Commit Graph

1774 Commits

Author SHA1 Message Date
Konstantin Belousov
1757d9699d Prefer struct sysentvec sv_psstrings to hardcoding FREEBSD32_PS_STRINGS
in the compat32 code. Use sv_usrstack instead of FREEBSD32_USRSTACK as well.

MFC after:	1 week
2010-08-07 11:57:13 +00:00
Konstantin Belousov
4605ef76e7 Add compat32 definition for (old) struct ostat.
MFC after:	1 week
2010-08-07 11:53:38 +00:00
Jung-uk Kim
a6d613a57f Do not block any I/O port on amd64. 2010-08-07 04:05:58 +00:00
Jung-uk Kim
d7a5fb634f Optimize interrupt vector lookup. There is no need to check the page table. 2010-08-07 03:45:45 +00:00
Jung-uk Kim
fc82156f95 Consistently use architecture specific macros. 2010-08-06 15:24:37 +00:00
Jung-uk Kim
f10776734f Fix allocation of multiple pages, which forgot to increase page number.
Particularly, it caused "vm86_addpage: overlap" panics under VirtualBox.
Add a safety check before freeing memory while I am here.
2010-08-06 15:04:01 +00:00
Jung-uk Kim
0a3493e5e7 Re-add flag register for output. Some BIOS calls actually use it to return
success/failure status.  Oops.
2010-08-05 19:30:57 +00:00
Jung-uk Kim
c5e960de35 Do not copy stack pointer and flags. These registers are unconditionally
destroyed from vm86_prepcall().
2010-08-05 19:12:35 +00:00
Jung-uk Kim
439f3d8b81 Implement a simple native VM86 backend for X86BIOS. Now i386 uses native
VM86 calls instead of the real mode emulator as a backend.  VM86 has been
proven reliable for very long time and it is actually few times faster than
emulation.  Increase maximum number of page table entries per VM86 context
from 3 to 8 pages.  It was (ridiculously) low and insufficient for new VM86
backend, which shares one context globally.  Slighly rearrange and clean up
the emulator backend to accommodate new code.  The only visible change here
is stack size, which is decreased from 64K to 4K bytes to sync. with VM86.
Actually, it seems there is no need for big stack in real mode.

MFC after:	1 month
2010-08-05 18:48:30 +00:00
Konstantin Belousov
64dc04de68 Copy inode birthtime to the struct stat32.
MFC after:	1 week
2010-08-04 14:38:20 +00:00
Konstantin Belousov
45b6fa3b54 Fix style.
MFC after:	1 week
2010-08-04 14:35:05 +00:00
Konstantin Belousov
34ab36a3dc When compat32 recvmsg(2) does not need to copy out control messages, set
msg_controllen to 0.

PR:	kern/149227
Submitted by:	Stef Walter <stef memberwebs com>
MFC after:	1 weeks
2010-08-03 11:23:44 +00:00
Alan Cox
2af6e14d39 Introduce exec_alloc_args(). The objective being to encapsulate the
details of the string buffer allocation in one place.

Eliminate the portion of the string buffer that was dedicated to storing
the interpreter name.  The pointer to the interpreter name can simply be
made to point to the appropriate argument string.

Reviewed by:	kib
2010-07-27 17:31:03 +00:00
Konstantin Belousov
a9c9e80bec Revert r210451, and the similar part of the r210431. The forward-declaration
for the enum tag when enum definition is not complete is not allowed by
C99, and is gcc extension.

Requested by:	stefanf
MFC after:	28 days
2010-07-26 12:52:44 +00:00
Alan Cox
9e4e511499 Change the order in which the file name, arguments, environment, and
shell command are stored in exec*()'s demand-paged string buffer.  For
a "buildworld" on an 8GB amd64 multiprocessor, the new order reduces
the number of global TLB shootdowns by 31%.  It also eliminates about
330k page faults on the kernel address space.

Change exec_shell_imgact() to use "args->begin_argv" consistently as
the start of the argument and environment strings.  Previously, it
would sometimes use "args->buf", which is the start of the overall
buffer, but no longer the start of the argument and environment
strings.  While I'm here, eliminate unnecessary passing of "&length"
to copystr(), where we don't actually care about the length of the
copied string.

Clean up the initialization of the exec map.  In particular, use the
correct size for an entry, and express that size in the same way that
is used when an entry is allocated.  The old size was one page too
large.  (This discrepancy originated in 2004 when I rewrote
exec_map_first_page() to use sf_buf_alloc() instead of the exec map
for mapping the first page of the executable.)

Reviewed by:	kib
2010-07-25 17:43:38 +00:00
Konstantin Belousov
0b53d1569e Remove the linux_exec_copyin_args(), freebsd32_exec_copyin_args() may
server as well. COMPAT_FREEBSD32 is a prerequisite for COMPAT_LINUX32.

Reviewed by:	alc
MFC after:	3 weeks
2010-07-23 21:30:33 +00:00
Alan Cox
69a8f9e3d1 Eliminate a little bit of duplicated code. 2010-07-23 18:58:27 +00:00
Edward Tomasz Napierala
49e134dfa6 Remove proc locking, it's not needed after r210132. 2010-07-17 15:52:11 +00:00
Edward Tomasz Napierala
c5dfcf4cc1 Make svr4(4) version of poll(2) use the same limit of file descriptors as the
usual poll(2) does, instead of checking resource limits.
2010-07-15 18:44:58 +00:00
Konstantin Belousov
67322a4cd2 Constify source argument for siginfo_to_siginfo32().
MFC after:	1 week
2010-07-04 11:43:53 +00:00
John Baldwin
ad6eec7b9e Tweak the in-kernel API for sending signals to threads:
- Rename tdsignal() to tdsendsignal() and make it private to kern_sig.c.
- Add tdsignal() and tdksignal() routines that mirror psignal() and
  pksignal() except that they accept a thread as an argument instead of
  a process.  They send a signal to a specific thread rather than to an
  individual process.

Reviewed by:	kib
2010-06-29 20:41:52 +00:00
Konstantin Belousov
13cedde2cb Regenerate 2010-06-28 18:17:21 +00:00
Konstantin Belousov
153ac44cf6 Count number of threads that enter and leave dynamically registered
syscalls. On the dynamic syscall deregistration, wait until all
threads leave the syscall code. This somewhat increases the safety
of the loadable modules unloading.

Reviewed by:	jhb
Tested by:	pho
MFC after:	1 month
2010-06-28 18:06:46 +00:00
Jung-uk Kim
362487c0ba Let x86bios_alloc() pass contigmalloc(9) flags. Use it to set M_WAITOK
from VESA BIOS initialization.  All other malloc(9) uses in the function is
blocking any way.
2010-06-23 17:20:51 +00:00
Ed Schouten
6c9cdb5860 ANSIfy prototypes in subr_usbd.c.
Clang generates the following warnings when building subr_usbd.c:

| subr_usbd.c:598:13: warning: promoted type 'int' of K&R function
|   parameter is not compatible with the parameter type 'uint8_t' (aka
|   'unsigned char') declared in a previous prototype
| subr_usbd.c:627:13: warning: promoted type 'int' of K&R function
|   parameter is not compatible with the parameter type 'uint8_t' (aka
|   'unsigned char') declared in a previous prototype
| subr_usbd.c:649:13: warning: promoted type 'int' of K&R function
|   parameter is not compatible with the parameter type 'uint8_t' (aka
|   'unsigned char') declared in a previous prototype

Instead of just ANSIfying these three prototypes, do it for the entire
file.

Spotted by:	clang
2010-06-12 12:19:08 +00:00
John Baldwin
3aa6d94e0c Update several places that iterate over CPUs to use CPU_FOREACH(). 2010-06-11 18:46:34 +00:00
Wojciech A. Koszek
eedfc35c5c 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
Konstantin Belousov
afe1a68827 Reorganize syscall entry and leave handling.
Extend struct sysvec with three new elements:
sv_fetch_syscall_args - the method to fetch syscall arguments from
  usermode into struct syscall_args. The structure is machine-depended
  (this might be reconsidered after all architectures are converted).
sv_set_syscall_retval - the method to set a return value for usermode
  from the syscall. It is a generalization of
  cpu_set_syscall_retval(9) to allow ABIs to override the way to set a
  return value.
sv_syscallnames - the table of syscall names.

Use sv_set_syscall_retval in kern_sigsuspend() instead of hardcoding
the call to cpu_set_syscall_retval().

The new functions syscallenter(9) and syscallret(9) are provided that
use sv_*syscall* pointers and contain the common repeated code from
the syscall() implementations for the architecture-specific syscall
trap handlers.

Syscallenter() fetches arguments, calls syscall implementation from
ABI sysent table, and set up return frame. The end of syscall
bookkeeping is done by syscallret().

Take advantage of single place for MI syscall handling code and
implement ptrace_lwpinfo pl_flags PL_FLAG_SCE, PL_FLAG_SCX and
PL_FLAG_EXEC. The SCE and SCX flags notify the debugger that the
thread is stopped at syscall entry or return point respectively.  The
EXEC flag augments SCX and notifies debugger that the process address
space was changed by one of exec(2)-family syscalls.

The i386, amd64, sparc64, sun4v, powerpc and ia64 syscall()s are
changed to use syscallenter()/syscallret(). MIPS and arm are not
converted and use the mostly unchanged syscall() implementation.

Reviewed by:	jhb, marcel, marius, nwhitehorn, stas
Tested by:	marcel (ia64), marius (sparc64), nwhitehorn (powerpc),
	stas (mips)
MFC after:	1 month
2010-05-23 18:32:02 +00:00
Alexander Leidinger
eddc400373 - #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
Jung-uk Kim
483191871b Reduce MD code further. At least, it compiles on ia64 now (but it is not
connected to build).  The idea/code was shamelessly taken from r207329.
2010-05-01 01:05:07 +00:00
Jung-uk Kim
2083bca542 Do not initialize mutex and return error if it cannot map memory. 2010-05-01 00:36:40 +00:00
Konstantin Belousov
05e06d1157 Provide compat32 shims for kinfo_proc sysctl. This allows 32bit ps(1) to
mostly work on 64bit host.

The work is based on an original patch submitted by emaste, obtained
from Sandvine's source tree.

Reviewed by:	jhb
MFC after:	1 week
2010-04-21 19:32:00 +00:00
Konstantin Belousov
9847e91b0b Extract the code to copy-out struct rusage32 from struct rusage
into the new function.

Reviewed by:	jhb
MFC after:	1 week
2010-04-21 19:28:01 +00:00
Ed Maste
0a6b4947d8 Linux puts a blank line between each CPU. 2010-04-14 13:44:22 +00:00
Bjoern A. Zeeb
f0564d3739 Add a forward declaration to silence a warning when compiling ia32_genassym.c.
Reviewed by:	kib
MFC after:	3 days
2010-04-03 12:34:32 +00:00
Alexander Leidinger
fc825cc346 Re-apply r205683 with some modifications:
Fix some bogus values in linprocfs.

  Submitted by:	Petr Salinger <Petr.Salinger@seznam.cz>
  Verified on:	GNU/kFreeBSD debian 8.0-1-686 (by submitter)
  PR:		144584

Reviewed by / discussed with:	kib, des, jhb, submitter
2010-04-02 06:50:28 +00:00
Ed Schouten
510ea843ba Rename st_*timespec fields to st_*tim for POSIX 2008 compliance.
A nice thing about POSIX 2008 is that it finally standardizes a way to
obtain file access/modification/change times in sub-second precision,
namely using struct timespec, which we already have for a very long
time. Unfortunately POSIX uses different names.

This commit adds compatibility macros, so existing code should still
build properly. Also change all source code in the kernel to work
without any of the compatibility macros. This makes it all a less
ambiguous.

I am also renaming st_birthtime to st_birthtim, even though it was a
local extension anyway. It seems Cygwin also has a st_birthtim.
2010-03-28 13:13:22 +00:00
Alexander Leidinger
2e2d757e51 Revert r205683 to resolve some code quality issues which do not affect the
build or use of linprocfs, before committing the reworked patch again.

Requested by:	des
2010-03-26 14:36:16 +00:00
Alexander Leidinger
6cdf931c46 Fix some bogus values in linprocfs.
Submitted by:	Petr Salinger <Petr.Salinger@seznam.cz>
Verified on:	GNU/kFreeBSD debian 8.0-1-686 (by submitter)
PR:		144584
2010-03-26 11:43:15 +00:00
Alexander Leidinger
90782c0a14 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
Jung-uk Kim
bfd026445f Revert accidentally committed initial real mode %sp change of r205347.
Note I am keeping %ds change because X.org int10 handler does it and
it seems reasonable.
2010-03-25 17:14:47 +00:00
Jung-uk Kim
d7312c88b4 Optimize real mode page table lookup. 2010-03-25 17:03:52 +00:00
Jung-uk Kim
ad51361a2c Fix stupid typos. Some VESA BIOSes directly call BIOS interrupt handlers
within the VBE interrupt handler.  Unfortunately it was causing real mode
page faults because we were fetching instructions from bogus addresses.
Pass me the pointyhat, please.

PR:		kern/144654
MFC after:	3 days
2010-03-25 15:56:04 +00:00
Nathan Whitehorn
a107d8aac9 Change the arguments of exec_setregs() so that it receives a pointer
to the image_params struct instead of several members of that struct
individually. This makes it easier to expand its arguments in the future
without touching all platforms.

Reviewed by:	jhb
2010-03-25 14:24:00 +00:00
John Baldwin
acfaf56efe Add missing Giant locking for the vfsconf list.
Submitted by:	kib
2010-03-24 14:20:37 +00:00
John Baldwin
663072c6d1 Implement /proc/filesystems.
Submitted by:	Fernando Apesteguia fernando.apesteguia (gmail)
2010-03-23 21:49:33 +00:00
Jung-uk Kim
c9cefec159 Support memory wraparound instead of high memory as VM86 mode does.
Suggested by:	delphij
2010-03-22 18:43:36 +00:00
Jung-uk Kim
a6c8a9c258 Fix i386 PAE kernel build.
Reported by:	tinderbox
2010-03-22 17:30:34 +00:00
Ed Schouten
0fef797f4a Actually make O_DIRECTORY work.
According to POSIX open() must return ENOTDIR when the path name does
not refer to a path name. Change vn_open() to respect this flag. This
also simplifies the Linuxolator a bit.
2010-03-21 20:43:23 +00:00
Jung-uk Kim
ef8201d39d - Map EBDA if available and add 64KB above 1MB (high memory), just in case.
- Print the initial memory map when bootverbose is set.
- Change the page fault address format from linear to %cs:%ip style.
- Move duplicate code into a newly added function.
- Add strictly aligned memory access for distant future. ;-)
2010-03-19 21:15:43 +00:00