Bug 144915 - stress/int8-repeat-out-of-bounds.js and stress/float32-repeat-out-of-bounds.js fail with eager compilation
Summary: stress/int8-repeat-out-of-bounds.js and stress/float32-repeat-out-of-bounds.j...
Status: RESOLVED INVALID
Alias: None
Product: WebKit
Classification: Unclassified
Component: JavaScriptCore (show other bugs)
Version: 528+ (Nightly build)
Hardware: Unspecified Unspecified
: P2 Minor
Assignee: Nobody
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2015-05-12 10:01 PDT by Basile Clement
Modified: 2015-05-12 12:27 PDT (History)
4 users (show)

See Also:

Attachments
Add an attachment (proposed patch, testcase, etc.)

Note You need to log in before you can comment on or make changes to this bug.
Description Basile Clement 2015-05-12 10:01:40 PDT 
...
Comment 1 Basile Clement 2015-05-12 10:04:10 PDT 
After looking at the test, this is probably a minor problem with the test making assumptions on the number of recompilations that is not correct with eager compilation.
Comment 2 Filip Pizlo 2015-05-12 10:38:56 PDT 
It's intentional. :-). The tests are asserting that we converge quickly and don't recompile too much. But with eager compilation, we will recompile more because we often don't get enough profiling in the first round. That's fine. I don't think we should try to make these tests pass in eager mode
Comment 3 Basile Clement 2015-05-12 12:27:01 PDT 
(In reply to comment #2)
> It's intentional. :-). The tests are asserting that we converge quickly and
> don't recompile too much. But with eager compilation, we will recompile more
> because we often don't get enough profiling in the first round. That's fine.
> I don't think we should try to make these tests pass in eager mode

OK, it just seemed weird that we had tests that would only succeed provided some options are within presupposed range, closing this then.