lkml.org 
[lkml]   [2006]   [Nov]   [5]   [last100]   RSS Feed
Views: [more markup]  [less markup]  [headers]  [forward] 
 
Messages in this thread
/
DateSun, 05 Nov 2006 00:01:57 -0800
FromZachary Amsden <>
SubjectRe: [rfc patch] i386: don't save eflags on task switch
Andi Kleen wrote:
>> For me, when compiled with -O2, it results in
>>
>> 	84
>> 	88
>> 	132
>>
>> which basically says: a "rdtsc->rdtsc" is 84 cycles, putting a "pushfl" in 
>> between is another _4_ cycles, and putting a "popfl" in between is about 
>> another 48 cycles. 
>> 
>
> This means we should definitely change restore_flags() to only STI, 
> never popf
> 

sti is expensive as well; iirc just as expensive on most processors as 
popf, although I don't have hard numbers to back this up on hand.  An 
_unlikely_ jcc + popf? is better than a sti, for sure, but a likely 
jcc+popf could cost more than a jcc+sti, depending on model.

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

\
 
 \ /
  Last update: 2006-11-05 08:05    [from the cache]
©2003-2008