svn commit: r313274 - head/sys/kern

Author: mjg Date: Sun Feb 5 06:51:45 2017 New Revision: 313274 URL: https://svnweb.freebsd.org/changeset/base/313274 Log: sx: add witness support missed in r313272 Modified: head/sys/kern/kern_sx.c Modified: head/sys/kern/kern_sx.c

svn commit: r313273 - head/sbin/kldload

Author: ngie Date: Sun Feb 5 05:36:51 2017 New Revision: 313273 URL: https://svnweb.freebsd.org/changeset/base/313273 Log: style(9) cleanup - Delete trailing whitespace - Fix alignment/variable sorting - Delete single-line enclosing braces MFC after:1 week Sponsored by:

svn commit: r313272 - in head/sys: kern sys

Author: mjg Date: Sun Feb 5 05:20:29 2017 New Revision: 313272 URL: https://svnweb.freebsd.org/changeset/base/313272 Log: sx: uninline slock/sunlock Shared locking routines explicitly read the value and test it. If the change attempt fails, they fall back to a regular function which

svn commit: r313271 - in head/sys: kern sys

Author: mjg Date: Sun Feb 5 04:54:20 2017 New Revision: 313271 URL: https://svnweb.freebsd.org/changeset/base/313271 Log: sx: switch to fcmpset Discussed with: jhb Tested by:pho (previous version) Modified: head/sys/kern/kern_sx.c head/sys/sys/sx.h Modified:

svn commit: r313270 - in head/sys: kern sys

Author: mjg Date: Sun Feb 5 04:53:13 2017 New Revision: 313270 URL: https://svnweb.freebsd.org/changeset/base/313270 Log: rwlock: switch to fcmpset Discussed with: jhb Tested by:pho Modified: head/sys/kern/kern_rwlock.c head/sys/sys/rwlock.h Modified:

Re: svn commit: r313037 - in head/sys: amd64/include kern mips/include net powerpc/include sparc64/include

Probably not related. But when I took short look to the patch to see what could go wrong, I walked into the following comment in _rm_wlock(): "Assumes rm->rm_writecpus update is visible on other CPUs before rm_cleanIPI is called." There is no explicit barrier to ensure it. However, there might be

svn commit: r313269 - in head/sys: kern sys

Author: mjg Date: Sun Feb 5 03:26:34 2017 New Revision: 313269 URL: https://svnweb.freebsd.org/changeset/base/313269 Log: mtx: switch to fcmpset The found value is passed to locking routines in order to reduce cacheline accesses. mtx_unlock grows an explicit check for regular

svn commit: r313268 - head/sys/kern

Author: mjg Date: Sun Feb 5 03:23:16 2017 New Revision: 313268 URL: https://svnweb.freebsd.org/changeset/base/313268 Log: vfs: use atomic_fcmpset in vfs_refcount_* Modified: head/sys/kern/vfs_subr.c Modified: head/sys/kern/vfs_subr.c

Re: svn commit: r313260 - head/sys/kern

On Sun, Feb 05, 2017 at 02:16:12AM +, Steven Hartland wrote: > Hi Mateusz could you improve on the commit message as it currently > describes what is changed, which can be obtained from the diff, but > not why? > > I hope on one feels like I'm trying to teach them to suck eggs, as I > know

svn commit: r313266 - head/sys/cddl/contrib/opensolaris/uts/common/dtrace

Author: markj Date: Sun Feb 5 02:47:34 2017 New Revision: 313266 URL: https://svnweb.freebsd.org/changeset/base/313266 Log: Ensure that the DOF string length is divisible by 2. It is an ASCII encoding of a hexadecimal representation of the DOF file used to enable anonymous tracing, so

Re: svn commit: r313260 - head/sys/kern

> On Feb 4, 2017, at 18:16, Steven Hartland > wrote: > > Hi Mateusz could you improve on the commit message as it currently describes > what is changed, which can be obtained from the diff, but not why? > > I hope on one feels like I'm trying to teach them to

svn commit: r313265 - head/cddl/contrib/opensolaris/cmd/dtrace/test/tst/common/usdt

Author: markj Date: Sun Feb 5 02:45:35 2017 New Revision: 313265 URL: https://svnweb.freebsd.org/changeset/base/313265 Log: Search for _DTRACE_VERSION in sys/sdt.h rather than unistd.h. MFC after:1 week Modified:

svn commit: r313264 - head/cddl/contrib/opensolaris/cmd/dtrace/test/tst/common/usdt

Author: markj Date: Sun Feb 5 02:44:48 2017 New Revision: 313264 URL: https://svnweb.freebsd.org/changeset/base/313264 Log: Avoid using Sun compiler-specific flags. MFC after:1 week Modified: head/cddl/contrib/opensolaris/cmd/dtrace/test/tst/common/usdt/tst.enabled2.ksh Modified:

svn commit: r313263 - head/cddl/contrib/opensolaris/lib/libdtrace/common

Author: markj Date: Sun Feb 5 02:44:08 2017 New Revision: 313263 URL: https://svnweb.freebsd.org/changeset/base/313263 Log: Fix a double free of libelf data buffers in the USDT link code. libdtrace needs to append to the input object files' string and symbol tables. Currently it does so

svn commit: r313262 - in head: cddl/contrib/opensolaris/lib/libdtrace/common sys/cddl/contrib/opensolaris/uts/common/dtrace sys/cddl/contrib/opensolaris/uts/common/sys sys/cddl/dev/dtrace

Author: markj Date: Sun Feb 5 02:39:12 2017 New Revision: 313262 URL: https://svnweb.freebsd.org/changeset/base/313262 Log: Use PC-relative relocations for USDT probe sites on i386 and amd64. When recording probe site addresses in the output DOF file, dtrace -G needs to emit relocations

svn commit: r313261 - head/sys/kern

Author: markj Date: Sun Feb 5 02:27:04 2017 New Revision: 313261 URL: https://svnweb.freebsd.org/changeset/base/313261 Log: Make witness_warn() always print to the console. witness_warn() either breaks into the debugger or panics the system, so its output should go to the console

Re: svn commit: r313260 - head/sys/kern

Hi Mateusz could you improve on the commit message as it currently describes what is changed, which can be obtained from the diff, but not why? I hope on one feels like I'm trying to teach them to suck eggs, as I know everyone here has a wealth of experience, but I strongly believe commit

svn commit: r313260 - head/sys/kern

Author: mjg Date: Sun Feb 5 01:40:27 2017 New Revision: 313260 URL: https://svnweb.freebsd.org/changeset/base/313260 Log: fd: switch fget_unlocked to atomic_fcmpset Modified: head/sys/kern/kern_descrip.c Modified: head/sys/kern/kern_descrip.c

Re: svn commit: r313191 - head/sbin/nvmecontrol

Thanks. -- Cheers, Cy Schubert FreeBSD UNIX: Web: http://www.FreeBSD.org The need of the many outweighs the greed of the few. In message , Warner Losh writes: >

Re: svn commit: r313191 - head/sbin/nvmecontrol

Sorry it took so long. Warner On Sat, Feb 4, 2017 at 6:26 PM, Cy Schubert wrote: > Thanks. > > > -- > Cheers, > Cy Schubert > FreeBSD UNIX: Web: http://www.FreeBSD.org > > The need of the many outweighs the

Re: svn commit: r313191 - head/sbin/nvmecontrol

Thanks! Fixed in r313259. Warner On Sat, Feb 4, 2017 at 6:16 PM, Cy Schubert wrote: > In message <201702040553.v145r1wb002...@repo.freebsd.org>, Warner Losh > writes: >> Author: imp >> Date: Sat Feb 4 05:53:00 2017 >> New Revision: 313191 >> URL:

svn commit: r313259 - head/sbin/nvmecontrol

Author: imp Date: Sun Feb 5 01:20:39 2017 New Revision: 313259 URL: https://svnweb.freebsd.org/changeset/base/313259 Log: Use ssize_t instead of uint32_t to prevent warnings about a comparison with different signs. Due to the promotion rules, this would only happen on 32-bit platforms.

Re: svn commit: r313191 - head/sbin/nvmecontrol

In message <201702040553.v145r1wb002...@repo.freebsd.org>, Warner Losh writes: > Author: imp > Date: Sat Feb 4 05:53:00 2017 > New Revision: 313191 > URL: https://svnweb.freebsd.org/changeset/base/313191 > > Log: > Implement 5 wdc-specific nvme control options for their HGST drives: >

svn commit: r313258 - head/sbin/nvmecontrol

Author: imp Date: Sun Feb 5 00:55:07 2017 New Revision: 313258 URL: https://svnweb.freebsd.org/changeset/base/313258 Log: Add the ability to dump log pages directly in binary to stdout. Update man page to include this flag, and an example of dumping a vendor-specific page while I'm here.

svn commit: r313257 - head/sbin/nvmecontrol

Author: imp Date: Sun Feb 5 00:45:02 2017 New Revision: 313257 URL: https://svnweb.freebsd.org/changeset/base/313257 Log: Add some descriptions to the man page for the supported log pages as well as the new wdc commands. Make wdc be an alias for hgst when specifying the vendor to use to

svn commit: r313254 - head/sys/riscv/include

Author: br Date: Sun Feb 5 00:32:12 2017 New Revision: 313254 URL: https://svnweb.freebsd.org/changeset/base/313254 Log: Implement atomic_fcmpset_*() for RISC-V. Requested by: mjg Sponsored by: DARPA, AFRL Differential Revision:https://reviews.freebsd.org/D9447 Modified:

svn commit: r313253 - head/sbin/ifconfig

Author: adrian Date: Sat Feb 4 20:57:09 2017 New Revision: 313253 URL: https://svnweb.freebsd.org/changeset/base/313253 Log: [net80211] fix quiet_duration parameter to match what is provided in the manpage. Modified: head/sbin/ifconfig/ifieee80211.c Modified:

svn commit: r313252 - head/sbin/nvmecontrol

Author: imp Date: Sat Feb 4 20:43:54 2017 New Revision: 313252 URL: https://svnweb.freebsd.org/changeset/base/313252 Log: Fix off by one error that truncated the serial number for filenames. Modified: head/sbin/nvmecontrol/wdc.c Modified: head/sbin/nvmecontrol/wdc.c

svn commit: r313251 - head/sbin/nvmecontrol

Author: imp Date: Sat Feb 4 20:43:45 2017 New Revision: 313251 URL: https://svnweb.freebsd.org/changeset/base/313251 Log: Fix a typo in usage string for unimplemented command. Modified: head/sbin/nvmecontrol/wdc.c Modified: head/sbin/nvmecontrol/wdc.c

Re: svn commit: r313037 - in head/sys: amd64/include kern mips/include net powerpc/include sparc64/include

Can you post an example of such panic? Only 2 MI pieces were changed, netisr and rmlock. I haven't seen problems on my own amd64/i386/arm testing of this, so a backtrace might help to narrow down the cause. On Sat, Feb 4, 2017 at 12:22 PM, Andreas Tobler wrote: > On

Re: svn commit: r313037 - in head/sys: amd64/include kern mips/include net powerpc/include sparc64/include

On 04.02.17 20:54, Jason Harmening wrote: I suspect this broke rmlocks for mips because the rmlock implementation takes the address of the per-CPU pc_rm_queue when building tracker lists. That address may be later accessed from another CPU and will then translate to the wrong physical region if

Re: svn commit: r313037 - in head/sys: amd64/include kern mips/include net powerpc/include sparc64/include

On Fri, 3 Feb 2017 21:29:33 -0800 Jason Harmening wrote: > Hi, > > I'm a bit confused as to how this change breaks MIPS. The new > function, get_pcpu() is intended to be used only to access the > per-cpu data pointer locally. It returns pcpup, which is the per-cpu >

Re: svn commit: r313037 - in head/sys: amd64/include kern mips/include net powerpc/include sparc64/include

I suspect this broke rmlocks for mips because the rmlock implementation takes the address of the per-CPU pc_rm_queue when building tracker lists. That address may be later accessed from another CPU and will then translate to the wrong physical region if the address was taken relative to the

svn commit: r313250 - in head/sys/dev: mmc sdhci

Author: marius Date: Sat Feb 4 19:35:38 2017 New Revision: 313250 URL: https://svnweb.freebsd.org/changeset/base/313250 Log: Fix some more overly long lines, whitespace and other bugs according to style(9) as well as spelling in comments. Modified: head/sys/dev/mmc/mmc.c

svn commit: r313249 - head/sys/vm

Author: kib Date: Sat Feb 4 19:16:19 2017 New Revision: 313249 URL: https://svnweb.freebsd.org/changeset/base/313249 Log: Style, use tab after #define. Reviewed by: alc Sponsored by: The FreeBSD Foundation MFC after:3 days Modified: head/sys/vm/vm_object.h Modified:

svn commit: r313248 - head/sys/net

Author: sbruno Date: Sat Feb 4 18:25:09 2017 New Revision: 313248 URL: https://svnweb.freebsd.org/changeset/base/313248 Log: Delete duplicate break. Modified: head/sys/net/iflib.c Modified: head/sys/net/iflib.c ==

svn commit: r313195 - in head/sbin: decryptcore savecore

Author: def Date: Sat Feb 4 14:10:16 2017 New Revision: 313195 URL: https://svnweb.freebsd.org/changeset/base/313195 Log: Fix bugs found by Coverity in decryptcore(8) and savecore(8): - Perform final decryption and write decrypted data in case of non-block aligned input data; - Use

svn commit: r313194 - in head/sys: arm/include arm64/include mips/include powerpc/include riscv/include sparc64/include sys x86/include

Author: kib Date: Sat Feb 4 12:26:38 2017 New Revision: 313194 URL: https://svnweb.freebsd.org/changeset/base/313194 Log: Define the vm_ooffset_t and vm_pindex_t types as machine-independend. The types are for the byte offset and page index in vm object. They are similar to off_t,

Re: svn commit: r313037 - in head/sys: amd64/include kern mips/include net powerpc/include sparc64/include

On 04.02.17 07:27, Jason Harmening wrote: It's hard to argue with that:) I've backed it out until we can figure out what's going on. Sorry for the breakage. For the record, powerpc64 was also affected. Andreas On Fri, Feb 3, 2017 at 9:54 PM, Kurt Lidl >