Like part a, but for `choices` messages rather than error messages.
MozReview-Commit-ID: 7dJ0NL2fUh5
--HG--
extra : rebase_source : 477f1364c0904bde78d54eae083bdb8e49ee5732
extra : histedit_source : 38c336b3a59481b6f2523798367159fb757c6485
For choices types, when one choice fails, we don't need the original error
string, since another choice may succeed, and we generate the final error
based on all of the options. Nevertheless, we spend a lot of time generating
JSON strings for the failed inputs in those cases, which adds up to about 12%
of the remaining overhead at this point.
MozReview-Commit-ID: 6nXBAv2W20V
--HG--
extra : rebase_source : 5894bc4b9e8d64ac9505f27240ea4fabfcb5f02f
extra : histedit_source : 0e8b5e0315abd672a57a60420453a1e0681c9df6
The Array and ArrayBuffer type checks we do in getBaseType add up to a
significant amount of overhead given the number of times we call them,
especially when X-ray overhead comes into play. These changes allow us to
avoid X-ray overhead altogether.
MozReview-Commit-ID: KlRuxeElIfp
--HG--
extra : rebase_source : c7f00fb8c35965476e7c7b888b6af36714c1323f
extra : histedit_source : fc559e665e60e9bbb688eebe6c6e6da5dacec748
This prevents naming conflicts if a paused or blocked download is retried from the original page.
MozReview-Commit-ID: 4rFZ5rP8saJ
--HG--
extra : rebase_source : c392fc62cb33b2c2d70a1c3a8a975ddf93d394ea
Version 4 of the Google Safe Browsing server will return a 404 if any of the
application reputation lists are requested on Android. As a result, we should
avoid these threat types from being sent along with ANDROID_PLATFORM.
MozReview-Commit-ID: 6TUBVxe455y
--HG--
extra : rebase_source : dee095c008f4d328f359c66d20f0cc2dfcd109f3
extra : source : 5d6338807c6b21b7672236cac01b13bd75155225
The FrameLoaderOwner interface has been implemented in WebIDL for several
years now, so these QIs are simply unnecessary overhead.
MozReview-Commit-ID: LAzvfm5Qhy0
--HG--
extra : rebase_source : 2495c07df21c474f5fabc257ff4db43b0d8047e4
In chrome process, we often know which url is going to be loaded.
As a performance optimization, we can start initiating network
connection before sending out the 'LoadURI' message to the content
process.
MozReview-Commit-ID: L79ylHOaxX8
--HG--
extra : rebase_source : cc3aeefedde21dc8eaaf37c67710980b0204f4dc
This is straightforward, with only two notable things.
- `#include "nsXPIDLString.h" is replaced with `#include "nsString.h"`
throughout, because all nsXPIDLString.h did was include nsString.h. The
exception is for files which already include nsString.h, in which case the
patch just removes the nsXPIDLString.h inclusion.
- The patch removes the |xpidl_string| gtest, but improves the |voided| test to
cover some of its ground, e.g. testing Adopt(nullptr).
--HG--
extra : rebase_source : 452cc4a08046a1adb1a8099a7e85a1917de5add8
We should not be declaring forward declarations for nsString classes directly,
instead we should use nsStringFwd.h. This will make changing the underlying
types easier.
--HG--
extra : rebase_source : b2c7554e8632f078167ff2f609392e63a136c299
This patch is mainly to add a probe to measure sw launch time. To do this, this
patch records the sw launch time when the sw is just spwaned and it's ready to
handle the incoming fetch event.
MozReview-Commit-ID: 3w5MNyhQNnd
--HG--
extra : rebase_source : 3228213d0ea6be1d23b9c49382f1f8d3c2f358f1
These are all easy cases where an nsXPIDLCString local variable is set via
getter_Copies() and then is null checked. The patch uses IsVoid() to replace
the null checks (and get() and EqualsLiteral() calls to replace any implicit
conversions).
--HG--
extra : rebase_source : 484ad42a7816b34b86afbe072e04ba131c1619c6