[Webkit-unassigned] [Bug 213307] REGRESSION(r263035): stress/get-prototype-of.js broken on s390x
bugzilla-daemon at webkit.org
bugzilla-daemon at webkit.org
Tue Jun 23 13:53:29 PDT 2020
https://bugs.webkit.org/show_bug.cgi?id=213307
Michael Catanzaro <mcatanzaro at gnome.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |shvaikalesh at gmail.com
See Also| |https://bugs.webkit.org/sho
| |w_bug.cgi?id=157972
Summary|REGRESSION(r263018-r263135) |REGRESSION(r263035):
|: |stress/get-prototype-of.js
|stress/get-prototype-of.js |broken on s390x
|broken on s390x |
--- Comment #2 from Michael Catanzaro <mcatanzaro at gnome.org> ---
OK, well turns out there is no need to bisect it: the problem is that the test is brand new. :P Should have checked for that first. So, from the backtrace, we can see the failing check is this one:
assert(getPrototypeOf(proxy) === proxyPrototype);
And here we reach the limit of my nonexistent JSC knowledge: I cannot begin to guess why getPrototypeOf(proxy) is not returning the expected result.k A few things I do know: s390x uses cloop only, so we can eliminate the entire JIT as a source of problems. (The tests are run with --no-jit.) It also uses 4 KB page size, which is nice as that eliminates potential for trouble there. The only major difference from x86_64/aarch64 likely to introduce software bugs is that s390x is big-endian, so it's likely that the bug is an endianness problem... somewhere. All other JSC stress tests are passing except for this new one (and memory-limited tests, since we run tests with --memory-limited).
--
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.webkit.org/pipermail/webkit-unassigned/attachments/20200623/489098d6/attachment-0001.htm>
More information about the webkit-unassigned
mailing list