Re: [PATCH] x86/fpu: Fix FPU register read access to the current task

Ingo Molnar wrote: > Good catch, this shows a bug in the new FPU code. > > Could you check whether the fix below solves the problem? Yes, it certainly does. Thanks. Tested-By: Bobby Powers yours, Bobby > > > From 47f01e8cc23f3d041f6b9fb97627369eaf75ba7

Re: [PATCH 000/208] big x86 FPU code rewrite

Hello, Ingo Molnar wrote: > Please have a look. I've been running this for ~ 2 weeks. I've only seen one issue, when emerging mesa 10.5.6: [May26 20:41] traps: aclocal-1.15[27452] trap invalid opcode ip:7f6331031ab0 sp:7ffe73ece880 error:0 in libperl.so.5.20.2[7f6330f18000+19e000] [ +0.51

[tip:x86/urgent] x86/fpu: Always restore_xinit_state() when use_eager_cpu()

Commit-ID: c88d47480d300eaad80c213d50c9bf6077fc49bc Gitweb: http://git.kernel.org/tip/c88d47480d300eaad80c213d50c9bf6077fc49bc Author: Bobby Powers AuthorDate: Mon, 27 Apr 2015 08:10:41 -0700 Committer: Ingo Molnar CommitDate: Wed, 6 May 2015 11:22:03 +0200 x86/fpu: Always

Re: [PATCH v2] x86/fpu: always restore_xinit_state() when use_eager_cpu()

015 at 11:19 AM, Oleg Nesterov wrote: > Thanks! > > On 04/27, Bobby Powers wrote: >> >> v2: switch used_math() -> tsk_used_math(tsk) to consistently use the >> grabbed tsk instead of current, like in the rest of flush_thread(). >> >> Oleg's commit

[tip:perf/urgent] tools lib api: Undefine _FORTIFY_SOURCE before setting it

Commit-ID: de28c15daf60e9625bece22f13a091fac8d05f1d Gitweb: http://git.kernel.org/tip/de28c15daf60e9625bece22f13a091fac8d05f1d Author: Bobby Powers AuthorDate: Tue, 21 Apr 2015 19:19:41 -0400 Committer: Arnaldo Carvalho de Melo CommitDate: Thu, 23 Apr 2015 17:08:23 -0300 tools lib api

[PATCH v2] x86/fpu: always restore_xinit_state() when use_eager_cpu()

ue. drop_init_fpu (now fpu_reset_state) calls restore_init_xstate() regardless of whether current used_math() - apply the same logic here. Fixes: f893959b ("x86/fpu: Don't abuse drop_init_fpu() in flush_thread()") Signed-off-by: Bobby Powers Acked-by: Oleg Nesterov Cc: Andi Kleen

Re: [PATCH] x86/fpu: always restore_xinit_state() when !use_eager_cpu()

Hello, On 4/27, Oleg Nesterov wrote: > ACK, but could you please fix the subject and update the changelog? > > The subject says "when !use_eager_cpu()", but we need to do this > if use_eager_cpu() == T. Whoops, now its my turn to not know what I was thinking. I will update the subject/changelog,

[PATCH] x86/fpu: always restore_xinit_state() when !use_eager_cpu()

. drop_init_fpu (now fpu_reset_state) calls restore_init_xstate() regardless of whether current used_math(). There was a lot of commentary on the initial patch - not sure I understand it all. Happy to get some pointers or be pointed to a better fix. Signed-off-by: Bobby Powers Cc: Oleg Nesterov Cc: Boris

Re: Crypto Update for 4.1

100644 > --- a/arch/x86/crypto/sha512-avx2-asm.S > +++ b/arch/x86/crypto/sha512-avx2-asm.S > @@ -79,7 +79,7 @@ NUM_BLKS= %rdx > c = %rcx > d = %r8 > e = %rdx > -y3 = %rdi > +y3 = %rsi > > TBL = %rbp > Teste

[PATCH v2] tools/lib/api: undefine _FORTIFY_SOURCE before setting it

avoid this, undefine _FORTIFY_SOURCE before (possibly re-)defining it in tools/lib/api. v2 applies cleanly on top of already pulled kbuild changes for 4.1-rc1. Signed-off-by: Bobby Powers Acked-by: Jiri Olsa --- tools/lib/api/Makefile | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) di

Re: [PATCH] tools/lib/api: undefine _FORTIFY_SOURCE before setting it

+CC linux-kbuild. Related to this Dirk's work [1]. On Sat, Apr 18, 2015 at 11:05 AM, Jiri Olsa wrote: > On Sat, Apr 18, 2015 at 10:46:20AM -0400, Bobby Powers wrote: >> Some toolchains (like Hardened Gentoo) define _FORTIFY_SOURCE in the >> built-in, default args. This c

Re: [PATCH] tools/lib/api/Makefile: Add feature check for _FORTIFY_SOURCE

Hi, Dirk Gouders wrote: > Yes, I was suggesting something similar (but without founded reasoning), > some time ago [1]. I submitted a patch for this a few days ago, but I didn't realize I should CC linux-kbuild@ (my bad): https://lkml.org/lkml/2015/4/18/71 yours, Bobby -- To unsubscribe from th

[PATCH] tools/lib/api: undefine _FORTIFY_SOURCE before setting it

avoid this, undefine _FORTIFY_SOURCE before (possibly re-)defining it in tools/lib/api. Signed-off-by: Bobby Powers --- tools/lib/api/Makefile | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/lib/api/Makefile b/tools/lib/api/Makefile index 36c08b1..87c72d1 100644 --- a/too

Re: Drop support for x86-32

On Fri, Aug 24, 2012 at 9:17 AM, wbrana wrote: > On 8/24/12, Martin Nybo Andersen wrote: >> Ahh..., so the development time saved by not supporting x86-32 in mainline >> can >> now be used by backporting new features to the forementioned long term >> tree? > new features won't be backported Did