Bug 155014

Summary: [css-grid] Initial support for implicit grid before explicit grid
Product: WebKit Reporter: Manuel Rego Casasnovas <rego>
Component: Layout and RenderingAssignee: Manuel Rego Casasnovas <rego>
Status: RESOLVED FIXED    
Severity: Normal CC: commit-queue, jfernandez, simon.fraser, svillar
Priority: P2    
Version: WebKit Nightly Build   
Hardware: Unspecified   
OS: Unspecified   
Bug Depends on:    
Bug Blocks: 153488, 155199    
Attachments:
Description Flags
Patch
none
Patch
none
Patch none

Manuel Rego Casasnovas
Reported 2016-03-04 04:02:26 PST
Next step is to port the following patch from Blink: https://codereview.chromium.org/1529083006/ Main change is the change in GridSpan to start storing ints instead of unsigned.
Attachments
Patch (55.02 KB, patch)
2016-03-07 15:19 PST, Manuel Rego Casasnovas
no flags
Patch (54.96 KB, patch)
2016-03-08 14:18 PST, Manuel Rego Casasnovas
no flags
Patch (55.21 KB, patch)
2016-03-09 01:04 PST, Manuel Rego Casasnovas
no flags
Manuel Rego Casasnovas
Comment 1 2016-03-07 15:19:05 PST
Darin Adler
Comment 2 2016-03-08 09:21:52 PST
Comment on attachment 273225 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=273225&action=review > Source/WebCore/rendering/RenderGrid.cpp:1167 > + coordinate.rows.translate(abs(m_smallestRowStart)); Please use std::abs, which won’t silently convert a floating point value, or a uint32_t or size_t, to an int, rather than abs, which despite its generic name *will* do those things. For example, abs(4294967295U) returns 1, while std::abs(4294967295U) gives us a compile time error because of overload ambiguity. And abs(-1.1) returns 1, while std::abs(-1.1) returns 1.1. Also note that the return value of abs and std::abs can be negative if it’s passed the most negative integer. Not sure if this is an issue for the correctness of our code. > Source/WebCore/rendering/RenderGrid.cpp:1169 > + coordinate.columns.translate(abs(m_smallestColumnStart)); Ditto. > Source/WebCore/rendering/RenderGrid.cpp:1237 > + m_grid.grow(maximumRowIndex + abs(m_smallestRowStart)); Ditto. > Source/WebCore/rendering/RenderGrid.cpp:1239 > + column.grow(maximumColumnIndex + abs(m_smallestColumnStart)); Ditto. > Source/WebCore/rendering/style/GridCoordinate.h:168 > + if (resolvedInitialPosition >= 0) > + m_resolvedInitialPosition = std::min(resolvedInitialPosition, kGridMaxTracks - 1); > + else > + m_resolvedInitialPosition = std::max(resolvedInitialPosition, -kGridMaxTracks); I normally prefer to write this in a way that brackets the value with the minimum and maximum values. More elegant than comparing with zero, I think. m_resolvedInitialPosition = std::max(-kGridMaxTracks, std::min(resolvedInitialPosition, kGridMaxTracks - 1)); > Source/WebCore/rendering/style/GridCoordinate.h:172 > + if (resolvedFinalPosition >= 0) > + m_resolvedFinalPosition = std::min(resolvedFinalPosition, kGridMaxTracks); > + else > + m_resolvedFinalPosition = std::max(resolvedFinalPosition, -kGridMaxTracks + 1); Ditto. > Source/WebCore/rendering/style/GridResolvedPosition.cpp:253 > unsigned resolvedPosition = abs(position.integerPosition()) - 1; Same issue with abs rather than std::abs.
Manuel Rego Casasnovas
Comment 3 2016-03-08 14:18:32 PST
Manuel Rego Casasnovas
Comment 4 2016-03-08 14:22:18 PST
Thanks for the review. (In reply to comment #2) > Comment on attachment 273225 [details] > Patch > > View in context: > https://bugs.webkit.org/attachment.cgi?id=273225&action=review > > > Source/WebCore/rendering/RenderGrid.cpp:1167 > > + coordinate.rows.translate(abs(m_smallestRowStart)); > > Please use std::abs, which won’t silently convert a floating point value, or > a uint32_t or size_t, to an int, rather than abs, which despite its generic > name *will* do those things. > > For example, abs(4294967295U) returns 1, while std::abs(4294967295U) gives > us a compile time error because of overload ambiguity. And abs(-1.1) returns > 1, while std::abs(-1.1) returns 1.1. > > Also note that the return value of abs and std::abs can be negative if it’s > passed the most negative integer. Not sure if this is an issue for the > correctness of our code. Thanks for the explanation, I've changed it to std::abs() in all the different places. I think we're safe regarding the most negative integer as we're setting the limit in -kGridMaxTracks. > > Source/WebCore/rendering/style/GridCoordinate.h:168 > > + if (resolvedInitialPosition >= 0) > > + m_resolvedInitialPosition = std::min(resolvedInitialPosition, kGridMaxTracks - 1); > > + else > > + m_resolvedInitialPosition = std::max(resolvedInitialPosition, -kGridMaxTracks); > > I normally prefer to write this in a way that brackets the value with the > minimum and maximum values. More elegant than comparing with zero, I think. > > m_resolvedInitialPosition = std::max(-kGridMaxTracks, > std::min(resolvedInitialPosition, kGridMaxTracks - 1)); Done.
WebKit Commit Bot
Comment 5 2016-03-08 15:11:53 PST
Comment on attachment 273340 [details] Patch Rejecting attachment 273340 [details] from commit-queue. Failed to run "['/Volumes/Data/EWS/WebKit/Tools/Scripts/webkit-patch', '--status-host=webkit-queues.webkit.org', '--bot-id=webkit-cq-02', 'build', '--no-clean', '--no-update', '--build-style=release', '--port=mac']" exit_code: 2 cwd: /Volumes/Data/EWS/WebKit Last 500 characters of output: emMac.dia -c /Volumes/Data/EWS/WebKit/Source/WebCore/history/mac/HistoryItemMac.mm -o /Volumes/Data/EWS/WebKit/WebKitBuild/WebCore.build/Release/WebCore.build/Objects-normal/x86_64/HistoryItemMac.o ** BUILD FAILED ** The following build commands failed: CompileC /Volumes/Data/EWS/WebKit/WebKitBuild/WebCore.build/Release/WebCore.build/Objects-normal/x86_64/GridResolvedPosition.o rendering/style/GridResolvedPosition.cpp normal x86_64 c++ com.apple.compilers.llvm.clang.1_0.compiler (1 failure) Full output: http://webkit-queues.webkit.org/results/943725
Manuel Rego Casasnovas
Comment 6 2016-03-09 01:04:57 PST
WebKit Commit Bot
Comment 7 2016-03-09 02:15:13 PST
Comment on attachment 273410 [details] Patch Clearing flags on attachment: 273410 Committed r197850: <http://trac.webkit.org/changeset/197850>
WebKit Commit Bot
Comment 8 2016-03-09 02:15:17 PST
All reviewed patches have been landed. Closing bug.
Note You need to log in before you can comment on or make changes to this bug.