[tip:x86/fpu] x86: Avoid unnecessary __clear_user() and xrstor in signal handling
in [Kernel]
|
Prev: 2.6.35-rc4 - include/linux/fdtable.h:85 invoked rcu_dereference_check() without protection!
Next: [PATCH]shmem: reduce one time of locking in pagefault
From: tip-bot for Suresh Siddha on 6 Jul 2010 20:40 Commit-ID: 8e221b6db4477643fefc885a97ea9889ac733140 Gitweb: http://git.kernel.org/tip/8e221b6db4477643fefc885a97ea9889ac733140 Author: Suresh Siddha <suresh.b.siddha(a)intel.com> AuthorDate: Tue, 22 Jun 2010 16:23:37 -0700 Committer: H. Peter Anvin <hpa(a)linux.intel.com> CommitDate: Tue, 6 Jul 2010 16:31:04 -0700 x86: Avoid unnecessary __clear_user() and xrstor in signal handling fxsave/xsave doesn't touch all the bytes in the memory layout used by these instructions. Specifically SW reserved (bytes 464..511) fields in the fxsave frame and the reserved fields in the xsave header. To present a clean context for the signal handling, just clear these fields instead of clearing the complete fxsave/xsave memory layout, when we dump these registers directly to the user signal frame. Also avoid the call to second xrstor (which inits the state not passed in the signal frame) in restore_user_xstate() if all the state has already been restored by the first xrstor. These changes improve the performance of signal handling(by ~3-5% as measured by the lat_sig). Signed-off-by: Suresh Siddha <suresh.b.siddha(a)intel.com> LKML-Reference: <1277249017.2847.85.camel(a)sbs-t61.sc.intel.com> Signed-off-by: H. Peter Anvin <hpa(a)linux.intel.com> --- arch/x86/include/asm/i387.h | 9 +++++++++ arch/x86/include/asm/xsave.h | 10 ++++++++++ arch/x86/kernel/xsave.c | 12 ++---------- 3 files changed, 21 insertions(+), 10 deletions(-) diff --git a/arch/x86/include/asm/i387.h b/arch/x86/include/asm/i387.h index c991b3a..0f1cf5d 100644 --- a/arch/x86/include/asm/i387.h +++ b/arch/x86/include/asm/i387.h @@ -127,6 +127,15 @@ static inline int fxsave_user(struct i387_fxsave_struct __user *fx) { int err; + /* + * Clear the bytes not touched by the fxsave and reserved + * for the SW usage. + */ + err = __clear_user(&fx->sw_reserved, + sizeof(struct _fpx_sw_bytes)); + if (unlikely(err)) + return -EFAULT; + asm volatile("1: rex64/fxsave (%[fx])\n\t" "2:\n" ".section .fixup,\"ax\"\n" diff --git a/arch/x86/include/asm/xsave.h b/arch/x86/include/asm/xsave.h index 2c4390c..30dfc81 100644 --- a/arch/x86/include/asm/xsave.h +++ b/arch/x86/include/asm/xsave.h @@ -59,6 +59,16 @@ static inline int fpu_xrstor_checking(struct fpu *fpu) static inline int xsave_user(struct xsave_struct __user *buf) { int err; + + /* + * Clear the xsave header first, so that reserved fields are + * initialized to zero. + */ + err = __clear_user(&buf->xsave_hdr, + sizeof(struct xsave_hdr_struct)); + if (unlikely(err)) + return -EFAULT; + __asm__ __volatile__("1: .byte " REX_PREFIX "0x0f,0xae,0x27\n" "2:\n" ".section .fixup,\"ax\"\n" diff --git a/arch/x86/kernel/xsave.c b/arch/x86/kernel/xsave.c index 37e68fc..6e73db1 100644 --- a/arch/x86/kernel/xsave.c +++ b/arch/x86/kernel/xsave.c @@ -91,14 +91,6 @@ int save_i387_xstate(void __user *buf) return 0; if (task_thread_info(tsk)->status & TS_USEDFPU) { - /* - * Start with clearing the user buffer. This will present a - * clean context for the bytes not touched by the fxsave/xsave. - */ - err = __clear_user(buf, sig_xstate_size); - if (err) - return err; - if (use_xsave()) err = xsave_user(buf); else @@ -184,8 +176,8 @@ static int restore_user_xstate(void __user *buf) * init the state skipped by the user. */ mask = pcntxt_mask & ~mask; - - xrstor_state(init_xstate_buf, mask); + if (unlikely(mask)) + xrstor_state(init_xstate_buf, mask); return 0; -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo(a)vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/ |