[Webkit-unassigned] [Bug 195025] MemoryPressureHandlerGeneric should use bmalloc when available

bugzilla-daemon at webkit.org bugzilla-daemon at webkit.org
Mon Feb 25 16:37:47 PST 2019


https://bugs.webkit.org/show_bug.cgi?id=195025

Ross Kirsling <ross.kirsling at sony.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |Basuke.Suzuki at sony.com,
                   |                            |don.olmstead at sony.com,
                   |                            |yoshiaki.jitsukawa at sony.com
                   |                            |, ysuzuki at apple.com

--- Comment #1 from Ross Kirsling <ross.kirsling at sony.com> ---
Looking at this code is making my head spin, but the main question seems to be whether we need a second timer or whether we can just implement MemoryFootprintGeneric (so that it gets called from the "cross-platform" 30-second timer at https://github.com/WebKit/webkit/blob/master/Source/WTF/wtf/MemoryPressureHandler.cpp#L63).

Windows has a second timer due to checking two different things for some reason (calculating the size of the "private working set" vs. calling QueryMemoryResourceNotification), but if we only need to check one thing (PerProcess<Scavenger>::get()->footprint()), then it seems like we wouldn't need a second timer. But then the first timer is only actually activated from here, so it's unclear whether this is really what we're looking for or not:
https://github.com/WebKit/webkit/blob/master/Source/WebKit/WebProcess/WebProcess.cpp#L307

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.webkit.org/pipermail/webkit-unassigned/attachments/20190226/19c810f1/attachment.html>


More information about the webkit-unassigned mailing list