Closed Bug 503141 Opened 16 years ago Closed 15 years ago

well-traced peacekeeper array benchmarks are gated by allocation and js_BoxDouble

Categories

(Core :: JavaScript Engine, defect)

x86
macOS
defect
Not set
normal

Tracking

()

RESOLVED FIXED

People

(Reporter: sayrer, Unassigned)

References

(Blocks 1 open bug)

Details

Attachments

(3 files)

Some of the ones we do really well on are gated by malloc/js_NewGCThing/js_NewUninitializedArray
Attached file peacekeeper array-pop testcase (deleted) —
Blocks: peacekeeper
I did some perf analysis of this on the flight back. I will write up my results. I have a bunch of patches, but a lot more work remains to be done.
Depends on: 503155
Depends on: 503156
Depends on: 503157
No longer depends on: 503155
Summary: well-traced peacekeeper array benchmarks are gated by allocation → well-traced peacekeeper array benchmarks are gated by allocation and js_BoxDouble
Depends on: 503160
No longer depends on: 503156
Depends on: 503156
Attached patch patch (deleted) — Splinter Review
Comment on attachment 387524 [details] [diff] [review] patch this is a good quick proof. I see about 10% faster on my box. Cleaned up patch should go in 503160.
/me still wishes for widen and narrow instead of promote and demote /be
This should be fixed by the fatval stuff, right?
(In reply to comment #7) > This should be fixed by the fatval stuff, right? I don't think so. Having said that, looks like this went in with bug 503160, so I'll close. Please reopen if you disagree.
Status: NEW → RESOLVED
Closed: 15 years ago
Resolution: --- → FIXED
(In reply to comment #8) > (In reply to comment #7) > > This should be fixed by the fatval stuff, right? > > I don't think so. Having said that, looks like this went in with bug 503160, > so I'll close. Please reopen if you disagree. For the record, I just tested with fatvals and saw no change.
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Creator:
Created:
Updated:
Size: