On 2016-01-15 7:44 PM, Trevor Saunders wrote:
On Fri, Jan 15, 2016 at 04:28:13PM -0800, Jonas Sicking wrote:
On Fri, Jan 15, 2016 at 11:41 AM, Joshua Cranmer 🐧 <pidgeo...@gmail.com> wrote:
On 1/15/2016 1:21 PM, Bobby Holley wrote:
Has anyone measured recently whether there's still a significant perf win
to making IIDs 32-bit? If we stop using them as a versioning tool, we
could
potentially relax our uniqueness requirements, and save a lot of
comparisons on each QI. Addon-compat would be tricky, but is potentially
solvable.
Are we still using nsISupports in a way that we expect it to be
ABI-compatible with IUnknown?
Last I checked (which was years ago), the accessibility code still
did. More specifically it implement MSCOM and XPCOM interfaces on the
same object.
Which is why we're forced to use stdcall and thus why we're using the
NS_IMETHOD* macros.
See also https://bugzilla.mozilla.org/show_bug.cgi?id=662348
The accessibility code still does that. we've moved away from it some,
but completely fixing it has never been really high priority. I've never
been totally clear on how that over loading manages to work, but I
wonder if changing to nsresult QueryInterface(int, void**) would keep
the two methods separate enough that it would be ok.
Hmm, it looks like there are already separate overloads now... See:
<https://dxr.mozilla.org/mozilla-central/source/accessible/windows/msaa/AccessibleWrap.cpp#88>
<https://dxr.mozilla.org/mozilla-central/source/accessible/windows/msaa/AccessibleWrap.cpp#110>
So maybe this is no longer an issue!
_______________________________________________
dev-platform mailing list
dev-platform@lists.mozilla.org
https://lists.mozilla.org/listinfo/dev-platform
