[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20120222154206.GA3405@redhat.com>
Date: Wed, 22 Feb 2012 10:42:07 -0500
From: Jason Baron <jbaron@...hat.com>
To: Steven Rostedt <rostedt@...dmis.org>
Cc: Ingo Molnar <mingo@...e.hu>, "H. Peter Anvin" <hpa@...or.com>,
a.p.zijlstra@...llo.nl, mathieu.desnoyers@...icios.com,
davem@...emloft.net, ddaney.cavm@...il.com,
akpm@...ux-foundation.org, linux-kernel@...r.kernel.org,
Linus Torvalds <torvalds@...ux-foundation.org>
Subject: Re: [PATCH 00/10] jump label: introduce very_[un]likely + cleanups +
docs
On Wed, Feb 22, 2012 at 10:12:13AM -0500, Steven Rostedt wrote:
> On Wed, 2012-02-22 at 15:56 +0100, Ingo Molnar wrote:
>
> > > Because it really just looks like a stronger "unlikely()" and
> > > fundamentally it really isn't. [...]
> >
> > Well, the fact is that right now it *is* a stronger unlikely()
> > on architectures that have jump-labels and it's mapped to
> > unlikely() on others.
> >
>
> Has gcc been fix to make it truly an unlikely case and remove the "jmp;
> jmp" problem of before? I'm still using gcc 4.6.0 which has the
> following code for a tracepoint (example is the
> trace_sched_migrate_task().
>
> 5b4a: e9 00 00 00 00 jmpq 5b4f <set_task_cpu+0x5e>
>
> The above is the jump label that turns into a nop at boot up.
>
> 5b4f: eb 19 jmp 5b6a <set_task_cpu+0x79>
>
> Here we jump over some of the trace code (this is the fast path)
>
> 5b51: 49 8b 7d 08 mov 0x8(%r13),%rdi
> 5b55: 44 89 e2 mov %r12d,%edx
> 5b58: 48 89 de mov %rbx,%rsi
> 5b5b: 41 ff 55 00 callq *0x0(%r13)
> 5b5f: 49 83 c5 10 add $0x10,%r13
> 5b63: 49 83 7d 00 00 cmpq $0x0,0x0(%r13)
> 5b68: eb 41 jmp 5bab <set_task_cpu+0xba>
>
> Below is the continuation of the fast path.
>
> 5b6a: 48 8b 43 08 mov 0x8(%rbx),%rax
> 5b6e: 44 39 60 18 cmp %r12d,0x18(%rax)
> 5b72: 74 0c je 5b80 <set_task_cpu+0x8f>
>
>
> Again, I'm using gcc 4.6.0 and maybe it has been fixed.
>
> -- Steve
>
>
Hi Steve,
Using 4.6.2, I don't see the 'jmp;jmp' issue. I have:
ffffffff810610c0 <set_task_cpu>:
ffffffff810610c0: 55 push %rbp
ffffffff810610c1: 48 89 e5 mov %rsp,%rbp
ffffffff810610c4: 48 81 ec e0 00 00 00 sub $0xe0,%rsp
ffffffff810610cb: 48 89 5d d8 mov %rbx,-0x28(%rbp)
ffffffff810610cf: 4c 89 65 e0 mov %r12,-0x20(%rbp)
ffffffff810610d3: 48 89 fb mov %rdi,%rbx
ffffffff810610d6: 4c 89 6d e8 mov %r13,-0x18(%rbp)
ffffffff810610da: 4c 89 75 f0 mov %r14,-0x10(%rbp)
ffffffff810610de: 41 89 f4 mov %esi,%r12d
ffffffff810610e1: 4c 89 7d f8 mov %r15,-0x8(%rbp)
no double jump here.
ffffffff810610e5: e9 00 00 00 00 jmpq ffffffff810610ea <set_task_cpu+0x2a>
ffffffff810610ea: 48 8b 43 08 mov 0x8(%rbx),%rax
ffffffff810610ee: 44 3b 60 18 cmp 0x18(%rax),%r12d
ffffffff810610f2: 74 0d je ffffffff81061101 <set_task_cpu+0x41>
ffffffff810610f4: 48 83 83 a8 00 00 00 addq $0x1,0xa8(%rbx)
ffffffff810610fb: 01
no double jump here.
ffffffff810610fc: e9 00 00 00 00 jmpq ffffffff81061101 <set_task_cpu+0x41>
ffffffff81061101: 48 8b 83 60 06 00 00 mov 0x660(%rbx),%rax
ffffffff81061108: 48 8b 50 40 mov 0x40(%rax),%rdx
ffffffff8106110c: 44 89 e0 mov %r12d,%eax
ffffffff8106110f: 48 8b 4a 28 mov 0x28(%rdx),%rcx
ffffffff81061113: 48 8b 0c c1 mov (%rcx,%rax,8),%rcx
ffffffff81061117: 48 89 8b 90 01 00 00 mov %rcx,0x190(%rbx)
ffffffff8106111e: 48 8b 52 20 mov 0x20(%rdx),%rdx
ffffffff81061122: 48 8b 04 c2 mov (%rdx,%rax,8),%rax
ffffffff81061126: 48 89 83 88 01 00 00 mov %rax,0x188(%rbx)
ffffffff8106112d: 48 8b 43 08 mov 0x8(%rbx),%rax
ffffffff81061131: 44 89 60 18 mov %r12d,0x18(%rax)
ffffffff81061135: 48 8b 5d d8 mov -0x28(%rbp),%rbx
ffffffff81061139: 4c 8b 65 e0 mov -0x20(%rbp),%r12
ffffffff8106113d: 4c 8b 6d e8 mov -0x18(%rbp),%r13
ffffffff81061141: 4c 8b 75 f0 mov -0x10(%rbp),%r14
ffffffff81061145: 4c 8b 7d f8 mov -0x8(%rbp),%r15
ffffffff81061149: c9 leaveq
ffffffff8106114a: c3 retq
....
Do you have 'CONFIG_CC_OPTIMIZE_FOR_SIZE' set?
Thanks,
-Jason
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists