Bug 123585
| Summary: | run-javascriptcore-tests is failing on Windows | ||
|---|---|---|---|
| Product: | WebKit | Reporter: | Ryosuke Niwa <rniwa> |
| Component: | Tools / Tests | Assignee: | Nobody <webkit-unassigned> |
| Status: | RESOLVED FIXED | ||
| Severity: | Normal | CC: | bfulgham, fpizlo, ggaren, mark.lam, oliver, roger_fong, webkit-bug-importer |
| Priority: | P1 | ||
| Version: | 528+ (Nightly build) | ||
| Hardware: | Unspecified | ||
| OS: | Unspecified | ||
| Bug Depends on: | 121972 | ||
| Bug Blocks: | |||
Ryosuke Niwa
e.g.
http://build.webkit.org/builders/Apple%20Win%207%20Release%20%28Tests%29/builds/39846
http://build.webkit.org/builders/Apple%20Win%207%20Release%20%28Tests%29/builds/39846/steps/jscore-test/logs/stdio
using PTY: False
program finished with exit code 254
elapsedTime=1.591203
| Attachments | ||
|---|---|---|
| Add attachment proposed patch, testcase, etc. |
Brent Fulgham
Are any of the JSC guys looking at these failures?
Geoffrey Garen
The first step here is to build locally and get the output from the failing run. The buildbot doesn't seem to have it.
Brent Fulgham
It looks like the issue is that testapi.exe fails immediately, stopping the tests from running at all.
If I run "run-javascriptcore-tests --debug --no-build --no-testapi" I get a clean run:
Results for Mozilla tests:
0 regressions found.
0 tests fixed.
OK.
Filip Pizlo
(In reply to comment #3)
> It looks like the issue is that testapi.exe fails immediately, stopping the tests from running at all.
>
> If I run "run-javascriptcore-tests --debug --no-build --no-testapi" I get a clean run:
>
> Results for Mozilla tests:
> 0 regressions found.
> 0 tests fixed.
> OK.
OK, though it would be great to port the jsc-stress-tests to Windows also. The Mozilla tests don't really exercise the system. jsc-stress-tests do, and they might be able to tell you more about why testapi is failing. They'll probably reveal other bugs that aren't currently covered by layout tests, mozilla tests, or anything else.
Porting run-jsc-stress-tests just involves some Windows script foo.
Brent Fulgham
(In reply to comment #4)
> (In reply to comment #3)
> > It looks like the issue is that testapi.exe fails immediately, stopping the tests from running at all.
> >
> > If I run "run-javascriptcore-tests --debug --no-build --no-testapi" I get a clean run:
> >
> > Results for Mozilla tests:
> > 0 regressions found.
> > 0 tests fixed.
> > OK.
>
> OK, though it would be great to port the jsc-stress-tests to Windows also. The Mozilla tests don't really exercise the system. jsc-stress-tests do, and they might be able to tell you more about why testapi is failing. They'll probably reveal other bugs that aren't currently covered by layout tests, mozilla tests, or anything else.
>
> Porting run-jsc-stress-tests just involves some Windows script foo.
I'll see about getting the stress tests running under https://bugs.webkit.org/show_bug.cgi?id=124801.
Brent Fulgham
These are now working properly. Closing.