Bug 155993 - bmalloc: page size should be configurable at runtime
Summary: bmalloc: page size should be configurable at runtime
Status: RESOLVED FIXED
Alias: None
Product: WebKit
Classification: Unclassified
Component: bmalloc (show other bugs)
Version: WebKit Nightly Build
Hardware: Unspecified Unspecified
: P2 Normal
Assignee: Geoffrey Garen
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2016-03-29 16:09 PDT by Geoffrey Garen
Modified: 2016-03-29 19:34 PDT (History)
5 users (show)

See Also:


Attachments
Patch (20.48 KB, patch)
2016-03-29 16:26 PDT, Geoffrey Garen
kling: review+
Details | Formatted Diff | Diff

Note You need to log in before you can comment on or make changes to this bug.
Description Geoffrey Garen 2016-03-29 16:09:38 PDT
bmalloc: page size should be configurable at runtime
Comment 1 Geoffrey Garen 2016-03-29 16:26:38 PDT
Created attachment 275146 [details]
Patch
Comment 2 Andreas Kling 2016-03-29 18:29:54 PDT
Comment on attachment 275146 [details]
Patch

r=me
Comment 3 Geoffrey Garen 2016-03-29 19:22:43 PDT
Committed r198821: <http://trac.webkit.org/changeset/198821>
Comment 4 Gavin Barraclough 2016-03-29 19:29:00 PDT
Comment on attachment 275146 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=275146&action=review

> Source/bmalloc/bmalloc/VMAllocate.h:51
> +    return sysconf(_SC_PAGESIZE);

Would it be worth caching the result here? - or are all interesting uses via Heap?
Comment 5 Gavin Barraclough 2016-03-29 19:29:32 PDT
r=me & Ellie Barraclough too.
Comment 6 Geoffrey Garen 2016-03-29 19:33:44 PDT
> > Source/bmalloc/bmalloc/VMAllocate.h:51
> > +    return sysconf(_SC_PAGESIZE);
> 
> Would it be worth caching the result here? - or are all interesting uses via
> Heap?

Hmmm... At first I didn't want to because threads. But I guess we can assume that writing a size_t is atomic.
Comment 7 Geoffrey Garen 2016-03-29 19:34:05 PDT
> r=me & Ellie Barraclough too.

\0/