitialization of cap_rights_t. - Convert most commonly used code paths to use preinitialized cap_rights_t. A 3.6% speedup in fstat was measured with this change. Reported by: mjg Reviewed by: oshogbo Approved by: sbruno MFC after: 1 month
2018-05-09 18:47:24 +00:00
error = fget(td, args->fd, &cap_ioctl_rights, &fp);
Change the cap_rights_t type from uint64_t to a structure that we can extend in the future in a backward compatible (API and ABI) way. The cap_rights_t represents capability rights. We used to use one bit to represent one right, but we are running out of spare bits. Currently the new structure provides place for 114 rights (so 50 more than the previous cap_rights_t), but it is possible to grow the structure to hold at least 285 rights, although we can make it even larger if 285 rights won't be enough. The structure definition looks like this: struct cap_rights { uint64_t cr_rights[CAP_RIGHTS_VERSION + 2]; }; The initial CAP_RIGHTS_VERSION is 0. The top two bits in the first element of the cr_rights[] array contain total number of elements in the array - 2. This means if those two bits are equal to 0, we have 2 array elements. The top two bits in all remaining array elements should be 0. The next five bits in all array elements contain array index. Only one bit is used and bit position in this five-bits range defines array index. This means there can be at most five array elements in the future. To define new right the CAPRIGHT() macro must be used. The macro takes two arguments - an array index and a bit to set, eg. #define CAP_PDKILL CAPRIGHT(1, 0x0000000000000800ULL) We still support aliases that combine few rights, but the rights have to belong to the same array element, eg: #define CAP_LOOKUP CAPRIGHT(0, 0x0000000000000400ULL) #define CAP_FCHMOD CAPRIGHT(0, 0x0000000000002000ULL) #define CAP_FCHMODAT (CAP_FCHMOD | CAP_LOOKUP) There is new API to manage the new cap_rights_t structure: cap_rights_t *cap_rights_init(cap_rights_t *rights, ...); void cap_rights_set(cap_rights_t *rights, ...); void cap_rights_clear(cap_rights_t *rights, ...); bool cap_rights_is_set(const cap_rights_t *rights, ...); bool cap_rights_is_valid(const cap_rights_t *rights); void cap_rights_merge(cap_rights_t *dst, const cap_rights_t *src); void cap_rights_remove(cap_rights_t *dst, const cap_rights_t *src); bool cap_rights_contains(const cap_rights_t *big, const cap_rights_t *little); Capability rights to the cap_rights_init(), cap_rights_set(), cap_rights_clear() and cap_rights_is_set() functions are provided by separating them with commas, eg: cap_rights_t rights; cap_rights_init(&rights, CAP_READ, CAP_WRITE, CAP_FSTAT); There is no need to terminate the list of rights, as those functions are actually macros that take care of the termination, eg: #define cap_rights_set(rights, ...) \ __cap_rights_set((rights), __VA_ARGS__, 0ULL) void __cap_rights_set(cap_rights_t *rights, ...); Thanks to using one bit as an array index we can assert in those functions that there are no two rights belonging to different array elements provided together. For example this is illegal and will be detected, because CAP_LOOKUP belongs to element 0 and CAP_PDKILL to element 1: cap_rights_init(&rights, CAP_LOOKUP | CAP_PDKILL); Providing several rights that belongs to the same array's element this way is correct, but is not advised. It should only be used for aliases definition. This commit also breaks compatibility with some existing Capsicum system calls, but I see no other way to do that. This should be fine as Capsicum is still experimental and this change is not going to 9.x. Sponsored by: The FreeBSD Foundation
2013-09-05 00:09:56 +00:00
if (error != 0)
Replace ffind_* with fget calls. Make fget MPsafe. Make fgetvp and fgetsock use the fget subsystem to reduce code bloat. Push giant down in fpathconf().
2002-01-14 00:13:45 +00:00
return (error);
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
switch (args->cmd & 0xffff) {
case LINUX_CDROMPAUSE:
args->cmd = CDIOCPAUSE;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = (sys_ioctl(td, (struct ioctl_args *)args));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_CDROMRESUME:
args->cmd = CDIOCRESUME;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = (sys_ioctl(td, (struct ioctl_args *)args));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_CDROMPLAYMSF:
args->cmd = CDIOCPLAYMSF;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = (sys_ioctl(td, (struct ioctl_args *)args));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_CDROMPLAYTRKIND:
args->cmd = CDIOCPLAYTRACKS;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = (sys_ioctl(td, (struct ioctl_args *)args));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_CDROMREADTOCHDR: {
struct ioc_toc_header th;
struct linux_cdrom_tochdr lth;
In continuation of early fileop credential changes, modify fo_ioctl() to accept an 'active_cred' argument reflecting the credential of the thread initiating the ioctl operation. - Change fo_ioctl() to accept active_cred; change consumers of the fo_ioctl() interface to generally pass active_cred from td->td_ucred. - In fifofs, initialize filetmp.f_cred to ap->a_cred so that the invocations of soo_ioctl() are provided access to the calling f_cred. Pass ap->a_td->td_ucred as the active_cred, but note that this is required because we don't yet distinguish file_cred and active_cred in invoking VOP's. - Update kqueue_ioctl() for its new argument. - Update pipe_ioctl() for its new argument, pass active_cred rather than td_ucred to MAC for authorization. - Update soo_ioctl() for its new argument. - Update vn_ioctl() for its new argument, use active_cred rather than td->td_ucred to authorize VOP_IOCTL() and the associated VOP_GETATTR(). Obtained from: TrustedBSD Project Sponsored by: DARPA, NAI Labs
2002-08-17 02:36:16 +00:00
error = fo_ioctl(fp, CDIOREADTOCHEADER, (caddr_t)&th,
td->td_ucred, td);
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
if (!error) {
lth.cdth_trk0 = th.starting_track;
lth.cdth_trk1 = th.ending_track;
More caddr_t removal, in conjunction with copy{in,out}(9) this time. Also clean up some egregious casts and incorrect use of sizeof.
2003-03-03 09:14:26 +00:00
copyout(&lth, (void *)args->arg, sizeof(lth));
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
}
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
}
case LINUX_CDROMREADTOCENTRY: {
Fix the VT_SETMODE/CDROMIOCTOCENTRY problem correctly. Reviewed by: tjr
2004-06-18 13:36:30 +00:00
struct linux_cdrom_tocentry lte;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
struct ioc_read_toc_single_entry irtse;
Fix the VT_SETMODE/CDROMIOCTOCENTRY problem correctly. Reviewed by: tjr
2004-06-18 13:36:30 +00:00
error = copyin((void *)args->arg, &lte, sizeof(lte));
if (error)
break;
irtse.address_format = lte.cdte_format;
irtse.track = lte.cdte_track;
error = fo_ioctl(fp, CDIOREADTOCENTRY, (caddr_t)&irtse,
td->td_ucred, td);
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
if (!error) {
lte.cdte_ctrl = irtse.entry.control;
lte.cdte_adr = irtse.entry.addr_type;
bsd_to_linux_msf_lba(irtse.address_format,
&irtse.entry.addr, &lte.cdte_addr);
Fix the VT_SETMODE/CDROMIOCTOCENTRY problem correctly. Reviewed by: tjr
2004-06-18 13:36:30 +00:00
error = copyout(&lte, (void *)args->arg, sizeof(lte));
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
}
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
}
case LINUX_CDROMSTOP:
args->cmd = CDIOCSTOP;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = (sys_ioctl(td, (struct ioctl_args *)args));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_CDROMSTART:
args->cmd = CDIOCSTART;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = (sys_ioctl(td, (struct ioctl_args *)args));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_CDROMEJECT:
args->cmd = CDIOCEJECT;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = (sys_ioctl(td, (struct ioctl_args *)args));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
/* LINUX_CDROMVOLCTRL */
case LINUX_CDROMSUBCHNL: {
struct linux_cdrom_subchnl sc;
struct ioc_read_subchannel bsdsc;
Boot away another stackgap (one of the lest ones in linuxlator/i386) by providing special version of CDIOCREADSUBCHANNEL ioctl(), which assumes that result has to be placed into kernel space not user space. In the long run more generic solution has to be designed WRT emulating various ioctl()s that operate on userspace buffers, but right now there is only one such ioctl() is emulated, so that it makes little sense. MFC after: 2 weeks
2005-01-30 08:12:37 +00:00
struct cd_sub_channel_info bsdinfo;
Remove the CDIOCREADSUBCHANNEL_SYSSPACE ioctl. This was added for emulation of Linux's CDROMSUBCHNL, but allows users with read access to a cd(4) device to overwrite kernel memory provided that the driver detects some media present. Reimplement CDROMSUBCHNL by bouncing the data from CDIOCREADSUBCHANNEL through the linux_cdrom_subchnl structure passed from userspace. admbugs: 768 Reported by: Alex Fortune Security: CVE-2019-5602 Security: FreeBSD-SA-19:11.cd_ioctl
2019-07-03 00:10:01 +00:00
error = copyin((void *)args->arg, &sc, sizeof(sc));
if (error)
break;
/*
* Invoke the native ioctl and bounce the returned data through
* the userspace buffer. This works because the Linux structure
* is the same size as our structures for the subchannel header
* and position data.
*/
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
bsdsc.address_format = CD_LBA_FORMAT;
bsdsc.data_format = CD_CURRENT_POSITION;
Avoid passing an uninitialized structure member to the real READSUBCHANNEL ioctl. This makes vmware work with SCSI CDROM drives. Approved by: jkh
2000-02-01 16:20:24 +00:00
bsdsc.track = 0;
Remove the CDIOCREADSUBCHANNEL_SYSSPACE ioctl. This was added for emulation of Linux's CDROMSUBCHNL, but allows users with read access to a cd(4) device to overwrite kernel memory provided that the driver detects some media present. Reimplement CDROMSUBCHNL by bouncing the data from CDIOCREADSUBCHANNEL through the linux_cdrom_subchnl structure passed from userspace. admbugs: 768 Reported by: Alex Fortune Security: CVE-2019-5602 Security: FreeBSD-SA-19:11.cd_ioctl
2019-07-03 00:10:01 +00:00
bsdsc.data_len = sizeof(sc);
bsdsc.data = (void *)args->arg;
error = fo_ioctl(fp, CDIOCREADSUBCHANNEL, (caddr_t)&bsdsc,
td->td_ucred, td);
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
if (error)
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Remove the CDIOCREADSUBCHANNEL_SYSSPACE ioctl. This was added for emulation of Linux's CDROMSUBCHNL, but allows users with read access to a cd(4) device to overwrite kernel memory provided that the driver detects some media present. Reimplement CDROMSUBCHNL by bouncing the data from CDIOCREADSUBCHANNEL through the linux_cdrom_subchnl structure passed from userspace. admbugs: 768 Reported by: Alex Fortune Security: CVE-2019-5602 Security: FreeBSD-SA-19:11.cd_ioctl
2019-07-03 00:10:01 +00:00
error = copyin((void *)args->arg, &bsdinfo, sizeof(bsdinfo));
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
if (error)
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Boot away another stackgap (one of the lest ones in linuxlator/i386) by providing special version of CDIOCREADSUBCHANNEL ioctl(), which assumes that result has to be placed into kernel space not user space. In the long run more generic solution has to be designed WRT emulating various ioctl()s that operate on userspace buffers, but right now there is only one such ioctl() is emulated, so that it makes little sense. MFC after: 2 weeks
2005-01-30 08:12:37 +00:00
sc.cdsc_audiostatus = bsdinfo.header.audio_status;
sc.cdsc_adr = bsdinfo.what.position.addr_type;
sc.cdsc_ctrl = bsdinfo.what.position.control;
sc.cdsc_trk = bsdinfo.what.position.track_number;
sc.cdsc_ind = bsdinfo.what.position.index_number;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
set_linux_cdrom_addr(&sc.cdsc_absaddr, sc.cdsc_format,
Boot away another stackgap (one of the lest ones in linuxlator/i386) by providing special version of CDIOCREADSUBCHANNEL ioctl(), which assumes that result has to be placed into kernel space not user space. In the long run more generic solution has to be designed WRT emulating various ioctl()s that operate on userspace buffers, but right now there is only one such ioctl() is emulated, so that it makes little sense. MFC after: 2 weeks
2005-01-30 08:12:37 +00:00
bsdinfo.what.position.absaddr.lba);
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
set_linux_cdrom_addr(&sc.cdsc_reladdr, sc.cdsc_format,
Boot away another stackgap (one of the lest ones in linuxlator/i386) by providing special version of CDIOCREADSUBCHANNEL ioctl(), which assumes that result has to be placed into kernel space not user space. In the long run more generic solution has to be designed WRT emulating various ioctl()s that operate on userspace buffers, but right now there is only one such ioctl() is emulated, so that it makes little sense. MFC after: 2 weeks
2005-01-30 08:12:37 +00:00
bsdinfo.what.position.reladdr.lba);
More caddr_t removal, in conjunction with copy{in,out}(9) this time. Also clean up some egregious casts and incorrect use of sizeof.
2003-03-03 09:14:26 +00:00
error = copyout(&sc, (void *)args->arg, sizeof(sc));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
}
/* LINUX_CDROMREADMODE2 */
/* LINUX_CDROMREADMODE1 */
We dont support CDROMREADAUDIO anymore.
2003-10-20 09:51:00 +00:00
/* LINUX_CDROMREADAUDIO */
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
/* LINUX_CDROMEJECT_SW */
/* LINUX_CDROMMULTISESSION */
/* LINUX_CDROM_GET_UPC */
case LINUX_CDROMRESET:
args->cmd = CDIOCRESET;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = (sys_ioctl(td, (struct ioctl_args *)args));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
/* LINUX_CDROMVOLREAD */
/* LINUX_CDROMREADRAW */
/* LINUX_CDROMREADCOOKED */
/* LINUX_CDROMSEEK */
/* LINUX_CDROMPLAYBLK */
/* LINUX_CDROMREADALL */
/* LINUX_CDROMCLOSETRAY */
/* LINUX_CDROMLOADFROMSLOT */
Implement DVD-ROM ioctls. PR: 26955 Submitted by: Boris Nikolaus (email unknown)
2001-11-18 18:49:07 +00:00
/* LINUX_CDROMGETSPINDOWN */
/* LINUX_CDROMSETSPINDOWN */
/* LINUX_CDROM_SET_OPTIONS */
/* LINUX_CDROM_CLEAR_OPTIONS */
/* LINUX_CDROM_SELECT_SPEED */
/* LINUX_CDROM_SELECT_DISC */
/* LINUX_CDROM_MEDIA_CHANGED */
/* LINUX_CDROM_DRIVE_STATUS */
/* LINUX_CDROM_DISC_STATUS */
/* LINUX_CDROM_CHANGER_NSLOTS */
/* LINUX_CDROM_LOCKDOOR */
/* LINUX_CDROM_DEBUG */
/* LINUX_CDROM_GET_CAPABILITY */
/* LINUX_CDROMAUDIOBUFSIZ */
case LINUX_DVD_READ_STRUCT: {
linux_ioctl_cdrom: reduce stack usage ... by moving two ~2KB structures from stack to heap allocation. I experienced stack overflow in linux emulation on i386 (8K stack) when LINUX_DVD_READ_STRUCT ioctl was performed on atapicam cd device and there was an error that resulted in additional quite heavy stack use in cam layer. Reviewed by: dchagin Approved by: jhb (mentor)
2009-05-27 15:23:12 +00:00
l_dvd_struct *lds;
struct dvd_struct *bds;
Implement DVD-ROM ioctls. PR: 26955 Submitted by: Boris Nikolaus (email unknown)
2001-11-18 18:49:07 +00:00
linux_ioctl_cdrom: reduce stack usage ... by moving two ~2KB structures from stack to heap allocation. I experienced stack overflow in linux emulation on i386 (8K stack) when LINUX_DVD_READ_STRUCT ioctl was performed on atapicam cd device and there was an error that resulted in additional quite heavy stack use in cam layer. Reviewed by: dchagin Approved by: jhb (mentor)
2009-05-27 15:23:12 +00:00
lds = malloc(sizeof(*lds), M_LINUX, M_WAITOK);
bds = malloc(sizeof(*bds), M_LINUX, M_WAITOK);
error = copyin((void *)args->arg, lds, sizeof(*lds));
Implement DVD-ROM ioctls. PR: 26955 Submitted by: Boris Nikolaus (email unknown)
2001-11-18 18:49:07 +00:00
if (error)
linux_ioctl_cdrom: reduce stack usage ... by moving two ~2KB structures from stack to heap allocation. I experienced stack overflow in linux emulation on i386 (8K stack) when LINUX_DVD_READ_STRUCT ioctl was performed on atapicam cd device and there was an error that resulted in additional quite heavy stack use in cam layer. Reviewed by: dchagin Approved by: jhb (mentor)
2009-05-27 15:23:12 +00:00
goto out;
error = linux_to_bsd_dvd_struct(lds, bds);
Implement DVD-ROM ioctls. PR: 26955 Submitted by: Boris Nikolaus (email unknown)
2001-11-18 18:49:07 +00:00
if (error)
linux_ioctl_cdrom: reduce stack usage ... by moving two ~2KB structures from stack to heap allocation. I experienced stack overflow in linux emulation on i386 (8K stack) when LINUX_DVD_READ_STRUCT ioctl was performed on atapicam cd device and there was an error that resulted in additional quite heavy stack use in cam layer. Reviewed by: dchagin Approved by: jhb (mentor)
2009-05-27 15:23:12 +00:00
goto out;
error = fo_ioctl(fp, DVDIOCREADSTRUCTURE, (caddr_t)bds,
In continuation of early fileop credential changes, modify fo_ioctl() to accept an 'active_cred' argument reflecting the credential of the thread initiating the ioctl operation. - Change fo_ioctl() to accept active_cred; change consumers of the fo_ioctl() interface to generally pass active_cred from td->td_ucred. - In fifofs, initialize filetmp.f_cred to ap->a_cred so that the invocations of soo_ioctl() are provided access to the calling f_cred. Pass ap->a_td->td_ucred as the active_cred, but note that this is required because we don't yet distinguish file_cred and active_cred in invoking VOP's. - Update kqueue_ioctl() for its new argument. - Update pipe_ioctl() for its new argument, pass active_cred rather than td_ucred to MAC for authorization. - Update soo_ioctl() for its new argument. - Update vn_ioctl() for its new argument, use active_cred rather than td->td_ucred to authorize VOP_IOCTL() and the associated VOP_GETATTR(). Obtained from: TrustedBSD Project Sponsored by: DARPA, NAI Labs
2002-08-17 02:36:16 +00:00
td->td_ucred, td);
Implement DVD-ROM ioctls. PR: 26955 Submitted by: Boris Nikolaus (email unknown)
2001-11-18 18:49:07 +00:00
if (error)
linux_ioctl_cdrom: reduce stack usage ... by moving two ~2KB structures from stack to heap allocation. I experienced stack overflow in linux emulation on i386 (8K stack) when LINUX_DVD_READ_STRUCT ioctl was performed on atapicam cd device and there was an error that resulted in additional quite heavy stack use in cam layer. Reviewed by: dchagin Approved by: jhb (mentor)
2009-05-27 15:23:12 +00:00
goto out;
error = bsd_to_linux_dvd_struct(bds, lds);
Implement DVD-ROM ioctls. PR: 26955 Submitted by: Boris Nikolaus (email unknown)
2001-11-18 18:49:07 +00:00
if (error)
linux_ioctl_cdrom: reduce stack usage ... by moving two ~2KB structures from stack to heap allocation. I experienced stack overflow in linux emulation on i386 (8K stack) when LINUX_DVD_READ_STRUCT ioctl was performed on atapicam cd device and there was an error that resulted in additional quite heavy stack use in cam layer. Reviewed by: dchagin Approved by: jhb (mentor)
2009-05-27 15:23:12 +00:00
goto out;
error = copyout(lds, (void *)args->arg, sizeof(*lds));
out:
free(bds, M_LINUX);
free(lds, M_LINUX);
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement DVD-ROM ioctls. PR: 26955 Submitted by: Boris Nikolaus (email unknown)
2001-11-18 18:49:07 +00:00
}
/* LINUX_DVD_WRITE_STRUCT */
case LINUX_DVD_AUTH: {
l_dvd_authinfo lda;
struct dvd_authinfo bda;
int bcode;
More caddr_t removal, in conjunction with copy{in,out}(9) this time. Also clean up some egregious casts and incorrect use of sizeof.
2003-03-03 09:14:26 +00:00
error = copyin((void *)args->arg, &lda, sizeof(lda));
Implement DVD-ROM ioctls. PR: 26955 Submitted by: Boris Nikolaus (email unknown)
2001-11-18 18:49:07 +00:00
if (error)
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement DVD-ROM ioctls. PR: 26955 Submitted by: Boris Nikolaus (email unknown)
2001-11-18 18:49:07 +00:00
error = linux_to_bsd_dvd_authinfo(&lda, &bcode, &bda);
if (error)
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
In continuation of early fileop credential changes, modify fo_ioctl() to accept an 'active_cred' argument reflecting the credential of the thread initiating the ioctl operation. - Change fo_ioctl() to accept active_cred; change consumers of the fo_ioctl() interface to generally pass active_cred from td->td_ucred. - In fifofs, initialize filetmp.f_cred to ap->a_cred so that the invocations of soo_ioctl() are provided access to the calling f_cred. Pass ap->a_td->td_ucred as the active_cred, but note that this is required because we don't yet distinguish file_cred and active_cred in invoking VOP's. - Update kqueue_ioctl() for its new argument. - Update pipe_ioctl() for its new argument, pass active_cred rather than td_ucred to MAC for authorization. - Update soo_ioctl() for its new argument. - Update vn_ioctl() for its new argument, use active_cred rather than td->td_ucred to authorize VOP_IOCTL() and the associated VOP_GETATTR(). Obtained from: TrustedBSD Project Sponsored by: DARPA, NAI Labs
2002-08-17 02:36:16 +00:00
error = fo_ioctl(fp, bcode, (caddr_t)&bda, td->td_ucred,
td);
Implement DVD-ROM ioctls. PR: 26955 Submitted by: Boris Nikolaus (email unknown)
2001-11-18 18:49:07 +00:00
if (error) {
if (lda.type == LINUX_DVD_HOST_SEND_KEY2) {
lda.type = LINUX_DVD_AUTH_FAILURE;
More caddr_t removal, in conjunction with copy{in,out}(9) this time. Also clean up some egregious casts and incorrect use of sizeof.
2003-03-03 09:14:26 +00:00
copyout(&lda, (void *)args->arg, sizeof(lda));
Implement DVD-ROM ioctls. PR: 26955 Submitted by: Boris Nikolaus (email unknown)
2001-11-18 18:49:07 +00:00
}
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement DVD-ROM ioctls. PR: 26955 Submitted by: Boris Nikolaus (email unknown)
2001-11-18 18:49:07 +00:00
}
error = bsd_to_linux_dvd_authinfo(&bda, &lda);
if (error)
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
More caddr_t removal, in conjunction with copy{in,out}(9) this time. Also clean up some egregious casts and incorrect use of sizeof.
2003-03-03 09:14:26 +00:00
error = copyout(&lda, (void *)args->arg, sizeof(lda));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement DVD-ROM ioctls. PR: 26955 Submitted by: Boris Nikolaus (email unknown)
2001-11-18 18:49:07 +00:00
}
Add the CAM 'SG' peripheral device. This device implements a subset of the Linux SCSI SG passthrough device API. The intention is to allow for both running of Linux apps that want to talk to /dev/sg* nodes, and to facilitate porting of apps from Linux to FreeBSD. As such, both native and linuxolator entry points and definitions are provided. Caveats: - This does not support the procfs and sysfs nodes that the Linux SG driver provides. Some Linux apps may rely on these for operation, others may only use them for informational purposes. - More ioctls need to be implemented. - Linux uses a naming scheme of "sg[a-z]" for devices, while FreeBSD uses a scheme of "sg[0-9]". Devfs aliasis (symlinks) are automatically created to link the two together. However, tools like camcontrol only see the native names. - Some operations were originally designed to return byte counts or other data directly as the syscall return value. The linuxolator doesn't appear to support this well, so this driver just punts for these cases. Now that the driver is in place, others are welcome to add missing functionality. Thanks to Roman Divacky for pushing this work along.
2007-04-07 19:40:58 +00:00
case LINUX_SCSI_GET_BUS_NUMBER:
Overhaul CAM SG driver IOCTL interfaces. Make it really work for native FreeBSD programs. Before this it was broken for years due to different number of pointer dereferences in Linux and FreeBSD IOCTL paths, permanently returning errors to FreeBSD programs. This change breaks the driver FreeBSD IOCTL ABI, making it more strict, but since it was not working any way -- who bother. Add shims for 32-bit programs on 64-bit host, translating the argument of the SG_IO IOCTL for both FreeBSD and Linux ABIs. With this change I was able to run 32-bit Linux sg3_utils tools and simple 32 and 64-bit FreeBSD test tools on both 32 and 64-bit FreeBSD systems. MFC after: 1 month
2014-06-02 19:53:53 +00:00
{
struct sg_scsi_id id;
error = fo_ioctl(fp, SG_GET_SCSI_ID, (caddr_t)&id,
td->td_ucred, td);
if (error)
break;
error = copyout(&id.channel, (void *)args->arg, sizeof(int));
break;
}
Add the CAM 'SG' peripheral device. This device implements a subset of the Linux SCSI SG passthrough device API. The intention is to allow for both running of Linux apps that want to talk to /dev/sg* nodes, and to facilitate porting of apps from Linux to FreeBSD. As such, both native and linuxolator entry points and definitions are provided. Caveats: - This does not support the procfs and sysfs nodes that the Linux SG driver provides. Some Linux apps may rely on these for operation, others may only use them for informational purposes. - More ioctls need to be implemented. - Linux uses a naming scheme of "sg[a-z]" for devices, while FreeBSD uses a scheme of "sg[0-9]". Devfs aliasis (symlinks) are automatically created to link the two together. However, tools like camcontrol only see the native names. - Some operations were originally designed to return byte counts or other data directly as the syscall return value. The linuxolator doesn't appear to support this well, so this driver just punts for these cases. Now that the driver is in place, others are welcome to add missing functionality. Thanks to Roman Divacky for pushing this work along.
2007-04-07 19:40:58 +00:00
case LINUX_SCSI_GET_IDLUN:
Overhaul CAM SG driver IOCTL interfaces. Make it really work for native FreeBSD programs. Before this it was broken for years due to different number of pointer dereferences in Linux and FreeBSD IOCTL paths, permanently returning errors to FreeBSD programs. This change breaks the driver FreeBSD IOCTL ABI, making it more strict, but since it was not working any way -- who bother. Add shims for 32-bit programs on 64-bit host, translating the argument of the SG_IO IOCTL for both FreeBSD and Linux ABIs. With this change I was able to run 32-bit Linux sg3_utils tools and simple 32 and 64-bit FreeBSD test tools on both 32 and 64-bit FreeBSD systems. MFC after: 1 month
2014-06-02 19:53:53 +00:00
{
struct sg_scsi_id id;
struct scsi_idlun idl;
error = fo_ioctl(fp, SG_GET_SCSI_ID, (caddr_t)&id,
td->td_ucred, td);
if (error)
break;
idl.dev_id = (id.scsi_id & 0xff) + ((id.lun & 0xff) << 8) +
((id.channel & 0xff) << 16) + ((id.host_no & 0xff) << 24);
idl.host_unique_id = id.host_no;
error = copyout(&idl, (void *)args->arg, sizeof(idl));
Add the CAM 'SG' peripheral device. This device implements a subset of the Linux SCSI SG passthrough device API. The intention is to allow for both running of Linux apps that want to talk to /dev/sg* nodes, and to facilitate porting of apps from Linux to FreeBSD. As such, both native and linuxolator entry points and definitions are provided. Caveats: - This does not support the procfs and sysfs nodes that the Linux SG driver provides. Some Linux apps may rely on these for operation, others may only use them for informational purposes. - More ioctls need to be implemented. - Linux uses a naming scheme of "sg[a-z]" for devices, while FreeBSD uses a scheme of "sg[0-9]". Devfs aliasis (symlinks) are automatically created to link the two together. However, tools like camcontrol only see the native names. - Some operations were originally designed to return byte counts or other data directly as the syscall return value. The linuxolator doesn't appear to support this well, so this driver just punts for these cases. Now that the driver is in place, others are welcome to add missing functionality. Thanks to Roman Divacky for pushing this work along.
2007-04-07 19:40:58 +00:00
break;
Overhaul CAM SG driver IOCTL interfaces. Make it really work for native FreeBSD programs. Before this it was broken for years due to different number of pointer dereferences in Linux and FreeBSD IOCTL paths, permanently returning errors to FreeBSD programs. This change breaks the driver FreeBSD IOCTL ABI, making it more strict, but since it was not working any way -- who bother. Add shims for 32-bit programs on 64-bit host, translating the argument of the SG_IO IOCTL for both FreeBSD and Linux ABIs. With this change I was able to run 32-bit Linux sg3_utils tools and simple 32 and 64-bit FreeBSD test tools on both 32 and 64-bit FreeBSD systems. MFC after: 1 month
2014-06-02 19:53:53 +00:00
}
Add the CAM 'SG' peripheral device. This device implements a subset of the Linux SCSI SG passthrough device API. The intention is to allow for both running of Linux apps that want to talk to /dev/sg* nodes, and to facilitate porting of apps from Linux to FreeBSD. As such, both native and linuxolator entry points and definitions are provided. Caveats: - This does not support the procfs and sysfs nodes that the Linux SG driver provides. Some Linux apps may rely on these for operation, others may only use them for informational purposes. - More ioctls need to be implemented. - Linux uses a naming scheme of "sg[a-z]" for devices, while FreeBSD uses a scheme of "sg[0-9]". Devfs aliasis (symlinks) are automatically created to link the two together. However, tools like camcontrol only see the native names. - Some operations were originally designed to return byte counts or other data directly as the syscall return value. The linuxolator doesn't appear to support this well, so this driver just punts for these cases. Now that the driver is in place, others are welcome to add missing functionality. Thanks to Roman Divacky for pushing this work along.
2007-04-07 19:40:58 +00:00
Implement DVD-ROM ioctls. PR: 26955 Submitted by: Boris Nikolaus (email unknown)
2001-11-18 18:49:07 +00:00
/* LINUX_CDROM_SEND_PACKET */
/* LINUX_CDROM_NEXT_WRITABLE */
/* LINUX_CDROM_LAST_WRITTEN */
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
default:
error = ENOIOCTL;
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
}
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
fdrop(fp, td);
return (error);
Implementation of the CDROMSUBCHNL ioctl.
1999-08-13 14:44:13 +00:00
}
- Add support for IPC_64 extensions into shmctl(2), semctl(2) and msgctl(2); - add wrappers for mmap2(2) and ftruncate64(2) system calls; - don't spam console with printf's when VFAT_READDIR_BOTH ioctl(2) is invoked; - add support for SOUND_MIXER_READ_STEREODEVS ioctl(2); - make msgctl(IPC_STAT) and IPC_SET actually working by converting from BSD msqid_ds to Linux and vice versa; - properly return EINVAL if semget(2) is called with nsems being negative. Reviewed by: marcel Approved by: marcel Tested with: LSB runtime test
2002-10-11 11:43:09 +00:00
static int
linux_ioctl_vfat(struct thread *td, struct linux_ioctl_args *args)
{
return (ENOTTY);
}
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
/*
* Sound related ioctls
*/
Implement SOUND_MIXER_INFO ioctl in compat layer.
2005-04-13 04:33:06 +00:00
struct linux_old_mixer_info {
char id[16];
char name[32];
};
The MI/MD split wasn't perfect and the MI files need hacks for the AlphaLinux compat bits. This will be better cleaned up soon. Agreed to what ever was necessary by: marcel
2000-11-01 19:48:35 +00:00
static u_int32_t dirbits[4] = { IOC_VOID, IOC_IN, IOC_OUT, IOC_INOUT };
- Handle mixer read ioctls correctly. They have the same group, number and argument size as their write counterparts and were handled as write ioctls. - Emulate some cdrom ioctls.
1999-04-29 04:37:57 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
#define SETDIR(c) (((c) & ~IOC_DIRMASK) | dirbits[args->cmd >> 30])
- Handle mixer read ioctls correctly. They have the same group, number and argument size as their write counterparts and were handled as write ioctls. - Emulate some cdrom ioctls.
1999-04-29 04:37:57 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
static int
KSE Milestone 2 Note ALL MODULES MUST BE RECOMPILED make the kernel aware that there are smaller units of scheduling than the process. (but only allow one thread per process at this time). This is functionally equivalent to teh previousl -current except that there is a thread associated with each process. Sorry john! (your next MFC will be a doosie!) Reviewed by: peter@freebsd.org, dillon@freebsd.org X-MFC after: ha ha ha ha
2001-09-12 08:38:13 +00:00
linux_ioctl_sound(struct thread *td, struct linux_ioctl_args *args)
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
{
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
switch (args->cmd & 0xffff) {
Add support for the SIOCGIFHWADDR ioctl, commonly used by license managers to obtain the host's ethernet address as a key. Note that this implementation takes the first hardware address for the first ethernet interface found, and disregards the interface name that may be passed in, as linux ethernet devices are all "ethX".
1997-06-02 06:31:49 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SOUND_MIXER_WRITE_VOLUME:
args->cmd = SETDIR(SOUND_MIXER_WRITE_VOLUME);
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
Add support for the SIOCGIFHWADDR ioctl, commonly used by license managers to obtain the host's ethernet address as a key. Note that this implementation takes the first hardware address for the first ethernet interface found, and disregards the interface name that may be passed in, as linux ethernet devices are all "ethX".
1997-06-02 06:31:49 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SOUND_MIXER_WRITE_BASS:
args->cmd = SETDIR(SOUND_MIXER_WRITE_BASS);
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
Add support for the SIOCGIFHWADDR ioctl, commonly used by license managers to obtain the host's ethernet address as a key. Note that this implementation takes the first hardware address for the first ethernet interface found, and disregards the interface name that may be passed in, as linux ethernet devices are all "ethX".
1997-06-02 06:31:49 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SOUND_MIXER_WRITE_TREBLE:
args->cmd = SETDIR(SOUND_MIXER_WRITE_TREBLE);
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SOUND_MIXER_WRITE_SYNTH:
args->cmd = SETDIR(SOUND_MIXER_WRITE_SYNTH);
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SOUND_MIXER_WRITE_PCM:
args->cmd = SETDIR(SOUND_MIXER_WRITE_PCM);
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SOUND_MIXER_WRITE_SPEAKER:
args->cmd = SETDIR(SOUND_MIXER_WRITE_SPEAKER);
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SOUND_MIXER_WRITE_LINE:
args->cmd = SETDIR(SOUND_MIXER_WRITE_LINE);
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SOUND_MIXER_WRITE_MIC:
args->cmd = SETDIR(SOUND_MIXER_WRITE_MIC);
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SOUND_MIXER_WRITE_CD:
args->cmd = SETDIR(SOUND_MIXER_WRITE_CD);
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SOUND_MIXER_WRITE_IMIX:
args->cmd = SETDIR(SOUND_MIXER_WRITE_IMIX);
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SOUND_MIXER_WRITE_ALTPCM:
args->cmd = SETDIR(SOUND_MIXER_WRITE_ALTPCM);
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SOUND_MIXER_WRITE_RECLEV:
args->cmd = SETDIR(SOUND_MIXER_WRITE_RECLEV);
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SOUND_MIXER_WRITE_IGAIN:
args->cmd = SETDIR(SOUND_MIXER_WRITE_IGAIN);
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SOUND_MIXER_WRITE_OGAIN:
args->cmd = SETDIR(SOUND_MIXER_WRITE_OGAIN);
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SOUND_MIXER_WRITE_LINE1:
args->cmd = SETDIR(SOUND_MIXER_WRITE_LINE1);
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SOUND_MIXER_WRITE_LINE2:
args->cmd = SETDIR(SOUND_MIXER_WRITE_LINE2);
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SOUND_MIXER_WRITE_LINE3:
args->cmd = SETDIR(SOUND_MIXER_WRITE_LINE3);
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
Implement SOUND_MIXER_INFO ioctl in compat layer.
2005-04-13 04:33:06 +00:00
case LINUX_SOUND_MIXER_INFO: {
/* Key on encoded length */
switch ((args->cmd >> 16) & 0x1fff) {
case 0x005c: { /* SOUND_MIXER_INFO */
Remake Linux' SOUND_MIXER_INFO IOCTL as a wrapper around new FreeBSD's one. Submitted by: Dmitry Luhtionov <dmitryluhtionov@gmail.com> MFC after: 3 days
2014-09-24 08:18:11 +00:00
args->cmd = SOUND_MIXER_INFO;
return (sys_ioctl(td, (struct ioctl_args *)args));
Implement SOUND_MIXER_INFO ioctl in compat layer.
2005-04-13 04:33:06 +00:00
}
case 0x0030: { /* SOUND_OLD_MIXER_INFO */
struct linux_old_mixer_info info;
bzero(&info, sizeof(info));
strncpy(info.id, "OSS", sizeof(info.id) - 1);
strncpy(info.name, "FreeBSD OSS Mixer", sizeof(info.name) - 1);
copyout(&info, (void *)args->arg, sizeof(info));
Fix the LINUX_SOUND_MIXER_INFO ioctl to return success after the information is set to FreeBSD. It had been falling through to the end of linux_ioctl_sound() and returning ENOIOCTL. Noticed when running the Linux ALSA amixer tool. Add a LINUX_SOUND_MIXER_READ_CAPS ioctl which is used by the Skype v2.1.0.81 binary. Reviewed by: gavin MFC after: 2 weeks
2010-12-30 02:18:04 +00:00
return (0);
Implement SOUND_MIXER_INFO ioctl in compat layer.
2005-04-13 04:33:06 +00:00
}
default:
return (ENOIOCTL);
}
break;
}
Add a stub for the Linux SOUND_MIXER_INFO ioctl (even though we don't actually implement it), as some applications, such as RealProducer, expect to be able to use it. PR: kern/65971 Submitted by: Matt Wright
2004-06-18 14:25:44 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_OSS_GETVERSION: {
- Change the linux_[gs]et_os{name, release, s_version}() functions to take a thread instead of a proc for their first argument. - Add a mutex to protect the system-wide Linux osname, osrelease, and oss_version variables. - Change linux_get_prison() to take a thread instead of a proc for its first argument and to use td_ucred rather than p_ucred. This is ok because a thread's prison does not change even though it's ucred might. - Also, change linux_get_prison() to return a struct prison * instead of a struct linux_prison * since it returns with the struct prison locked and this makes it easier to safely unlock the prison when we are done messing with it.
2003-03-13 22:45:43 +00:00
int version = linux_get_oss_version(td);
More caddr_t removal, in conjunction with copy{in,out}(9) this time. Also clean up some egregious casts and incorrect use of sizeof.
2003-03-03 09:14:26 +00:00
return (copyout(&version, (void *)args->arg, sizeof(int)));
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
}
My first shot at get sound to work on the emulator. Inspired by the work Amancio Hasty has done, but implemented somewhat differently.
1995-12-29 22:12:14 +00:00
- Add support for IPC_64 extensions into shmctl(2), semctl(2) and msgctl(2); - add wrappers for mmap2(2) and ftruncate64(2) system calls; - don't spam console with printf's when VFAT_READDIR_BOTH ioctl(2) is invoked; - add support for SOUND_MIXER_READ_STEREODEVS ioctl(2); - make msgctl(IPC_STAT) and IPC_SET actually working by converting from BSD msqid_ds to Linux and vice versa; - properly return EINVAL if semget(2) is called with nsems being negative. Reviewed by: marcel Approved by: marcel Tested with: LSB runtime test
2002-10-11 11:43:09 +00:00
case LINUX_SOUND_MIXER_READ_STEREODEVS:
args->cmd = SOUND_MIXER_READ_STEREODEVS;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
- Add support for IPC_64 extensions into shmctl(2), semctl(2) and msgctl(2); - add wrappers for mmap2(2) and ftruncate64(2) system calls; - don't spam console with printf's when VFAT_READDIR_BOTH ioctl(2) is invoked; - add support for SOUND_MIXER_READ_STEREODEVS ioctl(2); - make msgctl(IPC_STAT) and IPC_SET actually working by converting from BSD msqid_ds to Linux and vice versa; - properly return EINVAL if semget(2) is called with nsems being negative. Reviewed by: marcel Approved by: marcel Tested with: LSB runtime test
2002-10-11 11:43:09 +00:00
Fix the LINUX_SOUND_MIXER_INFO ioctl to return success after the information is set to FreeBSD. It had been falling through to the end of linux_ioctl_sound() and returning ENOIOCTL. Noticed when running the Linux ALSA amixer tool. Add a LINUX_SOUND_MIXER_READ_CAPS ioctl which is used by the Skype v2.1.0.81 binary. Reviewed by: gavin MFC after: 2 weeks
2010-12-30 02:18:04 +00:00
case LINUX_SOUND_MIXER_READ_CAPS:
args->cmd = SOUND_MIXER_READ_CAPS;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
Fix the LINUX_SOUND_MIXER_INFO ioctl to return success after the information is set to FreeBSD. It had been falling through to the end of linux_ioctl_sound() and returning ENOIOCTL. Noticed when running the Linux ALSA amixer tool. Add a LINUX_SOUND_MIXER_READ_CAPS ioctl which is used by the Skype v2.1.0.81 binary. Reviewed by: gavin MFC after: 2 weeks
2010-12-30 02:18:04 +00:00
Add stub for Linux SOUND_MIXER_READ_RECMASK, required by some Linux sound applications. PR: misc/27471 Submitted by: Gavin Atkinson (with cleanups)
2004-06-18 14:36:24 +00:00
case LINUX_SOUND_MIXER_READ_RECMASK:
args->cmd = SOUND_MIXER_READ_RECMASK;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
Add stub for Linux SOUND_MIXER_READ_RECMASK, required by some Linux sound applications. PR: misc/27471 Submitted by: Gavin Atkinson (with cleanups)
2004-06-18 14:36:24 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SOUND_MIXER_READ_DEVMASK:
args->cmd = SOUND_MIXER_READ_DEVMASK;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
MF22: Bring in some linux sound ioctl support which I committed to 2.2 for PR 7792 but did not bring forward. Submitted by: Avatar Liang <avatar@www.mmlab.cse.yzu.edu.tw> PR: 8656
1998-11-12 00:42:08 +00:00
Implement missing SOUND_MIXER_WRITE_RECSRC ioctl. PR: 22971 Tested by: dougb
2001-11-18 06:36:18 +00:00
case LINUX_SOUND_MIXER_WRITE_RECSRC:
args->cmd = SETDIR(SOUND_MIXER_WRITE_RECSRC);
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
Implement missing SOUND_MIXER_WRITE_RECSRC ioctl. PR: 22971 Tested by: dougb
2001-11-18 06:36:18 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SNDCTL_DSP_RESET:
args->cmd = SNDCTL_DSP_RESET;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
MF22: Bring in some linux sound ioctl support which I committed to 2.2 for PR 7792 but did not bring forward. Submitted by: Avatar Liang <avatar@www.mmlab.cse.yzu.edu.tw> PR: 8656
1998-11-12 00:42:08 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SNDCTL_DSP_SYNC:
args->cmd = SNDCTL_DSP_SYNC;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
MF22: Bring in some linux sound ioctl support which I committed to 2.2 for PR 7792 but did not bring forward. Submitted by: Avatar Liang <avatar@www.mmlab.cse.yzu.edu.tw> PR: 8656
1998-11-12 00:42:08 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SNDCTL_DSP_SPEED:
args->cmd = SNDCTL_DSP_SPEED;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
MF22: Bring in some linux sound ioctl support which I committed to 2.2 for PR 7792 but did not bring forward. Submitted by: Avatar Liang <avatar@www.mmlab.cse.yzu.edu.tw> PR: 8656
1998-11-12 00:42:08 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SNDCTL_DSP_STEREO:
args->cmd = SNDCTL_DSP_STEREO;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
MF22: Bring in some linux sound ioctl support which I committed to 2.2 for PR 7792 but did not bring forward. Submitted by: Avatar Liang <avatar@www.mmlab.cse.yzu.edu.tw> PR: 8656
1998-11-12 00:42:08 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SNDCTL_DSP_GETBLKSIZE: /* LINUX_SNDCTL_DSP_SETBLKSIZE */
args->cmd = SNDCTL_DSP_GETBLKSIZE;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
MF22: Bring in some linux sound ioctl support which I committed to 2.2 for PR 7792 but did not bring forward. Submitted by: Avatar Liang <avatar@www.mmlab.cse.yzu.edu.tw> PR: 8656
1998-11-12 00:42:08 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SNDCTL_DSP_SETFMT:
args->cmd = SNDCTL_DSP_SETFMT;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
MF22: Bring in some linux sound ioctl support which I committed to 2.2 for PR 7792 but did not bring forward. Submitted by: Avatar Liang <avatar@www.mmlab.cse.yzu.edu.tw> PR: 8656
1998-11-12 00:42:08 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SOUND_PCM_WRITE_CHANNELS:
args->cmd = SOUND_PCM_WRITE_CHANNELS;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
MF22: Bring in some linux sound ioctl support which I committed to 2.2 for PR 7792 but did not bring forward. Submitted by: Avatar Liang <avatar@www.mmlab.cse.yzu.edu.tw> PR: 8656
1998-11-12 00:42:08 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SOUND_PCM_WRITE_FILTER:
args->cmd = SOUND_PCM_WRITE_FILTER;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
MF22: Bring in some linux sound ioctl support which I committed to 2.2 for PR 7792 but did not bring forward. Submitted by: Avatar Liang <avatar@www.mmlab.cse.yzu.edu.tw> PR: 8656
1998-11-12 00:42:08 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SNDCTL_DSP_POST:
args->cmd = SNDCTL_DSP_POST;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
MF22: Bring in some linux sound ioctl support which I committed to 2.2 for PR 7792 but did not bring forward. Submitted by: Avatar Liang <avatar@www.mmlab.cse.yzu.edu.tw> PR: 8656
1998-11-12 00:42:08 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SNDCTL_DSP_SUBDIVIDE:
args->cmd = SNDCTL_DSP_SUBDIVIDE;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
MF22: Bring in some linux sound ioctl support which I committed to 2.2 for PR 7792 but did not bring forward. Submitted by: Avatar Liang <avatar@www.mmlab.cse.yzu.edu.tw> PR: 8656
1998-11-12 00:42:08 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SNDCTL_DSP_SETFRAGMENT:
args->cmd = SNDCTL_DSP_SETFRAGMENT;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
MF22: Bring in some linux sound ioctl support which I committed to 2.2 for PR 7792 but did not bring forward. Submitted by: Avatar Liang <avatar@www.mmlab.cse.yzu.edu.tw> PR: 8656
1998-11-12 00:42:08 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SNDCTL_DSP_GETFMTS:
args->cmd = SNDCTL_DSP_GETFMTS;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
MF22: Bring in some linux sound ioctl support which I committed to 2.2 for PR 7792 but did not bring forward. Submitted by: Avatar Liang <avatar@www.mmlab.cse.yzu.edu.tw> PR: 8656
1998-11-12 00:42:08 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SNDCTL_DSP_GETOSPACE:
args->cmd = SNDCTL_DSP_GETOSPACE;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
MF22: Bring in some linux sound ioctl support which I committed to 2.2 for PR 7792 but did not bring forward. Submitted by: Avatar Liang <avatar@www.mmlab.cse.yzu.edu.tw> PR: 8656
1998-11-12 00:42:08 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SNDCTL_DSP_GETISPACE:
args->cmd = SNDCTL_DSP_GETISPACE;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
MF22: Bring in some linux sound ioctl support which I committed to 2.2 for PR 7792 but did not bring forward. Submitted by: Avatar Liang <avatar@www.mmlab.cse.yzu.edu.tw> PR: 8656
1998-11-12 00:42:08 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SNDCTL_DSP_NONBLOCK:
args->cmd = SNDCTL_DSP_NONBLOCK;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
Added support for linux sound ioctls: LINUX_SNDCTL_DSP_GETOPTR LINUX_SNDCTL_DSP_GETIPTR LINUX_SNDCTL_DSP_SETTRIGGER LINUX_SNDCTL_DSP_GETCAPS With this rev level the linux realaudio player 5 and xquake should work.
1997-11-17 04:00:32 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SNDCTL_DSP_GETCAPS:
args->cmd = SNDCTL_DSP_GETCAPS;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
Added support for linux sound ioctls: LINUX_SNDCTL_DSP_GETOPTR LINUX_SNDCTL_DSP_GETIPTR LINUX_SNDCTL_DSP_SETTRIGGER LINUX_SNDCTL_DSP_GETCAPS With this rev level the linux realaudio player 5 and xquake should work.
1997-11-17 04:00:32 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SNDCTL_DSP_SETTRIGGER: /* LINUX_SNDCTL_GETTRIGGER */
args->cmd = SNDCTL_DSP_SETTRIGGER;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
Added support for linux sound ioctls: LINUX_SNDCTL_DSP_GETOPTR LINUX_SNDCTL_DSP_GETIPTR LINUX_SNDCTL_DSP_SETTRIGGER LINUX_SNDCTL_DSP_GETCAPS With this rev level the linux realaudio player 5 and xquake should work.
1997-11-17 04:00:32 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SNDCTL_DSP_GETIPTR:
args->cmd = SNDCTL_DSP_GETIPTR;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
Added support for linux sound ioctls: LINUX_SNDCTL_DSP_GETOPTR LINUX_SNDCTL_DSP_GETIPTR LINUX_SNDCTL_DSP_SETTRIGGER LINUX_SNDCTL_DSP_GETCAPS With this rev level the linux realaudio player 5 and xquake should work.
1997-11-17 04:00:32 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SNDCTL_DSP_GETOPTR:
args->cmd = SNDCTL_DSP_GETOPTR;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
My first shot at get sound to work on the emulator. Inspired by the work Amancio Hasty has done, but implemented somewhat differently.
1995-12-29 22:12:14 +00:00
Implement SNDCTL_DSP_SETDUPLEX. This may fix sound apps which want to use full duplex mode. Approved by: matk
2004-07-02 15:31:44 +00:00
case LINUX_SNDCTL_DSP_SETDUPLEX:
args->cmd = SNDCTL_DSP_SETDUPLEX;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
Implement SNDCTL_DSP_SETDUPLEX. This may fix sound apps which want to use full duplex mode. Approved by: matk
2004-07-02 15:31:44 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SNDCTL_DSP_GETODELAY:
args->cmd = SNDCTL_DSP_GETODELAY;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
My first shot at get sound to work on the emulator. Inspired by the work Amancio Hasty has done, but implemented somewhat differently.
1995-12-29 22:12:14 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SNDCTL_SEQ_RESET:
args->cmd = SNDCTL_SEQ_RESET;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
My first shot at get sound to work on the emulator. Inspired by the work Amancio Hasty has done, but implemented somewhat differently.
1995-12-29 22:12:14 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SNDCTL_SEQ_SYNC:
args->cmd = SNDCTL_SEQ_SYNC;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
My first shot at get sound to work on the emulator. Inspired by the work Amancio Hasty has done, but implemented somewhat differently.
1995-12-29 22:12:14 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SNDCTL_SYNTH_INFO:
args->cmd = SNDCTL_SYNTH_INFO;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
My first shot at get sound to work on the emulator. Inspired by the work Amancio Hasty has done, but implemented somewhat differently.
1995-12-29 22:12:14 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SNDCTL_SEQ_CTRLRATE:
args->cmd = SNDCTL_SEQ_CTRLRATE;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
add linuxulator wrapper for SNDCTL_DSP_GETODELAY
1999-11-29 23:03:34 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SNDCTL_SEQ_GETOUTCOUNT:
args->cmd = SNDCTL_SEQ_GETOUTCOUNT;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
My first shot at get sound to work on the emulator. Inspired by the work Amancio Hasty has done, but implemented somewhat differently.
1995-12-29 22:12:14 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SNDCTL_SEQ_GETINCOUNT:
args->cmd = SNDCTL_SEQ_GETINCOUNT;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
My first shot at get sound to work on the emulator. Inspired by the work Amancio Hasty has done, but implemented somewhat differently.
1995-12-29 22:12:14 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SNDCTL_SEQ_PERCMODE:
args->cmd = SNDCTL_SEQ_PERCMODE;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
My first shot at get sound to work on the emulator. Inspired by the work Amancio Hasty has done, but implemented somewhat differently.
1995-12-29 22:12:14 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SNDCTL_FM_LOAD_INSTR:
args->cmd = SNDCTL_FM_LOAD_INSTR;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
My first shot at get sound to work on the emulator. Inspired by the work Amancio Hasty has done, but implemented somewhat differently.
1995-12-29 22:12:14 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SNDCTL_SEQ_TESTMIDI:
args->cmd = SNDCTL_SEQ_TESTMIDI;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
My first shot at get sound to work on the emulator. Inspired by the work Amancio Hasty has done, but implemented somewhat differently.
1995-12-29 22:12:14 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SNDCTL_SEQ_RESETSAMPLES:
args->cmd = SNDCTL_SEQ_RESETSAMPLES;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
Add audio mixer ioctls. Only writing to the mixer is implemented.
1996-11-05 02:04:37 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SNDCTL_SEQ_NRSYNTHS:
args->cmd = SNDCTL_SEQ_NRSYNTHS;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
Add audio mixer ioctls. Only writing to the mixer is implemented.
1996-11-05 02:04:37 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SNDCTL_SEQ_NRMIDIS:
args->cmd = SNDCTL_SEQ_NRMIDIS;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
Add audio mixer ioctls. Only writing to the mixer is implemented.
1996-11-05 02:04:37 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SNDCTL_MIDI_INFO:
args->cmd = SNDCTL_MIDI_INFO;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
Add audio mixer ioctls. Only writing to the mixer is implemented.
1996-11-05 02:04:37 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SNDCTL_SEQ_TRESHOLD:
args->cmd = SNDCTL_SEQ_TRESHOLD;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
Add audio mixer ioctls. Only writing to the mixer is implemented.
1996-11-05 02:04:37 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SNDCTL_SYNTH_MEMAVL:
args->cmd = SNDCTL_SYNTH_MEMAVL;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
Fix LINUX_TIOC{S|G}SERIAL implementation. Both do not copy data in or out of kernel space. Remove the ioctl supporting functions, and move the actual code to the switch-statement. Now everybody can clearly see that the implementation is really poor. Also fix a typo in LINUX_TIOCGETD. The underlying function was given command TIOCSETD instead op TIOCGETD...
1999-08-14 13:26:44 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
}
Fix LINUX_TIOC{S|G}SERIAL implementation. Both do not copy data in or out of kernel space. Remove the ioctl supporting functions, and move the actual code to the switch-statement. Now everybody can clearly see that the implementation is really poor. Also fix a typo in LINUX_TIOCGETD. The underlying function was given command TIOCSETD instead op TIOCGETD...
1999-08-14 13:26:44 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
return (ENOIOCTL);
}
Implement rudumentry support for the linux TIOC[SG]ETSERIAL ioctl's. To complete this, some extra state has to be kept somewhere so that the B38400 flag in Linux can be correctly translated to/from either 38400, 57600 or 115200. Submitted by: Robert Sanders <rsanders@mindspring.com>
1996-03-10 22:30:53 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
/*
* Console related ioctls
*/
static int
KSE Milestone 2 Note ALL MODULES MUST BE RECOMPILED make the kernel aware that there are smaller units of scheduling than the process. (but only allow one thread per process at this time). This is functionally equivalent to teh previousl -current except that there is a thread associated with each process. Sorry john! (your next MFC will be a doosie!) Reviewed by: peter@freebsd.org, dillon@freebsd.org X-MFC after: ha ha ha ha
2001-09-12 08:38:13 +00:00
linux_ioctl_console(struct thread *td, struct linux_ioctl_args *args)
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
{
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
struct file *fp;
int error;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
Eliminate the overhead of gratuitous repeated reinitialization of cap_rights - Add macros to allow preinitialization of cap_rights_t. - Convert most commonly used code paths to use preinitialized cap_rights_t. A 3.6% speedup in fstat was measured with this change. Reported by: mjg Reviewed by: oshogbo Approved by: sbruno MFC after: 1 month
2018-05-09 18:47:24 +00:00
error = fget(td, args->fd, &cap_ioctl_rights, &fp);
Change the cap_rights_t type from uint64_t to a structure that we can extend in the future in a backward compatible (API and ABI) way. The cap_rights_t represents capability rights. We used to use one bit to represent one right, but we are running out of spare bits. Currently the new structure provides place for 114 rights (so 50 more than the previous cap_rights_t), but it is possible to grow the structure to hold at least 285 rights, although we can make it even larger if 285 rights won't be enough. The structure definition looks like this: struct cap_rights { uint64_t cr_rights[CAP_RIGHTS_VERSION + 2]; }; The initial CAP_RIGHTS_VERSION is 0. The top two bits in the first element of the cr_rights[] array contain total number of elements in the array - 2. This means if those two bits are equal to 0, we have 2 array elements. The top two bits in all remaining array elements should be 0. The next five bits in all array elements contain array index. Only one bit is used and bit position in this five-bits range defines array index. This means there can be at most five array elements in the future. To define new right the CAPRIGHT() macro must be used. The macro takes two arguments - an array index and a bit to set, eg. #define CAP_PDKILL CAPRIGHT(1, 0x0000000000000800ULL) We still support aliases that combine few rights, but the rights have to belong to the same array element, eg: #define CAP_LOOKUP CAPRIGHT(0, 0x0000000000000400ULL) #define CAP_FCHMOD CAPRIGHT(0, 0x0000000000002000ULL) #define CAP_FCHMODAT (CAP_FCHMOD | CAP_LOOKUP) There is new API to manage the new cap_rights_t structure: cap_rights_t *cap_rights_init(cap_rights_t *rights, ...); void cap_rights_set(cap_rights_t *rights, ...); void cap_rights_clear(cap_rights_t *rights, ...); bool cap_rights_is_set(const cap_rights_t *rights, ...); bool cap_rights_is_valid(const cap_rights_t *rights); void cap_rights_merge(cap_rights_t *dst, const cap_rights_t *src); void cap_rights_remove(cap_rights_t *dst, const cap_rights_t *src); bool cap_rights_contains(const cap_rights_t *big, const cap_rights_t *little); Capability rights to the cap_rights_init(), cap_rights_set(), cap_rights_clear() and cap_rights_is_set() functions are provided by separating them with commas, eg: cap_rights_t rights; cap_rights_init(&rights, CAP_READ, CAP_WRITE, CAP_FSTAT); There is no need to terminate the list of rights, as those functions are actually macros that take care of the termination, eg: #define cap_rights_set(rights, ...) \ __cap_rights_set((rights), __VA_ARGS__, 0ULL) void __cap_rights_set(cap_rights_t *rights, ...); Thanks to using one bit as an array index we can assert in those functions that there are no two rights belonging to different array elements provided together. For example this is illegal and will be detected, because CAP_LOOKUP belongs to element 0 and CAP_PDKILL to element 1: cap_rights_init(&rights, CAP_LOOKUP | CAP_PDKILL); Providing several rights that belongs to the same array's element this way is correct, but is not advised. It should only be used for aliases definition. This commit also breaks compatibility with some existing Capsicum system calls, but I see no other way to do that. This should be fine as Capsicum is still experimental and this change is not going to 9.x. Sponsored by: The FreeBSD Foundation
2013-09-05 00:09:56 +00:00
if (error != 0)
Replace ffind_* with fget calls. Make fget MPsafe. Make fgetvp and fgetsock use the fget subsystem to reduce code bloat. Push giant down in fpathconf().
2002-01-14 00:13:45 +00:00
return (error);
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
switch (args->cmd & 0xffff) {
case LINUX_KIOCSOUND:
args->cmd = KIOCSOUND;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = (sys_ioctl(td, (struct ioctl_args *)args));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_KDMKTONE:
args->cmd = KDMKTONE;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = (sys_ioctl(td, (struct ioctl_args *)args));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_KDGETLED:
args->cmd = KDGETLED;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = (sys_ioctl(td, (struct ioctl_args *)args));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_KDSETLED:
args->cmd = KDSETLED;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = (sys_ioctl(td, (struct ioctl_args *)args));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_KDSETMODE:
args->cmd = KDSETMODE;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = (sys_ioctl(td, (struct ioctl_args *)args));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_KDGETMODE:
args->cmd = KDGETMODE;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = (sys_ioctl(td, (struct ioctl_args *)args));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_KDGKBMODE:
args->cmd = KDGKBMODE;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = (sys_ioctl(td, (struct ioctl_args *)args));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_KDSKBMODE: {
int kbdmode;
switch (args->arg) {
case LINUX_KBD_RAW:
kbdmode = K_RAW;
break;
case LINUX_KBD_XLATE:
kbdmode = K_XLATE;
break;
case LINUX_KBD_MEDIUMRAW:
kbdmode = K_RAW;
break;
default:
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
fdrop(fp, td);
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
return (EINVAL);
}
In continuation of early fileop credential changes, modify fo_ioctl() to accept an 'active_cred' argument reflecting the credential of the thread initiating the ioctl operation. - Change fo_ioctl() to accept active_cred; change consumers of the fo_ioctl() interface to generally pass active_cred from td->td_ucred. - In fifofs, initialize filetmp.f_cred to ap->a_cred so that the invocations of soo_ioctl() are provided access to the calling f_cred. Pass ap->a_td->td_ucred as the active_cred, but note that this is required because we don't yet distinguish file_cred and active_cred in invoking VOP's. - Update kqueue_ioctl() for its new argument. - Update pipe_ioctl() for its new argument, pass active_cred rather than td_ucred to MAC for authorization. - Update soo_ioctl() for its new argument. - Update vn_ioctl() for its new argument, use active_cred rather than td->td_ucred to authorize VOP_IOCTL() and the associated VOP_GETATTR(). Obtained from: TrustedBSD Project Sponsored by: DARPA, NAI Labs
2002-08-17 02:36:16 +00:00
error = (fo_ioctl(fp, KDSKBMODE, (caddr_t)&kbdmode,
td->td_ucred, td));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
As described by the submitter: These patches enables us to play quake2 . Support linux keyboard ioctl for setting RAW, MEDIUMRAW and XLATE. Support linux virtual terminal operations: OPENQRY, GETMODE, SETMODE, GETSTATE, ACTIVATE, and WAITACTIVE. Submitted by: Amancio Hasty <hasty@rah.star-gate.com>
1997-12-15 06:09:11 +00:00
}
Initial support for using linux X servers under emulation - to use an XFree86 server, users need to create the following links in their /compat/linux/dev directory (assuming kernel configured with 4 VTs). lrwxrwxrwx 1 root wheel 7 Aug 30 22:59 tty0 -> console lrwxrwxrwx 1 root wheel 5 Aug 30 22:45 tty1 -> ttyv0 lrwxrwxrwx 1 root wheel 5 Aug 30 22:45 tty2 -> ttyv1 lrwxrwxrwx 1 root wheel 5 Aug 30 22:45 tty3 -> ttyv2 lrwxrwxrwx 1 root wheel 5 Aug 30 22:45 tty4 -> ttyv3 VT switching is still not yet supported. Attempting to switch VT currently will cause Xserver bus error. Submitted by: Chain Lee <chain@110.net>
1998-08-31 06:55:02 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_VT_OPENQRY:
args->cmd = VT_OPENQRY;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = (sys_ioctl(td, (struct ioctl_args *)args));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_VT_GETMODE:
args->cmd = VT_GETMODE;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = (sys_ioctl(td, (struct ioctl_args *)args));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_VT_SETMODE: {
Fix the VT_SETMODE/CDROMIOCTOCENTRY problem correctly. Reviewed by: tjr
2004-06-18 13:36:30 +00:00
struct vt_mode mode;
if ((error = copyin((void *)args->arg, &mode, sizeof(mode))))
break;
Convert signal number to native for VT_SETMODE ioctl and remove strange and invalid ISSIGVALID macro. The code has not been tested right way but it was originally broken.
2015-05-24 17:59:17 +00:00
if (LINUX_SIG_VALID(mode.relsig))
mode.relsig = linux_to_bsd_signal(mode.relsig);
else
mode.relsig = 0;
if (LINUX_SIG_VALID(mode.acqsig))
mode.acqsig = linux_to_bsd_signal(mode.acqsig);
else
mode.acqsig = 0;
/* XXX. Linux ignores frsig and set it to 0. */
mode.frsig = 0;
Fix the VT_SETMODE/CDROMIOCTOCENTRY problem correctly. Reviewed by: tjr
2004-06-18 13:36:30 +00:00
if ((error = copyout(&mode, (void *)args->arg, sizeof(mode))))
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
args->cmd = VT_SETMODE;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = (sys_ioctl(td, (struct ioctl_args *)args));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
}
Initial support for using linux X servers under emulation - to use an XFree86 server, users need to create the following links in their /compat/linux/dev directory (assuming kernel configured with 4 VTs). lrwxrwxrwx 1 root wheel 7 Aug 30 22:59 tty0 -> console lrwxrwxrwx 1 root wheel 5 Aug 30 22:45 tty1 -> ttyv0 lrwxrwxrwx 1 root wheel 5 Aug 30 22:45 tty2 -> ttyv1 lrwxrwxrwx 1 root wheel 5 Aug 30 22:45 tty3 -> ttyv2 lrwxrwxrwx 1 root wheel 5 Aug 30 22:45 tty4 -> ttyv3 VT switching is still not yet supported. Attempting to switch VT currently will cause Xserver bus error. Submitted by: Chain Lee <chain@110.net>
1998-08-31 06:55:02 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_VT_GETSTATE:
args->cmd = VT_GETACTIVE;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = (sys_ioctl(td, (struct ioctl_args *)args));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Initial support for using linux X servers under emulation - to use an XFree86 server, users need to create the following links in their /compat/linux/dev directory (assuming kernel configured with 4 VTs). lrwxrwxrwx 1 root wheel 7 Aug 30 22:59 tty0 -> console lrwxrwxrwx 1 root wheel 5 Aug 30 22:45 tty1 -> ttyv0 lrwxrwxrwx 1 root wheel 5 Aug 30 22:45 tty2 -> ttyv1 lrwxrwxrwx 1 root wheel 5 Aug 30 22:45 tty3 -> ttyv2 lrwxrwxrwx 1 root wheel 5 Aug 30 22:45 tty4 -> ttyv3 VT switching is still not yet supported. Attempting to switch VT currently will cause Xserver bus error. Submitted by: Chain Lee <chain@110.net>
1998-08-31 06:55:02 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_VT_RELDISP:
args->cmd = VT_RELDISP;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = (sys_ioctl(td, (struct ioctl_args *)args));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Initial support for using linux X servers under emulation - to use an XFree86 server, users need to create the following links in their /compat/linux/dev directory (assuming kernel configured with 4 VTs). lrwxrwxrwx 1 root wheel 7 Aug 30 22:59 tty0 -> console lrwxrwxrwx 1 root wheel 5 Aug 30 22:45 tty1 -> ttyv0 lrwxrwxrwx 1 root wheel 5 Aug 30 22:45 tty2 -> ttyv1 lrwxrwxrwx 1 root wheel 5 Aug 30 22:45 tty3 -> ttyv2 lrwxrwxrwx 1 root wheel 5 Aug 30 22:45 tty4 -> ttyv3 VT switching is still not yet supported. Attempting to switch VT currently will cause Xserver bus error. Submitted by: Chain Lee <chain@110.net>
1998-08-31 06:55:02 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_VT_ACTIVATE:
args->cmd = VT_ACTIVATE;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = (sys_ioctl(td, (struct ioctl_args *)args));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Initial support for using linux X servers under emulation - to use an XFree86 server, users need to create the following links in their /compat/linux/dev directory (assuming kernel configured with 4 VTs). lrwxrwxrwx 1 root wheel 7 Aug 30 22:59 tty0 -> console lrwxrwxrwx 1 root wheel 5 Aug 30 22:45 tty1 -> ttyv0 lrwxrwxrwx 1 root wheel 5 Aug 30 22:45 tty2 -> ttyv1 lrwxrwxrwx 1 root wheel 5 Aug 30 22:45 tty3 -> ttyv2 lrwxrwxrwx 1 root wheel 5 Aug 30 22:45 tty4 -> ttyv3 VT switching is still not yet supported. Attempting to switch VT currently will cause Xserver bus error. Submitted by: Chain Lee <chain@110.net>
1998-08-31 06:55:02 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_VT_WAITACTIVE:
args->cmd = VT_WAITACTIVE;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = (sys_ioctl(td, (struct ioctl_args *)args));
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
break;
Initial support for using linux X servers under emulation - to use an XFree86 server, users need to create the following links in their /compat/linux/dev directory (assuming kernel configured with 4 VTs). lrwxrwxrwx 1 root wheel 7 Aug 30 22:59 tty0 -> console lrwxrwxrwx 1 root wheel 5 Aug 30 22:45 tty1 -> ttyv0 lrwxrwxrwx 1 root wheel 5 Aug 30 22:45 tty2 -> ttyv1 lrwxrwxrwx 1 root wheel 5 Aug 30 22:45 tty3 -> ttyv2 lrwxrwxrwx 1 root wheel 5 Aug 30 22:45 tty4 -> ttyv3 VT switching is still not yet supported. Attempting to switch VT currently will cause Xserver bus error. Submitted by: Chain Lee <chain@110.net>
1998-08-31 06:55:02 +00:00
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
default:
error = ENOIOCTL;
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
}
Clean up whitespace, s/register //, refrain from strong urge to ANSIfy.
2003-03-02 15:56:49 +00:00
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
fdrop(fp, td);
return (error);
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
}
- Handle mixer read ioctls correctly. They have the same group, number and argument size as their write counterparts and were handled as write ioctls. - Emulate some cdrom ioctls.
1999-04-29 04:37:57 +00:00
Adding SIOCGIFNAME support in Linuxulator. This should silence the console warning associated with linux-opera: linux: pid 23492 (opera): ioctl fd=5, cmd=0x8910 ('\M^I',16) is not implemented linux: pid 23492 (opera): ioctl fd=28, cmd=0x8910 ('\M^I',16) is not implemented ... Reviewed by: kib, marcel, dchagin Tested with: linux-opera-12.16_3 MFC after: 1 month
2017-04-09 15:27:04 +00:00
/*
* Implement the SIOCGIFNAME ioctl
*/
static int
linux_ioctl_ifname(struct thread *td, struct l_ifreq *uifr)
{
struct l_ifreq ifr;
struct ifnet *ifp;
int error, ethno, index;
error = copyin(uifr, &ifr, sizeof(ifr));
if (error != 0)
return (error);
CURVNET_SET(TD_TO_VNET(curthread));
IFNET_RLOCK();
index = 1; /* ifr.ifr_ifindex starts from 1 */
ethno = 0;
error = ENODEV;
UDP: further performance improvements on tx Cumulative throughput while running 64 netperf -H $DUT -t UDP_STREAM -- -m 1 on a 2x8x2 SKL went from 1.1Mpps to 2.5Mpps Single stream throughput increases from 910kpps to 1.18Mpps Baseline: https://people.freebsd.org/~mmacy/2018.05.11/udpsender2.svg - Protect read access to global ifnet list with epoch https://people.freebsd.org/~mmacy/2018.05.11/udpsender3.svg - Protect short lived ifaddr references with epoch https://people.freebsd.org/~mmacy/2018.05.11/udpsender4.svg - Convert if_afdata read lock path to epoch https://people.freebsd.org/~mmacy/2018.05.11/udpsender5.svg A fix for the inpcbhash contention is pending sufficient time on a canary at LLNW. Reviewed by: gallatin Sponsored by: Limelight Networks Differential Revision: https://reviews.freebsd.org/D15409
2018-05-23 21:02:14 +00:00
CK_STAILQ_FOREACH(ifp, &V_ifnet, if_link) {
Adding SIOCGIFNAME support in Linuxulator. This should silence the console warning associated with linux-opera: linux: pid 23492 (opera): ioctl fd=5, cmd=0x8910 ('\M^I',16) is not implemented linux: pid 23492 (opera): ioctl fd=28, cmd=0x8910 ('\M^I',16) is not implemented ... Reviewed by: kib, marcel, dchagin Tested with: linux-opera-12.16_3 MFC after: 1 month
2017-04-09 15:27:04 +00:00
if (ifr.ifr_ifindex == index) {
if (IFP_IS_ETH(ifp))
snprintf(ifr.ifr_name, LINUX_IFNAMSIZ,
"eth%d", ethno);
else
strlcpy(ifr.ifr_name, ifp->if_xname,
LINUX_IFNAMSIZ);
error = 0;
break;
}
if (IFP_IS_ETH(ifp))
ethno++;
index++;
}
IFNET_RUNLOCK();
if (error == 0)
error = copyout(&ifr, uifr, sizeof(ifr));
CURVNET_RESTORE();
return (error);
}
Try to make Linux socket ioctls work. Up until now they've only *pretended* to work, but haven't really due to subtle differences in structs etc. This is still not perfect (some ioctls are still known not to work, while others haven't been tested at all), but it's enough to get Debian's ifconfig to produce relatively sane output. More work will be needed to get all ioctls (or at least a reasonable subset) working, and to support the Cisco Aironet config tool mentioned in the PR. PR: 26546 Submitted by: Doug Ambrisko <ambrisko@ambrisko.com>
2001-10-15 20:52:17 +00:00
/*
* Implement the SIOCGIFCONF ioctl
*/
static int
linux_ifconf(struct thread *td, struct ifconf *uifc)
{
Match the LINUX32's style with existing style Submitted by: Jung-uk Kim <jkim@niksun.com> Use positive, not negative logic.
2005-01-14 04:44:56 +00:00
#ifdef COMPAT_LINUX32
Changes to MI Linux emulation code necessary to run 32-bit Linux binaries on AMD64, and the general case where the emulated platform has different size pointers than we use natively: - declare certain structure members as l_uintptr_t and use the new PTRIN and PTROUT macros to convert to and from native pointers. - declare some structures __packed on amd64 when the layout would differ from that used on i386. - include <machine/../linux32/linux.h> instead of <machine/../linux/linux.h> if compiling with COMPAT_LINUX32. This will need to be revisited before 32-bit and 64-bit Linux emulation support can coexist in the same kernel. - other small scattered changes. This should be a no-op on i386 and Alpha.
2004-08-16 07:28:16 +00:00
struct l_ifconf ifc;
#else
Try to make Linux socket ioctls work. Up until now they've only *pretended* to work, but haven't really due to subtle differences in structs etc. This is still not perfect (some ioctls are still known not to work, while others haven't been tested at all), but it's enough to get Debian's ifconfig to produce relatively sane output. More work will be needed to get all ioctls (or at least a reasonable subset) working, and to support the Cisco Aironet config tool mentioned in the PR. PR: 26546 Submitted by: Doug Ambrisko <ambrisko@ambrisko.com>
2001-10-15 20:52:17 +00:00
struct ifconf ifc;
Changes to MI Linux emulation code necessary to run 32-bit Linux binaries on AMD64, and the general case where the emulated platform has different size pointers than we use natively: - declare certain structure members as l_uintptr_t and use the new PTRIN and PTROUT macros to convert to and from native pointers. - declare some structures __packed on amd64 when the layout would differ from that used on i386. - include <machine/../linux32/linux.h> instead of <machine/../linux/linux.h> if compiling with COMPAT_LINUX32. This will need to be revisited before 32-bit and 64-bit Linux emulation support can coexist in the same kernel. - other small scattered changes. This should be a no-op on i386 and Alpha.
2004-08-16 07:28:16 +00:00
#endif
Try to make Linux socket ioctls work. Up until now they've only *pretended* to work, but haven't really due to subtle differences in structs etc. This is still not perfect (some ioctls are still known not to work, while others haven't been tested at all), but it's enough to get Debian's ifconfig to produce relatively sane output. More work will be needed to get all ioctls (or at least a reasonable subset) working, and to support the Cisco Aironet config tool mentioned in the PR. PR: 26546 Submitted by: Doug Ambrisko <ambrisko@ambrisko.com>
2001-10-15 20:52:17 +00:00
struct l_ifreq ifr;
struct ifnet *ifp;
Have SIOCGIFCONF return all (if any) AF_INET addresses for the interfaces we encounter. In Linux, all addresses are returned for which gifconf handlers are installed. This boils down to AF_DECnet and AF_INET. We care mostly about AF_INET for now. Adding additional families is simple enough. Returning the addresses is important for RPC clients to function properly. Andrew found in some reference code that the logic that handles the retransmission looks for an interface that's up and has an AF_INET address. This obviously failed as we didn't return any addresses at all. Note also that with this change we don't return interfaces that don't have AF_INET addresses, whereas before we returned any interface present in the system. This is in line with what Linux does (modulo interfaces with only AF_DECnet addresses of course :-) Reported by: "Andrew Atrens" <atrens@nortelnetworks.com> MFC after: 1 week
2002-01-29 06:00:11 +00:00
struct ifaddr *ifa;
Rewrite linux_ifconf() to be more like ifconf() in net/if.c so that we do not call uiomove() while IFNET_RLOCK() is held. This eliminates the witness warning: Calling uiomove() with the following non-sleepable locks held: exclusive sleep mutex ifnet r = 0 (0xc096dd60) locked @ /usr/src/sys/modules/linux/../../compat/linux/linux_ioctl.c:2170 MFC after: 2 days
2005-08-27 14:44:10 +00:00
struct sbuf *sb;
int error, ethno, full = 0, valid_len, max_len;
Further fixes related to the interface renaming. Now that we properly translate the interface name passed to us, make sure we also translate correctly before we return the list of interfaces with the SIOCGIFCONF ioctl. It is common to use the interface names returned by that ioctl in further ioctls, such as SIOCGIFFLAGS. Remove linux_ifname as it is no longer used. Also remove ifname_bsd_to_linux as it cannot be used anymore now that linux_ifname is removed (was deadcode anyway). Reported and tested by: Andrew Atrens <atrens@nortelnetworks.com>
2002-01-10 05:36:36 +00:00
More caddr_t removal, in conjunction with copy{in,out}(9) this time. Also clean up some egregious casts and incorrect use of sizeof.
2003-03-03 09:14:26 +00:00
error = copyin(uifc, &ifc, sizeof(ifc));
Try to make Linux socket ioctls work. Up until now they've only *pretended* to work, but haven't really due to subtle differences in structs etc. This is still not perfect (some ioctls are still known not to work, while others haven't been tested at all), but it's enough to get Debian's ifconfig to produce relatively sane output. More work will be needed to get all ioctls (or at least a reasonable subset) working, and to support the Cisco Aironet config tool mentioned in the PR. PR: 26546 Submitted by: Doug Ambrisko <ambrisko@ambrisko.com>
2001-10-15 20:52:17 +00:00
if (error != 0)
return (error);
Rewrite linux_ifconf() to be more like ifconf() in net/if.c so that we do not call uiomove() while IFNET_RLOCK() is held. This eliminates the witness warning: Calling uiomove() with the following non-sleepable locks held: exclusive sleep mutex ifnet r = 0 (0xc096dd60) locked @ /usr/src/sys/modules/linux/../../compat/linux/linux_ioctl.c:2170 MFC after: 2 days
2005-08-27 14:44:10 +00:00
max_len = MAXPHYS - 1;
Fix a few panics in linuxulator + VIMAGE due to curvnet not being set. This change affects only options VIMAGE builds. Reviewed by: julian MFC after: 3 days
2009-08-28 22:51:07 +00:00
CURVNET_SET(TD_TO_VNET(td));
Linux applications expect to be able to call SIOCGIFCONF with an NULL ifc.ifc_buf pointer, to determine the expected buffer size. The submitted fix only takes account of interfaces with an AF_INET address configured. This could no doubt be improved. PR: kern/45753 Submitted by: Jacques Garrigue (with cleanups)
2004-06-18 14:06:46 +00:00
/* handle the 'request buffer size' case */
In sys/compat/linux/linux_ioctl.c, work around a warning when a pointer is compared to an integer, by casting the pointer to l_uintptr_t. No functional difference on both i386 and amd64. Reviewed by: ed, jhb MFC after: 1 week
2012-01-03 18:49:39 +00:00
if ((l_uintptr_t)ifc.ifc_buf == PTROUT(NULL)) {
Linux applications expect to be able to call SIOCGIFCONF with an NULL ifc.ifc_buf pointer, to determine the expected buffer size. The submitted fix only takes account of interfaces with an AF_INET address configured. This could no doubt be improved. PR: kern/45753 Submitted by: Jacques Garrigue (with cleanups)
2004-06-18 14:06:46 +00:00
ifc.ifc_len = 0;
Lock the ifnet list while iterating over it. Submitted by: julian MFC after: 3 days
2009-09-13 21:30:18 +00:00
IFNET_RLOCK();
UDP: further performance improvements on tx Cumulative throughput while running 64 netperf -H $DUT -t UDP_STREAM -- -m 1 on a 2x8x2 SKL went from 1.1Mpps to 2.5Mpps Single stream throughput increases from 910kpps to 1.18Mpps Baseline: https://people.freebsd.org/~mmacy/2018.05.11/udpsender2.svg - Protect read access to global ifnet list with epoch https://people.freebsd.org/~mmacy/2018.05.11/udpsender3.svg - Protect short lived ifaddr references with epoch https://people.freebsd.org/~mmacy/2018.05.11/udpsender4.svg - Convert if_afdata read lock path to epoch https://people.freebsd.org/~mmacy/2018.05.11/udpsender5.svg A fix for the inpcbhash contention is pending sufficient time on a canary at LLNW. Reviewed by: gallatin Sponsored by: Limelight Networks Differential Revision: https://reviews.freebsd.org/D15409
2018-05-23 21:02:14 +00:00
CK_STAILQ_FOREACH(ifp, &V_ifnet, if_link) {
ifnet: Replace if_addr_lock rwlock with epoch + mutex Run on LLNW canaries and tested by pho@ gallatin: Using a 14-core, 28-HTT single socket E5-2697 v3 with a 40GbE MLX5 based ConnectX 4-LX NIC, I see an almost 12% improvement in received packet rate, and a larger improvement in bytes delivered all the way to userspace. When the host receiving 64 streams of netperf -H $DUT -t UDP_STREAM -- -m 1, I see, using nstat -I mce0 1 before the patch: InMpps OMpps InGbs OGbs err TCP Est %CPU syscalls csw irq GBfree 4.98 0.00 4.42 0.00 4235592 33 83.80 4720653 2149771 1235 247.32 4.73 0.00 4.20 0.00 4025260 33 82.99 4724900 2139833 1204 247.32 4.72 0.00 4.20 0.00 4035252 33 82.14 4719162 2132023 1264 247.32 4.71 0.00 4.21 0.00 4073206 33 83.68 4744973 2123317 1347 247.32 4.72 0.00 4.21 0.00 4061118 33 80.82 4713615 2188091 1490 247.32 4.72 0.00 4.21 0.00 4051675 33 85.29 4727399 2109011 1205 247.32 4.73 0.00 4.21 0.00 4039056 33 84.65 4724735 2102603 1053 247.32 After the patch InMpps OMpps InGbs OGbs err TCP Est %CPU syscalls csw irq GBfree 5.43 0.00 4.20 0.00 3313143 33 84.96 5434214 1900162 2656 245.51 5.43 0.00 4.20 0.00 3308527 33 85.24 5439695 1809382 2521 245.51 5.42 0.00 4.19 0.00 3316778 33 87.54 5416028 1805835 2256 245.51 5.42 0.00 4.19 0.00 3317673 33 90.44 5426044 1763056 2332 245.51 5.42 0.00 4.19 0.00 3314839 33 88.11 5435732 1792218 2499 245.52 5.44 0.00 4.19 0.00 3293228 33 91.84 5426301 1668597 2121 245.52 Similarly, netperf reports 230Mb/s before the patch, and 270Mb/s after the patch Reviewed by: gallatin Sponsored by: Limelight Networks Differential Revision: https://reviews.freebsd.org/D15366
2018-05-18 20:13:34 +00:00
CK_STAILQ_FOREACH(ifa, &ifp->if_addrhead, ifa_link) {
Linux applications expect to be able to call SIOCGIFCONF with an NULL ifc.ifc_buf pointer, to determine the expected buffer size. The submitted fix only takes account of interfaces with an AF_INET address configured. This could no doubt be improved. PR: kern/45753 Submitted by: Jacques Garrigue (with cleanups)
2004-06-18 14:06:46 +00:00
struct sockaddr *sa = ifa->ifa_addr;
if (sa->sa_family == AF_INET)
ifc.ifc_len += sizeof(ifr);
}
}
Lock the ifnet list while iterating over it. Submitted by: julian MFC after: 3 days
2009-09-13 21:30:18 +00:00
IFNET_RUNLOCK();
Linux applications expect to be able to call SIOCGIFCONF with an NULL ifc.ifc_buf pointer, to determine the expected buffer size. The submitted fix only takes account of interfaces with an AF_INET address configured. This could no doubt be improved. PR: kern/45753 Submitted by: Jacques Garrigue (with cleanups)
2004-06-18 14:06:46 +00:00
error = copyout(&ifc, uifc, sizeof(ifc));
Fix a few panics in linuxulator + VIMAGE due to curvnet not being set. This change affects only options VIMAGE builds. Reviewed by: julian MFC after: 3 days
2009-08-28 22:51:07 +00:00
CURVNET_RESTORE();
Linux applications expect to be able to call SIOCGIFCONF with an NULL ifc.ifc_buf pointer, to determine the expected buffer size. The submitted fix only takes account of interfaces with an AF_INET address configured. This could no doubt be improved. PR: kern/45753 Submitted by: Jacques Garrigue (with cleanups)
2004-06-18 14:06:46 +00:00
return (error);
}
Fix a few panics in linuxulator + VIMAGE due to curvnet not being set. This change affects only options VIMAGE builds. Reviewed by: julian MFC after: 3 days
2009-08-28 22:51:07 +00:00
if (ifc.ifc_len <= 0) {
CURVNET_RESTORE();
Rewrite linux_ifconf() to be more like ifconf() in net/if.c so that we do not call uiomove() while IFNET_RLOCK() is held. This eliminates the witness warning: Calling uiomove() with the following non-sleepable locks held: exclusive sleep mutex ifnet r = 0 (0xc096dd60) locked @ /usr/src/sys/modules/linux/../../compat/linux/linux_ioctl.c:2170 MFC after: 2 days
2005-08-27 14:44:10 +00:00
return (EINVAL);
Fix a few panics in linuxulator + VIMAGE due to curvnet not being set. This change affects only options VIMAGE builds. Reviewed by: julian MFC after: 3 days
2009-08-28 22:51:07 +00:00
}
Try to make Linux socket ioctls work. Up until now they've only *pretended* to work, but haven't really due to subtle differences in structs etc. This is still not perfect (some ioctls are still known not to work, while others haven't been tested at all), but it's enough to get Debian's ifconfig to produce relatively sane output. More work will be needed to get all ioctls (or at least a reasonable subset) working, and to support the Cisco Aironet config tool mentioned in the PR. PR: 26546 Submitted by: Doug Ambrisko <ambrisko@ambrisko.com>
2001-10-15 20:52:17 +00:00
Rewrite linux_ifconf() to be more like ifconf() in net/if.c so that we do not call uiomove() while IFNET_RLOCK() is held. This eliminates the witness warning: Calling uiomove() with the following non-sleepable locks held: exclusive sleep mutex ifnet r = 0 (0xc096dd60) locked @ /usr/src/sys/modules/linux/../../compat/linux/linux_ioctl.c:2170 MFC after: 2 days
2005-08-27 14:44:10 +00:00
again:
Further fixes related to the interface renaming. Now that we properly translate the interface name passed to us, make sure we also translate correctly before we return the list of interfaces with the SIOCGIFCONF ioctl. It is common to use the interface names returned by that ioctl in further ioctls, such as SIOCGIFFLAGS. Remove linux_ifname as it is no longer used. Also remove ifname_bsd_to_linux as it cannot be used anymore now that linux_ifname is removed (was deadcode anyway). Reported and tested by: Andrew Atrens <atrens@nortelnetworks.com>
2002-01-10 05:36:36 +00:00
/* Keep track of eth interfaces */
ethno = 0;
Rewrite linux_ifconf() to be more like ifconf() in net/if.c so that we do not call uiomove() while IFNET_RLOCK() is held. This eliminates the witness warning: Calling uiomove() with the following non-sleepable locks held: exclusive sleep mutex ifnet r = 0 (0xc096dd60) locked @ /usr/src/sys/modules/linux/../../compat/linux/linux_ioctl.c:2170 MFC after: 2 days
2005-08-27 14:44:10 +00:00
if (ifc.ifc_len <= max_len) {
max_len = ifc.ifc_len;
full = 1;
}
sb = sbuf_new(NULL, NULL, max_len + 1, SBUF_FIXEDLEN);
max_len = 0;
valid_len = 0;
Further fixes related to the interface renaming. Now that we properly translate the interface name passed to us, make sure we also translate correctly before we return the list of interfaces with the SIOCGIFCONF ioctl. It is common to use the interface names returned by that ioctl in further ioctls, such as SIOCGIFFLAGS. Remove linux_ifname as it is no longer used. Also remove ifname_bsd_to_linux as it cannot be used anymore now that linux_ifname is removed (was deadcode anyway). Reported and tested by: Andrew Atrens <atrens@nortelnetworks.com>
2002-01-10 05:36:36 +00:00
Have SIOCGIFCONF return all (if any) AF_INET addresses for the interfaces we encounter. In Linux, all addresses are returned for which gifconf handlers are installed. This boils down to AF_DECnet and AF_INET. We care mostly about AF_INET for now. Adding additional families is simple enough. Returning the addresses is important for RPC clients to function properly. Andrew found in some reference code that the logic that handles the retransmission looks for an interface that's up and has an AF_INET address. This obviously failed as we didn't return any addresses at all. Note also that with this change we don't return interfaces that don't have AF_INET addresses, whereas before we returned any interface present in the system. This is in line with what Linux does (modulo interfaces with only AF_DECnet addresses of course :-) Reported by: "Andrew Atrens" <atrens@nortelnetworks.com> MFC after: 1 week
2002-01-29 06:00:11 +00:00
/* Return all AF_INET addresses of all interfaces */
Rework global locks for interface list and index management, correcting several critical bugs, including race conditions and lock order issues: Replace the single rwlock, ifnet_lock, with two locks, an rwlock and an sxlock. Either can be held to stablize the lists and indexes, but both are required to write. This allows the list to be held stable in both network interrupt contexts and sleepable user threads across sleeping memory allocations or device driver interactions. As before, writes to the interface list must occur from sleepable contexts. Reviewed by: bz, julian MFC after: 3 days
2009-08-23 20:40:19 +00:00
IFNET_RLOCK();
UDP: further performance improvements on tx Cumulative throughput while running 64 netperf -H $DUT -t UDP_STREAM -- -m 1 on a 2x8x2 SKL went from 1.1Mpps to 2.5Mpps Single stream throughput increases from 910kpps to 1.18Mpps Baseline: https://people.freebsd.org/~mmacy/2018.05.11/udpsender2.svg - Protect read access to global ifnet list with epoch https://people.freebsd.org/~mmacy/2018.05.11/udpsender3.svg - Protect short lived ifaddr references with epoch https://people.freebsd.org/~mmacy/2018.05.11/udpsender4.svg - Convert if_afdata read lock path to epoch https://people.freebsd.org/~mmacy/2018.05.11/udpsender5.svg A fix for the inpcbhash contention is pending sufficient time on a canary at LLNW. Reviewed by: gallatin Sponsored by: Limelight Networks Differential Revision: https://reviews.freebsd.org/D15409
2018-05-23 21:02:14 +00:00
CK_STAILQ_FOREACH(ifp, &V_ifnet, if_link) {
Rewrite linux_ifconf() to be more like ifconf() in net/if.c so that we do not call uiomove() while IFNET_RLOCK() is held. This eliminates the witness warning: Calling uiomove() with the following non-sleepable locks held: exclusive sleep mutex ifnet r = 0 (0xc096dd60) locked @ /usr/src/sys/modules/linux/../../compat/linux/linux_ioctl.c:2170 MFC after: 2 days
2005-08-27 14:44:10 +00:00
int addrs = 0;
Have SIOCGIFCONF return all (if any) AF_INET addresses for the interfaces we encounter. In Linux, all addresses are returned for which gifconf handlers are installed. This boils down to AF_DECnet and AF_INET. We care mostly about AF_INET for now. Adding additional families is simple enough. Returning the addresses is important for RPC clients to function properly. Andrew found in some reference code that the logic that handles the retransmission looks for an interface that's up and has an AF_INET address. This obviously failed as we didn't return any addresses at all. Note also that with this change we don't return interfaces that don't have AF_INET addresses, whereas before we returned any interface present in the system. This is in line with what Linux does (modulo interfaces with only AF_DECnet addresses of course :-) Reported by: "Andrew Atrens" <atrens@nortelnetworks.com> MFC after: 1 week
2002-01-29 06:00:11 +00:00
More caddr_t removal, in conjunction with copy{in,out}(9) this time. Also clean up some egregious casts and incorrect use of sizeof.
2003-03-03 09:14:26 +00:00
bzero(&ifr, sizeof(ifr));
Further fixes related to the interface renaming. Now that we properly translate the interface name passed to us, make sure we also translate correctly before we return the list of interfaces with the SIOCGIFCONF ioctl. It is common to use the interface names returned by that ioctl in further ioctls, such as SIOCGIFFLAGS. Remove linux_ifname as it is no longer used. Also remove ifname_bsd_to_linux as it cannot be used anymore now that linux_ifname is removed (was deadcode anyway). Reported and tested by: Andrew Atrens <atrens@nortelnetworks.com>
2002-01-10 05:36:36 +00:00
if (IFP_IS_ETH(ifp))
snprintf(ifr.ifr_name, LINUX_IFNAMSIZ, "eth%d",
ethno++);
else
Replace the if_name and if_unit members of struct ifnet with new members if_xname, if_dname, and if_dunit. if_xname is the name of the interface and if_dname/unit are the driver name and instance. This change paves the way for interface renaming and enhanced pseudo device creation and configuration symantics. Approved By: re (in principle) Reviewed By: njl, imp Tested On: i386, amd64, sparc64 Obtained From: NetBSD (if_xname)
2003-10-31 18:32:15 +00:00
strlcpy(ifr.ifr_name, ifp->if_xname, LINUX_IFNAMSIZ);
Have SIOCGIFCONF return all (if any) AF_INET addresses for the interfaces we encounter. In Linux, all addresses are returned for which gifconf handlers are installed. This boils down to AF_DECnet and AF_INET. We care mostly about AF_INET for now. Adding additional families is simple enough. Returning the addresses is important for RPC clients to function properly. Andrew found in some reference code that the logic that handles the retransmission looks for an interface that's up and has an AF_INET address. This obviously failed as we didn't return any addresses at all. Note also that with this change we don't return interfaces that don't have AF_INET addresses, whereas before we returned any interface present in the system. This is in line with what Linux does (modulo interfaces with only AF_DECnet addresses of course :-) Reported by: "Andrew Atrens" <atrens@nortelnetworks.com> MFC after: 1 week
2002-01-29 06:00:11 +00:00
/* Walk the address list */
ifnet: Replace if_addr_lock rwlock with epoch + mutex Run on LLNW canaries and tested by pho@ gallatin: Using a 14-core, 28-HTT single socket E5-2697 v3 with a 40GbE MLX5 based ConnectX 4-LX NIC, I see an almost 12% improvement in received packet rate, and a larger improvement in bytes delivered all the way to userspace. When the host receiving 64 streams of netperf -H $DUT -t UDP_STREAM -- -m 1, I see, using nstat -I mce0 1 before the patch: InMpps OMpps InGbs OGbs err TCP Est %CPU syscalls csw irq GBfree 4.98 0.00 4.42 0.00 4235592 33 83.80 4720653 2149771 1235 247.32 4.73 0.00 4.20 0.00 4025260 33 82.99 4724900 2139833 1204 247.32 4.72 0.00 4.20 0.00 4035252 33 82.14 4719162 2132023 1264 247.32 4.71 0.00 4.21 0.00 4073206 33 83.68 4744973 2123317 1347 247.32 4.72 0.00 4.21 0.00 4061118 33 80.82 4713615 2188091 1490 247.32 4.72 0.00 4.21 0.00 4051675 33 85.29 4727399 2109011 1205 247.32 4.73 0.00 4.21 0.00 4039056 33 84.65 4724735 2102603 1053 247.32 After the patch InMpps OMpps InGbs OGbs err TCP Est %CPU syscalls csw irq GBfree 5.43 0.00 4.20 0.00 3313143 33 84.96 5434214 1900162 2656 245.51 5.43 0.00 4.20 0.00 3308527 33 85.24 5439695 1809382 2521 245.51 5.42 0.00 4.19 0.00 3316778 33 87.54 5416028 1805835 2256 245.51 5.42 0.00 4.19 0.00 3317673 33 90.44 5426044 1763056 2332 245.51 5.42 0.00 4.19 0.00 3314839 33 88.11 5435732 1792218 2499 245.52 5.44 0.00 4.19 0.00 3293228 33 91.84 5426301 1668597 2121 245.52 Similarly, netperf reports 230Mb/s before the patch, and 270Mb/s after the patch Reviewed by: gallatin Sponsored by: Limelight Networks Differential Revision: https://reviews.freebsd.org/D15366
2018-05-18 20:13:34 +00:00
CK_STAILQ_FOREACH(ifa, &ifp->if_addrhead, ifa_link) {
Have SIOCGIFCONF return all (if any) AF_INET addresses for the interfaces we encounter. In Linux, all addresses are returned for which gifconf handlers are installed. This boils down to AF_DECnet and AF_INET. We care mostly about AF_INET for now. Adding additional families is simple enough. Returning the addresses is important for RPC clients to function properly. Andrew found in some reference code that the logic that handles the retransmission looks for an interface that's up and has an AF_INET address. This obviously failed as we didn't return any addresses at all. Note also that with this change we don't return interfaces that don't have AF_INET addresses, whereas before we returned any interface present in the system. This is in line with what Linux does (modulo interfaces with only AF_DECnet addresses of course :-) Reported by: "Andrew Atrens" <atrens@nortelnetworks.com> MFC after: 1 week
2002-01-29 06:00:11 +00:00
struct sockaddr *sa = ifa->ifa_addr;
if (sa->sa_family == AF_INET) {
ifr.ifr_addr.sa_family = LINUX_AF_INET;
memcpy(ifr.ifr_addr.sa_data, sa->sa_data,
sizeof(ifr.ifr_addr.sa_data));
Rewrite linux_ifconf() to be more like ifconf() in net/if.c so that we do not call uiomove() while IFNET_RLOCK() is held. This eliminates the witness warning: Calling uiomove() with the following non-sleepable locks held: exclusive sleep mutex ifnet r = 0 (0xc096dd60) locked @ /usr/src/sys/modules/linux/../../compat/linux/linux_ioctl.c:2170 MFC after: 2 days
2005-08-27 14:44:10 +00:00
sbuf_bcat(sb, &ifr, sizeof(ifr));
max_len += sizeof(ifr);
addrs++;
Have SIOCGIFCONF return all (if any) AF_INET addresses for the interfaces we encounter. In Linux, all addresses are returned for which gifconf handlers are installed. This boils down to AF_DECnet and AF_INET. We care mostly about AF_INET for now. Adding additional families is simple enough. Returning the addresses is important for RPC clients to function properly. Andrew found in some reference code that the logic that handles the retransmission looks for an interface that's up and has an AF_INET address. This obviously failed as we didn't return any addresses at all. Note also that with this change we don't return interfaces that don't have AF_INET addresses, whereas before we returned any interface present in the system. This is in line with what Linux does (modulo interfaces with only AF_DECnet addresses of course :-) Reported by: "Andrew Atrens" <atrens@nortelnetworks.com> MFC after: 1 week
2002-01-29 06:00:11 +00:00
}
Rewrite linux_ifconf() to be more like ifconf() in net/if.c so that we do not call uiomove() while IFNET_RLOCK() is held. This eliminates the witness warning: Calling uiomove() with the following non-sleepable locks held: exclusive sleep mutex ifnet r = 0 (0xc096dd60) locked @ /usr/src/sys/modules/linux/../../compat/linux/linux_ioctl.c:2170 MFC after: 2 days
2005-08-27 14:44:10 +00:00
Replace sbuf_overflowed() with sbuf_error(), which returns any error code associated with overflow or with the drain function. While this function is not expected to be used often, it produces more information in the form of an errno that sbuf_overflowed() did.
2010-09-10 16:42:16 +00:00
if (sbuf_error(sb) == 0)
Rewrite linux_ifconf() to be more like ifconf() in net/if.c so that we do not call uiomove() while IFNET_RLOCK() is held. This eliminates the witness warning: Calling uiomove() with the following non-sleepable locks held: exclusive sleep mutex ifnet r = 0 (0xc096dd60) locked @ /usr/src/sys/modules/linux/../../compat/linux/linux_ioctl.c:2170 MFC after: 2 days
2005-08-27 14:44:10 +00:00
valid_len = sbuf_len(sb);
}
if (addrs == 0) {
bzero((caddr_t)&ifr.ifr_addr, sizeof(ifr.ifr_addr));
sbuf_bcat(sb, &ifr, sizeof(ifr));
max_len += sizeof(ifr);
Replace sbuf_overflowed() with sbuf_error(), which returns any error code associated with overflow or with the drain function. While this function is not expected to be used often, it produces more information in the form of an errno that sbuf_overflowed() did.
2010-09-10 16:42:16 +00:00
if (sbuf_error(sb) == 0)
Rewrite linux_ifconf() to be more like ifconf() in net/if.c so that we do not call uiomove() while IFNET_RLOCK() is held. This eliminates the witness warning: Calling uiomove() with the following non-sleepable locks held: exclusive sleep mutex ifnet r = 0 (0xc096dd60) locked @ /usr/src/sys/modules/linux/../../compat/linux/linux_ioctl.c:2170 MFC after: 2 days
2005-08-27 14:44:10 +00:00
valid_len = sbuf_len(sb);
Have SIOCGIFCONF return all (if any) AF_INET addresses for the interfaces we encounter. In Linux, all addresses are returned for which gifconf handlers are installed. This boils down to AF_DECnet and AF_INET. We care mostly about AF_INET for now. Adding additional families is simple enough. Returning the addresses is important for RPC clients to function properly. Andrew found in some reference code that the logic that handles the retransmission looks for an interface that's up and has an AF_INET address. This obviously failed as we didn't return any addresses at all. Note also that with this change we don't return interfaces that don't have AF_INET addresses, whereas before we returned any interface present in the system. This is in line with what Linux does (modulo interfaces with only AF_DECnet addresses of course :-) Reported by: "Andrew Atrens" <atrens@nortelnetworks.com> MFC after: 1 week
2002-01-29 06:00:11 +00:00
}
Try to make Linux socket ioctls work. Up until now they've only *pretended* to work, but haven't really due to subtle differences in structs etc. This is still not perfect (some ioctls are still known not to work, while others haven't been tested at all), but it's enough to get Debian's ifconfig to produce relatively sane output. More work will be needed to get all ioctls (or at least a reasonable subset) working, and to support the Cisco Aironet config tool mentioned in the PR. PR: 26546 Submitted by: Doug Ambrisko <ambrisko@ambrisko.com>
2001-10-15 20:52:17 +00:00
}
SMP locking for ifnet list.
2002-12-22 05:35:03 +00:00
IFNET_RUNLOCK();
Try to make Linux socket ioctls work. Up until now they've only *pretended* to work, but haven't really due to subtle differences in structs etc. This is still not perfect (some ioctls are still known not to work, while others haven't been tested at all), but it's enough to get Debian's ifconfig to produce relatively sane output. More work will be needed to get all ioctls (or at least a reasonable subset) working, and to support the Cisco Aironet config tool mentioned in the PR. PR: 26546 Submitted by: Doug Ambrisko <ambrisko@ambrisko.com>
2001-10-15 20:52:17 +00:00
Rewrite linux_ifconf() to be more like ifconf() in net/if.c so that we do not call uiomove() while IFNET_RLOCK() is held. This eliminates the witness warning: Calling uiomove() with the following non-sleepable locks held: exclusive sleep mutex ifnet r = 0 (0xc096dd60) locked @ /usr/src/sys/modules/linux/../../compat/linux/linux_ioctl.c:2170 MFC after: 2 days
2005-08-27 14:44:10 +00:00
if (valid_len != max_len && !full) {
sbuf_delete(sb);
goto again;
}
Linuxolator whitespace cleanup A version of each of the MD files by necessity exists for each CPU architecture supported by the Linuxolator. Clean these up so that new architectures do not inherit whitespace issues. Clean up shared Linuxolator files while here. Sponsored by: Turing Robotic Industries Inc.
2018-02-05 17:29:12 +00:00
ifc.ifc_len = valid_len;
Rewrite linux_ifconf() to be more like ifconf() in net/if.c so that we do not call uiomove() while IFNET_RLOCK() is held. This eliminates the witness warning: Calling uiomove() with the following non-sleepable locks held: exclusive sleep mutex ifnet r = 0 (0xc096dd60) locked @ /usr/src/sys/modules/linux/../../compat/linux/linux_ioctl.c:2170 MFC after: 2 days
2005-08-27 14:44:10 +00:00
sbuf_finish(sb);
MFS security patches which seem to have accidentally not reached HEAD: Fix insufficient message length validation for EAP-TLS messages. Fix Linux compatibility layer input validation error. Security: FreeBSD-SA-12:07.hostapd Security: FreeBSD-SA-12:08.linux Security: CVE-2012-4445, CVE-2012-4576 With hat: so@
2012-11-23 01:48:31 +00:00
error = copyout(sbuf_data(sb), PTRIN(ifc.ifc_buf), ifc.ifc_len);
if (error == 0)
error = copyout(&ifc, uifc, sizeof(ifc));
Rewrite linux_ifconf() to be more like ifconf() in net/if.c so that we do not call uiomove() while IFNET_RLOCK() is held. This eliminates the witness warning: Calling uiomove() with the following non-sleepable locks held: exclusive sleep mutex ifnet r = 0 (0xc096dd60) locked @ /usr/src/sys/modules/linux/../../compat/linux/linux_ioctl.c:2170 MFC after: 2 days
2005-08-27 14:44:10 +00:00
sbuf_delete(sb);
Fix a few panics in linuxulator + VIMAGE due to curvnet not being set. This change affects only options VIMAGE builds. Reviewed by: julian MFC after: 3 days
2009-08-28 22:51:07 +00:00
CURVNET_RESTORE();
Further fixes related to the interface renaming. Now that we properly translate the interface name passed to us, make sure we also translate correctly before we return the list of interfaces with the SIOCGIFCONF ioctl. It is common to use the interface names returned by that ioctl in further ioctls, such as SIOCGIFFLAGS. Remove linux_ifname as it is no longer used. Also remove ifname_bsd_to_linux as it cannot be used anymore now that linux_ifname is removed (was deadcode anyway). Reported and tested by: Andrew Atrens <atrens@nortelnetworks.com>
2002-01-10 05:36:36 +00:00
Try to make Linux socket ioctls work. Up until now they've only *pretended* to work, but haven't really due to subtle differences in structs etc. This is still not perfect (some ioctls are still known not to work, while others haven't been tested at all), but it's enough to get Debian's ifconfig to produce relatively sane output. More work will be needed to get all ioctls (or at least a reasonable subset) working, and to support the Cisco Aironet config tool mentioned in the PR. PR: 26546 Submitted by: Doug Ambrisko <ambrisko@ambrisko.com>
2001-10-15 20:52:17 +00:00
return (error);
}
static int
linux_gifflags(struct thread *td, struct ifnet *ifp, struct l_ifreq *ifr)
{
l_short flags;
In order to reduce duplication between MD parts of the Linuxulator move bits that are MI out into the headers in compat/linux. For that remove bogus _packed attribute from struct l_sockaddr and use MI types for struct members. And continue to move into the linux_common module a code that is intended for both Linuxulator modules (both instruction set - 32 & 64 bit) or for external modules like linsysfs or linprocfs. To avoid header pollution introduce new sys/compat/linux_common.h header. Reviewed by: emaste MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D20137
2019-05-03 08:42:49 +00:00
linux_ifflags(ifp, &flags);
Try to make Linux socket ioctls work. Up until now they've only *pretended* to work, but haven't really due to subtle differences in structs etc. This is still not perfect (some ioctls are still known not to work, while others haven't been tested at all), but it's enough to get Debian's ifconfig to produce relatively sane output. More work will be needed to get all ioctls (or at least a reasonable subset) working, and to support the Cisco Aironet config tool mentioned in the PR. PR: 26546 Submitted by: Doug Ambrisko <ambrisko@ambrisko.com>
2001-10-15 20:52:17 +00:00
More caddr_t removal, in conjunction with copy{in,out}(9) this time. Also clean up some egregious casts and incorrect use of sizeof.
2003-03-03 09:14:26 +00:00
return (copyout(&flags, &ifr->ifr_flags, sizeof(flags)));
Try to make Linux socket ioctls work. Up until now they've only *pretended* to work, but haven't really due to subtle differences in structs etc. This is still not perfect (some ioctls are still known not to work, while others haven't been tested at all), but it's enough to get Debian's ifconfig to produce relatively sane output. More work will be needed to get all ioctls (or at least a reasonable subset) working, and to support the Cisco Aironet config tool mentioned in the PR. PR: 26546 Submitted by: Doug Ambrisko <ambrisko@ambrisko.com>
2001-10-15 20:52:17 +00:00
}
static int
Tweak the way we determine if an interface needs to have its name translated. Add some missing break statements in the socket ioctl switch. Check the return value from copyin() / copyout(). Fix some disorderings and misindentations. Support a couple more socket ioctls. Add missing break statements.
2001-10-20 00:01:26 +00:00
linux_gifhwaddr(struct ifnet *ifp, struct l_ifreq *ifr)
Try to make Linux socket ioctls work. Up until now they've only *pretended* to work, but haven't really due to subtle differences in structs etc. This is still not perfect (some ioctls are still known not to work, while others haven't been tested at all), but it's enough to get Debian's ifconfig to produce relatively sane output. More work will be needed to get all ioctls (or at least a reasonable subset) working, and to support the Cisco Aironet config tool mentioned in the PR. PR: 26546 Submitted by: Doug Ambrisko <ambrisko@ambrisko.com>
2001-10-15 20:52:17 +00:00
{
struct l_sockaddr lsa;
In order to reduce duplication between MD parts of the Linuxulator move bits that are MI out into the headers in compat/linux. For that remove bogus _packed attribute from struct l_sockaddr and use MI types for struct members. And continue to move into the linux_common module a code that is intended for both Linuxulator modules (both instruction set - 32 & 64 bit) or for external modules like linsysfs or linprocfs. To avoid header pollution introduce new sys/compat/linux_common.h header. Reviewed by: emaste MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D20137
2019-05-03 08:42:49 +00:00
if (linux_ifhwaddr(ifp, &lsa) != 0)
Try to make Linux socket ioctls work. Up until now they've only *pretended* to work, but haven't really due to subtle differences in structs etc. This is still not perfect (some ioctls are still known not to work, while others haven't been tested at all), but it's enough to get Debian's ifconfig to produce relatively sane output. More work will be needed to get all ioctls (or at least a reasonable subset) working, and to support the Cisco Aironet config tool mentioned in the PR. PR: 26546 Submitted by: Doug Ambrisko <ambrisko@ambrisko.com>
2001-10-15 20:52:17 +00:00
return (ENOENT);
In order to reduce duplication between MD parts of the Linuxulator move bits that are MI out into the headers in compat/linux. For that remove bogus _packed attribute from struct l_sockaddr and use MI types for struct members. And continue to move into the linux_common module a code that is intended for both Linuxulator modules (both instruction set - 32 & 64 bit) or for external modules like linsysfs or linprocfs. To avoid header pollution introduce new sys/compat/linux_common.h header. Reviewed by: emaste MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D20137
2019-05-03 08:42:49 +00:00
return (copyout(&lsa, &ifr->ifr_hwaddr, sizeof(lsa)));
Try to make Linux socket ioctls work. Up until now they've only *pretended* to work, but haven't really due to subtle differences in structs etc. This is still not perfect (some ioctls are still known not to work, while others haven't been tested at all), but it's enough to get Debian's ifconfig to produce relatively sane output. More work will be needed to get all ioctls (or at least a reasonable subset) working, and to support the Cisco Aironet config tool mentioned in the PR. PR: 26546 Submitted by: Doug Ambrisko <ambrisko@ambrisko.com>
2001-10-15 20:52:17 +00:00
}
Get rid of the need of COMPAT_43 in the linuxolator. Submitted by: Divacky Roman <xdivac02@stud.fit.vutbr.cz> Obtained from: DragonFly (some parts)
2006-03-18 18:20:17 +00:00
/*
* If we fault in bsd_to_linux_ifreq() then we will fault when we call
* the native ioctl(). Thus, we don't really need to check the return
* value of this function.
*/
static int
bsd_to_linux_ifreq(struct ifreq *arg)
{
struct ifreq ifr;
size_t ifr_len = sizeof(struct ifreq);
int error;
Linuxolator whitespace cleanup A version of each of the MD files by necessity exists for each CPU architecture supported by the Linuxolator. Clean these up so that new architectures do not inherit whitespace issues. Clean up shared Linuxolator files while here. Sponsored by: Turing Robotic Industries Inc.
2018-02-05 17:29:12 +00:00
Get rid of the need of COMPAT_43 in the linuxolator. Submitted by: Divacky Roman <xdivac02@stud.fit.vutbr.cz> Obtained from: DragonFly (some parts)
2006-03-18 18:20:17 +00:00
if ((error = copyin(arg, &ifr, ifr_len)))
return (error);
Linuxolator whitespace cleanup A version of each of the MD files by necessity exists for each CPU architecture supported by the Linuxolator. Clean these up so that new architectures do not inherit whitespace issues. Clean up shared Linuxolator files while here. Sponsored by: Turing Robotic Industries Inc.
2018-02-05 17:29:12 +00:00
Get rid of the need of COMPAT_43 in the linuxolator. Submitted by: Divacky Roman <xdivac02@stud.fit.vutbr.cz> Obtained from: DragonFly (some parts)
2006-03-18 18:20:17 +00:00
*(u_short *)&ifr.ifr_addr = ifr.ifr_addr.sa_family;
Linuxolator whitespace cleanup A version of each of the MD files by necessity exists for each CPU architecture supported by the Linuxolator. Clean these up so that new architectures do not inherit whitespace issues. Clean up shared Linuxolator files while here. Sponsored by: Turing Robotic Industries Inc.
2018-02-05 17:29:12 +00:00
Get rid of the need of COMPAT_43 in the linuxolator. Submitted by: Divacky Roman <xdivac02@stud.fit.vutbr.cz> Obtained from: DragonFly (some parts)
2006-03-18 18:20:17 +00:00
error = copyout(&ifr, arg, ifr_len);
return (error);
}
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
/*
* Socket related ioctls
*/
- Handle mixer read ioctls correctly. They have the same group, number and argument size as their write counterparts and were handled as write ioctls. - Emulate some cdrom ioctls.
1999-04-29 04:37:57 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
static int
KSE Milestone 2 Note ALL MODULES MUST BE RECOMPILED make the kernel aware that there are smaller units of scheduling than the process. (but only allow one thread per process at this time). This is functionally equivalent to teh previousl -current except that there is a thread associated with each process. Sorry john! (your next MFC will be a doosie!) Reviewed by: peter@freebsd.org, dillon@freebsd.org X-MFC after: ha ha ha ha
2001-09-12 08:38:13 +00:00
linux_ioctl_socket(struct thread *td, struct linux_ioctl_args *args)
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
{
Try to make Linux socket ioctls work. Up until now they've only *pretended* to work, but haven't really due to subtle differences in structs etc. This is still not perfect (some ioctls are still known not to work, while others haven't been tested at all), but it's enough to get Debian's ifconfig to produce relatively sane output. More work will be needed to get all ioctls (or at least a reasonable subset) working, and to support the Cisco Aironet config tool mentioned in the PR. PR: 26546 Submitted by: Doug Ambrisko <ambrisko@ambrisko.com>
2001-10-15 20:52:17 +00:00
char lifname[LINUX_IFNAMSIZ], ifname[IFNAMSIZ];
struct ifnet *ifp;
Deal with a few issues that cropped up following the recent changes to the code for translating socket and private ioctls: - Only perform socket ioctl translation if the file descriptor is a socket. - Treat socket ioctls on non-sockets specially, and for now assume that these are directed at a tap/vmnet device, so translate the ioctl numbers as appropriate (the way if_tap abuses some socket ioctls to pass non-ifreq data is utterly bogus, but this is how VMware on FreeBSD has always "worked"; I will deal with this later). - Add (untested) support for translating SIOCSIFADDR. - In all cases where we fail to translate an ioctl, return ENOIOCTL so that other handlers have a chance to do the translation. This should fix the "/dev/vmnet1: Invalid argument" errors that users of VMware were experiencing, though I have only verified this on RELENG_4. Submitted by: des (mostly) MFC after: 3 days
2001-11-19 15:43:50 +00:00
struct file *fp;
int error, type;
Try to make Linux socket ioctls work. Up until now they've only *pretended* to work, but haven't really due to subtle differences in structs etc. This is still not perfect (some ioctls are still known not to work, while others haven't been tested at all), but it's enough to get Debian's ifconfig to produce relatively sane output. More work will be needed to get all ioctls (or at least a reasonable subset) working, and to support the Cisco Aironet config tool mentioned in the PR. PR: 26546 Submitted by: Doug Ambrisko <ambrisko@ambrisko.com>
2001-10-15 20:52:17 +00:00
ifp = NULL;
error = 0;
Clean up whitespace, s/register //, refrain from strong urge to ANSIfy.
2003-03-02 15:56:49 +00:00
Eliminate the overhead of gratuitous repeated reinitialization of cap_rights - Add macros to allow preinitialization of cap_rights_t. - Convert most commonly used code paths to use preinitialized cap_rights_t. A 3.6% speedup in fstat was measured with this change. Reported by: mjg Reviewed by: oshogbo Approved by: sbruno MFC after: 1 month
2018-05-09 18:47:24 +00:00
error = fget(td, args->fd, &cap_ioctl_rights, &fp);
Change the cap_rights_t type from uint64_t to a structure that we can extend in the future in a backward compatible (API and ABI) way. The cap_rights_t represents capability rights. We used to use one bit to represent one right, but we are running out of spare bits. Currently the new structure provides place for 114 rights (so 50 more than the previous cap_rights_t), but it is possible to grow the structure to hold at least 285 rights, although we can make it even larger if 285 rights won't be enough. The structure definition looks like this: struct cap_rights { uint64_t cr_rights[CAP_RIGHTS_VERSION + 2]; }; The initial CAP_RIGHTS_VERSION is 0. The top two bits in the first element of the cr_rights[] array contain total number of elements in the array - 2. This means if those two bits are equal to 0, we have 2 array elements. The top two bits in all remaining array elements should be 0. The next five bits in all array elements contain array index. Only one bit is used and bit position in this five-bits range defines array index. This means there can be at most five array elements in the future. To define new right the CAPRIGHT() macro must be used. The macro takes two arguments - an array index and a bit to set, eg. #define CAP_PDKILL CAPRIGHT(1, 0x0000000000000800ULL) We still support aliases that combine few rights, but the rights have to belong to the same array element, eg: #define CAP_LOOKUP CAPRIGHT(0, 0x0000000000000400ULL) #define CAP_FCHMOD CAPRIGHT(0, 0x0000000000002000ULL) #define CAP_FCHMODAT (CAP_FCHMOD | CAP_LOOKUP) There is new API to manage the new cap_rights_t structure: cap_rights_t *cap_rights_init(cap_rights_t *rights, ...); void cap_rights_set(cap_rights_t *rights, ...); void cap_rights_clear(cap_rights_t *rights, ...); bool cap_rights_is_set(const cap_rights_t *rights, ...); bool cap_rights_is_valid(const cap_rights_t *rights); void cap_rights_merge(cap_rights_t *dst, const cap_rights_t *src); void cap_rights_remove(cap_rights_t *dst, const cap_rights_t *src); bool cap_rights_contains(const cap_rights_t *big, const cap_rights_t *little); Capability rights to the cap_rights_init(), cap_rights_set(), cap_rights_clear() and cap_rights_is_set() functions are provided by separating them with commas, eg: cap_rights_t rights; cap_rights_init(&rights, CAP_READ, CAP_WRITE, CAP_FSTAT); There is no need to terminate the list of rights, as those functions are actually macros that take care of the termination, eg: #define cap_rights_set(rights, ...) \ __cap_rights_set((rights), __VA_ARGS__, 0ULL) void __cap_rights_set(cap_rights_t *rights, ...); Thanks to using one bit as an array index we can assert in those functions that there are no two rights belonging to different array elements provided together. For example this is illegal and will be detected, because CAP_LOOKUP belongs to element 0 and CAP_PDKILL to element 1: cap_rights_init(&rights, CAP_LOOKUP | CAP_PDKILL); Providing several rights that belongs to the same array's element this way is correct, but is not advised. It should only be used for aliases definition. This commit also breaks compatibility with some existing Capsicum system calls, but I see no other way to do that. This should be fine as Capsicum is still experimental and this change is not going to 9.x. Sponsored by: The FreeBSD Foundation
2013-09-05 00:09:56 +00:00
if (error != 0)
Deal with a few issues that cropped up following the recent changes to the code for translating socket and private ioctls: - Only perform socket ioctl translation if the file descriptor is a socket. - Treat socket ioctls on non-sockets specially, and for now assume that these are directed at a tap/vmnet device, so translate the ioctl numbers as appropriate (the way if_tap abuses some socket ioctls to pass non-ifreq data is utterly bogus, but this is how VMware on FreeBSD has always "worked"; I will deal with this later). - Add (untested) support for translating SIOCSIFADDR. - In all cases where we fail to translate an ioctl, return ENOIOCTL so that other handlers have a chance to do the translation. This should fix the "/dev/vmnet1: Invalid argument" errors that users of VMware were experiencing, though I have only verified this on RELENG_4. Submitted by: des (mostly) MFC after: 3 days
2001-11-19 15:43:50 +00:00
return (error);
type = fp->f_type;
fdrop(fp, td);
if (type != DTYPE_SOCKET) {
/* not a socket - probably a tap / vmnet device */
switch (args->cmd) {
case LINUX_SIOCGIFADDR:
case LINUX_SIOCSIFADDR:
case LINUX_SIOCGIFFLAGS:
return (linux_ioctl_special(td, args));
default:
return (ENOIOCTL);
}
}
Try to make Linux socket ioctls work. Up until now they've only *pretended* to work, but haven't really due to subtle differences in structs etc. This is still not perfect (some ioctls are still known not to work, while others haven't been tested at all), but it's enough to get Debian's ifconfig to produce relatively sane output. More work will be needed to get all ioctls (or at least a reasonable subset) working, and to support the Cisco Aironet config tool mentioned in the PR. PR: 26546 Submitted by: Doug Ambrisko <ambrisko@ambrisko.com>
2001-10-15 20:52:17 +00:00
switch (args->cmd & 0xffff) {
Clean up whitespace, s/register //, refrain from strong urge to ANSIfy.
2003-03-02 15:56:49 +00:00
Try to make Linux socket ioctls work. Up until now they've only *pretended* to work, but haven't really due to subtle differences in structs etc. This is still not perfect (some ioctls are still known not to work, while others haven't been tested at all), but it's enough to get Debian's ifconfig to produce relatively sane output. More work will be needed to get all ioctls (or at least a reasonable subset) working, and to support the Cisco Aironet config tool mentioned in the PR. PR: 26546 Submitted by: Doug Ambrisko <ambrisko@ambrisko.com>
2001-10-15 20:52:17 +00:00
case LINUX_FIOGETOWN:
case LINUX_FIOSETOWN:
case LINUX_SIOCADDMULTI:
case LINUX_SIOCATMARK:
case LINUX_SIOCDELMULTI:
Adding SIOCGIFNAME support in Linuxulator. This should silence the console warning associated with linux-opera: linux: pid 23492 (opera): ioctl fd=5, cmd=0x8910 ('\M^I',16) is not implemented linux: pid 23492 (opera): ioctl fd=28, cmd=0x8910 ('\M^I',16) is not implemented ... Reviewed by: kib, marcel, dchagin Tested with: linux-opera-12.16_3 MFC after: 1 month
2017-04-09 15:27:04 +00:00
case LINUX_SIOCGIFNAME:
Try to make Linux socket ioctls work. Up until now they've only *pretended* to work, but haven't really due to subtle differences in structs etc. This is still not perfect (some ioctls are still known not to work, while others haven't been tested at all), but it's enough to get Debian's ifconfig to produce relatively sane output. More work will be needed to get all ioctls (or at least a reasonable subset) working, and to support the Cisco Aironet config tool mentioned in the PR. PR: 26546 Submitted by: Doug Ambrisko <ambrisko@ambrisko.com>
2001-10-15 20:52:17 +00:00
case LINUX_SIOCGIFCONF:
case LINUX_SIOCGPGRP:
case LINUX_SIOCSPGRP:
Implement LINUX_SIOCGIFCOUNT and LINUX_SIOCGIFINDEX/LINUX_SIOGIFINDEX. LINUX_SIOCGIFCOUNT just returns 0 since it is not implemented in the Linux 2.6.16. LINUX_SIOCGIFINDEX/LINUX_SIOGIFINDEX are mapped to the FreeBSD native SIOCGIFINDEX. Tested by: Peter Kostouros <kpeter@melbpc.org.au> Reviewed by: brooks, rpaulo (on net@) Submitted by: rdivacky MFC after: 1 week
2007-11-07 16:42:52 +00:00
case LINUX_SIOCGIFCOUNT:
Try to make Linux socket ioctls work. Up until now they've only *pretended* to work, but haven't really due to subtle differences in structs etc. This is still not perfect (some ioctls are still known not to work, while others haven't been tested at all), but it's enough to get Debian's ifconfig to produce relatively sane output. More work will be needed to get all ioctls (or at least a reasonable subset) working, and to support the Cisco Aironet config tool mentioned in the PR. PR: 26546 Submitted by: Doug Ambrisko <ambrisko@ambrisko.com>
2001-10-15 20:52:17 +00:00
/* these ioctls don't take an interface name */
break;
Clean up whitespace, s/register //, refrain from strong urge to ANSIfy.
2003-03-02 15:56:49 +00:00
Tweak the way we determine if an interface needs to have its name translated. Add some missing break statements in the socket ioctl switch. Check the return value from copyin() / copyout(). Fix some disorderings and misindentations. Support a couple more socket ioctls. Add missing break statements.
2001-10-20 00:01:26 +00:00
case LINUX_SIOCGIFFLAGS:
Try to make Linux socket ioctls work. Up until now they've only *pretended* to work, but haven't really due to subtle differences in structs etc. This is still not perfect (some ioctls are still known not to work, while others haven't been tested at all), but it's enough to get Debian's ifconfig to produce relatively sane output. More work will be needed to get all ioctls (or at least a reasonable subset) working, and to support the Cisco Aironet config tool mentioned in the PR. PR: 26546 Submitted by: Doug Ambrisko <ambrisko@ambrisko.com>
2001-10-15 20:52:17 +00:00
case LINUX_SIOCGIFADDR:
Deal with a few issues that cropped up following the recent changes to the code for translating socket and private ioctls: - Only perform socket ioctl translation if the file descriptor is a socket. - Treat socket ioctls on non-sockets specially, and for now assume that these are directed at a tap/vmnet device, so translate the ioctl numbers as appropriate (the way if_tap abuses some socket ioctls to pass non-ifreq data is utterly bogus, but this is how VMware on FreeBSD has always "worked"; I will deal with this later). - Add (untested) support for translating SIOCSIFADDR. - In all cases where we fail to translate an ioctl, return ENOIOCTL so that other handlers have a chance to do the translation. This should fix the "/dev/vmnet1: Invalid argument" errors that users of VMware were experiencing, though I have only verified this on RELENG_4. Submitted by: des (mostly) MFC after: 3 days
2001-11-19 15:43:50 +00:00
case LINUX_SIOCSIFADDR:
Try to make Linux socket ioctls work. Up until now they've only *pretended* to work, but haven't really due to subtle differences in structs etc. This is still not perfect (some ioctls are still known not to work, while others haven't been tested at all), but it's enough to get Debian's ifconfig to produce relatively sane output. More work will be needed to get all ioctls (or at least a reasonable subset) working, and to support the Cisco Aironet config tool mentioned in the PR. PR: 26546 Submitted by: Doug Ambrisko <ambrisko@ambrisko.com>
2001-10-15 20:52:17 +00:00
case LINUX_SIOCGIFDSTADDR:
Tweak the way we determine if an interface needs to have its name translated. Add some missing break statements in the socket ioctl switch. Check the return value from copyin() / copyout(). Fix some disorderings and misindentations. Support a couple more socket ioctls. Add missing break statements.
2001-10-20 00:01:26 +00:00
case LINUX_SIOCGIFBRDADDR:
Try to make Linux socket ioctls work. Up until now they've only *pretended* to work, but haven't really due to subtle differences in structs etc. This is still not perfect (some ioctls are still known not to work, while others haven't been tested at all), but it's enough to get Debian's ifconfig to produce relatively sane output. More work will be needed to get all ioctls (or at least a reasonable subset) working, and to support the Cisco Aironet config tool mentioned in the PR. PR: 26546 Submitted by: Doug Ambrisko <ambrisko@ambrisko.com>
2001-10-15 20:52:17 +00:00
case LINUX_SIOCGIFNETMASK:
Tweak the way we determine if an interface needs to have its name translated. Add some missing break statements in the socket ioctl switch. Check the return value from copyin() / copyout(). Fix some disorderings and misindentations. Support a couple more socket ioctls. Add missing break statements.
2001-10-20 00:01:26 +00:00
case LINUX_SIOCSIFNETMASK:
case LINUX_SIOCGIFMTU:
case LINUX_SIOCSIFMTU:
case LINUX_SIOCSIFNAME:
case LINUX_SIOCGIFHWADDR:
case LINUX_SIOCSIFHWADDR:
case LINUX_SIOCDEVPRIVATE:
case LINUX_SIOCDEVPRIVATE+1:
Implement LINUX_SIOCGIFCOUNT and LINUX_SIOCGIFINDEX/LINUX_SIOGIFINDEX. LINUX_SIOCGIFCOUNT just returns 0 since it is not implemented in the Linux 2.6.16. LINUX_SIOCGIFINDEX/LINUX_SIOGIFINDEX are mapped to the FreeBSD native SIOCGIFINDEX. Tested by: Peter Kostouros <kpeter@melbpc.org.au> Reviewed by: brooks, rpaulo (on net@) Submitted by: rdivacky MFC after: 1 week
2007-11-07 16:42:52 +00:00
case LINUX_SIOCGIFINDEX:
Try to make Linux socket ioctls work. Up until now they've only *pretended* to work, but haven't really due to subtle differences in structs etc. This is still not perfect (some ioctls are still known not to work, while others haven't been tested at all), but it's enough to get Debian's ifconfig to produce relatively sane output. More work will be needed to get all ioctls (or at least a reasonable subset) working, and to support the Cisco Aironet config tool mentioned in the PR. PR: 26546 Submitted by: Doug Ambrisko <ambrisko@ambrisko.com>
2001-10-15 20:52:17 +00:00
/* copy in the interface name and translate it. */
More caddr_t removal, in conjunction with copy{in,out}(9) this time. Also clean up some egregious casts and incorrect use of sizeof.
2003-03-03 09:14:26 +00:00
error = copyin((void *)args->arg, lifname, LINUX_IFNAMSIZ);
Tweak the way we determine if an interface needs to have its name translated. Add some missing break statements in the socket ioctl switch. Check the return value from copyin() / copyout(). Fix some disorderings and misindentations. Support a couple more socket ioctls. Add missing break statements.
2001-10-20 00:01:26 +00:00
if (error != 0)
return (error);
Fix kernel memory disclosure in linux_ioctl_socket strlcpy is used to copy a string into a buffer to be copied to userland, previously leaving uninitialized data after the terminating NUL. Zero the buffer first to avoid a kernel memory disclosure. admbugs: 765, 811 MFC after: 1 day Reported by: Ilja Van Sprundel <ivansprundel@ioactive.com> Reported by: Vlad Tsyrklevich Sponsored by: The FreeBSD Foundation
2018-04-04 19:58:25 +00:00
memset(ifname, 0, sizeof(ifname));
Get vnets from creds instead of threads where they're available, and from passed threads instead of curthread. Reviewed by: zec, julian Approved by: bz (mentor)
2009-06-15 19:01:53 +00:00
ifp = ifname_linux_to_bsd(td, lifname, ifname);
Try to make Linux socket ioctls work. Up until now they've only *pretended* to work, but haven't really due to subtle differences in structs etc. This is still not perfect (some ioctls are still known not to work, while others haven't been tested at all), but it's enough to get Debian's ifconfig to produce relatively sane output. More work will be needed to get all ioctls (or at least a reasonable subset) working, and to support the Cisco Aironet config tool mentioned in the PR. PR: 26546 Submitted by: Doug Ambrisko <ambrisko@ambrisko.com>
2001-10-15 20:52:17 +00:00
if (ifp == NULL)
return (EINVAL);
/*
* We need to copy it back out in case we pass the
* request on to our native ioctl(), which will expect
* the ifreq to be in user space and have the correct
* interface name.
*/
More caddr_t removal, in conjunction with copy{in,out}(9) this time. Also clean up some egregious casts and incorrect use of sizeof.
2003-03-03 09:14:26 +00:00
error = copyout(ifname, (void *)args->arg, IFNAMSIZ);
Tweak the way we determine if an interface needs to have its name translated. Add some missing break statements in the socket ioctl switch. Check the return value from copyin() / copyout(). Fix some disorderings and misindentations. Support a couple more socket ioctls. Add missing break statements.
2001-10-20 00:01:26 +00:00
if (error != 0)
return (error);
Try to make Linux socket ioctls work. Up until now they've only *pretended* to work, but haven't really due to subtle differences in structs etc. This is still not perfect (some ioctls are still known not to work, while others haven't been tested at all), but it's enough to get Debian's ifconfig to produce relatively sane output. More work will be needed to get all ioctls (or at least a reasonable subset) working, and to support the Cisco Aironet config tool mentioned in the PR. PR: 26546 Submitted by: Doug Ambrisko <ambrisko@ambrisko.com>
2001-10-15 20:52:17 +00:00
break;
Clean up whitespace, s/register //, refrain from strong urge to ANSIfy.
2003-03-02 15:56:49 +00:00
Try to make Linux socket ioctls work. Up until now they've only *pretended* to work, but haven't really due to subtle differences in structs etc. This is still not perfect (some ioctls are still known not to work, while others haven't been tested at all), but it's enough to get Debian's ifconfig to produce relatively sane output. More work will be needed to get all ioctls (or at least a reasonable subset) working, and to support the Cisco Aironet config tool mentioned in the PR. PR: 26546 Submitted by: Doug Ambrisko <ambrisko@ambrisko.com>
2001-10-15 20:52:17 +00:00
default:
return (ENOIOCTL);
}
- Handle mixer read ioctls correctly. They have the same group, number and argument size as their write counterparts and were handled as write ioctls. - Emulate some cdrom ioctls.
1999-04-29 04:37:57 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
switch (args->cmd & 0xffff) {
case LINUX_FIOSETOWN:
args->cmd = FIOSETOWN;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = sys_ioctl(td, (struct ioctl_args *)args);
Try to make Linux socket ioctls work. Up until now they've only *pretended* to work, but haven't really due to subtle differences in structs etc. This is still not perfect (some ioctls are still known not to work, while others haven't been tested at all), but it's enough to get Debian's ifconfig to produce relatively sane output. More work will be needed to get all ioctls (or at least a reasonable subset) working, and to support the Cisco Aironet config tool mentioned in the PR. PR: 26546 Submitted by: Doug Ambrisko <ambrisko@ambrisko.com>
2001-10-15 20:52:17 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SIOCSPGRP:
args->cmd = SIOCSPGRP;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = sys_ioctl(td, (struct ioctl_args *)args);
Try to make Linux socket ioctls work. Up until now they've only *pretended* to work, but haven't really due to subtle differences in structs etc. This is still not perfect (some ioctls are still known not to work, while others haven't been tested at all), but it's enough to get Debian's ifconfig to produce relatively sane output. More work will be needed to get all ioctls (or at least a reasonable subset) working, and to support the Cisco Aironet config tool mentioned in the PR. PR: 26546 Submitted by: Doug Ambrisko <ambrisko@ambrisko.com>
2001-10-15 20:52:17 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_FIOGETOWN:
args->cmd = FIOGETOWN;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = sys_ioctl(td, (struct ioctl_args *)args);
Try to make Linux socket ioctls work. Up until now they've only *pretended* to work, but haven't really due to subtle differences in structs etc. This is still not perfect (some ioctls are still known not to work, while others haven't been tested at all), but it's enough to get Debian's ifconfig to produce relatively sane output. More work will be needed to get all ioctls (or at least a reasonable subset) working, and to support the Cisco Aironet config tool mentioned in the PR. PR: 26546 Submitted by: Doug Ambrisko <ambrisko@ambrisko.com>
2001-10-15 20:52:17 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SIOCGPGRP:
args->cmd = SIOCGPGRP;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = sys_ioctl(td, (struct ioctl_args *)args);
Try to make Linux socket ioctls work. Up until now they've only *pretended* to work, but haven't really due to subtle differences in structs etc. This is still not perfect (some ioctls are still known not to work, while others haven't been tested at all), but it's enough to get Debian's ifconfig to produce relatively sane output. More work will be needed to get all ioctls (or at least a reasonable subset) working, and to support the Cisco Aironet config tool mentioned in the PR. PR: 26546 Submitted by: Doug Ambrisko <ambrisko@ambrisko.com>
2001-10-15 20:52:17 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SIOCATMARK:
args->cmd = SIOCATMARK;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = sys_ioctl(td, (struct ioctl_args *)args);
Try to make Linux socket ioctls work. Up until now they've only *pretended* to work, but haven't really due to subtle differences in structs etc. This is still not perfect (some ioctls are still known not to work, while others haven't been tested at all), but it's enough to get Debian's ifconfig to produce relatively sane output. More work will be needed to get all ioctls (or at least a reasonable subset) working, and to support the Cisco Aironet config tool mentioned in the PR. PR: 26546 Submitted by: Doug Ambrisko <ambrisko@ambrisko.com>
2001-10-15 20:52:17 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
/* LINUX_SIOCGSTAMP */
Adding SIOCGIFNAME support in Linuxulator. This should silence the console warning associated with linux-opera: linux: pid 23492 (opera): ioctl fd=5, cmd=0x8910 ('\M^I',16) is not implemented linux: pid 23492 (opera): ioctl fd=28, cmd=0x8910 ('\M^I',16) is not implemented ... Reviewed by: kib, marcel, dchagin Tested with: linux-opera-12.16_3 MFC after: 1 month
2017-04-09 15:27:04 +00:00
case LINUX_SIOCGIFNAME:
error = linux_ioctl_ifname(td, (struct l_ifreq *)args->arg);
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SIOCGIFCONF:
Try to make Linux socket ioctls work. Up until now they've only *pretended* to work, but haven't really due to subtle differences in structs etc. This is still not perfect (some ioctls are still known not to work, while others haven't been tested at all), but it's enough to get Debian's ifconfig to produce relatively sane output. More work will be needed to get all ioctls (or at least a reasonable subset) working, and to support the Cisco Aironet config tool mentioned in the PR. PR: 26546 Submitted by: Doug Ambrisko <ambrisko@ambrisko.com>
2001-10-15 20:52:17 +00:00
error = linux_ifconf(td, (struct ifconf *)args->arg);
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SIOCGIFFLAGS:
args->cmd = SIOCGIFFLAGS;
Try to make Linux socket ioctls work. Up until now they've only *pretended* to work, but haven't really due to subtle differences in structs etc. This is still not perfect (some ioctls are still known not to work, while others haven't been tested at all), but it's enough to get Debian's ifconfig to produce relatively sane output. More work will be needed to get all ioctls (or at least a reasonable subset) working, and to support the Cisco Aironet config tool mentioned in the PR. PR: 26546 Submitted by: Doug Ambrisko <ambrisko@ambrisko.com>
2001-10-15 20:52:17 +00:00
error = linux_gifflags(td, ifp, (struct l_ifreq *)args->arg);
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SIOCGIFADDR:
Get rid of the need of COMPAT_43 in the linuxolator. Submitted by: Divacky Roman <xdivac02@stud.fit.vutbr.cz> Obtained from: DragonFly (some parts)
2006-03-18 18:20:17 +00:00
args->cmd = SIOCGIFADDR;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = sys_ioctl(td, (struct ioctl_args *)args);
Get rid of the need of COMPAT_43 in the linuxolator. Submitted by: Divacky Roman <xdivac02@stud.fit.vutbr.cz> Obtained from: DragonFly (some parts)
2006-03-18 18:20:17 +00:00
bsd_to_linux_ifreq((struct ifreq *)args->arg);
Try to make Linux socket ioctls work. Up until now they've only *pretended* to work, but haven't really due to subtle differences in structs etc. This is still not perfect (some ioctls are still known not to work, while others haven't been tested at all), but it's enough to get Debian's ifconfig to produce relatively sane output. More work will be needed to get all ioctls (or at least a reasonable subset) working, and to support the Cisco Aironet config tool mentioned in the PR. PR: 26546 Submitted by: Doug Ambrisko <ambrisko@ambrisko.com>
2001-10-15 20:52:17 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
Deal with a few issues that cropped up following the recent changes to the code for translating socket and private ioctls: - Only perform socket ioctl translation if the file descriptor is a socket. - Treat socket ioctls on non-sockets specially, and for now assume that these are directed at a tap/vmnet device, so translate the ioctl numbers as appropriate (the way if_tap abuses some socket ioctls to pass non-ifreq data is utterly bogus, but this is how VMware on FreeBSD has always "worked"; I will deal with this later). - Add (untested) support for translating SIOCSIFADDR. - In all cases where we fail to translate an ioctl, return ENOIOCTL so that other handlers have a chance to do the translation. This should fix the "/dev/vmnet1: Invalid argument" errors that users of VMware were experiencing, though I have only verified this on RELENG_4. Submitted by: des (mostly) MFC after: 3 days
2001-11-19 15:43:50 +00:00
case LINUX_SIOCSIFADDR:
/* XXX probably doesn't work, included for completeness */
args->cmd = SIOCSIFADDR;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = sys_ioctl(td, (struct ioctl_args *)args);
Deal with a few issues that cropped up following the recent changes to the code for translating socket and private ioctls: - Only perform socket ioctl translation if the file descriptor is a socket. - Treat socket ioctls on non-sockets specially, and for now assume that these are directed at a tap/vmnet device, so translate the ioctl numbers as appropriate (the way if_tap abuses some socket ioctls to pass non-ifreq data is utterly bogus, but this is how VMware on FreeBSD has always "worked"; I will deal with this later). - Add (untested) support for translating SIOCSIFADDR. - In all cases where we fail to translate an ioctl, return ENOIOCTL so that other handlers have a chance to do the translation. This should fix the "/dev/vmnet1: Invalid argument" errors that users of VMware were experiencing, though I have only verified this on RELENG_4. Submitted by: des (mostly) MFC after: 3 days
2001-11-19 15:43:50 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SIOCGIFDSTADDR:
Get rid of the need of COMPAT_43 in the linuxolator. Submitted by: Divacky Roman <xdivac02@stud.fit.vutbr.cz> Obtained from: DragonFly (some parts)
2006-03-18 18:20:17 +00:00
args->cmd = SIOCGIFDSTADDR;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = sys_ioctl(td, (struct ioctl_args *)args);
Get rid of the need of COMPAT_43 in the linuxolator. Submitted by: Divacky Roman <xdivac02@stud.fit.vutbr.cz> Obtained from: DragonFly (some parts)
2006-03-18 18:20:17 +00:00
bsd_to_linux_ifreq((struct ifreq *)args->arg);
Try to make Linux socket ioctls work. Up until now they've only *pretended* to work, but haven't really due to subtle differences in structs etc. This is still not perfect (some ioctls are still known not to work, while others haven't been tested at all), but it's enough to get Debian's ifconfig to produce relatively sane output. More work will be needed to get all ioctls (or at least a reasonable subset) working, and to support the Cisco Aironet config tool mentioned in the PR. PR: 26546 Submitted by: Doug Ambrisko <ambrisko@ambrisko.com>
2001-10-15 20:52:17 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SIOCGIFBRDADDR:
Get rid of the need of COMPAT_43 in the linuxolator. Submitted by: Divacky Roman <xdivac02@stud.fit.vutbr.cz> Obtained from: DragonFly (some parts)
2006-03-18 18:20:17 +00:00
args->cmd = SIOCGIFBRDADDR;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = sys_ioctl(td, (struct ioctl_args *)args);
Get rid of the need of COMPAT_43 in the linuxolator. Submitted by: Divacky Roman <xdivac02@stud.fit.vutbr.cz> Obtained from: DragonFly (some parts)
2006-03-18 18:20:17 +00:00
bsd_to_linux_ifreq((struct ifreq *)args->arg);
Try to make Linux socket ioctls work. Up until now they've only *pretended* to work, but haven't really due to subtle differences in structs etc. This is still not perfect (some ioctls are still known not to work, while others haven't been tested at all), but it's enough to get Debian's ifconfig to produce relatively sane output. More work will be needed to get all ioctls (or at least a reasonable subset) working, and to support the Cisco Aironet config tool mentioned in the PR. PR: 26546 Submitted by: Doug Ambrisko <ambrisko@ambrisko.com>
2001-10-15 20:52:17 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SIOCGIFNETMASK:
Get rid of the need of COMPAT_43 in the linuxolator. Submitted by: Divacky Roman <xdivac02@stud.fit.vutbr.cz> Obtained from: DragonFly (some parts)
2006-03-18 18:20:17 +00:00
args->cmd = SIOCGIFNETMASK;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = sys_ioctl(td, (struct ioctl_args *)args);
Get rid of the need of COMPAT_43 in the linuxolator. Submitted by: Divacky Roman <xdivac02@stud.fit.vutbr.cz> Obtained from: DragonFly (some parts)
2006-03-18 18:20:17 +00:00
bsd_to_linux_ifreq((struct ifreq *)args->arg);
Try to make Linux socket ioctls work. Up until now they've only *pretended* to work, but haven't really due to subtle differences in structs etc. This is still not perfect (some ioctls are still known not to work, while others haven't been tested at all), but it's enough to get Debian's ifconfig to produce relatively sane output. More work will be needed to get all ioctls (or at least a reasonable subset) working, and to support the Cisco Aironet config tool mentioned in the PR. PR: 26546 Submitted by: Doug Ambrisko <ambrisko@ambrisko.com>
2001-10-15 20:52:17 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
Tweak the way we determine if an interface needs to have its name translated. Add some missing break statements in the socket ioctl switch. Check the return value from copyin() / copyout(). Fix some disorderings and misindentations. Support a couple more socket ioctls. Add missing break statements.
2001-10-20 00:01:26 +00:00
case LINUX_SIOCSIFNETMASK:
error = ENOIOCTL;
break;
Clean up whitespace, s/register //, refrain from strong urge to ANSIfy.
2003-03-02 15:56:49 +00:00
Tweak the way we determine if an interface needs to have its name translated. Add some missing break statements in the socket ioctl switch. Check the return value from copyin() / copyout(). Fix some disorderings and misindentations. Support a couple more socket ioctls. Add missing break statements.
2001-10-20 00:01:26 +00:00
case LINUX_SIOCGIFMTU:
args->cmd = SIOCGIFMTU;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = sys_ioctl(td, (struct ioctl_args *)args);
Tweak the way we determine if an interface needs to have its name translated. Add some missing break statements in the socket ioctl switch. Check the return value from copyin() / copyout(). Fix some disorderings and misindentations. Support a couple more socket ioctls. Add missing break statements.
2001-10-20 00:01:26 +00:00
break;
Clean up whitespace, s/register //, refrain from strong urge to ANSIfy.
2003-03-02 15:56:49 +00:00
Tweak the way we determine if an interface needs to have its name translated. Add some missing break statements in the socket ioctl switch. Check the return value from copyin() / copyout(). Fix some disorderings and misindentations. Support a couple more socket ioctls. Add missing break statements.
2001-10-20 00:01:26 +00:00
case LINUX_SIOCSIFMTU:
args->cmd = SIOCSIFMTU;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = sys_ioctl(td, (struct ioctl_args *)args);
Tweak the way we determine if an interface needs to have its name translated. Add some missing break statements in the socket ioctl switch. Check the return value from copyin() / copyout(). Fix some disorderings and misindentations. Support a couple more socket ioctls. Add missing break statements.
2001-10-20 00:01:26 +00:00
break;
Clean up whitespace, s/register //, refrain from strong urge to ANSIfy.
2003-03-02 15:56:49 +00:00
Tweak the way we determine if an interface needs to have its name translated. Add some missing break statements in the socket ioctl switch. Check the return value from copyin() / copyout(). Fix some disorderings and misindentations. Support a couple more socket ioctls. Add missing break statements.
2001-10-20 00:01:26 +00:00
case LINUX_SIOCSIFNAME:
error = ENOIOCTL;
break;
Clean up whitespace, s/register //, refrain from strong urge to ANSIfy.
2003-03-02 15:56:49 +00:00
Try to make Linux socket ioctls work. Up until now they've only *pretended* to work, but haven't really due to subtle differences in structs etc. This is still not perfect (some ioctls are still known not to work, while others haven't been tested at all), but it's enough to get Debian's ifconfig to produce relatively sane output. More work will be needed to get all ioctls (or at least a reasonable subset) working, and to support the Cisco Aironet config tool mentioned in the PR. PR: 26546 Submitted by: Doug Ambrisko <ambrisko@ambrisko.com>
2001-10-15 20:52:17 +00:00
case LINUX_SIOCGIFHWADDR:
Tweak the way we determine if an interface needs to have its name translated. Add some missing break statements in the socket ioctl switch. Check the return value from copyin() / copyout(). Fix some disorderings and misindentations. Support a couple more socket ioctls. Add missing break statements.
2001-10-20 00:01:26 +00:00
error = linux_gifhwaddr(ifp, (struct l_ifreq *)args->arg);
Try to make Linux socket ioctls work. Up until now they've only *pretended* to work, but haven't really due to subtle differences in structs etc. This is still not perfect (some ioctls are still known not to work, while others haven't been tested at all), but it's enough to get Debian's ifconfig to produce relatively sane output. More work will be needed to get all ioctls (or at least a reasonable subset) working, and to support the Cisco Aironet config tool mentioned in the PR. PR: 26546 Submitted by: Doug Ambrisko <ambrisko@ambrisko.com>
2001-10-15 20:52:17 +00:00
break;
- Handle mixer read ioctls correctly. They have the same group, number and argument size as their write counterparts and were handled as write ioctls. - Emulate some cdrom ioctls.
1999-04-29 04:37:57 +00:00
Tweak the way we determine if an interface needs to have its name translated. Add some missing break statements in the socket ioctl switch. Check the return value from copyin() / copyout(). Fix some disorderings and misindentations. Support a couple more socket ioctls. Add missing break statements.
2001-10-20 00:01:26 +00:00
case LINUX_SIOCSIFHWADDR:
error = ENOIOCTL;
break;
Clean up whitespace, s/register //, refrain from strong urge to ANSIfy.
2003-03-02 15:56:49 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SIOCADDMULTI:
args->cmd = SIOCADDMULTI;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = sys_ioctl(td, (struct ioctl_args *)args);
Try to make Linux socket ioctls work. Up until now they've only *pretended* to work, but haven't really due to subtle differences in structs etc. This is still not perfect (some ioctls are still known not to work, while others haven't been tested at all), but it's enough to get Debian's ifconfig to produce relatively sane output. More work will be needed to get all ioctls (or at least a reasonable subset) working, and to support the Cisco Aironet config tool mentioned in the PR. PR: 26546 Submitted by: Doug Ambrisko <ambrisko@ambrisko.com>
2001-10-15 20:52:17 +00:00
break;
- Handle mixer read ioctls correctly. They have the same group, number and argument size as their write counterparts and were handled as write ioctls. - Emulate some cdrom ioctls.
1999-04-29 04:37:57 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
case LINUX_SIOCDELMULTI:
args->cmd = SIOCDELMULTI;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = sys_ioctl(td, (struct ioctl_args *)args);
Try to make Linux socket ioctls work. Up until now they've only *pretended* to work, but haven't really due to subtle differences in structs etc. This is still not perfect (some ioctls are still known not to work, while others haven't been tested at all), but it's enough to get Debian's ifconfig to produce relatively sane output. More work will be needed to get all ioctls (or at least a reasonable subset) working, and to support the Cisco Aironet config tool mentioned in the PR. PR: 26546 Submitted by: Doug Ambrisko <ambrisko@ambrisko.com>
2001-10-15 20:52:17 +00:00
break;
Add support for the "device private" ioctls soon to be used by the an driver. Also slightly change the name translation policy - only rename interfaces that have the IFF_BROADCAST flag set. This is not perfect, but is closer to how Linux names network interfaces.
2001-10-19 01:38:10 +00:00
Implement LINUX_SIOCGIFCOUNT and LINUX_SIOCGIFINDEX/LINUX_SIOGIFINDEX. LINUX_SIOCGIFCOUNT just returns 0 since it is not implemented in the Linux 2.6.16. LINUX_SIOCGIFINDEX/LINUX_SIOGIFINDEX are mapped to the FreeBSD native SIOCGIFINDEX. Tested by: Peter Kostouros <kpeter@melbpc.org.au> Reviewed by: brooks, rpaulo (on net@) Submitted by: rdivacky MFC after: 1 week
2007-11-07 16:42:52 +00:00
case LINUX_SIOCGIFINDEX:
args->cmd = SIOCGIFINDEX;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = sys_ioctl(td, (struct ioctl_args *)args);
Implement LINUX_SIOCGIFCOUNT and LINUX_SIOCGIFINDEX/LINUX_SIOGIFINDEX. LINUX_SIOCGIFCOUNT just returns 0 since it is not implemented in the Linux 2.6.16. LINUX_SIOCGIFINDEX/LINUX_SIOGIFINDEX are mapped to the FreeBSD native SIOCGIFINDEX. Tested by: Peter Kostouros <kpeter@melbpc.org.au> Reviewed by: brooks, rpaulo (on net@) Submitted by: rdivacky MFC after: 1 week
2007-11-07 16:42:52 +00:00
break;
case LINUX_SIOCGIFCOUNT:
error = 0;
break;
Add support for the "device private" ioctls soon to be used by the an driver. Also slightly change the name translation policy - only rename interfaces that have the IFF_BROADCAST flag set. This is not perfect, but is closer to how Linux names network interfaces.
2001-10-19 01:38:10 +00:00
/*
* XXX This is slightly bogus, but these ioctls are currently
* XXX only used by the aironet (if_an) network driver.
*/
Tweak the way we determine if an interface needs to have its name translated. Add some missing break statements in the socket ioctl switch. Check the return value from copyin() / copyout(). Fix some disorderings and misindentations. Support a couple more socket ioctls. Add missing break statements.
2001-10-20 00:01:26 +00:00
case LINUX_SIOCDEVPRIVATE:
args->cmd = SIOCGPRIVATE_0;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = sys_ioctl(td, (struct ioctl_args *)args);
Tweak the way we determine if an interface needs to have its name translated. Add some missing break statements in the socket ioctl switch. Check the return value from copyin() / copyout(). Fix some disorderings and misindentations. Support a couple more socket ioctls. Add missing break statements.
2001-10-20 00:01:26 +00:00
break;
Clean up whitespace, s/register //, refrain from strong urge to ANSIfy.
2003-03-02 15:56:49 +00:00
Tweak the way we determine if an interface needs to have its name translated. Add some missing break statements in the socket ioctl switch. Check the return value from copyin() / copyout(). Fix some disorderings and misindentations. Support a couple more socket ioctls. Add missing break statements.
2001-10-20 00:01:26 +00:00
case LINUX_SIOCDEVPRIVATE+1:
args->cmd = SIOCGPRIVATE_1;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = sys_ioctl(td, (struct ioctl_args *)args);
Tweak the way we determine if an interface needs to have its name translated. Add some missing break statements in the socket ioctl switch. Check the return value from copyin() / copyout(). Fix some disorderings and misindentations. Support a couple more socket ioctls. Add missing break statements.
2001-10-20 00:01:26 +00:00
break;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
}
- Handle mixer read ioctls correctly. They have the same group, number and argument size as their write counterparts and were handled as write ioctls. - Emulate some cdrom ioctls.
1999-04-29 04:37:57 +00:00
Try to make Linux socket ioctls work. Up until now they've only *pretended* to work, but haven't really due to subtle differences in structs etc. This is still not perfect (some ioctls are still known not to work, while others haven't been tested at all), but it's enough to get Debian's ifconfig to produce relatively sane output. More work will be needed to get all ioctls (or at least a reasonable subset) working, and to support the Cisco Aironet config tool mentioned in the PR. PR: 26546 Submitted by: Doug Ambrisko <ambrisko@ambrisko.com>
2001-10-15 20:52:17 +00:00
if (ifp != NULL)
/* restore the original interface name */
More caddr_t removal, in conjunction with copy{in,out}(9) this time. Also clean up some egregious casts and incorrect use of sizeof.
2003-03-03 09:14:26 +00:00
copyout(lifname, (void *)args->arg, LINUX_IFNAMSIZ);
Try to make Linux socket ioctls work. Up until now they've only *pretended* to work, but haven't really due to subtle differences in structs etc. This is still not perfect (some ioctls are still known not to work, while others haven't been tested at all), but it's enough to get Debian's ifconfig to produce relatively sane output. More work will be needed to get all ioctls (or at least a reasonable subset) working, and to support the Cisco Aironet config tool mentioned in the PR. PR: 26546 Submitted by: Doug Ambrisko <ambrisko@ambrisko.com>
2001-10-15 20:52:17 +00:00
return (error);
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
}
- Handle mixer read ioctls correctly. They have the same group, number and argument size as their write counterparts and were handled as write ioctls. - Emulate some cdrom ioctls.
1999-04-29 04:37:57 +00:00
Add support for the "device private" ioctls soon to be used by the an driver. Also slightly change the name translation policy - only rename interfaces that have the IFF_BROADCAST flag set. This is not perfect, but is closer to how Linux names network interfaces.
2001-10-19 01:38:10 +00:00
/*
* Device private ioctl handler
*/
static int
linux_ioctl_private(struct thread *td, struct linux_ioctl_args *args)
{
struct file *fp;
Deal with a few issues that cropped up following the recent changes to the code for translating socket and private ioctls: - Only perform socket ioctl translation if the file descriptor is a socket. - Treat socket ioctls on non-sockets specially, and for now assume that these are directed at a tap/vmnet device, so translate the ioctl numbers as appropriate (the way if_tap abuses some socket ioctls to pass non-ifreq data is utterly bogus, but this is how VMware on FreeBSD has always "worked"; I will deal with this later). - Add (untested) support for translating SIOCSIFADDR. - In all cases where we fail to translate an ioctl, return ENOIOCTL so that other handlers have a chance to do the translation. This should fix the "/dev/vmnet1: Invalid argument" errors that users of VMware were experiencing, though I have only verified this on RELENG_4. Submitted by: des (mostly) MFC after: 3 days
2001-11-19 15:43:50 +00:00
int error, type;
Add support for the "device private" ioctls soon to be used by the an driver. Also slightly change the name translation policy - only rename interfaces that have the IFF_BROADCAST flag set. This is not perfect, but is closer to how Linux names network interfaces.
2001-10-19 01:38:10 +00:00
Eliminate the overhead of gratuitous repeated reinitialization of cap_rights - Add macros to allow preinitialization of cap_rights_t. - Convert most commonly used code paths to use preinitialized cap_rights_t. A 3.6% speedup in fstat was measured with this change. Reported by: mjg Reviewed by: oshogbo Approved by: sbruno MFC after: 1 month
2018-05-09 18:47:24 +00:00
error = fget(td, args->fd, &cap_ioctl_rights, &fp);
Change the cap_rights_t type from uint64_t to a structure that we can extend in the future in a backward compatible (API and ABI) way. The cap_rights_t represents capability rights. We used to use one bit to represent one right, but we are running out of spare bits. Currently the new structure provides place for 114 rights (so 50 more than the previous cap_rights_t), but it is possible to grow the structure to hold at least 285 rights, although we can make it even larger if 285 rights won't be enough. The structure definition looks like this: struct cap_rights { uint64_t cr_rights[CAP_RIGHTS_VERSION + 2]; }; The initial CAP_RIGHTS_VERSION is 0. The top two bits in the first element of the cr_rights[] array contain total number of elements in the array - 2. This means if those two bits are equal to 0, we have 2 array elements. The top two bits in all remaining array elements should be 0. The next five bits in all array elements contain array index. Only one bit is used and bit position in this five-bits range defines array index. This means there can be at most five array elements in the future. To define new right the CAPRIGHT() macro must be used. The macro takes two arguments - an array index and a bit to set, eg. #define CAP_PDKILL CAPRIGHT(1, 0x0000000000000800ULL) We still support aliases that combine few rights, but the rights have to belong to the same array element, eg: #define CAP_LOOKUP CAPRIGHT(0, 0x0000000000000400ULL) #define CAP_FCHMOD CAPRIGHT(0, 0x0000000000002000ULL) #define CAP_FCHMODAT (CAP_FCHMOD | CAP_LOOKUP) There is new API to manage the new cap_rights_t structure: cap_rights_t *cap_rights_init(cap_rights_t *rights, ...); void cap_rights_set(cap_rights_t *rights, ...); void cap_rights_clear(cap_rights_t *rights, ...); bool cap_rights_is_set(const cap_rights_t *rights, ...); bool cap_rights_is_valid(const cap_rights_t *rights); void cap_rights_merge(cap_rights_t *dst, const cap_rights_t *src); void cap_rights_remove(cap_rights_t *dst, const cap_rights_t *src); bool cap_rights_contains(const cap_rights_t *big, const cap_rights_t *little); Capability rights to the cap_rights_init(), cap_rights_set(), cap_rights_clear() and cap_rights_is_set() functions are provided by separating them with commas, eg: cap_rights_t rights; cap_rights_init(&rights, CAP_READ, CAP_WRITE, CAP_FSTAT); There is no need to terminate the list of rights, as those functions are actually macros that take care of the termination, eg: #define cap_rights_set(rights, ...) \ __cap_rights_set((rights), __VA_ARGS__, 0ULL) void __cap_rights_set(cap_rights_t *rights, ...); Thanks to using one bit as an array index we can assert in those functions that there are no two rights belonging to different array elements provided together. For example this is illegal and will be detected, because CAP_LOOKUP belongs to element 0 and CAP_PDKILL to element 1: cap_rights_init(&rights, CAP_LOOKUP | CAP_PDKILL); Providing several rights that belongs to the same array's element this way is correct, but is not advised. It should only be used for aliases definition. This commit also breaks compatibility with some existing Capsicum system calls, but I see no other way to do that. This should be fine as Capsicum is still experimental and this change is not going to 9.x. Sponsored by: The FreeBSD Foundation
2013-09-05 00:09:56 +00:00
if (error != 0)
Deal with a few issues that cropped up following the recent changes to the code for translating socket and private ioctls: - Only perform socket ioctl translation if the file descriptor is a socket. - Treat socket ioctls on non-sockets specially, and for now assume that these are directed at a tap/vmnet device, so translate the ioctl numbers as appropriate (the way if_tap abuses some socket ioctls to pass non-ifreq data is utterly bogus, but this is how VMware on FreeBSD has always "worked"; I will deal with this later). - Add (untested) support for translating SIOCSIFADDR. - In all cases where we fail to translate an ioctl, return ENOIOCTL so that other handlers have a chance to do the translation. This should fix the "/dev/vmnet1: Invalid argument" errors that users of VMware were experiencing, though I have only verified this on RELENG_4. Submitted by: des (mostly) MFC after: 3 days
2001-11-19 15:43:50 +00:00
return (error);
type = fp->f_type;
fdrop(fp, td);
Add support for the "device private" ioctls soon to be used by the an driver. Also slightly change the name translation policy - only rename interfaces that have the IFF_BROADCAST flag set. This is not perfect, but is closer to how Linux names network interfaces.
2001-10-19 01:38:10 +00:00
if (type == DTYPE_SOCKET)
return (linux_ioctl_socket(td, args));
Deal with a few issues that cropped up following the recent changes to the code for translating socket and private ioctls: - Only perform socket ioctl translation if the file descriptor is a socket. - Treat socket ioctls on non-sockets specially, and for now assume that these are directed at a tap/vmnet device, so translate the ioctl numbers as appropriate (the way if_tap abuses some socket ioctls to pass non-ifreq data is utterly bogus, but this is how VMware on FreeBSD has always "worked"; I will deal with this later). - Add (untested) support for translating SIOCSIFADDR. - In all cases where we fail to translate an ioctl, return ENOIOCTL so that other handlers have a chance to do the translation. This should fix the "/dev/vmnet1: Invalid argument" errors that users of VMware were experiencing, though I have only verified this on RELENG_4. Submitted by: des (mostly) MFC after: 3 days
2001-11-19 15:43:50 +00:00
return (ENOIOCTL);
}
Add an ioctl handler for the DRM. This removes the need for the DRM_LINUX option, which has been a source of frustration for many users.
2003-04-24 23:36:35 +00:00
/*
* DRM ioctl handler (sys/dev/drm)
*/
static int
linux_ioctl_drm(struct thread *td, struct linux_ioctl_args *args)
{
args->cmd = SETDIR(args->cmd);
Linuxulator: apply style(9) to return Sponsored by: Turing Robotic Industries Inc.
2018-03-12 15:35:24 +00:00
return (sys_ioctl(td, (struct ioctl_args *)args));
Add an ioctl handler for the DRM. This removes the need for the DRM_LINUX option, which has been a source of frustration for many users.
2003-04-24 23:36:35 +00:00
}
Overhaul CAM SG driver IOCTL interfaces. Make it really work for native FreeBSD programs. Before this it was broken for years due to different number of pointer dereferences in Linux and FreeBSD IOCTL paths, permanently returning errors to FreeBSD programs. This change breaks the driver FreeBSD IOCTL ABI, making it more strict, but since it was not working any way -- who bother. Add shims for 32-bit programs on 64-bit host, translating the argument of the SG_IO IOCTL for both FreeBSD and Linux ABIs. With this change I was able to run 32-bit Linux sg3_utils tools and simple 32 and 64-bit FreeBSD test tools on both 32 and 64-bit FreeBSD systems. MFC after: 1 month
2014-06-02 19:53:53 +00:00
#ifdef COMPAT_LINUX32
Add the CAM 'SG' peripheral device. This device implements a subset of the Linux SCSI SG passthrough device API. The intention is to allow for both running of Linux apps that want to talk to /dev/sg* nodes, and to facilitate porting of apps from Linux to FreeBSD. As such, both native and linuxolator entry points and definitions are provided. Caveats: - This does not support the procfs and sysfs nodes that the Linux SG driver provides. Some Linux apps may rely on these for operation, others may only use them for informational purposes. - More ioctls need to be implemented. - Linux uses a naming scheme of "sg[a-z]" for devices, while FreeBSD uses a scheme of "sg[0-9]". Devfs aliasis (symlinks) are automatically created to link the two together. However, tools like camcontrol only see the native names. - Some operations were originally designed to return byte counts or other data directly as the syscall return value. The linuxolator doesn't appear to support this well, so this driver just punts for these cases. Now that the driver is in place, others are welcome to add missing functionality. Thanks to Roman Divacky for pushing this work along.
2007-04-07 19:40:58 +00:00
static int
Overhaul CAM SG driver IOCTL interfaces. Make it really work for native FreeBSD programs. Before this it was broken for years due to different number of pointer dereferences in Linux and FreeBSD IOCTL paths, permanently returning errors to FreeBSD programs. This change breaks the driver FreeBSD IOCTL ABI, making it more strict, but since it was not working any way -- who bother. Add shims for 32-bit programs on 64-bit host, translating the argument of the SG_IO IOCTL for both FreeBSD and Linux ABIs. With this change I was able to run 32-bit Linux sg3_utils tools and simple 32 and 64-bit FreeBSD test tools on both 32 and 64-bit FreeBSD systems. MFC after: 1 month
2014-06-02 19:53:53 +00:00
linux_ioctl_sg_io(struct thread *td, struct linux_ioctl_args *args)
Add the CAM 'SG' peripheral device. This device implements a subset of the Linux SCSI SG passthrough device API. The intention is to allow for both running of Linux apps that want to talk to /dev/sg* nodes, and to facilitate porting of apps from Linux to FreeBSD. As such, both native and linuxolator entry points and definitions are provided. Caveats: - This does not support the procfs and sysfs nodes that the Linux SG driver provides. Some Linux apps may rely on these for operation, others may only use them for informational purposes. - More ioctls need to be implemented. - Linux uses a naming scheme of "sg[a-z]" for devices, while FreeBSD uses a scheme of "sg[0-9]". Devfs aliasis (symlinks) are automatically created to link the two together. However, tools like camcontrol only see the native names. - Some operations were originally designed to return byte counts or other data directly as the syscall return value. The linuxolator doesn't appear to support this well, so this driver just punts for these cases. Now that the driver is in place, others are welcome to add missing functionality. Thanks to Roman Divacky for pushing this work along.
2007-04-07 19:40:58 +00:00
{
Overhaul CAM SG driver IOCTL interfaces. Make it really work for native FreeBSD programs. Before this it was broken for years due to different number of pointer dereferences in Linux and FreeBSD IOCTL paths, permanently returning errors to FreeBSD programs. This change breaks the driver FreeBSD IOCTL ABI, making it more strict, but since it was not working any way -- who bother. Add shims for 32-bit programs on 64-bit host, translating the argument of the SG_IO IOCTL for both FreeBSD and Linux ABIs. With this change I was able to run 32-bit Linux sg3_utils tools and simple 32 and 64-bit FreeBSD test tools on both 32 and 64-bit FreeBSD systems. MFC after: 1 month
2014-06-02 19:53:53 +00:00
struct sg_io_hdr io;
struct sg_io_hdr32 io32;
Add the CAM 'SG' peripheral device. This device implements a subset of the Linux SCSI SG passthrough device API. The intention is to allow for both running of Linux apps that want to talk to /dev/sg* nodes, and to facilitate porting of apps from Linux to FreeBSD. As such, both native and linuxolator entry points and definitions are provided. Caveats: - This does not support the procfs and sysfs nodes that the Linux SG driver provides. Some Linux apps may rely on these for operation, others may only use them for informational purposes. - More ioctls need to be implemented. - Linux uses a naming scheme of "sg[a-z]" for devices, while FreeBSD uses a scheme of "sg[0-9]". Devfs aliasis (symlinks) are automatically created to link the two together. However, tools like camcontrol only see the native names. - Some operations were originally designed to return byte counts or other data directly as the syscall return value. The linuxolator doesn't appear to support this well, so this driver just punts for these cases. Now that the driver is in place, others are welcome to add missing functionality. Thanks to Roman Divacky for pushing this work along.
2007-04-07 19:40:58 +00:00
struct file *fp;
int error;
Eliminate the overhead of gratuitous repeated reinitialization of cap_rights - Add macros to allow preinitialization of cap_rights_t. - Convert most commonly used code paths to use preinitialized cap_rights_t. A 3.6% speedup in fstat was measured with this change. Reported by: mjg Reviewed by: oshogbo Approved by: sbruno MFC after: 1 month
2018-05-09 18:47:24 +00:00
error = fget(td, args->fd, &cap_ioctl_rights, &fp);
Change the cap_rights_t type from uint64_t to a structure that we can extend in the future in a backward compatible (API and ABI) way. The cap_rights_t represents capability rights. We used to use one bit to represent one right, but we are running out of spare bits. Currently the new structure provides place for 114 rights (so 50 more than the previous cap_rights_t), but it is possible to grow the structure to hold at least 285 rights, although we can make it even larger if 285 rights won't be enough. The structure definition looks like this: struct cap_rights { uint64_t cr_rights[CAP_RIGHTS_VERSION + 2]; }; The initial CAP_RIGHTS_VERSION is 0. The top two bits in the first element of the cr_rights[] array contain total number of elements in the array - 2. This means if those two bits are equal to 0, we have 2 array elements. The top two bits in all remaining array elements should be 0. The next five bits in all array elements contain array index. Only one bit is used and bit position in this five-bits range defines array index. This means there can be at most five array elements in the future. To define new right the CAPRIGHT() macro must be used. The macro takes two arguments - an array index and a bit to set, eg. #define CAP_PDKILL CAPRIGHT(1, 0x0000000000000800ULL) We still support aliases that combine few rights, but the rights have to belong to the same array element, eg: #define CAP_LOOKUP CAPRIGHT(0, 0x0000000000000400ULL) #define CAP_FCHMOD CAPRIGHT(0, 0x0000000000002000ULL) #define CAP_FCHMODAT (CAP_FCHMOD | CAP_LOOKUP) There is new API to manage the new cap_rights_t structure: cap_rights_t *cap_rights_init(cap_rights_t *rights, ...); void cap_rights_set(cap_rights_t *rights, ...); void cap_rights_clear(cap_rights_t *rights, ...); bool cap_rights_is_set(const cap_rights_t *rights, ...); bool cap_rights_is_valid(const cap_rights_t *rights); void cap_rights_merge(cap_rights_t *dst, const cap_rights_t *src); void cap_rights_remove(cap_rights_t *dst, const cap_rights_t *src); bool cap_rights_contains(const cap_rights_t *big, const cap_rights_t *little); Capability rights to the cap_rights_init(), cap_rights_set(), cap_rights_clear() and cap_rights_is_set() functions are provided by separating them with commas, eg: cap_rights_t rights; cap_rights_init(&rights, CAP_READ, CAP_WRITE, CAP_FSTAT); There is no need to terminate the list of rights, as those functions are actually macros that take care of the termination, eg: #define cap_rights_set(rights, ...) \ __cap_rights_set((rights), __VA_ARGS__, 0ULL) void __cap_rights_set(cap_rights_t *rights, ...); Thanks to using one bit as an array index we can assert in those functions that there are no two rights belonging to different array elements provided together. For example this is illegal and will be detected, because CAP_LOOKUP belongs to element 0 and CAP_PDKILL to element 1: cap_rights_init(&rights, CAP_LOOKUP | CAP_PDKILL); Providing several rights that belongs to the same array's element this way is correct, but is not advised. It should only be used for aliases definition. This commit also breaks compatibility with some existing Capsicum system calls, but I see no other way to do that. This should be fine as Capsicum is still experimental and this change is not going to 9.x. Sponsored by: The FreeBSD Foundation
2013-09-05 00:09:56 +00:00
if (error != 0) {
Add the CAM 'SG' peripheral device. This device implements a subset of the Linux SCSI SG passthrough device API. The intention is to allow for both running of Linux apps that want to talk to /dev/sg* nodes, and to facilitate porting of apps from Linux to FreeBSD. As such, both native and linuxolator entry points and definitions are provided. Caveats: - This does not support the procfs and sysfs nodes that the Linux SG driver provides. Some Linux apps may rely on these for operation, others may only use them for informational purposes. - More ioctls need to be implemented. - Linux uses a naming scheme of "sg[a-z]" for devices, while FreeBSD uses a scheme of "sg[0-9]". Devfs aliasis (symlinks) are automatically created to link the two together. However, tools like camcontrol only see the native names. - Some operations were originally designed to return byte counts or other data directly as the syscall return value. The linuxolator doesn't appear to support this well, so this driver just punts for these cases. Now that the driver is in place, others are welcome to add missing functionality. Thanks to Roman Divacky for pushing this work along.
2007-04-07 19:40:58 +00:00
printf("sg_linux_ioctl: fget returned %d\n", error);
return (error);
}
Overhaul CAM SG driver IOCTL interfaces. Make it really work for native FreeBSD programs. Before this it was broken for years due to different number of pointer dereferences in Linux and FreeBSD IOCTL paths, permanently returning errors to FreeBSD programs. This change breaks the driver FreeBSD IOCTL ABI, making it more strict, but since it was not working any way -- who bother. Add shims for 32-bit programs on 64-bit host, translating the argument of the SG_IO IOCTL for both FreeBSD and Linux ABIs. With this change I was able to run 32-bit Linux sg3_utils tools and simple 32 and 64-bit FreeBSD test tools on both 32 and 64-bit FreeBSD systems. MFC after: 1 month
2014-06-02 19:53:53 +00:00
if ((error = copyin((void *)args->arg, &io32, sizeof(io32))) != 0)
goto out;
CP(io32, io, interface_id);
CP(io32, io, dxfer_direction);
CP(io32, io, cmd_len);
CP(io32, io, mx_sb_len);
CP(io32, io, iovec_count);
CP(io32, io, dxfer_len);
PTRIN_CP(io32, io, dxferp);
PTRIN_CP(io32, io, cmdp);
PTRIN_CP(io32, io, sbp);
CP(io32, io, timeout);
CP(io32, io, flags);
CP(io32, io, pack_id);
PTRIN_CP(io32, io, usr_ptr);
CP(io32, io, status);
CP(io32, io, masked_status);
CP(io32, io, msg_status);
CP(io32, io, sb_len_wr);
CP(io32, io, host_status);
CP(io32, io, driver_status);
CP(io32, io, resid);
CP(io32, io, duration);
CP(io32, io, info);
if ((error = fo_ioctl(fp, SG_IO, (caddr_t)&io, td->td_ucred, td)) != 0)
goto out;
CP(io, io32, interface_id);
CP(io, io32, dxfer_direction);
CP(io, io32, cmd_len);
CP(io, io32, mx_sb_len);
CP(io, io32, iovec_count);
CP(io, io32, dxfer_len);
PTROUT_CP(io, io32, dxferp);
PTROUT_CP(io, io32, cmdp);
PTROUT_CP(io, io32, sbp);
CP(io, io32, timeout);
CP(io, io32, flags);
CP(io, io32, pack_id);
PTROUT_CP(io, io32, usr_ptr);
CP(io, io32, status);
CP(io, io32, masked_status);
CP(io, io32, msg_status);
CP(io, io32, sb_len_wr);
CP(io, io32, host_status);
CP(io, io32, driver_status);
CP(io, io32, resid);
CP(io, io32, duration);
CP(io, io32, info);
error = copyout(&io32, (void *)args->arg, sizeof(io32));
out:
Add the CAM 'SG' peripheral device. This device implements a subset of the Linux SCSI SG passthrough device API. The intention is to allow for both running of Linux apps that want to talk to /dev/sg* nodes, and to facilitate porting of apps from Linux to FreeBSD. As such, both native and linuxolator entry points and definitions are provided. Caveats: - This does not support the procfs and sysfs nodes that the Linux SG driver provides. Some Linux apps may rely on these for operation, others may only use them for informational purposes. - More ioctls need to be implemented. - Linux uses a naming scheme of "sg[a-z]" for devices, while FreeBSD uses a scheme of "sg[0-9]". Devfs aliasis (symlinks) are automatically created to link the two together. However, tools like camcontrol only see the native names. - Some operations were originally designed to return byte counts or other data directly as the syscall return value. The linuxolator doesn't appear to support this well, so this driver just punts for these cases. Now that the driver is in place, others are welcome to add missing functionality. Thanks to Roman Divacky for pushing this work along.
2007-04-07 19:40:58 +00:00
fdrop(fp, td);
return (error);
}
Overhaul CAM SG driver IOCTL interfaces. Make it really work for native FreeBSD programs. Before this it was broken for years due to different number of pointer dereferences in Linux and FreeBSD IOCTL paths, permanently returning errors to FreeBSD programs. This change breaks the driver FreeBSD IOCTL ABI, making it more strict, but since it was not working any way -- who bother. Add shims for 32-bit programs on 64-bit host, translating the argument of the SG_IO IOCTL for both FreeBSD and Linux ABIs. With this change I was able to run 32-bit Linux sg3_utils tools and simple 32 and 64-bit FreeBSD test tools on both 32 and 64-bit FreeBSD systems. MFC after: 1 month
2014-06-02 19:53:53 +00:00
#endif
static int
linux_ioctl_sg(struct thread *td, struct linux_ioctl_args *args)
{
switch (args->cmd) {
case LINUX_SG_GET_VERSION_NUM:
args->cmd = SG_GET_VERSION_NUM;
break;
case LINUX_SG_SET_TIMEOUT:
args->cmd = SG_SET_TIMEOUT;
break;
case LINUX_SG_GET_TIMEOUT:
args->cmd = SG_GET_TIMEOUT;
break;
case LINUX_SG_IO:
args->cmd = SG_IO;
#ifdef COMPAT_LINUX32
return (linux_ioctl_sg_io(td, args));
#endif
break;
case LINUX_SG_GET_RESERVED_SIZE:
args->cmd = SG_GET_RESERVED_SIZE;
break;
case LINUX_SG_GET_SCSI_ID:
args->cmd = SG_GET_SCSI_ID;
break;
- Add support for SG_GET_SG_TABLESIZE IOCTL to report that we don't support scatter/gather lists. - Return error for still unsupported SG 3.x API read/write calls. MFC after: 1 month
2014-06-04 12:05:47 +00:00
case LINUX_SG_GET_SG_TABLESIZE:
args->cmd = SG_GET_SG_TABLESIZE;
break;
Overhaul CAM SG driver IOCTL interfaces. Make it really work for native FreeBSD programs. Before this it was broken for years due to different number of pointer dereferences in Linux and FreeBSD IOCTL paths, permanently returning errors to FreeBSD programs. This change breaks the driver FreeBSD IOCTL ABI, making it more strict, but since it was not working any way -- who bother. Add shims for 32-bit programs on 64-bit host, translating the argument of the SG_IO IOCTL for both FreeBSD and Linux ABIs. With this change I was able to run 32-bit Linux sg3_utils tools and simple 32 and 64-bit FreeBSD test tools on both 32 and 64-bit FreeBSD systems. MFC after: 1 month
2014-06-02 19:53:53 +00:00
default:
return (ENODEV);
}
return (sys_ioctl(td, (struct ioctl_args *)args));
}
Add the CAM 'SG' peripheral device. This device implements a subset of the Linux SCSI SG passthrough device API. The intention is to allow for both running of Linux apps that want to talk to /dev/sg* nodes, and to facilitate porting of apps from Linux to FreeBSD. As such, both native and linuxolator entry points and definitions are provided. Caveats: - This does not support the procfs and sysfs nodes that the Linux SG driver provides. Some Linux apps may rely on these for operation, others may only use them for informational purposes. - More ioctls need to be implemented. - Linux uses a naming scheme of "sg[a-z]" for devices, while FreeBSD uses a scheme of "sg[0-9]". Devfs aliasis (symlinks) are automatically created to link the two together. However, tools like camcontrol only see the native names. - Some operations were originally designed to return byte counts or other data directly as the syscall return value. The linuxolator doesn't appear to support this well, so this driver just punts for these cases. Now that the driver is in place, others are welcome to add missing functionality. Thanks to Roman Divacky for pushing this work along.
2007-04-07 19:40:58 +00:00
This is v4l support for the linuxulator. This allows to access FreeBSD native devices which support the v4l API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd driver. Not tested is firmware upload, framebuffer stuff and video tuner stuff due to lack of hardware. The clipping part (VIDIOCSWIN) needs a little bit of further work (partly in progress, but can not be tested due to lack of a suitable device). The submitter tested this sucessfully with Skype and flash apps on amd64 and i386 with the multimedia/pwcbsd driver. Submitted by: J.R. Oldroyd <fbsd@opal.com>
2009-12-04 21:06:54 +00:00
/*
* Video4Linux (V4L) ioctl handler
*/
static int
linux_to_bsd_v4l_tuner(struct l_video_tuner *lvt, struct video_tuner *vt)
{
vt->tuner = lvt->tuner;
strlcpy(vt->name, lvt->name, LINUX_VIDEO_TUNER_NAME_SIZE);
vt->rangelow = lvt->rangelow; /* possible long size conversion */
vt->rangehigh = lvt->rangehigh; /* possible long size conversion */
vt->flags = lvt->flags;
vt->mode = lvt->mode;
vt->signal = lvt->signal;
return (0);
}
static int
bsd_to_linux_v4l_tuner(struct video_tuner *vt, struct l_video_tuner *lvt)
{
lvt->tuner = vt->tuner;
strlcpy(lvt->name, vt->name, LINUX_VIDEO_TUNER_NAME_SIZE);
lvt->rangelow = vt->rangelow; /* possible long size conversion */
lvt->rangehigh = vt->rangehigh; /* possible long size conversion */
lvt->flags = vt->flags;
lvt->mode = vt->mode;
lvt->signal = vt->signal;
return (0);
}
- #ifdef out the cliplist part, skype seems like using an uninitialized variable and can cause problems, without the cliplist handling it works without problems - improve the cliplist error handling - fix VIDIOCGTUNER and VIDIOCSMICROCODE (still no hardware available to test) Submitted by: J.R. Oldroyd <jr@opal.com> X-MFC after: soon (together with all the v4l stuff)
2010-05-03 14:19:58 +00:00
#ifdef COMPAT_LINUX_V4L_CLIPLIST
This is v4l support for the linuxulator. This allows to access FreeBSD native devices which support the v4l API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd driver. Not tested is firmware upload, framebuffer stuff and video tuner stuff due to lack of hardware. The clipping part (VIDIOCSWIN) needs a little bit of further work (partly in progress, but can not be tested due to lack of a suitable device). The submitter tested this sucessfully with Skype and flash apps on amd64 and i386 with the multimedia/pwcbsd driver. Submitted by: J.R. Oldroyd <fbsd@opal.com>
2009-12-04 21:06:54 +00:00
static int
linux_to_bsd_v4l_clip(struct l_video_clip *lvc, struct video_clip *vc)
{
vc->x = lvc->x;
vc->y = lvc->y;
vc->width = lvc->width;
vc->height = lvc->height;
vc->next = PTRIN(lvc->next); /* possible pointer size conversion */
return (0);
}
- #ifdef out the cliplist part, skype seems like using an uninitialized variable and can cause problems, without the cliplist handling it works without problems - improve the cliplist error handling - fix VIDIOCGTUNER and VIDIOCSMICROCODE (still no hardware available to test) Submitted by: J.R. Oldroyd <jr@opal.com> X-MFC after: soon (together with all the v4l stuff)
2010-05-03 14:19:58 +00:00
#endif
This is v4l support for the linuxulator. This allows to access FreeBSD native devices which support the v4l API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd driver. Not tested is firmware upload, framebuffer stuff and video tuner stuff due to lack of hardware. The clipping part (VIDIOCSWIN) needs a little bit of further work (partly in progress, but can not be tested due to lack of a suitable device). The submitter tested this sucessfully with Skype and flash apps on amd64 and i386 with the multimedia/pwcbsd driver. Submitted by: J.R. Oldroyd <fbsd@opal.com>
2009-12-04 21:06:54 +00:00
static int
linux_to_bsd_v4l_window(struct l_video_window *lvw, struct video_window *vw)
{
vw->x = lvw->x;
vw->y = lvw->y;
vw->width = lvw->width;
vw->height = lvw->height;
vw->chromakey = lvw->chromakey;
vw->flags = lvw->flags;
vw->clips = PTRIN(lvw->clips); /* possible pointer size conversion */
vw->clipcount = lvw->clipcount;
return (0);
}
static int
bsd_to_linux_v4l_window(struct video_window *vw, struct l_video_window *lvw)
{
linuxulator: fix stack memory disclosure in linux_ioctl_v4l admbugs: 765 Reported by: Vlad Tsyrklevich <vlad@tsyrklevich.net> Reviewed by: andrew MFC after: 1 day Security: Kernel stack memory disclosure Sponsored by: The FreeBSD Foundation
2019-01-21 16:19:02 +00:00
memset(lvw, 0, sizeof(*lvw));
This is v4l support for the linuxulator. This allows to access FreeBSD native devices which support the v4l API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd driver. Not tested is firmware upload, framebuffer stuff and video tuner stuff due to lack of hardware. The clipping part (VIDIOCSWIN) needs a little bit of further work (partly in progress, but can not be tested due to lack of a suitable device). The submitter tested this sucessfully with Skype and flash apps on amd64 and i386 with the multimedia/pwcbsd driver. Submitted by: J.R. Oldroyd <fbsd@opal.com>
2009-12-04 21:06:54 +00:00
lvw->x = vw->x;
lvw->y = vw->y;
lvw->width = vw->width;
lvw->height = vw->height;
lvw->chromakey = vw->chromakey;
lvw->flags = vw->flags;
lvw->clips = PTROUT(vw->clips); /* possible pointer size conversion */
lvw->clipcount = vw->clipcount;
return (0);
}
static int
linux_to_bsd_v4l_buffer(struct l_video_buffer *lvb, struct video_buffer *vb)
{
vb->base = PTRIN(lvb->base); /* possible pointer size conversion */
vb->height = lvb->height;
vb->width = lvb->width;
vb->depth = lvb->depth;
vb->bytesperline = lvb->bytesperline;
return (0);
}
static int
bsd_to_linux_v4l_buffer(struct video_buffer *vb, struct l_video_buffer *lvb)
{
lvb->base = PTROUT(vb->base); /* possible pointer size conversion */
lvb->height = vb->height;
lvb->width = vb->width;
lvb->depth = vb->depth;
lvb->bytesperline = vb->bytesperline;
return (0);
}
static int
linux_to_bsd_v4l_code(struct l_video_code *lvc, struct video_code *vc)
{
strlcpy(vc->loadwhat, lvc->loadwhat, LINUX_VIDEO_CODE_LOADWHAT_SIZE);
vc->datasize = lvc->datasize;
vc->data = PTRIN(lvc->data); /* possible pointer size conversion */
return (0);
}
- #ifdef out the cliplist part, skype seems like using an uninitialized variable and can cause problems, without the cliplist handling it works without problems - improve the cliplist error handling - fix VIDIOCGTUNER and VIDIOCSMICROCODE (still no hardware available to test) Submitted by: J.R. Oldroyd <jr@opal.com> X-MFC after: soon (together with all the v4l stuff)
2010-05-03 14:19:58 +00:00
#ifdef COMPAT_LINUX_V4L_CLIPLIST
This is v4l support for the linuxulator. This allows to access FreeBSD native devices which support the v4l API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd driver. Not tested is firmware upload, framebuffer stuff and video tuner stuff due to lack of hardware. The clipping part (VIDIOCSWIN) needs a little bit of further work (partly in progress, but can not be tested due to lack of a suitable device). The submitter tested this sucessfully with Skype and flash apps on amd64 and i386 with the multimedia/pwcbsd driver. Submitted by: J.R. Oldroyd <fbsd@opal.com>
2009-12-04 21:06:54 +00:00
static int
This is v4l support for the linuxulator. This allows to access FreeBSD native devices which support the v4l API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd driver. Not tested is firmware upload, framebuffer stuff and video tuner stuff due to lack of hardware. The clipping part (VIDIOCSWIN) needs a little bit of further work (partly in progress, but can not be tested due to lack of a suitable device). The submitter tested this sucessfully with Skype and flash apps on amd64 and i386 with the multimedia/pwcbsd driver. Submitted by: J.R. Oldroyd <fbsd@opal.com>
2010-01-15 14:58:19 +00:00
linux_v4l_clip_copy(void *lvc, struct video_clip **ppvc)
This is v4l support for the linuxulator. This allows to access FreeBSD native devices which support the v4l API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd driver. Not tested is firmware upload, framebuffer stuff and video tuner stuff due to lack of hardware. The clipping part (VIDIOCSWIN) needs a little bit of further work (partly in progress, but can not be tested due to lack of a suitable device). The submitter tested this sucessfully with Skype and flash apps on amd64 and i386 with the multimedia/pwcbsd driver. Submitted by: J.R. Oldroyd <fbsd@opal.com>
2009-12-04 21:06:54 +00:00
{
This is v4l support for the linuxulator. This allows to access FreeBSD native devices which support the v4l API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd driver. Not tested is firmware upload, framebuffer stuff and video tuner stuff due to lack of hardware. The clipping part (VIDIOCSWIN) needs a little bit of further work (partly in progress, but can not be tested due to lack of a suitable device). The submitter tested this sucessfully with Skype and flash apps on amd64 and i386 with the multimedia/pwcbsd driver. Submitted by: J.R. Oldroyd <fbsd@opal.com>
2010-01-15 14:58:19 +00:00
int error;
This is v4l support for the linuxulator. This allows to access FreeBSD native devices which support the v4l API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd driver. Not tested is firmware upload, framebuffer stuff and video tuner stuff due to lack of hardware. The clipping part (VIDIOCSWIN) needs a little bit of further work (partly in progress, but can not be tested due to lack of a suitable device). The submitter tested this sucessfully with Skype and flash apps on amd64 and i386 with the multimedia/pwcbsd driver. Submitted by: J.R. Oldroyd <fbsd@opal.com>
2009-12-04 21:06:54 +00:00
struct video_clip vclip;
struct l_video_clip l_vclip;
This is v4l support for the linuxulator. This allows to access FreeBSD native devices which support the v4l API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd driver. Not tested is firmware upload, framebuffer stuff and video tuner stuff due to lack of hardware. The clipping part (VIDIOCSWIN) needs a little bit of further work (partly in progress, but can not be tested due to lack of a suitable device). The submitter tested this sucessfully with Skype and flash apps on amd64 and i386 with the multimedia/pwcbsd driver. Submitted by: J.R. Oldroyd <fbsd@opal.com>
2010-01-15 14:58:19 +00:00
error = copyin(lvc, &l_vclip, sizeof(l_vclip));
if (error) return (error);
linux_to_bsd_v4l_clip(&l_vclip, &vclip);
/* XXX: If there can be no concurrency: s/M_NOWAIT/M_WAITOK/ */
if ((*ppvc = malloc(sizeof(**ppvc), M_LINUX, M_NOWAIT)) == NULL)
Correct proper nouns in the Linuxulator - Capitalize Linux - Spell FreeBSD out in full - Address some style(9) on changed lines Sponsored by: Turing Robotic Industries Inc.
2018-02-22 02:24:17 +00:00
return (ENOMEM); /* XXX: Linux has no ENOMEM here. */
Fix some problems which may lead to a panic: - right order of src and dst in memcpy - NULL out the clips after freeing to prevent an accident Noticed by: hselasky
2010-03-26 08:42:11 +00:00
memcpy(*ppvc, &vclip, sizeof(vclip));
This is v4l support for the linuxulator. This allows to access FreeBSD native devices which support the v4l API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd driver. Not tested is firmware upload, framebuffer stuff and video tuner stuff due to lack of hardware. The clipping part (VIDIOCSWIN) needs a little bit of further work (partly in progress, but can not be tested due to lack of a suitable device). The submitter tested this sucessfully with Skype and flash apps on amd64 and i386 with the multimedia/pwcbsd driver. Submitted by: J.R. Oldroyd <fbsd@opal.com>
2010-01-15 14:58:19 +00:00
(*ppvc)->next = NULL;
This is v4l support for the linuxulator. This allows to access FreeBSD native devices which support the v4l API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd driver. Not tested is firmware upload, framebuffer stuff and video tuner stuff due to lack of hardware. The clipping part (VIDIOCSWIN) needs a little bit of further work (partly in progress, but can not be tested due to lack of a suitable device). The submitter tested this sucessfully with Skype and flash apps on amd64 and i386 with the multimedia/pwcbsd driver. Submitted by: J.R. Oldroyd <fbsd@opal.com>
2009-12-04 21:06:54 +00:00
return (0);
}
static int
linux_v4l_cliplist_free(struct video_window *vw)
{
struct video_clip **ppvc;
struct video_clip **ppvc_next;
for (ppvc = &(vw->clips); *ppvc != NULL; ppvc = ppvc_next) {
ppvc_next = &((*ppvc)->next);
free(*ppvc, M_LINUX);
}
Fix some problems which may lead to a panic: - right order of src and dst in memcpy - NULL out the clips after freeing to prevent an accident Noticed by: hselasky
2010-03-26 08:42:11 +00:00
vw->clips = NULL;
This is v4l support for the linuxulator. This allows to access FreeBSD native devices which support the v4l API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd driver. Not tested is firmware upload, framebuffer stuff and video tuner stuff due to lack of hardware. The clipping part (VIDIOCSWIN) needs a little bit of further work (partly in progress, but can not be tested due to lack of a suitable device). The submitter tested this sucessfully with Skype and flash apps on amd64 and i386 with the multimedia/pwcbsd driver. Submitted by: J.R. Oldroyd <fbsd@opal.com>
2009-12-04 21:06:54 +00:00
return (0);
}
This is v4l support for the linuxulator. This allows to access FreeBSD native devices which support the v4l API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd driver. Not tested is firmware upload, framebuffer stuff and video tuner stuff due to lack of hardware. The clipping part (VIDIOCSWIN) needs a little bit of further work (partly in progress, but can not be tested due to lack of a suitable device). The submitter tested this sucessfully with Skype and flash apps on amd64 and i386 with the multimedia/pwcbsd driver. Submitted by: J.R. Oldroyd <fbsd@opal.com>
2010-01-15 14:58:19 +00:00
static int
linux_v4l_cliplist_copy(struct l_video_window *lvw, struct video_window *vw)
{
int error;
int clipcount;
void *plvc;
struct video_clip **ppvc;
/*
* XXX: The cliplist is used to pass in a list of clipping
* rectangles or, if clipcount == VIDEO_CLIP_BITMAP, a
* clipping bitmap. Some Linux apps, however, appear to
* leave cliplist and clips uninitialized. In any case,
* the cliplist is not used by pwc(4), at the time of
* writing, FreeBSD's only V4L driver. When a driver
* that uses the cliplist is developed, this code may
* need re-examiniation.
*/
error = 0;
clipcount = vw->clipcount;
if (clipcount == VIDEO_CLIP_BITMAP) {
/*
* In this case, the pointer (clips) is overloaded
* to be a "void *" to a bitmap, therefore there
* is no struct video_clip to copy now.
*/
} else if (clipcount > 0 && clipcount <= 16384) {
/*
* Clips points to list of clip rectangles, so
* copy the list.
*
* XXX: Upper limit of 16384 was used here to try to
* avoid cases when clipcount and clips pointer
* are uninitialized and therefore have high random
* values, as is the case in the Linux Skype
* application. The value 16384 was chosen as that
* is what is used in the Linux stradis(4) MPEG
* decoder driver, the only place we found an
* example of cliplist use.
*/
plvc = PTRIN(lvw->clips);
- #ifdef out the cliplist part, skype seems like using an uninitialized variable and can cause problems, without the cliplist handling it works without problems - improve the cliplist error handling - fix VIDIOCGTUNER and VIDIOCSMICROCODE (still no hardware available to test) Submitted by: J.R. Oldroyd <jr@opal.com> X-MFC after: soon (together with all the v4l stuff)
2010-05-03 14:19:58 +00:00
vw->clips = NULL;
This is v4l support for the linuxulator. This allows to access FreeBSD native devices which support the v4l API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd driver. Not tested is firmware upload, framebuffer stuff and video tuner stuff due to lack of hardware. The clipping part (VIDIOCSWIN) needs a little bit of further work (partly in progress, but can not be tested due to lack of a suitable device). The submitter tested this sucessfully with Skype and flash apps on amd64 and i386 with the multimedia/pwcbsd driver. Submitted by: J.R. Oldroyd <fbsd@opal.com>
2010-01-15 14:58:19 +00:00
ppvc = &(vw->clips);
while (clipcount-- > 0) {
compat/linux: for pointers replace 0 with NULL. plvc is a pointer, no functional change. Found with devel/coccinelle.
2016-04-15 16:21:13 +00:00
if (plvc == NULL) {
This is v4l support for the linuxulator. This allows to access FreeBSD native devices which support the v4l API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd driver. Not tested is firmware upload, framebuffer stuff and video tuner stuff due to lack of hardware. The clipping part (VIDIOCSWIN) needs a little bit of further work (partly in progress, but can not be tested due to lack of a suitable device). The submitter tested this sucessfully with Skype and flash apps on amd64 and i386 with the multimedia/pwcbsd driver. Submitted by: J.R. Oldroyd <fbsd@opal.com>
2010-01-15 14:58:19 +00:00
error = EFAULT;
break;
- #ifdef out the cliplist part, skype seems like using an uninitialized variable and can cause problems, without the cliplist handling it works without problems - improve the cliplist error handling - fix VIDIOCGTUNER and VIDIOCSMICROCODE (still no hardware available to test) Submitted by: J.R. Oldroyd <jr@opal.com> X-MFC after: soon (together with all the v4l stuff)
2010-05-03 14:19:58 +00:00
} else {
error = linux_v4l_clip_copy(plvc, ppvc);
if (error) {
linux_v4l_cliplist_free(vw);
break;
}
This is v4l support for the linuxulator. This allows to access FreeBSD native devices which support the v4l API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd driver. Not tested is firmware upload, framebuffer stuff and video tuner stuff due to lack of hardware. The clipping part (VIDIOCSWIN) needs a little bit of further work (partly in progress, but can not be tested due to lack of a suitable device). The submitter tested this sucessfully with Skype and flash apps on amd64 and i386 with the multimedia/pwcbsd driver. Submitted by: J.R. Oldroyd <fbsd@opal.com>
2010-01-15 14:58:19 +00:00
}
ppvc = &((*ppvc)->next);
Linuxolator whitespace cleanup A version of each of the MD files by necessity exists for each CPU architecture supported by the Linuxolator. Clean these up so that new architectures do not inherit whitespace issues. Clean up shared Linuxolator files while here. Sponsored by: Turing Robotic Industries Inc.
2018-02-05 17:29:12 +00:00
plvc = PTRIN(((struct l_video_clip *) plvc)->next);
This is v4l support for the linuxulator. This allows to access FreeBSD native devices which support the v4l API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd driver. Not tested is firmware upload, framebuffer stuff and video tuner stuff due to lack of hardware. The clipping part (VIDIOCSWIN) needs a little bit of further work (partly in progress, but can not be tested due to lack of a suitable device). The submitter tested this sucessfully with Skype and flash apps on amd64 and i386 with the multimedia/pwcbsd driver. Submitted by: J.R. Oldroyd <fbsd@opal.com>
2010-01-15 14:58:19 +00:00
}
} else {
/*
* clipcount == 0 or negative (but not VIDEO_CLIP_BITMAP)
* Force cliplist to null.
*/
vw->clipcount = 0;
vw->clips = NULL;
}
return (error);
}
- #ifdef out the cliplist part, skype seems like using an uninitialized variable and can cause problems, without the cliplist handling it works without problems - improve the cliplist error handling - fix VIDIOCGTUNER and VIDIOCSMICROCODE (still no hardware available to test) Submitted by: J.R. Oldroyd <jr@opal.com> X-MFC after: soon (together with all the v4l stuff)
2010-05-03 14:19:58 +00:00
#endif
This is v4l support for the linuxulator. This allows to access FreeBSD native devices which support the v4l API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd driver. Not tested is firmware upload, framebuffer stuff and video tuner stuff due to lack of hardware. The clipping part (VIDIOCSWIN) needs a little bit of further work (partly in progress, but can not be tested due to lack of a suitable device). The submitter tested this sucessfully with Skype and flash apps on amd64 and i386 with the multimedia/pwcbsd driver. Submitted by: J.R. Oldroyd <fbsd@opal.com>
2009-12-04 21:06:54 +00:00
static int
linux_ioctl_v4l(struct thread *td, struct linux_ioctl_args *args)
{
struct file *fp;
int error;
struct video_tuner vtun;
struct video_window vwin;
struct video_buffer vbuf;
struct video_code vcode;
struct l_video_tuner l_vtun;
struct l_video_window l_vwin;
struct l_video_buffer l_vbuf;
struct l_video_code l_vcode;
switch (args->cmd & 0xffff) {
case LINUX_VIDIOCGCAP: args->cmd = VIDIOCGCAP; break;
case LINUX_VIDIOCGCHAN: args->cmd = VIDIOCGCHAN; break;
case LINUX_VIDIOCSCHAN: args->cmd = VIDIOCSCHAN; break;
case LINUX_VIDIOCGTUNER:
Change the cap_rights_t type from uint64_t to a structure that we can extend in the future in a backward compatible (API and ABI) way. The cap_rights_t represents capability rights. We used to use one bit to represent one right, but we are running out of spare bits. Currently the new structure provides place for 114 rights (so 50 more than the previous cap_rights_t), but it is possible to grow the structure to hold at least 285 rights, although we can make it even larger if 285 rights won't be enough. The structure definition looks like this: struct cap_rights { uint64_t cr_rights[CAP_RIGHTS_VERSION + 2]; }; The initial CAP_RIGHTS_VERSION is 0. The top two bits in the first element of the cr_rights[] array contain total number of elements in the array - 2. This means if those two bits are equal to 0, we have 2 array elements. The top two bits in all remaining array elements should be 0. The next five bits in all array elements contain array index. Only one bit is used and bit position in this five-bits range defines array index. This means there can be at most five array elements in the future. To define new right the CAPRIGHT() macro must be used. The macro takes two arguments - an array index and a bit to set, eg. #define CAP_PDKILL CAPRIGHT(1, 0x0000000000000800ULL) We still support aliases that combine few rights, but the rights have to belong to the same array element, eg: #define CAP_LOOKUP CAPRIGHT(0, 0x0000000000000400ULL) #define CAP_FCHMOD CAPRIGHT(0, 0x0000000000002000ULL) #define CAP_FCHMODAT (CAP_FCHMOD | CAP_LOOKUP) There is new API to manage the new cap_rights_t structure: cap_rights_t *cap_rights_init(cap_rights_t *rights, ...); void cap_rights_set(cap_rights_t *rights, ...); void cap_rights_clear(cap_rights_t *rights, ...); bool cap_rights_is_set(const cap_rights_t *rights, ...); bool cap_rights_is_valid(const cap_rights_t *rights); void cap_rights_merge(cap_rights_t *dst, const cap_rights_t *src); void cap_rights_remove(cap_rights_t *dst, const cap_rights_t *src); bool cap_rights_contains(const cap_rights_t *big, const cap_rights_t *little); Capability rights to the cap_rights_init(), cap_rights_set(), cap_rights_clear() and cap_rights_is_set() functions are provided by separating them with commas, eg: cap_rights_t rights; cap_rights_init(&rights, CAP_READ, CAP_WRITE, CAP_FSTAT); There is no need to terminate the list of rights, as those functions are actually macros that take care of the termination, eg: #define cap_rights_set(rights, ...) \ __cap_rights_set((rights), __VA_ARGS__, 0ULL) void __cap_rights_set(cap_rights_t *rights, ...); Thanks to using one bit as an array index we can assert in those functions that there are no two rights belonging to different array elements provided together. For example this is illegal and will be detected, because CAP_LOOKUP belongs to element 0 and CAP_PDKILL to element 1: cap_rights_init(&rights, CAP_LOOKUP | CAP_PDKILL); Providing several rights that belongs to the same array's element this way is correct, but is not advised. It should only be used for aliases definition. This commit also breaks compatibility with some existing Capsicum system calls, but I see no other way to do that. This should be fine as Capsicum is still experimental and this change is not going to 9.x. Sponsored by: The FreeBSD Foundation
2013-09-05 00:09:56 +00:00
error = fget(td, args->fd,
Eliminate the overhead of gratuitous repeated reinitialization of cap_rights - Add macros to allow preinitialization of cap_rights_t. - Convert most commonly used code paths to use preinitialized cap_rights_t. A 3.6% speedup in fstat was measured with this change. Reported by: mjg Reviewed by: oshogbo Approved by: sbruno MFC after: 1 month
2018-05-09 18:47:24 +00:00
&cap_ioctl_rights, &fp);
Change the cap_rights_t type from uint64_t to a structure that we can extend in the future in a backward compatible (API and ABI) way. The cap_rights_t represents capability rights. We used to use one bit to represent one right, but we are running out of spare bits. Currently the new structure provides place for 114 rights (so 50 more than the previous cap_rights_t), but it is possible to grow the structure to hold at least 285 rights, although we can make it even larger if 285 rights won't be enough. The structure definition looks like this: struct cap_rights { uint64_t cr_rights[CAP_RIGHTS_VERSION + 2]; }; The initial CAP_RIGHTS_VERSION is 0. The top two bits in the first element of the cr_rights[] array contain total number of elements in the array - 2. This means if those two bits are equal to 0, we have 2 array elements. The top two bits in all remaining array elements should be 0. The next five bits in all array elements contain array index. Only one bit is used and bit position in this five-bits range defines array index. This means there can be at most five array elements in the future. To define new right the CAPRIGHT() macro must be used. The macro takes two arguments - an array index and a bit to set, eg. #define CAP_PDKILL CAPRIGHT(1, 0x0000000000000800ULL) We still support aliases that combine few rights, but the rights have to belong to the same array element, eg: #define CAP_LOOKUP CAPRIGHT(0, 0x0000000000000400ULL) #define CAP_FCHMOD CAPRIGHT(0, 0x0000000000002000ULL) #define CAP_FCHMODAT (CAP_FCHMOD | CAP_LOOKUP) There is new API to manage the new cap_rights_t structure: cap_rights_t *cap_rights_init(cap_rights_t *rights, ...); void cap_rights_set(cap_rights_t *rights, ...); void cap_rights_clear(cap_rights_t *rights, ...); bool cap_rights_is_set(const cap_rights_t *rights, ...); bool cap_rights_is_valid(const cap_rights_t *rights); void cap_rights_merge(cap_rights_t *dst, const cap_rights_t *src); void cap_rights_remove(cap_rights_t *dst, const cap_rights_t *src); bool cap_rights_contains(const cap_rights_t *big, const cap_rights_t *little); Capability rights to the cap_rights_init(), cap_rights_set(), cap_rights_clear() and cap_rights_is_set() functions are provided by separating them with commas, eg: cap_rights_t rights; cap_rights_init(&rights, CAP_READ, CAP_WRITE, CAP_FSTAT); There is no need to terminate the list of rights, as those functions are actually macros that take care of the termination, eg: #define cap_rights_set(rights, ...) \ __cap_rights_set((rights), __VA_ARGS__, 0ULL) void __cap_rights_set(cap_rights_t *rights, ...); Thanks to using one bit as an array index we can assert in those functions that there are no two rights belonging to different array elements provided together. For example this is illegal and will be detected, because CAP_LOOKUP belongs to element 0 and CAP_PDKILL to element 1: cap_rights_init(&rights, CAP_LOOKUP | CAP_PDKILL); Providing several rights that belongs to the same array's element this way is correct, but is not advised. It should only be used for aliases definition. This commit also breaks compatibility with some existing Capsicum system calls, but I see no other way to do that. This should be fine as Capsicum is still experimental and this change is not going to 9.x. Sponsored by: The FreeBSD Foundation
2013-09-05 00:09:56 +00:00
if (error != 0)
This is v4l support for the linuxulator. This allows to access FreeBSD native devices which support the v4l API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd driver. Not tested is firmware upload, framebuffer stuff and video tuner stuff due to lack of hardware. The clipping part (VIDIOCSWIN) needs a little bit of further work (partly in progress, but can not be tested due to lack of a suitable device). The submitter tested this sucessfully with Skype and flash apps on amd64 and i386 with the multimedia/pwcbsd driver. Submitted by: J.R. Oldroyd <fbsd@opal.com>
2009-12-04 21:06:54 +00:00
return (error);
- #ifdef out the cliplist part, skype seems like using an uninitialized variable and can cause problems, without the cliplist handling it works without problems - improve the cliplist error handling - fix VIDIOCGTUNER and VIDIOCSMICROCODE (still no hardware available to test) Submitted by: J.R. Oldroyd <jr@opal.com> X-MFC after: soon (together with all the v4l stuff)
2010-05-03 14:19:58 +00:00
error = copyin((void *) args->arg, &l_vtun, sizeof(l_vtun));
if (error) {
fdrop(fp, td);
return (error);
}
linux_to_bsd_v4l_tuner(&l_vtun, &vtun);
This is v4l support for the linuxulator. This allows to access FreeBSD native devices which support the v4l API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd driver. Not tested is firmware upload, framebuffer stuff and video tuner stuff due to lack of hardware. The clipping part (VIDIOCSWIN) needs a little bit of further work (partly in progress, but can not be tested due to lack of a suitable device). The submitter tested this sucessfully with Skype and flash apps on amd64 and i386 with the multimedia/pwcbsd driver. Submitted by: J.R. Oldroyd <fbsd@opal.com>
2009-12-04 21:06:54 +00:00
error = fo_ioctl(fp, VIDIOCGTUNER, &vtun, td->td_ucred, td);
if (!error) {
bsd_to_linux_v4l_tuner(&vtun, &l_vtun);
error = copyout(&l_vtun, (void *) args->arg,
sizeof(l_vtun));
}
fdrop(fp, td);
return (error);
case LINUX_VIDIOCSTUNER:
Change the cap_rights_t type from uint64_t to a structure that we can extend in the future in a backward compatible (API and ABI) way. The cap_rights_t represents capability rights. We used to use one bit to represent one right, but we are running out of spare bits. Currently the new structure provides place for 114 rights (so 50 more than the previous cap_rights_t), but it is possible to grow the structure to hold at least 285 rights, although we can make it even larger if 285 rights won't be enough. The structure definition looks like this: struct cap_rights { uint64_t cr_rights[CAP_RIGHTS_VERSION + 2]; }; The initial CAP_RIGHTS_VERSION is 0. The top two bits in the first element of the cr_rights[] array contain total number of elements in the array - 2. This means if those two bits are equal to 0, we have 2 array elements. The top two bits in all remaining array elements should be 0. The next five bits in all array elements contain array index. Only one bit is used and bit position in this five-bits range defines array index. This means there can be at most five array elements in the future. To define new right the CAPRIGHT() macro must be used. The macro takes two arguments - an array index and a bit to set, eg. #define CAP_PDKILL CAPRIGHT(1, 0x0000000000000800ULL) We still support aliases that combine few rights, but the rights have to belong to the same array element, eg: #define CAP_LOOKUP CAPRIGHT(0, 0x0000000000000400ULL) #define CAP_FCHMOD CAPRIGHT(0, 0x0000000000002000ULL) #define CAP_FCHMODAT (CAP_FCHMOD | CAP_LOOKUP) There is new API to manage the new cap_rights_t structure: cap_rights_t *cap_rights_init(cap_rights_t *rights, ...); void cap_rights_set(cap_rights_t *rights, ...); void cap_rights_clear(cap_rights_t *rights, ...); bool cap_rights_is_set(const cap_rights_t *rights, ...); bool cap_rights_is_valid(const cap_rights_t *rights); void cap_rights_merge(cap_rights_t *dst, const cap_rights_t *src); void cap_rights_remove(cap_rights_t *dst, const cap_rights_t *src); bool cap_rights_contains(const cap_rights_t *big, const cap_rights_t *little); Capability rights to the cap_rights_init(), cap_rights_set(), cap_rights_clear() and cap_rights_is_set() functions are provided by separating them with commas, eg: cap_rights_t rights; cap_rights_init(&rights, CAP_READ, CAP_WRITE, CAP_FSTAT); There is no need to terminate the list of rights, as those functions are actually macros that take care of the termination, eg: #define cap_rights_set(rights, ...) \ __cap_rights_set((rights), __VA_ARGS__, 0ULL) void __cap_rights_set(cap_rights_t *rights, ...); Thanks to using one bit as an array index we can assert in those functions that there are no two rights belonging to different array elements provided together. For example this is illegal and will be detected, because CAP_LOOKUP belongs to element 0 and CAP_PDKILL to element 1: cap_rights_init(&rights, CAP_LOOKUP | CAP_PDKILL); Providing several rights that belongs to the same array's element this way is correct, but is not advised. It should only be used for aliases definition. This commit also breaks compatibility with some existing Capsicum system calls, but I see no other way to do that. This should be fine as Capsicum is still experimental and this change is not going to 9.x. Sponsored by: The FreeBSD Foundation
2013-09-05 00:09:56 +00:00
error = fget(td, args->fd,
Eliminate the overhead of gratuitous repeated reinitialization of cap_rights - Add macros to allow preinitialization of cap_rights_t. - Convert most commonly used code paths to use preinitialized cap_rights_t. A 3.6% speedup in fstat was measured with this change. Reported by: mjg Reviewed by: oshogbo Approved by: sbruno MFC after: 1 month
2018-05-09 18:47:24 +00:00
&cap_ioctl_rights, &fp);
Change the cap_rights_t type from uint64_t to a structure that we can extend in the future in a backward compatible (API and ABI) way. The cap_rights_t represents capability rights. We used to use one bit to represent one right, but we are running out of spare bits. Currently the new structure provides place for 114 rights (so 50 more than the previous cap_rights_t), but it is possible to grow the structure to hold at least 285 rights, although we can make it even larger if 285 rights won't be enough. The structure definition looks like this: struct cap_rights { uint64_t cr_rights[CAP_RIGHTS_VERSION + 2]; }; The initial CAP_RIGHTS_VERSION is 0. The top two bits in the first element of the cr_rights[] array contain total number of elements in the array - 2. This means if those two bits are equal to 0, we have 2 array elements. The top two bits in all remaining array elements should be 0. The next five bits in all array elements contain array index. Only one bit is used and bit position in this five-bits range defines array index. This means there can be at most five array elements in the future. To define new right the CAPRIGHT() macro must be used. The macro takes two arguments - an array index and a bit to set, eg. #define CAP_PDKILL CAPRIGHT(1, 0x0000000000000800ULL) We still support aliases that combine few rights, but the rights have to belong to the same array element, eg: #define CAP_LOOKUP CAPRIGHT(0, 0x0000000000000400ULL) #define CAP_FCHMOD CAPRIGHT(0, 0x0000000000002000ULL) #define CAP_FCHMODAT (CAP_FCHMOD | CAP_LOOKUP) There is new API to manage the new cap_rights_t structure: cap_rights_t *cap_rights_init(cap_rights_t *rights, ...); void cap_rights_set(cap_rights_t *rights, ...); void cap_rights_clear(cap_rights_t *rights, ...); bool cap_rights_is_set(const cap_rights_t *rights, ...); bool cap_rights_is_valid(const cap_rights_t *rights); void cap_rights_merge(cap_rights_t *dst, const cap_rights_t *src); void cap_rights_remove(cap_rights_t *dst, const cap_rights_t *src); bool cap_rights_contains(const cap_rights_t *big, const cap_rights_t *little); Capability rights to the cap_rights_init(), cap_rights_set(), cap_rights_clear() and cap_rights_is_set() functions are provided by separating them with commas, eg: cap_rights_t rights; cap_rights_init(&rights, CAP_READ, CAP_WRITE, CAP_FSTAT); There is no need to terminate the list of rights, as those functions are actually macros that take care of the termination, eg: #define cap_rights_set(rights, ...) \ __cap_rights_set((rights), __VA_ARGS__, 0ULL) void __cap_rights_set(cap_rights_t *rights, ...); Thanks to using one bit as an array index we can assert in those functions that there are no two rights belonging to different array elements provided together. For example this is illegal and will be detected, because CAP_LOOKUP belongs to element 0 and CAP_PDKILL to element 1: cap_rights_init(&rights, CAP_LOOKUP | CAP_PDKILL); Providing several rights that belongs to the same array's element this way is correct, but is not advised. It should only be used for aliases definition. This commit also breaks compatibility with some existing Capsicum system calls, but I see no other way to do that. This should be fine as Capsicum is still experimental and this change is not going to 9.x. Sponsored by: The FreeBSD Foundation
2013-09-05 00:09:56 +00:00
if (error != 0)
This is v4l support for the linuxulator. This allows to access FreeBSD native devices which support the v4l API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd driver. Not tested is firmware upload, framebuffer stuff and video tuner stuff due to lack of hardware. The clipping part (VIDIOCSWIN) needs a little bit of further work (partly in progress, but can not be tested due to lack of a suitable device). The submitter tested this sucessfully with Skype and flash apps on amd64 and i386 with the multimedia/pwcbsd driver. Submitted by: J.R. Oldroyd <fbsd@opal.com>
2009-12-04 21:06:54 +00:00
return (error);
error = copyin((void *) args->arg, &l_vtun, sizeof(l_vtun));
if (error) {
fdrop(fp, td);
return (error);
}
linux_to_bsd_v4l_tuner(&l_vtun, &vtun);
- #ifdef out the cliplist part, skype seems like using an uninitialized variable and can cause problems, without the cliplist handling it works without problems - improve the cliplist error handling - fix VIDIOCGTUNER and VIDIOCSMICROCODE (still no hardware available to test) Submitted by: J.R. Oldroyd <jr@opal.com> X-MFC after: soon (together with all the v4l stuff)
2010-05-03 14:19:58 +00:00
error = fo_ioctl(fp, VIDIOCSTUNER, &vtun, td->td_ucred, td);
This is v4l support for the linuxulator. This allows to access FreeBSD native devices which support the v4l API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd driver. Not tested is firmware upload, framebuffer stuff and video tuner stuff due to lack of hardware. The clipping part (VIDIOCSWIN) needs a little bit of further work (partly in progress, but can not be tested due to lack of a suitable device). The submitter tested this sucessfully with Skype and flash apps on amd64 and i386 with the multimedia/pwcbsd driver. Submitted by: J.R. Oldroyd <fbsd@opal.com>
2009-12-04 21:06:54 +00:00
fdrop(fp, td);
return (error);
case LINUX_VIDIOCGPICT: args->cmd = VIDIOCGPICT; break;
case LINUX_VIDIOCSPICT: args->cmd = VIDIOCSPICT; break;
case LINUX_VIDIOCCAPTURE: args->cmd = VIDIOCCAPTURE; break;
case LINUX_VIDIOCGWIN:
Change the cap_rights_t type from uint64_t to a structure that we can extend in the future in a backward compatible (API and ABI) way. The cap_rights_t represents capability rights. We used to use one bit to represent one right, but we are running out of spare bits. Currently the new structure provides place for 114 rights (so 50 more than the previous cap_rights_t), but it is possible to grow the structure to hold at least 285 rights, although we can make it even larger if 285 rights won't be enough. The structure definition looks like this: struct cap_rights { uint64_t cr_rights[CAP_RIGHTS_VERSION + 2]; }; The initial CAP_RIGHTS_VERSION is 0. The top two bits in the first element of the cr_rights[] array contain total number of elements in the array - 2. This means if those two bits are equal to 0, we have 2 array elements. The top two bits in all remaining array elements should be 0. The next five bits in all array elements contain array index. Only one bit is used and bit position in this five-bits range defines array index. This means there can be at most five array elements in the future. To define new right the CAPRIGHT() macro must be used. The macro takes two arguments - an array index and a bit to set, eg. #define CAP_PDKILL CAPRIGHT(1, 0x0000000000000800ULL) We still support aliases that combine few rights, but the rights have to belong to the same array element, eg: #define CAP_LOOKUP CAPRIGHT(0, 0x0000000000000400ULL) #define CAP_FCHMOD CAPRIGHT(0, 0x0000000000002000ULL) #define CAP_FCHMODAT (CAP_FCHMOD | CAP_LOOKUP) There is new API to manage the new cap_rights_t structure: cap_rights_t *cap_rights_init(cap_rights_t *rights, ...); void cap_rights_set(cap_rights_t *rights, ...); void cap_rights_clear(cap_rights_t *rights, ...); bool cap_rights_is_set(const cap_rights_t *rights, ...); bool cap_rights_is_valid(const cap_rights_t *rights); void cap_rights_merge(cap_rights_t *dst, const cap_rights_t *src); void cap_rights_remove(cap_rights_t *dst, const cap_rights_t *src); bool cap_rights_contains(const cap_rights_t *big, const cap_rights_t *little); Capability rights to the cap_rights_init(), cap_rights_set(), cap_rights_clear() and cap_rights_is_set() functions are provided by separating them with commas, eg: cap_rights_t rights; cap_rights_init(&rights, CAP_READ, CAP_WRITE, CAP_FSTAT); There is no need to terminate the list of rights, as those functions are actually macros that take care of the termination, eg: #define cap_rights_set(rights, ...) \ __cap_rights_set((rights), __VA_ARGS__, 0ULL) void __cap_rights_set(cap_rights_t *rights, ...); Thanks to using one bit as an array index we can assert in those functions that there are no two rights belonging to different array elements provided together. For example this is illegal and will be detected, because CAP_LOOKUP belongs to element 0 and CAP_PDKILL to element 1: cap_rights_init(&rights, CAP_LOOKUP | CAP_PDKILL); Providing several rights that belongs to the same array's element this way is correct, but is not advised. It should only be used for aliases definition. This commit also breaks compatibility with some existing Capsicum system calls, but I see no other way to do that. This should be fine as Capsicum is still experimental and this change is not going to 9.x. Sponsored by: The FreeBSD Foundation
2013-09-05 00:09:56 +00:00
error = fget(td, args->fd,
Eliminate the overhead of gratuitous repeated reinitialization of cap_rights - Add macros to allow preinitialization of cap_rights_t. - Convert most commonly used code paths to use preinitialized cap_rights_t. A 3.6% speedup in fstat was measured with this change. Reported by: mjg Reviewed by: oshogbo Approved by: sbruno MFC after: 1 month
2018-05-09 18:47:24 +00:00
&cap_ioctl_rights, &fp);
Change the cap_rights_t type from uint64_t to a structure that we can extend in the future in a backward compatible (API and ABI) way. The cap_rights_t represents capability rights. We used to use one bit to represent one right, but we are running out of spare bits. Currently the new structure provides place for 114 rights (so 50 more than the previous cap_rights_t), but it is possible to grow the structure to hold at least 285 rights, although we can make it even larger if 285 rights won't be enough. The structure definition looks like this: struct cap_rights { uint64_t cr_rights[CAP_RIGHTS_VERSION + 2]; }; The initial CAP_RIGHTS_VERSION is 0. The top two bits in the first element of the cr_rights[] array contain total number of elements in the array - 2. This means if those two bits are equal to 0, we have 2 array elements. The top two bits in all remaining array elements should be 0. The next five bits in all array elements contain array index. Only one bit is used and bit position in this five-bits range defines array index. This means there can be at most five array elements in the future. To define new right the CAPRIGHT() macro must be used. The macro takes two arguments - an array index and a bit to set, eg. #define CAP_PDKILL CAPRIGHT(1, 0x0000000000000800ULL) We still support aliases that combine few rights, but the rights have to belong to the same array element, eg: #define CAP_LOOKUP CAPRIGHT(0, 0x0000000000000400ULL) #define CAP_FCHMOD CAPRIGHT(0, 0x0000000000002000ULL) #define CAP_FCHMODAT (CAP_FCHMOD | CAP_LOOKUP) There is new API to manage the new cap_rights_t structure: cap_rights_t *cap_rights_init(cap_rights_t *rights, ...); void cap_rights_set(cap_rights_t *rights, ...); void cap_rights_clear(cap_rights_t *rights, ...); bool cap_rights_is_set(const cap_rights_t *rights, ...); bool cap_rights_is_valid(const cap_rights_t *rights); void cap_rights_merge(cap_rights_t *dst, const cap_rights_t *src); void cap_rights_remove(cap_rights_t *dst, const cap_rights_t *src); bool cap_rights_contains(const cap_rights_t *big, const cap_rights_t *little); Capability rights to the cap_rights_init(), cap_rights_set(), cap_rights_clear() and cap_rights_is_set() functions are provided by separating them with commas, eg: cap_rights_t rights; cap_rights_init(&rights, CAP_READ, CAP_WRITE, CAP_FSTAT); There is no need to terminate the list of rights, as those functions are actually macros that take care of the termination, eg: #define cap_rights_set(rights, ...) \ __cap_rights_set((rights), __VA_ARGS__, 0ULL) void __cap_rights_set(cap_rights_t *rights, ...); Thanks to using one bit as an array index we can assert in those functions that there are no two rights belonging to different array elements provided together. For example this is illegal and will be detected, because CAP_LOOKUP belongs to element 0 and CAP_PDKILL to element 1: cap_rights_init(&rights, CAP_LOOKUP | CAP_PDKILL); Providing several rights that belongs to the same array's element this way is correct, but is not advised. It should only be used for aliases definition. This commit also breaks compatibility with some existing Capsicum system calls, but I see no other way to do that. This should be fine as Capsicum is still experimental and this change is not going to 9.x. Sponsored by: The FreeBSD Foundation
2013-09-05 00:09:56 +00:00
if (error != 0)
This is v4l support for the linuxulator. This allows to access FreeBSD native devices which support the v4l API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd driver. Not tested is firmware upload, framebuffer stuff and video tuner stuff due to lack of hardware. The clipping part (VIDIOCSWIN) needs a little bit of further work (partly in progress, but can not be tested due to lack of a suitable device). The submitter tested this sucessfully with Skype and flash apps on amd64 and i386 with the multimedia/pwcbsd driver. Submitted by: J.R. Oldroyd <fbsd@opal.com>
2009-12-04 21:06:54 +00:00
return (error);
error = fo_ioctl(fp, VIDIOCGWIN, &vwin, td->td_ucred, td);
if (!error) {
bsd_to_linux_v4l_window(&vwin, &l_vwin);
error = copyout(&l_vwin, (void *) args->arg,
sizeof(l_vwin));
}
fdrop(fp, td);
return (error);
case LINUX_VIDIOCSWIN:
Change the cap_rights_t type from uint64_t to a structure that we can extend in the future in a backward compatible (API and ABI) way. The cap_rights_t represents capability rights. We used to use one bit to represent one right, but we are running out of spare bits. Currently the new structure provides place for 114 rights (so 50 more than the previous cap_rights_t), but it is possible to grow the structure to hold at least 285 rights, although we can make it even larger if 285 rights won't be enough. The structure definition looks like this: struct cap_rights { uint64_t cr_rights[CAP_RIGHTS_VERSION + 2]; }; The initial CAP_RIGHTS_VERSION is 0. The top two bits in the first element of the cr_rights[] array contain total number of elements in the array - 2. This means if those two bits are equal to 0, we have 2 array elements. The top two bits in all remaining array elements should be 0. The next five bits in all array elements contain array index. Only one bit is used and bit position in this five-bits range defines array index. This means there can be at most five array elements in the future. To define new right the CAPRIGHT() macro must be used. The macro takes two arguments - an array index and a bit to set, eg. #define CAP_PDKILL CAPRIGHT(1, 0x0000000000000800ULL) We still support aliases that combine few rights, but the rights have to belong to the same array element, eg: #define CAP_LOOKUP CAPRIGHT(0, 0x0000000000000400ULL) #define CAP_FCHMOD CAPRIGHT(0, 0x0000000000002000ULL) #define CAP_FCHMODAT (CAP_FCHMOD | CAP_LOOKUP) There is new API to manage the new cap_rights_t structure: cap_rights_t *cap_rights_init(cap_rights_t *rights, ...); void cap_rights_set(cap_rights_t *rights, ...); void cap_rights_clear(cap_rights_t *rights, ...); bool cap_rights_is_set(const cap_rights_t *rights, ...); bool cap_rights_is_valid(const cap_rights_t *rights); void cap_rights_merge(cap_rights_t *dst, const cap_rights_t *src); void cap_rights_remove(cap_rights_t *dst, const cap_rights_t *src); bool cap_rights_contains(const cap_rights_t *big, const cap_rights_t *little); Capability rights to the cap_rights_init(), cap_rights_set(), cap_rights_clear() and cap_rights_is_set() functions are provided by separating them with commas, eg: cap_rights_t rights; cap_rights_init(&rights, CAP_READ, CAP_WRITE, CAP_FSTAT); There is no need to terminate the list of rights, as those functions are actually macros that take care of the termination, eg: #define cap_rights_set(rights, ...) \ __cap_rights_set((rights), __VA_ARGS__, 0ULL) void __cap_rights_set(cap_rights_t *rights, ...); Thanks to using one bit as an array index we can assert in those functions that there are no two rights belonging to different array elements provided together. For example this is illegal and will be detected, because CAP_LOOKUP belongs to element 0 and CAP_PDKILL to element 1: cap_rights_init(&rights, CAP_LOOKUP | CAP_PDKILL); Providing several rights that belongs to the same array's element this way is correct, but is not advised. It should only be used for aliases definition. This commit also breaks compatibility with some existing Capsicum system calls, but I see no other way to do that. This should be fine as Capsicum is still experimental and this change is not going to 9.x. Sponsored by: The FreeBSD Foundation
2013-09-05 00:09:56 +00:00
error = fget(td, args->fd,
Eliminate the overhead of gratuitous repeated reinitialization of cap_rights - Add macros to allow preinitialization of cap_rights_t. - Convert most commonly used code paths to use preinitialized cap_rights_t. A 3.6% speedup in fstat was measured with this change. Reported by: mjg Reviewed by: oshogbo Approved by: sbruno MFC after: 1 month
2018-05-09 18:47:24 +00:00
&cap_ioctl_rights, &fp);
Change the cap_rights_t type from uint64_t to a structure that we can extend in the future in a backward compatible (API and ABI) way. The cap_rights_t represents capability rights. We used to use one bit to represent one right, but we are running out of spare bits. Currently the new structure provides place for 114 rights (so 50 more than the previous cap_rights_t), but it is possible to grow the structure to hold at least 285 rights, although we can make it even larger if 285 rights won't be enough. The structure definition looks like this: struct cap_rights { uint64_t cr_rights[CAP_RIGHTS_VERSION + 2]; }; The initial CAP_RIGHTS_VERSION is 0. The top two bits in the first element of the cr_rights[] array contain total number of elements in the array - 2. This means if those two bits are equal to 0, we have 2 array elements. The top two bits in all remaining array elements should be 0. The next five bits in all array elements contain array index. Only one bit is used and bit position in this five-bits range defines array index. This means there can be at most five array elements in the future. To define new right the CAPRIGHT() macro must be used. The macro takes two arguments - an array index and a bit to set, eg. #define CAP_PDKILL CAPRIGHT(1, 0x0000000000000800ULL) We still support aliases that combine few rights, but the rights have to belong to the same array element, eg: #define CAP_LOOKUP CAPRIGHT(0, 0x0000000000000400ULL) #define CAP_FCHMOD CAPRIGHT(0, 0x0000000000002000ULL) #define CAP_FCHMODAT (CAP_FCHMOD | CAP_LOOKUP) There is new API to manage the new cap_rights_t structure: cap_rights_t *cap_rights_init(cap_rights_t *rights, ...); void cap_rights_set(cap_rights_t *rights, ...); void cap_rights_clear(cap_rights_t *rights, ...); bool cap_rights_is_set(const cap_rights_t *rights, ...); bool cap_rights_is_valid(const cap_rights_t *rights); void cap_rights_merge(cap_rights_t *dst, const cap_rights_t *src); void cap_rights_remove(cap_rights_t *dst, const cap_rights_t *src); bool cap_rights_contains(const cap_rights_t *big, const cap_rights_t *little); Capability rights to the cap_rights_init(), cap_rights_set(), cap_rights_clear() and cap_rights_is_set() functions are provided by separating them with commas, eg: cap_rights_t rights; cap_rights_init(&rights, CAP_READ, CAP_WRITE, CAP_FSTAT); There is no need to terminate the list of rights, as those functions are actually macros that take care of the termination, eg: #define cap_rights_set(rights, ...) \ __cap_rights_set((rights), __VA_ARGS__, 0ULL) void __cap_rights_set(cap_rights_t *rights, ...); Thanks to using one bit as an array index we can assert in those functions that there are no two rights belonging to different array elements provided together. For example this is illegal and will be detected, because CAP_LOOKUP belongs to element 0 and CAP_PDKILL to element 1: cap_rights_init(&rights, CAP_LOOKUP | CAP_PDKILL); Providing several rights that belongs to the same array's element this way is correct, but is not advised. It should only be used for aliases definition. This commit also breaks compatibility with some existing Capsicum system calls, but I see no other way to do that. This should be fine as Capsicum is still experimental and this change is not going to 9.x. Sponsored by: The FreeBSD Foundation
2013-09-05 00:09:56 +00:00
if (error != 0)
This is v4l support for the linuxulator. This allows to access FreeBSD native devices which support the v4l API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd driver. Not tested is firmware upload, framebuffer stuff and video tuner stuff due to lack of hardware. The clipping part (VIDIOCSWIN) needs a little bit of further work (partly in progress, but can not be tested due to lack of a suitable device). The submitter tested this sucessfully with Skype and flash apps on amd64 and i386 with the multimedia/pwcbsd driver. Submitted by: J.R. Oldroyd <fbsd@opal.com>
2009-12-04 21:06:54 +00:00
return (error);
error = copyin((void *) args->arg, &l_vwin, sizeof(l_vwin));
if (error) {
fdrop(fp, td);
return (error);
}
linux_to_bsd_v4l_window(&l_vwin, &vwin);
- #ifdef out the cliplist part, skype seems like using an uninitialized variable and can cause problems, without the cliplist handling it works without problems - improve the cliplist error handling - fix VIDIOCGTUNER and VIDIOCSMICROCODE (still no hardware available to test) Submitted by: J.R. Oldroyd <jr@opal.com> X-MFC after: soon (together with all the v4l stuff)
2010-05-03 14:19:58 +00:00
#ifdef COMPAT_LINUX_V4L_CLIPLIST
This is v4l support for the linuxulator. This allows to access FreeBSD native devices which support the v4l API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd driver. Not tested is firmware upload, framebuffer stuff and video tuner stuff due to lack of hardware. The clipping part (VIDIOCSWIN) needs a little bit of further work (partly in progress, but can not be tested due to lack of a suitable device). The submitter tested this sucessfully with Skype and flash apps on amd64 and i386 with the multimedia/pwcbsd driver. Submitted by: J.R. Oldroyd <fbsd@opal.com>
2009-12-04 21:06:54 +00:00
error = linux_v4l_cliplist_copy(&l_vwin, &vwin);
This is v4l support for the linuxulator. This allows to access FreeBSD native devices which support the v4l API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd driver. Not tested is firmware upload, framebuffer stuff and video tuner stuff due to lack of hardware. The clipping part (VIDIOCSWIN) needs a little bit of further work (partly in progress, but can not be tested due to lack of a suitable device). The submitter tested this sucessfully with Skype and flash apps on amd64 and i386 with the multimedia/pwcbsd driver. Submitted by: J.R. Oldroyd <fbsd@opal.com>
2010-01-15 14:58:19 +00:00
if (error) {
fdrop(fp, td);
return (error);
}
- #ifdef out the cliplist part, skype seems like using an uninitialized variable and can cause problems, without the cliplist handling it works without problems - improve the cliplist error handling - fix VIDIOCGTUNER and VIDIOCSMICROCODE (still no hardware available to test) Submitted by: J.R. Oldroyd <jr@opal.com> X-MFC after: soon (together with all the v4l stuff)
2010-05-03 14:19:58 +00:00
#endif
This is v4l support for the linuxulator. This allows to access FreeBSD native devices which support the v4l API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd driver. Not tested is firmware upload, framebuffer stuff and video tuner stuff due to lack of hardware. The clipping part (VIDIOCSWIN) needs a little bit of further work (partly in progress, but can not be tested due to lack of a suitable device). The submitter tested this sucessfully with Skype and flash apps on amd64 and i386 with the multimedia/pwcbsd driver. Submitted by: J.R. Oldroyd <fbsd@opal.com>
2010-01-15 14:58:19 +00:00
error = fo_ioctl(fp, VIDIOCSWIN, &vwin, td->td_ucred, td);
This is v4l support for the linuxulator. This allows to access FreeBSD native devices which support the v4l API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd driver. Not tested is firmware upload, framebuffer stuff and video tuner stuff due to lack of hardware. The clipping part (VIDIOCSWIN) needs a little bit of further work (partly in progress, but can not be tested due to lack of a suitable device). The submitter tested this sucessfully with Skype and flash apps on amd64 and i386 with the multimedia/pwcbsd driver. Submitted by: J.R. Oldroyd <fbsd@opal.com>
2009-12-04 21:06:54 +00:00
fdrop(fp, td);
- #ifdef out the cliplist part, skype seems like using an uninitialized variable and can cause problems, without the cliplist handling it works without problems - improve the cliplist error handling - fix VIDIOCGTUNER and VIDIOCSMICROCODE (still no hardware available to test) Submitted by: J.R. Oldroyd <jr@opal.com> X-MFC after: soon (together with all the v4l stuff)
2010-05-03 14:19:58 +00:00
#ifdef COMPAT_LINUX_V4L_CLIPLIST
This is v4l support for the linuxulator. This allows to access FreeBSD native devices which support the v4l API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd driver. Not tested is firmware upload, framebuffer stuff and video tuner stuff due to lack of hardware. The clipping part (VIDIOCSWIN) needs a little bit of further work (partly in progress, but can not be tested due to lack of a suitable device). The submitter tested this sucessfully with Skype and flash apps on amd64 and i386 with the multimedia/pwcbsd driver. Submitted by: J.R. Oldroyd <fbsd@opal.com>
2009-12-04 21:06:54 +00:00
linux_v4l_cliplist_free(&vwin);
- #ifdef out the cliplist part, skype seems like using an uninitialized variable and can cause problems, without the cliplist handling it works without problems - improve the cliplist error handling - fix VIDIOCGTUNER and VIDIOCSMICROCODE (still no hardware available to test) Submitted by: J.R. Oldroyd <jr@opal.com> X-MFC after: soon (together with all the v4l stuff)
2010-05-03 14:19:58 +00:00
#endif
This is v4l support for the linuxulator. This allows to access FreeBSD native devices which support the v4l API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd driver. Not tested is firmware upload, framebuffer stuff and video tuner stuff due to lack of hardware. The clipping part (VIDIOCSWIN) needs a little bit of further work (partly in progress, but can not be tested due to lack of a suitable device). The submitter tested this sucessfully with Skype and flash apps on amd64 and i386 with the multimedia/pwcbsd driver. Submitted by: J.R. Oldroyd <fbsd@opal.com>
2009-12-04 21:06:54 +00:00
return (error);
case LINUX_VIDIOCGFBUF:
Change the cap_rights_t type from uint64_t to a structure that we can extend in the future in a backward compatible (API and ABI) way. The cap_rights_t represents capability rights. We used to use one bit to represent one right, but we are running out of spare bits. Currently the new structure provides place for 114 rights (so 50 more than the previous cap_rights_t), but it is possible to grow the structure to hold at least 285 rights, although we can make it even larger if 285 rights won't be enough. The structure definition looks like this: struct cap_rights { uint64_t cr_rights[CAP_RIGHTS_VERSION + 2]; }; The initial CAP_RIGHTS_VERSION is 0. The top two bits in the first element of the cr_rights[] array contain total number of elements in the array - 2. This means if those two bits are equal to 0, we have 2 array elements. The top two bits in all remaining array elements should be 0. The next five bits in all array elements contain array index. Only one bit is used and bit position in this five-bits range defines array index. This means there can be at most five array elements in the future. To define new right the CAPRIGHT() macro must be used. The macro takes two arguments - an array index and a bit to set, eg. #define CAP_PDKILL CAPRIGHT(1, 0x0000000000000800ULL) We still support aliases that combine few rights, but the rights have to belong to the same array element, eg: #define CAP_LOOKUP CAPRIGHT(0, 0x0000000000000400ULL) #define CAP_FCHMOD CAPRIGHT(0, 0x0000000000002000ULL) #define CAP_FCHMODAT (CAP_FCHMOD | CAP_LOOKUP) There is new API to manage the new cap_rights_t structure: cap_rights_t *cap_rights_init(cap_rights_t *rights, ...); void cap_rights_set(cap_rights_t *rights, ...); void cap_rights_clear(cap_rights_t *rights, ...); bool cap_rights_is_set(const cap_rights_t *rights, ...); bool cap_rights_is_valid(const cap_rights_t *rights); void cap_rights_merge(cap_rights_t *dst, const cap_rights_t *src); void cap_rights_remove(cap_rights_t *dst, const cap_rights_t *src); bool cap_rights_contains(const cap_rights_t *big, const cap_rights_t *little); Capability rights to the cap_rights_init(), cap_rights_set(), cap_rights_clear() and cap_rights_is_set() functions are provided by separating them with commas, eg: cap_rights_t rights; cap_rights_init(&rights, CAP_READ, CAP_WRITE, CAP_FSTAT); There is no need to terminate the list of rights, as those functions are actually macros that take care of the termination, eg: #define cap_rights_set(rights, ...) \ __cap_rights_set((rights), __VA_ARGS__, 0ULL) void __cap_rights_set(cap_rights_t *rights, ...); Thanks to using one bit as an array index we can assert in those functions that there are no two rights belonging to different array elements provided together. For example this is illegal and will be detected, because CAP_LOOKUP belongs to element 0 and CAP_PDKILL to element 1: cap_rights_init(&rights, CAP_LOOKUP | CAP_PDKILL); Providing several rights that belongs to the same array's element this way is correct, but is not advised. It should only be used for aliases definition. This commit also breaks compatibility with some existing Capsicum system calls, but I see no other way to do that. This should be fine as Capsicum is still experimental and this change is not going to 9.x. Sponsored by: The FreeBSD Foundation
2013-09-05 00:09:56 +00:00
error = fget(td, args->fd,
Eliminate the overhead of gratuitous repeated reinitialization of cap_rights - Add macros to allow preinitialization of cap_rights_t. - Convert most commonly used code paths to use preinitialized cap_rights_t. A 3.6% speedup in fstat was measured with this change. Reported by: mjg Reviewed by: oshogbo Approved by: sbruno MFC after: 1 month
2018-05-09 18:47:24 +00:00
&cap_ioctl_rights, &fp);
Change the cap_rights_t type from uint64_t to a structure that we can extend in the future in a backward compatible (API and ABI) way. The cap_rights_t represents capability rights. We used to use one bit to represent one right, but we are running out of spare bits. Currently the new structure provides place for 114 rights (so 50 more than the previous cap_rights_t), but it is possible to grow the structure to hold at least 285 rights, although we can make it even larger if 285 rights won't be enough. The structure definition looks like this: struct cap_rights { uint64_t cr_rights[CAP_RIGHTS_VERSION + 2]; }; The initial CAP_RIGHTS_VERSION is 0. The top two bits in the first element of the cr_rights[] array contain total number of elements in the array - 2. This means if those two bits are equal to 0, we have 2 array elements. The top two bits in all remaining array elements should be 0. The next five bits in all array elements contain array index. Only one bit is used and bit position in this five-bits range defines array index. This means there can be at most five array elements in the future. To define new right the CAPRIGHT() macro must be used. The macro takes two arguments - an array index and a bit to set, eg. #define CAP_PDKILL CAPRIGHT(1, 0x0000000000000800ULL) We still support aliases that combine few rights, but the rights have to belong to the same array element, eg: #define CAP_LOOKUP CAPRIGHT(0, 0x0000000000000400ULL) #define CAP_FCHMOD CAPRIGHT(0, 0x0000000000002000ULL) #define CAP_FCHMODAT (CAP_FCHMOD | CAP_LOOKUP) There is new API to manage the new cap_rights_t structure: cap_rights_t *cap_rights_init(cap_rights_t *rights, ...); void cap_rights_set(cap_rights_t *rights, ...); void cap_rights_clear(cap_rights_t *rights, ...); bool cap_rights_is_set(const cap_rights_t *rights, ...); bool cap_rights_is_valid(const cap_rights_t *rights); void cap_rights_merge(cap_rights_t *dst, const cap_rights_t *src); void cap_rights_remove(cap_rights_t *dst, const cap_rights_t *src); bool cap_rights_contains(const cap_rights_t *big, const cap_rights_t *little); Capability rights to the cap_rights_init(), cap_rights_set(), cap_rights_clear() and cap_rights_is_set() functions are provided by separating them with commas, eg: cap_rights_t rights; cap_rights_init(&rights, CAP_READ, CAP_WRITE, CAP_FSTAT); There is no need to terminate the list of rights, as those functions are actually macros that take care of the termination, eg: #define cap_rights_set(rights, ...) \ __cap_rights_set((rights), __VA_ARGS__, 0ULL) void __cap_rights_set(cap_rights_t *rights, ...); Thanks to using one bit as an array index we can assert in those functions that there are no two rights belonging to different array elements provided together. For example this is illegal and will be detected, because CAP_LOOKUP belongs to element 0 and CAP_PDKILL to element 1: cap_rights_init(&rights, CAP_LOOKUP | CAP_PDKILL); Providing several rights that belongs to the same array's element this way is correct, but is not advised. It should only be used for aliases definition. This commit also breaks compatibility with some existing Capsicum system calls, but I see no other way to do that. This should be fine as Capsicum is still experimental and this change is not going to 9.x. Sponsored by: The FreeBSD Foundation
2013-09-05 00:09:56 +00:00
if (error != 0)
This is v4l support for the linuxulator. This allows to access FreeBSD native devices which support the v4l API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd driver. Not tested is firmware upload, framebuffer stuff and video tuner stuff due to lack of hardware. The clipping part (VIDIOCSWIN) needs a little bit of further work (partly in progress, but can not be tested due to lack of a suitable device). The submitter tested this sucessfully with Skype and flash apps on amd64 and i386 with the multimedia/pwcbsd driver. Submitted by: J.R. Oldroyd <fbsd@opal.com>
2009-12-04 21:06:54 +00:00
return (error);
error = fo_ioctl(fp, VIDIOCGFBUF, &vbuf, td->td_ucred, td);
if (!error) {
bsd_to_linux_v4l_buffer(&vbuf, &l_vbuf);
error = copyout(&l_vbuf, (void *) args->arg,
sizeof(l_vbuf));
}
fdrop(fp, td);
return (error);
case LINUX_VIDIOCSFBUF:
Change the cap_rights_t type from uint64_t to a structure that we can extend in the future in a backward compatible (API and ABI) way. The cap_rights_t represents capability rights. We used to use one bit to represent one right, but we are running out of spare bits. Currently the new structure provides place for 114 rights (so 50 more than the previous cap_rights_t), but it is possible to grow the structure to hold at least 285 rights, although we can make it even larger if 285 rights won't be enough. The structure definition looks like this: struct cap_rights { uint64_t cr_rights[CAP_RIGHTS_VERSION + 2]; }; The initial CAP_RIGHTS_VERSION is 0. The top two bits in the first element of the cr_rights[] array contain total number of elements in the array - 2. This means if those two bits are equal to 0, we have 2 array elements. The top two bits in all remaining array elements should be 0. The next five bits in all array elements contain array index. Only one bit is used and bit position in this five-bits range defines array index. This means there can be at most five array elements in the future. To define new right the CAPRIGHT() macro must be used. The macro takes two arguments - an array index and a bit to set, eg. #define CAP_PDKILL CAPRIGHT(1, 0x0000000000000800ULL) We still support aliases that combine few rights, but the rights have to belong to the same array element, eg: #define CAP_LOOKUP CAPRIGHT(0, 0x0000000000000400ULL) #define CAP_FCHMOD CAPRIGHT(0, 0x0000000000002000ULL) #define CAP_FCHMODAT (CAP_FCHMOD | CAP_LOOKUP) There is new API to manage the new cap_rights_t structure: cap_rights_t *cap_rights_init(cap_rights_t *rights, ...); void cap_rights_set(cap_rights_t *rights, ...); void cap_rights_clear(cap_rights_t *rights, ...); bool cap_rights_is_set(const cap_rights_t *rights, ...); bool cap_rights_is_valid(const cap_rights_t *rights); void cap_rights_merge(cap_rights_t *dst, const cap_rights_t *src); void cap_rights_remove(cap_rights_t *dst, const cap_rights_t *src); bool cap_rights_contains(const cap_rights_t *big, const cap_rights_t *little); Capability rights to the cap_rights_init(), cap_rights_set(), cap_rights_clear() and cap_rights_is_set() functions are provided by separating them with commas, eg: cap_rights_t rights; cap_rights_init(&rights, CAP_READ, CAP_WRITE, CAP_FSTAT); There is no need to terminate the list of rights, as those functions are actually macros that take care of the termination, eg: #define cap_rights_set(rights, ...) \ __cap_rights_set((rights), __VA_ARGS__, 0ULL) void __cap_rights_set(cap_rights_t *rights, ...); Thanks to using one bit as an array index we can assert in those functions that there are no two rights belonging to different array elements provided together. For example this is illegal and will be detected, because CAP_LOOKUP belongs to element 0 and CAP_PDKILL to element 1: cap_rights_init(&rights, CAP_LOOKUP | CAP_PDKILL); Providing several rights that belongs to the same array's element this way is correct, but is not advised. It should only be used for aliases definition. This commit also breaks compatibility with some existing Capsicum system calls, but I see no other way to do that. This should be fine as Capsicum is still experimental and this change is not going to 9.x. Sponsored by: The FreeBSD Foundation
2013-09-05 00:09:56 +00:00
error = fget(td, args->fd,
Eliminate the overhead of gratuitous repeated reinitialization of cap_rights - Add macros to allow preinitialization of cap_rights_t. - Convert most commonly used code paths to use preinitialized cap_rights_t. A 3.6% speedup in fstat was measured with this change. Reported by: mjg Reviewed by: oshogbo Approved by: sbruno MFC after: 1 month
2018-05-09 18:47:24 +00:00
&cap_ioctl_rights, &fp);
Change the cap_rights_t type from uint64_t to a structure that we can extend in the future in a backward compatible (API and ABI) way. The cap_rights_t represents capability rights. We used to use one bit to represent one right, but we are running out of spare bits. Currently the new structure provides place for 114 rights (so 50 more than the previous cap_rights_t), but it is possible to grow the structure to hold at least 285 rights, although we can make it even larger if 285 rights won't be enough. The structure definition looks like this: struct cap_rights { uint64_t cr_rights[CAP_RIGHTS_VERSION + 2]; }; The initial CAP_RIGHTS_VERSION is 0. The top two bits in the first element of the cr_rights[] array contain total number of elements in the array - 2. This means if those two bits are equal to 0, we have 2 array elements. The top two bits in all remaining array elements should be 0. The next five bits in all array elements contain array index. Only one bit is used and bit position in this five-bits range defines array index. This means there can be at most five array elements in the future. To define new right the CAPRIGHT() macro must be used. The macro takes two arguments - an array index and a bit to set, eg. #define CAP_PDKILL CAPRIGHT(1, 0x0000000000000800ULL) We still support aliases that combine few rights, but the rights have to belong to the same array element, eg: #define CAP_LOOKUP CAPRIGHT(0, 0x0000000000000400ULL) #define CAP_FCHMOD CAPRIGHT(0, 0x0000000000002000ULL) #define CAP_FCHMODAT (CAP_FCHMOD | CAP_LOOKUP) There is new API to manage the new cap_rights_t structure: cap_rights_t *cap_rights_init(cap_rights_t *rights, ...); void cap_rights_set(cap_rights_t *rights, ...); void cap_rights_clear(cap_rights_t *rights, ...); bool cap_rights_is_set(const cap_rights_t *rights, ...); bool cap_rights_is_valid(const cap_rights_t *rights); void cap_rights_merge(cap_rights_t *dst, const cap_rights_t *src); void cap_rights_remove(cap_rights_t *dst, const cap_rights_t *src); bool cap_rights_contains(const cap_rights_t *big, const cap_rights_t *little); Capability rights to the cap_rights_init(), cap_rights_set(), cap_rights_clear() and cap_rights_is_set() functions are provided by separating them with commas, eg: cap_rights_t rights; cap_rights_init(&rights, CAP_READ, CAP_WRITE, CAP_FSTAT); There is no need to terminate the list of rights, as those functions are actually macros that take care of the termination, eg: #define cap_rights_set(rights, ...) \ __cap_rights_set((rights), __VA_ARGS__, 0ULL) void __cap_rights_set(cap_rights_t *rights, ...); Thanks to using one bit as an array index we can assert in those functions that there are no two rights belonging to different array elements provided together. For example this is illegal and will be detected, because CAP_LOOKUP belongs to element 0 and CAP_PDKILL to element 1: cap_rights_init(&rights, CAP_LOOKUP | CAP_PDKILL); Providing several rights that belongs to the same array's element this way is correct, but is not advised. It should only be used for aliases definition. This commit also breaks compatibility with some existing Capsicum system calls, but I see no other way to do that. This should be fine as Capsicum is still experimental and this change is not going to 9.x. Sponsored by: The FreeBSD Foundation
2013-09-05 00:09:56 +00:00
if (error != 0)
This is v4l support for the linuxulator. This allows to access FreeBSD native devices which support the v4l API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd driver. Not tested is firmware upload, framebuffer stuff and video tuner stuff due to lack of hardware. The clipping part (VIDIOCSWIN) needs a little bit of further work (partly in progress, but can not be tested due to lack of a suitable device). The submitter tested this sucessfully with Skype and flash apps on amd64 and i386 with the multimedia/pwcbsd driver. Submitted by: J.R. Oldroyd <fbsd@opal.com>
2009-12-04 21:06:54 +00:00
return (error);
error = copyin((void *) args->arg, &l_vbuf, sizeof(l_vbuf));
if (error) {
fdrop(fp, td);
return (error);
}
linux_to_bsd_v4l_buffer(&l_vbuf, &vbuf);
error = fo_ioctl(fp, VIDIOCSFBUF, &vbuf, td->td_ucred, td);
fdrop(fp, td);
return (error);
case LINUX_VIDIOCKEY: args->cmd = VIDIOCKEY; break;
case LINUX_VIDIOCGFREQ: args->cmd = VIDIOCGFREQ; break;
case LINUX_VIDIOCSFREQ: args->cmd = VIDIOCSFREQ; break;
case LINUX_VIDIOCGAUDIO: args->cmd = VIDIOCGAUDIO; break;
case LINUX_VIDIOCSAUDIO: args->cmd = VIDIOCSAUDIO; break;
case LINUX_VIDIOCSYNC: args->cmd = VIDIOCSYNC; break;
case LINUX_VIDIOCMCAPTURE: args->cmd = VIDIOCMCAPTURE; break;
case LINUX_VIDIOCGMBUF: args->cmd = VIDIOCGMBUF; break;
case LINUX_VIDIOCGUNIT: args->cmd = VIDIOCGUNIT; break;
case LINUX_VIDIOCGCAPTURE: args->cmd = VIDIOCGCAPTURE; break;
case LINUX_VIDIOCSCAPTURE: args->cmd = VIDIOCSCAPTURE; break;
case LINUX_VIDIOCSPLAYMODE: args->cmd = VIDIOCSPLAYMODE; break;
case LINUX_VIDIOCSWRITEMODE: args->cmd = VIDIOCSWRITEMODE; break;
case LINUX_VIDIOCGPLAYINFO: args->cmd = VIDIOCGPLAYINFO; break;
case LINUX_VIDIOCSMICROCODE:
Change the cap_rights_t type from uint64_t to a structure that we can extend in the future in a backward compatible (API and ABI) way. The cap_rights_t represents capability rights. We used to use one bit to represent one right, but we are running out of spare bits. Currently the new structure provides place for 114 rights (so 50 more than the previous cap_rights_t), but it is possible to grow the structure to hold at least 285 rights, although we can make it even larger if 285 rights won't be enough. The structure definition looks like this: struct cap_rights { uint64_t cr_rights[CAP_RIGHTS_VERSION + 2]; }; The initial CAP_RIGHTS_VERSION is 0. The top two bits in the first element of the cr_rights[] array contain total number of elements in the array - 2. This means if those two bits are equal to 0, we have 2 array elements. The top two bits in all remaining array elements should be 0. The next five bits in all array elements contain array index. Only one bit is used and bit position in this five-bits range defines array index. This means there can be at most five array elements in the future. To define new right the CAPRIGHT() macro must be used. The macro takes two arguments - an array index and a bit to set, eg. #define CAP_PDKILL CAPRIGHT(1, 0x0000000000000800ULL) We still support aliases that combine few rights, but the rights have to belong to the same array element, eg: #define CAP_LOOKUP CAPRIGHT(0, 0x0000000000000400ULL) #define CAP_FCHMOD CAPRIGHT(0, 0x0000000000002000ULL) #define CAP_FCHMODAT (CAP_FCHMOD | CAP_LOOKUP) There is new API to manage the new cap_rights_t structure: cap_rights_t *cap_rights_init(cap_rights_t *rights, ...); void cap_rights_set(cap_rights_t *rights, ...); void cap_rights_clear(cap_rights_t *rights, ...); bool cap_rights_is_set(const cap_rights_t *rights, ...); bool cap_rights_is_valid(const cap_rights_t *rights); void cap_rights_merge(cap_rights_t *dst, const cap_rights_t *src); void cap_rights_remove(cap_rights_t *dst, const cap_rights_t *src); bool cap_rights_contains(const cap_rights_t *big, const cap_rights_t *little); Capability rights to the cap_rights_init(), cap_rights_set(), cap_rights_clear() and cap_rights_is_set() functions are provided by separating them with commas, eg: cap_rights_t rights; cap_rights_init(&rights, CAP_READ, CAP_WRITE, CAP_FSTAT); There is no need to terminate the list of rights, as those functions are actually macros that take care of the termination, eg: #define cap_rights_set(rights, ...) \ __cap_rights_set((rights), __VA_ARGS__, 0ULL) void __cap_rights_set(cap_rights_t *rights, ...); Thanks to using one bit as an array index we can assert in those functions that there are no two rights belonging to different array elements provided together. For example this is illegal and will be detected, because CAP_LOOKUP belongs to element 0 and CAP_PDKILL to element 1: cap_rights_init(&rights, CAP_LOOKUP | CAP_PDKILL); Providing several rights that belongs to the same array's element this way is correct, but is not advised. It should only be used for aliases definition. This commit also breaks compatibility with some existing Capsicum system calls, but I see no other way to do that. This should be fine as Capsicum is still experimental and this change is not going to 9.x. Sponsored by: The FreeBSD Foundation
2013-09-05 00:09:56 +00:00
error = fget(td, args->fd,
Eliminate the overhead of gratuitous repeated reinitialization of cap_rights - Add macros to allow preinitialization of cap_rights_t. - Convert most commonly used code paths to use preinitialized cap_rights_t. A 3.6% speedup in fstat was measured with this change. Reported by: mjg Reviewed by: oshogbo Approved by: sbruno MFC after: 1 month
2018-05-09 18:47:24 +00:00
&cap_ioctl_rights, &fp);
Change the cap_rights_t type from uint64_t to a structure that we can extend in the future in a backward compatible (API and ABI) way. The cap_rights_t represents capability rights. We used to use one bit to represent one right, but we are running out of spare bits. Currently the new structure provides place for 114 rights (so 50 more than the previous cap_rights_t), but it is possible to grow the structure to hold at least 285 rights, although we can make it even larger if 285 rights won't be enough. The structure definition looks like this: struct cap_rights { uint64_t cr_rights[CAP_RIGHTS_VERSION + 2]; }; The initial CAP_RIGHTS_VERSION is 0. The top two bits in the first element of the cr_rights[] array contain total number of elements in the array - 2. This means if those two bits are equal to 0, we have 2 array elements. The top two bits in all remaining array elements should be 0. The next five bits in all array elements contain array index. Only one bit is used and bit position in this five-bits range defines array index. This means there can be at most five array elements in the future. To define new right the CAPRIGHT() macro must be used. The macro takes two arguments - an array index and a bit to set, eg. #define CAP_PDKILL CAPRIGHT(1, 0x0000000000000800ULL) We still support aliases that combine few rights, but the rights have to belong to the same array element, eg: #define CAP_LOOKUP CAPRIGHT(0, 0x0000000000000400ULL) #define CAP_FCHMOD CAPRIGHT(0, 0x0000000000002000ULL) #define CAP_FCHMODAT (CAP_FCHMOD | CAP_LOOKUP) There is new API to manage the new cap_rights_t structure: cap_rights_t *cap_rights_init(cap_rights_t *rights, ...); void cap_rights_set(cap_rights_t *rights, ...); void cap_rights_clear(cap_rights_t *rights, ...); bool cap_rights_is_set(const cap_rights_t *rights, ...); bool cap_rights_is_valid(const cap_rights_t *rights); void cap_rights_merge(cap_rights_t *dst, const cap_rights_t *src); void cap_rights_remove(cap_rights_t *dst, const cap_rights_t *src); bool cap_rights_contains(const cap_rights_t *big, const cap_rights_t *little); Capability rights to the cap_rights_init(), cap_rights_set(), cap_rights_clear() and cap_rights_is_set() functions are provided by separating them with commas, eg: cap_rights_t rights; cap_rights_init(&rights, CAP_READ, CAP_WRITE, CAP_FSTAT); There is no need to terminate the list of rights, as those functions are actually macros that take care of the termination, eg: #define cap_rights_set(rights, ...) \ __cap_rights_set((rights), __VA_ARGS__, 0ULL) void __cap_rights_set(cap_rights_t *rights, ...); Thanks to using one bit as an array index we can assert in those functions that there are no two rights belonging to different array elements provided together. For example this is illegal and will be detected, because CAP_LOOKUP belongs to element 0 and CAP_PDKILL to element 1: cap_rights_init(&rights, CAP_LOOKUP | CAP_PDKILL); Providing several rights that belongs to the same array's element this way is correct, but is not advised. It should only be used for aliases definition. This commit also breaks compatibility with some existing Capsicum system calls, but I see no other way to do that. This should be fine as Capsicum is still experimental and this change is not going to 9.x. Sponsored by: The FreeBSD Foundation
2013-09-05 00:09:56 +00:00
if (error != 0)
This is v4l support for the linuxulator. This allows to access FreeBSD native devices which support the v4l API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd driver. Not tested is firmware upload, framebuffer stuff and video tuner stuff due to lack of hardware. The clipping part (VIDIOCSWIN) needs a little bit of further work (partly in progress, but can not be tested due to lack of a suitable device). The submitter tested this sucessfully with Skype and flash apps on amd64 and i386 with the multimedia/pwcbsd driver. Submitted by: J.R. Oldroyd <fbsd@opal.com>
2009-12-04 21:06:54 +00:00
return (error);
error = copyin((void *) args->arg, &l_vcode, sizeof(l_vcode));
if (error) {
fdrop(fp, td);
return (error);
}
linux_to_bsd_v4l_code(&l_vcode, &vcode);
- #ifdef out the cliplist part, skype seems like using an uninitialized variable and can cause problems, without the cliplist handling it works without problems - improve the cliplist error handling - fix VIDIOCGTUNER and VIDIOCSMICROCODE (still no hardware available to test) Submitted by: J.R. Oldroyd <jr@opal.com> X-MFC after: soon (together with all the v4l stuff)
2010-05-03 14:19:58 +00:00
error = fo_ioctl(fp, VIDIOCSMICROCODE, &vcode, td->td_ucred, td);
This is v4l support for the linuxulator. This allows to access FreeBSD native devices which support the v4l API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd driver. Not tested is firmware upload, framebuffer stuff and video tuner stuff due to lack of hardware. The clipping part (VIDIOCSWIN) needs a little bit of further work (partly in progress, but can not be tested due to lack of a suitable device). The submitter tested this sucessfully with Skype and flash apps on amd64 and i386 with the multimedia/pwcbsd driver. Submitted by: J.R. Oldroyd <fbsd@opal.com>
2009-12-04 21:06:54 +00:00
fdrop(fp, td);
return (error);
case LINUX_VIDIOCGVBIFMT: args->cmd = VIDIOCGVBIFMT; break;
case LINUX_VIDIOCSVBIFMT: args->cmd = VIDIOCSVBIFMT; break;
default: return (ENOIOCTL);
}
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = sys_ioctl(td, (struct ioctl_args *)args);
This is v4l support for the linuxulator. This allows to access FreeBSD native devices which support the v4l API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd driver. Not tested is firmware upload, framebuffer stuff and video tuner stuff due to lack of hardware. The clipping part (VIDIOCSWIN) needs a little bit of further work (partly in progress, but can not be tested due to lack of a suitable device). The submitter tested this sucessfully with Skype and flash apps on amd64 and i386 with the multimedia/pwcbsd driver. Submitted by: J.R. Oldroyd <fbsd@opal.com>
2009-12-04 21:06:54 +00:00
return (error);
}
Deal with a few issues that cropped up following the recent changes to the code for translating socket and private ioctls: - Only perform socket ioctl translation if the file descriptor is a socket. - Treat socket ioctls on non-sockets specially, and for now assume that these are directed at a tap/vmnet device, so translate the ioctl numbers as appropriate (the way if_tap abuses some socket ioctls to pass non-ifreq data is utterly bogus, but this is how VMware on FreeBSD has always "worked"; I will deal with this later). - Add (untested) support for translating SIOCSIFADDR. - In all cases where we fail to translate an ioctl, return ENOIOCTL so that other handlers have a chance to do the translation. This should fix the "/dev/vmnet1: Invalid argument" errors that users of VMware were experiencing, though I have only verified this on RELENG_4. Submitted by: des (mostly) MFC after: 3 days
2001-11-19 15:43:50 +00:00
/*
* Special ioctl handler
*/
static int
linux_ioctl_special(struct thread *td, struct linux_ioctl_args *args)
{
int error;
switch (args->cmd) {
case LINUX_SIOCGIFADDR:
args->cmd = SIOCGIFADDR;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = sys_ioctl(td, (struct ioctl_args *)args);
Deal with a few issues that cropped up following the recent changes to the code for translating socket and private ioctls: - Only perform socket ioctl translation if the file descriptor is a socket. - Treat socket ioctls on non-sockets specially, and for now assume that these are directed at a tap/vmnet device, so translate the ioctl numbers as appropriate (the way if_tap abuses some socket ioctls to pass non-ifreq data is utterly bogus, but this is how VMware on FreeBSD has always "worked"; I will deal with this later). - Add (untested) support for translating SIOCSIFADDR. - In all cases where we fail to translate an ioctl, return ENOIOCTL so that other handlers have a chance to do the translation. This should fix the "/dev/vmnet1: Invalid argument" errors that users of VMware were experiencing, though I have only verified this on RELENG_4. Submitted by: des (mostly) MFC after: 3 days
2001-11-19 15:43:50 +00:00
break;
case LINUX_SIOCSIFADDR:
args->cmd = SIOCSIFADDR;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = sys_ioctl(td, (struct ioctl_args *)args);
Deal with a few issues that cropped up following the recent changes to the code for translating socket and private ioctls: - Only perform socket ioctl translation if the file descriptor is a socket. - Treat socket ioctls on non-sockets specially, and for now assume that these are directed at a tap/vmnet device, so translate the ioctl numbers as appropriate (the way if_tap abuses some socket ioctls to pass non-ifreq data is utterly bogus, but this is how VMware on FreeBSD has always "worked"; I will deal with this later). - Add (untested) support for translating SIOCSIFADDR. - In all cases where we fail to translate an ioctl, return ENOIOCTL so that other handlers have a chance to do the translation. This should fix the "/dev/vmnet1: Invalid argument" errors that users of VMware were experiencing, though I have only verified this on RELENG_4. Submitted by: des (mostly) MFC after: 3 days
2001-11-19 15:43:50 +00:00
break;
case LINUX_SIOCGIFFLAGS:
args->cmd = SIOCGIFFLAGS;
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = sys_ioctl(td, (struct ioctl_args *)args);
Deal with a few issues that cropped up following the recent changes to the code for translating socket and private ioctls: - Only perform socket ioctl translation if the file descriptor is a socket. - Treat socket ioctls on non-sockets specially, and for now assume that these are directed at a tap/vmnet device, so translate the ioctl numbers as appropriate (the way if_tap abuses some socket ioctls to pass non-ifreq data is utterly bogus, but this is how VMware on FreeBSD has always "worked"; I will deal with this later). - Add (untested) support for translating SIOCSIFADDR. - In all cases where we fail to translate an ioctl, return ENOIOCTL so that other handlers have a chance to do the translation. This should fix the "/dev/vmnet1: Invalid argument" errors that users of VMware were experiencing, though I have only verified this on RELENG_4. Submitted by: des (mostly) MFC after: 3 days
2001-11-19 15:43:50 +00:00
break;
default:
error = ENOIOCTL;
}
return (error);
Add support for the "device private" ioctls soon to be used by the an driver. Also slightly change the name translation policy - only rename interfaces that have the IFF_BROADCAST flag set. This is not perfect, but is closer to how Linux names network interfaces.
2001-10-19 01:38:10 +00:00
}
This is v4l2 support for the linuxulator. This allows to access FreeBSD native devices which support the v4l2 API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd or multimedia/webcamd supplied drivers. Submitted by: nox MFC after: 1 month
2011-05-04 09:05:39 +00:00
static int
linux_to_bsd_v4l2_standard(struct l_v4l2_standard *lvstd, struct v4l2_standard *vstd)
{
vstd->index = lvstd->index;
vstd->id = lvstd->id;
linux_ioctl: Refactor some v4l2 struct converters According to the C standard, it is invalid to copy beyond the end of an object, even if that object is obviously a member of a larger object (a struct, in this case). Appease the standard and Coverity by refactoring the copy in a straightforward way. No functional change. Reported by: Coverity (CWE-120) CIDs: 1007819, 1007820, 1007821, 1007822, 1009668, 1009669 Security: no (false positive detection) Sponsored by: Dell EMC Isilon
2017-04-13 17:34:51 +00:00
CTASSERT(sizeof(vstd->name) == sizeof(lvstd->name));
memcpy(vstd->name, lvstd->name, sizeof(vstd->name));
vstd->frameperiod = lvstd->frameperiod;
vstd->framelines = lvstd->framelines;
CTASSERT(sizeof(vstd->reserved) == sizeof(lvstd->reserved));
memcpy(vstd->reserved, lvstd->reserved, sizeof(vstd->reserved));
This is v4l2 support for the linuxulator. This allows to access FreeBSD native devices which support the v4l2 API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd or multimedia/webcamd supplied drivers. Submitted by: nox MFC after: 1 month
2011-05-04 09:05:39 +00:00
return (0);
}
static int
bsd_to_linux_v4l2_standard(struct v4l2_standard *vstd, struct l_v4l2_standard *lvstd)
{
lvstd->index = vstd->index;
lvstd->id = vstd->id;
linux_ioctl: Refactor some v4l2 struct converters According to the C standard, it is invalid to copy beyond the end of an object, even if that object is obviously a member of a larger object (a struct, in this case). Appease the standard and Coverity by refactoring the copy in a straightforward way. No functional change. Reported by: Coverity (CWE-120) CIDs: 1007819, 1007820, 1007821, 1007822, 1009668, 1009669 Security: no (false positive detection) Sponsored by: Dell EMC Isilon
2017-04-13 17:34:51 +00:00
CTASSERT(sizeof(vstd->name) == sizeof(lvstd->name));
memcpy(lvstd->name, vstd->name, sizeof(lvstd->name));
lvstd->frameperiod = vstd->frameperiod;
lvstd->framelines = vstd->framelines;
CTASSERT(sizeof(vstd->reserved) == sizeof(lvstd->reserved));
memcpy(lvstd->reserved, vstd->reserved, sizeof(lvstd->reserved));
This is v4l2 support for the linuxulator. This allows to access FreeBSD native devices which support the v4l2 API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd or multimedia/webcamd supplied drivers. Submitted by: nox MFC after: 1 month
2011-05-04 09:05:39 +00:00
return (0);
}
static int
linux_to_bsd_v4l2_buffer(struct l_v4l2_buffer *lvb, struct v4l2_buffer *vb)
{
vb->index = lvb->index;
vb->type = lvb->type;
vb->bytesused = lvb->bytesused;
vb->flags = lvb->flags;
vb->field = lvb->field;
vb->timestamp.tv_sec = lvb->timestamp.tv_sec;
vb->timestamp.tv_usec = lvb->timestamp.tv_usec;
memcpy(&vb->timecode, &lvb->timecode, sizeof (lvb->timecode));
vb->sequence = lvb->sequence;
vb->memory = lvb->memory;
if (lvb->memory == V4L2_MEMORY_USERPTR)
/* possible pointer size conversion */
vb->m.userptr = (unsigned long)PTRIN(lvb->m.userptr);
else
vb->m.offset = lvb->m.offset;
vb->length = lvb->length;
vb->input = lvb->input;
vb->reserved = lvb->reserved;
return (0);
}
static int
bsd_to_linux_v4l2_buffer(struct v4l2_buffer *vb, struct l_v4l2_buffer *lvb)
{
lvb->index = vb->index;
lvb->type = vb->type;
lvb->bytesused = vb->bytesused;
lvb->flags = vb->flags;
lvb->field = vb->field;
lvb->timestamp.tv_sec = vb->timestamp.tv_sec;
lvb->timestamp.tv_usec = vb->timestamp.tv_usec;
memcpy(&lvb->timecode, &vb->timecode, sizeof (vb->timecode));
lvb->sequence = vb->sequence;
lvb->memory = vb->memory;
if (vb->memory == V4L2_MEMORY_USERPTR)
/* possible pointer size conversion */
lvb->m.userptr = PTROUT(vb->m.userptr);
else
lvb->m.offset = vb->m.offset;
lvb->length = vb->length;
lvb->input = vb->input;
lvb->reserved = vb->reserved;
return (0);
}
static int
linux_to_bsd_v4l2_format(struct l_v4l2_format *lvf, struct v4l2_format *vf)
{
vf->type = lvf->type;
if (lvf->type == V4L2_BUF_TYPE_VIDEO_OVERLAY
#ifdef V4L2_BUF_TYPE_VIDEO_OUTPUT_OVERLAY
|| lvf->type == V4L2_BUF_TYPE_VIDEO_OUTPUT_OVERLAY
#endif
)
/*
* XXX TODO - needs 32 -> 64 bit conversion:
* (unused by webcams?)
*/
Linuxulator: apply style(9) to return Sponsored by: Turing Robotic Industries Inc.
2018-03-12 15:35:24 +00:00
return (EINVAL);
This is v4l2 support for the linuxulator. This allows to access FreeBSD native devices which support the v4l2 API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd or multimedia/webcamd supplied drivers. Submitted by: nox MFC after: 1 month
2011-05-04 09:05:39 +00:00
memcpy(&vf->fmt, &lvf->fmt, sizeof(vf->fmt));
Linuxulator: apply style(9) to return Sponsored by: Turing Robotic Industries Inc.
2018-03-12 15:35:24 +00:00
return (0);
This is v4l2 support for the linuxulator. This allows to access FreeBSD native devices which support the v4l2 API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd or multimedia/webcamd supplied drivers. Submitted by: nox MFC after: 1 month
2011-05-04 09:05:39 +00:00
}
static int
bsd_to_linux_v4l2_format(struct v4l2_format *vf, struct l_v4l2_format *lvf)
{
lvf->type = vf->type;
if (vf->type == V4L2_BUF_TYPE_VIDEO_OVERLAY
#ifdef V4L2_BUF_TYPE_VIDEO_OUTPUT_OVERLAY
|| vf->type == V4L2_BUF_TYPE_VIDEO_OUTPUT_OVERLAY
#endif
)
/*
* XXX TODO - needs 32 -> 64 bit conversion:
* (unused by webcams?)
*/
Linuxulator: apply style(9) to return Sponsored by: Turing Robotic Industries Inc.
2018-03-12 15:35:24 +00:00
return (EINVAL);
This is v4l2 support for the linuxulator. This allows to access FreeBSD native devices which support the v4l2 API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd or multimedia/webcamd supplied drivers. Submitted by: nox MFC after: 1 month
2011-05-04 09:05:39 +00:00
memcpy(&lvf->fmt, &vf->fmt, sizeof(vf->fmt));
Linuxulator: apply style(9) to return Sponsored by: Turing Robotic Industries Inc.
2018-03-12 15:35:24 +00:00
return (0);
This is v4l2 support for the linuxulator. This allows to access FreeBSD native devices which support the v4l2 API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd or multimedia/webcamd supplied drivers. Submitted by: nox MFC after: 1 month
2011-05-04 09:05:39 +00:00
}
static int
linux_ioctl_v4l2(struct thread *td, struct linux_ioctl_args *args)
{
struct file *fp;
int error;
struct v4l2_format vformat;
struct l_v4l2_format l_vformat;
struct v4l2_standard vstd;
struct l_v4l2_standard l_vstd;
struct l_v4l2_buffer l_vbuf;
struct v4l2_buffer vbuf;
struct v4l2_input vinp;
switch (args->cmd & 0xffff) {
case LINUX_VIDIOC_RESERVED:
case LINUX_VIDIOC_LOG_STATUS:
if ((args->cmd & IOC_DIRMASK) != LINUX_IOC_VOID)
Linuxulator: apply style(9) to return Sponsored by: Turing Robotic Industries Inc.
2018-03-12 15:35:24 +00:00
return (ENOIOCTL);
This is v4l2 support for the linuxulator. This allows to access FreeBSD native devices which support the v4l2 API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd or multimedia/webcamd supplied drivers. Submitted by: nox MFC after: 1 month
2011-05-04 09:05:39 +00:00
args->cmd = (args->cmd & 0xffff) | IOC_VOID;
break;
case LINUX_VIDIOC_OVERLAY:
case LINUX_VIDIOC_STREAMON:
case LINUX_VIDIOC_STREAMOFF:
case LINUX_VIDIOC_S_STD:
case LINUX_VIDIOC_S_TUNER:
case LINUX_VIDIOC_S_AUDIO:
case LINUX_VIDIOC_S_AUDOUT:
case LINUX_VIDIOC_S_MODULATOR:
case LINUX_VIDIOC_S_FREQUENCY:
case LINUX_VIDIOC_S_CROP:
case LINUX_VIDIOC_S_JPEGCOMP:
case LINUX_VIDIOC_S_PRIORITY:
case LINUX_VIDIOC_DBG_S_REGISTER:
case LINUX_VIDIOC_S_HW_FREQ_SEEK:
case LINUX_VIDIOC_SUBSCRIBE_EVENT:
case LINUX_VIDIOC_UNSUBSCRIBE_EVENT:
args->cmd = (args->cmd & ~IOC_DIRMASK) | IOC_IN;
break;
case LINUX_VIDIOC_QUERYCAP:
case LINUX_VIDIOC_G_STD:
case LINUX_VIDIOC_G_AUDIO:
case LINUX_VIDIOC_G_INPUT:
case LINUX_VIDIOC_G_OUTPUT:
case LINUX_VIDIOC_G_AUDOUT:
case LINUX_VIDIOC_G_JPEGCOMP:
case LINUX_VIDIOC_QUERYSTD:
case LINUX_VIDIOC_G_PRIORITY:
case LINUX_VIDIOC_QUERY_DV_PRESET:
args->cmd = (args->cmd & ~IOC_DIRMASK) | IOC_OUT;
break;
case LINUX_VIDIOC_ENUM_FMT:
case LINUX_VIDIOC_REQBUFS:
case LINUX_VIDIOC_G_PARM:
case LINUX_VIDIOC_S_PARM:
case LINUX_VIDIOC_G_CTRL:
case LINUX_VIDIOC_S_CTRL:
case LINUX_VIDIOC_G_TUNER:
case LINUX_VIDIOC_QUERYCTRL:
case LINUX_VIDIOC_QUERYMENU:
case LINUX_VIDIOC_S_INPUT:
case LINUX_VIDIOC_S_OUTPUT:
case LINUX_VIDIOC_ENUMOUTPUT:
case LINUX_VIDIOC_G_MODULATOR:
case LINUX_VIDIOC_G_FREQUENCY:
case LINUX_VIDIOC_CROPCAP:
case LINUX_VIDIOC_G_CROP:
case LINUX_VIDIOC_ENUMAUDIO:
case LINUX_VIDIOC_ENUMAUDOUT:
case LINUX_VIDIOC_G_SLICED_VBI_CAP:
#ifdef VIDIOC_ENUM_FRAMESIZES
case LINUX_VIDIOC_ENUM_FRAMESIZES:
case LINUX_VIDIOC_ENUM_FRAMEINTERVALS:
case LINUX_VIDIOC_ENCODER_CMD:
case LINUX_VIDIOC_TRY_ENCODER_CMD:
#endif
case LINUX_VIDIOC_DBG_G_REGISTER:
case LINUX_VIDIOC_DBG_G_CHIP_IDENT:
case LINUX_VIDIOC_ENUM_DV_PRESETS:
case LINUX_VIDIOC_S_DV_PRESET:
case LINUX_VIDIOC_G_DV_PRESET:
case LINUX_VIDIOC_S_DV_TIMINGS:
case LINUX_VIDIOC_G_DV_TIMINGS:
args->cmd = (args->cmd & ~IOC_DIRMASK) | IOC_INOUT;
break;
case LINUX_VIDIOC_G_FMT:
case LINUX_VIDIOC_S_FMT:
case LINUX_VIDIOC_TRY_FMT:
error = copyin((void *)args->arg, &l_vformat, sizeof(l_vformat));
if (error)
return (error);
Change the cap_rights_t type from uint64_t to a structure that we can extend in the future in a backward compatible (API and ABI) way. The cap_rights_t represents capability rights. We used to use one bit to represent one right, but we are running out of spare bits. Currently the new structure provides place for 114 rights (so 50 more than the previous cap_rights_t), but it is possible to grow the structure to hold at least 285 rights, although we can make it even larger if 285 rights won't be enough. The structure definition looks like this: struct cap_rights { uint64_t cr_rights[CAP_RIGHTS_VERSION + 2]; }; The initial CAP_RIGHTS_VERSION is 0. The top two bits in the first element of the cr_rights[] array contain total number of elements in the array - 2. This means if those two bits are equal to 0, we have 2 array elements. The top two bits in all remaining array elements should be 0. The next five bits in all array elements contain array index. Only one bit is used and bit position in this five-bits range defines array index. This means there can be at most five array elements in the future. To define new right the CAPRIGHT() macro must be used. The macro takes two arguments - an array index and a bit to set, eg. #define CAP_PDKILL CAPRIGHT(1, 0x0000000000000800ULL) We still support aliases that combine few rights, but the rights have to belong to the same array element, eg: #define CAP_LOOKUP CAPRIGHT(0, 0x0000000000000400ULL) #define CAP_FCHMOD CAPRIGHT(0, 0x0000000000002000ULL) #define CAP_FCHMODAT (CAP_FCHMOD | CAP_LOOKUP) There is new API to manage the new cap_rights_t structure: cap_rights_t *cap_rights_init(cap_rights_t *rights, ...); void cap_rights_set(cap_rights_t *rights, ...); void cap_rights_clear(cap_rights_t *rights, ...); bool cap_rights_is_set(const cap_rights_t *rights, ...); bool cap_rights_is_valid(const cap_rights_t *rights); void cap_rights_merge(cap_rights_t *dst, const cap_rights_t *src); void cap_rights_remove(cap_rights_t *dst, const cap_rights_t *src); bool cap_rights_contains(const cap_rights_t *big, const cap_rights_t *little); Capability rights to the cap_rights_init(), cap_rights_set(), cap_rights_clear() and cap_rights_is_set() functions are provided by separating them with commas, eg: cap_rights_t rights; cap_rights_init(&rights, CAP_READ, CAP_WRITE, CAP_FSTAT); There is no need to terminate the list of rights, as those functions are actually macros that take care of the termination, eg: #define cap_rights_set(rights, ...) \ __cap_rights_set((rights), __VA_ARGS__, 0ULL) void __cap_rights_set(cap_rights_t *rights, ...); Thanks to using one bit as an array index we can assert in those functions that there are no two rights belonging to different array elements provided together. For example this is illegal and will be detected, because CAP_LOOKUP belongs to element 0 and CAP_PDKILL to element 1: cap_rights_init(&rights, CAP_LOOKUP | CAP_PDKILL); Providing several rights that belongs to the same array's element this way is correct, but is not advised. It should only be used for aliases definition. This commit also breaks compatibility with some existing Capsicum system calls, but I see no other way to do that. This should be fine as Capsicum is still experimental and this change is not going to 9.x. Sponsored by: The FreeBSD Foundation
2013-09-05 00:09:56 +00:00
error = fget(td, args->fd,
Eliminate the overhead of gratuitous repeated reinitialization of cap_rights - Add macros to allow preinitialization of cap_rights_t. - Convert most commonly used code paths to use preinitialized cap_rights_t. A 3.6% speedup in fstat was measured with this change. Reported by: mjg Reviewed by: oshogbo Approved by: sbruno MFC after: 1 month
2018-05-09 18:47:24 +00:00
&cap_ioctl_rights, &fp);
Change the cap_rights_t type from uint64_t to a structure that we can extend in the future in a backward compatible (API and ABI) way. The cap_rights_t represents capability rights. We used to use one bit to represent one right, but we are running out of spare bits. Currently the new structure provides place for 114 rights (so 50 more than the previous cap_rights_t), but it is possible to grow the structure to hold at least 285 rights, although we can make it even larger if 285 rights won't be enough. The structure definition looks like this: struct cap_rights { uint64_t cr_rights[CAP_RIGHTS_VERSION + 2]; }; The initial CAP_RIGHTS_VERSION is 0. The top two bits in the first element of the cr_rights[] array contain total number of elements in the array - 2. This means if those two bits are equal to 0, we have 2 array elements. The top two bits in all remaining array elements should be 0. The next five bits in all array elements contain array index. Only one bit is used and bit position in this five-bits range defines array index. This means there can be at most five array elements in the future. To define new right the CAPRIGHT() macro must be used. The macro takes two arguments - an array index and a bit to set, eg. #define CAP_PDKILL CAPRIGHT(1, 0x0000000000000800ULL) We still support aliases that combine few rights, but the rights have to belong to the same array element, eg: #define CAP_LOOKUP CAPRIGHT(0, 0x0000000000000400ULL) #define CAP_FCHMOD CAPRIGHT(0, 0x0000000000002000ULL) #define CAP_FCHMODAT (CAP_FCHMOD | CAP_LOOKUP) There is new API to manage the new cap_rights_t structure: cap_rights_t *cap_rights_init(cap_rights_t *rights, ...); void cap_rights_set(cap_rights_t *rights, ...); void cap_rights_clear(cap_rights_t *rights, ...); bool cap_rights_is_set(const cap_rights_t *rights, ...); bool cap_rights_is_valid(const cap_rights_t *rights); void cap_rights_merge(cap_rights_t *dst, const cap_rights_t *src); void cap_rights_remove(cap_rights_t *dst, const cap_rights_t *src); bool cap_rights_contains(const cap_rights_t *big, const cap_rights_t *little); Capability rights to the cap_rights_init(), cap_rights_set(), cap_rights_clear() and cap_rights_is_set() functions are provided by separating them with commas, eg: cap_rights_t rights; cap_rights_init(&rights, CAP_READ, CAP_WRITE, CAP_FSTAT); There is no need to terminate the list of rights, as those functions are actually macros that take care of the termination, eg: #define cap_rights_set(rights, ...) \ __cap_rights_set((rights), __VA_ARGS__, 0ULL) void __cap_rights_set(cap_rights_t *rights, ...); Thanks to using one bit as an array index we can assert in those functions that there are no two rights belonging to different array elements provided together. For example this is illegal and will be detected, because CAP_LOOKUP belongs to element 0 and CAP_PDKILL to element 1: cap_rights_init(&rights, CAP_LOOKUP | CAP_PDKILL); Providing several rights that belongs to the same array's element this way is correct, but is not advised. It should only be used for aliases definition. This commit also breaks compatibility with some existing Capsicum system calls, but I see no other way to do that. This should be fine as Capsicum is still experimental and this change is not going to 9.x. Sponsored by: The FreeBSD Foundation
2013-09-05 00:09:56 +00:00
if (error)
This is v4l2 support for the linuxulator. This allows to access FreeBSD native devices which support the v4l2 API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd or multimedia/webcamd supplied drivers. Submitted by: nox MFC after: 1 month
2011-05-04 09:05:39 +00:00
return (error);
if (linux_to_bsd_v4l2_format(&l_vformat, &vformat) != 0)
error = EINVAL;
else if ((args->cmd & 0xffff) == LINUX_VIDIOC_G_FMT)
error = fo_ioctl(fp, VIDIOC_G_FMT, &vformat,
td->td_ucred, td);
else if ((args->cmd & 0xffff) == LINUX_VIDIOC_S_FMT)
error = fo_ioctl(fp, VIDIOC_S_FMT, &vformat,
td->td_ucred, td);
else
error = fo_ioctl(fp, VIDIOC_TRY_FMT, &vformat,
td->td_ucred, td);
bsd_to_linux_v4l2_format(&vformat, &l_vformat);
copyout(&l_vformat, (void *)args->arg, sizeof(l_vformat));
fdrop(fp, td);
return (error);
case LINUX_VIDIOC_ENUMSTD:
error = copyin((void *)args->arg, &l_vstd, sizeof(l_vstd));
if (error)
return (error);
linux_to_bsd_v4l2_standard(&l_vstd, &vstd);
Change the cap_rights_t type from uint64_t to a structure that we can extend in the future in a backward compatible (API and ABI) way. The cap_rights_t represents capability rights. We used to use one bit to represent one right, but we are running out of spare bits. Currently the new structure provides place for 114 rights (so 50 more than the previous cap_rights_t), but it is possible to grow the structure to hold at least 285 rights, although we can make it even larger if 285 rights won't be enough. The structure definition looks like this: struct cap_rights { uint64_t cr_rights[CAP_RIGHTS_VERSION + 2]; }; The initial CAP_RIGHTS_VERSION is 0. The top two bits in the first element of the cr_rights[] array contain total number of elements in the array - 2. This means if those two bits are equal to 0, we have 2 array elements. The top two bits in all remaining array elements should be 0. The next five bits in all array elements contain array index. Only one bit is used and bit position in this five-bits range defines array index. This means there can be at most five array elements in the future. To define new right the CAPRIGHT() macro must be used. The macro takes two arguments - an array index and a bit to set, eg. #define CAP_PDKILL CAPRIGHT(1, 0x0000000000000800ULL) We still support aliases that combine few rights, but the rights have to belong to the same array element, eg: #define CAP_LOOKUP CAPRIGHT(0, 0x0000000000000400ULL) #define CAP_FCHMOD CAPRIGHT(0, 0x0000000000002000ULL) #define CAP_FCHMODAT (CAP_FCHMOD | CAP_LOOKUP) There is new API to manage the new cap_rights_t structure: cap_rights_t *cap_rights_init(cap_rights_t *rights, ...); void cap_rights_set(cap_rights_t *rights, ...); void cap_rights_clear(cap_rights_t *rights, ...); bool cap_rights_is_set(const cap_rights_t *rights, ...); bool cap_rights_is_valid(const cap_rights_t *rights); void cap_rights_merge(cap_rights_t *dst, const cap_rights_t *src); void cap_rights_remove(cap_rights_t *dst, const cap_rights_t *src); bool cap_rights_contains(const cap_rights_t *big, const cap_rights_t *little); Capability rights to the cap_rights_init(), cap_rights_set(), cap_rights_clear() and cap_rights_is_set() functions are provided by separating them with commas, eg: cap_rights_t rights; cap_rights_init(&rights, CAP_READ, CAP_WRITE, CAP_FSTAT); There is no need to terminate the list of rights, as those functions are actually macros that take care of the termination, eg: #define cap_rights_set(rights, ...) \ __cap_rights_set((rights), __VA_ARGS__, 0ULL) void __cap_rights_set(cap_rights_t *rights, ...); Thanks to using one bit as an array index we can assert in those functions that there are no two rights belonging to different array elements provided together. For example this is illegal and will be detected, because CAP_LOOKUP belongs to element 0 and CAP_PDKILL to element 1: cap_rights_init(&rights, CAP_LOOKUP | CAP_PDKILL); Providing several rights that belongs to the same array's element this way is correct, but is not advised. It should only be used for aliases definition. This commit also breaks compatibility with some existing Capsicum system calls, but I see no other way to do that. This should be fine as Capsicum is still experimental and this change is not going to 9.x. Sponsored by: The FreeBSD Foundation
2013-09-05 00:09:56 +00:00
error = fget(td, args->fd,
Eliminate the overhead of gratuitous repeated reinitialization of cap_rights - Add macros to allow preinitialization of cap_rights_t. - Convert most commonly used code paths to use preinitialized cap_rights_t. A 3.6% speedup in fstat was measured with this change. Reported by: mjg Reviewed by: oshogbo Approved by: sbruno MFC after: 1 month
2018-05-09 18:47:24 +00:00
&cap_ioctl_rights, &fp);
Change the cap_rights_t type from uint64_t to a structure that we can extend in the future in a backward compatible (API and ABI) way. The cap_rights_t represents capability rights. We used to use one bit to represent one right, but we are running out of spare bits. Currently the new structure provides place for 114 rights (so 50 more than the previous cap_rights_t), but it is possible to grow the structure to hold at least 285 rights, although we can make it even larger if 285 rights won't be enough. The structure definition looks like this: struct cap_rights { uint64_t cr_rights[CAP_RIGHTS_VERSION + 2]; }; The initial CAP_RIGHTS_VERSION is 0. The top two bits in the first element of the cr_rights[] array contain total number of elements in the array - 2. This means if those two bits are equal to 0, we have 2 array elements. The top two bits in all remaining array elements should be 0. The next five bits in all array elements contain array index. Only one bit is used and bit position in this five-bits range defines array index. This means there can be at most five array elements in the future. To define new right the CAPRIGHT() macro must be used. The macro takes two arguments - an array index and a bit to set, eg. #define CAP_PDKILL CAPRIGHT(1, 0x0000000000000800ULL) We still support aliases that combine few rights, but the rights have to belong to the same array element, eg: #define CAP_LOOKUP CAPRIGHT(0, 0x0000000000000400ULL) #define CAP_FCHMOD CAPRIGHT(0, 0x0000000000002000ULL) #define CAP_FCHMODAT (CAP_FCHMOD | CAP_LOOKUP) There is new API to manage the new cap_rights_t structure: cap_rights_t *cap_rights_init(cap_rights_t *rights, ...); void cap_rights_set(cap_rights_t *rights, ...); void cap_rights_clear(cap_rights_t *rights, ...); bool cap_rights_is_set(const cap_rights_t *rights, ...); bool cap_rights_is_valid(const cap_rights_t *rights); void cap_rights_merge(cap_rights_t *dst, const cap_rights_t *src); void cap_rights_remove(cap_rights_t *dst, const cap_rights_t *src); bool cap_rights_contains(const cap_rights_t *big, const cap_rights_t *little); Capability rights to the cap_rights_init(), cap_rights_set(), cap_rights_clear() and cap_rights_is_set() functions are provided by separating them with commas, eg: cap_rights_t rights; cap_rights_init(&rights, CAP_READ, CAP_WRITE, CAP_FSTAT); There is no need to terminate the list of rights, as those functions are actually macros that take care of the termination, eg: #define cap_rights_set(rights, ...) \ __cap_rights_set((rights), __VA_ARGS__, 0ULL) void __cap_rights_set(cap_rights_t *rights, ...); Thanks to using one bit as an array index we can assert in those functions that there are no two rights belonging to different array elements provided together. For example this is illegal and will be detected, because CAP_LOOKUP belongs to element 0 and CAP_PDKILL to element 1: cap_rights_init(&rights, CAP_LOOKUP | CAP_PDKILL); Providing several rights that belongs to the same array's element this way is correct, but is not advised. It should only be used for aliases definition. This commit also breaks compatibility with some existing Capsicum system calls, but I see no other way to do that. This should be fine as Capsicum is still experimental and this change is not going to 9.x. Sponsored by: The FreeBSD Foundation
2013-09-05 00:09:56 +00:00
if (error)
This is v4l2 support for the linuxulator. This allows to access FreeBSD native devices which support the v4l2 API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd or multimedia/webcamd supplied drivers. Submitted by: nox MFC after: 1 month
2011-05-04 09:05:39 +00:00
return (error);
error = fo_ioctl(fp, VIDIOC_ENUMSTD, (caddr_t)&vstd,
td->td_ucred, td);
if (error) {
fdrop(fp, td);
return (error);
}
bsd_to_linux_v4l2_standard(&vstd, &l_vstd);
error = copyout(&l_vstd, (void *)args->arg, sizeof(l_vstd));
fdrop(fp, td);
return (error);
case LINUX_VIDIOC_ENUMINPUT:
/*
* The Linux struct l_v4l2_input differs only in size,
* it has no padding at the end.
*/
error = copyin((void *)args->arg, &vinp,
sizeof(struct l_v4l2_input));
if (error != 0)
return (error);
Change the cap_rights_t type from uint64_t to a structure that we can extend in the future in a backward compatible (API and ABI) way. The cap_rights_t represents capability rights. We used to use one bit to represent one right, but we are running out of spare bits. Currently the new structure provides place for 114 rights (so 50 more than the previous cap_rights_t), but it is possible to grow the structure to hold at least 285 rights, although we can make it even larger if 285 rights won't be enough. The structure definition looks like this: struct cap_rights { uint64_t cr_rights[CAP_RIGHTS_VERSION + 2]; }; The initial CAP_RIGHTS_VERSION is 0. The top two bits in the first element of the cr_rights[] array contain total number of elements in the array - 2. This means if those two bits are equal to 0, we have 2 array elements. The top two bits in all remaining array elements should be 0. The next five bits in all array elements contain array index. Only one bit is used and bit position in this five-bits range defines array index. This means there can be at most five array elements in the future. To define new right the CAPRIGHT() macro must be used. The macro takes two arguments - an array index and a bit to set, eg. #define CAP_PDKILL CAPRIGHT(1, 0x0000000000000800ULL) We still support aliases that combine few rights, but the rights have to belong to the same array element, eg: #define CAP_LOOKUP CAPRIGHT(0, 0x0000000000000400ULL) #define CAP_FCHMOD CAPRIGHT(0, 0x0000000000002000ULL) #define CAP_FCHMODAT (CAP_FCHMOD | CAP_LOOKUP) There is new API to manage the new cap_rights_t structure: cap_rights_t *cap_rights_init(cap_rights_t *rights, ...); void cap_rights_set(cap_rights_t *rights, ...); void cap_rights_clear(cap_rights_t *rights, ...); bool cap_rights_is_set(const cap_rights_t *rights, ...); bool cap_rights_is_valid(const cap_rights_t *rights); void cap_rights_merge(cap_rights_t *dst, const cap_rights_t *src); void cap_rights_remove(cap_rights_t *dst, const cap_rights_t *src); bool cap_rights_contains(const cap_rights_t *big, const cap_rights_t *little); Capability rights to the cap_rights_init(), cap_rights_set(), cap_rights_clear() and cap_rights_is_set() functions are provided by separating them with commas, eg: cap_rights_t rights; cap_rights_init(&rights, CAP_READ, CAP_WRITE, CAP_FSTAT); There is no need to terminate the list of rights, as those functions are actually macros that take care of the termination, eg: #define cap_rights_set(rights, ...) \ __cap_rights_set((rights), __VA_ARGS__, 0ULL) void __cap_rights_set(cap_rights_t *rights, ...); Thanks to using one bit as an array index we can assert in those functions that there are no two rights belonging to different array elements provided together. For example this is illegal and will be detected, because CAP_LOOKUP belongs to element 0 and CAP_PDKILL to element 1: cap_rights_init(&rights, CAP_LOOKUP | CAP_PDKILL); Providing several rights that belongs to the same array's element this way is correct, but is not advised. It should only be used for aliases definition. This commit also breaks compatibility with some existing Capsicum system calls, but I see no other way to do that. This should be fine as Capsicum is still experimental and this change is not going to 9.x. Sponsored by: The FreeBSD Foundation
2013-09-05 00:09:56 +00:00
error = fget(td, args->fd,
Eliminate the overhead of gratuitous repeated reinitialization of cap_rights - Add macros to allow preinitialization of cap_rights_t. - Convert most commonly used code paths to use preinitialized cap_rights_t. A 3.6% speedup in fstat was measured with this change. Reported by: mjg Reviewed by: oshogbo Approved by: sbruno MFC after: 1 month
2018-05-09 18:47:24 +00:00
&cap_ioctl_rights, &fp);
Change the cap_rights_t type from uint64_t to a structure that we can extend in the future in a backward compatible (API and ABI) way. The cap_rights_t represents capability rights. We used to use one bit to represent one right, but we are running out of spare bits. Currently the new structure provides place for 114 rights (so 50 more than the previous cap_rights_t), but it is possible to grow the structure to hold at least 285 rights, although we can make it even larger if 285 rights won't be enough. The structure definition looks like this: struct cap_rights { uint64_t cr_rights[CAP_RIGHTS_VERSION + 2]; }; The initial CAP_RIGHTS_VERSION is 0. The top two bits in the first element of the cr_rights[] array contain total number of elements in the array - 2. This means if those two bits are equal to 0, we have 2 array elements. The top two bits in all remaining array elements should be 0. The next five bits in all array elements contain array index. Only one bit is used and bit position in this five-bits range defines array index. This means there can be at most five array elements in the future. To define new right the CAPRIGHT() macro must be used. The macro takes two arguments - an array index and a bit to set, eg. #define CAP_PDKILL CAPRIGHT(1, 0x0000000000000800ULL) We still support aliases that combine few rights, but the rights have to belong to the same array element, eg: #define CAP_LOOKUP CAPRIGHT(0, 0x0000000000000400ULL) #define CAP_FCHMOD CAPRIGHT(0, 0x0000000000002000ULL) #define CAP_FCHMODAT (CAP_FCHMOD | CAP_LOOKUP) There is new API to manage the new cap_rights_t structure: cap_rights_t *cap_rights_init(cap_rights_t *rights, ...); void cap_rights_set(cap_rights_t *rights, ...); void cap_rights_clear(cap_rights_t *rights, ...); bool cap_rights_is_set(const cap_rights_t *rights, ...); bool cap_rights_is_valid(const cap_rights_t *rights); void cap_rights_merge(cap_rights_t *dst, const cap_rights_t *src); void cap_rights_remove(cap_rights_t *dst, const cap_rights_t *src); bool cap_rights_contains(const cap_rights_t *big, const cap_rights_t *little); Capability rights to the cap_rights_init(), cap_rights_set(), cap_rights_clear() and cap_rights_is_set() functions are provided by separating them with commas, eg: cap_rights_t rights; cap_rights_init(&rights, CAP_READ, CAP_WRITE, CAP_FSTAT); There is no need to terminate the list of rights, as those functions are actually macros that take care of the termination, eg: #define cap_rights_set(rights, ...) \ __cap_rights_set((rights), __VA_ARGS__, 0ULL) void __cap_rights_set(cap_rights_t *rights, ...); Thanks to using one bit as an array index we can assert in those functions that there are no two rights belonging to different array elements provided together. For example this is illegal and will be detected, because CAP_LOOKUP belongs to element 0 and CAP_PDKILL to element 1: cap_rights_init(&rights, CAP_LOOKUP | CAP_PDKILL); Providing several rights that belongs to the same array's element this way is correct, but is not advised. It should only be used for aliases definition. This commit also breaks compatibility with some existing Capsicum system calls, but I see no other way to do that. This should be fine as Capsicum is still experimental and this change is not going to 9.x. Sponsored by: The FreeBSD Foundation
2013-09-05 00:09:56 +00:00
if (error != 0)
This is v4l2 support for the linuxulator. This allows to access FreeBSD native devices which support the v4l2 API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd or multimedia/webcamd supplied drivers. Submitted by: nox MFC after: 1 month
2011-05-04 09:05:39 +00:00
return (error);
error = fo_ioctl(fp, VIDIOC_ENUMINPUT, (caddr_t)&vinp,
td->td_ucred, td);
if (error) {
fdrop(fp, td);
return (error);
}
error = copyout(&vinp, (void *)args->arg,
sizeof(struct l_v4l2_input));
fdrop(fp, td);
return (error);
case LINUX_VIDIOC_QUERYBUF:
case LINUX_VIDIOC_QBUF:
case LINUX_VIDIOC_DQBUF:
error = copyin((void *)args->arg, &l_vbuf, sizeof(l_vbuf));
if (error)
return (error);
Change the cap_rights_t type from uint64_t to a structure that we can extend in the future in a backward compatible (API and ABI) way. The cap_rights_t represents capability rights. We used to use one bit to represent one right, but we are running out of spare bits. Currently the new structure provides place for 114 rights (so 50 more than the previous cap_rights_t), but it is possible to grow the structure to hold at least 285 rights, although we can make it even larger if 285 rights won't be enough. The structure definition looks like this: struct cap_rights { uint64_t cr_rights[CAP_RIGHTS_VERSION + 2]; }; The initial CAP_RIGHTS_VERSION is 0. The top two bits in the first element of the cr_rights[] array contain total number of elements in the array - 2. This means if those two bits are equal to 0, we have 2 array elements. The top two bits in all remaining array elements should be 0. The next five bits in all array elements contain array index. Only one bit is used and bit position in this five-bits range defines array index. This means there can be at most five array elements in the future. To define new right the CAPRIGHT() macro must be used. The macro takes two arguments - an array index and a bit to set, eg. #define CAP_PDKILL CAPRIGHT(1, 0x0000000000000800ULL) We still support aliases that combine few rights, but the rights have to belong to the same array element, eg: #define CAP_LOOKUP CAPRIGHT(0, 0x0000000000000400ULL) #define CAP_FCHMOD CAPRIGHT(0, 0x0000000000002000ULL) #define CAP_FCHMODAT (CAP_FCHMOD | CAP_LOOKUP) There is new API to manage the new cap_rights_t structure: cap_rights_t *cap_rights_init(cap_rights_t *rights, ...); void cap_rights_set(cap_rights_t *rights, ...); void cap_rights_clear(cap_rights_t *rights, ...); bool cap_rights_is_set(const cap_rights_t *rights, ...); bool cap_rights_is_valid(const cap_rights_t *rights); void cap_rights_merge(cap_rights_t *dst, const cap_rights_t *src); void cap_rights_remove(cap_rights_t *dst, const cap_rights_t *src); bool cap_rights_contains(const cap_rights_t *big, const cap_rights_t *little); Capability rights to the cap_rights_init(), cap_rights_set(), cap_rights_clear() and cap_rights_is_set() functions are provided by separating them with commas, eg: cap_rights_t rights; cap_rights_init(&rights, CAP_READ, CAP_WRITE, CAP_FSTAT); There is no need to terminate the list of rights, as those functions are actually macros that take care of the termination, eg: #define cap_rights_set(rights, ...) \ __cap_rights_set((rights), __VA_ARGS__, 0ULL) void __cap_rights_set(cap_rights_t *rights, ...); Thanks to using one bit as an array index we can assert in those functions that there are no two rights belonging to different array elements provided together. For example this is illegal and will be detected, because CAP_LOOKUP belongs to element 0 and CAP_PDKILL to element 1: cap_rights_init(&rights, CAP_LOOKUP | CAP_PDKILL); Providing several rights that belongs to the same array's element this way is correct, but is not advised. It should only be used for aliases definition. This commit also breaks compatibility with some existing Capsicum system calls, but I see no other way to do that. This should be fine as Capsicum is still experimental and this change is not going to 9.x. Sponsored by: The FreeBSD Foundation
2013-09-05 00:09:56 +00:00
error = fget(td, args->fd,
Eliminate the overhead of gratuitous repeated reinitialization of cap_rights - Add macros to allow preinitialization of cap_rights_t. - Convert most commonly used code paths to use preinitialized cap_rights_t. A 3.6% speedup in fstat was measured with this change. Reported by: mjg Reviewed by: oshogbo Approved by: sbruno MFC after: 1 month
2018-05-09 18:47:24 +00:00
&cap_ioctl_rights, &fp);
Change the cap_rights_t type from uint64_t to a structure that we can extend in the future in a backward compatible (API and ABI) way. The cap_rights_t represents capability rights. We used to use one bit to represent one right, but we are running out of spare bits. Currently the new structure provides place for 114 rights (so 50 more than the previous cap_rights_t), but it is possible to grow the structure to hold at least 285 rights, although we can make it even larger if 285 rights won't be enough. The structure definition looks like this: struct cap_rights { uint64_t cr_rights[CAP_RIGHTS_VERSION + 2]; }; The initial CAP_RIGHTS_VERSION is 0. The top two bits in the first element of the cr_rights[] array contain total number of elements in the array - 2. This means if those two bits are equal to 0, we have 2 array elements. The top two bits in all remaining array elements should be 0. The next five bits in all array elements contain array index. Only one bit is used and bit position in this five-bits range defines array index. This means there can be at most five array elements in the future. To define new right the CAPRIGHT() macro must be used. The macro takes two arguments - an array index and a bit to set, eg. #define CAP_PDKILL CAPRIGHT(1, 0x0000000000000800ULL) We still support aliases that combine few rights, but the rights have to belong to the same array element, eg: #define CAP_LOOKUP CAPRIGHT(0, 0x0000000000000400ULL) #define CAP_FCHMOD CAPRIGHT(0, 0x0000000000002000ULL) #define CAP_FCHMODAT (CAP_FCHMOD | CAP_LOOKUP) There is new API to manage the new cap_rights_t structure: cap_rights_t *cap_rights_init(cap_rights_t *rights, ...); void cap_rights_set(cap_rights_t *rights, ...); void cap_rights_clear(cap_rights_t *rights, ...); bool cap_rights_is_set(const cap_rights_t *rights, ...); bool cap_rights_is_valid(const cap_rights_t *rights); void cap_rights_merge(cap_rights_t *dst, const cap_rights_t *src); void cap_rights_remove(cap_rights_t *dst, const cap_rights_t *src); bool cap_rights_contains(const cap_rights_t *big, const cap_rights_t *little); Capability rights to the cap_rights_init(), cap_rights_set(), cap_rights_clear() and cap_rights_is_set() functions are provided by separating them with commas, eg: cap_rights_t rights; cap_rights_init(&rights, CAP_READ, CAP_WRITE, CAP_FSTAT); There is no need to terminate the list of rights, as those functions are actually macros that take care of the termination, eg: #define cap_rights_set(rights, ...) \ __cap_rights_set((rights), __VA_ARGS__, 0ULL) void __cap_rights_set(cap_rights_t *rights, ...); Thanks to using one bit as an array index we can assert in those functions that there are no two rights belonging to different array elements provided together. For example this is illegal and will be detected, because CAP_LOOKUP belongs to element 0 and CAP_PDKILL to element 1: cap_rights_init(&rights, CAP_LOOKUP | CAP_PDKILL); Providing several rights that belongs to the same array's element this way is correct, but is not advised. It should only be used for aliases definition. This commit also breaks compatibility with some existing Capsicum system calls, but I see no other way to do that. This should be fine as Capsicum is still experimental and this change is not going to 9.x. Sponsored by: The FreeBSD Foundation
2013-09-05 00:09:56 +00:00
if (error)
This is v4l2 support for the linuxulator. This allows to access FreeBSD native devices which support the v4l2 API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd or multimedia/webcamd supplied drivers. Submitted by: nox MFC after: 1 month
2011-05-04 09:05:39 +00:00
return (error);
linux_to_bsd_v4l2_buffer(&l_vbuf, &vbuf);
if ((args->cmd & 0xffff) == LINUX_VIDIOC_QUERYBUF)
error = fo_ioctl(fp, VIDIOC_QUERYBUF, &vbuf,
td->td_ucred, td);
else if ((args->cmd & 0xffff) == LINUX_VIDIOC_QBUF)
error = fo_ioctl(fp, VIDIOC_QBUF, &vbuf,
td->td_ucred, td);
else
error = fo_ioctl(fp, VIDIOC_DQBUF, &vbuf,
td->td_ucred, td);
bsd_to_linux_v4l2_buffer(&vbuf, &l_vbuf);
copyout(&l_vbuf, (void *)args->arg, sizeof(l_vbuf));
fdrop(fp, td);
return (error);
/*
* XXX TODO - these need 32 -> 64 bit conversion:
* (are any of them needed for webcams?)
*/
case LINUX_VIDIOC_G_FBUF:
case LINUX_VIDIOC_S_FBUF:
case LINUX_VIDIOC_G_EXT_CTRLS:
case LINUX_VIDIOC_S_EXT_CTRLS:
case LINUX_VIDIOC_TRY_EXT_CTRLS:
case LINUX_VIDIOC_DQEVENT:
default: return (ENOIOCTL);
}
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = sys_ioctl(td, (struct ioctl_args *)args);
This is v4l2 support for the linuxulator. This allows to access FreeBSD native devices which support the v4l2 API from processes running within the linuxulator, e.g. skype or flash can access the multimedia/pwcbsd or multimedia/webcamd supplied drivers. Submitted by: nox MFC after: 1 month
2011-05-04 09:05:39 +00:00
return (error);
}
Let us to use our libusb(3) in Linuxolator. With this change, Linux binaries can work with our libusb(3) when it's compiled against our header files on GNU/Linux system -- this solves the problem with differences between /dev layouts. With ported libusb(3), I am able to use my USB JTAG cable with Linux binaries that support it. Reviewed by: thompsa
2010-01-18 22:46:06 +00:00
/*
Bring USB fixes for linux(4). Intention of this commit is to let us take a full advantage of libusb(8) ported to Linux. This decreases a possibility of getting any collisions within ioctl() "command" space, especially with relation to LINUX_SNDCTL_SEQ... stuff. Basically, we provide commands, that will be mapped in the kernel to correct ones and forward those to the USB layer. Port enabling functionality brought with this patch is here: http://www.freebsd.org/cgi/query-pr.cgi?pr=146895 Bump __FreeBSD_version to catch, since which version installing a port makes sense. This patch should bring no regressions. So far, only i386 is tested. Tested by: thompsa@ Reviewed by: thompsa@ OKed by: netchild@
2010-05-24 07:04:00 +00:00
* Support for emulators/linux-libusb. This port uses FBSD_LUSB* macros
* instead of USB* ones. This lets us to provide correct values for cmd.
* 0xffffffe0 -- 0xffffffff range seemed to be the least collision-prone.
Let us to use our libusb(3) in Linuxolator. With this change, Linux binaries can work with our libusb(3) when it's compiled against our header files on GNU/Linux system -- this solves the problem with differences between /dev layouts. With ported libusb(3), I am able to use my USB JTAG cable with Linux binaries that support it. Reviewed by: thompsa
2010-01-18 22:46:06 +00:00
*/
static int
linux_ioctl_fbsd_usb(struct thread *td, struct linux_ioctl_args *args)
{
Bring USB fixes for linux(4). Intention of this commit is to let us take a full advantage of libusb(8) ported to Linux. This decreases a possibility of getting any collisions within ioctl() "command" space, especially with relation to LINUX_SNDCTL_SEQ... stuff. Basically, we provide commands, that will be mapped in the kernel to correct ones and forward those to the USB layer. Port enabling functionality brought with this patch is here: http://www.freebsd.org/cgi/query-pr.cgi?pr=146895 Bump __FreeBSD_version to catch, since which version installing a port makes sense. This patch should bring no regressions. So far, only i386 is tested. Tested by: thompsa@ Reviewed by: thompsa@ OKed by: netchild@
2010-05-24 07:04:00 +00:00
int error;
Let us to use our libusb(3) in Linuxolator. With this change, Linux binaries can work with our libusb(3) when it's compiled against our header files on GNU/Linux system -- this solves the problem with differences between /dev layouts. With ported libusb(3), I am able to use my USB JTAG cable with Linux binaries that support it. Reviewed by: thompsa
2010-01-18 22:46:06 +00:00
Bring USB fixes for linux(4). Intention of this commit is to let us take a full advantage of libusb(8) ported to Linux. This decreases a possibility of getting any collisions within ioctl() "command" space, especially with relation to LINUX_SNDCTL_SEQ... stuff. Basically, we provide commands, that will be mapped in the kernel to correct ones and forward those to the USB layer. Port enabling functionality brought with this patch is here: http://www.freebsd.org/cgi/query-pr.cgi?pr=146895 Bump __FreeBSD_version to catch, since which version installing a port makes sense. This patch should bring no regressions. So far, only i386 is tested. Tested by: thompsa@ Reviewed by: thompsa@ OKed by: netchild@
2010-05-24 07:04:00 +00:00
error = 0;
switch (args->cmd) {
case FBSD_LUSB_DEVICEENUMERATE:
args->cmd = USB_DEVICEENUMERATE;
break;
case FBSD_LUSB_DEV_QUIRK_ADD:
args->cmd = USB_DEV_QUIRK_ADD;
break;
case FBSD_LUSB_DEV_QUIRK_GET:
args->cmd = USB_DEV_QUIRK_GET;
break;
case FBSD_LUSB_DEV_QUIRK_REMOVE:
args->cmd = USB_DEV_QUIRK_REMOVE;
break;
case FBSD_LUSB_DO_REQUEST:
args->cmd = USB_DO_REQUEST;
break;
case FBSD_LUSB_FS_CLEAR_STALL_SYNC:
args->cmd = USB_FS_CLEAR_STALL_SYNC;
break;
case FBSD_LUSB_FS_CLOSE:
args->cmd = USB_FS_CLOSE;
break;
case FBSD_LUSB_FS_COMPLETE:
args->cmd = USB_FS_COMPLETE;
break;
case FBSD_LUSB_FS_INIT:
args->cmd = USB_FS_INIT;
break;
case FBSD_LUSB_FS_OPEN:
args->cmd = USB_FS_OPEN;
break;
case FBSD_LUSB_FS_START:
args->cmd = USB_FS_START;
break;
case FBSD_LUSB_FS_STOP:
args->cmd = USB_FS_STOP;
break;
case FBSD_LUSB_FS_UNINIT:
args->cmd = USB_FS_UNINIT;
break;
case FBSD_LUSB_GET_CONFIG:
args->cmd = USB_GET_CONFIG;
break;
case FBSD_LUSB_GET_DEVICEINFO:
args->cmd = USB_GET_DEVICEINFO;
break;
case FBSD_LUSB_GET_DEVICE_DESC:
args->cmd = USB_GET_DEVICE_DESC;
break;
case FBSD_LUSB_GET_FULL_DESC:
args->cmd = USB_GET_FULL_DESC;
break;
case FBSD_LUSB_GET_IFACE_DRIVER:
args->cmd = USB_GET_IFACE_DRIVER;
break;
case FBSD_LUSB_GET_PLUGTIME:
args->cmd = USB_GET_PLUGTIME;
break;
case FBSD_LUSB_GET_POWER_MODE:
args->cmd = USB_GET_POWER_MODE;
break;
case FBSD_LUSB_GET_REPORT_DESC:
args->cmd = USB_GET_REPORT_DESC;
break;
case FBSD_LUSB_GET_REPORT_ID:
args->cmd = USB_GET_REPORT_ID;
break;
case FBSD_LUSB_GET_TEMPLATE:
args->cmd = USB_GET_TEMPLATE;
break;
case FBSD_LUSB_IFACE_DRIVER_ACTIVE:
args->cmd = USB_IFACE_DRIVER_ACTIVE;
break;
case FBSD_LUSB_IFACE_DRIVER_DETACH:
args->cmd = USB_IFACE_DRIVER_DETACH;
break;
case FBSD_LUSB_QUIRK_NAME_GET:
args->cmd = USB_QUIRK_NAME_GET;
break;
case FBSD_LUSB_READ_DIR:
args->cmd = USB_READ_DIR;
break;
case FBSD_LUSB_SET_ALTINTERFACE:
args->cmd = USB_SET_ALTINTERFACE;
break;
case FBSD_LUSB_SET_CONFIG:
args->cmd = USB_SET_CONFIG;
break;
case FBSD_LUSB_SET_IMMED:
args->cmd = USB_SET_IMMED;
break;
case FBSD_LUSB_SET_POWER_MODE:
args->cmd = USB_SET_POWER_MODE;
break;
case FBSD_LUSB_SET_TEMPLATE:
args->cmd = USB_SET_TEMPLATE;
break;
Add some missing LIBUSB IOCTL conversion codes.
2013-07-14 10:13:01 +00:00
case FBSD_LUSB_FS_OPEN_STREAM:
args->cmd = USB_FS_OPEN_STREAM;
break;
case FBSD_LUSB_GET_DEV_PORT_PATH:
args->cmd = USB_GET_DEV_PORT_PATH;
break;
case FBSD_LUSB_GET_POWER_USAGE:
args->cmd = USB_GET_POWER_USAGE;
break;
Add support for the device statistics IOCTL, needed by the coming linux_libusb upgrade. MFC after: 3 days Sponsored by: Mellanox Technologies
2020-03-10 15:56:49 +00:00
case FBSD_LUSB_DEVICESTATS:
args->cmd = USB_DEVICESTATS;
break;
Bring USB fixes for linux(4). Intention of this commit is to let us take a full advantage of libusb(8) ported to Linux. This decreases a possibility of getting any collisions within ioctl() "command" space, especially with relation to LINUX_SNDCTL_SEQ... stuff. Basically, we provide commands, that will be mapped in the kernel to correct ones and forward those to the USB layer. Port enabling functionality brought with this patch is here: http://www.freebsd.org/cgi/query-pr.cgi?pr=146895 Bump __FreeBSD_version to catch, since which version installing a port makes sense. This patch should bring no regressions. So far, only i386 is tested. Tested by: thompsa@ Reviewed by: thompsa@ OKed by: netchild@
2010-05-24 07:04:00 +00:00
default:
error = ENOIOCTL;
}
if (error != ENOIOCTL)
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
error = sys_ioctl(td, (struct ioctl_args *)args);
Bring USB fixes for linux(4). Intention of this commit is to let us take a full advantage of libusb(8) ported to Linux. This decreases a possibility of getting any collisions within ioctl() "command" space, especially with relation to LINUX_SNDCTL_SEQ... stuff. Basically, we provide commands, that will be mapped in the kernel to correct ones and forward those to the USB layer. Port enabling functionality brought with this patch is here: http://www.freebsd.org/cgi/query-pr.cgi?pr=146895 Bump __FreeBSD_version to catch, since which version installing a port makes sense. This patch should bring no regressions. So far, only i386 is tested. Tested by: thompsa@ Reviewed by: thompsa@ OKed by: netchild@
2010-05-24 07:04:00 +00:00
return (error);
Let us to use our libusb(3) in Linuxolator. With this change, Linux binaries can work with our libusb(3) when it's compiled against our header files on GNU/Linux system -- this solves the problem with differences between /dev layouts. With ported libusb(3), I am able to use my USB JTAG cable with Linux binaries that support it. Reviewed by: thompsa
2010-01-18 22:46:06 +00:00
}
Add Evdev ioctl handler to the Linuxulator. PR: 218627 Submitted by: Jan Kokemüller Reported by: Jan Kokemüller MFC after: 1 week
2017-04-23 07:43:50 +00:00
/*
* Some evdev ioctls must be translated.
* - EVIOCGMTSLOTS is a IOC_READ ioctl on Linux although it has input data
* (must be IOC_INOUT on FreeBSD).
* - On Linux, EVIOCGRAB, EVIOCREVOKE and EVIOCRMFF are defined as _IOW with
* an int argument. You don't pass an int pointer to the ioctl(), however,
* but just the int directly. On FreeBSD, they are defined as _IOWINT for
* this to work.
*/
static int
linux_ioctl_evdev(struct thread *td, struct linux_ioctl_args *args)
{
struct file *fp;
clockid_t clock;
int error;
args->cmd = SETDIR(args->cmd);
switch (args->cmd) {
case (EVIOCGRAB & ~IOC_DIRMASK) | IOC_IN:
args->cmd = EVIOCGRAB;
break;
case (EVIOCREVOKE & ~IOC_DIRMASK) | IOC_IN:
args->cmd = EVIOCREVOKE;
break;
case (EVIOCRMFF & ~IOC_DIRMASK) | IOC_IN:
args->cmd = EVIOCRMFF;
break;
case EVIOCSCLOCKID: {
error = copyin(PTRIN(args->arg), &clock, sizeof(clock));
if (error != 0)
return (error);
if (clock & ~(LINUX_IOCTL_EVDEV_CLK))
return (EINVAL);
error = linux_to_native_clockid(&clock, clock);
if (error != 0)
return (error);
error = fget(td, args->fd,
Eliminate the overhead of gratuitous repeated reinitialization of cap_rights - Add macros to allow preinitialization of cap_rights_t. - Convert most commonly used code paths to use preinitialized cap_rights_t. A 3.6% speedup in fstat was measured with this change. Reported by: mjg Reviewed by: oshogbo Approved by: sbruno MFC after: 1 month
2018-05-09 18:47:24 +00:00
&cap_ioctl_rights, &fp);
Add Evdev ioctl handler to the Linuxulator. PR: 218627 Submitted by: Jan Kokemüller Reported by: Jan Kokemüller MFC after: 1 week
2017-04-23 07:43:50 +00:00
if (error != 0)
return (error);
error = fo_ioctl(fp, EVIOCSCLOCKID, &clock, td->td_ucred, td);
fdrop(fp, td);
return (error);
}
default:
break;
}
if (IOCBASECMD(args->cmd) ==
((EVIOCGMTSLOTS(0) & ~IOC_DIRMASK) | IOC_OUT))
args->cmd = (args->cmd & ~IOC_DIRMASK) | IOC_INOUT;
return (sys_ioctl(td, (struct ioctl_args *)args));
}
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
/*
* main ioctl syscall function
*/
- Handle mixer read ioctls correctly. They have the same group, number and argument size as their write counterparts and were handled as write ioctls. - Emulate some cdrom ioctls.
1999-04-29 04:37:57 +00:00
linux: fix ioctl performance for termios TCGETS et al are frequently issued by Linux binaries while the previous code avoidably ping-pongs a global sx lock and serializes on Giant. Note that even with the fix the common case will serialize on a per-tty lock.
2020-07-04 06:25:41 +00:00
static int
linux_ioctl_fallback(struct thread *td, struct linux_ioctl_args *args)
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
{
struct file *fp;
On amd64 both Linux compat modules, linux.ko and linux64.ko, provide linux_ioctl_(un)register_handler that allows other driver modules to register ioctl handlers. The ioctl syscall implementation in each Linux compat module iterates over the list of handlers and forwards the call to the appropriate driver. Because the registration functions have the same name in each module it is not possible for a driver to support both 32 and 64 bit linux compatibility. Move the list of ioctl handlers to linux_common.ko so it is shared by both Linux modules and all drivers receive both 32 and 64 bit ioctl calls with one registration. These ioctl handlers normally forward the call to the FreeBSD ioctl handler which can handle both 32 and 64 bit. Keep the special COMPAT_LINUX32 ioctl handlers in linux.ko in a separate list for now and let the ioctl syscall iterate over that list first. Later, COMPAT_LINUX32 support can be added to the 64 bit ioctl handlers via a runtime check for ILP32 like is done for COMPAT_FREEBSD32 and then this separate list would disappear again. That is a much bigger effort however and this commit is meant to be MFCable. This enables linux64 support in x11/nvidia-driver*. PR: 206711 Reviewed by: kib MFC after: 3 days
2018-11-06 13:51:08 +00:00
struct linux_ioctl_handler_element *he;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
int error, cmd;
Eliminate the overhead of gratuitous repeated reinitialization of cap_rights - Add macros to allow preinitialization of cap_rights_t. - Convert most commonly used code paths to use preinitialized cap_rights_t. A 3.6% speedup in fstat was measured with this change. Reported by: mjg Reviewed by: oshogbo Approved by: sbruno MFC after: 1 month
2018-05-09 18:47:24 +00:00
error = fget(td, args->fd, &cap_ioctl_rights, &fp);
Change the cap_rights_t type from uint64_t to a structure that we can extend in the future in a backward compatible (API and ABI) way. The cap_rights_t represents capability rights. We used to use one bit to represent one right, but we are running out of spare bits. Currently the new structure provides place for 114 rights (so 50 more than the previous cap_rights_t), but it is possible to grow the structure to hold at least 285 rights, although we can make it even larger if 285 rights won't be enough. The structure definition looks like this: struct cap_rights { uint64_t cr_rights[CAP_RIGHTS_VERSION + 2]; }; The initial CAP_RIGHTS_VERSION is 0. The top two bits in the first element of the cr_rights[] array contain total number of elements in the array - 2. This means if those two bits are equal to 0, we have 2 array elements. The top two bits in all remaining array elements should be 0. The next five bits in all array elements contain array index. Only one bit is used and bit position in this five-bits range defines array index. This means there can be at most five array elements in the future. To define new right the CAPRIGHT() macro must be used. The macro takes two arguments - an array index and a bit to set, eg. #define CAP_PDKILL CAPRIGHT(1, 0x0000000000000800ULL) We still support aliases that combine few rights, but the rights have to belong to the same array element, eg: #define CAP_LOOKUP CAPRIGHT(0, 0x0000000000000400ULL) #define CAP_FCHMOD CAPRIGHT(0, 0x0000000000002000ULL) #define CAP_FCHMODAT (CAP_FCHMOD | CAP_LOOKUP) There is new API to manage the new cap_rights_t structure: cap_rights_t *cap_rights_init(cap_rights_t *rights, ...); void cap_rights_set(cap_rights_t *rights, ...); void cap_rights_clear(cap_rights_t *rights, ...); bool cap_rights_is_set(const cap_rights_t *rights, ...); bool cap_rights_is_valid(const cap_rights_t *rights); void cap_rights_merge(cap_rights_t *dst, const cap_rights_t *src); void cap_rights_remove(cap_rights_t *dst, const cap_rights_t *src); bool cap_rights_contains(const cap_rights_t *big, const cap_rights_t *little); Capability rights to the cap_rights_init(), cap_rights_set(), cap_rights_clear() and cap_rights_is_set() functions are provided by separating them with commas, eg: cap_rights_t rights; cap_rights_init(&rights, CAP_READ, CAP_WRITE, CAP_FSTAT); There is no need to terminate the list of rights, as those functions are actually macros that take care of the termination, eg: #define cap_rights_set(rights, ...) \ __cap_rights_set((rights), __VA_ARGS__, 0ULL) void __cap_rights_set(cap_rights_t *rights, ...); Thanks to using one bit as an array index we can assert in those functions that there are no two rights belonging to different array elements provided together. For example this is illegal and will be detected, because CAP_LOOKUP belongs to element 0 and CAP_PDKILL to element 1: cap_rights_init(&rights, CAP_LOOKUP | CAP_PDKILL); Providing several rights that belongs to the same array's element this way is correct, but is not advised. It should only be used for aliases definition. This commit also breaks compatibility with some existing Capsicum system calls, but I see no other way to do that. This should be fine as Capsicum is still experimental and this change is not going to 9.x. Sponsored by: The FreeBSD Foundation
2013-09-05 00:09:56 +00:00
if (error != 0)
Replace ffind_* with fget calls. Make fget MPsafe. Make fgetvp and fgetsock use the fget subsystem to reduce code bloat. Push giant down in fpathconf().
2002-01-14 00:13:45 +00:00
return (error);
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
if ((fp->f_flag & (FREAD|FWRITE)) == 0) {
fdrop(fp, td);
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
return (EBADF);
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
}
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
/* Iterate over the ioctl handlers */
cmd = args->cmd & 0xffff;
- Protect the list of linux ioctl handlers with an sx lock. - Hold Giant while calling linux ioctl handlers for now as they aren't all known to be MPSAFE yet. - Mark linux_ioctl() MPSAFE.
2006-07-06 21:42:36 +00:00
sx_slock(&linux_ioctl_sx);
mtx_lock(&Giant);
On amd64 both Linux compat modules, linux.ko and linux64.ko, provide linux_ioctl_(un)register_handler that allows other driver modules to register ioctl handlers. The ioctl syscall implementation in each Linux compat module iterates over the list of handlers and forwards the call to the appropriate driver. Because the registration functions have the same name in each module it is not possible for a driver to support both 32 and 64 bit linux compatibility. Move the list of ioctl handlers to linux_common.ko so it is shared by both Linux modules and all drivers receive both 32 and 64 bit ioctl calls with one registration. These ioctl handlers normally forward the call to the FreeBSD ioctl handler which can handle both 32 and 64 bit. Keep the special COMPAT_LINUX32 ioctl handlers in linux.ko in a separate list for now and let the ioctl syscall iterate over that list first. Later, COMPAT_LINUX32 support can be added to the 64 bit ioctl handlers via a runtime check for ILP32 like is done for COMPAT_FREEBSD32 and then this separate list would disappear again. That is a much bigger effort however and this commit is meant to be MFCable. This enables linux64 support in x11/nvidia-driver*. PR: 206711 Reviewed by: kib MFC after: 3 days
2018-11-06 13:51:08 +00:00
#ifdef COMPAT_LINUX32
TAILQ_FOREACH(he, &linux32_ioctl_handlers, list) {
if (cmd >= he->low && cmd <= he->high) {
error = (*he->func)(td, args);
if (error != ENOIOCTL) {
mtx_unlock(&Giant);
sx_sunlock(&linux_ioctl_sx);
fdrop(fp, td);
return (error);
}
}
}
#endif
TAILQ_FOREACH(he, &linux_ioctl_handlers, list) {
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
if (cmd >= he->low && cmd <= he->high) {
KSE Milestone 2 Note ALL MODULES MUST BE RECOMPILED make the kernel aware that there are smaller units of scheduling than the process. (but only allow one thread per process at this time). This is functionally equivalent to teh previousl -current except that there is a thread associated with each process. Sorry john! (your next MFC will be a doosie!) Reviewed by: peter@freebsd.org, dillon@freebsd.org X-MFC after: ha ha ha ha
2001-09-12 08:38:13 +00:00
error = (*he->func)(td, args);
fix file descriptor leak. Submitted by: Mark Santcroos <marks@ripe.net>
2002-02-20 17:06:37 +00:00
if (error != ENOIOCTL) {
- Protect the list of linux ioctl handlers with an sx lock. - Hold Giant while calling linux ioctl handlers for now as they aren't all known to be MPSAFE yet. - Mark linux_ioctl() MPSAFE.
2006-07-06 21:42:36 +00:00
mtx_unlock(&Giant);
sx_sunlock(&linux_ioctl_sx);
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
fdrop(fp, td);
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
return (error);
fix file descriptor leak. Submitted by: Mark Santcroos <marks@ripe.net>
2002-02-20 17:06:37 +00:00
}
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
}
}
- Protect the list of linux ioctl handlers with an sx lock. - Hold Giant while calling linux ioctl handlers for now as they aren't all known to be MPSAFE yet. - Mark linux_ioctl() MPSAFE.
2006-07-06 21:42:36 +00:00
mtx_unlock(&Giant);
sx_sunlock(&linux_ioctl_sx);
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
fdrop(fp, td);
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
The latest cp tool is trying to use the btrfs clone operation that is implemented via ioctl interface. First of all return ENOTSUP for this operation as a cp fallback to usual method in that case. Secondly, do not print out the message about unimplemented operation.
2015-05-24 18:12:04 +00:00
switch (args->cmd & 0xffff) {
case LINUX_BTRFS_IOC_CLONE:
Return ENOTSUP for Linux FS_IOC_FIEMAP ioctl. Linux man(1) calls it for no good reason; this avoids the console spam (eg '(man): ioctl fd=4, cmd=0x660b ('f',11) is not implemented'). Reviewed by: emaste MFC after: 2 weeks Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D20690
2019-07-04 20:16:04 +00:00
case LINUX_FS_IOC_FIEMAP:
The latest cp tool is trying to use the btrfs clone operation that is implemented via ioctl interface. First of all return ENOTSUP for this operation as a cp fallback to usual method in that case. Secondly, do not print out the message about unimplemented operation.
2015-05-24 18:12:04 +00:00
return (ENOTSUP);
default:
linux_msg(td, "ioctl fd=%d, cmd=0x%x ('%c',%d) is not implemented",
args->fd, (int)(args->cmd & 0xffff),
(int)(args->cmd & 0xff00) >> 8, (int)(args->cmd & 0xff));
break;
}
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
return (EINVAL);
}
linux: fix ioctl performance for termios TCGETS et al are frequently issued by Linux binaries while the previous code avoidably ping-pongs a global sx lock and serializes on Giant. Note that even with the fix the common case will serialize on a per-tty lock.
2020-07-04 06:25:41 +00:00
int
linux_ioctl(struct thread *td, struct linux_ioctl_args *args)
{
struct linux_ioctl_handler *handler;
int error, cmd, i;
cmd = args->cmd & 0xffff;
/*
* array of ioctls known at compilation time. Elides a lot of work on
* each call compared to the list variant. Everything frequently used
* should be moved here.
*
* Arguably the magic creating the list should create an array instead.
*
* For now just a linear scan.
*/
for (i = 0; i < nitems(linux_ioctls); i++) {
handler = &linux_ioctls[i];
if (cmd >= handler->low && cmd <= handler->high) {
error = (*handler->func)(td, args);
if (error != ENOIOCTL) {
return (error);
}
}
}
return (linux_ioctl_fallback(td, args));
}
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
int
linux_ioctl_register_handler(struct linux_ioctl_handler *h)
{
On amd64 both Linux compat modules, linux.ko and linux64.ko, provide linux_ioctl_(un)register_handler that allows other driver modules to register ioctl handlers. The ioctl syscall implementation in each Linux compat module iterates over the list of handlers and forwards the call to the appropriate driver. Because the registration functions have the same name in each module it is not possible for a driver to support both 32 and 64 bit linux compatibility. Move the list of ioctl handlers to linux_common.ko so it is shared by both Linux modules and all drivers receive both 32 and 64 bit ioctl calls with one registration. These ioctl handlers normally forward the call to the FreeBSD ioctl handler which can handle both 32 and 64 bit. Keep the special COMPAT_LINUX32 ioctl handlers in linux.ko in a separate list for now and let the ioctl syscall iterate over that list first. Later, COMPAT_LINUX32 support can be added to the 64 bit ioctl handlers via a runtime check for ILP32 like is done for COMPAT_FREEBSD32 and then this separate list would disappear again. That is a much bigger effort however and this commit is meant to be MFCable. This enables linux64 support in x11/nvidia-driver*. PR: 206711 Reviewed by: kib MFC after: 3 days
2018-11-06 13:51:08 +00:00
struct linux_ioctl_handler_element *he, *cur;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
if (h == NULL || h->func == NULL)
return (EINVAL);
/*
* Reuse the element if the handler is already on the list, otherwise
* create a new element.
*/
- Protect the list of linux ioctl handlers with an sx lock. - Hold Giant while calling linux ioctl handlers for now as they aren't all known to be MPSAFE yet. - Mark linux_ioctl() MPSAFE.
2006-07-06 21:42:36 +00:00
sx_xlock(&linux_ioctl_sx);
On amd64 both Linux compat modules, linux.ko and linux64.ko, provide linux_ioctl_(un)register_handler that allows other driver modules to register ioctl handlers. The ioctl syscall implementation in each Linux compat module iterates over the list of handlers and forwards the call to the appropriate driver. Because the registration functions have the same name in each module it is not possible for a driver to support both 32 and 64 bit linux compatibility. Move the list of ioctl handlers to linux_common.ko so it is shared by both Linux modules and all drivers receive both 32 and 64 bit ioctl calls with one registration. These ioctl handlers normally forward the call to the FreeBSD ioctl handler which can handle both 32 and 64 bit. Keep the special COMPAT_LINUX32 ioctl handlers in linux.ko in a separate list for now and let the ioctl syscall iterate over that list first. Later, COMPAT_LINUX32 support can be added to the 64 bit ioctl handlers via a runtime check for ILP32 like is done for COMPAT_FREEBSD32 and then this separate list would disappear again. That is a much bigger effort however and this commit is meant to be MFCable. This enables linux64 support in x11/nvidia-driver*. PR: 206711 Reviewed by: kib MFC after: 3 days
2018-11-06 13:51:08 +00:00
TAILQ_FOREACH(he, &linux_ioctl_handlers, list) {
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
if (he->func == h->func)
break;
}
if (he == NULL) {
Retire the MALLOC and FREE macros. They are an abomination unto style(9). MFC after: 3 months
2008-10-23 15:53:51 +00:00
he = malloc(sizeof(*he),
Back out M_* changes, per decision of the TRB. Approved by: trb
2003-02-19 05:47:46 +00:00
M_LINUX, M_WAITOK);
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
he->func = h->func;
} else
On amd64 both Linux compat modules, linux.ko and linux64.ko, provide linux_ioctl_(un)register_handler that allows other driver modules to register ioctl handlers. The ioctl syscall implementation in each Linux compat module iterates over the list of handlers and forwards the call to the appropriate driver. Because the registration functions have the same name in each module it is not possible for a driver to support both 32 and 64 bit linux compatibility. Move the list of ioctl handlers to linux_common.ko so it is shared by both Linux modules and all drivers receive both 32 and 64 bit ioctl calls with one registration. These ioctl handlers normally forward the call to the FreeBSD ioctl handler which can handle both 32 and 64 bit. Keep the special COMPAT_LINUX32 ioctl handlers in linux.ko in a separate list for now and let the ioctl syscall iterate over that list first. Later, COMPAT_LINUX32 support can be added to the 64 bit ioctl handlers via a runtime check for ILP32 like is done for COMPAT_FREEBSD32 and then this separate list would disappear again. That is a much bigger effort however and this commit is meant to be MFCable. This enables linux64 support in x11/nvidia-driver*. PR: 206711 Reviewed by: kib MFC after: 3 days
2018-11-06 13:51:08 +00:00
TAILQ_REMOVE(&linux_ioctl_handlers, he, list);
Clean up whitespace, s/register //, refrain from strong urge to ANSIfy.
2003-03-02 15:56:49 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
/* Initialize range information. */
he->low = h->low;
he->high = h->high;
he->span = h->high - h->low + 1;
/* Add the element to the list, sorted on span. */
On amd64 both Linux compat modules, linux.ko and linux64.ko, provide linux_ioctl_(un)register_handler that allows other driver modules to register ioctl handlers. The ioctl syscall implementation in each Linux compat module iterates over the list of handlers and forwards the call to the appropriate driver. Because the registration functions have the same name in each module it is not possible for a driver to support both 32 and 64 bit linux compatibility. Move the list of ioctl handlers to linux_common.ko so it is shared by both Linux modules and all drivers receive both 32 and 64 bit ioctl calls with one registration. These ioctl handlers normally forward the call to the FreeBSD ioctl handler which can handle both 32 and 64 bit. Keep the special COMPAT_LINUX32 ioctl handlers in linux.ko in a separate list for now and let the ioctl syscall iterate over that list first. Later, COMPAT_LINUX32 support can be added to the 64 bit ioctl handlers via a runtime check for ILP32 like is done for COMPAT_FREEBSD32 and then this separate list would disappear again. That is a much bigger effort however and this commit is meant to be MFCable. This enables linux64 support in x11/nvidia-driver*. PR: 206711 Reviewed by: kib MFC after: 3 days
2018-11-06 13:51:08 +00:00
TAILQ_FOREACH(cur, &linux_ioctl_handlers, list) {
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
if (cur->span > he->span) {
TAILQ_INSERT_BEFORE(cur, he, list);
- Protect the list of linux ioctl handlers with an sx lock. - Hold Giant while calling linux ioctl handlers for now as they aren't all known to be MPSAFE yet. - Mark linux_ioctl() MPSAFE.
2006-07-06 21:42:36 +00:00
sx_xunlock(&linux_ioctl_sx);
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
return (0);
}
}
On amd64 both Linux compat modules, linux.ko and linux64.ko, provide linux_ioctl_(un)register_handler that allows other driver modules to register ioctl handlers. The ioctl syscall implementation in each Linux compat module iterates over the list of handlers and forwards the call to the appropriate driver. Because the registration functions have the same name in each module it is not possible for a driver to support both 32 and 64 bit linux compatibility. Move the list of ioctl handlers to linux_common.ko so it is shared by both Linux modules and all drivers receive both 32 and 64 bit ioctl calls with one registration. These ioctl handlers normally forward the call to the FreeBSD ioctl handler which can handle both 32 and 64 bit. Keep the special COMPAT_LINUX32 ioctl handlers in linux.ko in a separate list for now and let the ioctl syscall iterate over that list first. Later, COMPAT_LINUX32 support can be added to the 64 bit ioctl handlers via a runtime check for ILP32 like is done for COMPAT_FREEBSD32 and then this separate list would disappear again. That is a much bigger effort however and this commit is meant to be MFCable. This enables linux64 support in x11/nvidia-driver*. PR: 206711 Reviewed by: kib MFC after: 3 days
2018-11-06 13:51:08 +00:00
TAILQ_INSERT_TAIL(&linux_ioctl_handlers, he, list);
- Protect the list of linux ioctl handlers with an sx lock. - Hold Giant while calling linux ioctl handlers for now as they aren't all known to be MPSAFE yet. - Mark linux_ioctl() MPSAFE.
2006-07-06 21:42:36 +00:00
sx_xunlock(&linux_ioctl_sx);
- Handle mixer read ioctls correctly. They have the same group, number and argument size as their write counterparts and were handled as write ioctls. - Emulate some cdrom ioctls.
1999-04-29 04:37:57 +00:00
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
return (0);
}
int
linux_ioctl_unregister_handler(struct linux_ioctl_handler *h)
{
On amd64 both Linux compat modules, linux.ko and linux64.ko, provide linux_ioctl_(un)register_handler that allows other driver modules to register ioctl handlers. The ioctl syscall implementation in each Linux compat module iterates over the list of handlers and forwards the call to the appropriate driver. Because the registration functions have the same name in each module it is not possible for a driver to support both 32 and 64 bit linux compatibility. Move the list of ioctl handlers to linux_common.ko so it is shared by both Linux modules and all drivers receive both 32 and 64 bit ioctl calls with one registration. These ioctl handlers normally forward the call to the FreeBSD ioctl handler which can handle both 32 and 64 bit. Keep the special COMPAT_LINUX32 ioctl handlers in linux.ko in a separate list for now and let the ioctl syscall iterate over that list first. Later, COMPAT_LINUX32 support can be added to the 64 bit ioctl handlers via a runtime check for ILP32 like is done for COMPAT_FREEBSD32 and then this separate list would disappear again. That is a much bigger effort however and this commit is meant to be MFCable. This enables linux64 support in x11/nvidia-driver*. PR: 206711 Reviewed by: kib MFC after: 3 days
2018-11-06 13:51:08 +00:00
struct linux_ioctl_handler_element *he;
if (h == NULL || h->func == NULL)
return (EINVAL);
sx_xlock(&linux_ioctl_sx);
TAILQ_FOREACH(he, &linux_ioctl_handlers, list) {
if (he->func == h->func) {
TAILQ_REMOVE(&linux_ioctl_handlers, he, list);
sx_xunlock(&linux_ioctl_sx);
free(he, M_LINUX);
return (0);
}
}
sx_xunlock(&linux_ioctl_sx);
return (EINVAL);
}
#ifdef COMPAT_LINUX32
int
linux32_ioctl_register_handler(struct linux_ioctl_handler *h)
{
struct linux_ioctl_handler_element *he, *cur;
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
if (h == NULL || h->func == NULL)
return (EINVAL);
On amd64 both Linux compat modules, linux.ko and linux64.ko, provide linux_ioctl_(un)register_handler that allows other driver modules to register ioctl handlers. The ioctl syscall implementation in each Linux compat module iterates over the list of handlers and forwards the call to the appropriate driver. Because the registration functions have the same name in each module it is not possible for a driver to support both 32 and 64 bit linux compatibility. Move the list of ioctl handlers to linux_common.ko so it is shared by both Linux modules and all drivers receive both 32 and 64 bit ioctl calls with one registration. These ioctl handlers normally forward the call to the FreeBSD ioctl handler which can handle both 32 and 64 bit. Keep the special COMPAT_LINUX32 ioctl handlers in linux.ko in a separate list for now and let the ioctl syscall iterate over that list first. Later, COMPAT_LINUX32 support can be added to the 64 bit ioctl handlers via a runtime check for ILP32 like is done for COMPAT_FREEBSD32 and then this separate list would disappear again. That is a much bigger effort however and this commit is meant to be MFCable. This enables linux64 support in x11/nvidia-driver*. PR: 206711 Reviewed by: kib MFC after: 3 days
2018-11-06 13:51:08 +00:00
/*
* Reuse the element if the handler is already on the list, otherwise
* create a new element.
*/
- Protect the list of linux ioctl handlers with an sx lock. - Hold Giant while calling linux ioctl handlers for now as they aren't all known to be MPSAFE yet. - Mark linux_ioctl() MPSAFE.
2006-07-06 21:42:36 +00:00
sx_xlock(&linux_ioctl_sx);
On amd64 both Linux compat modules, linux.ko and linux64.ko, provide linux_ioctl_(un)register_handler that allows other driver modules to register ioctl handlers. The ioctl syscall implementation in each Linux compat module iterates over the list of handlers and forwards the call to the appropriate driver. Because the registration functions have the same name in each module it is not possible for a driver to support both 32 and 64 bit linux compatibility. Move the list of ioctl handlers to linux_common.ko so it is shared by both Linux modules and all drivers receive both 32 and 64 bit ioctl calls with one registration. These ioctl handlers normally forward the call to the FreeBSD ioctl handler which can handle both 32 and 64 bit. Keep the special COMPAT_LINUX32 ioctl handlers in linux.ko in a separate list for now and let the ioctl syscall iterate over that list first. Later, COMPAT_LINUX32 support can be added to the 64 bit ioctl handlers via a runtime check for ILP32 like is done for COMPAT_FREEBSD32 and then this separate list would disappear again. That is a much bigger effort however and this commit is meant to be MFCable. This enables linux64 support in x11/nvidia-driver*. PR: 206711 Reviewed by: kib MFC after: 3 days
2018-11-06 13:51:08 +00:00
TAILQ_FOREACH(he, &linux32_ioctl_handlers, list) {
if (he->func == h->func)
break;
}
if (he == NULL) {
he = malloc(sizeof(*he), M_LINUX, M_WAITOK);
he->func = h->func;
} else
TAILQ_REMOVE(&linux32_ioctl_handlers, he, list);
/* Initialize range information. */
he->low = h->low;
he->high = h->high;
he->span = h->high - h->low + 1;
/* Add the element to the list, sorted on span. */
TAILQ_FOREACH(cur, &linux32_ioctl_handlers, list) {
if (cur->span > he->span) {
TAILQ_INSERT_BEFORE(cur, he, list);
sx_xunlock(&linux_ioctl_sx);
return (0);
}
}
TAILQ_INSERT_TAIL(&linux32_ioctl_handlers, he, list);
sx_xunlock(&linux_ioctl_sx);
return (0);
}
int
linux32_ioctl_unregister_handler(struct linux_ioctl_handler *h)
{
struct linux_ioctl_handler_element *he;
if (h == NULL || h->func == NULL)
return (EINVAL);
sx_xlock(&linux_ioctl_sx);
TAILQ_FOREACH(he, &linux32_ioctl_handlers, list) {
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
if (he->func == h->func) {
On amd64 both Linux compat modules, linux.ko and linux64.ko, provide linux_ioctl_(un)register_handler that allows other driver modules to register ioctl handlers. The ioctl syscall implementation in each Linux compat module iterates over the list of handlers and forwards the call to the appropriate driver. Because the registration functions have the same name in each module it is not possible for a driver to support both 32 and 64 bit linux compatibility. Move the list of ioctl handlers to linux_common.ko so it is shared by both Linux modules and all drivers receive both 32 and 64 bit ioctl calls with one registration. These ioctl handlers normally forward the call to the FreeBSD ioctl handler which can handle both 32 and 64 bit. Keep the special COMPAT_LINUX32 ioctl handlers in linux.ko in a separate list for now and let the ioctl syscall iterate over that list first. Later, COMPAT_LINUX32 support can be added to the 64 bit ioctl handlers via a runtime check for ILP32 like is done for COMPAT_FREEBSD32 and then this separate list would disappear again. That is a much bigger effort however and this commit is meant to be MFCable. This enables linux64 support in x11/nvidia-driver*. PR: 206711 Reviewed by: kib MFC after: 3 days
2018-11-06 13:51:08 +00:00
TAILQ_REMOVE(&linux32_ioctl_handlers, he, list);
- Protect the list of linux ioctl handlers with an sx lock. - Hold Giant while calling linux ioctl handlers for now as they aren't all known to be MPSAFE yet. - Mark linux_ioctl() MPSAFE.
2006-07-06 21:42:36 +00:00
sx_xunlock(&linux_ioctl_sx);
Retire the MALLOC and FREE macros. They are an abomination unto style(9). MFC after: 3 months
2008-10-23 15:53:51 +00:00
free(he, M_LINUX);
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
return (0);
}
- Handle mixer read ioctls correctly. They have the same group, number and argument size as their write counterparts and were handled as write ioctls. - Emulate some cdrom ioctls.
1999-04-29 04:37:57 +00:00
}
- Protect the list of linux ioctl handlers with an sx lock. - Hold Giant while calling linux ioctl handlers for now as they aren't all known to be MPSAFE yet. - Mark linux_ioctl() MPSAFE.
2006-07-06 21:42:36 +00:00
sx_xunlock(&linux_ioctl_sx);
Implement pluggable ioctl handlers. Other modules can register and unregister ioctl handlers to extend the ioctls known by the Linuxulator. A recent application is the vmware port. The Linuxulator itself uses the new interface to register its handlers as well. Handlers for the following types of ioctls have been defined: cdrom console (=keyboard and VT handling) socket sound termio All ioctl related defines and declarations have been moved to a new file (linux_ioctl.h), except for the pluggable ioctl handler interface definition. While there, cleanup linux.h some more. linux.h and linux_ioctl.[ch] have been made to conform to style(9) as much as possible. Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
return (EINVAL);
}
On amd64 both Linux compat modules, linux.ko and linux64.ko, provide linux_ioctl_(un)register_handler that allows other driver modules to register ioctl handlers. The ioctl syscall implementation in each Linux compat module iterates over the list of handlers and forwards the call to the appropriate driver. Because the registration functions have the same name in each module it is not possible for a driver to support both 32 and 64 bit linux compatibility. Move the list of ioctl handlers to linux_common.ko so it is shared by both Linux modules and all drivers receive both 32 and 64 bit ioctl calls with one registration. These ioctl handlers normally forward the call to the FreeBSD ioctl handler which can handle both 32 and 64 bit. Keep the special COMPAT_LINUX32 ioctl handlers in linux.ko in a separate list for now and let the ioctl syscall iterate over that list first. Later, COMPAT_LINUX32 support can be added to the 64 bit ioctl handlers via a runtime check for ILP32 like is done for COMPAT_FREEBSD32 and then this separate list would disappear again. That is a much bigger effort however and this commit is meant to be MFCable. This enables linux64 support in x11/nvidia-driver*. PR: 206711 Reviewed by: kib MFC after: 3 days
2018-11-06 13:51:08 +00:00
#endif
Reference in New Issue Copy Permalink