[webkit-changes] [239370] trunk
Title: [239370] trunk
Revision 239370
Author ross.kirsl...@sony.com
Date 2018-12-18 22:37:46 -0800 (Tue, 18 Dec 2018)
Log Message
Error message for `-x ** y` contains a typo.
https://bugs.webkit.org/show_bug.cgi?id=192832
Reviewed by Saam Barati.
JSTests:
* ChakraCore/test/UnitTestFramework/UnitTestFramework.js:
(assert.assert.return.throws):
* stress/pow-expects-update-_expression_-on-lhs.js:
(throw.new.Error):
Update test expectations which match against the exact error message.
Source/_javascript_Core:
* parser/Parser.cpp:
(JSC::Parser::parseBinaryExpression):
Make corrections to error message: "amiguous" -> "ambiguous", "parenthesis" -> "parentheses"
Modified Paths
trunk/JSTests/ChakraCore/test/UnitTestFramework/UnitTestFramework.js
trunk/JSTests/ChangeLog
trunk/JSTests/stress/pow-expects-update-_expression_-on-lhs.js
trunk/Source/_javascript_Core/ChangeLog
trunk/Source/_javascript_Core/parser/Parser.cpp
Diff
Modified: trunk/JSTests/ChakraCore/test/UnitTestFramework/UnitTestFramework.js (239369 => 239370)
--- trunk/JSTests/ChakraCore/test/UnitTestFramework/UnitTestFramework.js 2018-12-19 06:10:28 UTC (rev 239369)
+++ trunk/JSTests/ChakraCore/test/UnitTestFramework/UnitTestFramework.js 2018-12-19 06:37:46 UTC (rev 239370)
@@ -366,7 +366,7 @@
},
{
regexp: /Invalid unary operator on the left hand side of exponentiation \(\*\*\) operator/,
-replStr: "Unexpected token '**'. Amiguous unary _expression_ in the left hand side of the exponentiation _expression_; parenthesis must be used to disambiguate the _expression_."
+replStr: "Unexpected token '**'. Ambiguous unary _expression_ in the left hand side of the exponentiation _expression_; parentheses must be used to disambiguate the _expression_."
},
{
regexp: /Assignment to read-only properties is not allowed in strict mode/,
Modified: trunk/JSTests/ChangeLog (239369 => 239370)
--- trunk/JSTests/ChangeLog 2018-12-19 06:10:28 UTC (rev 239369)
+++ trunk/JSTests/ChangeLog 2018-12-19 06:37:46 UTC (rev 239370)
@@ -1,3 +1,16 @@
+2018-12-18 Ross Kirsling
+
+Error message for `-x ** y` contains a typo.
+https://bugs.webkit.org/show_bug.cgi?id=192832
+
+Reviewed by Saam Barati.
+
+* ChakraCore/test/UnitTestFramework/UnitTestFramework.js:
+(assert.assert.return.throws):
+* stress/pow-expects-update-_expression_-on-lhs.js:
+(throw.new.Error):
+Update test expectations which match against the exact error message.
+
2018-12-18 Mark Lam
Gardening: test options fix.
Modified: trunk/JSTests/stress/pow-expects-update-_expression_-on-lhs.js (239369 => 239370)
--- trunk/JSTests/stress/pow-expects-update-_expression_-on-lhs.js 2018-12-19 06:10:28 UTC (rev 239369)
+++ trunk/JSTests/stress/pow-expects-update-_expression_-on-lhs.js 2018-12-19 06:37:46 UTC (rev 239370)
@@ -38,7 +38,7 @@
{
return ${token} a ** b;
}
-`, `SyntaxError: Unexpected token '**'. Amiguous unary _expression_ in the left hand side of the exponentiation _expression_; parenthesis must be used to disambiguate the _expression_.`);
+`, `SyntaxError: Unexpected token '**'. Ambiguous unary _expression_ in the left hand side of the exponentiation _expression_; parentheses must be used to disambiguate the _expression_.`);
}
}
Modified: trunk/Source/_javascript_Core/ChangeLog (239369 => 239370)
--- trunk/Source/_javascript_Core/ChangeLog 2018-12-19 06:10:28 UTC (rev 239369)
+++ trunk/Source/_javascript_Core/ChangeLog 2018-12-19 06:37:46 UTC (rev 239370)
@@ -1,3 +1,14 @@
+2018-12-18 Ross Kirsling
+
+Error message for `-x ** y` contains a typo.
+https://bugs.webkit.org/show_bug.cgi?id=192832
+
+Reviewed by Saam Barati.
+
+* parser/Parser.cpp:
+(JSC::Parser::parseBinaryExpression):
+Make corrections to error message: "amiguous" -> "ambiguous", "parenthesis" -> "parentheses"
+
2018-12-18 Saam Barati
Update ARM64EHash
Modified: trunk/Source/_javascript_Core/parser/Parser.cpp (239369 => 239370)
--- trunk/Source/_javascript_Core/parser/Parser.cpp 2018-12-19 06:10:28 UTC (rev 239369)
+++ trunk/Source/_javascript_Core/parser/Parser.cpp 2018-12-19 06:37:46 UTC (rev 239370)
@@ -3886,7 +3886,7 @@
// Check this.
// If the binary operator <> has higher precedence than one of "**", this check does not work.
// But it's OK for ** because the operator "**" has the highest operator precedence in the binary operators.
-failIfTrue(match(POW) && isUnaryOpExcludingUpdateOp(leadingTokenTypeForUnaryExpression), "Amiguous unary _expression_ in the left hand side of the exponentiation _expression_; parenthesis must be used to disambiguate the _expression_");
+failIfTrue(match(POW) &&
[webkit-changes] [239369] trunk
Title: [239369] trunk
Revision 239369
Author wenson_hs...@apple.com
Date 2018-12-18 22:10:28 -0800 (Tue, 18 Dec 2018)
Log Message
[iOS] A copied text selection is pasted as a web archive attachment in the entry view in Messages
https://bugs.webkit.org/show_bug.cgi?id=192842
Reviewed by Tim Horton.
Source/WebCore:
Temporarily revert a behavior change introduced by r238661, where we now add "com.apple.webarchive" as a
registered UTI when dragging or copying a text selection. This broke the Messages app on iOS, which currently
inserts a copied or dragged text selection from WebKit-based views as a web archive file attachment. A fix for
this is internally tracked in .
* platform/ios/PlatformPasteboardIOS.mm:
(WebCore::PlatformPasteboard::write):
Tools:
Adjusts an existing API test to verify that "com.apple.webarchive" is not one of the types registered to the
item provider when dragging a rich text selection.
* TestWebKitAPI/Tests/ios/DragAndDropTestsIOS.mm:
Modified Paths
trunk/Source/WebCore/ChangeLog
trunk/Source/WebCore/platform/ios/PlatformPasteboardIOS.mm
trunk/Tools/ChangeLog
trunk/Tools/TestWebKitAPI/Tests/ios/DragAndDropTestsIOS.mm
Diff
Modified: trunk/Source/WebCore/ChangeLog (239368 => 239369)
--- trunk/Source/WebCore/ChangeLog 2018-12-19 05:31:11 UTC (rev 239368)
+++ trunk/Source/WebCore/ChangeLog 2018-12-19 06:10:28 UTC (rev 239369)
@@ -1,3 +1,19 @@
+2018-12-18 Wenson Hsieh
+
+[iOS] A copied text selection is pasted as a web archive attachment in the entry view in Messages
+https://bugs.webkit.org/show_bug.cgi?id=192842
+
+
+Reviewed by Tim Horton.
+
+Temporarily revert a behavior change introduced by r238661, where we now add "com.apple.webarchive" as a
+registered UTI when dragging or copying a text selection. This broke the Messages app on iOS, which currently
+inserts a copied or dragged text selection from WebKit-based views as a web archive file attachment. A fix for
+this is internally tracked in .
+
+* platform/ios/PlatformPasteboardIOS.mm:
+(WebCore::PlatformPasteboard::write):
+
2018-12-18 Michael Catanzaro
Unreviewed, add a missing UNUSED_PARAM()
Modified: trunk/Source/WebCore/platform/ios/PlatformPasteboardIOS.mm (239368 => 239369)
--- trunk/Source/WebCore/platform/ios/PlatformPasteboardIOS.mm 2018-12-19 05:31:11 UTC (rev 239368)
+++ trunk/Source/WebCore/platform/ios/PlatformPasteboardIOS.mm 2018-12-19 06:10:28 UTC (rev 239369)
@@ -369,10 +369,13 @@
if (content.dataInWebArchiveFormat) {
auto webArchiveData = content.dataInWebArchiveFormat->createNSData();
-#if !PLATFORM(IOSMAC)
-[representationsToRegister addData:webArchiveData.get() forType:WebArchivePboardType];
+#if PLATFORM(IOSMAC)
+NSString *webArchiveType = (__bridge NSString *)kUTTypeWebArchive;
+#else
+// FIXME: We should additionally register "com.apple.webarchive" once is fixed.
+NSString *webArchiveType = WebArchivePboardType;
#endif
-[representationsToRegister addData:webArchiveData.get() forType:(__bridge NSString *)kUTTypeWebArchive];
+[representationsToRegister addData:webArchiveData.get() forType:webArchiveType];
}
if (content.dataInAttributedStringFormat) {
Modified: trunk/Tools/ChangeLog (239368 => 239369)
--- trunk/Tools/ChangeLog 2018-12-19 05:31:11 UTC (rev 239368)
+++ trunk/Tools/ChangeLog 2018-12-19 06:10:28 UTC (rev 239369)
@@ -1,3 +1,16 @@
+2018-12-18 Wenson Hsieh
+
+[iOS] A copied text selection is pasted as a web archive attachment in the entry view in Messages
+https://bugs.webkit.org/show_bug.cgi?id=192842
+
+
+Reviewed by Tim Horton.
+
+Adjusts an existing API test to verify that "com.apple.webarchive" is not one of the types registered to the
+item provider when dragging a rich text selection.
+
+* TestWebKitAPI/Tests/ios/DragAndDropTestsIOS.mm:
+
2018-12-18 David Quesada
REGRESSION: [iOS] API test TestWebKitAPI._WKDownload.OriginatingWebView is a flaky failure
Modified: trunk/Tools/TestWebKitAPI/Tests/ios/DragAndDropTestsIOS.mm (239368 => 239369)
--- trunk/Tools/TestWebKitAPI/Tests/ios/DragAndDropTestsIOS.mm 2018-12-19 05:31:11 UTC (rev 239368)
+++ trunk/Tools/TestWebKitAPI/Tests/ios/DragAndDropTestsIOS.mm 2018-12-19 06:10:28 UTC (rev 239369)
@@ -372,6 +372,9 @@
checkSelectionRectsWithLogging(@[ makeCGRectValue(1, 201, 961, 227) ], [simulator finalSelectionRects]);
checkRichTextTypePrecedesPlainTextType(simulator.get());
EXPECT_TRUE([simulator lastKnownDropProposal].precise);
+
+// FIXME: Once is fixed, we should add "com.apple.webarchive" as a registered pasteboard type and rebaseline this expectation.
+EXPECT_FALSE([[[simulator sourceItemProviders].firstObject registeredTypeIdentifiers] containsObject:(__bridge NSString *)kUTTypeWebArchive]);
}
TEST(DragAndDropTests, ContentEditableToTextarea)
[webkit-changes] [239368] trunk/Source/WebCore
Title: [239368] trunk/Source/WebCore Revision 239368 Author mcatanz...@igalia.com Date 2018-12-18 21:31:11 -0800 (Tue, 18 Dec 2018) Log Message Unreviewed, add a missing UNUSED_PARAM() * Modules/mediarecorder/MediaRecorder.cpp: (WebCore::MediaRecorder::getPrivateImpl): Modified Paths trunk/Source/WebCore/ChangeLog trunk/Source/WebCore/Modules/mediarecorder/MediaRecorder.cpp Diff Modified: trunk/Source/WebCore/ChangeLog (239367 => 239368) --- trunk/Source/WebCore/ChangeLog 2018-12-19 05:01:04 UTC (rev 239367) +++ trunk/Source/WebCore/ChangeLog 2018-12-19 05:31:11 UTC (rev 239368) @@ -1,3 +1,10 @@ +2018-12-18 Michael Catanzaro + +Unreviewed, add a missing UNUSED_PARAM() + +* Modules/mediarecorder/MediaRecorder.cpp: +(WebCore::MediaRecorder::getPrivateImpl): + 2018-12-18 Justin Michaud Update CSS Properties and Values API to use new cycle fallback behaviour Modified: trunk/Source/WebCore/Modules/mediarecorder/MediaRecorder.cpp (239367 => 239368) --- trunk/Source/WebCore/Modules/mediarecorder/MediaRecorder.cpp 2018-12-19 05:01:04 UTC (rev 239367) +++ trunk/Source/WebCore/Modules/mediarecorder/MediaRecorder.cpp 2018-12-19 05:31:11 UTC (rev 239368) @@ -66,8 +66,10 @@ #if PLATFORM(COCOA) return MediaRecorderPrivateAVFImpl::create(stream); +#else +UNUSED_PARAM(stream); +return nullptr; #endif -return nullptr; } MediaRecorder::MediaRecorder(Document& document, Ref&& stream, std::unique_ptr&& privateImpl, Options&& option) ___ webkit-changes mailing list webkit-changes@lists.webkit.org https://lists.webkit.org/mailman/listinfo/webkit-changes
[webkit-changes] [239367] trunk/Source/WebKit
Title: [239367] trunk/Source/WebKit
Revision 239367
Author achristen...@apple.com
Date 2018-12-18 21:01:04 -0800 (Tue, 18 Dec 2018)
Log Message
REGRESSION(r239134) iOS safe browsing warning unable to show details
https://bugs.webkit.org/show_bug.cgi?id=192837
Reviewed by Tim Horton.
* UIProcess/Cocoa/WKSafeBrowsingWarning.mm:
(-[WKSafeBrowsingWarning addContent]):
I had a fragile design that relied on the internal view heirarchy structure of UIScrollView,
which I changed in r239134 by introducing a layout which changed where its scrollbars were in the view heirarchy.
I now have a more robust design that saves a WeakObjCPtr to the view I need to traverse to.
Testing is problematic because this had to do with scrollbars and UIScrollView internals which seem to behave
differently in TestWebKitAPI, but I verified manually that this fixes the issue.
Modified Paths
trunk/Source/WebKit/ChangeLog
trunk/Source/WebKit/UIProcess/Cocoa/WKSafeBrowsingWarning.h
trunk/Source/WebKit/UIProcess/Cocoa/WKSafeBrowsingWarning.mm
Diff
Modified: trunk/Source/WebKit/ChangeLog (239366 => 239367)
--- trunk/Source/WebKit/ChangeLog 2018-12-19 04:35:34 UTC (rev 239366)
+++ trunk/Source/WebKit/ChangeLog 2018-12-19 05:01:04 UTC (rev 239367)
@@ -1,3 +1,18 @@
+2018-12-18 Alex Christensen
+
+REGRESSION(r239134) iOS safe browsing warning unable to show details
+https://bugs.webkit.org/show_bug.cgi?id=192837
+
+Reviewed by Tim Horton.
+
+* UIProcess/Cocoa/WKSafeBrowsingWarning.mm:
+(-[WKSafeBrowsingWarning addContent]):
+I had a fragile design that relied on the internal view heirarchy structure of UIScrollView,
+which I changed in r239134 by introducing a layout which changed where its scrollbars were in the view heirarchy.
+I now have a more robust design that saves a WeakObjCPtr to the view I need to traverse to.
+Testing is problematic because this had to do with scrollbars and UIScrollView internals which seem to behave
+differently in TestWebKitAPI, but I verified manually that this fixes the issue.
+
2018-12-18 Fujii Hironori
[Win][Clang] Fix compilation warnings under Source/WebKit directory
Modified: trunk/Source/WebKit/UIProcess/Cocoa/WKSafeBrowsingWarning.h (239366 => 239367)
--- trunk/Source/WebKit/UIProcess/Cocoa/WKSafeBrowsingWarning.h 2018-12-19 04:35:34 UTC (rev 239366)
+++ trunk/Source/WebKit/UIProcess/Cocoa/WKSafeBrowsingWarning.h 2018-12-19 05:01:04 UTC (rev 239367)
@@ -43,9 +43,11 @@
OBJC_CLASS WKSafeBrowsingTextView;
#if PLATFORM(MAC)
+using ViewType = NSView;
using RectType = NSRect;
@interface WKSafeBrowsingWarning : NSView
#else
+using ViewType = UIView;
using RectType = CGRect;
@interface WKSafeBrowsingWarning : UIScrollView
#endif
@@ -54,6 +56,7 @@
CompletionHandler&&)> _completionHandler;
RefPtr _warning;
WeakObjCPtr _details;
+WeakObjCPtr _box;
}
- (instancetype)initWithFrame:(RectType)frame safeBrowsingWarning:(const WebKit::SafeBrowsingWarning&)warning completionHandler:(CompletionHandler&&)>&&)completionHandler;
Modified: trunk/Source/WebKit/UIProcess/Cocoa/WKSafeBrowsingWarning.mm (239366 => 239367)
--- trunk/Source/WebKit/UIProcess/Cocoa/WKSafeBrowsingWarning.mm 2018-12-19 04:35:34 UTC (rev 239366)
+++ trunk/Source/WebKit/UIProcess/Cocoa/WKSafeBrowsingWarning.mm 2018-12-19 05:01:04 UTC (rev 239367)
@@ -50,7 +50,6 @@
using TextViewType = NSTextView;
using ButtonType = NSButton;
using AlignmentType = NSLayoutAttribute;
-using ViewType = NSView;
using SizeType = NSSize;
#else
using ColorType = UIColor;
@@ -58,7 +57,6 @@
using TextViewType = UITextView;
using ButtonType = UIButton;
using AlignmentType = UIStackViewAlignment;
-using ViewType = UIView;
using SizeType = CGSize;
#endif
@@ -299,6 +297,7 @@
auto showDetails = makeButton(WarningItem::ShowDetailsButton, self, @selector(showDetailsClicked));
auto goBack = makeButton(WarningItem::GoBackButton, self, @selector(goBackClicked));
auto box = [[ViewType new] autorelease];
+_box = box;
setBackground(box, colorForItem(WarningItem::BoxBackground, self));
box.layer.cornerRadius = boxCornerRadius;
@@ -367,7 +366,7 @@
- (void)showDetailsClicked
{
-ViewType *box = self.subviews.lastObject;
+ViewType *box = _box.get().get();
ButtonType *showDetails = box.subviews.lastObject;
[showDetails removeFromSuperview];
___
webkit-changes mailing list
webkit-changes@lists.webkit.org
https://lists.webkit.org/mailman/listinfo/webkit-changes
[webkit-changes] [239366] trunk/Source/WebKit
Title: [239366] trunk/Source/WebKit
Revision 239366
Author hironori.fu...@sony.com
Date 2018-12-18 20:35:34 -0800 (Tue, 18 Dec 2018)
Log Message
[Win][Clang] Fix compilation warnings under Source/WebKit directory
https://bugs.webkit.org/show_bug.cgi?id=192695
Reviewed by Alex Christensen.
* NetworkProcess/cache/NetworkCacheData.cpp:
(makeSalt): Enclosed by #if !OS(WINDOWS).
* NetworkProcess/cache/NetworkCacheFileSystem.cpp:
(WebKit::NetworkCache::directoryEntryType): Ditto.
* Platform/win/ModuleWin.cpp:
(WebKit::Module::platformFunctionPointer const): Cast a function pointer with reinterpret_cast().
* UIProcess/DrawingAreaProxyImpl.cpp:
(WebKit::DrawingAreaProxyImpl::DrawingMonitor::DrawingMonitor):
Moved the initializer of m_webPage in order to be encluded by #if PLATFORM(GTK).
* UIProcess/DrawingAreaProxyImpl.h: Ditto.
* UIProcess/Launcher/win/ProcessLauncherWin.cpp:
(WebKit::processName): Removed the duplicated 'const' type qualifier.
* UIProcess/win/WebInspectorProxyWin.cpp:
(WebKit::WebInspectorProxy::platformAttach): Removed an unused variable.
(WebKit::WebInspectorProxy::platformDetach): Ditto.
* UIProcess/win/WebPopupMenuProxyWin.cpp: Ditto.
* UIProcess/win/WebView.cpp:
(WebKit::WebView::paint): Ditto.
(WebKit::WebPopupMenu::setUpPlatformData): Ditto.
* UIProcess/win/WebPopupMenuProxyWin.h: Marked override methods with 'override'.
* WebProcess/WebCoreSupport/curl/WebFrameNetworkingContext.h: Ditto.
* WebProcess/WebCoreSupport/win/WebPopupMenuWin.cpp: Removed an unused variable.
Modified Paths
trunk/Source/WebKit/ChangeLog
trunk/Source/WebKit/NetworkProcess/cache/NetworkCacheData.cpp
trunk/Source/WebKit/NetworkProcess/cache/NetworkCacheFileSystem.cpp
trunk/Source/WebKit/Platform/win/ModuleWin.cpp
trunk/Source/WebKit/UIProcess/DrawingAreaProxyImpl.cpp
trunk/Source/WebKit/UIProcess/DrawingAreaProxyImpl.h
trunk/Source/WebKit/UIProcess/Launcher/win/ProcessLauncherWin.cpp
trunk/Source/WebKit/UIProcess/win/WebInspectorProxyWin.cpp
trunk/Source/WebKit/UIProcess/win/WebPopupMenuProxyWin.cpp
trunk/Source/WebKit/UIProcess/win/WebPopupMenuProxyWin.h
trunk/Source/WebKit/UIProcess/win/WebView.cpp
trunk/Source/WebKit/WebProcess/WebCoreSupport/curl/WebFrameNetworkingContext.h
trunk/Source/WebKit/WebProcess/WebCoreSupport/win/WebPopupMenuWin.cpp
Diff
Modified: trunk/Source/WebKit/ChangeLog (239365 => 239366)
--- trunk/Source/WebKit/ChangeLog 2018-12-19 03:21:15 UTC (rev 239365)
+++ trunk/Source/WebKit/ChangeLog 2018-12-19 04:35:34 UTC (rev 239366)
@@ -1,3 +1,33 @@
+2018-12-18 Fujii Hironori
+
+[Win][Clang] Fix compilation warnings under Source/WebKit directory
+https://bugs.webkit.org/show_bug.cgi?id=192695
+
+Reviewed by Alex Christensen.
+
+* NetworkProcess/cache/NetworkCacheData.cpp:
+(makeSalt): Enclosed by #if !OS(WINDOWS).
+* NetworkProcess/cache/NetworkCacheFileSystem.cpp:
+(WebKit::NetworkCache::directoryEntryType): Ditto.
+* Platform/win/ModuleWin.cpp:
+(WebKit::Module::platformFunctionPointer const): Cast a function pointer with reinterpret_cast().
+* UIProcess/DrawingAreaProxyImpl.cpp:
+(WebKit::DrawingAreaProxyImpl::DrawingMonitor::DrawingMonitor):
+Moved the initializer of m_webPage in order to be encluded by #if PLATFORM(GTK).
+* UIProcess/DrawingAreaProxyImpl.h: Ditto.
+* UIProcess/Launcher/win/ProcessLauncherWin.cpp:
+(WebKit::processName): Removed the duplicated 'const' type qualifier.
+* UIProcess/win/WebInspectorProxyWin.cpp:
+(WebKit::WebInspectorProxy::platformAttach): Removed an unused variable.
+(WebKit::WebInspectorProxy::platformDetach): Ditto.
+* UIProcess/win/WebPopupMenuProxyWin.cpp: Ditto.
+* UIProcess/win/WebView.cpp:
+(WebKit::WebView::paint): Ditto.
+(WebKit::WebPopupMenu::setUpPlatformData): Ditto.
+* UIProcess/win/WebPopupMenuProxyWin.h: Marked override methods with 'override'.
+* WebProcess/WebCoreSupport/curl/WebFrameNetworkingContext.h: Ditto.
+* WebProcess/WebCoreSupport/win/WebPopupMenuWin.cpp: Removed an unused variable.
+
2018-12-18 Wenson Hsieh
[macOS] fast/forms/datalist/datalist-textinput-suggestions-order.html sometimes crashes after r239337
Modified: trunk/Source/WebKit/NetworkProcess/cache/NetworkCacheData.cpp (239365 => 239366)
--- trunk/Source/WebKit/NetworkProcess/cache/NetworkCacheData.cpp 2018-12-19 03:21:15 UTC (rev 239365)
+++ trunk/Source/WebKit/NetworkProcess/cache/NetworkCacheData.cpp 2018-12-19 04:35:34 UTC (rev 239366)
@@ -142,6 +142,7 @@
return !memcmp(a.data(), b.data(), a.size());
}
+#if !OS(WINDOWS)
static Salt makeSalt()
{
Salt salt;
@@ -150,6 +151,7 @@
*reinterpret_cast([4]) = cryptographicallyRandomNumber();
return salt;
}
+#endif
std::optional readOrMakeSalt(const String& path)
{
Modified: trunk/Source/WebKit/NetworkProcess/cache/NetworkCacheFileSystem.cpp (239365 => 239366)
---
[webkit-changes] [239364] trunk/Source/JavaScriptCore
Title: [239364] trunk/Source/_javascript_Core
Revision 239364
Author sbar...@apple.com
Date 2018-12-18 18:27:00 -0800 (Tue, 18 Dec 2018)
Log Message
Update ARM64EHash
https://bugs.webkit.org/show_bug.cgi?id=192823
Reviewed by Mark Lam.
* assembler/ARM64Assembler.h:
(JSC::ARM64Assembler::ARM64Assembler):
* assembler/AbstractMacroAssembler.h:
(JSC::AbstractMacroAssembler::AbstractMacroAssembler):
* assembler/AssemblerBuffer.h:
(JSC::ARM64EHash::update):
(JSC::ARM64EHash::finalHash const):
(JSC::AssemblerBuffer::AssemblerBuffer):
(JSC::AssemblerBuffer::putIntegralUnchecked):
(JSC::ARM64EHash::ARM64EHash): Deleted.
(JSC::ARM64EHash::hash const): Deleted.
(JSC::ARM64EHash::randomSeed const): Deleted.
* assembler/LinkBuffer.cpp:
(JSC::LinkBuffer::copyCompactAndLinkCode):
Modified Paths
trunk/Source/_javascript_Core/ChangeLog
trunk/Source/_javascript_Core/assembler/ARM64Assembler.h
trunk/Source/_javascript_Core/assembler/AbstractMacroAssembler.h
trunk/Source/_javascript_Core/assembler/AssemblerBuffer.h
trunk/Source/_javascript_Core/assembler/LinkBuffer.cpp
Diff
Modified: trunk/Source/_javascript_Core/ChangeLog (239363 => 239364)
--- trunk/Source/_javascript_Core/ChangeLog 2018-12-19 02:11:58 UTC (rev 239363)
+++ trunk/Source/_javascript_Core/ChangeLog 2018-12-19 02:27:00 UTC (rev 239364)
@@ -1,3 +1,26 @@
+2018-12-18 Saam Barati
+
+Update ARM64EHash
+https://bugs.webkit.org/show_bug.cgi?id=192823
+
+
+Reviewed by Mark Lam.
+
+* assembler/ARM64Assembler.h:
+(JSC::ARM64Assembler::ARM64Assembler):
+* assembler/AbstractMacroAssembler.h:
+(JSC::AbstractMacroAssembler::AbstractMacroAssembler):
+* assembler/AssemblerBuffer.h:
+(JSC::ARM64EHash::update):
+(JSC::ARM64EHash::finalHash const):
+(JSC::AssemblerBuffer::AssemblerBuffer):
+(JSC::AssemblerBuffer::putIntegralUnchecked):
+(JSC::ARM64EHash::ARM64EHash): Deleted.
+(JSC::ARM64EHash::hash const): Deleted.
+(JSC::ARM64EHash::randomSeed const): Deleted.
+* assembler/LinkBuffer.cpp:
+(JSC::LinkBuffer::copyCompactAndLinkCode):
+
2018-12-18 Mark Lam
JSON.stringify() should throw OOM on StringBuilder overflows.
Modified: trunk/Source/_javascript_Core/assembler/ARM64Assembler.h (239363 => 239364)
--- trunk/Source/_javascript_Core/assembler/ARM64Assembler.h 2018-12-19 02:11:58 UTC (rev 239363)
+++ trunk/Source/_javascript_Core/assembler/ARM64Assembler.h 2018-12-19 02:27:00 UTC (rev 239364)
@@ -326,17 +326,9 @@
static constexpr bool isZr(RegisterID reg) { return ARM64Registers::isZr(reg); }
public:
-ARM64Assembler(
-#if CPU(ARM64E)
-unsigned randomNumber
-#endif
-)
+ARM64Assembler()
: m_indexOfLastWatchpoint(INT_MIN)
, m_indexOfTailOfLastWatchpoint(INT_MIN)
-#if CPU(ARM64E)
-, m_buffer(randomNumber)
-#endif
-
{
}
Modified: trunk/Source/_javascript_Core/assembler/AbstractMacroAssembler.h (239363 => 239364)
--- trunk/Source/_javascript_Core/assembler/AbstractMacroAssembler.h 2018-12-19 02:11:58 UTC (rev 239363)
+++ trunk/Source/_javascript_Core/assembler/AbstractMacroAssembler.h 2018-12-19 02:27:00 UTC (rev 239364)
@@ -986,11 +986,7 @@
protected:
AbstractMacroAssembler()
: m_randomSource(0)
-#if CPU(ARM64E)
-, m_assembler(random())
-#else
, m_assembler()
-#endif
{
invalidateAllTempRegisters();
}
Modified: trunk/Source/_javascript_Core/assembler/AssemblerBuffer.h (239363 => 239364)
--- trunk/Source/_javascript_Core/assembler/AssemblerBuffer.h 2018-12-19 02:11:58 UTC (rev 239363)
+++ trunk/Source/_javascript_Core/assembler/AssemblerBuffer.h 2018-12-19 02:27:00 UTC (rev 239364)
@@ -149,34 +149,31 @@
#if CPU(ARM64E)
class ARM64EHash {
public:
-ARM64EHash(unsigned randomNumber)
-: m_hash(randomNumber)
-, m_randomSeed(randomNumber)
-{ }
-ALWAYS_INLINE void update(unsigned value, uintptr_t index)
+ARM64EHash() = default;
+ALWAYS_INLINE void update(uint32_t value)
{
-m_hash = tagInt((static_cast(value) + m_hash) ^ (m_hash >> 32), static_cast(index));
+uint64_t input = value ^ m_hash;
+uint64_t a = static_cast(tagInt(input, static_cast(0)) >> 39);
+uint64_t b = tagInt(input, static_cast(0xb7e151628aed2a6a)) >> 23;
+m_hash = a | b;
}
-uintptr_t hash() const { return m_hash; }
-unsigned randomSeed() const { return m_randomSeed; }
+uint32_t finalHash() const
+{
+uint64_t hash = m_hash;
+uint64_t a = static_cast(tagInt(hash, static_cast(0xbf7158809cf4f3c7)) >> 39);
+uint64_t b = tagInt(hash, static_cast(0x62e7160f38b4da56)) >> 23;
+return static_cast(a | b);
+}
private:
-uintptr_t m_hash;
-unsigned m_randomSeed;
+
[webkit-changes] [239363] trunk/Source/WebKit
Title: [239363] trunk/Source/WebKit
Revision 239363
Author wenson_hs...@apple.com
Date 2018-12-18 18:11:58 -0800 (Tue, 18 Dec 2018)
Log Message
[macOS] fast/forms/datalist/datalist-textinput-suggestions-order.html sometimes crashes after r239337
https://bugs.webkit.org/show_bug.cgi?id=192836
Reviewed by Tim Horton.
r239337 exposed a subtle issue in the implementation of `-[WKDataListSuggestionsView updateWithInformation:]`,
wherein we might end up recomputing information for table view cells at rows that no longer exist. This currently
happens because we first set the datalist's `_suggestions`, then change the table view's frame, and then finally
reload the table view; however, in the case where the number of suggestions decreased and the frame changed
(which prompts AppKit to ask us again for table cell views), we'll attempt to access an out-of-bounds index in
the `_suggestions` vector.
To address this, we change our suggestions and immediately reload the table view before updating `NSTableView`'s
frame, so that we only provide table view cells for rows that actually correspond to items in `_suggestions`.
* UIProcess/mac/WebDataListSuggestionsDropdownMac.mm:
(-[WKDataListSuggestionsView updateWithInformation:]):
Modified Paths
trunk/Source/WebKit/ChangeLog
trunk/Source/WebKit/UIProcess/mac/WebDataListSuggestionsDropdownMac.mm
Diff
Modified: trunk/Source/WebKit/ChangeLog (239362 => 239363)
--- trunk/Source/WebKit/ChangeLog 2018-12-19 01:25:52 UTC (rev 239362)
+++ trunk/Source/WebKit/ChangeLog 2018-12-19 02:11:58 UTC (rev 239363)
@@ -1,3 +1,24 @@
+2018-12-18 Wenson Hsieh
+
+[macOS] fast/forms/datalist/datalist-textinput-suggestions-order.html sometimes crashes after r239337
+https://bugs.webkit.org/show_bug.cgi?id=192836
+
+
+Reviewed by Tim Horton.
+
+r239337 exposed a subtle issue in the implementation of `-[WKDataListSuggestionsView updateWithInformation:]`,
+wherein we might end up recomputing information for table view cells at rows that no longer exist. This currently
+happens because we first set the datalist's `_suggestions`, then change the table view's frame, and then finally
+reload the table view; however, in the case where the number of suggestions decreased and the frame changed
+(which prompts AppKit to ask us again for table cell views), we'll attempt to access an out-of-bounds index in
+the `_suggestions` vector.
+
+To address this, we change our suggestions and immediately reload the table view before updating `NSTableView`'s
+frame, so that we only provide table view cells for rows that actually correspond to items in `_suggestions`.
+
+* UIProcess/mac/WebDataListSuggestionsDropdownMac.mm:
+(-[WKDataListSuggestionsView updateWithInformation:]):
+
2018-12-18 Jiewen Tan
[Mac] Layout Test http/wpt/webauthn/public-key-credential-create-success-hid.https.html and http/wpt/webauthn/public-key-credential-get-success-hid.https.html are flaky
Modified: trunk/Source/WebKit/UIProcess/mac/WebDataListSuggestionsDropdownMac.mm (239362 => 239363)
--- trunk/Source/WebKit/UIProcess/mac/WebDataListSuggestionsDropdownMac.mm 2018-12-19 01:25:52 UTC (rev 239362)
+++ trunk/Source/WebKit/UIProcess/mac/WebDataListSuggestionsDropdownMac.mm 2018-12-19 02:11:58 UTC (rev 239363)
@@ -346,9 +346,10 @@
- (void)updateWithInformation:(WebCore::DataListSuggestionInformation&&)information
{
_suggestions = WTFMove(information.suggestions);
+[_table reload];
+
[_enclosingWindow setFrame:[self dropdownRectForElementRect:information.elementRect] display:YES];
[_table setVisibleRect:[_enclosingWindow frame]];
-[_table reload];
}
- (void)moveSelectionByDirection:(const String&)direction
___
webkit-changes mailing list
webkit-changes@lists.webkit.org
https://lists.webkit.org/mailman/listinfo/webkit-changes
[webkit-changes] [239362] trunk/Source/WebKit
Title: [239362] trunk/Source/WebKit
Revision 239362
Author jiewen_...@apple.com
Date 2018-12-18 17:25:52 -0800 (Tue, 18 Dec 2018)
Log Message
[Mac] Layout Test http/wpt/webauthn/public-key-credential-create-success-hid.https.html and http/wpt/webauthn/public-key-credential-get-success-hid.https.html are flaky
https://bugs.webkit.org/show_bug.cgi?id=192061
Reviewed by Dewei Zhu.
Part 4.
Add some additional temporary logging info to determine if the time out value passed to the timer is right.
Once the bug is determined and fixed, we should remove all logging added in this patch.
* UIProcess/WebAuthentication/AuthenticatorManager.cpp:
(WebKit::AuthenticatorManager::initTimeOutTimer):
Modified Paths
trunk/Source/WebKit/ChangeLog
trunk/Source/WebKit/UIProcess/WebAuthentication/AuthenticatorManager.cpp
Diff
Modified: trunk/Source/WebKit/ChangeLog (239361 => 239362)
--- trunk/Source/WebKit/ChangeLog 2018-12-19 00:23:21 UTC (rev 239361)
+++ trunk/Source/WebKit/ChangeLog 2018-12-19 01:25:52 UTC (rev 239362)
@@ -1,3 +1,18 @@
+2018-12-18 Jiewen Tan
+
+[Mac] Layout Test http/wpt/webauthn/public-key-credential-create-success-hid.https.html and http/wpt/webauthn/public-key-credential-get-success-hid.https.html are flaky
+https://bugs.webkit.org/show_bug.cgi?id=192061
+
+Reviewed by Dewei Zhu.
+
+Part 4.
+
+Add some additional temporary logging info to determine if the time out value passed to the timer is right.
+Once the bug is determined and fixed, we should remove all logging added in this patch.
+
+* UIProcess/WebAuthentication/AuthenticatorManager.cpp:
+(WebKit::AuthenticatorManager::initTimeOutTimer):
+
2018-12-18 Sihui Liu
Clean up IndexedDB files between tests
Modified: trunk/Source/WebKit/UIProcess/WebAuthentication/AuthenticatorManager.cpp (239361 => 239362)
--- trunk/Source/WebKit/UIProcess/WebAuthentication/AuthenticatorManager.cpp 2018-12-19 00:23:21 UTC (rev 239361)
+++ trunk/Source/WebKit/UIProcess/WebAuthentication/AuthenticatorManager.cpp 2018-12-19 01:25:52 UTC (rev 239362)
@@ -226,7 +226,7 @@
unsigned timeOutInMsValue = std::min(maxTimeOutValue, timeOutInMs.value_or(maxTimeOutValue));
// FIXME(192061)
-LOG_ERROR("Start timer.");
+LOG_ERROR("Start timer. %d", timeOutInMsValue);
m_requestTimeOutTimer.startOneShot(Seconds::fromMilliseconds(timeOutInMsValue));
}
___
webkit-changes mailing list
webkit-changes@lists.webkit.org
https://lists.webkit.org/mailman/listinfo/webkit-changes
[webkit-changes] [239361] trunk/Source/WebCore
Title: [239361] trunk/Source/WebCore Revision 239361 Author dba...@webkit.org Date 2018-12-18 16:23:21 -0800 (Tue, 18 Dec 2018) Log Message Wrong value for key property in keydown and keyup events generated holding Control key https://bugs.webkit.org/show_bug.cgi?id=192788 Reviewed by Wenson Hsieh. Similar to what we do on Mac, compute the DOM key property from the characters ignoring modifier keys input string when the Control key is held down. * platform/ios/PlatformEventFactoryIOS.mm: (WebCore::keyForKeyEvent): * platform/mac/PlatformEventFactoryMac.mm: (WebCore::keyForKeyEvent): Modified Paths trunk/Source/WebCore/ChangeLog trunk/Source/WebCore/platform/ios/PlatformEventFactoryIOS.mm trunk/Source/WebCore/platform/mac/PlatformEventFactoryMac.mm Diff Modified: trunk/Source/WebCore/ChangeLog (239360 => 239361) --- trunk/Source/WebCore/ChangeLog 2018-12-19 00:02:45 UTC (rev 239360) +++ trunk/Source/WebCore/ChangeLog 2018-12-19 00:23:21 UTC (rev 239361) @@ -1,3 +1,19 @@ +2018-12-18 Daniel Bates + +Wrong value for key property in keydown and keyup events generated holding Control key +https://bugs.webkit.org/show_bug.cgi?id=192788 + + +Reviewed by Wenson Hsieh. + +Similar to what we do on Mac, compute the DOM key property from the characters ignoring +modifier keys input string when the Control key is held down. + +* platform/ios/PlatformEventFactoryIOS.mm: +(WebCore::keyForKeyEvent): +* platform/mac/PlatformEventFactoryMac.mm: +(WebCore::keyForKeyEvent): + 2018-12-18 Sihui Liu Clean up IndexedDB files between tests Modified: trunk/Source/WebCore/platform/ios/PlatformEventFactoryIOS.mm (239360 => 239361) --- trunk/Source/WebCore/platform/ios/PlatformEventFactoryIOS.mm 2018-12-19 00:02:45 UTC (rev 239360) +++ trunk/Source/WebCore/platform/ios/PlatformEventFactoryIOS.mm 2018-12-19 00:23:21 UTC (rev 239361) @@ -190,7 +190,13 @@ } } -NSString *characters = event.characters; +// If more than one key is being pressed and the key combination includes one or more modifier keys +// that result in the key no longer producing a printable character (e.g., Control + a), then the +// key value should be the printable key value that would have been produced if the key had been +// typed with the default keyboard layout with no modifier keys except for Shift and AltGr applied. +// See . +bool isControlDown = event.modifierFlags & WebEventFlagMaskControlKey; +NSString *characters = isControlDown ? event.charactersIgnoringModifiers : event.characters; auto length = [characters length]; // characters return an empty string for dead keys. // https://developer.apple.com/reference/appkit/nsevent/1534183-characters Modified: trunk/Source/WebCore/platform/mac/PlatformEventFactoryMac.mm (239360 => 239361) --- trunk/Source/WebCore/platform/mac/PlatformEventFactoryMac.mm 2018-12-19 00:02:45 UTC (rev 239360) +++ trunk/Source/WebCore/platform/mac/PlatformEventFactoryMac.mm 2018-12-19 00:23:21 UTC (rev 239361) @@ -260,7 +260,7 @@ // that result in the key no longer producing a printable character (e.g., Control + a), then the // key value should be the printable key value that would have been produced if the key had been // typed with the default keyboard layout with no modifier keys except for Shift and AltGr applied. -// https://w3c.github.io/uievents/#keys-guidelines +// See . bool isControlDown = ([event modifierFlags] & NSEventModifierFlagControl); NSString *s = isControlDown ? [event charactersIgnoringModifiers] : [event characters]; auto length = [s length]; ___ webkit-changes mailing list webkit-changes@lists.webkit.org https://lists.webkit.org/mailman/listinfo/webkit-changes
[webkit-changes] [239360] trunk/JSTests
Title: [239360] trunk/JSTests Revision 239360 Author mark@apple.com Date 2018-12-18 16:02:45 -0800 (Tue, 18 Dec 2018) Log Message Gardening: test options fix. https://bugs.webkit.org/show_bug.cgi?id=192822 Unreviewed. * stress/json-stringify-string-builder-overflow.js: Modified Paths trunk/JSTests/ChangeLog trunk/JSTests/stress/json-stringify-string-builder-overflow.js Diff Modified: trunk/JSTests/ChangeLog (239359 => 239360) --- trunk/JSTests/ChangeLog 2018-12-18 23:59:34 UTC (rev 239359) +++ trunk/JSTests/ChangeLog 2018-12-19 00:02:45 UTC (rev 239360) @@ -1,5 +1,14 @@ 2018-12-18 Mark Lam +Gardening: test options fix. +https://bugs.webkit.org/show_bug.cgi?id=192822 + +Unreviewed. + +* stress/json-stringify-string-builder-overflow.js: + +2018-12-18 Mark Lam + JSON.stringify() should throw OOM on StringBuilder overflows. https://bugs.webkit.org/show_bug.cgi?id=192822 Modified: trunk/JSTests/stress/json-stringify-string-builder-overflow.js (239359 => 239360) --- trunk/JSTests/stress/json-stringify-string-builder-overflow.js 2018-12-18 23:59:34 UTC (rev 239359) +++ trunk/JSTests/stress/json-stringify-string-builder-overflow.js 2018-12-19 00:02:45 UTC (rev 239360) @@ -1,4 +1,4 @@ -//@ slow +//@ slow! //@ skip if $architecture != "arm64" and $architecture != "x86-64" var exception; ___ webkit-changes mailing list webkit-changes@lists.webkit.org https://lists.webkit.org/mailman/listinfo/webkit-changes
[webkit-changes] [239359] trunk/Tools
Title: [239359] trunk/Tools
Revision 239359
Author david_ques...@apple.com
Date 2018-12-18 15:59:34 -0800 (Tue, 18 Dec 2018)
Log Message
REGRESSION: [iOS] API test TestWebKitAPI._WKDownload.OriginatingWebView is a flaky failure
https://bugs.webkit.org/show_bug.cgi?id=192810
Reviewed by Alex Christensen.
* TestWebKitAPI/Tests/WebKitCocoa/Download.mm:
(-[OriginatingWebViewDownloadDelegate _downloadDidStart:]):
After releasing our reference to the web view, wait until the next turn of the runloop
before checking that the originatingWebView has been zeroed, since the web view might
still be referenced in the autorelease pool. This _WKDownload unit test isn't about
testing general memory management of WKWebViews, only that _WKDownload does not retain
its web view via the 'originatingWebView' property.
Modified Paths
trunk/Tools/ChangeLog
trunk/Tools/TestWebKitAPI/Tests/WebKitCocoa/Download.mm
Diff
Modified: trunk/Tools/ChangeLog (239358 => 239359)
--- trunk/Tools/ChangeLog 2018-12-18 23:56:23 UTC (rev 239358)
+++ trunk/Tools/ChangeLog 2018-12-18 23:59:34 UTC (rev 239359)
@@ -1,3 +1,19 @@
+2018-12-18 David Quesada
+
+REGRESSION: [iOS] API test TestWebKitAPI._WKDownload.OriginatingWebView is a flaky failure
+https://bugs.webkit.org/show_bug.cgi?id=192810
+
+
+Reviewed by Alex Christensen.
+
+* TestWebKitAPI/Tests/WebKitCocoa/Download.mm:
+(-[OriginatingWebViewDownloadDelegate _downloadDidStart:]):
+After releasing our reference to the web view, wait until the next turn of the runloop
+before checking that the originatingWebView has been zeroed, since the web view might
+still be referenced in the autorelease pool. This _WKDownload unit test isn't about
+testing general memory management of WKWebViews, only that _WKDownload does not retain
+its web view via the 'originatingWebView' property.
+
2018-12-18 Sihui Liu
Clean up IndexedDB files between tests
Modified: trunk/Tools/TestWebKitAPI/Tests/WebKitCocoa/Download.mm (239358 => 239359)
--- trunk/Tools/TestWebKitAPI/Tests/WebKitCocoa/Download.mm 2018-12-18 23:56:23 UTC (rev 239358)
+++ trunk/Tools/TestWebKitAPI/Tests/WebKitCocoa/Download.mm 2018-12-18 23:59:34 UTC (rev 239359)
@@ -41,6 +41,7 @@
#import
#import
#import
+#import
#import
#import
@@ -253,13 +254,13 @@
- (void)_downloadDidStart:(_WKDownload *)download
{
-@autoreleasepool {
-EXPECT_EQ([download originatingWebView], _webView);
-}
+EXPECT_EQ([download originatingWebView], _webView);
+_webView = nullptr;
-_webView = nullptr;
-EXPECT_NULL([download originatingWebView]);
-isDone = true;
+WTF::callOnMainThread([download = retainPtr(download)] {
+EXPECT_NULL([download originatingWebView]);
+isDone = true;
+});
}
@end
___
webkit-changes mailing list
webkit-changes@lists.webkit.org
https://lists.webkit.org/mailman/listinfo/webkit-changes
[webkit-changes] [239358] trunk
Title: [239358] trunk Revision 239358 Author sihui_...@apple.com Date 2018-12-18 15:56:23 -0800 (Tue, 18 Dec 2018) Log Message Clean up IndexedDB files between tests https://bugs.webkit.org/show_bug.cgi?id=192796 Reviewed by Geoffrey Garen. Source/WebCore: We should clean up the IndexedDB files between tests to make sure each test is independent of others. This patch also fixes some issues in IDB. Covered by existing tests. * Modules/indexeddb/server/IDBServer.cpp: (WebCore::IDBServer::IDBServer::closeAndDeleteDatabasesModifiedSince): We should shut down all open databases instead of databases from open database connections before deleting files, because database starts accessing files before connection to database is established. * Modules/indexeddb/server/UniqueIDBDatabase.cpp: (WebCore::IDBServer::UniqueIDBDatabase::shutdownForClose): We should shutdown database after tasks in queue are completed, because tasks have pointer of UniqueIDBDatabase and UniqueIDBDatabase can be destructed after shutdown. (WebCore::IDBServer::UniqueIDBDatabase::didDeleteBackingStore): didDeleteBackingStore can be posted to main thread after immediateCloseForUserDelete, and timer should not be invoked during the hard close. (WebCore::IDBServer::UniqueIDBDatabase::handleDatabaseOperations): Tasks like didOpenBackingStore could be posted from database thread to main thread after immediateCloseForUserDelete, but we know the backing store will be deleted soon, so no need to handle any database operation. (WebCore::IDBServer::UniqueIDBDatabase::performPrefetchCursor): performPrefetchCursor needs to be aware of whether UniqueIDBDatabase is being closed, so that it will not access m_backingStore when m_backingStore may already be deleted. (WebCore::IDBServer::UniqueIDBDatabase::immediateCloseForUserDelete): immediateCloseForUserDelete does not handle transactions that are in the process of commit or abort. m_objectStoreTransactionCounts and m_objectStoreWriteTransactions may be used by those transactions in transactionCompleted, so they do not need to be cleared here. Source/WebKit: * UIProcess/API/C/WKWebsiteDataStoreRef.cpp: (WKWebsiteDataStoreRemoveAllIndexedDatabasesSync): * UIProcess/API/C/WKWebsiteDataStoreRef.h: Tools: * DumpRenderTree/mac/DumpRenderTree.mm: (runTest): * WebKitTestRunner/TestController.cpp: (WTR::TestController::resetStateToConsistentValues): (WTR::RemoveAllIndexedDatabasesCallbackContext::RemoveAllIndexedDatabasesCallbackContext): (WTR::RemoveAllIndexedDatabasesCallback): (WTR::TestController::ClearIndexedDatabases): * WebKitTestRunner/TestController.h: Modified Paths trunk/Source/WebCore/ChangeLog trunk/Source/WebCore/Modules/indexeddb/server/IDBServer.cpp trunk/Source/WebCore/Modules/indexeddb/server/UniqueIDBDatabase.cpp trunk/Source/WebKit/ChangeLog trunk/Source/WebKit/UIProcess/API/C/WKWebsiteDataStoreRef.cpp trunk/Source/WebKit/UIProcess/API/C/WKWebsiteDataStoreRef.h trunk/Tools/ChangeLog trunk/Tools/DumpRenderTree/mac/DumpRenderTree.mm trunk/Tools/WebKitTestRunner/TestController.cpp trunk/Tools/WebKitTestRunner/TestController.h trunk/Tools/WebKitTestRunner/TestInvocation.cpp Diff Modified: trunk/Source/WebCore/ChangeLog (239357 => 239358) --- trunk/Source/WebCore/ChangeLog 2018-12-18 23:38:54 UTC (rev 239357) +++ trunk/Source/WebCore/ChangeLog 2018-12-18 23:56:23 UTC (rev 239358) @@ -1,3 +1,44 @@ +2018-12-18 Sihui Liu + +Clean up IndexedDB files between tests +https://bugs.webkit.org/show_bug.cgi?id=192796 + +Reviewed by Geoffrey Garen. + +We should clean up the IndexedDB files between tests to make sure each test is independent of others. + +This patch also fixes some issues in IDB. + +Covered by existing tests. + +* Modules/indexeddb/server/IDBServer.cpp: +(WebCore::IDBServer::IDBServer::closeAndDeleteDatabasesModifiedSince): +We should shut down all open databases instead of databases from open database connections before deleting +files, because database starts accessing files before connection to database is established. + +* Modules/indexeddb/server/UniqueIDBDatabase.cpp: +(WebCore::IDBServer::UniqueIDBDatabase::shutdownForClose): +We should shutdown database after tasks in queue are completed, because tasks have pointer of UniqueIDBDatabase +and UniqueIDBDatabase can be destructed after shutdown. + +(WebCore::IDBServer::UniqueIDBDatabase::didDeleteBackingStore): +didDeleteBackingStore can be posted to main thread after immediateCloseForUserDelete, and timer should not be +invoked during the hard close. + +(WebCore::IDBServer::UniqueIDBDatabase::handleDatabaseOperations): +Tasks like didOpenBackingStore could be posted from database thread to main thread after +immediateCloseForUserDelete, but we know the backing store will be deleted soon, so no need to handle any +database operation. + +
[webkit-changes] [239357] trunk
Title: [239357] trunk
Revision 239357
Author mmaxfi...@apple.com
Date 2018-12-18 15:38:54 -0800 (Tue, 18 Dec 2018)
Log Message
Thick overlines and line-throughs grow in the wrong direction
https://bugs.webkit.org/show_bug.cgi?id=192264
Reviewed by Dean Jackson.
Source/WebCore:
Overlines should grow upward, and line-throughs should stay centered.
Test: fast/css3-text/css3-text-decoration/text-decoration-thicknes-overline-grow-direction.html
* rendering/TextDecorationPainter.cpp:
(WebCore::TextDecorationPainter::paintTextDecoration):
* style/InlineTextBoxStyle.cpp:
(WebCore::visualOverflowForDecorations):
LayoutTests:
I can't figure out a way to test the line-through, so this just tests the overline.
* fast/css3-text/css3-text-decoration/text-decoration-thicknes-overline-grow-direction-expected.html: Added.
* fast/css3-text/css3-text-decoration/text-decoration-thicknes-overline-grow-direction.html: Added.
Modified Paths
trunk/LayoutTests/ChangeLog
trunk/Source/WebCore/ChangeLog
trunk/Source/WebCore/rendering/TextDecorationPainter.cpp
trunk/Source/WebCore/style/InlineTextBoxStyle.cpp
Added Paths
trunk/LayoutTests/fast/css3-text/css3-text-decoration/text-decoration-thicknes-overline-grow-direction-expected.html
trunk/LayoutTests/fast/css3-text/css3-text-decoration/text-decoration-thicknes-overline-grow-direction.html
Diff
Modified: trunk/LayoutTests/ChangeLog (239356 => 239357)
--- trunk/LayoutTests/ChangeLog 2018-12-18 23:02:16 UTC (rev 239356)
+++ trunk/LayoutTests/ChangeLog 2018-12-18 23:38:54 UTC (rev 239357)
@@ -1,3 +1,15 @@
+2018-12-18 Myles C. Maxfield
+
+Thick overlines and line-throughs grow in the wrong direction
+https://bugs.webkit.org/show_bug.cgi?id=192264
+
+Reviewed by Dean Jackson.
+
+I can't figure out a way to test the line-through, so this just tests the overline.
+
+* fast/css3-text/css3-text-decoration/text-decoration-thicknes-overline-grow-direction-expected.html: Added.
+* fast/css3-text/css3-text-decoration/text-decoration-thicknes-overline-grow-direction.html: Added.
+
2018-12-18 Justin Fan
[WebGPU] BindGroupLayout and Device::createBindGroupLayout
Added: trunk/LayoutTests/fast/css3-text/css3-text-decoration/text-decoration-thicknes-overline-grow-direction-expected.html (0 => 239357)
--- trunk/LayoutTests/fast/css3-text/css3-text-decoration/text-decoration-thicknes-overline-grow-direction-expected.html (rev 0)
+++ trunk/LayoutTests/fast/css3-text/css3-text-decoration/text-decoration-thicknes-overline-grow-direction-expected.html 2018-12-18 23:38:54 UTC (rev 239357)
@@ -0,0 +1,10 @@
+
+
+
+
+
+
+This test makes sure that overlines grow upward. The test passes if the top line below has the default thickness.
+
+
+
Added: trunk/LayoutTests/fast/css3-text/css3-text-decoration/text-decoration-thicknes-overline-grow-direction.html (0 => 239357)
--- trunk/LayoutTests/fast/css3-text/css3-text-decoration/text-decoration-thicknes-overline-grow-direction.html (rev 0)
+++ trunk/LayoutTests/fast/css3-text/css3-text-decoration/text-decoration-thicknes-overline-grow-direction.html 2018-12-18 23:38:54 UTC (rev 239357)
@@ -0,0 +1,10 @@
+
+
+
+
+
+
+This test makes sure that overlines grow upward. The test passes if the top line below has the default thickness.
+
+
+
Modified: trunk/Source/WebCore/ChangeLog (239356 => 239357)
--- trunk/Source/WebCore/ChangeLog 2018-12-18 23:02:16 UTC (rev 239356)
+++ trunk/Source/WebCore/ChangeLog 2018-12-18 23:38:54 UTC (rev 239357)
@@ -1,3 +1,19 @@
+2018-12-18 Myles C. Maxfield
+
+Thick overlines and line-throughs grow in the wrong direction
+https://bugs.webkit.org/show_bug.cgi?id=192264
+
+Reviewed by Dean Jackson.
+
+Overlines should grow upward, and line-throughs should stay centered.
+
+Test: fast/css3-text/css3-text-decoration/text-decoration-thicknes-overline-grow-direction.html
+
+* rendering/TextDecorationPainter.cpp:
+(WebCore::TextDecorationPainter::paintTextDecoration):
+* style/InlineTextBoxStyle.cpp:
+(WebCore::visualOverflowForDecorations):
+
2018-12-18 Ryosuke Niwa
Some iOS app crash in FrameLoader::checkCompleted
Modified: trunk/Source/WebCore/rendering/TextDecorationPainter.cpp (239356 => 239357)
--- trunk/Source/WebCore/rendering/TextDecorationPainter.cpp 2018-12-18 23:02:16 UTC (rev 239356)
+++ trunk/Source/WebCore/rendering/TextDecorationPainter.cpp 2018-12-18 23:38:54 UTC (rev 239357)
@@ -296,12 +296,15 @@
if (m_decorations.contains(TextDecoration::Overline)) {
float wavyOffset = m_styles.overlineStyle == TextDecorationStyle::Wavy ? m_wavyOffset : 0;
FloatRect rect(localOrigin, FloatSize(m_width, textDecorationThickness));
-rect.move(0, -wavyOffset);
+float autoTextDecorationThickness =
[webkit-changes] [239356] trunk/Tools
Title: [239356] trunk/Tools
Revision 239356
Author achristen...@apple.com
Date 2018-12-18 15:02:16 -0800 (Tue, 18 Dec 2018)
Log Message
Fix API test introduced in r239339 on iOS.
https://bugs.webkit.org/show_bug.cgi?id=192675
* TestWebKitAPI/Tests/WebKitCocoa/SafeBrowsing.mm:
(TEST):
The iOS implementation of the safe browsing warning is a little different.
It does some element creation when it is added to the window so it knows how big it is.
To test these elements, we need to simulate adding it to the window.
A similar technique is done in other iOS API tests, such as SafeBrowsing.ShowWarningSPI.
Modified Paths
trunk/Tools/ChangeLog
trunk/Tools/TestWebKitAPI/Tests/WebKitCocoa/SafeBrowsing.mm
Diff
Modified: trunk/Tools/ChangeLog (239355 => 239356)
--- trunk/Tools/ChangeLog 2018-12-18 22:24:33 UTC (rev 239355)
+++ trunk/Tools/ChangeLog 2018-12-18 23:02:16 UTC (rev 239356)
@@ -1,3 +1,15 @@
+2018-12-18 Alex Christensen
+
+Fix API test introduced in r239339 on iOS.
+https://bugs.webkit.org/show_bug.cgi?id=192675
+
+* TestWebKitAPI/Tests/WebKitCocoa/SafeBrowsing.mm:
+(TEST):
+The iOS implementation of the safe browsing warning is a little different.
+It does some element creation when it is added to the window so it knows how big it is.
+To test these elements, we need to simulate adding it to the window.
+A similar technique is done in other iOS API tests, such as SafeBrowsing.ShowWarningSPI.
+
2018-12-18 Jonathan Bedard
webkitpy: Ignore device type for test when using --force
Modified: trunk/Tools/TestWebKitAPI/Tests/WebKitCocoa/SafeBrowsing.mm (239355 => 239356)
--- trunk/Tools/TestWebKitAPI/Tests/WebKitCocoa/SafeBrowsing.mm 2018-12-18 22:24:33 UTC (rev 239355)
+++ trunk/Tools/TestWebKitAPI/Tests/WebKitCocoa/SafeBrowsing.mm 2018-12-18 23:02:16 UTC (rev 239356)
@@ -318,6 +318,9 @@
[webView loadHTMLString:@"alert('loaded')" baseURL:simpleURL.get()];
while (![webView _safeBrowsingWarning])
TestWebKitAPI::Util::spinRunLoop();
+#if !PLATFORM(MAC)
+[[webView _safeBrowsingWarning] didMoveToWindow];
+#endif
visitUnsafeSite([webView _safeBrowsingWarning]);
TestWebKitAPI::Util::run();
EXPECT_FALSE(!![webView _safeBrowsingWarning]);
@@ -326,6 +329,9 @@
[webView evaluateJavaScript:[NSString stringWithFormat:@"window.location='%@'", simple2URL.get()] completionHandler:nil];
while (![webView _safeBrowsingWarning])
TestWebKitAPI::Util::spinRunLoop();
+#if !PLATFORM(MAC)
+[[webView _safeBrowsingWarning] didMoveToWindow];
+#endif
return webView;
};
___
webkit-changes mailing list
webkit-changes@lists.webkit.org
https://lists.webkit.org/mailman/listinfo/webkit-changes
[webkit-changes] [239355] trunk
Title: [239355] trunk
Revision 239355
Author mark@apple.com
Date 2018-12-18 14:24:33 -0800 (Tue, 18 Dec 2018)
Log Message
JSON.stringify() should throw OOM on StringBuilder overflows.
https://bugs.webkit.org/show_bug.cgi?id=192822
Reviewed by Saam Barati.
JSTests:
* stress/json-stringify-string-builder-overflow.js: Added.
Source/_javascript_Core:
* runtime/JSONObject.cpp:
(JSC::Stringifier::stringify):
(JSC::Stringifier::appendStringifiedValue):
(JSC::Stringifier::Holder::appendNextProperty):
Modified Paths
trunk/JSTests/ChangeLog
trunk/Source/_javascript_Core/ChangeLog
trunk/Source/_javascript_Core/runtime/JSONObject.cpp
Added Paths
trunk/JSTests/stress/json-stringify-string-builder-overflow.js
Diff
Modified: trunk/JSTests/ChangeLog (239354 => 239355)
--- trunk/JSTests/ChangeLog 2018-12-18 22:14:20 UTC (rev 239354)
+++ trunk/JSTests/ChangeLog 2018-12-18 22:24:33 UTC (rev 239355)
@@ -1,3 +1,13 @@
+2018-12-18 Mark Lam
+
+JSON.stringify() should throw OOM on StringBuilder overflows.
+https://bugs.webkit.org/show_bug.cgi?id=192822
+
+
+Reviewed by Saam Barati.
+
+* stress/json-stringify-string-builder-overflow.js: Added.
+
2018-12-18 Ross Kirsling
Redeclaration of var over let/const/class should be a syntax error.
Added: trunk/JSTests/stress/json-stringify-string-builder-overflow.js (0 => 239355)
--- trunk/JSTests/stress/json-stringify-string-builder-overflow.js (rev 0)
+++ trunk/JSTests/stress/json-stringify-string-builder-overflow.js 2018-12-18 22:24:33 UTC (rev 239355)
@@ -0,0 +1,29 @@
+//@ slow
+//@ skip if $architecture != "arm64" and $architecture != "x86-64"
+
+var exception;
+try {
+var str = JSON.stringify({
+'a1': {
+'a2': {
+'a3': {
+'a4': {
+'a5': {
+'a6': 'AA'
+}
+}
+}
+}
+}
+}, function (key, value) {
+var val = {
+'A': true,
+};
+return val;
+}, 1);
+} catch (e) {
+exception = e;
+}
+
+if (exception != "Error: Out of memory")
+throw "FAILED";
Modified: trunk/Source/_javascript_Core/ChangeLog (239354 => 239355)
--- trunk/Source/_javascript_Core/ChangeLog 2018-12-18 22:14:20 UTC (rev 239354)
+++ trunk/Source/_javascript_Core/ChangeLog 2018-12-18 22:24:33 UTC (rev 239355)
@@ -1,3 +1,16 @@
+2018-12-18 Mark Lam
+
+JSON.stringify() should throw OOM on StringBuilder overflows.
+https://bugs.webkit.org/show_bug.cgi?id=192822
+
+
+Reviewed by Saam Barati.
+
+* runtime/JSONObject.cpp:
+(JSC::Stringifier::stringify):
+(JSC::Stringifier::appendStringifiedValue):
+(JSC::Stringifier::Holder::appendNextProperty):
+
2018-12-18 Ross Kirsling
Redeclaration of var over let/const/class should be a syntax error.
Modified: trunk/Source/_javascript_Core/runtime/JSONObject.cpp (239354 => 239355)
--- trunk/Source/_javascript_Core/runtime/JSONObject.cpp 2018-12-18 22:14:20 UTC (rev 239354)
+++ trunk/Source/_javascript_Core/runtime/JSONObject.cpp 2018-12-18 22:24:33 UTC (rev 239355)
@@ -269,7 +269,7 @@
JSObject* object = nullptr;
if (isCallableReplacer()) {
object = constructEmptyObject(m_exec);
-RETURN_IF_EXCEPTION(scope, jsNull());
+RETURN_IF_EXCEPTION(scope, jsUndefined());
object->putDirect(vm, vm.propertyNames->emptyIdentifier, value);
}
@@ -276,10 +276,13 @@
StringBuilder result(StringBuilder::OverflowHandler::RecordOverflow);
Holder root(Holder::RootHolder, object);
auto stringifyResult = appendStringifiedValue(result, value, root, emptyPropertyName);
-EXCEPTION_ASSERT(!scope.exception() || (stringifyResult != StringifySucceeded));
+RETURN_IF_EXCEPTION(scope, jsUndefined());
+if (UNLIKELY(result.hasOverflowed())) {
+throwOutOfMemoryError(m_exec, scope);
+return jsUndefined();
+}
if (UNLIKELY(stringifyResult != StringifySucceeded))
return jsUndefined();
-
RELEASE_AND_RETURN(scope, jsString(m_exec, result.toString()));
}
@@ -359,10 +362,6 @@
const String& string = asString(value)->value(m_exec);
RETURN_IF_EXCEPTION(scope, StringifyFailed);
builder.appendQuotedJSONString(string);
-if (UNLIKELY(builder.hasOverflowed())) {
-throwOutOfMemoryError(m_exec, scope);
-return StringifyFailed;
-}
return StringifySucceeded;
}
@@ -391,6 +390,9 @@
return StringifyFailedDueToUndefinedOrSymbolValue;
}
+if (UNLIKELY(builder.hasOverflowed()))
+return StringifyFailed;
+
// Handle cycle detection, and put the holder on the stack.
for (unsigned i = 0; i < m_holderStack.size(); i++) {
if (m_holderStack[i].object() == object) {
@@ -410,6 +412,8
[webkit-changes] [239354] trunk
Title: [239354] trunk
Revision 239354
Author ross.kirsl...@sony.com
Date 2018-12-18 14:14:20 -0800 (Tue, 18 Dec 2018)
Log Message
Redeclaration of var over let/const/class should be a syntax error.
https://bugs.webkit.org/show_bug.cgi?id=192298
Reviewed by Keith Miller.
JSTests:
* test262.yaml:
* test262/expectations.yaml:
Mark 46 tests as passing.
* stress/block-scope-redeclarations.js:
Add some new tests.
* stress/for-in-invalidate-context-weird-assignments.js:
* stress/for-in-tests.js:
Replace tests for outdated behavior with tests for SyntaxError.
* ChakraCore/test/LetConst/defer3.baseline-jsc:
* ChakraCore/test/LetConst/letvar.baseline-jsc:
Update expectations.
Source/_javascript_Core:
>From https://tc39.github.io/ecma262/#sec-block-static-semantics-early-errors:
It is a Syntax Error if any element of the LexicallyDeclaredNames of StatementList also occurs in the
VarDeclaredNames of StatementList.
Accordingly, this patch ensures that { let x; { var x; } } and { { var x; } let x; } are syntax errors.
For the "var after" scenario:
When checking for existing lexically-declared names, we can't simply check the current var scope;
we need to check *all* enclosing scopes up to (and including) the current var scope. In so doing,
we must also avoid violating the Annex B.3.5 condition that allows `try {} catch (e) { var e; }`.
For the "var before" scenario:
We ensure that lexical scopes always keep track of the vars being hoisted over them; this gives us
a simple way to check the current block's var-declared names prior to making a lexical declaration.
* parser/Parser.cpp:
(JSC::Parser::parseTryStatement):
* parser/Parser.h:
(JSC::Scope::Scope):
(JSC::Scope::setIsSimpleCatchParameterScope): Added.
(JSC::Scope::isSimpleCatchParameterScope): Added.
(JSC::Scope::declareVariable):
(JSC::Scope::addVariableBeingHoisted): Added.
(JSC::Scope::declareLexicalVariable):
(JSC::Scope::hasDeclaredVariable):
(JSC::Scope::hasLexicallyDeclaredVariable): Added.
(JSC::Parser::declareHoistedVariable): Added.
(JSC::Parser::declareVariable):
Modified Paths
trunk/JSTests/ChakraCore/test/LetConst/defer3.baseline-jsc
trunk/JSTests/ChakraCore/test/LetConst/letvar.baseline-jsc
trunk/JSTests/ChangeLog
trunk/JSTests/stress/for-in-invalidate-context-weird-assignments.js
trunk/JSTests/stress/for-in-tests.js
trunk/JSTests/test262/expectations.yaml
trunk/JSTests/test262.yaml
trunk/Source/_javascript_Core/ChangeLog
trunk/Source/_javascript_Core/parser/Parser.cpp
trunk/Source/_javascript_Core/parser/Parser.h
Added Paths
trunk/JSTests/stress/block-scope-redeclarations.js
Diff
Modified: trunk/JSTests/ChakraCore/test/LetConst/defer3.baseline-jsc (239353 => 239354)
--- trunk/JSTests/ChakraCore/test/LetConst/defer3.baseline-jsc 2018-12-18 22:03:07 UTC (rev 239353)
+++ trunk/JSTests/ChakraCore/test/LetConst/defer3.baseline-jsc 2018-12-18 22:14:20 UTC (rev 239354)
@@ -28,11 +28,11 @@
SyntaxError: Cannot declare a let variable twice: 'x'.
SyntaxError: Cannot declare a var variable that shadows a let/const/class variable: 'x'.
SyntaxError: Cannot declare a var variable that shadows a let/const/class variable: 'x'.
-Syntax check succeeded
-Syntax check succeeded
-Syntax check succeeded
-Syntax check succeeded
SyntaxError: Cannot declare a const variable twice: 'x'.
SyntaxError: Cannot declare a let variable twice: 'x'.
SyntaxError: Cannot declare a var variable that shadows a let/const/class variable: 'x'.
SyntaxError: Cannot declare a var variable that shadows a let/const/class variable: 'x'.
+SyntaxError: Cannot declare a const variable twice: 'x'.
+SyntaxError: Cannot declare a let variable twice: 'x'.
+SyntaxError: Cannot declare a var variable that shadows a let/const/class variable: 'x'.
+SyntaxError: Cannot declare a var variable that shadows a let/const/class variable: 'x'.
Modified: trunk/JSTests/ChakraCore/test/LetConst/letvar.baseline-jsc (239353 => 239354)
--- trunk/JSTests/ChakraCore/test/LetConst/letvar.baseline-jsc 2018-12-18 22:03:07 UTC (rev 239353)
+++ trunk/JSTests/ChakraCore/test/LetConst/letvar.baseline-jsc 2018-12-18 22:14:20 UTC (rev 239354)
@@ -1,8 +1,3 @@
SyntaxError: Cannot declare a var variable that shadows a let/const/class variable: 'x'.
-var x
-undefined
-let y
-var y
-undefined
-let x
-undefined
+SyntaxError: Cannot declare a var variable that shadows a let/const/class variable: 'x'.
+SyntaxError: Cannot declare a let variable twice: 'x'.
Modified: trunk/JSTests/ChangeLog (239353 => 239354)
--- trunk/JSTests/ChangeLog 2018-12-18 22:03:07 UTC (rev 239353)
+++ trunk/JSTests/ChangeLog 2018-12-18 22:14:20 UTC (rev 239354)
@@ -1,3 +1,25 @@
+2018-12-18 Ross Kirsling
+
+Redeclaration of var over let/const/class should be a syntax error.
+https://bugs.webkit.org/show_bug.cgi?id=192298
+
+Reviewed by Keith Miller.
+
+* test262.yaml:
+* test262/expectations.yaml:
+Mark 46 tests as passing.
+
+*
[webkit-changes] [239353] trunk/Source/WebCore
Title: [239353] trunk/Source/WebCore
Revision 239353
Author rn...@webkit.org
Date 2018-12-18 14:03:07 -0800 (Tue, 18 Dec 2018)
Log Message
Some iOS app crash in FrameLoader::checkCompleted
https://bugs.webkit.org/show_bug.cgi?id=192804
Reviewed by Tim Horton.
It's possible for the main thread to call into WebCore / UIWebView selectors while Web thread
is trying to send a delegate message. Disable the release assertion while this is happening
so that iOS app would not crash.
Unfortunately no new test as there is no way to easily test UIWebView in iOS,
and this requires a race between the web thread & the main thread.
* dom/ScriptDisallowedScope.h:
(WebCore::ScriptDisallowedScope::InMainThread::isScriptAllowed):
* platform/ios/wak/WebCoreThread.h:
* platform/ios/wak/WebCoreThread.mm:
(WebThreadDelegateMessageScope::WebThreadDelegateMessageScope):
(WebThreadDelegateMessageScope::~WebThreadDelegateMessageScope):
(SendDelegateMessage):
Modified Paths
trunk/Source/WebCore/ChangeLog
trunk/Source/WebCore/dom/ScriptDisallowedScope.h
trunk/Source/WebCore/platform/ios/wak/WebCoreThread.h
trunk/Source/WebCore/platform/ios/wak/WebCoreThread.mm
Diff
Modified: trunk/Source/WebCore/ChangeLog (239352 => 239353)
--- trunk/Source/WebCore/ChangeLog 2018-12-18 21:54:24 UTC (rev 239352)
+++ trunk/Source/WebCore/ChangeLog 2018-12-18 22:03:07 UTC (rev 239353)
@@ -1,3 +1,26 @@
+2018-12-18 Ryosuke Niwa
+
+Some iOS app crash in FrameLoader::checkCompleted
+https://bugs.webkit.org/show_bug.cgi?id=192804
+
+
+Reviewed by Tim Horton.
+
+It's possible for the main thread to call into WebCore / UIWebView selectors while Web thread
+is trying to send a delegate message. Disable the release assertion while this is happening
+so that iOS app would not crash.
+
+Unfortunately no new test as there is no way to easily test UIWebView in iOS,
+and this requires a race between the web thread & the main thread.
+
+* dom/ScriptDisallowedScope.h:
+(WebCore::ScriptDisallowedScope::InMainThread::isScriptAllowed):
+* platform/ios/wak/WebCoreThread.h:
+* platform/ios/wak/WebCoreThread.mm:
+(WebThreadDelegateMessageScope::WebThreadDelegateMessageScope):
+(WebThreadDelegateMessageScope::~WebThreadDelegateMessageScope):
+(SendDelegateMessage):
+
2018-12-18 David Kilzer
clang-tidy: Use const reference for MediaTime parameter to prevent object copy
Modified: trunk/Source/WebCore/dom/ScriptDisallowedScope.h (239352 => 239353)
--- trunk/Source/WebCore/dom/ScriptDisallowedScope.h 2018-12-18 21:54:24 UTC (rev 239352)
+++ trunk/Source/WebCore/dom/ScriptDisallowedScope.h 2018-12-18 22:03:07 UTC (rev 239353)
@@ -26,6 +26,10 @@
#include "ContainerNode.h"
#include
+#if PLATFORM(IOS_FAMILY)
+#include "WebCoreThread.h"
+#endif
+
namespace WebCore {
class ScriptDisallowedScope {
@@ -86,7 +90,11 @@
static bool isScriptAllowed()
{
ASSERT(isMainThread());
+#if PLATFORM(IOS_FAMILY)
+return !s_count || webThreadDelegateMessageScopeCount;
+#else
return !s_count;
+#endif
}
};
Modified: trunk/Source/WebCore/platform/ios/wak/WebCoreThread.h (239352 => 239353)
--- trunk/Source/WebCore/platform/ios/wak/WebCoreThread.h 2018-12-18 21:54:24 UTC (rev 239352)
+++ trunk/Source/WebCore/platform/ios/wak/WebCoreThread.h 2018-12-18 22:03:07 UTC (rev 239353)
@@ -44,6 +44,7 @@
} WebThreadContext;
extern volatile bool webThreadShouldYield;
+extern volatile unsigned webThreadDelegateMessageScopeCount;
#ifdef __OBJC__
@class NSRunLoop;
Modified: trunk/Source/WebCore/platform/ios/wak/WebCoreThread.mm (239352 => 239353)
--- trunk/Source/WebCore/platform/ios/wak/WebCoreThread.mm 2018-12-18 21:54:24 UTC (rev 239352)
+++ trunk/Source/WebCore/platform/ios/wak/WebCoreThread.mm 2018-12-18 22:03:07 UTC (rev 239353)
@@ -132,6 +132,8 @@
static NSMutableArray* sAsyncDelegates = nil;
+WEBCORE_EXPORT volatile unsigned webThreadDelegateMessageScopeCount = 0;
+
static inline void SendMessage(NSInvocation* invocation)
{
[invocation invoke];
@@ -171,6 +173,16 @@
#endif
}
+class WebThreadDelegateMessageScope {
+public:
+WebThreadDelegateMessageScope() { ++webThreadDelegateMessageScopeCount; }
+~WebThreadDelegateMessageScope()
+{
+ASSERT(webThreadDelegateMessageScopeCount);
+--webThreadDelegateMessageScopeCount;
+}
+};
+
static void SendDelegateMessage(NSInvocation* invocation)
{
if (!WebThreadIsCurrent()) {
@@ -194,6 +206,7 @@
#endif
{
+WebThreadDelegateMessageScope delegateScope;
// Code block created to scope JSC::JSLock::DropAllLocks outside of WebThreadLock()
JSC::JSLock::DropAllLocks dropAllLocks(WebCore::commonVM());
_WebThreadUnlock();
___
webkit-changes mailing list
[webkit-changes] [239352] trunk/Source
Title: [239352] trunk/Source
Revision 239352
Author ddkil...@apple.com
Date 2018-12-18 13:54:24 -0800 (Tue, 18 Dec 2018)
Log Message
clang-tidy: Use const reference for MediaTime parameter to prevent object copy
Reviewed by Mark Lam.
Source/_javascript_Core:
* runtime/JSCJSValue.h:
(JSC::jsNumber):
Source/WebCore:
* bindings/js/JSDOMConvertNumbers.h:
(WebCore::JSConverter::convert):
Modified Paths
trunk/Source/_javascript_Core/ChangeLog
trunk/Source/_javascript_Core/runtime/JSCJSValue.h
trunk/Source/WebCore/ChangeLog
trunk/Source/WebCore/bindings/js/JSDOMConvertNumbers.h
Diff
Modified: trunk/Source/_javascript_Core/ChangeLog (239351 => 239352)
--- trunk/Source/_javascript_Core/ChangeLog 2018-12-18 21:51:56 UTC (rev 239351)
+++ trunk/Source/_javascript_Core/ChangeLog 2018-12-18 21:54:24 UTC (rev 239352)
@@ -1,3 +1,13 @@
+2018-12-18 David Kilzer
+
+clang-tidy: Use const reference for MediaTime parameter to prevent object copy
+
+
+Reviewed by Mark Lam.
+
+* runtime/JSCJSValue.h:
+(JSC::jsNumber):
+
2018-12-18 Saam Barati
Unreviewed. Appease a build error where we weren't using a variable defined in an "if"
Modified: trunk/Source/_javascript_Core/runtime/JSCJSValue.h (239351 => 239352)
--- trunk/Source/_javascript_Core/runtime/JSCJSValue.h 2018-12-18 21:51:56 UTC (rev 239351)
+++ trunk/Source/_javascript_Core/runtime/JSCJSValue.h 2018-12-18 21:54:24 UTC (rev 239352)
@@ -551,7 +551,7 @@
return JSValue(d);
}
-ALWAYS_INLINE JSValue jsNumber(MediaTime t)
+ALWAYS_INLINE JSValue jsNumber(const MediaTime& t)
{
return jsNumber(t.toDouble());
}
Modified: trunk/Source/WebCore/ChangeLog (239351 => 239352)
--- trunk/Source/WebCore/ChangeLog 2018-12-18 21:51:56 UTC (rev 239351)
+++ trunk/Source/WebCore/ChangeLog 2018-12-18 21:54:24 UTC (rev 239352)
@@ -1,3 +1,13 @@
+2018-12-18 David Kilzer
+
+clang-tidy: Use const reference for MediaTime parameter to prevent object copy
+
+
+Reviewed by Mark Lam.
+
+* bindings/js/JSDOMConvertNumbers.h:
+(WebCore::JSConverter::convert):
+
2018-12-18 Justin Fan
[WebGPU] BindGroupLayout and Device::createBindGroupLayout
Modified: trunk/Source/WebCore/bindings/js/JSDOMConvertNumbers.h (239351 => 239352)
--- trunk/Source/WebCore/bindings/js/JSDOMConvertNumbers.h 2018-12-18 21:51:56 UTC (rev 239351)
+++ trunk/Source/WebCore/bindings/js/JSDOMConvertNumbers.h 2018-12-18 21:54:24 UTC (rev 239352)
@@ -389,7 +389,7 @@
}
// Add overload for MediaTime.
-static JSC::JSValue convert(MediaTime value)
+static JSC::JSValue convert(const MediaTime& value)
{
return JSC::jsNumber(JSC::purifyNaN(value.toDouble()));
}
___
webkit-changes mailing list
webkit-changes@lists.webkit.org
https://lists.webkit.org/mailman/listinfo/webkit-changes
[webkit-changes] [239351] trunk
Title: [239351] trunk
Revision 239351
Author justin_...@apple.com
Date 2018-12-18 13:51:56 -0800 (Tue, 18 Dec 2018)
Log Message
[WebGPU] BindGroupLayout and Device::createBindGroupLayout
https://bugs.webkit.org/show_bug.cgi?id=192817
Reviewed by Dean Jackson.
Update bind-group-layouts to test new functionality.
Source/WebCore:
Implement the emtpy WebGPUBindGroupLayout interface, and enable creation via WebGPUDevice::createBindGroupLayout:
* Modules/webgpu/WebGPUBindGroupLayout.cpp: Added.
(WebCore::WebGPUBindGroupLayout::create):
(WebCore::WebGPUBindGroupLayout::WebGPUBindGroupLayout):
* Modules/webgpu/WebGPUBindGroupLayout.h: Added.
* Modules/webgpu/WebGPUBindGroupLayout.idl: Added. Empty interface for now.
* Modules/webgpu/WebGPUDevice.cpp:
(WebCore::WebGPUDevice::createBindGroupLayout const): Added.
* Modules/webgpu/WebGPUDevice.h:
* Modules/webgpu/WebGPUDevice.idl:
* platform/graphics/gpu/GPUBindGroupLayout.cpp: Added.
(WebCore::GPUBindGroupLayout::tryCreate):
(WebCore::GPUBindGroupLayout::GPUBindGroupLayout):
* platform/graphics/gpu/GPUBindGroupLayout.h: Added.
* platform/graphics/gpu/GPUDevice.cpp:
(WebCore::GPUDevice::tryCreateBindGroupLayout const): Added.
* platform/graphics/gpu/GPUDevice.h:
Add files and symbols to project:
* CMakeLists.txt:
* DerivedSources.make:
* Sources.txt:
* WebCore.xcodeproj/project.pbxproj:
* bindings/js/WebCoreBuiltinNames.h:
Missing includes that were previously provided via UnifiedSources:
* platform/mediarecorder/cocoa/MediaRecorderPrivateWriterCocoa.h:
* platform/sql/SQLiteFileSystem.h:
LayoutTests:
* webgpu/bind-group-layouts-expected.txt:
* webgpu/bind-group-layouts.html:
Modified Paths
trunk/LayoutTests/ChangeLog
trunk/LayoutTests/webgpu/bind-group-layouts-expected.txt
trunk/LayoutTests/webgpu/bind-group-layouts.html
trunk/Source/WebCore/CMakeLists.txt
trunk/Source/WebCore/ChangeLog
trunk/Source/WebCore/DerivedSources.make
trunk/Source/WebCore/Modules/webgpu/WebGPUDevice.cpp
trunk/Source/WebCore/Modules/webgpu/WebGPUDevice.h
trunk/Source/WebCore/Modules/webgpu/WebGPUDevice.idl
trunk/Source/WebCore/Sources.txt
trunk/Source/WebCore/WebCore.xcodeproj/project.pbxproj
trunk/Source/WebCore/bindings/js/WebCoreBuiltinNames.h
trunk/Source/WebCore/platform/graphics/gpu/GPUDevice.cpp
trunk/Source/WebCore/platform/graphics/gpu/GPUDevice.h
trunk/Source/WebCore/platform/mediarecorder/cocoa/MediaRecorderPrivateWriterCocoa.h
trunk/Source/WebCore/platform/sql/SQLiteFileSystem.h
Added Paths
trunk/Source/WebCore/Modules/webgpu/WebGPUBindGroupLayout.cpp
trunk/Source/WebCore/Modules/webgpu/WebGPUBindGroupLayout.h
trunk/Source/WebCore/Modules/webgpu/WebGPUBindGroupLayout.idl
trunk/Source/WebCore/platform/graphics/gpu/GPUBindGroupLayout.cpp
trunk/Source/WebCore/platform/graphics/gpu/GPUBindGroupLayout.h
Diff
Modified: trunk/LayoutTests/ChangeLog (239350 => 239351)
--- trunk/LayoutTests/ChangeLog 2018-12-18 21:51:06 UTC (rev 239350)
+++ trunk/LayoutTests/ChangeLog 2018-12-18 21:51:56 UTC (rev 239351)
@@ -1,3 +1,15 @@
+2018-12-18 Justin Fan
+
+[WebGPU] BindGroupLayout and Device::createBindGroupLayout
+https://bugs.webkit.org/show_bug.cgi?id=192817
+
+Reviewed by Dean Jackson.
+
+Update bind-group-layouts to test new functionality.
+
+* webgpu/bind-group-layouts-expected.txt:
+* webgpu/bind-group-layouts.html:
+
2018-12-18 Ryan Haddad
Unreviewed test gardening after r239306 and r239338.
Modified: trunk/LayoutTests/webgpu/bind-group-layouts-expected.txt (239350 => 239351)
--- trunk/LayoutTests/webgpu/bind-group-layouts-expected.txt 2018-12-18 21:51:06 UTC (rev 239350)
+++ trunk/LayoutTests/webgpu/bind-group-layouts-expected.txt 2018-12-18 21:51:56 UTC (rev 239351)
@@ -1,3 +1,4 @@
-PASS Create WebGPUBindGroupLayoutDescriptor.
+PASS Create a basic WebGPUBindGroupLayoutDescriptor.
+PASS Create a basic WebGPUBindGroupLayout with a WebGPUDevice.
Modified: trunk/LayoutTests/webgpu/bind-group-layouts.html (239350 => 239351)
--- trunk/LayoutTests/webgpu/bind-group-layouts.html 2018-12-18 21:51:06 UTC (rev 239350)
+++ trunk/LayoutTests/webgpu/bind-group-layouts.html 2018-12-18 21:51:56 UTC (rev 239351)
@@ -1,7 +1,8 @@
-Create WebGPUBindGroupLayoutDescriptor.
+Create WebGPUBindGroupLayout.
+