Closed Bug 453668 Opened 17 years ago Closed 14 years ago

TM: Regressions with JIT enabled for Chrome

Categories

(Core :: JavaScript Engine, defect)

defect
Not set
normal

Tracking

()

RESOLVED WONTFIX

People

(Reporter: gal, Unassigned)

References

Details

(Keywords: meta)

Please track regressions when JIT enabled for Chrome here. This is currently not recommended and has lower priority than polishing the JIT for content.
Depends on: 451986
Depends on: 453644
Blocks: 451602
Keywords: meta
Depends on: 451951
Depends on: 451899
Depends on: 454590
Depends on: 455646
Not a blocker as this is just an investigation.
Depends on: 454669
Depends on: 455173
Depends on: 465247
Depends on: 470223
I'm seeing major memory leakage when I enable JIT for chrome. Is this known or should I file a bug on it?
scratch, file a bug if you can not find an already filed one that matches the leaks you have found.
Depends on: 470030, 481302
Depends on: 481804
Will chrome jit ever be enabled for chrome by default? Post-3.5? Are there real significant perf gains? Just curious.
Fennec reported visible improvements (i.e. during scrolling). Their UI is more JS-heavy than FF as far as I can tell.
To be honest, in the nightly Shiretoko builds, I can't tell a difference with jit.chrome on or off. So I just leave it on until something crashes it.
I don't know much about the FF JS code base but since JS used to be pretty slow I am sure we go to great length avoiding computationally intensive code in the UI. I don't think having the JIT turned on for chrome is a huge priority, but it will definitively shift the cost around and allow to do more things in JS than before for future code.
Obsolete with the removal of tracejit.
Status: NEW → RESOLVED
Closed: 14 years ago
Resolution: --- → WONTFIX
You need to log in before you can comment on or make changes to this bug.