'auto' tracks now also gets an item's min/max-content contribution
(in 2.2/2.3) when sizing under a min/max-content constraint.
(Previously, 'auto' only got the min-size contribution under 2.1)
https://drafts.csswg.org/css-grid/#algo-content
Now it returns the min/max-content contribution for intrinsic sizing.
Otherwise, i.e. during Reflow, return the min-size contribution.
(Previously, it always returned the min-size contribution.)
https://drafts.csswg.org/css-grid/#algo-content
DETAILS ON THE CLEANUP:
* In nsFrameTraversal.cpp:
- Whitespace fix.
- Remove a useless nullptr assignment.
- Use "=" to initialize a nsCOMPtr, per coding style guideline ("initialize variables with nsFoo aFoo = bFoo and not nsFoo aFoo(bFoo)").
* In nsPresShell.cpp:
- Wrap a longish assignment to 2 lines (since next patch will make the first line longer, which would push the whole thing over 80 characters if it weren't wrapped).
- Add braces around "if" body.
* In nsTextFrame.cpp:
- Remove a useless nullptr assignment.
- Move the "metrics->GetThebesFontGroup()" call slightly earlier, so it happens *before* we've implicitly transferred ownership the object pointed to by "metrics".
- Add a code-comment.
MozReview-Commit-ID: LG6vgOmM9MK
--HG--
extra : rebase_source : 7754d4edefb09ae7992275b9dcb7e84371955d9c
The operator new is infallable. Also change the return value to void
accordingly. The only caller doesn't even check the return value.
MozReview-Commit-ID: 3whf7s1d35q
--HG--
extra : rebase_source : 6c3a7010f77dad0ee5f50408205fc3a5418dacc9
To see the flag in action, export GECKO_BLOCK_DEBUG_FLAGS="float-manager" and
open a page containing float elements. For example:
$ GECKO_BLOCK_DEBUG_FLAGS="float-manager" ./mach run layout/reftests/floats/orthogonal-floats-1a.html
MozReview-Commit-ID: 7Rb45ophewf
--HG--
extra : rebase_source : f40703a2ec448bf39a4a49e98fcd50f5550bbdd3
Adds a hidden layout.show_previous_page pref for non-e10s that, when set to false,
will blank out the content area in between page loads instead of showing the
previous page.
MozReview-Commit-ID: 1EyDk4IabeK
--HG--
extra : rebase_source : efd4c745b2c249b1f57faedd8a3d5080c7445893
Per bug 1299753 comment 20, SetFlags() is designed for flag propagation
to continuations during block reflow. To avoid misuse, I expand
SetFlags() directly in the only reasonable caller nsBlackFrame::Init(),
and replace other usages by AddStateBits().
MozReview-Commit-ID: GsbE2Z0Rps1
--HG--
extra : rebase_source : 72a64e9218870d638f67d1b586f533cd7d16c491
This takes the second hunk of https://hg.mozilla.org/mozilla-central/rev/82c3b4b81d82 only, but instead of the "in displayport" list we are adding to the "approximately visible" list. The "in displayport" list doesn't exist after the backouts that need to be done.
This ensures that all visible frames are in the approximately visible list.
By not taking the first chunk of the aforementioned changeset we don't remove frames that aren't painted from the approximately visible list. So the list only grows during a paint. But the next time the we do a full update of approximately visible images it can shrink.
We need this because the video element uses visibility to start/stop decoding (bug 1282710, bug 1299065) and we want frames that are painted to for sure be marked as visible.
nsCSSFrameConstructor::ConstructNonScrollableBlock() has logic to
determine whether to create a block formatting context for a block
frame. I refactor the function to make it reusable by
nsCSSFrameConstructor::ConstructDetailsFrame().
Also, make NS_NewBlockFrame() accept two arguments as other frame
factory functions so that it could be pointed by BlockFrameCreationFunc.
NS_NewBlockFormattingContext is changed accordingly.
The construction for a scrollable DetailsFrame will be further revised
in Part 3.
MozReview-Commit-ID: 8TwG9YMyGva
--HG--
extra : rebase_source : fffdd974df81a809a607491d2534aa8dd2d13ab1