SVNews

NOTE: This service is experimental and subject to change! Use at your own risk!

Year:   Month: - or - Filter committers: - or -
Filter paths:
Include diffs if smaller than lines.

2015 - 02 - 27   (today)
02:56:58 - r279349
(1 hour 2 minutes ago)
by imp
Affects:  /head/sys/kern/sched_ule.c
Create sched_rand() and move the LCG code into that. Call this when
we need randomness in ULE. This removes random() call from the
rebalance interval code.

Submitted by: Harrison Grundy
Differential Revision: https://reviews.freebsd.org/D1968
02:53:44 - r279348
(1 hour 5 minutes ago)
by jamie
Affects: 
(2 files)
/stable/10/usr.sbin/jls/jls.c
MFC r279083:

  Fix the logic for skipping parameters (with -s) that have "jailsys"
  parents (such as host.hostname); these were being skipped all the time.
  That it went this long without anyone noticing is a sign that this feature
  isn't actually used by anyone, but it's there so it might as well work.

MFC r279123:

  Allow for parameters added with the JP_OPT flag to not exist.
  That's why the flag exists in the first place.
02:50:01 - r279347
(1 hour 9 minutes ago)
by jamie
Affects: 
(3 files)
/stable/10/usr.sbin/jls/jls.8
/stable/10/usr.sbin/jls/jls.c
MFC r279081:

  Allow parameters listed on the command line to override the -v option,
  instead of crashing.

PR:             197701
02:44:12 - r279346
(1 hour 15 minutes ago)
by ken
Affects:  /head/lib/libdevstat/devstat.c
Remove an obsolete comment in devstat(3) about the accuracy of the
milliseconds per transaction (DSM_MS_PER_TRANSACTION) calculation.

The comment was accurate many years ago when the kernel didn't
record I/O times on a per-I/O basis, but now that we do collect
that information in most areas, it isn't correct.

The milliseconds per transaction values are correct, assuming the
I/O duration has been recorded.

Sponsored by:   Spectra Logic
MFC after:      3 days
02:21:52 - r279345
(1 hour 37 minutes ago)
by jchandra
Affects:  /head/sys/boot/fdt/dts/mips/xlp-basic.dts
/head/sys/mips/nlm/xlp_pci.c
Move PCI bus below simplebus for Broadcom XLP

This will enable us to do common allocation code for memory and
interrupts for SoC devices as well as PCI devices.
01:59:29 - r279344
(2 hours ago)
by pfg
Affects:  /head/lib/libc/net/ip6opt.c
Hint out check for unsigned negative values.

On FreeBSD socklen_t is unsigned so the check negative len
in inet6_opt_append() is redundant and likely to be optimized
away by the compiler.

On other operating systems this is not necessarily so, and
in the future we may want to sign it so leave the check in
but place it in a secondary position as a subtle indication
that the bogus check is intentional.

Discussed with: rpaulo

CID:    1017783
01:20:17 - r279343
(2 hours 39 minutes ago)
by glebius
Affects:  /projects/ifnet/sys/net/if.c
Use mbufq_drain() instead of pasting it.
01:15:23 - r279342
(2 hours 44 minutes ago)
by glebius
Affects:  /head/sys/net/if_var.h
Hide struct ifmultiaddr under _KERNEL, too.
00:57:09 - r279341
(3 hours 2 minutes ago)
by jchandra
Affects:  /head/sys/mips/nlm/interrupt.h
/head/sys/mips/nlm/intr_machdep.c
/head/sys/mips/nlm/xlp_pci.c
Improve additional interrupt ACK for Broadcom XLP

Handling some interrupts in XLP (like PCIe and SATA) involves writing to
vendor specific registers as part of interrupt acknowledgement.

This was earlier done with xlp_establish_intr(), but a better solution
is to provide a function xlp_set_bus_ack() that can be used with
cpu_establish_hardintr(). This will allow platform initialization code to
setup these ACKs without changing the standrard drivers.
00:54:53 - r279340
(3 hours 4 minutes ago)
by glebius
Affects: 
(9 files)
/projects/ifnet/sys/dev/*
/projects/ifnet/sys/net/*
/projects/ifnet/sys/netinet/in.c
Historically all interfaces have their if_init() method, which initializes
final resources and brings it up. Usually, the driver code calls its
xxx_init() method internally, when being brought IFF_UP via SIOCSIFFLAGS.
However, for IPv4 there historically existed another quite ugly code path:

sys_ioctl(SIOCAIFADDR) -> .. -> in_control(SIOCAIFADDR) ->
-> if_ioctl(SIOCSIFADDR) -> ether_ioctl(SIOCSIFADDR) -> if_init().

Note that a legitimate SIOCAIFADDR is temporarily substituted with
obsoleted SIOCSIFADDR. Moreover, the pointer provided with command
is not ifreq pointer, but ifaddr. Finally, the IFF_UP was set by
ether_ioctl() blindly, without executing if_up().

Unentangle that:
- All xxx_init()s are no longer a driver methods, but a static functions.
  The only legitimate way of upping an interface is SIOCSIFFLAGS.
- in_control() calls SIOCSIFFLAGS if it sees interface not IFF_UP.
- arp_ifinit() moves to in_control(), out of Ethernet code.
- ether_ioctl() no longer contains #ifdef INET hack.

Sponsored by:   Netflix
Sponsored by:   Nginx, Inc.
00:26:17 - r279339
(3 hours 33 minutes ago)
by glebius
Affects:  /projects/ifnet/sys/net/if.c
Inline if_route() and if_unroute().
2015 - 02 - 26   (yesterday)
23:05:46 - r279338
(4 hours 53 minutes ago)
by ian
Affects:  /head/sys/arm/include/atomic.h
Add casting to make atomic ops work for pointers.  (Apparently nobody has
ever done atomic ops on pointers before now on arm).

Submitted by:   Svatopluk Kraus <onwahe@gmail.com>
22:46:01 - r279337
(5 hours 13 minutes ago)
by dim
Affects:  /projects/clang360-import/lib/clang/include/llvm/Config/config.h
/projects/clang360-import/lib/clang/include/llvm/Config/llvm-config.h
Minor updates to llvm's pregenerated config headers.
22:22:06 - r279336
(5 hours 37 minutes ago)
by ken
Affects:  /head/sys/dev/mpr/mpr_sas.c
/head/sys/dev/mps/mps_sas.c
Add FreeBSD stable/10 version checks for the availability of the
CDAI_FLAG_NONE advanced information CCB flag.

Support for the flag was merged to stable/10 in r279329, and the
__FreeBSD_version in stable/10 was bumped to 1001510.

Check for that version in the mps(4) and mpr(4) drivers when determining
whether to use the flag.

Sponsored by:   Spectra Logic
MFC after:      3 days
21:30:40 - r279335
(6 hours 29 minutes ago)
by dchagin
Affects: 
(11 files)
/user/dchagin/lemul/sys/compat/linprocfs/linprocfs.c
/user/dchagin/lemul/sys/fs/*
/user/dchagin/lemul/sys/modules/procfs/Makefile
Add preliminary support of /proc/[pid]/fd. On Linux this is a
subdirectory containing one entry for each file which the process
has open, named by its file descriptor, and which is a symbolic
link to the actual file.
For now only regular file descriptors are supported.
21:15:02 - r279334
(6 hours 44 minutes ago)
by dchagin
Affects:  /user/dchagin/lemul/sys/kern/kern_descrip.c
/user/dchagin/lemul/sys/sys/filedesc.h
Export fdhold() & fddrop() needed by an upcoming pseudofs change.
21:13:33 - r279333
(6 hours 46 minutes ago)
by dchagin
Affects:  /user/dchagin/lemul/sys/fs/pseudofs/pseudofs.c
/user/dchagin/lemul/sys/modules/pseudofs/Makefile
Hide vfs.pfs.trace variable if it is not used.
21:12:22 - r279332
(6 hours 47 minutes ago)
by dchagin
Affects: 
(13 files)
/user/dchagin/lemul/sys/amd64/linux/*
/user/dchagin/lemul/sys/amd64/linux32/*
/user/dchagin/lemul/sys/i386/linux/*
Regen for r279331 (syncfs).
21:10:03 - r279331
(6 hours 49 minutes ago)
by dchagin
Affects: 
(7 files)
/user/dchagin/lemul/sys/amd64/*
/user/dchagin/lemul/sys/compat/linux/linux_stats.c
/user/dchagin/lemul/sys/i386/linux/*
Implement Linux specific syncfs() system call.
20:59:18 - r279330
(7 hours ago)
by nwhitehorn
Affects:  /head/bin/sh/expand.c
Fix unitialized variable that broke sh on PowerPC starting with r278826.
20:46:16 - r279329
(7 hours 13 minutes ago)
by ken
Affects: 
(8 files)
/stable/10/sbin/camcontrol/camcontrol.c
/stable/10/sys/*
MFC r278964:

The __FreeBSD_version was changed to 1001510 to be appropriate for
stable/10.

I will followup with a commit to mpr(4) and mps(4) in head to reflect
the stable/10 __FreeBSD_version and merge the change back to stable/10.

  ------------------------------------------------------------------------
  r278964 | ken | 2015-02-18 11:30:19 -0700 (Wed, 18 Feb 2015) | 46 lines

  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
20:02:29 - r279328
(7 hours 57 minutes ago)
by emaste
Affects:  /head/Makefile.inc1
Support CROSS_BINUTILS_PREFIX with in-tree compiler

Reviewed by:    bapt, imp
Sponsored by:   The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D1974
19:56:48 - r279327
(8 hours 2 minutes ago)
by rpaulo
Affects: 
(2 files)
/stable/10/etc/network.subr
MFC r278933:
  Fix a typo in ipv6_down().

  Submitted by: Ashutosh Kumar AK0037447 at TechMahindra.com
16:39:57 - r279326
(11 hours 19 minutes ago)
by kib
Affects:  /head/lib/libstdthreads/mtx.c
Use pthread_mutex_trylock(3) to implement mtx_trylock(3).

Noted and tested by:    Vineela <vineela_17@yahoo.com>
PR:     198050
Sponsored by:   The FreeBSD Foundation
MFC after:      1 week
16:05:09 - r279325
(11 hours 54 minutes ago)
by royger
Affects:  /head/sys/x86/xen/xen_intr.c
xen/intr: fix fallout from r278854

r278854 introduced a race in the event channel handling code. We must make
sure that the pending bit is cleared before executing the filter, or else we
might miss other events that would be injected after the filter has ran but
before the pending bit is cleared.

While there also mask event channels while FreeBSD executes the ithread
bound to that event channel. This refrains Xen from injecting more
interrupts while the ithread has not finished it's work.

Sponsored by: Citrix Systems R&D
Reported by: sbruno, robak
Tested by: robak
15:59:45 - r279324
(11 hours 59 minutes ago)
by ae
Affects:  /head/sbin/geom/class/part/geom_part.c
When gpart(8) is trying automatically determine the first available
block of free space after existing partition, take into account
provider's stripeoffset, since the result will be adjusted to this
value.

PR:             197989
MFC after:      1 week
15:48:20 - r279323
(12 hours 11 minutes ago)
by trasz
Affects:  /head/usr.sbin/uefisign/Makefile
Use LIBADD.

MFC after:      1 month
Sponsored by:   The FreeBSD Foundation
14:48:28 - r279322
(13 hours 11 minutes ago)
by pho
Affects:  /user/pho/stress2/misc/core5.sh
Added a regression test.

Sponsored by:    EMC / Isilon storage division
14:22:27 - r279321
(13 hours 37 minutes ago)
by trasz
Affects:  /head/share/examples/uefisign/uefikeys
Make the uefikeys script output slightly more obvious.

MFC after:      1 month
Sponsored by:   The FreeBSD Foundation
12:51:05 - r279320
(15 hours 8 minutes ago)
by mav
Affects:  /head/sys/dev/ahci/ahci.c
For some uniformity move ahci_ch_init() call under the lock.

Submitted by:   Dmitry Luhtionov <dmitryluhtionov@gmail.com>
MFC after:      2 weeks
11:02:40 - r279319
(16 hours 57 minutes ago)
by kib
Affects:  /head/sys/x86/include/apicvar.h
/head/sys/x86/x86/io_apic.c
/head/sys/x86/x86/local_apic.c
Implements EOI suppression mode, where LAPIC on EOI command for
level-triggered interrupt does not broadcast the EOI message to all
APICs in the system.  Instead, interrupt handler must follow LAPIC EOI
with IOAPIC EOI.  For modern IOAPICs, the later is done by writing to
EOIR register.  Otherwise, Intel provided Linux with a trick of
temporary switching the pin config to edge and then back to level.

Detect presence of EOIR register by reading IO-APIC version.  The
summary table in the comments was taken from the Linux kernel.  For
Intel, newer IO-APICs are only briefly documented as part of the
ICH/PCH datasheet.  According to the BKDG and chipset documentation,
AMD LAPICs do not provide EOI suppression, althought IO-APICs do
declare version 0x21 and implement EOIR.

The trick to temporary switch pin to edge mode to clear IRR was tested
on modern chipset, by pretending that EOIR is not present, i.e. by
forcing io_haseoi to zero.

Tunable hw.lapic_eoi_suppression disables the optimization.

Reviewed by:    neel
Tested by:      pho
Review: https://reviews.freebsd.org/D1943
Sponsored by:   The FreeBSD Foundation
MFC after:      2 months
09:42:03 - r279318
(18 hours 17 minutes ago)
by kib
Affects:  /head/lib/libstdthreads/thrd.c
Check that the pointer to the thread return value is not NULL before
dereferencing. NULL is allowed by C11 and must be handled.

Reported and tested by: Vineela <vineela_17@yahoo.com>
PR:     198038
Sponsored by:   The FreeBSD Foundation
MFC after:      1 week
09:31:25 - r279317
(18 hours 28 minutes ago)
by trasz
Affects: 
(4 files)
/head/etc/mtree/BSD.usr.dist
/head/share/examples/Makefile
/head/share/examples/uefisign/uefikeys
Add key/cert generation script for uefisign(8).

(Forgot about Relnotes in the commit that added uefisign(8), so set it here.)

MFC after:      1 month
Relnotes:       yes
Sponsored by:   The FreeBSD Foundation
09:16:36 - r279316
(18 hours 43 minutes ago)
by trasz
Affects:  /head/share/man/man8/uefi.8
Add uefisign(8) reference to uefi(8) manual page.

MFC after:      1 month
Sponsored by:   The FreeBSD Foundation
09:15:24 - r279315
(18 hours 44 minutes ago)
by trasz
Affects: 
(9 files)
/head/usr.sbin/Makefile
/head/usr.sbin/uefisign/*
Add uefisign(8), UEFI Secure Boot signing utility.

MFC after:      1 month
Sponsored by:   The FreeBSD Foundation
09:08:48 - r279314
(18 hours 50 minutes ago)
by trasz
Affects:  /head/usr.sbin/ctld/parse.y
Add missing error check.

MFC after:      1 month
Sponsored by:   The FreeBSD Foundation
07:51:43 - r279313
(20 hours 8 minutes ago)
by dim
Affects: 
(9 files)
/projects/clang360-import/contrib/libcxxrt/unwind-arm.h
/projects/clang360-import/sys/arm/arm/disassem.c
/projects/clang360-import/sys/arm/ti/*
Merge ^/head r279309 through r279312.
07:47:35 - r279312
(20 hours 12 minutes ago)
by dim
Affects: 
(4 files)
/head/sys/arm/arm/disassem.c
/head/sys/arm/ti/*
Fix a number of -Wcast-qual warnings under sys/arm.  No functional
change.

Submitted by:   andrew
MFC after:      3 days
07:45:40 - r279311
(20 hours 14 minutes ago)
by dim
Affects:  /head/sys/arm/ti/am335x/am335x_rtc.c
In sys/arm/ti/am335x/am335x_rtc.c, fix a clang 3.6.0 warning about
am33x_rtc_softc::sc_irq_res (which is an array) never being NULL.

Submitted by:   andrew
MFC after:      3 days
07:42:16 - r279310
(20 hours 17 minutes ago)
by dim
Affects:  /head/contrib/libcxxrt/unwind-arm.h
Since newer versions of compiler-rt require unwind.h, and we want to use
the copy in libcxxrt for it, fix the arm-specific header to define the
_Unwind_Action type.

Submitted by:   andrew
MFC after:      3 days
07:26:56 - r279309
(20 hours 32 minutes ago)
by dim
Affects: 
(181 files)
/projects/clang360-import/*
Merge ^/head r279163 through r279308.
07:23:52 - r279308
(20 hours 35 minutes ago)
by jhibbits
Affects:  /projects/pmac_pmu/sys/powerpc/powermac/pmu.c
Use the power management subsystem, overloading APM, for suspending, instead of
a sysctl.

This unifies things with the x86 behavior.
07:20:05 - r279307
(20 hours 39 minutes ago)
by dim
Affects:  /head/contrib/libcxxrt/dwarf_eh.h
Make libcxxrt's parsing of DWARF exception handling tables work on
architectures with strict alignment, by using memcpy() instead of
directly reading fields.

Reported by:    Daisuke Aoyama <aoyama@peach.ne.jp>
Reviewed by:    imp, bapt
Tested by:      bapt
MFC after:      3 days
Differential Revision: https://reviews.freebsd.org/D1967
02:22:47 - r279306
(1 day 1 hour ago)
by jchandra
Affects: 
(5 files)
/head/sys/boot/fdt/dts/mips/xlp-basic.dts
/head/sys/mips/nlm/*
Remove run-time allocation of XLP IRQs

Follow the same static IRQ to Interrupt Table Entry mapping as the other
OS supported on XLP.
02:05:45 - r279305
(1 day 1 hour ago)
by jchandra
Affects:  /head/sys/boot/fdt/dts/mips/xlp-basic.dts
/head/sys/mips/nlm/intr_machdep.c
Add netlogic,xlp-pic as interrupt controller for XLP

Add an empty driver for netlogic,xlp-pic to ensure that the device tree
is correct and has an interrupt controller.
01:53:24 - r279304
(1 day 2 hours ago)
by jchandra
Affects:  /head/sys/mips/nlm/hal/pic.h
/head/sys/mips/nlm/interrupt.h
/head/sys/mips/nlm/xlp.h
Fix up interrupt definitions for Broadcom XLP

Gather all the IRQ definitions to interrupt.h. Earlier these were in xlp.h
and pic.h.  Update the definition of XLP_IRQ_IS_PICINTR to check for last
irq as well.
2015 - 02 - 25   (2 days ago)
22:41:27 - r279303
(1 day 5 hours ago)
by emaste
Affects: 
(7 files)
/stable/9/contrib/llvm/tools/clang/include/clang/*
/stable/9/contrib/llvm/tools/clang/lib/Driver/*
Merge upstream Clang revision 211785:
  This commit implements the -fuse-ld= option, so that the user
  can specify -fuse-ld=bfd to use ld.bfd.

  This commit re-applies r194328 with some test case changes.
  It seems that r194328 was breaking macosx or mingw build
  because clang can't find ld.bfd or ld.gold in the given sysroot.
  We should use -B to specify the executable search path instead.

  Patch originally by David Chisnall.

This is a merge from stable/10 rather than MFC as this is change was
already included in Clang 3.5 in HEAD.

MFS-10:         r279302
Sponsored by:   The FreeBSD Foundation
22:32:32 - r279302
(1 day 5 hours ago)
by emaste
Affects: 
(6 files)
/stable/10/contrib/llvm/tools/clang/include/clang/*
/stable/10/contrib/llvm/tools/clang/lib/Driver/*
Implement the -fuse-ld= option.

Merge upstream Clang revision 211785:
  This commit implements the -fuse-ld= option, so that the user
  can specify -fuse-ld=bfd to use ld.bfd.

  This commit re-applies r194328 with some test case changes.
  It seems that r194328 was breaking macosx or mingw build
  because clang can't find ld.bfd or ld.gold in the given sysroot.
  We should use -B to specify the executable search path instead.

  Patch originally by David Chisnall.

This is a direct commit to stable/10 as this is change is already
included in Clang 3.5 in HEAD. The patch is also reworked slightly for
Clang 3.4.1.

Reviewed by:    dim
Sponsored by:   The FreeBSD Foundation
22:12:37 - r279301
(1 day 5 hours ago)
by emaste
Affects:  /head/contrib/binutils/bfd/dwarf2.c
GNU nm: Avoid NULL dereference

bfd_dwarf2_find_line() calls find_line() with NULL functionname_ptr,
which resulted in a crash on certain ELF objects.

This change was implemented independently from upstream binutils, but
I have checked that the crash does not happen there.

MFC after:      1 week
Sponsored by:   The FreeBSD Foundation
21:59:03 - r279300
(1 day 6 hours ago)
by adrian
Affects:  /head/sys/kern/subr_taskqueue.c
/head/sys/sys/taskqueue.h
Remove taskqueue_start_threads_pinned(); there's noa generic cpuset version of
this.

Sponsored by:   Norse Corp, Inc.
21:44:53 - r279299
(1 day 6 hours ago)
by adrian
Affects: 
(4 files)
/head/sys/dev/e1000/if_igb.c
/head/sys/dev/ixgbe/ixgbe.c
/head/sys/dev/ixl/*
Migrate using CPU_ZERO() + CPU_SET() -> CPU_SETOF().

Tested:

* ixgbe, igb, RSS enabled

Submitted by:   jhb
Sponsored by:   Norse Corp, Inc.
21:43:09 - r279298
(1 day 6 hours ago)
by emaste
Affects:  /head/contrib/elftoolchain/nm/nm.c
nm: avoid crash in print_lineno if func->name is NULL

This can occur when DW_AT_specification is used to refer to another DIE
that provides the actual DW_AT_name string. For example:

< 3><0x00000086> DW_TAG_subprogram
                   DW_AT_name              PrettyStackTraceEntry
...
< 1><0x00002cf4> DW_TAG_subprogram
                   DW_AT_specification     <0x00000086>

We will need to add support for DW_AT_specification, but in the interim
we should not segfault.

Obtained from:  Elftoolchain (r3170)
Sponsored by:   The FreeBSD Foundation
21:10:03 - r279297
(1 day 6 hours ago)
by hselasky
Affects: 
(4 files)
/head/usr.bin/unifdef/*
Update to upstream version 2.10

The most notable new feature is support for definition files.

Obtained from:  http://dotat.at/prog/unifdef
MFC after:      1 week
20:47:25 - r279296
(1 day 7 hours ago)
by delphij
Affects:  /head/usr.bin/enigma/enigma.c
Explicitly crypt_set_format("des") and bail out if we
can't.  This would prevent problem when we changed the
default crypt(3) algorithm or removed it in the future.

PR:             bin/189958
MFC after:      1 month
18:50:24 - r279295
(1 day 9 hours ago)
by dim
Affects: 
(6 files)
/projects/clang360-import/contrib/llvm/*
/projects/clang360-import/lib/clang/include/clang/Basic/Version.inc
Merge llvm 3.6.0 final from ^/vendor/llvm/dist, merge clang 3.6.0 final
from ^/vendor/clang/dist, and resolve conflicts.
18:27:53 - r279294
(1 day 9 hours ago)
by dim
Affects:  /vendor/clang/clang-release_360-r230434
Tag clang tags/RELEASE_360/final r230434 (effectively, 3.6.0 release).
18:26:53 - r279293
(1 day 9 hours ago)
by dim
Affects:  /vendor/clang/dist/docs/ReleaseNotes.rst
/vendor/clang/dist/lib/Basic/Version.cpp
Vendor import of clang RELEASE_360/final tag r230434 (effectively, 3.6.0
release):
https://llvm.org/svn/llvm-project/cfe/tags/RELEASE_360/final@230434
18:26:13 - r279292
(1 day 9 hours ago)
by dim
Affects:  /vendor/llvm/llvm-release_360-r230434
Tag llvm tags/RELEASE_360/final r230434 (effectively, 3.6.0 release).
18:25:34 - r279291
(1 day 9 hours ago)
by dim
Affects: 
(8 files)
/vendor/llvm/dist/*
Vendor import of llvm RELEASE_360/final tag r230434 (effectively, 3.6.0
release):
https://llvm.org/svn/llvm-project/llvm/tags/RELEASE_360/final@230434
17:54:18 - r279290
(1 day 10 hours ago)
by dim
Affects:  /stable/10/contrib/llvm/patches/patch-r279289-clang-r199571-fix-string-literal-assertion.diff
/stable/9/contrib/llvm/patches/patch-r279289-clang-r199571-fix-string-literal-assertion.diff
Add clang patches corresponding to r279289.
17:27:02 - r279289
(1 day 10 hours ago)
by dim
Affects:  /stable/10/contrib/llvm/tools/clang/lib/AST/Expr.cpp
/stable/9/contrib/llvm/tools/clang/lib/AST/Expr.cpp
Pull in r199571 from upstream clang trunk (by Ted Kremenek):

  Harden InitListExpr::isStringLiteralInit() against getInit()
  returning null.

  This led to a crash on invalid code (sorry, no good test case).

  Fixes <rdar://problem/15831804>.

This fixes an assertion when compiling certain incorrect code, as
reported upstream in http://llvm.org/PR22684 .

Direct commit to stable/10 and stable/9, since head has clang 3.5.1,
which already includes this change.

Reported by:    hbowden@securelabsllc.com
17:06:27 - r279288
(1 day 10 hours ago)
by marcel
Affects:  /head/usr.bin/netstat/if.c
Fix "netstat -hdw 1" output.

Reported by: np@
16:44:40 - r279287
(1 day 11 hours ago)
by gjb
Affects:  /stable/10/sys/sys/param.h
/stable/8/sys/sys/param.h
/stable/9/sys/sys/param.h
Bump __FreeBSD_version after FreeBSD-EN-15:01.vt,
FreeBSD-EN-15:02.openssl, FreeBSD-EN-15:03.freebsd-update,
FreeBSD-SA-15:04.igmp, FreeBSD-SA-15:05.bind

Sponsored by:   The FreeBSD Foundation
16:44:07 - r279286
(1 day 11 hours ago)
by kib
Affects:  /head/sys/x86/acpica/madt.c
For now, disable x2APIC mode when Xen is detected, even if CPU
declares support for it.  Newer versions of Xen works fine with x2APIC
code, but e.g. Xen 4.2 delivers GPF on the LAPIC MSR write, despite
x2APIC mode being known to hypervisor.

Discussed with: royger
Sponsored by:   The FreeBSD Foundation
16:36:44 - r279285
(1 day 11 hours ago)
by gjb
Affects: 
(6 files)
/stable/10/release/doc/share/xml/*
/stable/8/release/doc/share/xml/*
/stable/9/release/doc/share/xml/*
Document FreeBSD-EN-15:01.vt, FreeBSD-EN-15:02.openssl,
FreeBSD-EN-15:03.freebsd-update, FreeBSD-SA-15:04.igmp,
FreeBSD-SA-15:05.bind

Sponsored by:   The FreeBSD Foundation
16:18:26 - r279284
(1 day 11 hours ago)
by kib
Affects:  /head/lib/libthr/thread/thr_mutex.c
Propagate errors from _thr_umutex_unlock2 through mutex_unlock_common.

Errors from _thr_umutex_unlock2 should "never happen" in normal
circumstances.  If they do, however, return them to the application
so it can fail early and loudly.  Hiding the errors will only delay
the inevitable failure, making it harder to find and diagnose.

Submitted by:   Eric van Gyzen <eric_van_gyzen@dell.com>
Obtained from:  Dell Inc.
PR:     198914
MFC after:      1 week
16:17:16 - r279283
(1 day 11 hours ago)
by kib
Affects:  /head/sys/kern/kern_umtx.c
When failing to claim ownership of a umtx_pi, restore the umutex owner
to its previous, unowned state.  This avoids compounding an existing
problem of inconsistent ownership.

Submitted by:   Eric van Gyzen <eric_van_gyzen@dell.com>
Obtained from:  Dell Inc.
PR:     198914
MFC after:      1 week
16:12:56 - r279282
(1 day 11 hours ago)
by kib
Affects:  /head/sys/kern/kern_umtx.c
When unlocking a contested PI pthread mutex, if the queue of waiters
is empty, look up the umtx_pi and disown it if the current thread owns it.
This can happen if a signal or timeout removed the last waiter from
the queue, but there is still a thread in do_lock_pi() holding a reference
on the umtx_pi.  The unlocking thread might not own the umtx_pi in this case,
but if it does, it must disown it to keep the ownership consistent between
the umtx_pi and the umutex.

Submitted by:   Eric van Gyzen <eric_van_gyzen@dell.com>
        with advice from: Elliott Rabe and Jim Muchow, also at Dell Inc.
Obtained from:  Dell Inc.
PR:     198914
13:58:43 - r279281
(1 day 14 hours ago)
by hselasky
Affects:  /head/sys/netinet/ip_output.c
Fix a special case in ip_fragment() to produce a more sensible chain
of packets. When the data payload length excluding any headers, of an
outgoing IPv4 packet exceeds PAGE_SIZE bytes, a special case in
ip_fragment() can kick in to optimise the outgoing payload(s). The
code which was added in r98849 as part of zero copy socket support
assumes that the beginning of any MTU sized payload is aligned to
where a MBUF's "m_data" pointer points. This is not always the case
and can sometimes cause large IPv4 packets, as part of ping replies,
to be split more than needed.

Instead of iterating the MBUFs to figure out how much data is in the
current chain, use the value already in the "m_pkthdr.len" field of
the first MBUF in the chain.

Reviewed by:            ken @
Differential Revision:  https://reviews.freebsd.org/D1893
MFC after:              2 weeks
Sponsored by:           Mellanox Technologies
12:26:45 - r279280
(1 day 15 hours ago)
by hselasky
Affects: 
(3 files)
/stable/9/sys/dev/usb/controller/musb_otg.c
MFC r278850:
Handle VBUS error interrupts.

PR:             190471
12:24:24 - r279279
(1 day 15 hours ago)
by hselasky
Affects: 
(2 files)
/stable/10/sys/dev/usb/controller/musb_otg.c
MFC r278850:
Handle VBUS error interrupts.

PR:             190471
10:18:11 - r279278
(1 day 17 hours ago)
by mav
Affects: 
(5 files)
/head/sys/geom/raid/*
Replace constant with proper sizeof().

Submitted by:   Dmitry Luhtionov <dmitryluhtionov@gmail.com>
MFC after:      2 weeks
10:10:40 - r279277
(1 day 17 hours ago)
by mav
Affects:  /head/usr.sbin/ctld/ctld.c
Fix memory leak on incorrect initiator portal.

Submitted by:   Dmitry Luhtionov <dmitryluhtionov@gmail.com>
MFC after:      3 days
10:07:39 - r279276
(1 day 17 hours ago)
by mav
Affects:  /head/usr.sbin/ctld/kernel.c
Add checks for malloc() failures.

Submitted by:   Dmitry Luhtionov <dmitryluhtionov@gmail.com>
MFC after:      3 days
09:57:05 - r279275
(1 day 18 hours ago)
by mav
Affects:  /head/sys/boot/fdt/fdt_loader_cmd.c
Fix potential NULL dereference.

Submitted by:   Dmitry Luhtionov <dmitryluhtionov@gmail.com>
MFC after:      2 weeks
09:21:59 - r279274
(1 day 18 hours ago)
by kib
Affects: 
(2 files)
/stable/9/sys/kern/sysv_shm.c
MFC r278963:
If malloc() sleeps, Giant is dropped.  Recheck for another thread
doing our work.

Remove unneeded check for failed M_WAITOK allocation.
09:21:04 - r279273
(1 day 18 hours ago)
by mav
Affects: 
(5 files)
/stable/10/sys/cam/ctl/*
/stable/10/sys/cam/scsi/scsi_all.h
MFC r278584: Add support for General Statistics and Performance log page.

CTL already collects most of statistics reported there, so why not.
09:19:26 - r279272
(1 day 18 hours ago)
by kib
Affects: 
(2 files)
/stable/10/sys/kern/sysv_shm.c
MFC r278963:
If malloc() sleeps, Giant is dropped.  Recheck for another thread
doing our work.

Remove unneeded check for failed M_WAITOK allocation.
08:39:48 - r279271
(1 day 19 hours ago)
by tijl
Affects: 
(2 files)
/stable/10/contrib/binutils/ld/scripttempl/elf.sc
MFC r278586:

Fix ldscripts such that ld(1) collects the .fini_array section in the same
order as the .init_array section.  Finalisation routines need to be called
in the opposite order as their corresponding initialisation routines but
rtld(1) handles that by calling the function pointers in .fini_array in
reverse order.

Reviewed by:    kib
08:35:00 - r279270
(1 day 19 hours ago)
by hselasky
Affects:  /head/sys/dev/usb/usb_hub.c
Add ugen keyword to USB pnpinfo. Remove extra space from existing pnpinfo.

PR:             198015
PR:             198019
PR:             198026
MFC after:      1 week
07:49:15 - r279269
(1 day 20 hours ago)
by cperciva
Affects:  /head/lib/libc/stdlib/random.3
Adjust wording slightly to emphasize that random(3) should not be used for
any applications which need unpredictable random numbers, not merely those
which are cryptographic in nature.

If you work for a lottery and you're using random(3) to select the winning
numbers, please let me know.
06:20:42 - r279268
(1 day 21 hours ago)
by arybchik
Affects:  /head/sys/conf/files
sfxge: add to config files

Sponsored by:   Solarflare Communications, Inc.
Approved by:    gnn (mentor)
06:19:23 - r279267
(1 day 21 hours ago)
by arybchik
Affects:  /head/sys/dev/sfxge/sfxge_port.c
sfxge: use tab to indent instead of spaces

Sponsored by:   Solarflare Communications, Inc.
Approved by:    gnn (mentor)
06:19:00 - r279266
(1 day 21 hours ago)
by arybchik
Affects:  /head/sys/dev/sfxge/common/efx_impl.h
sfxge: correct limit for number of Rx queues

Sponsored by:   Solarflare Communications, Inc.
Approved by:    gnn (mentor)
05:56:54 - r279265
(1 day 22 hours ago)
by delphij
Affects: 
(875 files)
/releng/8.4/*
/releng/9.3/*
Fix integer overflow in IGMP protocol. [SA-15:04]

Fix BIND remote denial of service vulnerability. [SA-15:05]

Fix vt(4) crash with improper ioctl parameters. [EN-15:01]

Updated base system OpenSSL to 0.9.8zd. [EN-15:02]

Fix freebsd-update libraries update ordering issue. [EN-15:03]

Approved by:    so
05:56:16 - r279264
(1 day 22 hours ago)
by delphij
Affects: 
(1143 files)
/releng/10.0/*
/releng/10.1/*
Fix integer overflow in IGMP protocol. [SA-15:04]

Fix vt(4) crash with improper ioctl parameters. [EN-15:01]

Updated base system OpenSSL to 1.0.1l. [EN-15:02]

Fix freebsd-update libraries update ordering issue. [EN-15:03]

Approved by:    so
05:43:02 - r279263
(1 day 22 hours ago)
by delphij
Affects:  /stable/10/sys/netinet/igmp.c
/stable/8/sys/netinet/igmp.c
/stable/9/sys/netinet/igmp.c
Instant MFC:

Fix integer overflow in IGMP protocol.

Security:       FreeBSD-SA-15:04.igmp
Security:       CVE-2015-1414
Found by:       Mateusz Kocielski, Logicaltrust
Analyzed by:    Marek Kroemeke, Mateusz Kocielski (shm@NetBSD.org) and
                22733db72ab3ed94b5f8a1ffcde850251fe6f466
Submited by:    Mariusz Zaborski <oshogbo@FreeBSD.org>
Reviewed by:    bms
Approved by:    so
05:42:59 - r279262
(1 day 22 hours ago)
by delphij
Affects:  /head/sys/netinet/igmp.c
Fix integer overflow in IGMP protocol.

Security:       FreeBSD-SA-15:04.igmp
Security:       CVE-2015-1414
Found by:       Mateusz Kocielski, Logicaltrust
Analyzed by:    Marek Kroemeke, Mateusz Kocielski (shm@NetBSD.org) and
                22733db72ab3ed94b5f8a1ffcde850251fe6f466
Submited by:    Mariusz Zaborski <oshogbo@FreeBSD.org>
Reviewed by:    bms
04:30:23 - r279261
(1 day 23 hours ago)
by ken
Affects:  /head/lib/libmt/mtlib.c
/head/usr.bin/mt/mt.c
Fix several problems found by Coverity.

lib/libmt/mtlib.c:
        In mt_start_element(), make sure we don't overflow the
        cur_sb array.  CID 1271325

usr.bin/mt/mt.c:
        In main(), bzero the mt_com structure so that we aren't
        using any uninitialized stack variables.  CID 1271319

        In mt_param(), only allow one -s and one -p argument.  This
        will prevent a memory leak caused by overwriting the
        param_name and/or param_value variables.  CID 1271320 and
        CID 1271322

        To make things simpler in mt_param(), make sure there
        there is only one exit path for the function.  Make sure
        the arguments are explicitly freed.

Sponsored by:   Spectra Logic
Pointed out by: emaste
MFC after:      1 month
00:06:25 - r279260
(2 days 3 hours ago)
by glebius
Affects: 
(99 files)
/projects/sendfile/*
Merge head r261119 through r279258.
2015 - 02 - 24   (3 days ago)
23:50:10 - r279259
(2 days 4 hours ago)
by nwhitehorn
Affects: 
(911 files)
/projects/powernv/*
IFC
22:28:44 - r279258
(2 days 5 hours ago)
by marius
Affects: 
(6 files)
/stable/10/lib/libc/sparc64/sys/*
/stable/10/lib/libc/sys/Makefile.inc
/stable/10/sys/sparc64/include/asm.h
MFC: r278870

Unbreak sparc64 after r276630 (MFCed to stable/10 in r277317) by calling
__sparc_sigtramp_setup signal trampoline as part of the MD __sys_sigaction
again.

Submitted by:   kib (initial versions)
22:27:02 - r279257
(2 days 5 hours ago)
by gjb
Affects:  /head/release/Makefile.vm
/head/release/tools/gce-package.sh
Add a CLOUDPACKAGE variable to contain a list of targets for
post-install packaging cloud provider images.

Add a 'gce-package.sh' script to generate the final output
image ready for upload to the GCE platform.  Right now, this
is the only image that has a specific output format (GNU-tar),
and this implementation is expected to be temporary.

This is not directly connected to the other release targets.

MFC after:      1 week
X-MFC-with:     r279249, r279250
Sponsored by:   The FreeBSD Foundation
22:22:06 - r279256
(2 days 5 hours ago)
by brd
Affects:  /head/usr.sbin/pw/tests/pw_useradd.sh
Fix a race condition by allowing up to 5 seconds of difference between the time
stamps.

PR:             197861
Approved by:    will
22:17:12 - r279255
(2 days 5 hours ago)
by adrian
Affects: 
(4 files)
/head/sys/dev/e1000/if_igb.c
/head/sys/dev/ixgbe/ixgbe.c
/head/sys/dev/ixl/*
Change uses of taskqueue_start_threads_pinned() ->
taskqueue_start_threads_cpuset()

Differential Revision:  https://reviews.freebsd.org/D1897
Reviewed by:    jfv
22:11:07 - r279254
(2 days 5 hours ago)
by emaste
Affects:  /stable/10/sys/boot/amd64/boot1.efi/boot1.c
MFC part of r273865: fix boot1.efi for block size != 512

r273865 is part of the work for supporting 4Kn drives, but it turns out
the underlying bug can actually cause corruption of the UEFI system
table in any case where block size is not 512.

Relevant portion of the original commit message:

  convert boot1.efi to corrrectly calculate the lba for what the
  media reports and convert the size based on what FreeBSD uses.
  existing code would use the 512 byte lba and convert the
  using 4K byte size.

PR:             197881
Reviewed by:    Chris Ruffin
22:07:42 - r279253
(2 days 5 hours ago)
by slm
Affects: 
(23 files)
/head/sys/dev/mps/*
- Updated all files with 2015 Avago copyright, and updated LSI's copyright
  dates.

- Changed all of the PCI device strings from LSI to Avago Technologies (LSI).

- Added a sysctl variable to control how StartStopUnit behavior works. User can
  select to spin down disks based on if disk is SSD or HDD.

- Inquiry data is required to tell if a disk will support SSU at shutdown or
  not. Due to the addition of mpssas_async, which gets Advanced Info but not
  Inquiry data, the setting of supports_SSU was moved to the
  mpssas_scsiio_complete function, which snoops for any Inquiry commands. And,
  since disks are shutdown as a target and not a LUN, this process was
  simplified by basing it on targets and not LUNs.

- Added a sysctl variable that sets the amount of time to retry after sending a
  failed SATA ID command. This helps with some bad disks and large disks that
  require a lot of time to spin up. Part of this change was to add a callout to
  handle timeouts with the SATA ID command. The callout function is called
  mpssas_ata_id_timeout(). (Fixes PR 191348)

- Changed the way resets work by allowing I/O to continue to devices that are
  not currently under a reset condition. This uses devq's instead of simq's and
  makes use of the MPSSAS_TARGET_INRESET flag. This change also adds a function
  called mpssas_prepare_tm().

- Some changes were made to reduce code duplication when getting a SAS address
  for a SATA disk.

- Fixed some formatting and whitespace.

- Bump version of mps driver to 20.00.00.00-fbsd

PR:             191348
Reviewed by:    ken, scottl
Approved by:    ken, scottl
MFC after:      2 weeks
21:37:20 - r279252
(2 days 6 hours ago)
by nwhitehorn
Affects: 
(7 files)
/head/sys/powerpc/*
New pmap implementation for 64-bit PowerPC processors. The main focus of
this change is to improve concurrency:
- Drop global state stored in the shadow overflow page table (and all other
  global state)
- Remove all global locks
- Use per-PTE lock bits to allow parallel page insertion
- Reconstruct state when requested for evicted PTEs instead of buffering
  it during overflow

This drops total wall time for make buildworld on a 32-thread POWER8 system
by a factor of two and system time by a factor of three, providing performance
20% better than similarly clocked Core i7 Xeons per-core. Performance on
smaller SMP systems, where PMAP lock contention was not as much of an issue,
is nearly unchanged.

Tested on:      POWER8, POWER5+, G5 UP, G5 SMP (64-bit and 32-bit kernels)
Merged from:    user/nwhitehorn/ppc64-pmap-rework
Looked over by: jhibbits, andreast
MFC after:      3 months
Relnotes:       yes
Sponsored by:   FreeBSD Foundation
21:31:13 - r279251
(2 days 6 hours ago)
by np
Affects:  /head/sys/dev/cxgbe/t4_netmap.c
cxgbe(4): allow tx hardware checksumming on the netmap interface.

It is disabled by default but users can set IFCAP_TXCSUM on the
netmap ifnet (ifconfig ncxl0 txcsum) to override netmap and force
the hardware to calculate and insert proper IP and L4 checksums in
outbound frames.

MFC after:      2 weeks
21:16:57 - r279250
(2 days 6 hours ago)
by gjb
Affects:  /head/release/Makefile.vm
Remove a stale comment.

MFC after:      1 week
X-MFC-with:     r279249
Sponsored by:   The FreeBSD Foundation
21:06:25 - r279249
(2 days 6 hours ago)
by gjb
Affects:  /head/release/Makefile
/head/release/Makefile.vm
Add 'cloudware-install' target to handle copying cloud
disk images to a directory within DESTDIR.

Add CLOUDINSTALL variable to automatically generate the
actual list of install targets.  While here, rename the
cloud-specific names from 'vm-' to 'cw-' to avoid naming
collisions.

Add AZURE_DISK, GCE_DISK, and OPENSTACK_DISK output file
variables.

Add 'cloudware-install' to the 'install' target.

MFC after:      1 week
Sponsored by:   The FreeBSD Foundation
19:27:51 - r279248
(2 days 8 hours ago)
by emaste
Affects:  /head/etc/Makefile
Unconditionally install debug directory hierarchy

This avoids various failure modes (e.g., when building and installing a
single binary with debug data on a system that otherwise does not have
it enabled).

It is also consistent with the way other directory hierarchies are
handled (e.g. share/man).

Reviewed by:    imp
Differential Revision:  https://reviews.freebsd.org/D1962
19:03:31 - r279247
(2 days 8 hours ago)
by sjg
Affects:  /head/share/mk/src.sys.mk
If sys.mk were found via the magic path .../share/mk
replace it with the absolute path of .PARSEDIR, so that sub-makes
launched from objdirs (eg kernel) can still find the correct mk files.

Reviewed by:    obrien
18:40:10 - r279246
(2 days 9 hours ago)
by np
Affects:  /head/sys/dev/cxgbe/adapter.h
/head/sys/dev/cxgbe/t4_netmap.c
/head/sys/dev/cxgbe/t4_sge.c
cxgbe(4): set up congestion management for netmap rx queues.

The hw.cxgbe.cong_drop knob controls the response of the chip when
netmap queues are congested.
18:32:17 - r279245
(2 days 9 hours ago)
by np
Affects:  /head/sys/dev/cxgbe/t4_netmap.c
cxgbe(4): do not set the netmap rxq interrupts on a hair-trigger.

MFC after:      2 weeks
18:22:24 - r279244
(2 days 9 hours ago)
by np
Affects:  /head/sys/dev/cxgbe/t4_netmap.c
cxgbe(4): wait for the hardware to catch up before destroying a netmap txq.

MFC after:      2 weeks
18:19:25 - r279243
(2 days 9 hours ago)
by np
Affects:  /head/sys/dev/cxgbe/t4_netmap.c
cxgbe(4): request an automatic tx update when a netmap txq idles.

MFC after:      2 weeks
17:36:44 - r279242
(2 days 10 hours ago)
by hselasky
Affects:  /head/sys/sys/queue.h
Pass macro arguments properly.

MFC after:      1 week
17:29:04 - r279241
(2 days 10 hours ago)
by hselasky
Affects:  /head/sys/sys/queue.h
Fix order of arguments in the TRACEBUF_INITIALIZER macro so that we
can define QUEUE_MACRO_DEBUG to debug list problems.

MFC after:      1 week
16:45:16 - r279240
(2 days 11 hours ago)
by pfg
Affects: 
(4 files)
/head/lib/msun/src/*
msun: drop unnecessary checks (part 2).

Drop an unnecessary check in some calculations. The check
would have Coverity falsely conclude that a variable could
be left undefined.
15:35:26 - r279239
(2 days 12 hours ago)
by br
Affects:  /head/sys/dev/uart/uart_dev_pl011.c
Enable 'receive timeout' interrupt allowing us to not
loose 'rx buffer full' event.

This fixes operation on ARMv8 Foundation Model.

Sponsored by:   DARPA, AFRL
15:07:09 - r279238
(2 days 12 hours ago)
by nwhitehorn
Affects: 
(2 files)
/user/nwhitehorn/ppc64-pmap-rework/pseries/phyp_vscsi.c
IFC
13:07:31 - r279237
(2 days 14 hours ago)
by kib
Affects:  /head/sys/kern/kern_sig.c
Keep a reference on the coredump vnode for vn_fullpath() call.  Do it
by moving vn_close() after the point where notification is sent.

Reported by:    sbruno
Tested by:      pho, sbruno
Sponsored by:   The FreeBSD Foundation
12:57:03 - r279236
(2 days 15 hours ago)
by zbb
Affects:  /head/sys/netinet/ip.h
Change struct attribute to avoid aligned operations mismatch

Previous __alignment(4) allowed compiler to assume that operations are
performed on aligned region. On ARM processor, this led to alignment fault
as shown below:
trapframe: 0xda9e5b10
FSR=00000001, FAR=a67b680e, spsr=60000113
r0 =00000000, r1 =00000068, r2 =0000007c, r3 =00000000
r4 =a67b6826, r5 =a67b680e, r6 =00000014, r7 =00000068
r8 =00000068, r9 =da9e5bd0, r10=00000011, r11=da9e5c10
r12=da9e5be0, ssp=da9e5b60, slr=a054f164, pc =a054f2cc
<...>
udp_input+0x264: ldmia r5, {r0-r3, r6}
udp_input+0x268: stmia r12, {r0-r3, r6}

This was due to instructions which do not support unaligned access,
whereas for __alignment(2) compiler replaced ldmia/stmia with some
logically equivalent memcpy operations.
In fact, the assumption that 'struct ip' is always 4-byte aligned
is definitely false, as we have no impact on data alignment of packet
stream received.

Another possible solution would be to explicitely perform memcpy()
on objects of 'struct ip' type, which, however, would suffer from
performance drop, and be merely a problem hiding.

Please, note that this has nothing to do with
ARM32_DISABLE_ALIGNMENT_FAULTS option, but is related strictly to
compiler behaviour.

Submitted by:  Wojciech Macek <wma@semihalf.com>
Reviewed by:   glebius, ian
Obtained from: Semihalf
12:31:08 - r279235
(2 days 15 hours ago)
by zbb
Affects:  /head/sys/arm/arm/gic.c
Fix endianness on FDT read in ARM GIC

Submitted by:  Jakub Palider <jpa@semihalf.com>
Reviewed by:   ian, nwhitehorn
Obtained from: Semihalf
10:35:07 - r279234
(2 days 17 hours ago)
by ae
Affects:  /head/sys/netipsec/ipsec.h
/head/sys/netipsec/key.c
Fix possible memory leak and several races in the IPsec policy management
code.

Resurrect the state field in the struct secpolicy, it has
IPSEC_SPSTATE_ALIVE value when security policy linked in the chain,
and IPSEC_SPSTATE_DEAD value in all other cases. This field protects
from trying to unlink one security policy several times from the different
threads.

Take additional reference in the key_flush_spd() to be sure that policy
won't be freed from the different thread while we are sending SPDEXPIRE message.

Add KEY_FREESP() call to the key_unlink() to release additional reference
that we take when use key_getsp*() functions.

Differential Revision:  https://reviews.freebsd.org/D1914
Tested by:              Emeric POUPON <emeric.poupon at stormshield dot eu>
Reviewed by:    hrs
Sponsored by:   Yandex LLC
08:53:47 - r279233
(2 days 19 hours ago)
by hselasky
Affects:  /head/sys/dev/usb/controller/xhci.c
Ensure that the XHCI driver will refresh the control endpoint settings
when re-enumerating a FULL speed device. Else the wrong max packet
setting might be used when trying to re-enumerate a FULL speed device.

MFC after:      3 days
06:20:50 - r279232
(2 days 21 hours ago)
by luigi
Affects:  /head/sys/dev/ixl/if_ixl.c
/head/sys/dev/ixl/ixl_txrx.c
/head/sys/dev/netmap/if_ixl_netmap.h
Add native netmap support to ixl.
Preliminary tests indicate 32 Mpps on tx, 24 Mpps on rx
with source and receiver on two different ports of the same 40G card.
Optimizations are likely possible.
The code follows closely the one for ixgbe so i do not
expect stability issues.

Hardware kindly supplied by Intel.

Reviewed by:    Jack Vogel
MFC after:      1 week
06:09:31 - r279231
(2 days 21 hours ago)
by arybchik
Affects:  /head/sys/dev/sfxge/sfxge_tx.c
/head/sys/dev/sfxge/sfxge_tx.h
sfxge: add put-list high watermark

It is interesting to know how long put-list grows.

Sponsored by:   Solarflare Communications, Inc.
Approved by:    gnn (mentor)
06:08:55 - r279230
(2 days 21 hours ago)
by arybchik
Affects:  /head/sys/dev/sfxge/sfxge_dma.c
sfxge: use goto to cleanup to avoid duplicate cleanup code

Sponsored by:   Solarflare Communications, Inc.
Approved by:    gnn (mentor)
05:43:16 - r279229
(2 days 22 hours ago)
by ken
Affects:  /head/sys/cam/scsi/scsi_sa.c
Fix printf format warnings on sparc64 and mips.

Sponsored by:   Spectra Logic
MFC after:      1 month
05:35:15 - r279228
(2 days 22 hours ago)
by neel
Affects: 
(4 files)
/head/sys/amd64/vmm/intel/*
Always emulate MSR_PAT on Intel processors and don't rely on PAT save/restore
capability of VT-x. This lets bhyve run nested in older VMware versions that
don't support the PAT save/restore capability.

Note that the actual value programmed by the guest in MSR_PAT is irrelevant
because bhyve sets the 'Ignore PAT' bit in the nested PTE.

Reported by:    marcel
Tested by:      Leon Dang (ldang@nahannisys.com)
Sponsored by:   Nahanni Systems
MFC after:      2 weeks
05:15:40 - r279227
(2 days 22 hours ago)
by neel
Affects:  /head/usr.sbin/bhyve/xmsr.c
Emulate MSR 0xC0011024 when running on AMD processors.

OpenBSD guests test bit 0 of this MSR to detect whether the workaround for
erratum 721 has been applied.

Reported by:    Jason Tubnor (jason@tubnor.net)
MFC after:      1 week
04:05:32 - r279226
(2 days 23 hours ago)
by gnn
Affects:  /svnadmin/conf/mentors
Free Andrew Rybchenko from mentorship.
02:04:16 - r279225
(3 days 1 hour ago)
by neel
Affects: 
(4 files)
/head/usr.sbin/bhyve/*
Add "-u" option to bhyve(8) to indicate that the RTC should maintain UTC time.

The default remains localtime for compatibility with the original device model
in bhyve(8). This is required for OpenBSD guests which assume that the RTC
keeps UTC time.

Reviewed by:    grehan
Pointed out by: Jason Tubnor (jason@tubnor.net)
MFC after:      2 weeks
01:46:43 - r279224
(3 days 2 hours ago)
by kib
Affects: 
(3 files)
/stable/10/sys/kern/vfs_mount.c
/stable/10/usr.sbin/mountd/mountd.c
MFC r278523:
In mountd, silence a race with the parallel unmount.
01:00:46 - r279223
(3 days 2 hours ago)
by yongari
Affects:  /head/sys/dev/bce/if_bcereg.h
Correct a typo.

Reported by:    jmg
2015 - 02 - 23   (4 days ago)
22:58:51 - r279222
(3 days 5 hours ago)
by grembo
Affects:  /head/sys/boot/i386/libi386/biosmem.c
Improve memory detection in biosmem.c

- Add a quirk to allow ignoring e820 extended memory detection.
- Improve memory detection through e801.
- Add bootloader command "biosmem".

See differential revision for more details.

Reviewed by:    jhb, adrian
Approved by:    adrian
Differential Revision:  https://reviews.freebsd.org/D1741
22:50:43 - r279221
(3 days 5 hours ago)
by ken
Affects:  /head/sys/sys/param.h
Bump __FreeBSD_version for the mtio(4) / sa(4) ioctl and API additions
in revision 279219.

Sponsored by:   Spectra Logic
MFC after:      1 month
22:31:39 - r279220
(3 days 5 hours ago)
by grehan
Affects:  /head/usr.sbin/bhyve/pci_ahci.c
Don't close a block context if it couldn't be opened,
for example if the backing file doesn't exist,
avoiding a null deref.

Reviewed by:    neel
MFC after:      1 week.
21:59:30 - r279219
(3 days 6 hours ago)
by ken
Affects: 
(19 files)
/head/*
Significant upgrades to sa(4) and mt(1).

The primary focus of these changes is to modernize FreeBSD's
tape infrastructure so that we can take advantage of some of the
features of modern tape drives and allow support for LTFS.

Significant changes and new features include:

 o sa(4) driver status and parameter information is now exported via an
   XML structure.  This will allow for changes and improvements later
   on that will not break userland applications.  The old MTIOCGET
   status ioctl remains, so applications using the existing interface
   will not break.

 o 'mt status' now reports drive-reported tape position information
   as well as the previously available calculated tape position
   information.  These numbers will be different at times, because
   the drive-reported block numbers are relative to BOP (Beginning
   of Partition), but the block numbers calculated previously via
   sa(4) (and still provided) are relative to the last filemark.
   Both numbers are now provided.  'mt status' now also shows the
   drive INQUIRY information, serial number and any position flags
   (BOP, EOT, etc.) provided with the tape position information.
   'mt status -v' adds information on the maximum possible I/O size,
   and the underlying values used to calculate it.

 o The extra sa(4) /dev entries (/dev/saN.[0-3]) have been removed.

   The extra devices were originally added as place holders for
   density-specific device nodes.  Some OSes (NetBSD, NetApp's OnTap
   and Solaris) have had device nodes that, when you write to them,
   will automatically select a given density for particular tape drives.

   This is a convenient way of switching densities, but it was never
   implemented in FreeBSD.  Only the device nodes were there, and that
   sometimes confused users.

   For modern tape devices, the density is generally not selectable
   (e.g. with LTO) or defaults to the highest availble density when
   the tape is rewritten from BOT (e.g. TS11X0).  So, for most users,
   density selection won't be necessary.  If they do need to select
   the density, it is easy enough to use 'mt density' to change it.

 o Protection information is now supported.  This is either a
   Reed-Solomon CRC or CRC32 that is included at the end of each block
   read and written.  On write, the tape drive verifies the CRC, and
   on read, the tape drive provides a CRC for the userland application
   to verify.

 o New, extensible tape driver parameter get/set interface.

 o Density reporting information.  For drives that support it,
   'mt getdensity' will show detailed information on what formats the
   tape drive supports, and what formats the tape drive supports.

 o Some mt(1) functionality moved into a new mt(3) library so that
   external applications can reuse the code.

 o The new mt(3) library includes helper routines to aid in parsing
   the XML output of the sa(4) driver, and build a tree of driver
   metadata.

 o Support for the MTLOAD (load a tape in the drive) and MTWEOFI
   (write filemark immediate) ioctls needed by IBM's LTFS
   implementation.

 o Improve device departure behavior for the sa(4) driver.  The previous
   implementation led to hangs when the device was open.

 o This has been tested on the following types of drives:
        IBM TS1150
        IBM TS1140
        IBM LTO-6
        IBM LTO-5
        HP LTO-2
        Seagate DDS-4
        Quantum DLT-4000
        Exabyte 8505
        Sony DDS-2

contrib/groff/tmac/doc-syms,
share/mk/bsd.libnames.mk,
lib/Makefile,
        Add libmt.

lib/libmt/Makefile,
lib/libmt/mt.3,
lib/libmt/mtlib.c,
lib/libmt/mtlib.h,
        New mt(3) library that contains functions moved from mt(1) and
        new functions needed to interact with the updated sa(4) driver.

        This includes XML parser helper functions that application writers
        can use when writing code to query tape parameters.

rescue/rescue/Makefile:
        Add -lmt to CRUNCH_LIBS.

src/share/man/man4/mtio.4
        Clarify this man page a bit, and since it contains what is
        essentially the mtio.h header file, add new ioctls and structure
        definitions from mtio.h.

src/share/man/man4/sa.4
        Update BUGS and maintainer section.

sys/cam/scsi/scsi_all.c,
sys/cam/scsi/scsi_all.h:
        Add SCSI SECURITY PROTOCOL IN/OUT CDB definitions and CDB building
        functions.

sys/cam/scsi/scsi_sa.c
sys/cam/scsi/scsi_sa.h
        Many tape driver changes, largely outlined above.

        Increase the sa(4) driver read/write timeout from 4 to 32
        minutes.  This is based on the recommended values for IBM LTO
        5/6 drives.  This may also avoid timeouts for other tape
        hardware that can take a long time to do retries and error
        recovery.  Longer term, a better way to handle this is to ask
        the drive for recommended timeout values using the REPORT
        SUPPORTED OPCODES command.  Modern IBM and Oracle tape drives
        at least support that command, and it would allow for more
        accurate timeout values.

        Add XML status generation.  This is done with a series of
        macros to eliminate as much duplicate code as possible.  The
        new XML-based status values are reported through the new
        MTIOCEXTGET ioctl.

        Add XML driver parameter reporting, using the new MTIOCPARAMGET
        ioctl.

        Add a new driver parameter setting interface, using the new
        MTIOCPARAMSET and MTIOCSETLIST ioctls.

        Add a new MTIOCRBLIM ioctl to get block limits information.

        Add CCB/CDB building routines scsi_locate_16, scsi_locate_10,
        and scsi_read_position_10().

        scsi_locate_10 implements the LOCATE command, as does the
        existing scsi_set_position() command.  It just supports
        additional arguments and features.  If/when we figure out a
        good way to provide backward compatibility for older
        applications using the old function API, we can just revamp
        scsi_set_position().  The same goes for
        scsi_read_position_10() and the existing scsi_read_position()
        function.

        Revamp sasetpos() to take the new mtlocate structure as an
        argument.  It now will use either scsi_locate_10() or
        scsi_locate_16(), depending upon the arguments the user
        supplies.  As before, once we change position we don't have a
        clear idea of what the current logical position of the tape
        drive is.

        For tape drives that support long form position data, we
        read the current position and store that for later reporting
        after changing the position.  This should help applications
        like Bacula speed tape access under FreeBSD once they are
        modified to support the new ioctls.

        Add a new quirk, SA_QUIRK_NO_LONG_POS, that is set for all
        drives that report SCSI-2 or older, as well as drives that
        report an Illegal Request type error for READ POSITION with
        the long format.  So we should automatically detect drives
        that don't support the long form and stop asking for it after
        an initial try.

        Add a partition number to the sa(4) softc.

        Improve device departure handling. The previous implementation
        led to hangs when the device was open.

        If an application had the sa(4) driver open, and attempted to
        close it after it went away, the cam_periph_release() call in
        saclose() would cause the periph to get destroyed because that
        was the last reference to it.  Because destroy_dev() was
        called from the sa(4) driver's cleanup routine (sacleanup()),
        and would block waiting for the close to happen, a deadlock
        would result.

        So instead of calling destroy_dev() from the cleanup routine,
        call destroy_dev_sched_cb() from saoninvalidate() and wait for
        the callback.

        Acquire a reference for devfs in saregister(), and release it
        in the new sadevgonecb() routine when all devfs devices for
        the particular sa(4) driver instance are gone.

        Add a new function, sasetupdev(), to centralize setting
        per-instance devfs device parameters instead of repeating the
        code in saregister().

        Add an open count to the softc, so we know how many
        peripheral driver references are a result of open
        sessions.

        Add the D_TRACKCLOSE flag to the cdevsw flags so
        that we get a 1:1 mapping of open to close calls
        instead of a N:1 mapping.

        This should be a no-op for everything except the
        control device, since we don't allow more than one
        open on non-control devices.

        However, since we do allow multiple opens on the
        control device, the combination of the open count
        and the D_TRACKCLOSE flag should result in an
        accurate peripheral driver reference count, and an
        accurate open count.

        The accurate open count allows us to release all
        peripheral driver references that are the result
        of open contexts once we get the callback from devfs.

sys/sys/mtio.h:
        Add a number of new mt(4) ioctls and the requisite data
        structures.  None of the existing interfaces been removed
        or changed.

        This includes definitions for the following new ioctls:

        MTIOCRBLIM      /* get block limits */
        MTIOCEXTLOCATE  /* seek to position */
        MTIOCEXTGET     /* get tape status */
        MTIOCPARAMGET   /* get tape params */
        MTIOCPARAMSET   /* set tape params */
        MTIOCSETLIST    /* set N params */

usr.bin/mt/Makefile:
        mt(1) now depends on libmt, libsbuf and libbsdxml.

usr.bin/mt/mt.1:
        Document new mt(1) features and subcommands.

usr.bin/mt/mt.c:
        Implement support for mt(1) subcommands that need to
        use getopt(3) for their arguments.

        Implement a new 'mt status' command to replace the old
        'mt status' command.  The old status command has been
        renamed 'ostatus'.

        The new status function uses the MTIOCEXTGET ioctl, and
        therefore parses the XML data to determine drive status.
        The -x argument to 'mt status' allows the user to dump out
        the raw XML reported by the kernel.

        The new status display is mostly the same as the old status
        display, except that it doesn't print the redundant density
        mode information, and it does print the current partition
        number and position flags.

        Add a new command, 'mt locate', that will supersede the
        old 'mt setspos' and 'mt sethpos' commands.  'mt locate'
        implements all of the functionality of the MTIOCEXTLOCATE
        ioctl, and allows the user to change the logical position
        of the tape drive in a number of ways.  (Partition,
        block number, file number, set mark number, end of data.)
        The immediate bit and the explicit address bits are
        implemented, but not documented in the man page.

        Add a new 'mt weofi' command to use the new MTWEOFI ioctl.
        This allows the user to ask the drive to write a filemark
        without waiting around for the operation to complete.

        Add a new 'mt getdensity' command that gets the XML-based
        tape drive density report from the sa(4) driver and displays
        it.  This uses the SCSI REPORT DENSITY SUPPORT command
        to get comprehensive information from the tape drive about
        what formats it is able to read and write.

        Add a new 'mt protect' command that allows getting and setting
        tape drive protection information.  The protection information
        is a CRC tacked on to the end of every read/write from and to
        the tape drive.

Sponsored by:   Spectra Logic
MFC after:      1 month
21:16:02 - r279218
(3 days 6 hours ago)
by jhb
Affects: 
(2 files)
/stable/10/libexec/rtld-elf/rtld.c
MFC 275412:
The runtime linker needs to include a path to itself in the link map
it exports to the debugger.  It currently has two choices: it can use
a compiled-in path (/libexec/ld-elf.so.1) or it can use the path stored
in the interpreter path in the binary being executed.  The runtime linker
currently prefers the second.  However, this is usually wrong for compat32
binaries since the binary specifies the path of rtld on a 32-bit system
(/libexec/ld-elf.so.1) instead of the actual path (/libexec/ld-elf32.so.1).
For now, always assume the compiled in path (/libexec/ld-elf32.so.1) as
the rtld path and ignore the path in the binary for the 32-bit runtime
linker.
20:38:00 - r279217
(3 days 7 hours ago)
by nwhitehorn
Affects:  /head/sys/powerpc/pseries/phyp_vscsi.c
Fix race in interrupt handling that could cause IO to hang up under heavy
load.
20:36:07 - r279216
(3 days 7 hours ago)
by glebius
Affects: 
(8704 files)
/projects/sendfile/*
Merge head r258543 through r279213.
20:09:05 - r279215
(3 days 7 hours ago)
by ian
Affects:  /head/sys/arm/arm/cpufunc_asm_armv7.S
There is no reason to do i+dcache writeback and invalidate when changing
the translation table (this may be left over from armv5 days).  It's
especially bad to do so using a cache operation that isn't coherent on
SMP systems.

Submitted by:   Michal Meloun
19:36:31 - r279214
(3 days 8 hours ago)
by dchagin
Affects:  /user/dchagin/lemul/sys/compat/linux/linux_misc.c
Properly check tv_nsec value. The tv_nsec field can also be one
of the special value UTIME_NOW or UTIME_OMIT.
18:57:09 - r279213
(3 days 9 hours ago)
by glebius
Affects:  /head/sys/sys/mbuf.h
A lot of current code in network stack expects mbufs not having
m_nextpkt pointer, assuming that if there is one, then this is
a packet batch. Thus, mbufq_dequeue() needs to clear it on mbuf
that is being returned.
18:55:26 - r279212
(3 days 9 hours ago)
by glebius
Affects:  /head/sys/dev/xen/netfront/netfront.c
Previous version of mbufq were fine initialized by M_ZERO, while
new one require explicti initialization.

Reported by:    royger
18:38:41 - r279211
(3 days 9 hours ago)
by jhb
Affects: 
(16 files)
/stable/10/contrib/binutils/*
/stable/10/sys/*
/stable/10/usr.bin/gcore/elfcore.c
MFC 274817,274878,276801,276840,278976:
Improve support for XSAVE with debuggers.
- Dump an NT_X86_XSTATE note if XSAVE is in use. This note is designed
  to match what Linux does in that 1) it dumps the entire XSAVE area
  including the fxsave state, and 2) it stashes a copy of the current
  xsave mask in the unused padding between the fxsave state and the
  xstate header at the same location used by Linux.
- Teach readelf() to recognize NT_X86_XSTATE notes.
- Change PT_GET/SETXSTATE to take the entire XSAVE state instead of
  only the extra portion. This avoids having to always make two
  ptrace() calls to get or set the full XSAVE state.
- Add a PT_GET_XSTATE_INFO which returns the length of the current
  XSTATE save area (so the size of the buffer needed for PT_GETXSTATE)
  and the current XSAVE mask (%xcr0).
17:01:38 - r279210
(3 days 10 hours ago)
by hselasky
Affects:  /head/sys/dev/usb/controller/dwc_otg.c
/head/sys/dev/usb/controller/dwc_otgreg.h
Add support for the DWC OTG v2 chipset found in the STM32F4 series of
processors. Make sure we pullup the data lines in device mode when we
power on the port.

MFC after:      1 week
15:24:43 - r279209
(3 days 12 hours ago)
by ae
Affects:  /head/sys/kern/uipc_socket.c
soreceive_generic() still has similar KASSERT(), therefore instead of
remove KASSERT(), change it to check mbuf isn't NULL.

Suggested by:   kib
MFC after:      1 week
15:05:39 - r279208
(3 days 12 hours ago)
by dchagin
Affects:  /user/dchagin/lemul/sys/compat/linux/linux_socket.c
Since FreeBSD supports SOCK_CLOEXEC & SOCK_NONBLOCK options
remove its emulation via fcntl call from Linuxulator.

XXX. Commit mostly for test
14:42:14 - r279207
(3 days 13 hours ago)
by dchagin
Affects:  /user/dchagin/lemul/sys/compat/linux/linux_socket.c
Fix recvmsg() after r279203.
13:41:35 - r279206
(3 days 14 hours ago)
by ae
Affects:  /head/sys/kern/uipc_socket.c
In some cases soreceive_dgram() can return no data, but has control
message. This can happen when application is sending packets too big
for the path MTU and recvmsg() will return zero (indicating no data)
but there will be a cmsghdr with cmsg_type set to IPV6_PATHMTU.
Remove KASSERT() which does NULL pointer dereference in such case.
Also call m_freem() only when m isn't NULL.

PR:             197882
MFC after:      1 week
Sponsored by:   Yandex LLC
12:54:46 - r279205
(3 days 15 hours ago)
by hselasky
Affects:  /head/sys/ofed/include/linux/kernel.h
/head/sys/ofed/include/linux/log2.h
Macro fixes:
- Add missing "order_base_2()" macro.
- Fix BUILD_BUG_ON() macro.

MFC after:      1 month
Sponsored by:   Mellanox Technologies
09:40:59 - r279204
(3 days 18 hours ago)
by dchagin
Affects: 
(13 files)
/user/dchagin/lemul/sys/amd64/linux/*
/user/dchagin/lemul/sys/amd64/linux32/*
/user/dchagin/lemul/sys/i386/linux/*
Regen for r279203 (recvmmsg, sendmmsg).
09:38:28 - r279203
(3 days 18 hours ago)
by dchagin
Affects: 
(8 files)
/user/dchagin/lemul/sys/amd64/*
/user/dchagin/lemul/sys/compat/linux/*
/user/dchagin/lemul/sys/i386/linux/*
Implement recvmmsg() and sendmmsg() system calls.
09:37:10 - r279202
(3 days 18 hours ago)
by dchagin
Affects: 
(4 files)
/user/dchagin/lemul/sys/amd64/*
/user/dchagin/lemul/sys/compat/linux/linux_socket.h
/user/dchagin/lemul/sys/i386/linux/linux.h
Reduce duplication between MD Linux code by moving msg related
struct definitions out into the compat/linux/linux_socket.h
08:45:42 - r279201
(3 days 19 hours ago)
by pluknet
Affects: 
(2 files)
/stable/10/usr.bin/kdump/kdump.c
MFC r278857:
   kdump: sendfile(2) "flags" argument needs casting on 64-bit platforms.

Sponsored by:   Nginx, Inc.
07:35:27 - r279200
(3 days 20 hours ago)
by ngie
Affects: 
(175 files)
/projects/building-blocks/*
MFhead @ r279199
07:28:31 - r279199
(3 days 20 hours ago)
by luigi
Affects:  /head/sys/dev/netmap/netmap_freebsd.c
add MODULE_VERSION, needed to track module dependencies

MFC after:      3 days
06:46:04 - r279198
(3 days 21 hours ago)
by ngie
Affects:  /head/Makefile.inc1
Fill in missing dependencies for dtrace related tools so the bootstrap-tools
compiles properly on older hosts

Pointyhat to: me
X-MFC with: r279197
MFC after: 2 weeks
05:28:17 - r279197
(3 days 22 hours ago)
by ngie
Affects:  /head/Makefile.inc1
Parallelize building bootstrap-tools

Differential Revision: https://reviews.freebsd.org/D1901
MFC after: 2 weeks
Reviewed by: ian
No serious objections from: imp
01:24:10 - r279196
(4 days 2 hours ago)
by markj
Affects: 
(2 files)
/stable/10/sys/kern/kern_ctf.c
MFC r278983:
Free the zlib stream once the CTF section is decompressed.

MFC r278984:
Remove unnecessary checks for a NULL return value from malloc.

MFC r279089:
Let vn_rdwr() check for short reads.
01:17:45 - r279195
(4 days 2 hours ago)
by kib
Affects: 
(2 files)
/stable/10/sys/dev/acpica/acpi_pcib.c
MFC r278871:
Array cannot be NULL, remove always true comparision.
2015 - 02 - 22   (5 days ago)
22:53:51 - r279194
(4 days 5 hours ago)
by dim
Affects: 
(77 files)
/projects/clang360-import/contrib/compiler-rt/lib/*
Update compiler-rt to trunk r230183.  This has some of our patches
imported, so we have just a few small diffs against upstream left.
22:44:30 - r279193
(4 days 5 hours ago)
by dim
Affects:  /vendor/compiler-rt/compiler-rt-r230183
Tag compiler-rt trunk r230183.
22:43:40 - r279192
(4 days 5 hours ago)
by dim
Affects: 
(127 files)
/vendor/compiler-rt/dist/*
Import compiler-rt trunk r230183.

https://llvm.org/svn/llvm-project/compiler-rt/trunk@230183
22:31:36 - r279191
(4 days 5 hours ago)
by dim
Affects: 
(150 files)
/projects/clang360-import/contrib/compiler-rt/*
/projects/clang360-import/lib/libclang_rt/*
/projects/clang360-import/lib/libcompiler_rt/Makefile
Update compiler-rt to trunk r228651.  This enables using Address
Sanitizer and Undefined Behavior Sanitizer with clang 3.6.0.
22:00:53 - r279190
(4 days 5 hours ago)
by nwhitehorn
Affects: 
(14 files)
/user/nwhitehorn/ppc64-pmap-rework/*
IFC after VSX kernel code
21:40:27 - r279189
(4 days 6 hours ago)
by nwhitehorn
Affects: 
(13 files)
/head/sys/powerpc/*
Kernel support for the Vector-Scalar eXtension (VSX) found on the POWER7
and POWER8. This instruction set unifies the 32 64-bit scalar floating
point registers with the 32 128-bit vector registers into a single bank
of 64 128-bit registers. Kernel support mostly amounts to saving and
restoring the wider version of the floating point registers and making
sure that both scalar FP and vector registers are enabled once a VSX
instruction is executed. get_mcontext() and friends currently cannot
see the high bits, which will require a little more work.

As the system compiler (GCC 4.2) does not support VSX, making use of this
from userland requires either newer GCC or clang.

Relnotes:       yes
Sponsored by:   FreeBSD Foundation
21:32:57 - r279188
(4 days 6 hours ago)
by jilles
Affects:  /head/bin/sh/sh.1
sh: Add details about importing the environment and initializing OPTIND.
20:52:29 - r279187
(4 days 7 hours ago)
by nwhitehorn
Affects:  /head/contrib/binutils/opcodes/ppc-opc.c
Add some opcodes for assembling forthcoming VSX (Vector-Scalar eXtension)
support in the kernel. Userspace programs are expected to rely on LLVM's
integrated assembler or newer binutils.
20:16:44 - r279186
(4 days 7 hours ago)
by pfg
Affects:  /head/lib/libc/gen/setmode.3
/head/lib/libc/gen/setmode.c
setmode(3): Make sure that setmode sets errno on failure.

Our man page already documented this partially but now
we have some consistent behavior.

PR:             136669
Obtained from:  NetBSD (CVS rev. 1.31, 1.33)
Relnotes:       yes
MFC after:      3 weeks
19:42:50 - r279185
(4 days 8 hours ago)
by maxim
Affects:  /head/share/man/man5/periodic.conf.5
o Typo: securiy -> security.

PR:             197927
Submitted by:   mike.543@comcast.net
MFC after:      1 week
19:25:57 - r279184
(4 days 8 hours ago)
by arybchik
Affects: 
(5 files)
/head/sys/dev/sfxge/*
sfxge: implement if_get_counter callback

Sponsored by:   Solarflare Communications, Inc.
Approved by:    gnn (mentor)
19:24:08 - r279183
(4 days 8 hours ago)
by arybchik
Affects:  /head/sys/dev/sfxge/common/efx.h
/head/sys/dev/sfxge/common/efx_regs.h
/head/sys/dev/sfxge/common/efx_tx.c
sfxge: add common code support for changing TX queue pace

To delay packets from a particular TX queue by a particular time, write a value
into the TX Pace table s.t. pace time <= TX Pace Clock Period * (2 ^ pace value)
- the TX pace clock is 1/13 of the system clock, so its period should be 104 or
52 ns depending on whether turbo mode is active.

EFX_TX_PACE_CLOCK_BASE added by me.

Submitted by:   Mark Spender <mspender at solarflare.com>
Sponsored by:   Solarflare Communications, Inc.
Approved by:    gnn (mentor)
19:13:52 - r279182
(4 days 8 hours ago)
by arybchik
Affects: 
(5 files)
/head/sys/dev/sfxge/common/*
/head/sys/dev/sfxge/sfxge_ev.c
sfxge: correct event queue interrupt moderation timer quanta

Submitted by:   Andrew Lee <alee at solarflare.com>
Sponsored by:   Solarflare Communications, Inc.
Approved by:    gnn (mentor)
18:59:00 - r279181
(4 days 9 hours ago)
by arybchik
Affects:  /head/share/man/man4/sfxge.4
sfxge: add indefinite article and update timestamp

Sponsored by:   Solarflare Communications, Inc.
Approved by:    gnn (mentor)
18:57:28 - r279180
(4 days 9 hours ago)
by arybchik
Affects:  /head/sys/modules/sfxge/Makefile
sfxge: list header in SRCS

Sponsored by:   Solarflare Communications, Inc.
Approved by:    gnn (mentor)
18:56:03 - r279179
(4 days 9 hours ago)
by arybchik
Affects:  /head/sys/dev/sfxge/sfxge_rx.c
/head/sys/dev/sfxge/sfxge_tx.c
sfxge: DMA allocated memory is set to zeros because of BUS_DMA_ZERO flag

It is not required to set it to zeros once again.

Sponsored by:   Solarflare Communications, Inc.
Approved by:    gnn (mentor)
18:54:16 - r279178
(4 days 9 hours ago)
by arybchik
Affects:  /head/sys/dev/sfxge/common/efsys.h
sfxge: do no allow EFSYS_MEM_ALLOC sleep

It solves locking problem when EFSYS_MEM_ALLOC is called in
the context holding a mutex (not allowed to sleep).
E.g. on interface bring up or multicast addresses addition.

Sponsored by:   Solarflare Communications, Inc.
Approved by:    gnn (mentor)
18:52:15 - r279177
(4 days 9 hours ago)
by arybchik
Affects:  /head/sys/dev/sfxge/sfxge_ev.c
sfxge: assert event queue lock in event handlers

It is useful to highlight lock context.

Sponsored by:   Solarflare Communications, Inc.
Approved by:    gnn (mentor)
18:50:59 - r279176
(4 days 9 hours ago)
by arybchik
Affects:  /head/sys/dev/sfxge/sfxge_dma.c
sfxge: pass correct address to free allocated memory in the case of load error

It is one more place missed in the previous fix.
Most likely is was just memory leak on the error handling path since
typically efsys_mem_t is filled in by zeros on allocation.

Sponsored by:   Solarflare Communications, Inc.
Approved by:    gnn (mentor)
18:48:54 - r279175
(4 days 9 hours ago)
by arybchik
Affects:  /head/sys/dev/sfxge/common/efsys.h
sfxge: using bus_space_*_stream_* API for better portability

Host-bus byte order translation is not requred.

Submitted by:   Artem V. Andreev <Artem.Andreev at oktetlabs.ru>
Sponsored by:   Solarflare Communications, Inc.
Approved by:    gnn (mentor)
18:47:30 - r279174
(4 days 9 hours ago)
by arybchik
Affects:  /head/sys/dev/sfxge/common/siena_mon.c
sfxge: add missing Siena sensors to common code

Sponsored by:   Solarflare Communications, Inc.
Approved by:    gnn (mentor)
18:46:11 - r279173
(4 days 9 hours ago)
by arybchik
Affects:  /head/sys/dev/sfxge/common/efx.h
/head/sys/dev/sfxge/common/siena_nvram.c
sfxge: add missing common code NVRAM types and map from MCDI

Sponsored by:   Solarflare Communications, Inc.
Approved by:    gnn (mentor)
18:44:37 - r279172
(4 days 9 hours ago)
by arybchik
Affects:  /head/sys/dev/sfxge/common/siena_flash.h
sfxge: add new identities to Siena static config

Submitted by:   Andrew Jackson <ajackson at solarflare.com>
Sponsored by:   Solarflare Communications, Inc.
Approved by:    gnn (mentor)
18:04:32 - r279171
(4 days 9 hours ago)
by pho
Affects:  /user/pho/stress2/misc/graid0.sh
/user/pho/stress2/misc/graid1.sh
/user/pho/stress2/misc/graid3.sh
Handle graid not being available.

Sponsored by:    EMC / Isilon storage division
17:50:39 - r279170
(4 days 10 hours ago)
by dchagin
Affects: 
(10 files)
/user/dchagin/lemul/sys/amd64/linux32/*
/user/dchagin/lemul/sys/i386/linux/*
Regen for r279178. (epoll_pwait)
17:50:19 - r279169
(4 days 10 hours ago)
by pho
Affects:  /user/pho/stress2/misc/nullfs3.sh
/user/pho/stress2/misc/nullfs5.sh
/user/pho/stress2/misc/union.sh
Use $mntpoint as mount point.

Sponsored by:    EMC / Isilon storage division
17:48:48 - r279168
(4 days 10 hours ago)
by dchagin
Affects: 
(6 files)
/user/dchagin/lemul/sys/amd64/*
/user/dchagin/lemul/sys/compat/linux/linux_event.c
/user/dchagin/lemul/sys/i386/linux/*
Implement epoll_pwait() system call.
17:45:09 - r279167
(4 days 10 hours ago)
by pho
Affects: 
(17 files)
/user/pho/stress2/misc/*
Remove trailing whitespace.

Sponsored by:    EMC / Isilon storage division
16:28:24 - r279166
(4 days 11 hours ago)
by dim
Affects:  /projects/clang360-import/contrib/llvm/patches/patch-08-llvm-r227089-fix-mips-i128.diff
/projects/clang360-import/contrib/llvm/patches/patch-08-llvm-r230058-indirectbrs-assert.diff
/projects/clang360-import/contrib/llvm/patches/patch-09-llvm-r230058-indirectbrs-assert.diff
Belatedly add llvm patch corresponding to r278367.
16:17:18 - r279165
(4 days 11 hours ago)
by dchagin
Affects:  /user/dchagin/lemul/sys/compat/linux/linux.h
Whoops, LINUX_SIGTBLSZ is used. Always forget about linux_common module.
16:08:27 - r279164
(4 days 11 hours ago)
by dchagin
Affects:  /user/dchagin/lemul/sys/compat/linux/linux_ioctl.c
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.
16:04:37 - r279163
(4 days 11 hours ago)
by dim
Affects: 
(131 files)
/projects/clang360-import/*
Merge ^/head r279023 through r279162.
15:56:16 - r279162
(4 days 12 hours ago)
by dim
Affects:  /head/contrib/llvm/patches/patch-32-llvm-r230058-indirectbrs-assert.diff
Add llvm patch corresponding to r279161.
15:51:49 - r279161
(4 days 12 hours ago)
by dim
Affects:  /head/contrib/llvm/lib/Transforms/Scalar/LoopRotation.cpp
Pull in r230058 from upstream llvm trunk (by Benjamin Kramer):

  LoopRotate: When reconstructing loop simplify form don't split edges
  from indirectbrs.

  Yet another chapter in the endless story. While this looks like we
  leave the loop in a non-canonical state this replicates the logic in
  LoopSimplify so it doesn't diverge from the canonical form in any way.

  http://llvm.org/PR21968

This fixes a "Cannot split critical edge from IndirectBrInst" assertion
failure when building the devel/radare2 port.

PR:             195480, 196987
MFC after:      3 days
15:48:36 - r279160
(4 days 12 hours ago)
by dchagin
Affects:  /user/dchagin/lemul/sys/compat/linux/linux.h
Fix a bug introduced in r279107. Linux supports signal numbers from
1 to 64 incl. To prevent such errors in the future remove LINUX_NSIG
constant which is similar to SIGRTMAX and use last everywhere.
Remove LINUX_SIGTBLSZ constant which is not used anymore.
15:30:08 - r279159
(4 days 12 hours ago)
by kevlo
Affects: 
(2 files)
/stable/10/share/man/man4/wlan.4
MFC r278840:
Xref the following in wlan(4):

- rsu(4)
- urtwn(4)
15:28:49 - r279158
(4 days 12 hours ago)
by kevlo
Affects: 
(2 files)
/stable/10/sys/dev/mii/mii.c
MFC r277481:
Typo: ivalid -> invalid.
15:27:02 - r279157
(4 days 12 hours ago)
by kevlo
Affects: 
(7 files)
/stable/10/share/man/man4/ral.4
/stable/10/sys/contrib/dev/ral/*
/stable/10/sys/dev/ral/*
MFC r278551:
Add preliminary support for the Ralink RT5390 and RT5392 chipsets.
Committed over the D-Link DWA-525 rev A2 on amd64 with WPA.
14:45:00 - r279156
(4 days 13 hours ago)
by pfg
Affects:  /head/usr.sbin/pmccontrol/pmccontrol.c
pmccontrol(8): Replace malloc() + memset() with calloc()

PMCC_OP_IGNORE happens to be zero, so just use calloc() here.
14:03:40 - r279155
(4 days 13 hours ago)
by pho
Affects:  /user/pho/stress2/misc/nfs10.sh
/user/pho/stress2/misc/nfs11.sh
/user/pho/stress2/misc/nullfs8.sh
Added "retrycnt" to nfs mount option.

Sponsored by:    EMC / Isilon storage division
13:36:44 - r279154
(4 days 14 hours ago)
by jilles
Affects:  /head/contrib/netbsd-tests/lib/libc/gen/t_nice.c
/head/lib/libc/gen/nice.3
/head/lib/libc/gen/nice.c
nice(): Correct return value and [EPERM] error.

PR:             189821
Obtained from:  NetBSD
Relnotes:       yes
11:11:05 - r279153
(4 days 16 hours ago)
by andrew
Affects:  /head/sys/arm/broadcom/bcm2835/bcm2835_mbox.c
Add support to the bcm2835 mailbox driver to work before interrupts are
enabled. This will be needed to enable the power on devices early on in the
boot process.
08:54:25 - r279152
(4 days 19 hours ago)
by dchagin
Affects: 
(13 files)
/user/dchagin/lemul/sys/amd64/linux/*
/user/dchagin/lemul/sys/amd64/linux32/*
/user/dchagin/lemul/sys/i386/linux/*
Regen for r279151 (utimensat).
08:52:53 - r279151
(4 days 19 hours ago)
by dchagin
Affects: 
(8 files)
/user/dchagin/lemul/sys/amd64/*
/user/dchagin/lemul/sys/compat/linux/*
/user/dchagin/lemul/sys/i386/linux/*
Add utimensat() system call.

The patch developed by Jilles Tjoelker and Andrew Wilcox and
adopted for lemul branch by me.
08:50:05 - r279150
(4 days 19 hours ago)
by cperciva
Affects: 
(5 files)
/head/ObsoleteFiles.inc
/head/games/*
/head/tools/build/mk/OptionalObsoleteFiles.inc
Step 2 of eliminating the "games" distribution: Remove punch card
utilities.

I was originally planning on removing the phase-of-moon (pom), clock
(grdc), and caesar cipher (caesar, rot13) utilities as well, but after
I committed r278616 I received an astonishing volume of email informing
me that those are still being widely used.  Much to my relief, nobody
reported continuing to use the punch card utilities in production.

The final step will be to merge src/games into src/usr.bin.

This change will not be MFCed.
08:25:08 - r279149
(4 days 19 hours ago)
by dchagin
Affects: 
(480 files)
/user/dchagin/lemul/*
Merge from head.
08:03:44 - r279148
(4 days 19 hours ago)
by dchagin
Affects: 
(8 files)
/user/dchagin/lemul/sys/*
The kernel sends signals to the processes via ABI specific sv_sendsig method.
Native ABI do not need signal conversion, only emulators may want this. Usually
emulators implements its own sv_sendsig method. For now only ibcs2 emulator does
not have own sv_sendsig implementation and depends on native sendsig() method.
So, remove any extra attempts to convert signal numbers from native sendsig()
methods except from i386 where ibsc2 is living.
07:20:04 - r279147
(4 days 20 hours ago)
by arybchik
Affects:  /head/sys/dev/sfxge/sfxge_tx.c
sfxge: TxQ block level should use EFX_TXQ_LIMIT as maximum TxQ size

Sponsored by:   Solarflare Communications, Inc.
Approved by:    gnn (mentor)
07:18:38 - r279146
(4 days 20 hours ago)
by arybchik
Affects:  /head/sys/dev/sfxge/common/siena_mon.c
sfxge: use sparse index to retrieve sensor value

Submitted by:   Andrew Jackson <ajackson at solarflare.com>
Sponsored by:   Solarflare Communications, Inc.
Approved by:    gnn (mentor)
07:17:35 - r279145
(4 days 20 hours ago)
by arybchik
Affects:  /head/sys/dev/sfxge/common/efsys.h
sfxge: use macros to acquire and release BAR locks everywhere

It was the result of invalid merge of the patch from out-of-tree driver.

Sponsored by:   Solarflare Communications, Inc.
Approved by:    gnn (mentor)
07:16:11 - r279144
(4 days 20 hours ago)
by arybchik
Affects:  /head/sys/dev/sfxge/sfxge_tx.c
sfxge: packet can't require more than n_dma_seg descriptors plus one

Use remaining number of DMA segment instead of maximum number in mapping
when checking space for one more TSO segment packet.

Sponsored by:   Solarflare Communications, Inc.
Approved by:    gnn (mentor)
07:14:12 - r279143
(4 days 20 hours ago)
by arybchik
Affects:  /head/sys/dev/sfxge/common/efx_mcdi.c
sfxge: use common definitions of MC shared memory offsets and PDU length

Submitted by:   Ben Hutchings
Sponsored by:   Solarflare Communications, Inc.
Approved by:    gnn (mentor)
07:11:55 - r279142
(4 days 20 hours ago)
by arybchik
Affects:  /head/sys/dev/sfxge/common/efx_regs_ef10.h
sfxge: synchronize EF10 registers header

Sponsored by:   Solarflare Communications, Inc.
Approved by:    gnn (mentor)
07:08:57 - r279141
(4 days 20 hours ago)
by arybchik
Affects: 
(9 files)
/head/sys/dev/sfxge/common/*
/head/sys/dev/sfxge/sfxge_tx.c
sfxge: style fixes and cleanup

Sync endif comment with conditional.
BOOTROM and SIENA_BOOTROM are the same, but highlight that it is Siena.
Restore commented out assertion.
Sync comments with out-of-tree driver.

Sponsored by:   Solarflare Communications, Inc.
Approved by:    gnn (mentor)
06:22:10 - r279140
(4 days 21 hours ago)
by dchagin
Affects:  /user/dchagin/lemul/sys/compat/linprocfs/linprocfs.c
Simplify linprocfs_doprocenviron(). Remove extra proc visibility checks
and initialize pn_vis by well known procfs_candebug().
04:50:47 - r279139
(4 days 23 hours ago)
by marcel
Affects:  /head/usr.bin/mkimg/Makefile
/head/usr.bin/mkimg/mkimg.1
/head/usr.bin/mkimg/mkimg.c
Add the -c option for specifying the capacity of the disk image. When
a capcity is given, no partitions are required. When no partitions are
given, no scheme needs to be specified either. This makes it possible
to create an entirely empty disk image. To add an empty partitioning
table, specify the scheme.

Bump the version to 20150222.
03:30:02 - r279138
(5 days ago)
by nwhitehorn
Affects:  /user/nwhitehorn/ppc64-pmap-rework/aim/mmu_oea64.c
/user/nwhitehorn/ppc64-pmap-rework/aim/moea64_native.c
Fix operation of 32-bit kernels on 64-bit CPUs. I'm not sure this mode is
even really useful, but the fixes aren't so bad.
03:04:40 - r279137
(5 days ago)
by nwhitehorn
Affects: 
(5 files)
/user/nwhitehorn/ppc64-pmap-rework/powermac/*
/user/nwhitehorn/ppc64-pmap-rework/ps3/ps3_syscons.c
MFC @ r279136
02:59:53 - r279136
(5 days ago)
by nwhitehorn
Affects:  /head/sys/powerpc/ps3/ps3_syscons.c
Allow use of higher-resolution (e.g. 1920x1080) framebuffers on PS3.

MFC after:      1 month
02:49:27 - r279135
(5 days 1 hour ago)
by nwhitehorn
Affects:  /user/nwhitehorn/ppc64-pmap-rework/ps3/mmu_ps3.c
The fact that our PMAP code now can run in parallel has exposed the fact
that the PS3 hypervisor's page table code does not have such capabilities.
Slap a lock on all interactions with it.

This fixes occasional lockups presumably caused by conflicting TLB
invalidations.
02:16:24 - r279134
(5 days 1 hour ago)
by pfg
Affects: 
(3 files)
/stable/9/sys/fs/ext2fs/ext2_vfsops.c
MFC     r278790, r278802:
Initialize the allocation of variables related to the ext2 allocator.

Use malloc to clear the values and initialize e2fs_contigdirs
during allocation.  free() e2fs_contigdirs upon error.

While here clean up small style issues.
02:14:49 - r279133
(5 days 1 hour ago)
by pfg
Affects: 
(2 files)
/stable/10/sys/fs/ext2fs/ext2_vfsops.c
MFC     r278790, r278802:
Initialize the allocation of variables related to the ext2 allocator.

Use malloc to clear the values and initialize e2fs_contigdirs
during allocation.  free() e2fs_contigdirs upon error.

While here clean up small style issues.
01:43:30 - r279132
(5 days 2 hours ago)
by pfg
Affects: 
(3 files)
/stable/9/sys/fs/ext2fs/ext2_htree.c
MFC     r278791:
Reuse value of cursize instead of recalculating.

Reported by:    Clang static checker
01:42:45 - r279131
(5 days 2 hours ago)
by pfg
Affects: 
(2 files)
/stable/10/sys/fs/ext2fs/ext2_htree.c
MFC     r278791:
Reuse value of cursize instead of recalculating.

Reported by:    Clang static checker
01:32:37 - r279130
(5 days 2 hours ago)
by pfg
Affects: 
(2 files)
/stable/9/lib/libc/gen/ulimit.c
MFC     r278803, r278905:
ulimit(3): Fix broken check.

The existing implementation had a broken comparison that could
overflow and return confusing values.  Replace this with a check
that avoids the overflow before it happens.

Consistently return a maximum value also on the case of negative
arguments since negative is considered an overflow and means
infinity for our current setrlimit().

New revamped version is credited to Bruce Evans.

CID:            1199295
01:31:28 - r279129
(5 days 2 hours ago)
by pfg
Affects: 
(2 files)
/stable/10/lib/libc/gen/ulimit.c
MFC     r278803, r278905:
ulimit(3): Fix broken check.

The existing implementation had a broken comparison that could
overflow and return confusing values.  Replace this with a check
that avoids the overflow before it happens.

Consistently return a maximum value also on the case of negative
arguments since negative is considered an overflow and means
infinity for our current setrlimit().

New revamped version is credited to Bruce Evans.

CID:            1199295
01:20:49 - r279128
(5 days 2 hours ago)
by marcel
Affects:  /head/usr.bin/mkimg/mkimg.c
/head/usr.bin/mkimg/scheme.c
Don't require a scheme if no partitions are given. Change the code
to handle that case. Note that we still require partitions, so the
change is effectively a no-op.
01:15:09 - r279127
(5 days 2 hours ago)
by pfg
Affects: 
(4 files)
/head/lib/msun/src/*
msun: drop unnecessary checks.

Drop an unnecessary check in some calculations. The check
would have Coverity falsely conclude that variables could
be left undefined.

Discussed with: kargl, bde
Reviewed by:    bde
01:01:02 - r279126
(5 days 2 hours ago)
by marcel
Affects:  /head/usr.bin/mkimg/scheme.c
In scheme_write(), don't overwrite the end parameter with the
return value of image_get_size(). As it so happens, the only
caller of scheme_write() passes exactly that.
00:58:10 - r279125
(5 days 3 hours ago)
by marcel
Affects:  /head/usr.bin/mkimg/mkimg.c
Remove extraneous parenthesis in assignment.
00:01:42 - r279124
(5 days 3 hours ago)
by marcel
Affects:  /head/usr.bin/netstat/netstat.1
Document that netstat(1) supports libxo(3).
00:00:10 - r279123
(5 days 3 hours ago)
by jamie
Affects:  /head/usr.sbin/jls/jls.c
Allow for parameters added with the JP_OPT flag to not exist.
That's why the flag exists in the first place.

MFC after:      1 week
2015 - 02 - 21   (6 days ago)
23:47:20 - r279122
(5 days 4 hours ago)
by marcel
Affects: 
(18 files)
/head/usr.bin/netstat/*
Convert netstat to use libxo.

Obtained from:  Phil Shafer <phil@juniper.net>
Ported to -current by: alfred@ (mostly), Kim Shrier
Formatting: marcel@
Sponsored by:   Juniper Networks, Inc.
23:39:43 - r279121
(5 days 4 hours ago)
by ngie
Affects:  /head/bin/pkill/tests/pgrep-j_test.sh
/head/bin/pkill/tests/pkill-j_test.sh
Disguising stderr from jls generally considered harmful

Undisguising it dumps out the following logspew:

jls: unknown parameter: allow

PR: 191019
23:21:40 - r279120
(5 days 4 hours ago)
by nwhitehorn
Affects:  /user/nwhitehorn/ppc64-pmap-rework/ps3/mmu_ps3.c
Fix reinsertion of pages into secondary PTE group. Something here is still
causing instability under high load, but this is a step forward at least.
23:20:05 - r279119
(5 days 4 hours ago)
by nwhitehorn
Affects:  /head/sys/kern/link_elf.c
Make kernel ELF image parsing not crash for kernels running at locations
other than their link address.
23:04:18 - r279118
(5 days 4 hours ago)
by nwhitehorn
Affects:  /user/nwhitehorn/ppc64-pmap-rework/aim/mmu_oea64.c
/user/nwhitehorn/ppc64-pmap-rework/aim/moea64_native.c
Keep overflow statistics updated.
22:38:32 - r279117
(5 days 5 hours ago)
by kib
Affects:  /head/sys/x86/iommu/busdma_dmar.c
Revert r276949 and redo the fix for PCIe/PCI bridges, which do not
follow specification and do not provide PCIe capability.

Verify if the port above such bridge is downstream PCIe (or root port)
and treat the bridge as PCIe/PCI then.  This allows to avoid
maintaining the table of device ids for bridges without capability,
while still calculate correct request originator for devices behind
the bridge.

Submitted by:   Jason Harmening <jason.harmening@gmail.com>
MFC after:      1 week
22:27:57 - r279116
(5 days 5 hours ago)
by dchagin
Affects: 
(4 files)
/user/dchagin/lemul/sys/i386/linux/*
Finish r279107. Fix build on i386.
22:25:24 - r279115
(5 days 5 hours ago)
by jmg
Affects:  /head/share/man/man9/microuptime.9
back Xref get_cyclecount to have it get more exposure...
22:19:59 - r279114
(5 days 5 hours ago)
by ian
Affects:  /head/sys/arm/include/atomic.h
Correct a comment which was exactly backwards from reality.
22:10:14 - r279113
(5 days 5 hours ago)
by dchagin
Affects:  /user/dchagin/lemul/sys/fs/procfs/procfs_mem.c
Remove extra proc visibility check as it has been done
by pn_vis which is pointed to the procfs_candebug().
21:59:04 - r279112
(5 days 6 hours ago)
by nwhitehorn
Affects:  /user/nwhitehorn/ppc64-pmap-rework/aim/moea64_native.c
Simplicity and correctness improvements to native page table access.
21:28:45 - r279111
(5 days 6 hours ago)
by dchagin
Affects:  /user/dchagin/lemul/sys/compat/linprocfs/linprocfs.c
Fix linprocfs build on i386 - specify proper format for int64_t.
21:12:12 - r279110
(5 days 6 hours ago)
by dchagin
Affects:  /user/dchagin/lemul/sys/compat/linux/linux_timer.c
Convert Linux signal number to the FreeBSD.
21:10:10 - r279109
(5 days 6 hours ago)
by nwhitehorn
Affects:  /user/nwhitehorn/ppc64-pmap-rework/aim/mmu_oea64.c
/user/nwhitehorn/ppc64-pmap-rework/aim/moea64_native.c
Fix locking problems on Apple G5s.
20:41:50 - r279108
(5 days 7 hours ago)
by dchagin
Affects:  /user/dchagin/lemul/sys/compat/linprocfs/linprocfs.c
Convert Linux sigsets before showing.
Linux kernel displays sigset always as 16x4 bit mask.
20:40:40 - r279107
(5 days 7 hours ago)
by dchagin
Affects: 
(16 files)
/user/dchagin/lemul/sys/*
Rework signal code to allow using it by other modules, like linprocfs:

1. Linux sigset always 64 bit on all platforms. In order to move Linux
sigset code to the linux_common module define it as 64 bit int. Move
Linux sigset manipulation routines to the MI path.

2. Move Linux signal number definitions to the MI path. In geneeral they
are the same on all platforms except for a few signals.

3. Map Linux RT signals to the FreeBSD RT signals and hide signal conversion
tables to avoid conversion errors.

4. Emulate Linux SIGPWR signal via FreeBSD SIGRTMIN signal which is outside
of allowed on Linux signal numbers.

PR:             197216
15:18:25 - r279106
(5 days 12 hours ago)
by pho
Affects:  /user/pho/stress2/misc/lockf2.sh
/user/pho/stress2/misc/lockf3.sh
ADded test scenario for lockf(1) and lockf(3).

Sponsored by:    EMC / Isilon storage division
15:15:16 - r279105
(5 days 12 hours ago)
by pho
Affects:  /user/pho/stress2/misc/callout_reset_on.sh
/user/pho/stress2/misc/signal.sh
Use /tmp as working directory.

Sponsored by:    EMC / Isilon storage division
15:02:27 - r279104
(5 days 12 hours ago)
by pfg
Affects:  /head/lib/libc/regex/engine.c
Prevent NULL pointer de-reference.

As a follow up to r279090, if dp hasn't been defined, we
shouldn't attempt to do an optimization here.
13:00:52 - r279103
(5 days 14 hours ago)
by gavin
Affects:  /head/UPDATING
Typo: effect -> affect.
10:10:26 - r279102
(5 days 17 hours ago)
by dchagin
Affects:  /user/dchagin/lemul/sys/compat/linprocfs/linprocfs.c
Add a resid bounds checking to doprocauxv() method.
09:25:53 - r279101
(5 days 18 hours ago)
by dchagin
Affects:  /user/dchagin/lemul/sys/compat/linprocfs/linprocfs.c
Print vdso and stack names to the /proc/self/maps only for Linux procs.
09:17:31 - r279100
(5 days 18 hours ago)
by dchagin
Affects:  /user/dchagin/lemul/sys/compat/linprocfs/linprocfs.c
Reimplement /proc/<pid>/auxv using RAW pfs method.
09:13:32 - r279099
(5 days 18 hours ago)
by dchagin
Affects:  /user/dchagin/lemul/sys/fs/pseudofs/pseudofs_vnops.c
Revert r278821. Its not so trivial to implement pfs fill methods
with autosized sbuf. Will use RAW flag for auxv emulation.
06:28:31 - r279098
(5 days 21 hours ago)
by arybchik
Affects: 
(4 files)
/head/sys/dev/sfxge/common/*
sfxge: allow TX and RX queue limits to be changed

Before the common code had hard coded limits on the IDs RXQs and TXQs could
be created with which were suited for the Windows driver with VMQ, and so
would prevent queues with IDs greater than or equal to 259 (for TXQs) or 768
(for RXQs) from being created. This change allows the limits to be set in
efsys.h, so that all 1024 queues can be created during new manftest tests.
Also, the descriptor cache sizes were also hard coded to values suited to
the smaller queue counts, and so it was necessary to make them configurable
as well.

Submitted by:   Mark Spender <mspender at solarflare.com>
Sponsored by:   Solarflare Communications, Inc.
Approved by:    gnn (mentor)
06:27:45 - r279097
(5 days 21 hours ago)
by arybchik
Affects:  /head/sys/dev/sfxge/common/efx_filter.c
sfxge: check allocations are non-NULL before freeing them

Caught when efx_filter_init() failed and called efx_filter_fini() in the
teardown path.

Submitted by:   Andrew Lee <alee at solarflare.com>
Sponsored by:   Solarflare Communications, Inc.
Approved by:    gnn (mentor)
06:27:16 - r279096
(5 days 21 hours ago)
by arybchik
Affects:  /head/sys/dev/sfxge/common/efx_phy.c
sfxge: keep fw and driver view of autoneg parameters consistent

Previously the driver's view was the expected outcome of any
reconfiguration even if that reconfiguration failed.

Submitted by:   Ben Horgan
Sponsored by:   Solarflare Communications, Inc.
Approved by:    gnn (mentor)
06:26:10 - r279095
(5 days 21 hours ago)
by arybchik
Affects:  /head/sys/dev/sfxge/common/siena_nic.c
sfxge: never set RX_DESCQ_EN during self-test

We must not enable RX queues with random parameters when they are
mapped into a VF with an untrusted driver.  It's probably not a good
idea to do this anyway, so take this bit out of the table test masks.

Submitted by:   Ben Hutchings
Sponsored by:   Solarflare Communications, Inc.
Approved by:    gnn (mentor)
06:24:48 - r279094
(5 days 21 hours ago)
by arybchik
Affects:  /head/sys/dev/sfxge/sfxge_tx.c
/head/sys/dev/sfxge/sfxge_tx.h
sfxge: more accurate calculation of maximum number of TSO segments

Sponsored by:   Solarflare Communications, Inc.
Approved by:    gnn (mentor)
06:05:51 - r279093
(5 days 21 hours ago)
by markj
Affects:  /head/etc/netstart
/head/share/man/man5/rc.conf.5
Purge references to the mrouted rc script, which was removed in r275299.

PR:     195785

 


Powered by Python FreeBSD support by secnetix GmbH & Co. KG

Page generated in 43 ms, 257 revisions printed. Current time is 2015-02-27 03:59:43. All times are in UTC/GMT.