dim [Thu, 19 Feb 2015 22:20:19 +0000 (22:20 +0000)]
Merge llvm 3.6.0rc4 from ^/vendor/llvm/dist, merge clang 3.6.0rc4 from
^/vendor/clang/dist, resolve conflicts, and update patches.
dim [Thu, 19 Feb 2015 21:10:01 +0000 (21:10 +0000)]
Merging ^/head r278916 through r279022.
dim [Thu, 19 Feb 2015 20:55:17 +0000 (20:55 +0000)]
Vendor import of llvm RELEASE_360/rc4 tag r229772 (effectively, 3.6.0 RC4):
https://llvm.org/svn/llvm-project/llvm/tags/RELEASE_360/rc4@229772
imp [Thu, 19 Feb 2015 20:22:30 +0000 (20:22 +0000)]
Only disable gvn on clang 3.5 and newer.
glebius [Thu, 19 Feb 2015 19:36:54 +0000 (19:36 +0000)]
Burn bridges to FreeBSD 7.x IGMP stats.
ed [Thu, 19 Feb 2015 18:56:39 +0000 (18:56 +0000)]
Make syslogd work in case shutdown() is POSIX-ly correct.
On POSIX conformant systems, shutdown() should return ENOTCONN when not
connected. We attempted to fix this once (kern/84761), but this change
got backed out because it 'breaks code' (r150155).
I just reapplied the patch and indeed, syslogd fails on startup. Make it
easier to re-enable this change in the future by paching up syslogd to
do the right thing.
MFC after: 3 weeks
Sponsored by: Nuxi
glebius [Thu, 19 Feb 2015 17:03:13 +0000 (17:03 +0000)]
Use KTR_COMPILE=(KTR_ALL) for LINTs, to get more code coverage.
andrew [Thu, 19 Feb 2015 16:34:48 +0000 (16:34 +0000)]
Adda function to iterate over the cpu nodes in the OFW or FDT data. This
will be used with arm64 to find which cpus to enable, and could also be
used with 32-bit arm and mips for the same purpose.
Differential Revision: https://reviews.freebsd.org/D1825
Sponsored by: The FreeBSD Foundation
kib [Thu, 19 Feb 2015 15:41:23 +0000 (15:41 +0000)]
Fix build with KTR after r278978.
andrew [Thu, 19 Feb 2015 12:47:48 +0000 (12:47 +0000)]
Add support to get the cpu ID from its device driver in a generic way.
This will be needed by arm64 to find the value to pass to the psci (Power
State Coordination Interface) driver, among other things, used to enable
cores.
Differential Revision: https://reviews.freebsd.org/D1824
Reviewed by: imp
Sponsored by: The FreeBSD Foundation
andrew [Thu, 19 Feb 2015 12:20:21 +0000 (12:20 +0000)]
Use the ARM unwinder with dtrace to extract the stack when asked. With this
dtrace is able to display a stack trace similar to the one below.
# dtrace -p 603 -n 'tcp:kernel::receive { stack(); }'
0 70 :receive
kernel`ip_input+0x140
kernel`netisr_dispatch_src+0xb8
kernel`ether_demux+0x1c4
kernel`ether_nh_input+0x3a8
kernel`netisr_dispatch_src+0xb8
kernel`ether_input+0x60
kernel`cpsw_intr_rx+0xac
kernel`intr_event_execute_handlers+0x128
kernel`ithread_loop+0xb4
kernel`fork_exit+0x84
kernel`swi_exit
kernel`swi_exit
Tested by: gnn
Sponsored by: ABT Systems Ltd
andrew [Thu, 19 Feb 2015 12:06:57 +0000 (12:06 +0000)]
Allow the ARM unwinder to work through modules. This will be used to add
support for unwinding from dtrace.
Tested by: gnn (with dtrace)
Sponsored by: ABT Systems Ltd
bz [Thu, 19 Feb 2015 11:48:00 +0000 (11:48 +0000)]
Try to unbreak NOIP and NOINET6 LINT builds after r278886
by placing appropriate #ifdefs around otherwise unused variables
or sections with functions called which are not available without
IPv6 support in the kernel.
ngie [Thu, 19 Feb 2015 11:17:36 +0000 (11:17 +0000)]
Unbreak freshly installed worlds by properly "commenting" out ifmcstat
Pointyhat to: glebius
glebius [Thu, 19 Feb 2015 06:27:14 +0000 (06:27 +0000)]
Temporarily disconnect ifmcstat(8) from build, to make world buildable
ifmcstat(8) noses in kernel memory too much, and thus is very tentative
to any changes in kernel.
I will rewrite it to use some API instead of libkvm(3) and connect back
to build.
glebius [Thu, 19 Feb 2015 06:24:27 +0000 (06:24 +0000)]
Widen _KERNEL ifdef to hide more kernel network stack structures from userland.
gjb [Thu, 19 Feb 2015 05:20:59 +0000 (05:20 +0000)]
Fix a grammar nit.
Sponsored by: The FreeBSD Foundation
markj [Thu, 19 Feb 2015 03:32:48 +0000 (03:32 +0000)]
Remove unnecessary checks for a return value of NULL from M_WAITOK
allocations.
MFC after: 3 days
markj [Thu, 19 Feb 2015 03:29:46 +0000 (03:29 +0000)]
Free the zlib stream after expanding a compressed CTF section.
Note that this memory would only be leaked once, since CTF info for a kld
file is cached after the first access.
MFC after: 3 days
jmg [Thu, 19 Feb 2015 01:51:17 +0000 (01:51 +0000)]
fix spelling, add comma and remove BUGS section.. it provided no useful
information, and is not really bugs, but limitations for other reasons...
glebius [Thu, 19 Feb 2015 01:21:23 +0000 (01:21 +0000)]
Use new struct mbufq instead of struct ifqueue to manage packet queues in
IPv6 multicast code.
Sponsored by: Netflix
Sponsored by: Nginx, Inc.
glebius [Thu, 19 Feb 2015 01:21:02 +0000 (01:21 +0000)]
Use new struct mbufq instead of struct ifqueue to manage packet queues in
IPv4 multicast code.
Sponsored by: Netflix
Sponsored by: Nginx, Inc.
glebius [Thu, 19 Feb 2015 01:19:42 +0000 (01:19 +0000)]
Provide a set of inline functions to manage simple mbuf(9) queues, based
on queue(3)'s STAILQ. Utilize them in cxgb(4) and Xen, deleting home
grown implementations.
Sponsored by: Netflix
Sponsored by: Nginx, Inc.
jhb [Wed, 18 Feb 2015 23:34:03 +0000 (23:34 +0000)]
Ensure that the supplied data length is large enough to hold the base
FPU state to avoid passing a negative length to fpusetregs() / npxsetregs().
Differential Revision: https://reviews.freebsd.org/D1861
Reviewed by: kib, emaste
jkim [Wed, 18 Feb 2015 23:10:15 +0000 (23:10 +0000)]
Fix bootstrap on systems with old yacc with small MAXTABLE.
jkim [Wed, 18 Feb 2015 20:33:00 +0000 (20:33 +0000)]
Merge ACPICA
20141107 and
20150204.
grehan [Wed, 18 Feb 2015 19:27:51 +0000 (19:27 +0000)]
Restore the ability to use clang as an external compiler. This was
inadvertently removed when support for external GCC was added.
Deprecate XFLAGS in favour of the newer XCFLAGS/XCXXFLAGS.
Tested with: make universe, make CROSS_COMPILER_PREFIX=/usr/bin/ buildworld
Reviewed by: imp, bapt
ken [Wed, 18 Feb 2015 18:30:19 +0000 (18:30 +0000)]
Make sure that the flags for the XPT_DEV_ADVINFO CCB are initialized
properly.
If there is garbage in the flags field, it can sometimes include a
set CDAI_FLAG_STORE flag, which may cause either an error or
perhaps result in overwriting the field that was intended to be
read.
sys/cam/cam_ccb.h:
Add a new flag to the XPT_DEV_ADVINFO CCB, CDAI_FLAG_NONE,
that callers can use to set the flags field when no store
is desired.
sys/cam/scsi/scsi_enc_ses.c:
In ses_setphyspath_callback(), explicitly set the
XPT_DEV_ADVINFO flags to CDAI_FLAG_NONE when fetching the
physical path information. Instead of ORing in the
CDAI_FLAG_STORE flag when storing the physical path, set
the flags field to CDAI_FLAG_STORE.
sys/cam/scsi/scsi_sa.c:
Set the XPT_DEV_ADVINFO flags field to CDAI_FLAG_NONE when
fetching extended inquiry information.
sys/cam/scsi/scsi_da.c:
When storing extended READ CAPACITY information, set the
XPT_DEV_ADVINFO flags field to CDAI_FLAG_STORE instead of
ORing it into a field that isn't initialized.
sys/dev/mpr/mpr_sas.c,
sys/dev/mps/mps_sas.c:
When fetching extended READ CAPACITY information, set the
XPT_DEV_ADVINFO flags field to CDAI_FLAG_NONE instead of
setting it to 0.
sbin/camcontrol/camcontrol.c:
When fetching a device ID, set the XPT_DEV_ADVINFO flags
field to CDAI_FLAG_NONE instead of 0.
sys/sys/param.h:
Bump __FreeBSD_version to 1100061 for the new XPT_DEV_ADVINFO
CCB flag, CDAI_FLAG_NONE.
Sponsored by: Spectra Logic
MFC after: 1 week
kib [Wed, 18 Feb 2015 18:12:06 +0000 (18:12 +0000)]
If malloc() sleeps, Giant is dropped. Recheck for another thread
doing our work.
Remove unneeded check for failed M_WAITOK allocation.
Reviewed by: alc
Sponsored by: The FreeBSD Foundation
MFC after: 1 week
marcel [Wed, 18 Feb 2015 17:28:55 +0000 (17:28 +0000)]
Add LIBXO.
imp [Wed, 18 Feb 2015 15:25:25 +0000 (15:25 +0000)]
Clean up a bit of the INET/INET6 mess wrt options.
imp [Wed, 18 Feb 2015 15:25:22 +0000 (15:25 +0000)]
Cleanup the way we include some files (in this case EISA) based on
kernel config. Also tidy up the ed options.
imp [Wed, 18 Feb 2015 15:25:19 +0000 (15:25 +0000)]
Populate new KERN_OPTS from all the opt_*.h files in
KERNBUILDDIR. Come up with some sensible defaults (though listing them
in kmod.mk may be unwise -- we have no easy way to know what are the
best sensible defaults for everything so we just catch the big stuff).
Append SRCS.${opt} for each option in KERN_OPTS to SRCS to allow easy
conditional compilation. Append any notion of KERN_OPTS_EXTRA to the
list of kernel opts.
Differential Revision: https://reviews.freebsd.org/D1530
imp [Wed, 18 Feb 2015 14:33:33 +0000 (14:33 +0000)]
Add copyright statement I should have had on these files in 2006,
based on the spibus_if.m file.
mjg [Wed, 18 Feb 2015 13:37:28 +0000 (13:37 +0000)]
filedesc: obtain a stable copy of credentials in fget_unlocked
This was broken in r278930.
While here tidy up fget_mmap to use fdp from local var instead of obtaining
the same pointer from td.
kib [Wed, 18 Feb 2015 10:51:48 +0000 (10:51 +0000)]
Fix UP build.
Sponsored by: The FreeBSD Foundation
MFC after: 2 months
delphij [Wed, 18 Feb 2015 08:24:00 +0000 (08:24 +0000)]
Bump revision again after my Fortuna change.
delphij [Wed, 18 Feb 2015 08:21:51 +0000 (08:21 +0000)]
- fortuna.c: catch up with r278927 and fix a buffer overflow by using the
temporary buffer when remaining space is not enough to hold
a whole block.
- yarrow.c: add a comment that we intend to change the code and remove
memcpy's in the future. (*)
Requested by: markm (*)
Reviewed by: markm
Approved by: so (self)
jhibbits [Wed, 18 Feb 2015 07:34:32 +0000 (07:34 +0000)]
Match the right backlight driver.
Some ATI-based PowerBooks use the string 'mnca' in the backlight controller
device tree entry, so account for this and don't use nVidia when it's not an
nVidia device.
MFC after: 3 weeks
jhibbits [Wed, 18 Feb 2015 06:53:40 +0000 (06:53 +0000)]
Don't set the write bit if we're just reading.
Also fix a couple typos.
MFC after: 3 weeks
arybchik [Wed, 18 Feb 2015 06:26:07 +0000 (06:26 +0000)]
sfxge: fix broken MCDI_EV_FIELD() macro
Submitted by: Andrew Lee <alee at solarflare.com>
Sponsored by: Solarflare Communications, Inc.
Approved by: gnn (mentor)
arybchik [Wed, 18 Feb 2015 06:23:16 +0000 (06:23 +0000)]
sfxge: support variable-length response to MCDI GET_BOARD_CFG
Allocate the minimum or maximum response length for GET_BOARD_CFG as
appropriate. When looking up firmware subtypes by partition ID,
check the ID against the actual response length.
Merge of the patch made by Ben Hutchings in 2011.
Sponsored by: Solarflare Communications, Inc.
Approved by: gnn (mentor)
arybchik [Wed, 18 Feb 2015 06:21:59 +0000 (06:21 +0000)]
sfxge: add driver context member with number of event queues
Mainly to unify with similar member for transmit and receive queues.
It will be used in the future for resources allocation processing.
Sponsored by: Solarflare Communications, Inc.
Approved by: gnn (mentor)
arybchik [Wed, 18 Feb 2015 06:20:59 +0000 (06:20 +0000)]
sfxge: add driver context member with number of receive queues
Mainly to unify with similar member for transmit queues.
Sponsored by: Solarflare Communications, Inc.
Approved by: gnn (mentor)
arybchik [Wed, 18 Feb 2015 06:19:35 +0000 (06:19 +0000)]
sfxge: add driver context member with number of transmit queues
Sponsored by: Solarflare Communications, Inc.
Approved by: gnn (mentor)
arybchik [Wed, 18 Feb 2015 06:18:51 +0000 (06:18 +0000)]
sfxge: add TCP segment size to sfxge_tso_state
It avoids access to m_pkthdr when TSO packet is started and also makes
tso_start_new_packet() function smaller.
Sponsored by: Solarflare Communications, Inc.
Approved by: gnn (mentor)
imp [Wed, 18 Feb 2015 05:53:04 +0000 (05:53 +0000)]
On my Lenovo T400, a Atheros 2413 has a problem powering up
sometimes. It will power up wrong and identify itself badly:
cardbus0: <network, ethernet> at device 0.0 (no driver attached)
cardbus0: <simple comms, UART> at device 0.1 (no driver attached)
cardbus0: <old, non-VGA display device> at device 0.2 (no driver attached)
cardbus0: <old, non-VGA display device> at device 0.3 (no driver attached)
cardbus0: <old, non-VGA display device> at device 0.4 (no driver attached)
cardbus0: <old, non-VGA display device> at device 0.5 (no driver attached)
cardbus0: <old, non-VGA display device> at device 0.6 (no driver attached)
cardbus0: <old, non-VGA display device> at device 0.7 (no driver attached)
All the higher numbered functions (.2 and above) have a config space
of all 0's. This smells a bit like a special debug mode, but the
current atheros driver doesn't cope. It is unclear if this card is
just a flake, or if we're doing something wrong in the power-up
sequence.
Put a work around into the code that tests for this rather unusual
condition. If we power a CardBus device up, and the device says it is
multi-function, and any of the functions have a 0 device ID, try the
power-up sequence again.
markj [Wed, 18 Feb 2015 03:54:54 +0000 (03:54 +0000)]
Remove drti.o's dependency on libelf. This makes it possible to add DTrace
probes to userland programs and libraries without also needing to link
libelf.
dtrace -G places the __SUNW_dof symbol at the beginning of the DOF (DTrace
probe and provider metdata) section in the generated object file; drti.o
now just uses this symbol to locate the section. A complication occurs
when multiple dtrace-generated object files are linked together, since the
__SUNW_dof symbol defined in each file is global. This is handled by
using objcopy(1) to convert __SUNW_dof to a local symbol once drti.o has
been linked with the generated object file. Upstream, this is done using a
linker feature not present in GNU ld.
Differential Revision: https://reviews.freebsd.org/D1757
Reviewed by: rpaulo
MFC after: 1 month
Relnotes: yes
rpaulo [Wed, 18 Feb 2015 03:46:43 +0000 (03:46 +0000)]
Fix a typo in ipv6_down().
We weren't skipping extraneous entries in the ifconfig | grep inet6 case.
Submitted by: Ashutosh Kumar AK0037447 at TechMahindra.com
MFC after: 1 week
pfg [Wed, 18 Feb 2015 03:33:17 +0000 (03:33 +0000)]
libc: clean some set-but-not-used errors.
These were found by gcc 5.0 on Dragonfly BSD, however I
made no attempt to silence the false positives.
Obtained from: DragonFly (
cf515c3a6f3a8964ad592e524442bc628f8ed63b)
mjg [Tue, 17 Feb 2015 23:54:06 +0000 (23:54 +0000)]
filedesc: simplify fget_unlocked & friends
Introduce fget_fcntl which performs appropriate checks when needed.
This removes a branch from fget_unlocked.
Introduce fget_mmap dealing with cap_rights_to_vmprot conversion.
This removes a branch from _fget.
Modify fget_unlocked to pass sequence counter to interested callers so
that they can perform their own checks and make sure the result was
otained from stable & current state.
Reviewed by: silence on -hackers
jmg [Tue, 17 Feb 2015 23:41:08 +0000 (23:41 +0000)]
bump the rev... ZFS is the first fall out of this bug, but there might
be others...
pfg [Tue, 17 Feb 2015 23:20:19 +0000 (23:20 +0000)]
ministat(1): replace malloc + memset with calloc.
Reviewed by: phk
jmg [Tue, 17 Feb 2015 23:14:26 +0000 (23:14 +0000)]
Fix a bug where this function overflowed it's buffer... This was
causing ZFS panics on boot...
This is purely reviewed and tested by peter.
Reviewed by: peter
Approved by: so (implicit), peter
rpaulo [Tue, 17 Feb 2015 23:13:45 +0000 (23:13 +0000)]
release: use xz via pipe when compressing the tarballs.
libarchive(3) doesn't support the new liblzma API yet, but this change
allows us to enable multi-threaded xz compression.
``make release'' should now finish in half the time on a machine with
several cores and fast disks (our typical build server).
This behaviour only applies when building a release and it doesn't
affect buildworld/installworld. To disable threaded xz compression,
set XZ_THREADS=1.
Reviewed by: gjb
Tested by: gjb
glebius [Tue, 17 Feb 2015 22:33:22 +0000 (22:33 +0000)]
Even more fixes to !INET and !INET6 kernels.
In collaboration with: pluknet
hselasky [Tue, 17 Feb 2015 21:59:15 +0000 (21:59 +0000)]
Fix compilation of LINT-NOINET kernel target after r278886.
Sponsored by: Mellanox Technologies
MFC after: 1 month
jmg [Tue, 17 Feb 2015 21:18:17 +0000 (21:18 +0000)]
note that you should update your kernel due to RNG issue, and what
versions you should upgrade...
Suggested by: Ryan Stone
pfg [Tue, 17 Feb 2015 21:12:45 +0000 (21:12 +0000)]
du(1): replace malloc + memset with calloc.
glebius [Tue, 17 Feb 2015 20:52:51 +0000 (20:52 +0000)]
Use anonymous unions and structs to organize shared space in mbuf(9),
instead of preprocessor macros.
This will make debugger output of 'print *m' exactly match the names
we use in code, making life of a kernel hacker way more pleasant. And
this also allows to rename struct_m_ext back to m_ext.
loos [Tue, 17 Feb 2015 20:37:21 +0000 (20:37 +0000)]
Make use of the newly introduced macros.
Update the copyright.
loos [Tue, 17 Feb 2015 20:29:42 +0000 (20:29 +0000)]
Replace hardcoded sizes by sizeof() and nitems().
Set all the sysctls as CTLFLAG_MPSAFE.
loos [Tue, 17 Feb 2015 20:08:40 +0000 (20:08 +0000)]
Fix the display of negative temperatures.
Fix the setting of TOS (overtemperature shutdown) and THYST (hysteresis
temperature).
dim [Tue, 17 Feb 2015 19:53:41 +0000 (19:53 +0000)]
Merge ^/head r278756 through r278915.
loos [Tue, 17 Feb 2015 19:38:16 +0000 (19:38 +0000)]
Set the #address-cells and #size-cells properties on I2C controllers of
Raspberry-pi.
This fixes the issues of dtc(1) while compiling I2C addresses set in "reg"
property.
MFC after: 2 weeks
glebius [Tue, 17 Feb 2015 19:32:11 +0000 (19:32 +0000)]
Use anonymous unions to add possibility to put mbufs into queue(3)
STAILQs and SLISTs using the same structure field as good old m_next
and m_nextpkt linkage occupy.
New code is encouraged to use queue(3) macros, instead of implementing
the wheel. However, better not to have a mixture of old style and
queue(3) in one file or subsystem.
Reviewed by: rwatson, rrs, rpaulo
Differential Revision: D1499
glebius [Tue, 17 Feb 2015 19:27:14 +0000 (19:27 +0000)]
Globally enable -fms-extensions when building kernel with gcc, and remove
this option from all modules that enable it theirselves.
In C mode -fms-extensions option enables anonymous structs and unions,
allowing us to use this C11 feature in kernel. Of course, clang supports
it without any extra options.
Reviewed by: dim
emaste [Tue, 17 Feb 2015 18:33:17 +0000 (18:33 +0000)]
lldb: workaround to permit cross-arch core file debugging
FreeBSD core files have no section table and thus LLDB's OS and vendor
detection logic does not work. If we encounter such an ELF file, update
an unknown OS to match the host.
This is not really the correct way to handle this, but more extensive
rework of ObjectFileELF will be needed and this change restores cross-
arch core debugging until that can be completed.
imp [Tue, 17 Feb 2015 18:10:46 +0000 (18:10 +0000)]
hd64570.h was installed, so add it here now that it is gone.
Submitted by: ae@
jmg [Tue, 17 Feb 2015 17:37:00 +0000 (17:37 +0000)]
When the new random adaptor code was brought it in r273872, a call to
randomdev_init_reader to change read_random over to the newly installed
adaptor was missed. This means both read_random and arc4random (seeded
from read_random) were not returning very random data. This also
effects userland arc4random as it is seeded from kernel arc4random.
The random devices are uneffected and have returned good randomness
since the change.
All keys generated with a kernel of r273872 must be regenerated with
a kernel with this patch. Keys generated may be predictable.
Remove the warning as log is too early to print anything, and it would
always get printed due to early use of arc4random...
Reviewed by: delphij, markm
Approved by: so (delphij)
pfg [Tue, 17 Feb 2015 16:01:00 +0000 (16:01 +0000)]
ulimit(3): simplify.
rlim_t is at least as large as long, so we don't need the
extra variable to keep the intermediate step. We don't
need the volatile either.
The code was tested on i386 and amd64.
Suggested by: bde
X-MFC with: r278803
emaste [Tue, 17 Feb 2015 15:19:58 +0000 (15:19 +0000)]
Update elftoolchain to upstream revision 3163
Most of our changes have now been committed upstream, so this change is
largely bookkeeping.
Sponsored by: The FreeBSD Foundation
brd [Tue, 17 Feb 2015 14:48:16 +0000 (14:48 +0000)]
Add tests for account and password expiration.
Approved by: will
jilles [Tue, 17 Feb 2015 13:12:54 +0000 (13:12 +0000)]
compress,gzip,xz: Preserve timestamps with nanosecond precision.
andrew [Tue, 17 Feb 2015 13:09:20 +0000 (13:09 +0000)]
Pull the ARM ddb unwind code out to a new file. This will allow it to be
used by other places that expect to unwind the stack, e.g. dtrace and
stack(9).
As I have written most of this code I'm changing the license to the
standard FreeBSD license. I have received approval from the other
developers who have changed any of the affected code.
Approved by: ian, imp, rpaulo, eadler (all license change)
hselasky [Tue, 17 Feb 2015 10:00:15 +0000 (10:00 +0000)]
Fix compilation of the SDP driver and a compile warning after r278886.
Also fix the kernel build rule for mlx4_exp.c.
This fixes the LINT kernel target for amd64.
Sponsored by: Mellanox Technologies
MFC after: 1 month
ngie [Tue, 17 Feb 2015 09:31:58 +0000 (09:31 +0000)]
Add the mnt_lockref field to the ddb(4) 'show mount' command
MFC after: 1 week
Differential Revision: https://reviews.freebsd.org/D1688
Submitted by: Conrad Meyer <conrad.meyer@isilon.com>
Sponsored by: EMC / Isilon Storage Division
hselasky [Tue, 17 Feb 2015 08:57:36 +0000 (08:57 +0000)]
Fix compilation when DEBUG is defined.
Sponsored by: Mellanox Technologies
MFC after: 1 month
kib [Tue, 17 Feb 2015 08:54:03 +0000 (08:54 +0000)]
Restore the extern qualifier on __cleanup.
Sponsored by: The FreeBSD Foundation
MFC after: 3 days
ngie [Tue, 17 Feb 2015 08:50:26 +0000 (08:50 +0000)]
Some minor style(9) fixes (whitespace + comment)
MFC after: 3 days
hselasky [Tue, 17 Feb 2015 08:40:27 +0000 (08:40 +0000)]
Update the infiniband stack to Mellanox's OFED version 2.1.
Highlights:
- Multiple verbs API updates
- Support for RoCE, RDMA over ethernet
All hardware drivers depending on the common infiniband stack has been
updated aswell.
Discussed with: np @
Sponsored by: Mellanox Technologies
MFC after: 1 month
hselasky [Tue, 17 Feb 2015 07:52:50 +0000 (07:52 +0000)]
Try to resolve infinite interrupts by clearing an undocumented
interrupt status bit. According to the UHCI controller specification
the host controller halted interrupt is non-maskable.
PR: 156596
Tested by: adrian @
MFC after: 1 week
imp [Tue, 17 Feb 2015 05:10:41 +0000 (05:10 +0000)]
Remove compat code for pre-FreeBSD 7 systems.
imp [Tue, 17 Feb 2015 05:07:38 +0000 (05:07 +0000)]
This should have been GC'd 6 years ago when ar(4) was removed.
adrian [Tue, 17 Feb 2015 02:35:06 +0000 (02:35 +0000)]
Implement taskqueue_start_threads_cpuset().
This is a more generic version of taskqueue_start_threads_pinned()
which only supports a single cpuid.
This originally came from John Baldwin <jhb@> who implemented it
as part of a push towards NUMA awareness in drivers. I started implementing
something similar for RSS and NUMA, then found he already did it.
I'd like to axe taskqueue_start_threads_pinned() so it doesn't become
part of a longer-term API. (Read: hps@ wants to MFC things, and
if I don't do this soon, he'll MFC what's here. :-)
I have a follow-up commit which converts the intel drivers over
to using the cpuset version of this function, so we can eventually
nuke the the pinned version.
Tested:
* igb, ixgbe
Obtained from: jhbbsd
nwhitehorn [Tue, 17 Feb 2015 01:23:38 +0000 (01:23 +0000)]
Having the TOC pointer in kernel dbeugger printouts is useful.
glebius [Mon, 16 Feb 2015 23:50:53 +0000 (23:50 +0000)]
- Improve INET/INET6 scope.
- style(9) declarations.
- Make couple of local functions static.
rpaulo [Mon, 16 Feb 2015 23:29:56 +0000 (23:29 +0000)]
Fix a typo when invoking tar(1) to create the debug distributions.
kib [Mon, 16 Feb 2015 22:18:43 +0000 (22:18 +0000)]
Array cannot be NULL, remove always true comparision. ACPI spec
identifies the tested condition for _PRT as "BYTE value of 0", so the
remaining part of the conditionals is sufficient.
Sponsored by: The FreeBSD Foundation
MFC after: 1 week
marius [Mon, 16 Feb 2015 22:13:03 +0000 (22:13 +0000)]
Unbreak sparc64 after r276630 by calling __sparc_sigtramp_setup signal
trampoline as part of the MD __sys_sigaction again.
Submitted by: kib (initial versions)
MFC after: 3 days
kib [Mon, 16 Feb 2015 21:56:19 +0000 (21:56 +0000)]
Initialize x2APIC mode on the resume path before accessing LAPIC.
Remove unneeded disable of LAPIC in the native_lapic_xapic_mode(). We
attempt to send wakeup IPI on the resume path right after BSP wakeup,
so disabling is wrong.
Reported and tested by: glebius, "Ranjan1018 ." <214748mv@gmail.com>
Sponsored by: The FreeBSD Foundation
MFC after: 2 months
glebius [Mon, 16 Feb 2015 21:52:28 +0000 (21:52 +0000)]
Toss declarations to fix regular build and NO_INET6 build.
pfg [Mon, 16 Feb 2015 21:35:29 +0000 (21:35 +0000)]
rexec(3): prevent uninitialized access to "port" variable.
CID: 1018716 (and clang static checker)
hselasky [Mon, 16 Feb 2015 21:26:16 +0000 (21:26 +0000)]
Define standard formatting strings to print GIDs
in a separate header file.
Sponsored by: Mellanox Technologies
MFC after: 3 days
hselasky [Mon, 16 Feb 2015 21:22:56 +0000 (21:22 +0000)]
The kasprintf() function cannot be inlined due to using a variable
number of arguments. Move it to a C-file in the linuxapi module to
make the function usable.
Sponsored by: Mellanox Technologies
MFC after: 3 days
sbruno [Mon, 16 Feb 2015 20:55:37 +0000 (20:55 +0000)]
Add example for upcoming aarch64/arm64 targets.
loos [Mon, 16 Feb 2015 19:06:31 +0000 (19:06 +0000)]
Add a mask to match only the relative base address of BSC controllers.
This should fix the attach of BSC on RPI2 (based on andrew@ dmesg).
jmg [Mon, 16 Feb 2015 18:43:52 +0000 (18:43 +0000)]
remove NULL check as M_WAITOK will not return NULL
Reviewed by: jhb
Sponsored by: FreeBSD Foundation
glebius [Mon, 16 Feb 2015 18:33:33 +0000 (18:33 +0000)]
Commit a miss from r278843.
Pointy hat to: glebius
brd [Mon, 16 Feb 2015 18:06:24 +0000 (18:06 +0000)]
Fix build.
Approved by: gibbs