From: Michal Hocko <mhocko@kernel.org> To: LKML <linux-kernel@vger.kernel.org> Cc: Peter Zijlstra <peterz@infradead.org>, Ingo Molnar <mingo@redhat.com>, Thomas Gleixner <tglx@linutronix.de>, "H. Peter Anvin" <hpa@zytor.com>, "David S. Miller" <davem@davemloft.net>, Tony Luck <tony.luck@intel.com>, Andrew Morton <akpm@linux-foundation.org>, Chris Zankel <chris@zankel.net>, Max Filippov <jcmvbkbc@gmail.com>, x86@kernel.org, linux-alpha@vger.kernel.org, linux-ia64@vger.kernel.org, linux-s390@vger.kernel.org, linux-sh@vger.kernel.org, sparclinux@vger.kernel.org, linux-xtensa@linux-xtensa.org, linux-arch@vger.kernel.org, Michal Hocko <mhocko@suse.com> Subject: [RFC 10/12] x86, rwsem: simplify __down_write Date: Tue, 02 Feb 2016 20:19:27 +0000 [thread overview] Message-ID: <1454444369-2146-11-git-send-email-mhocko@kernel.org> (raw) In-Reply-To: <1454444369-2146-1-git-send-email-mhocko@kernel.org> From: Michal Hocko <mhocko@suse.com> x86 implementation of __down_write is using inline asm to optimize the code flow. This however requires that it has go over an additional hop for the slow path call_rwsem_down_write_failed which has to save_common_regs/restore_common_regs to preserve the calling convention. This, however doesn't add much because the fast path only saves one register push/pop (rdx) when compared to the generic implementation: Before: 0000000000000019 <down_write>: 19: e8 00 00 00 00 callq 1e <down_write+0x5> 1e: 55 push %rbp 1f: 48 ba 01 00 00 00 ff movabs $0xffffffff00000001,%rdx 26: ff ff ff 29: 48 89 f8 mov %rdi,%rax 2c: 48 89 e5 mov %rsp,%rbp 2f: f0 48 0f c1 10 lock xadd %rdx,(%rax) 34: 85 d2 test %edx,%edx 36: 74 05 je 3d <down_write+0x24> 38: e8 00 00 00 00 callq 3d <down_write+0x24> 3d: 65 48 8b 04 25 00 00 mov %gs:0x0,%rax 44: 00 00 46: 5d pop %rbp 47: 48 89 47 38 mov %rax,0x38(%rdi) 4b: c3 retq After: 0000000000000019 <down_write>: 19: e8 00 00 00 00 callq 1e <down_write+0x5> 1e: 55 push %rbp 1f: 48 b8 01 00 00 00 ff movabs $0xffffffff00000001,%rax 26: ff ff ff 29: 48 89 e5 mov %rsp,%rbp 2c: 53 push %rbx 2d: 48 89 fb mov %rdi,%rbx 30: f0 48 0f c1 07 lock xadd %rax,(%rdi) 35: 48 85 c0 test %rax,%rax 38: 74 05 je 3f <down_write+0x26> 3a: e8 00 00 00 00 callq 3f <down_write+0x26> 3f: 65 48 8b 04 25 00 00 mov %gs:0x0,%rax 46: 00 00 48: 48 89 43 38 mov %rax,0x38(%rbx) 4c: 5b pop %rbx 4d: 5d pop %rbp 4e: c3 retq This doesn't seem to justify the code obfuscation and complexity. Use the generic implementation instead. Signed-off-by: Michal Hocko <mhocko@suse.com> --- arch/x86/include/asm/rwsem.h | 17 +++++------------ arch/x86/lib/rwsem.S | 9 --------- 2 files changed, 5 insertions(+), 21 deletions(-) diff --git a/arch/x86/include/asm/rwsem.h b/arch/x86/include/asm/rwsem.h index d79a218675bc..1b5e89b3643d 100644 --- a/arch/x86/include/asm/rwsem.h +++ b/arch/x86/include/asm/rwsem.h @@ -102,18 +102,11 @@ static inline int __down_read_trylock(struct rw_semaphore *sem) static inline void __down_write(struct rw_semaphore *sem) { long tmp; - asm volatile("# beginning down_write\n\t" - LOCK_PREFIX " xadd %1,(%2)\n\t" - /* adds 0xffff0001, returns the old value */ - " test " __ASM_SEL(%w1,%k1) "," __ASM_SEL(%w1,%k1) "\n\t" - /* was the active mask 0 before? */ - " jz 1f\n" - " call call_rwsem_down_write_failed\n" - "1:\n" - "# ending down_write" - : "+m" (sem->count), "=d" (tmp) - : "a" (sem), "1" (RWSEM_ACTIVE_WRITE_BIAS) - : "memory", "cc"); + + tmp = atomic_long_add_return(RWSEM_ACTIVE_WRITE_BIAS, + (atomic_long_t *)&sem->count); + if (unlikely(tmp != RWSEM_ACTIVE_WRITE_BIAS)) + rwsem_down_write_failed(sem); } /* diff --git a/arch/x86/lib/rwsem.S b/arch/x86/lib/rwsem.S index 40027db99140..ea5c7c177483 100644 --- a/arch/x86/lib/rwsem.S +++ b/arch/x86/lib/rwsem.S @@ -57,7 +57,6 @@ * is also the input argument to these helpers) * * The following can clobber %rdx because the asm clobbers it: - * call_rwsem_down_write_failed * call_rwsem_wake * but %rdi, %rsi, %rcx, %r8-r11 always need saving. */ @@ -93,14 +92,6 @@ ENTRY(call_rwsem_down_read_failed) ret ENDPROC(call_rwsem_down_read_failed) -ENTRY(call_rwsem_down_write_failed) - save_common_regs - movq %rax,%rdi - call rwsem_down_write_failed - restore_common_regs - ret -ENDPROC(call_rwsem_down_write_failed) - ENTRY(call_rwsem_wake) /* do nothing if still outstanding active readers */ __ASM_HALF_SIZE(dec) %__ASM_HALF_REG(dx) -- 2.7.0
WARNING: multiple messages have this Message-ID (diff)
From: Michal Hocko <mhocko@kernel.org> To: LKML <linux-kernel@vger.kernel.org> Cc: Peter Zijlstra <peterz@infradead.org>, Ingo Molnar <mingo@redhat.com>, Thomas Gleixner <tglx@linutronix.de>, "H. Peter Anvin" <hpa@zytor.com>, "David S. Miller" <davem@davemloft.net>, Tony Luck <tony.luck@intel.com>, Andrew Morton <akpm@linux-foundation.org>, Chris Zankel <chris@zankel.net>, Max Filippov <jcmvbkbc@gmail.com>, x86@kernel.org, linux-alpha@vger.kernel.org, linux-ia64@vger.kernel.org, linux-s390@vger.kernel.org, linux-sh@vger.kernel.org, sparclinux@vger.kernel.org, linux-xtensa@linux-xtensa.org, linux-arch@vger.kernel.org, Michal Hocko <mhocko@suse.com> Subject: [RFC 10/12] x86, rwsem: simplify __down_write Date: Tue, 2 Feb 2016 21:19:27 +0100 [thread overview] Message-ID: <1454444369-2146-11-git-send-email-mhocko@kernel.org> (raw) In-Reply-To: <1454444369-2146-1-git-send-email-mhocko@kernel.org> From: Michal Hocko <mhocko@suse.com> x86 implementation of __down_write is using inline asm to optimize the code flow. This however requires that it has go over an additional hop for the slow path call_rwsem_down_write_failed which has to save_common_regs/restore_common_regs to preserve the calling convention. This, however doesn't add much because the fast path only saves one register push/pop (rdx) when compared to the generic implementation: Before: 0000000000000019 <down_write>: 19: e8 00 00 00 00 callq 1e <down_write+0x5> 1e: 55 push %rbp 1f: 48 ba 01 00 00 00 ff movabs $0xffffffff00000001,%rdx 26: ff ff ff 29: 48 89 f8 mov %rdi,%rax 2c: 48 89 e5 mov %rsp,%rbp 2f: f0 48 0f c1 10 lock xadd %rdx,(%rax) 34: 85 d2 test %edx,%edx 36: 74 05 je 3d <down_write+0x24> 38: e8 00 00 00 00 callq 3d <down_write+0x24> 3d: 65 48 8b 04 25 00 00 mov %gs:0x0,%rax 44: 00 00 46: 5d pop %rbp 47: 48 89 47 38 mov %rax,0x38(%rdi) 4b: c3 retq After: 0000000000000019 <down_write>: 19: e8 00 00 00 00 callq 1e <down_write+0x5> 1e: 55 push %rbp 1f: 48 b8 01 00 00 00 ff movabs $0xffffffff00000001,%rax 26: ff ff ff 29: 48 89 e5 mov %rsp,%rbp 2c: 53 push %rbx 2d: 48 89 fb mov %rdi,%rbx 30: f0 48 0f c1 07 lock xadd %rax,(%rdi) 35: 48 85 c0 test %rax,%rax 38: 74 05 je 3f <down_write+0x26> 3a: e8 00 00 00 00 callq 3f <down_write+0x26> 3f: 65 48 8b 04 25 00 00 mov %gs:0x0,%rax 46: 00 00 48: 48 89 43 38 mov %rax,0x38(%rbx) 4c: 5b pop %rbx 4d: 5d pop %rbp 4e: c3 retq This doesn't seem to justify the code obfuscation and complexity. Use the generic implementation instead. Signed-off-by: Michal Hocko <mhocko@suse.com> --- arch/x86/include/asm/rwsem.h | 17 +++++------------ arch/x86/lib/rwsem.S | 9 --------- 2 files changed, 5 insertions(+), 21 deletions(-) diff --git a/arch/x86/include/asm/rwsem.h b/arch/x86/include/asm/rwsem.h index d79a218675bc..1b5e89b3643d 100644 --- a/arch/x86/include/asm/rwsem.h +++ b/arch/x86/include/asm/rwsem.h @@ -102,18 +102,11 @@ static inline int __down_read_trylock(struct rw_semaphore *sem) static inline void __down_write(struct rw_semaphore *sem) { long tmp; - asm volatile("# beginning down_write\n\t" - LOCK_PREFIX " xadd %1,(%2)\n\t" - /* adds 0xffff0001, returns the old value */ - " test " __ASM_SEL(%w1,%k1) "," __ASM_SEL(%w1,%k1) "\n\t" - /* was the active mask 0 before? */ - " jz 1f\n" - " call call_rwsem_down_write_failed\n" - "1:\n" - "# ending down_write" - : "+m" (sem->count), "=d" (tmp) - : "a" (sem), "1" (RWSEM_ACTIVE_WRITE_BIAS) - : "memory", "cc"); + + tmp = atomic_long_add_return(RWSEM_ACTIVE_WRITE_BIAS, + (atomic_long_t *)&sem->count); + if (unlikely(tmp != RWSEM_ACTIVE_WRITE_BIAS)) + rwsem_down_write_failed(sem); } /* diff --git a/arch/x86/lib/rwsem.S b/arch/x86/lib/rwsem.S index 40027db99140..ea5c7c177483 100644 --- a/arch/x86/lib/rwsem.S +++ b/arch/x86/lib/rwsem.S @@ -57,7 +57,6 @@ * is also the input argument to these helpers) * * The following can clobber %rdx because the asm clobbers it: - * call_rwsem_down_write_failed * call_rwsem_wake * but %rdi, %rsi, %rcx, %r8-r11 always need saving. */ @@ -93,14 +92,6 @@ ENTRY(call_rwsem_down_read_failed) ret ENDPROC(call_rwsem_down_read_failed) -ENTRY(call_rwsem_down_write_failed) - save_common_regs - movq %rax,%rdi - call rwsem_down_write_failed - restore_common_regs - ret -ENDPROC(call_rwsem_down_write_failed) - ENTRY(call_rwsem_wake) /* do nothing if still outstanding active readers */ __ASM_HALF_SIZE(dec) %__ASM_HALF_REG(dx) -- 2.7.0
next prev parent reply other threads:[~2016-02-02 20:19 UTC|newest] Thread overview: 64+ messages / expand[flat|nested] mbox.gz Atom feed top 2016-02-02 20:19 [RFC 0/12] introduce down_write_killable for rw_semaphore Michal Hocko 2016-02-02 20:19 ` Michal Hocko 2016-02-02 20:19 ` [RFC 01/12] locking, rwsem: get rid of __down_write_nested Michal Hocko 2016-02-02 20:19 ` Michal Hocko 2016-02-02 20:19 ` [RFC 02/12] locking, rwsem: drop explicit memory barriers Michal Hocko 2016-02-02 20:19 ` Michal Hocko 2016-02-02 20:19 ` [RFC 03/12] locking, rwsem: introduce basis for down_write_killable Michal Hocko 2016-02-02 20:19 ` Michal Hocko 2016-02-02 20:19 ` [RFC 04/12] alpha, rwsem: provide __down_write_killable Michal Hocko 2016-02-02 20:19 ` Michal Hocko 2016-02-02 20:19 ` [RFC 05/12] ia64, " Michal Hocko 2016-02-02 20:19 ` Michal Hocko 2016-02-02 20:19 ` [RFC 06/12] s390, " Michal Hocko 2016-02-02 20:19 ` Michal Hocko 2016-02-02 20:19 ` [RFC 07/12] sh, " Michal Hocko 2016-02-02 20:19 ` Michal Hocko 2016-02-03 11:19 ` Sergei Shtylyov 2016-02-03 11:19 ` Sergei Shtylyov 2016-02-03 12:11 ` Michal Hocko 2016-02-03 12:11 ` Michal Hocko 2016-02-02 20:19 ` [RFC 08/12] sparc, " Michal Hocko 2016-02-02 20:19 ` Michal Hocko 2016-02-02 20:19 ` [RFC 09/12] xtensa, " Michal Hocko 2016-02-02 20:19 ` Michal Hocko 2016-02-02 20:19 ` Michal Hocko [this message] 2016-02-02 20:19 ` [RFC 10/12] x86, rwsem: simplify __down_write Michal Hocko 2016-02-03 8:10 ` Ingo Molnar 2016-02-03 8:10 ` Ingo Molnar 2016-02-03 12:10 ` Michal Hocko 2016-02-03 12:10 ` Michal Hocko 2016-06-03 16:13 ` Peter Zijlstra 2016-06-03 16:13 ` Peter Zijlstra 2016-06-03 22:34 ` Peter Zijlstra 2016-06-03 22:34 ` Peter Zijlstra 2016-06-09 14:40 ` David Howells 2016-06-09 14:40 ` David Howells 2016-06-09 17:36 ` Peter Zijlstra 2016-06-09 17:36 ` Peter Zijlstra 2016-06-10 16:39 ` Paul E. McKenney 2016-06-10 16:39 ` Paul E. McKenney 2016-02-02 20:19 ` [RFC 11/12] x86, rwsem: provide __down_write_killable Michal Hocko 2016-02-02 20:19 ` Michal Hocko 2016-02-17 16:41 ` [RFC 11/12 v1] " Michal Hocko 2016-02-17 16:41 ` Michal Hocko 2016-02-17 16:41 ` Michal Hocko 2016-02-17 16:41 ` Michal Hocko 2016-02-02 20:19 ` [RFC 12/12] locking, rwsem: provide down_write_killable Michal Hocko 2016-02-02 20:19 ` Michal Hocko 2016-02-19 12:15 ` [RFC 0/12] introduce down_write_killable for rw_semaphore Michal Hocko 2016-02-19 12:15 ` Michal Hocko 2016-03-09 12:18 ` Ingo Molnar 2016-03-09 12:18 ` Ingo Molnar 2016-03-09 12:56 ` Michal Hocko 2016-03-09 12:56 ` Michal Hocko 2016-03-09 13:17 ` Ingo Molnar 2016-03-09 13:17 ` Ingo Molnar 2016-03-09 13:28 ` Michal Hocko 2016-03-09 13:28 ` Michal Hocko 2016-03-09 13:43 ` Ingo Molnar 2016-03-09 13:43 ` Ingo Molnar 2016-03-09 14:41 ` Michal Hocko 2016-03-09 14:41 ` Michal Hocko 2016-03-10 10:24 ` Ingo Molnar 2016-03-10 10:24 ` Ingo Molnar
Reply instructions: You may reply publicly to this message via plain-text email using any one of the following methods: * Save the following mbox file, import it into your mail client, and reply-to-all from there: mbox Avoid top-posting and favor interleaved quoting: https://en.wikipedia.org/wiki/Posting_style#Interleaved_style * Reply using the --to, --cc, and --in-reply-to switches of git-send-email(1): git send-email \ --in-reply-to=1454444369-2146-11-git-send-email-mhocko@kernel.org \ --to=mhocko@kernel.org \ --cc=akpm@linux-foundation.org \ --cc=chris@zankel.net \ --cc=davem@davemloft.net \ --cc=hpa@zytor.com \ --cc=jcmvbkbc@gmail.com \ --cc=linux-alpha@vger.kernel.org \ --cc=linux-arch@vger.kernel.org \ --cc=linux-ia64@vger.kernel.org \ --cc=linux-kernel@vger.kernel.org \ --cc=linux-s390@vger.kernel.org \ --cc=linux-sh@vger.kernel.org \ --cc=linux-xtensa@linux-xtensa.org \ --cc=mhocko@suse.com \ --cc=mingo@redhat.com \ --cc=peterz@infradead.org \ --cc=sparclinux@vger.kernel.org \ --cc=tglx@linutronix.de \ --cc=tony.luck@intel.com \ --cc=x86@kernel.org \ /path/to/YOUR_REPLY https://kernel.org/pub/software/scm/git/docs/git-send-email.html * If your mail client supports setting the In-Reply-To header via mailto: links, try the mailto: linkBe sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes, see mirroring instructions on how to clone and mirror all data and code used by this external index.