"USING VLA'S IS ACTIVELY STUPID! It generates much more code, and much
slower code (and more fragile code), than just using a fixed key size
would have done." -- Linus Torvalds
We never use jemalloc without MOZ_MEMORY surrounding code (mostly a
FreeBSD concession which doesn't work anyway). This allows us to get rid
of the "poor man's mutex" (CPU-based spinlock) in mozjemalloc.
This basic validation is hard-set to always-on, so no point in keeping
it conditional since we basically always want this rudimentary pointer
validation to be done.
forming a monotonically increasing sequence (e.g. with position:sticky)
Relative positioning can cause table parts to move, which can cause
issues with the cursor position to know which rows can be skipped.
To make this work, use the max difference between the frame's rect and
the union of the frame's "normal" position and the overflow rect to
cover the area of relatively positioned elements even if they are out of
order.
This resolves#1309.
list collisions when processing the background color of a table.
This is effectively #146 Part 3c to go along with the other 2 there.
This resolves#1308.
- Check for undefined/null regex flags (because...)
- Make it throw a typeerror instead of syntax error on non-global
- Generalize JS error messages for these checks.
* The langpack is not generated via the package target but IS for mozpackage target using old behavior
* Add mach command to specifically generate the langpack when needed without the rest of the packaging routine (mach langpack)
* This adds what will eventually be a rewritten packaging routine while maintaining backwards compatibility with the original mozilla routine
* Changes the build target and installer makefile to use the alternative packaging
* Adds build target to specifically invoke the original mozilla routine
* Update mach commands accordingly
I didn't include this in my initial PR because I assumed this part
was not relevant to UXP. I was asked to submit this while the MailNews rebase
was going on, but I was late with my patch and didn't quite grasp that this
specifically was still needed.
My counterpart at Mozilla, Petr Sumbara, has gotten his version of this
approved.
https://bugzilla.mozilla.org/show_bug.cgi?id=1568473
I made a typo in commit 687a789 when updating the ifdef style I used to comply with UXP standards. The typo I made resulted in a compiler warning I failed to notice, so I was asked to tag issue #457 when submitting the PR. I also fixed some trailing whitespace I apparently left behind in the file.
MOZ_ASSERT changed to NS_WARNING. Also correctly handle case where last chunk ends with \r. Tested to make sure that regression identified in Bug 1494764 comment 10 remains fixed and that non-chunked and chunked messages are handled correctly including when \r\n is split between chunks.
Tag #1273
