Re: eth_type_trans(): Re: [Bug #11308] tbench regression on eachkernel release from 2.6.22 -> 2.6.28

[Linus Torvalds]

On Mon, 17 Nov 2008, Ingo Molnar wrote:
> ffffffff8049e2ae:        0 	0f b7 c0             	movzwl %ax,%eax
> ffffffff8049e2b1:        0 	3d ff 05 00 00       	cmp    $0x5ff,%eax
> ffffffff8049e2b6:      468 	7f 18                	jg     ffffffff8049e2d0 <eth_type_trans+0xbb>
> ffffffff8049e2b8:        0 	48 8b 83 d8 00 00 00 	mov    0xd8(%rbx),%rax
> ffffffff8049e2bf:        0 	b9 00 01 00 00       	mov    $0x100,%ecx
> ffffffff8049e2c4:        0 	66 83 38 ff          	cmpw   $0xffffffffffffffff,(%rax)
> ffffffff8049e2c8:        0 	b8 00 04 00 00       	mov    $0x400,%eax
> ffffffff8049e2cd:        0 	0f 45 c8             	cmovne %eax,%ecx
> ffffffff8049e2d0:        0 	5b                   	pop    %rbx
> ffffffff8049e2d1:    85064 	5d                   	pop    %rbp
> ffffffff8049e2d2:    63776 	41 5c                	pop    %r12
> ffffffff8049e2d4:        1 	89 c8                	mov    %ecx,%eax
> ffffffff8049e2d6:      474 	c3                   	retq   
> 
> small function, big bang - 1.7% of the total overhead.
> 
> 90% of this function's cost is in the closing sequence. My guess would 
> be that it originates from ffffffff8049e2ae (the branch after that is 
> not taken), which corresponds to this source code context:

I would actually suspect that branch mispredicts may be an issue.

If that thing falls out of the branch prediction table (which it could 
easily do), then a forward branch will be predicted as "not taken". And if 
it then turns out that the _common_ case is the other way around, the 
incorrectly predicted destination is often the one that shows up in 
profiles.

Giving gcc likely()/unlikely() hints usually doesn't much help, I'm 
afraid. It _can_ make a difference, but often not for -Os in particular.

		Linus
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/