112245 Commits

Author SHA1 Message Date
adrian
24494b0ddf Remove bus space fdt for MIPS.
This was originall done by kan@.

Submitted by:	Stanislav Galabov <sgalabov@gmail.com>
Reviewed by:	kan
Differential Revision:	https://reviews.freebsd.org/D5184
2016-02-11 06:19:32 +00:00
adrian
93a38902f4 Convert MIPS nexus and mips_pic to BUS_PASS
Submitted by:	Stanislav Galabov <sgalabov@gmail.com>
Reviewed by:	kan
Differential Revision:	https://reviews.freebsd.org/D5196
2016-02-11 06:15:43 +00:00
adrian
570389571c Teach the MIPS ticker to attach itself properly when using INTRNG.
Submitted by:	Stanislav Galabov <sgalabov@gmail.com>
Reviewed by:	kan
Differential Revision:	https://reviews.freebsd.org/D5183
2016-02-11 06:13:53 +00:00
adrian
dd3b3e6ad6 Begin the MIPS_INTRNG support.
This is a prelude to intr-ng support for MIPS boards that need it -
notably the CI20 port from kan@ that's upcoming, but also work that
Stanislav is doing for the Mediatek platforms.

This is the initial platform dependent bits in include/intr.h, some
#defines for the nexus code for the intrng initialisation/runtime
bits, some changed naming (which I'll fix later to be the same, much
like what I did for ARM intr-ng) in exception.S, and the first cut
at a PIC.

Stanislav and I refactored out the common code for intrng support,
so the mips intrng definitions are quite small (sys/mips/include/intr.h.)

This is all work done by kan@, which stanislav has been cherry picking
into common code for his mediatek chipset work.

Tested:

* Carambola2 - no regressions (not intr-ng though!)

Submitted by:	Stanislav Galabov <sgalabov@gmail.com>
Reviewed by:	kan (original author)
Differential Revision:	https://reviews.freebsd.org/D5182
2016-02-11 06:09:27 +00:00
pfg
cdb1cc5394 Ext4: fix handling of files with sparse blocks before extent's index.
This is ongoing work from Damjan Jovanovic to improve ext4 read support
with sparse files:

Keep track of the first and last block in each extent as it descends down
the extent tree, thus being able to work out that some blocks are sparse
earlier. This solves an issue on r293680.

In ext4_bmapext() start supporting the runb parameter, which appears to be
the number of adjacent blocks prior to the block being converted in the
same way that runp is the number of blocks after, speding up random access
to mmaped files.

PR:	206652
2016-02-11 00:34:11 +00:00
kib
5fefd35412 Remove useless checks for NULL before calling free(9), in the kernel
elf linkers.

Found by:	Related PVS-Studio diagnostic
Sponsored by:	The FreeBSD Foundation
MFC after:	1 week
2016-02-10 21:35:00 +00:00
kib
fd3e233ddd Finish r173600. There is no need to test a condition if both cases
result in the same value.

Found by:	PVS-Studio
Sponsored by:	The FreeBSD Foundation
MFC after:	1 week
2016-02-10 21:16:37 +00:00
cem
954cde6c76 ntb_hw(4): Print correct PAT name for non-WC/WB types mapped at load
Sponsored by:	EMC / Isilon Storage Division
2016-02-10 20:49:22 +00:00
cem
1e46ac5ade ntb_hw(4): Allow any x86 PAT caching flags for MW defaults
Replace the hw.ntb.enable_writecombine tunable with
hw.ntb.default_mw_pat.  It can be set with several specific numerical
values to select a caching type.  Any bogus value is treated as
Uncacheable (UC).

The ntb_mw_set_wc() KPI has removed the restriction that the selected
mode must be one of UC, WC, or WB.

Sponsored by:	EMC / Isilon Storage Division
2016-02-10 20:28:28 +00:00
trasz
e75b8cbe27 Add a kern.icl.drivers sysctl, to retrieve the list of registered
ICL drivers.

MFC after:	1 month
Sponsored by:	The FreeBSD Foundation
2016-02-10 19:01:26 +00:00
glebius
bb999f3ee9 Garbage collect unused arguments of m_init(). 2016-02-10 18:54:18 +00:00
glebius
3148d24a20 Garbage collect m_getclr(). 2016-02-10 18:36:51 +00:00
trasz
97d8d42022 Remove stray semicolons from the iSCSI code.
MFC after:	1 month
Sponsored by:	The FreeBSD Foundation
2016-02-10 17:55:39 +00:00
andrew
352dc89128 Explicitly include <sys/systm.h>. Previously we were pulling it in due to
header pollution.
2016-02-10 16:49:20 +00:00
andrew
8491692074 Fix a logic inversion when checking for ARM_INTRNG
Pointy hat to:	andrew
2016-02-10 16:44:20 +00:00
andrew
3f190828b4 Only use fdt_pic_table when not using ARM_INTRNG 2016-02-10 16:39:02 +00:00
andrew
1e16bdfe11 Use OF_getencprop to read a property, it already handles the endian
conversion so we don't need to pass the data through fdt32_to_cpu.
2016-02-10 15:56:52 +00:00
arybchik
983d20d844 sfxge: implement SIOCGI2C to read information from phy modules
The IOCTL is used by 'ifconfig -v' to show SFP+/QSFP+ information
including inventory information and dianostics (temperature, light
levels, voltage etc).

Reviewed by:    gnn,melifaro
Sponsored by:   Solarflare Communications, Inc.
MFC after:      2 days
Differential Revision: https://reviews.freebsd.org/D5240
2016-02-10 12:14:56 +00:00
andrew
bfa435406f Revert an arm64 change that sneaked in with r295464. 2016-02-10 10:28:33 +00:00
andrew
4cf4fddec8 Update of the Allwinner drivers to:
* Use the Linux compat string
 * Use EARLY_DRIVER_MODULE to attach at the right time
 * Add a generic A10 kernel config file
 * A20 now use generic_timer
 * Add two new dts files for Olimex boards
 * Update our custom DTS file for A10 and A20 to use the same compatible
   property names as the vendor ones.

Submitted by:	Emmanuel Vadot <manu@bidouilliste.com>
Differential Revision:	https://reviews.freebsd.org/D4792
2016-02-10 09:19:29 +00:00
hselasky
517ee62210 Correct PCI device description.
Submitted by:	Dmitry Luhtionov <dmitryluhtionov@gmail.com>
2016-02-10 08:03:10 +00:00
adrian
6d8e3abce2 Break out the shared bits of the arm intrng definitions into sys/intr.h;
leave the machine dependent bits in sys/arm/.

This is in preparation for MIPS INTRNG work.

Submitted by:	Stanislav Galabov <sgalabov@gmail.com>
2016-02-10 04:43:08 +00:00
glebius
b3c4f0ddbf Include sys/_task.h into uma_int.h, so that taskqueue.h isn't a
requirement for uma_int.h.

Suggested by:	jhb
2016-02-09 20:22:35 +00:00
andrew
f8539e89c5 Remove unneeded whitespace, and fix a license for a new file. Extracted
from a larger patch.

Submitted by:	Emmanuel Vadot <manu@bidouilliste.com>
X-Differential Revision:	https://reviews.freebsd.org/D4792
2016-02-09 18:19:24 +00:00
hselasky
4a8d8f7fba To support userspace audio daemons like Virtual OSS, /dev/sndstat is
made writeable by the root user. Userspace audio daemons can add or
update an entry in /dev/sndstat by doing a single system write call to
any /dev/sndstat file descriptor handle. When the audio daemon closes the
file handle or is killed the entry disappears.

While at it, cleanup the sound status code a bit:
- keep the device list sorted to avoid sorting the list every time a
  /dev/sndstat read request is made.
- factor out locking into a pair of locking macros.
- use the sound status lock to protect all per file handle states,
  when generating the output for /dev/sndstat and when removing or
  adding sound status devices. This way sndstat_acquire() and
  sndstat_release() become superfluous and can be removed.

Reviewed by:		mav @
Differential Revision:	https://reviews.freebsd.org/D5191
2016-02-09 17:09:14 +00:00
andrew
327efa1f37 Update our copy of the Linux dts files to be in sync with Linux 4.5-rc1. We
now uave a full copy of the arm device tree sources to help when adding
support for newer boards.

Sponsored by:	ABT Systems Ltd
2016-02-09 16:42:32 +00:00
kib
736e078495 Rename P_KTHREAD struct proc p_flag to P_KPROC.
I left as is an apparent bug in ntoskrnl_var.h:AT_PASSIVE_LEVEL()
definition.

Suggested by:	jhb
Sponsored by:	The FreeBSD Foundation
2016-02-09 16:30:16 +00:00
andrew
6b818b6371 Use designated initialisers for the db disassembler interface to help with
finding the struct when searching for one of its members.
2016-02-09 10:18:22 +00:00
andrew
9b8ae2bdc3 Don't set the offset when loading the kernel on the arm loader.efi. The
copyin and copyout code handle virtual addresses such that they will take
a virtual address and convert it into a valid physical address. It may
also mean we fail to boot as the elf files load address could be 0.

Sponsored by:	ABT Systems Ltd
2016-02-09 09:39:30 +00:00
wma
db0cfe8402 Ignore invalid page descriptors in ARM64 pmap_mincore
Prevent the function from null-pointer-dereference when unexisting
    mapping is being processed.

Obtained from:         Semihalf
Sponsored by:          Cavium
Approved by:           cognet (mentor)
Reviewed by:           zbb, cognet
Differential revision: https://reviews.freebsd.org/D5228
2016-02-09 06:26:27 +00:00
adrian
93fc03e6d8 Teach ofw_bus_parse_xref_list_alloc to be able to return the length of the parsed list.
Currently, there is no easy way to know in advance how many entries a list parsed by
ofw_bus_parse_xref_list_alloc() in sys/dev/ofw/ofw_bus_subr.c has.

This patch:

* teaches the existing function about handling idx == -1 and returning how big
  the set is; then renames it as _internal;
* create a new function that asserts idx != -1, so the old API is maintained;
* add a new function that returns just the list length.

Submitted by:	Stanislav Galabov <sgalabov@gmail.com>
Differential Revision:	https://reviews.freebsd.org/D5043
2016-02-09 03:35:40 +00:00
andrew
ae802bf627 Stop clang from generating movt/movw sequences. These don't get relocated
so will access data from an unrelocated address. This is only needed for
self relocating code on ARMv7, however this is true for both ubldr and
loader.efi, the only two loaders we support on ARMv7.

While here also force the fpu to be none as is done in libstand.

Sponsored by:	ABT Systems Ltd
2016-02-09 00:15:41 +00:00
andrew
cf9bbb5d8a Ensure the bss is aligned to a 4-byte address as we use 4-byte aligned
stores to clear it.

While here reduce the alignment of the data from 4k to 16 byte aligned.
This should be more than enough, without wasting too much space.

Sponsored by:	ABT Systems Ltd
2016-02-09 00:01:43 +00:00
jhb
f6d0271729 Call kthread_exit() rather than kproc_exit() for a premature kthread exit.
Kernel threads (and processes) are supposed to call kthread_exit() (or
kproc_exit()) to terminate.  However, the kernel includes a fallback in
fork_exit() to force a kthread exit if a kernel thread's "main" routine
returns.  This fallback was added back when the kernel only had processes
and was not updated to call kthread_exit() instead of kproc_exit() when
threads were added to the kernel.

This mistake was particular exciting when the errant thread belonged to
proc0.  Due to the missing P_KTHREAD flag the fallback did not kick in
and instead tried to return to userland via whatever garbage was in the
trapframe.  With P_KTHREAD set it tried to terminate proc0 resulting in
other amusements.

PR:		204999
MFC after:	1 week
2016-02-08 23:11:23 +00:00
jhb
9fbc164e2b Mark proc0 as a kernel process via the P_KTHREAD flag.
All other kernel processes have this flag set and all threads in proc0
(including thread0) have the similar TDP_KTHREAD flag set.

PR:		204999
Submitted by:	Oliver Pinter @ HardenedBSD
Reviewed by:	kib
MFC after:	1 week
2016-02-08 23:06:27 +00:00
ken
63a60c271f Fix the SCSI Extended INQUIRY probe case when an error is returned
and a retry is scheduled.

Instead of leaving the device queue frozen, unfreeze the device queue so
that the retry can happen.

Sponsored by:	Spectra Logic
MFC after:	3 days
2016-02-08 22:13:08 +00:00
bdrewery
fd3581bd1c kmod.mk: Ensure ILINKS are created before building.
This allows skipping 'make depend' or running 'make clean all' without
getting a flip-flopping dependency due to the exists() just below.
Otherwise an error is encountered, such as:
  fatal error: 'machine/endian.h' file not found.

Sponsored by:	EMC / Isilon Storage Division
2016-02-08 20:58:52 +00:00
bdrewery
5f89cdb598 kmod.mk: Support a beforebuild target such as bsd.prog.mk has.
Sponsored by:	EMC / Isilon Storage Division
2016-02-08 20:57:12 +00:00
pfg
7a22e1974d ficl: Replace rand(3) with random(3).
Be a little more consistent with random(3) and push an
unsigned value.

Again, this has no effect as this code doesn't get compiled
for the boot code.
2016-02-08 20:03:14 +00:00
pfg
ea701e02f2 ficl: Replace rand(3) with random(3).
While the later is a better random generator than the former, the main
reason of the change is that random() has a better chance to work with
libstand(3).

At this time we don't include random number generators in bootforth
so this has no effect.
2016-02-08 19:45:55 +00:00
imp
cee34e3864 Implement -P for boot loader. It's a bit easier to implement here than
in boot1, like is normally done. When a keyboard appears in the UEFI
device tree, assume -D -h, just like on a BIOS boot.

# It is unclear if an ACPI keyboard appearing in the tree means there's
# a real keyboard or not. A USB keyboard doesn't seem to appear unless
# it is really there.

Differential Revision: https://reviews.freebsd.org/D5223
2016-02-08 19:34:17 +00:00
emaste
020d1b050f Restore kgzldr on i386 only
I previously disconnected kgzdr based on a misunderstanding.

I'd still like to transition to supporting only the loader(8)-based
boot path for handling compressed kernels, but that can follow the
standard deprecation procedure.

This reverts r291113.

Requested by:	dteske
2016-02-08 16:59:52 +00:00
kib
2ed1e2991e Remove the assert which outlived its usefulness, and, by default,
disable compilation of the code which made it possible to call
stop_all_proc() from usermode at all.

Move the comment to the preamble of stop_all_proc() and reword it to
give overview of the function intent.

proc0 has P_HADTHREADS flag set due to kthread_add(), but no
P_KTHREAD, which triggered the assert, which does not serve a purpose
now.

Reported by:	Oliver Pinter
Sponsored by:	The FreeBSD Foundation
MFC after:	2 weeks
2016-02-08 10:54:27 +00:00
jilles
b2791f185c semget(): Check for [EEXIST] error first.
Although POSIX literally permits failing with [EINVAL] if IPC_CREAT and
IPC_EXCL were both passed, the semaphore set already exists and has fewer
semaphores than nsems, this does not allow an application to retry safely:
if the [EINVAL] is actually because of the semmsl limit, an infinite loop
would result.

PR:		206927
2016-02-07 22:12:39 +00:00
pfg
8d82ef5201 Minor grammar fix in comment. 2016-02-07 16:18:12 +00:00
pfg
d7b2b433b4 Revert r295359:
CID 1018688 is a false positive.

The initialization is done by calling vn_start_write(... &mp, flags).
mp is only an output parameter unless (flags & V_MNTREF), and fdesc
doesn't put V_MNTREF in flags.

Pointed out by:	bde
2016-02-07 15:40:01 +00:00
pfg
b42dfac655 msdosfs_rename: yet another unused value.
As with r295355, it seems to be left over from a cleanup
in r33548. The code is not in NetBSD either.

Thanks to bde for checking out the history.
2016-02-07 15:36:16 +00:00
pfg
29ef016884 cd9660: Drop an unnecessary check for NULL.
This was unnecessary and also confused Coverity.

Confirmed on:	NetBSD
CID:		978558
2016-02-07 03:48:40 +00:00
pfg
0bbadbe82b fdesc_setattr: unitialized pointer read
CID:	1018688
2016-02-07 01:09:38 +00:00
mckusick
3ddfa78dde Clarify a comment in kern_openat() about the use of falloc_noinstall().
Suggested by: Steve Jacobson
2016-02-07 01:04:47 +00:00