Author: kan
Date: Mon Jan 27 00:14:51 2020
New Revision: 357161
URL: https://svnweb.freebsd.org/changeset/base/357161
Log:
Enter net epoch in msk_tick.
One more instance of if_input being called outside of
interrupt, by means of msk_handle_events.
Differential Revision: https://revi
Author: kan
Date: Wed Oct 2 16:08:01 2019
New Revision: 353011
URL: https://svnweb.freebsd.org/changeset/base/353011
Log:
Revert r352953: Convert pnmatch to single element array in regexec calls
Requested by: cem
Modified:
head/usr.bin/killall/killall.c
head/usr.bin/split/split.c
Mod
Author: kan
Date: Wed Oct 2 06:15:30 2019
New Revision: 352953
URL: https://svnweb.freebsd.org/changeset/base/352953
Log:
Convert pnmatch to single element array in regexec calls
The regexec function is declared as taking an array of regmatch_t
elements, and passing in the pointer to sin
Author: kan
Date: Mon Dec 25 04:48:39 2017
New Revision: 327173
URL: https://svnweb.freebsd.org/changeset/base/327173
Log:
Do pass removing some write-only variables from the kernel.
This reduces noise when kernel is compiled by newer GCC versions,
such as one used by external toolchain p
Author: kan
Date: Sun Dec 24 16:39:57 2017
New Revision: 327163
URL: https://svnweb.freebsd.org/changeset/base/327163
Log:
Remove write-only opt and useless optlen variables.
This squashes the warning gebnerated by GCC 6.x. Since
variables that are now removed had come documentation
val
Author: kan
Date: Sat Dec 23 17:55:19 2017
New Revision: 327114
URL: https://svnweb.freebsd.org/changeset/base/327114
Log:
Reverse the check to allocate the buffer if cached pointer is NULL.
Reviewed by: kib
Differential Revision: https://reviews.freebsd.org/D13596
Modified:
head/sys/k
Author: kan
Date: Sat Dec 23 16:49:57 2017
New Revision: 327113
URL: https://svnweb.freebsd.org/changeset/base/327113
Log:
Remove dead store to local variable.
Modified:
head/sys/kern/kern_shutdown.c
Modified: head/sys/kern/kern_shutdown.c
Author: kan
Date: Sat Dec 23 16:45:26 2017
New Revision: 327112
URL: https://svnweb.freebsd.org/changeset/base/327112
Log:
Silence clang analyzer false positive.
clang does not know that two lookup calls will return the same
pointer, so it assumes correctly that using the old pointer
af
Author: kan
Date: Sat Dec 23 16:45:24 2017
New Revision: 327111
URL: https://svnweb.freebsd.org/changeset/base/327111
Log:
Do not pass NULL pointer to copyout in if_clone_list.
Sometimes caller is only interested in how many clones
are there and NULL pointer is passed for the destination
Author: kan
Date: Sat Dec 23 16:23:58 2017
New Revision: 327108
URL: https://svnweb.freebsd.org/changeset/base/327108
Log:
Do not double free the memory in if_clone.
if_clone_attach function will drop the reference on failure which will
free the if_clone structure. No need to do it secon
Author: kan
Date: Sat Dec 23 16:24:02 2017
New Revision: 327110
URL: https://svnweb.freebsd.org/changeset/base/327110
Log:
Remove dead initialization of the inode pointer.
The pointer gets initialized again later in the code. This also
improves code style(9).
Modified:
head/sys/ufs/ffs
Author: kan
Date: Sat Dec 23 16:24:00 2017
New Revision: 327109
URL: https://svnweb.freebsd.org/changeset/base/327109
Log:
Remove some trailing whitespace.
Reviewed by: glebius, ae
Differential Revision: https://reviews.freebsd.org/D10386
Modified:
head/sys/net/if_clone.c
Modified: he
Author: kan
Date: Thu Apr 13 14:23:27 2017
New Revision: 316757
URL: https://svnweb.freebsd.org/changeset/base/316757
Log:
Use proper fields to check for interrupt trigger mode.
Modified:
head/sys/arm/arm/gic.c
head/sys/arm64/arm64/gic_v3.c
Modified: head/sys/arm/arm/gic.c
Author: kan
Date: Sun Apr 9 03:50:48 2017
New Revision: 316649
URL: https://svnweb.freebsd.org/changeset/base/316649
Log:
Add missing double quote to fix r316635 commit.
Modified:
head/lib/libz/Makefile
Modified: head/lib/libz/Makefile
===
Author: kan
Date: Fri Apr 7 22:58:28 2017
New Revision: 316626
URL: https://svnweb.freebsd.org/changeset/base/316626
Log:
Define 'lr' as x30 on aarch64
GNU toolchain does not recognize LR as standard register alias,
but clang does. Use of #define will work on both. Place the
definition
Author: kan
Date: Fri Apr 7 22:58:20 2017
New Revision: 316624
URL: https://svnweb.freebsd.org/changeset/base/316624
Log:
Do not use b.cs instruction to jump to cerror.
The conditional jump can only be performed to targets up to 1MB in
either direction and does not work too well when lin
Author: kan
Date: Fri Apr 7 22:58:31 2017
New Revision: 316627
URL: https://svnweb.freebsd.org/changeset/base/316627
Log:
Use int to receive the return value of getopt function.
getopt returns int and not char, so assigning the value to
char is not ideal, especially on platforms with uns
Author: kan
Date: Fri Apr 7 22:58:34 2017
New Revision: 316628
URL: https://svnweb.freebsd.org/changeset/base/316628
Log:
Silence GCC warning by initializing the local variable.
GCC 6.3 is unable to trace all code paths that lead to
this variable being left uninitialized and correlate th
Author: kan
Date: Fri Apr 7 22:58:25 2017
New Revision: 316625
URL: https://svnweb.freebsd.org/changeset/base/316625
Log:
Do not use -msoft-float with intention of disabling FP on aarch64
GNU GCC does does recognise it as a valid option and we already
use -mgeneral-regs-only that has the
Author: kan
Date: Thu Mar 16 00:49:59 2017
New Revision: 315338
URL: https://svnweb.freebsd.org/changeset/base/315338
Log:
Add cavium octeon 'bootoctlinux' boot argument support
While there, parse u-boot provided command line arguments
for supported switches and update boothowto appropria
Author: kan
Date: Mon Feb 6 14:00:28 2017
New Revision: 313341
URL: https://svnweb.freebsd.org/changeset/base/313341
Log:
Use 64bit store instruction in atomic_fcmpset_64.
Reported by: br
Modified:
head/sys/mips/include/atomic.h
Modified: head/sys/mips/include/atomic.h
y pc_rm_queue is a cyclic list
consisting of just one element pc_rm_queue, the loop terminating
condition will never be met, as &pc->pc_rm_qeue value will be very
different from the one that pc->pc_rm_queue.rmq_next was set at
initialisation time.
This affects MIPS in SMP configurati
id array. I
therefore officially am requesting this change to be reverted until
reasonable solution is found to unbreak architectures that use wired
TLBs to access local per-CPU data.
--
Alexander Kabaev
pgpA1EtpIy7w7.pgp
Description: Цифровая подпись OpenPGP
esired than other? I am not against the
change, but I need to understand the reasoning behind it better. Since
atomic_cmpset retries too, it will have to be adjusted as well.
--
Alexander Kabaev
pgpVlTYDug3ZT.pgp
Description: Цифровая подпись OpenPGP
Author: kan
Date: Wed Feb 1 05:00:34 2017
New Revision: 313040
URL: https://svnweb.freebsd.org/changeset/base/313040
Log:
Add atomic_fcmpset_*() inlines for MIPS
atomic_fcmpset_*() is analogous to atomic_cmpset(), but saves off the
read value from the target memory location into the 'old
Author: kan
Date: Thu Jan 12 18:05:12 2017
New Revision: 311993
URL: https://svnweb.freebsd.org/changeset/base/311993
Log:
Fix typo in r311971.
Reported by: ohartmann at walstatt.org
Modified:
head/sys/dev/nand/nand_geom.c
Modified: head/sys/dev/nand/nand_geom.c
===
Author: kan
Date: Fri Jan 6 00:07:36 2017
New Revision: 311463
URL: https://svnweb.freebsd.org/changeset/base/311463
Log:
Remove redundant cache initialization in JZ4780 SMP startup code
This was done out of pure paranoia when hunting for bugs in cache
and is not really required.
Modifi
Author: kan
Date: Thu Jan 5 19:17:14 2017
New Revision: 311458
URL: https://svnweb.freebsd.org/changeset/base/311458
Log:
Use compiler driver to link BERI boot loaders
Do not hardcode elf64-tradbigmips as output format in BERI linker scrips.
Unfortunately, in-tree toolchain and external
Author: kan
Date: Wed Jan 4 18:36:30 2017
New Revision: 311299
URL: https://svnweb.freebsd.org/changeset/base/311299
Log:
Remove duplicate iic and iicbus devices from JZ4780 kernel
Modified:
head/sys/mips/conf/JZ4780
Modified: head/sys/mips/conf/JZ4780
==
Author: kan
Date: Fri Dec 30 17:12:41 2016
New Revision: 310829
URL: https://svnweb.freebsd.org/changeset/base/310829
Log:
Restore status quo: mips64 does not need subr_sfbuf.c
Reported by: br
Modified:
head/sys/conf/files.mips
Modified: head/sys/conf/files.mips
mpile+link
as you suggest. It feeds .o file to to the compiler driver which
then invokes ld with all the correct parameters for the target object
format.
--
Alexander Kabaev
pgpcFsm1wwEwq.pgp
Description: Цифровая подпись OpenPGP
Author: kan
Date: Fri Dec 30 00:34:52 2016
New Revision: 310794
URL: https://svnweb.freebsd.org/changeset/base/310794
Log:
Support mips[*]hf variants in config files
Recognize new MACHINE_ARCH names now as we have added hardfloat support.
Switch JZ4780 to mipselhf and remove all uses of T
Author: kan
Date: Thu Dec 29 21:36:04 2016
New Revision: 310790
URL: https://svnweb.freebsd.org/changeset/base/310790
Log:
Use TARGET_ARCH instead of MACHINE_ARCH for MIPS kernel
MACHINE_ARCH is overwritten by config file and will not
contain -hf suffix, so uname -p reported by kernel wil
Author: kan
Date: Thu Dec 29 21:30:52 2016
New Revision: 310789
URL: https://svnweb.freebsd.org/changeset/base/310789
Log:
Use compiler driver to build relocatable object
This works better with external toolchains where LD
will not necessarily defailt to emulation we want.
Compiler driv
Author: kan
Date: Thu Dec 29 20:11:50 2016
New Revision: 310784
URL: https://svnweb.freebsd.org/changeset/base/310784
Log:
Use mips_dcache_wbinv_range instead of mips_dcache_wb_range on CPU_XBURST
Ingenic CPUs treat plain cache writeback as local-only operation and do
nothing if that is a
Author: kan
Date: Wed Dec 28 19:40:59 2016
New Revision: 310722
URL: https://svnweb.freebsd.org/changeset/base/310722
Log:
Do not use read-modify-write on MSC control register.
The register is write-only, so just write only bits we want.
Submitted by: jmcneill
Modified:
head/sys/mip
Author: kan
Date: Wed Dec 28 13:48:00 2016
New Revision: 310700
URL: https://svnweb.freebsd.org/changeset/base/310700
Log:
Be more conservative when enabling write-combining on MIPS
Some MIPS revisions do implement uncached-accelerate caching
attribute, but place extra requirement on acce
ange that does not map UA to
WC unconditionally and will let people who need write combining to
enable it on a case by case basis. For now, only Ingenic XBursts are
doing so as these are ones I have tested.
--
Alexander Kabaev
pgp4owHqbim3S.pgp
Description: Цифровая подпись OpenPGP
Author: kan
Date: Wed Dec 28 02:55:26 2016
New Revision: 310650
URL: https://svnweb.freebsd.org/changeset/base/310650
Log:
Implement pmap_change_attr and related APIs on MIPS
On platforms that have uncached-accelerate cache attribute, map it
to VM_MEMATTR_WRITE_COMBINING. Otherwise, leave
@@
+/*-
+ * Copyright 2016 Alexander Kabaev
+ * All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *notice, this
Author: kan
Date: Fri Dec 23 19:30:14 2016
New Revision: 310486
URL: https://svnweb.freebsd.org/changeset/base/310486
Log:
Pass memattrs to fb device so that user mappings can inherit them
Reviewed by: jmcneill
Modified:
head/sys/mips/ingenic/jz4780_lcd.c
Modified: head/sys/mips/ingeni
Author: kan
Date: Sun Nov 20 01:59:22 2016
New Revision: 308872
URL: https://svnweb.freebsd.org/changeset/base/308872
Log:
Use ofw_gpiobus instead of plain gpiobus for CI20
Requesed by: mmel
Modified:
head/sys/mips/ingenic/jz4780_gpio.c
Modified: head/sys/mips/ingenic/jz4780_gpio.c
Author: kan
Date: Fri Nov 18 16:23:31 2016
New Revision: 308807
URL: https://svnweb.freebsd.org/changeset/base/308807
Log:
Subject: [PATCH] Pass MACHINE_ARCH on command line for MIPS kernels.
While there, make param.h guess proper MACHINE_ARCH on hardfloat targets
correctly as well, so to
Author: kan
Date: Wed Nov 16 03:24:20 2016
New Revision: 308713
URL: https://svnweb.freebsd.org/changeset/base/308713
Log:
Compile trampoline with soft-float on MIPS, to match the rest of the kernel
Core kernel is always compiled with -msoft-float on all of our platforms,
make sure we fol
Author: kan
Date: Wed Nov 16 03:21:49 2016
New Revision: 308712
URL: https://svnweb.freebsd.org/changeset/base/308712
Log:
Make MIPS startup assembly files use neutral float ABI.
This allows these files to be used with hard and softfloat targets
with no special flags passed to the compile
Author: kan
Date: Wed Nov 16 03:19:36 2016
New Revision: 308711
URL: https://svnweb.freebsd.org/changeset/base/308711
Log:
Set endianness and floating point flags explicitly for MIPS targets
The tree can be build with an external toolchain that will not
necessarily default to desired sett
if (callout_stop(&lle->lle_timer) > 0)
LLE_REMREF(lle);
This results in panics as described in:
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=210884
--
Alexander Kabaev
pgpCIQAIpPC7t.pgp
Description: Цифровая подпись OpenPGP
1a5a649815318391789e977a198
is a bit more complete version that only enables hardfloat in routines
that need it, not in random place of the file and for the rest of it.
--
Alexander Kabaev
pgpIVPympkI4f.pgp
Description: Цифровая подпись OpenPGP
Author: kan
Date: Wed May 4 01:56:37 2016
New Revision: 299034
URL: https://svnweb.freebsd.org/changeset/base/299034
Log:
Fix nvram2env after bus_space_fdt removal.
The generic bus space does the mapping now, no need to pre-translate
phys addresses to KSEG1 anymore.
Submitted by: Hi
Author: kan
Date: Sun Mar 27 23:19:14 2016
New Revision: 297327
URL: https://svnweb.freebsd.org/changeset/base/297327
Log:
Do not override top level CFLAGS in libexec/dma
Modified:
head/libexec/dma/Makefile.inc
Modified: head/libexec/dma/Makefile.inc
=
Author: kan
Date: Sun Mar 27 23:19:21 2016
New Revision: 297329
URL: https://svnweb.freebsd.org/changeset/base/297329
Log:
Use correct response bits for MMC_RSP_R4-R7 types
Modified:
head/sys/dev/mmc/mmcreg.h
Modified: head/sys/dev/mmc/mmcreg.h
===
Author: kan
Date: Sun Mar 27 23:19:20 2016
New Revision: 297328
URL: https://svnweb.freebsd.org/changeset/base/297328
Log:
Do nothing if requested clk frequency is already correct.
Reviewed by: mmel
Modified:
head/sys/dev/extres/clk/clk.c
Modified: head/sys/dev/extres/clk/clk.c
Nope, mips shares no IPI code with ARM and probably shouldn't too,
since IPI delivery and mapping between IPI types and IRQs are very
platform specific.
--
Alexander Kabaev
pgpp2Qgb3hV6x.pgp
Description: Цифровая подпись OpenPGP
On Wed, 27 Jan 2016 20:20:37 + (UTC)
Alexander Kabaev wrote:
> Author: kan
> Date: Wed Jan 27 20:20:37 2016
> New Revision: 294936
> URL: https://svnweb.freebsd.org/changeset/base/294936
>
> Log:
> Do not unlock rtld_phdr_lock over callback invocations.
>
Author: kan
Date: Wed Jan 27 20:20:37 2016
New Revision: 294936
URL: https://svnweb.freebsd.org/changeset/base/294936
Log:
Do not unlock rtld_phdr_lock over callback invocations.
The dl_iterate_phdr consumer code in libgcc does not expect multiple
callbacks running concurrently. This was
Author: kan
Date: Wed Jan 27 20:20:35 2016
New Revision: 294935
URL: https://svnweb.freebsd.org/changeset/base/294935
Log:
Make .debug file for libgcc_s.so.1 more useful.
The files compiled into libgcc_s.so.1 did not have -g on
compiler command line, making generated .debug quite
pointl
On Sat, 23 Jan 2016 12:00:23 +
Antoine Brodin wrote:
> On Thu, Jan 21, 2016 at 11:36 PM, Alexander Kabaev
> wrote:
> > On Thu, 21 Jan 2016 22:20:12 +0100
> > Antoine Brodin wrote:
> >
> >> On Thu, Jan 21, 2016 at 12:26 AM, Alexander Kabaev
> >>
On Thu, 21 Jan 2016 22:20:12 +0100
Antoine Brodin wrote:
> On Thu, Jan 21, 2016 at 12:26 AM, Alexander Kabaev
> wrote:
> > Author: kan
> > Date: Wed Jan 20 23:26:35 2016
> > New Revision: 294470
> > URL: https://svnweb.freebsd.org/changeset/base/294470
> >
&g
Author: kan
Date: Wed Jan 20 23:26:35 2016
New Revision: 294470
URL: https://svnweb.freebsd.org/changeset/base/294470
Log:
Fix initlist_add_object invocation parameters.
The tail parameter should point to the last object for
which dependencies should be processed. In most cases,
this is
f defined(NO_CLEANDIR)
For some definition of 'correct'. It definitely is not for those who
have -DNOCLEAN wired into their finger memory and above mentioned
decade did not help with that any.
I think some of the common spellings were left as is in great NO_
revolution for a
Author: kan
Date: Fri Oct 9 18:39:55 2015
New Revision: 289076
URL: https://svnweb.freebsd.org/changeset/base/289076
Log:
Add definitions for MIPS TLS relocations to elftoolchain.
This makes our readelf more useful when looking for TLS-related
issues.
Modified:
head/contrib/elftoolcha
Author: kan
Date: Fri Oct 9 18:39:52 2015
New Revision: 289075
URL: https://svnweb.freebsd.org/changeset/base/289075
Log:
Remove some trailing space.
Modified:
head/libexec/rtld-elf/powerpc64/reloc.c
Modified: head/libexec/rtld-elf/powerpc64/reloc.c
=
On Wed, 02 Sep 2015 16:57:01 +0300
Alexander Motin wrote:
> On 02.09.2015 16:36, Alexander Kabaev wrote:
> > On Wed, 2 Sep 2015 16:20:34 +0300
> > Gleb Smirnoff wrote:
> >
> >> On Wed, Sep 02, 2015 at 09:08:25AM -0400, Alexander Kabaev wrote:
> >> A> &
On Wed, 2 Sep 2015 16:20:34 +0300
Gleb Smirnoff wrote:
> On Wed, Sep 02, 2015 at 09:08:25AM -0400, Alexander Kabaev wrote:
> A> > A> > Modified: head/Makefile.inc1
> A> > A> >
> =
On Wed, 2 Sep 2015 16:20:34 +0300
Gleb Smirnoff wrote:
> On Wed, Sep 02, 2015 at 09:08:25AM -0400, Alexander Kabaev wrote:
> A> >
> A> > Does that happen to you, or do you just suppose that this is
> A> > going to happen looking at the code?
> A> >
>
On Wed, 2 Sep 2015 15:50:24 +0300
Gleb Smirnoff wrote:
> Alexander,
>
> On Tue, Sep 01, 2015 at 10:44:38PM -0400, Alexander Kabaev wrote:
> A> On Tue, 1 Sep 2015 11:59:12 + (UTC)
> A> > Log:
> A> > Not only build with buildworld, but also install with
&g
On Thu, Aug 27, 2015 at 12:04 PM, Pedro Giffuni wrote:
> Hello kan;
>
> On 08/27/15 09:00, Alexander Kabaev wrote:
>
>> Author: kan
>> Date: Thu Aug 27 14:00:23 2015
>> New Revision: 287206
>> URL: https://svnweb.freebsd.org/changeset/base/287206
>>
>
Author: kan
Date: Thu Aug 27 14:00:23 2015
New Revision: 287206
URL: https://svnweb.freebsd.org/changeset/base/287206
Log:
Repair sys/cdefs.h enough to be usable with GCC 5.x
The __alloc_size and __alloc_align need to be defined to
nothingness for lint, but the existing check is deficient
Author: kan
Date: Thu Aug 27 14:00:16 2015
New Revision: 287204
URL: https://svnweb.freebsd.org/changeset/base/287204
Log:
Unbreak nvi message catalog generation for 8 bit locales.
Feeding any file encoded in 8 bit locales such as KOI8-RU
to sort utility running under UTF-8 locale produce
Author: kan
Date: Thu Aug 27 14:00:20 2015
New Revision: 287205
URL: https://svnweb.freebsd.org/changeset/base/287205
Log:
Make ncurses build with GCC 5.0 and up
Merge the end result of two upstream changes:
Original fix from 20141206:
+ modify MKlib_gen.sh to work around change in
On Sun, 9 Aug 2015 09:37:13 +0200
Ed Schouten wrote:
> Hi Alexander,
>
> 2015-08-08 21:05 GMT+02:00 Alexander Kabaev :
> > this commit broke more than just syslogd. rtsol and rtsold both are
> > victims of the change as they are trying to shutdown unconnected
> > raw
work Subscribers: bms, mjg, imp
> Differential Revision: https://reviews.freebsd.org/D3039
>
Hi Ed,
this commit broke more than just syslogd. rtsol and rtsold both are
victims of the change as they are trying to shutdown unconnected
raw sockets to make then 'send-only' and fail.
--
Alexander Kabaev
pgps0YItNxzam.pgp
Description: OpenPGP digital signature
ng it for port bringups.
>
> Thanks,
>
>
> -adrian
>
IIRC, FreeBSD's internal GCC defines __FreeBSD_cc_version, while stock
one does not.
--
Alexander Kabaev
pgpqLH71LqYuS.pgp
Description: OpenPGP digital signature
Author: kan
Date: Wed Jun 17 04:26:48 2015
New Revision: 284494
URL: https://svnweb.freebsd.org/changeset/base/284494
Log:
Continue ubreaking ``env LANG=ru_RU.KOI8-R ls -l''.
File names are in current locale as well.
Modified:
head/bin/ls/print.c
Modified: head/bin/ls/print.c
==
his in printlong:
- xo_emit("{ke:name}", name);
+ xo_emit("{ke:name/%hs}", name)
- xo_emit("{dk:name}", name);
+ xo_emit("{dk:name/%hs}", name);
--
Alexander Kabaev
pgpLIhePiAmRE.pgp
Description: OpenPGP digital signature
Author: kan
Date: Sat Jun 13 15:14:39 2015
New Revision: 284340
URL: https://svnweb.freebsd.org/changeset/base/284340
Log:
Unbreak libxo's handling of characters not representable in current locale
The xo_format_string_direct function loops forever never advancing the
processed string poi
On Fri, 12 Jun 2015 20:43:09 -0400
Alexander Kabaev wrote:
> On Wed, 10 Jun 2015 01:27:39 + (UTC)
> Marcel Moolenaar wrote:
>
> > Author: marcel
> > Date: Wed Jun 10 01:27:38 2015
> > New Revision: 284198
> > URL: https://svnweb.freebsd.org/changeset/base/2
Phil Shafer
> Sponsored by: Juniper Networks, Inc.
>
This broke all code that pipes output of the ls command to pipeline,
such as 'ls | wc -l'. ls never exits and never output anything. Is
there any purpose to libxo other than breaking stuff, which it ach
On Wed, 20 May 2015 09:54:45 -0700
Matthew Ahrens wrote:
> On Wed, May 20, 2015 at 9:00 AM, Alexander Kabaev
> wrote:
>
> > On Fri, 15 May 2015 13:50:38 + (UTC)
> > John Baldwin wrote:
> >
> > > Author: jhb
> > > Date: Fri May 15 13:50:37 201
ssed.
It is debatable just how correct ZFS code in its expectations, but I
think this commit should probably be reverted until either ZFS is
changed not to expect cv modifiable by waking threads or until
alternative solution is found to the cv_waiters overflow issue fixed by
this commit.
--
Alexander Kabaev
pgpESgZs_K4y1.pgp
Description: OpenPGP digital signature
ix(dev, &num_vectors) != 0) {
Huh, Linux just falls back to as many vectors as it can and just
allocates them between per-cpu queues in a round-robin manner. I think
is is a better approach than what we have here, would you consider it?
--
Alexander Kabaev
pgphqx9dYSivT.pgp
Description: OpenPGP digital signature
Author: kan
Date: Sat Feb 28 22:24:45 2015
New Revision: 279412
URL: https://svnweb.freebsd.org/changeset/base/279412
Log:
Do not leak 'copy' buffer if bmap_truncate_indirect fails.
Reported by: Brainy Code Scanner, by Maxime Villard.
MFC after: 2 weeks
Modified:
head/sys/fs/nandfs/bma
Author: kan
Date: Sat Feb 28 20:30:29 2015
New Revision: 279405
URL: https://svnweb.freebsd.org/changeset/base/279405
Log:
Avoid closing unallocated socket in case socreate fails.
Found by: Brainy Code Scanner
Reported by: Maxime Villard
MFC after: 2 weeks
Modified:
head/sys/nfs/krp
Author: kan
Date: Sat Feb 28 20:30:25 2015
New Revision: 279404
URL: https://svnweb.freebsd.org/changeset/base/279404
Log:
Avoid lookup of CODESET aliases using uninitialized path
We do not use iconv.alias file, so avoid using the vestiges
of the code that do.
Differential Revision:
On Sat, 08 Nov 2014 22:09:38 +0200
Andriy Gapon wrote:
> On 08/11/2014 19:00, Alexander Kabaev wrote:
> > On Thu, 6 Nov 2014 22:46:40 + (UTC) Mark Johnston
> > wrote:
> >
> >> Author: markj Date: Thu Nov 6 22:46:40 2014 New Revision: 274200
> >> UR
get rid of the debug symbols?
IMHO, we should only pass -g to ctfconvert if WITH_CTF is set _and_ user
requested debug symbols explicitly. Your change seems to be breaking
that.
--
Alexander Kabaev
pgp15e_jzijDZ.pgp
Description: OpenPGP digital signature
do a 'sysctl kern.random' and post the result?
>
> Do you have anything random-related in your rc.conf?
>
> Could you please do a verbose boot on a kernel with "options
> RANDOM_DEBUG" set and send me the dmesg output from a failing box.
>
> M
>
> &g
s- like it's not getting enough entropy now?
>
>
>
> -adrian
>
Me too, on amd64 box. The machine gets hung for several minutes after
'writing entropy file' until random is unblocked. Anecdotally, banging
on the keyboard trying to help the entropy gathering along seems to
speed the process up.
--
Alexander Kabaev
pgprcVla2H9qP.pgp
Description: OpenPGP digital signature
Author: kan
Date: Sat Sep 27 16:50:24 2014
New Revision: 272215
URL: http://svnweb.freebsd.org/changeset/base/272215
Log:
Do not pass whole descriptor block size as align to fwdma_malloc_multiseg
Do not pass wrong alignment value to fwdma_malloc_multiseg and ultimately
to contigalloc. In
Author: kan
Date: Sat Sep 27 16:50:21 2014
New Revision: 272214
URL: http://svnweb.freebsd.org/changeset/base/272214
Log:
Remove obsolete compatibility glue and improve firewire code readability.
Commit my version of style(9) pass over the firewire code. Now that
other people have started
Author: kan
Date: Sun Sep 14 00:02:37 2014
New Revision: 271561
URL: http://svnweb.freebsd.org/changeset/base/271561
Log:
Add delay to Octeon MDIO access routines.
Prevent saturattion of the bus by constant polling which in
extreme cases can cause interface lockup. This makes FreeBSD
ma
Author: kan
Date: Sun Sep 14 00:02:40 2014
New Revision: 271562
URL: http://svnweb.freebsd.org/changeset/base/271562
Log:
Fix RTC clock writes on many Octeon boards.
The struct clocktime uses 0-based week day number, so back out part of
r229161 by gonzo, which actually broke the RTC clock
Author: kan
Date: Thu Aug 21 15:10:10 2014
New Revision: 270277
URL: http://svnweb.freebsd.org/changeset/base/270277
Log:
Add guards to ptrdiff_t definition in include/stddef.h
Back in 2011 obrien has added the #define macro in sys/sys/stddef.h to
guard ptrdiff_t. Add similar protection t
g_dev_attrchanged(cp, "GEOM::physpath");
> >
>
> This part looks unrelated.
>
Commit says that it deals with consequences of make_dev_alias_p
failing. Preventing trap happening on said failure seems to be
pretty much related to the stated goal.
--
Alexander Kabaev
signature.asc
Description: PGP signature
Author: kan
Date: Mon Jun 23 03:45:39 2014
New Revision: 267762
URL: http://svnweb.freebsd.org/changeset/base/267762
Log:
Restore the check for non-NULL dmatag in sndbuf_free.
The sound drivers that use own buffer management can use sndbuf_setup
and not do any busdma allocation, so the dr
Author: kan
Date: Fri Jun 20 01:45:03 2014
New Revision: 267660
URL: http://svnweb.freebsd.org/changeset/base/267660
Log:
Set target->sbp field to valid value when sbp device is created.
The sbp_cam_detach_target can be called from sbp_post_explore function
on the first target that is not
On Thu, 13 Feb 2014 10:11:27 +
David Chisnall wrote:
> On 13 Feb 2014, at 01:04, Alexander Kabaev wrote:
>
> > The refusal to use tools that are there precisely to help to help
> > with the binary compatibility in favor of mindless library bumps is
> > just sad
t standards non-compliance (std::pair having an
> > explicit constructor) causing real problems.
>
> I see. I am not a big fan of library version bump but standard
> compliance is also very important for us, too. Hmm... :-(
>
The refusal to use tools that are there precisely to
On Mon, 3 Feb 2014 18:54:46 +
David Chisnall wrote:
> On 3 Feb 2014, at 18:32, Alexander Kabaev wrote:
>
> > More than likely. It does appear libc++ does not go through same
> > pains to maintain ABI stable as libstdc++ does. The lack of all and
> > any symbol ve
ith ABI stability, they simply can't enforce
it at the moment even if they wanted to.
--
Alexander Kabaev
signature.asc
Description: PGP signature
1 - 100 of 158 matches
Mail list logo
