[Webkit-unassigned] [Bug 154964] New: Slow JSC stress tests fail on 32 bit debug bots

bugzilla-daemon at webkit.org bugzilla-daemon at webkit.org
Thu Mar 3 03:29:42 PST 2016


            Bug ID: 154964
           Summary: Slow JSC stress tests fail on 32 bit debug bots
    Classification: Unclassified
           Product: WebKit
           Version: Other
          Hardware: Unspecified
                OS: Unspecified
            Status: NEW
          Severity: Critical
          Priority: P1
         Component: New Bugs
          Assignee: webkit-unassigned at lists.webkit.org
          Reporter: ossy at webkit.org
                CC: ap at webkit.org, fpizlo at apple.com, rniwa at webkit.org
            Blocks: 154858

http://trac.webkit.org/changeset/197442 disabled FTL JIT testing 
on 32 bit bots, because there is no FTL JIT on any 32 bit platforms.

But 25-35 tests started to fail on these bots:
- https://build.webkit.org/builders/Apple%20El%20Capitan%2032-bit%20JSC%20%28BuildAndTest%29/builds/1658
- https://build.webkit.org/builders/Apple%20Yosemite%2032-bit%20JSC%20%28BuildAndTest%29

The problem is that these tests seems to be very slow in debug mode,
but they didn't fail previously because jsc-stress-tests script
duplicated the timeout in case of FTL JIT is enabled:

if $enableFTL and ENV["JSCTEST_timeout"]
    # Currently, using the FTL is a performance regression particularly in real
    # (i.e. non-loopy) benchmarks. Account for this in the timeout.
    ENV["JSCTEST_timeout"] = (ENV["JSCTEST_timeout"].to_i * 2).to_s

It seems we should duplicate timeout in debug mode too.

You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.webkit.org/pipermail/webkit-unassigned/attachments/20160303/0184ed0b/attachment.html>

More information about the webkit-unassigned mailing list