[Webkit-unassigned] [Bug 160497] MemoryPressureHandler shouldn't know how to release WebCore memory

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=160497

Michael Catanzaro <mcatanzaro at igalia.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
 Attachment #291638|review?                     |review+
              Flags|                            |

--- Comment #25 from Michael Catanzaro <mcatanzaro at igalia.com> ---
Comment on attachment 291638
  --> https://bugs.webkit.org/attachment.cgi?id=291638
Fix the build

View in context: https://bugs.webkit.org/attachment.cgi?id=291638&action=review

Andreas liked the initial version of the patch (see comment #2), and I think not much has changed; I think we can count that as owner approval if there are not any objections. It looks like a good fix for these layering violations. A couple questions:

> Source/WebCore/platform/MemoryPressureHandler.cpp:62
> +    if (!m_lowMemoryHandler)
> +        return;

Wouldn't an early return here be a bug? Shouldn't there be an assertion here?

> Source/WebCore/platform/MemoryPressureHandler.h:173
> +    void (^m_releaseMemoryBlock)() { nullptr };

I know this was here before, but... what is this???

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.webkit.org/pipermail/webkit-unassigned/attachments/20161101/00e5fedb/attachment.html>