Commit graph

15 commits

Author SHA1 Message Date
Ms2ger
466faac2a5 Update rustc to revision 3dcd2157403163789aaf21a9ab3c4d30a7c6494d. 2014-12-17 15:19:45 -05:00
Jack Moffitt
d1b433a3b3 Rust upgrade to rustc hash b03a2755193cd756583bcf5831cf4545d75ecb8a 2014-11-13 11:17:43 +10:00
Patrick Walton
08fc7c2795 layout: Make incremental reflow more fine-grained by introducing "reflow
out-of-flow" and "reconstruct flow" damage bits.

This is needed for good performance on the maze solver.
2014-10-31 12:24:40 -07:00
Clark Gaebel
a8f80b89f4 layout: Implement flow tree dumping with RUST_LOG=debug is on.
r? @pcwalton
2014-10-28 10:14:12 -07:00
Clark Gaebel
5cd47c7670 Clear reflow flags after reflow. 2014-10-22 17:46:28 -07:00
Clark Gaebel
81bd3cbd9d Fixes the table_percentage_width_a.html reftest with incremental reflow turned on. 2014-10-20 11:13:31 -07:00
Glenn Watson
076495db94 Use opts as a global, to avoid cloning and passing the struct all over the code. 2014-10-20 10:48:47 +10:00
Clark Gaebel
2faf5b270f Fix image_dynamic_remove reftest with incremental layout turned out
This also adds some extra debugging infrastructure which I found useful tracking
this bug down. A regression in the br reftests is also uncovered by this patch,
which I'll work on fixing next.

r? @pcwalton
2014-10-17 11:11:14 -07:00
Clark Gaebel
7368d42225 Removes duplicate CSS selector matching logic.
Now that DOM/Flow traversals have been refactored out, the `recalc_style_for_subtree`
function in `css/matching.rs` can be removed, in lieu of just running the standard
`recalc_style_for_node` and `construct_flows` traversals sequentially. Now we
no longer have the maintenance headache of duplicating selector matching logic
in two places! \o/

r? @pcwalton
2014-10-14 16:28:29 -07:00
bors-servo
56989b8dec auto merge of #3640 : cgaebel/servo/incremental-flow-construction, r=pcwalton
This also hides the not-yet-working parts of incremental reflow behind a runtime
flag. As I get the failing reftests passing, I'll send pull requests for them one
by one.
2014-10-14 16:51:30 -06:00
Patrick Walton
5f8d3f72d8 layout: Introduce support for legacy presentational attributes to selector
matching, and use it for `<input size>` and `<td width>`.

This implements a general framework for legacy presentational attributes
to the DOM and style calculation, so that adding more of them later will
be straightforward.
2014-10-14 12:44:09 -07:00
Clark Gaebel
f552e2f750 try to reset flows which need reflow, since reflow isn't yet idempotent 2014-10-14 10:33:46 -07:00
Patrick Walton
bffaad118e layout: Rewrite clipping to be per-display-item instead of having
a separate `ClipDisplayItem`.

We push down clipping areas during absolute position calculation. This
makes display items into a flat list, improving cache locality. It
dramatically simplifies the code all around.

Because we need to push down clip rects even for absolutely-positioned
children of non-absolutely-positioned flows, this patch alters the
parallel traversal to compute absolute positions for
absolutely-positioned children at the same time it computes absolute
positions for other children. This doesn't seem to break anything either
in theory (since the overall order remains correct) or in practice. It
simplifies the parallel traversal code quite a bit.

See the relevant Gecko bug:
https://bugzilla.mozilla.org/show_bug.cgi?id=615734
2014-10-13 21:40:36 -07:00
Patrick Walton
2a790d06dd Use Gecko's simpler Bloom filter instead of one based on hash
stretching.

This preserves the usage of the Bloom filter throughout style recalc,
but the implementation is rewritten. Provides a 15% improvement on
Guardians of the Galaxy.
2014-10-10 17:02:27 -07:00
Clark Gaebel
24bff2416b Factors out DOM traversal, keeping the code in parallel free of traversal-specific logic.
DOM traversals and Flow traversals look very similar. This patch unifies them
with the preorder/postorder pattern. Hopefully, it also opens the door for writing
the traversal code only once, instead of the duplication we have today.
2014-10-10 14:55:18 -04:00