[Webkit-unassigned] [Bug 113862] Negative zero checks cause unnecessary speculation failures on SunSpider on ARMv7

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=113862


Roman Zhuykov <zhroma at ispras.ru> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
 Attachment #196848|0                           |1
        is obsolete|                            |
 Attachment #197263|0                           |1
        is obsolete|                            |
 Attachment #199746|                            |review?, commit-queue?
               Flag|                            |




--- Comment #10 from Roman Zhuykov <zhroma at ispras.ru>  2013-04-25 13:57:58 PST ---
Created an attachment (id=199746)
 --> (https://bugs.webkit.org/attachment.cgi?id=199746&action=review)
patch with smaller layout tests

(In reply to comment #9)
> (From update of attachment 197263 [details])
> View in context: https://bugs.webkit.org/attachment.cgi?id=197263&action=review
> 
> r=me
> 
> cq- because of the layout test issue. Any committer can fix the layout test and then land, or you can post a new patch, and I can cq+ it.
> 
> > LayoutTests/fast/js/regress/script-tests/negative-zero-divide.js:6
> > +  for (i = 0; i < 100000; i++) {
> 
> Please change this to 100. That's our de facto limit for getting code to run in the DFG, without making regression tests run too long.

Thank you for your review. Unfortunately, "i < 100" is not enough. In script-tests directory there are 95 tests, 71 of them have loops with more than 10000 iterations, 55 tests have more than 100000 loop iterations. I've just tested my negative-zero-negate.js on ARMv7 Linux. The test starts to fail (without my patch) only with "i < 1094" or more loop iterations. Checking with DFG_ENABLE_DEBUG_VERBOSE=1 shows, that there are no DFG compilations when I use "i < 1093" or smaller values. I have created the new patch with "i < 2001". This is fast enough, and give some reserve if DFG-starting parameters would change later. Is patch now OK?

-- 
Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.