stress/sampling-profiler-deep-stack.js fails on ARM 32bit
authorsbarati@apple.com <sbarati@apple.com@268f45cc-cd09-0410-ab3c-d52691b4dbfc>
Tue, 16 Feb 2016 20:50:46 +0000 (20:50 +0000)
committersbarati@apple.com <sbarati@apple.com@268f45cc-cd09-0410-ab3c-d52691b4dbfc>
Tue, 16 Feb 2016 20:50:46 +0000 (20:50 +0000)
commit7a19551cc48f4e1bec67d5f5d06bcf19e7dc9da3
treec618c617e8c4debb053773146cf7e7fad515e5db
parent3ce3b4f2f9fbd48650fb4b80092f9151094cf419
stress/sampling-profiler-deep-stack.js fails on ARM 32bit
https://bugs.webkit.org/show_bug.cgi?id=154255
<rdar://problem/24662996>

Reviewed by Mark Lam.

The bug here wasn't in the implementation of the sampling profiler
itself. Rather, it was a bug in the test. JSC wasn't spending a lot
of time in a function that the test assumed a lot of time was spent in.
That's because the DFG was doing a good job at optimizing the function
at the leaf of the recursion. Because of that, we often wouldn't sample it.
I fixed this by making the leaf function do more work.

* tests/stress/sampling-profiler-deep-stack.js:
(platformSupportsSamplingProfiler.foo):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@196650 268f45cc-cd09-0410-ab3c-d52691b4dbfc
Source/JavaScriptCore/ChangeLog
Source/JavaScriptCore/tests/stress/sampling-profiler-deep-stack.js