Bug 224817 - FullGCActivityCallback should use the percentage of pages uncompressed in RAM to determine deferral.
Summary: FullGCActivityCallback should use the percentage of pages uncompressed in RAM...
Status: RESOLVED FIXED
Alias: None
Product: WebKit
Classification: Unclassified
Component: New Bugs (show other bugs)
Version: WebKit Nightly Build
Hardware: Unspecified Unspecified
: P2 Normal
Assignee: Keith Miller
URL:
Keywords: InRadar
Depends on:
Blocks:
 
Reported: 2021-04-20 09:17 PDT by Keith Miller
Modified: 2021-04-20 15:42 PDT (History)
11 users (show)

See Also:

Attachments
Patch (10.40 KB, patch)
2021-04-20 09:53 PDT, Keith Miller 		no flags Details | Formatted Diff | Diff
Patch (10.72 KB, patch)
2021-04-20 10:18 PDT, Keith Miller 		no flags Details | Formatted Diff | Diff
Patch (13.32 KB, patch)
2021-04-20 10:57 PDT, Keith Miller 		ews-feeder: commit-queue-
Details | Formatted Diff | Diff
Patch (13.46 KB, patch)
2021-04-20 11:17 PDT, Keith Miller 		no flags Details | Formatted Diff | Diff
Patch (13.53 KB, patch)
2021-04-20 11:37 PDT, Keith Miller 		no flags Details | Formatted Diff | Diff
Patch (13.86 KB, patch)
2021-04-20 11:58 PDT, Keith Miller 		no flags Details | Formatted Diff | Diff
Patch for landing (13.94 KB, patch)
2021-04-20 14:33 PDT, Keith Miller 		no flags Details | Formatted Diff | Diff
Show Obsolete (6) View All 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 Keith Miller 2021-04-20 09:17:50 PDT 
FullGCActivityCallback should use the percentage of pages uncompressed in RAM to determine deferral.
Comment 1 Keith Miller 2021-04-20 09:53:48 PDT 
Created attachment 426561 [details]
Patch
Comment 2 Keith Miller 2021-04-20 10:18:08 PDT 
Created attachment 426563 [details]
Patch
Comment 3 Keith Miller 2021-04-20 10:57:45 PDT 
Created attachment 426569 [details]
Patch
Comment 4 Keith Miller 2021-04-20 11:17:40 PDT 
Created attachment 426571 [details]
Patch
Comment 5 Darin Adler 2021-04-20 11:37:26 PDT 
Comment on attachment 426571 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=426571&action=review

> Source/WTF/ChangeLog:8
> +        Add a noexecpt flavor of FunctionTraits.

Why? Where is noexcept coming into play? Some std namespace function?
Comment 6 Keith Miller 2021-04-20 11:37:52 PDT 
Created attachment 426575 [details]
Patch
Comment 7 Keith Miller 2021-04-20 11:40:30 PDT 
Comment on attachment 426571 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=426571&action=review

>> Source/WTF/ChangeLog:8
>> +        Add a noexecpt flavor of FunctionTraits.
> 
> Why? Where is noexcept coming into play? Some std namespace function?

mincore is marked noexcept on linux apparently, which is part of the reason I was getting build failures there.
Comment 8 Keith Miller 2021-04-20 11:58:40 PDT 
Created attachment 426577 [details]
Patch
Comment 9 Darin Adler 2021-04-20 13:08:56 PDT 
Comment on attachment 426577 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=426577&action=review

> Source/WTF/ChangeLog:8
> +        Add a noexecpt flavor of FunctionTraits.

I suggest you fix the typo here and mention mincore as the reason.
Comment 10 Keith Miller 2021-04-20 13:09:49 PDT 
Comment on attachment 426577 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=426577&action=review

>> Source/WTF/ChangeLog:8
>> +        Add a noexecpt flavor of FunctionTraits.
> 
> I suggest you fix the typo here and mention mincore as the reason.

Sure, will do both before landing.
Comment 11 Keith Miller 2021-04-20 13:10:44 PDT 
rdar://76425456
Comment 12 Keith Miller 2021-04-20 13:13:00 PDT 
Comment on attachment 426577 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=426577&action=review

In a follow up patch I'll make sure we account for PreciseAllocations too.

> Source/JavaScriptCore/heap/BlockDirectory.cpp:80
> +        // FIXME: We could cache this in bulk (e.g. 25 MB chunks) but we haven't seen any data that it actually matters.

Actually, this probably shouldn't be marked as a FIXME since we're not actually going to fix this until we say data saying it matters.
Comment 13 Keith Miller 2021-04-20 14:33:16 PDT 
Created attachment 426593 [details]
Patch for landing
Comment 14 EWS 2021-04-20 15:42:13 PDT 
Committed r276324 (236804@main): <https://commits.webkit.org/236804@main>

All reviewed patches have been landed. Closing bug and clearing flags on attachment 426593 [details].