[webkit-changes] [WebKit/WebKit] 531333: Versioning.

MyahCobbs noreply at github.com
Mon Nov 27 16:35:06 PST 2023


  Branch: refs/heads/safari-7616.1.21-branch
  Home:   https://github.com/WebKit/WebKit
  Commit: 531333346b57efbe2c97197307b45c801b55aae3
      https://github.com/WebKit/WebKit/commit/531333346b57efbe2c97197307b45c801b55aae3
  Author: Russell Epstein <repstein at apple.com>
  Date:   2023-06-26 (Mon, 26 Jun 2023)

  Changed paths:
    M Configurations/Version.xcconfig

  Log Message:
  -----------
  Versioning.

WebKit-7616.1.21

Canonical link: https://commits.webkit.org/265500.1@safari-7616.1.21-branch


  Commit: 055a74191a3e9275077b1ab80cfcc68e97b91310
      https://github.com/WebKit/WebKit/commit/055a74191a3e9275077b1ab80cfcc68e97b91310
  Author: Chris Dumez <cdumez at apple.com>
  Date:   2023-06-26 (Mon, 26 Jun 2023)

  Changed paths:
    M Source/WebKit/Shared/Cocoa/ArgumentCodersCocoa.mm
    M Tools/TestWebKitAPI/Tests/WebKitCocoa/IPCTestingAPI.mm

  Log Message:
  -----------
  Cherry-pick 0ba9e6c4473f. rdar://problem/111161160

    WebPage_LoadRequest IPC fails decoding in PingDuoDuo app
    https://bugs.webkit.org/show_bug.cgi?id=258486
    rdar://111161160

    Reviewed by Tim Horton.

    WebPage_LoadRequest IPC was failing decoding in PingDuoDuo app.
    The issue was due to getting a WKSecureCodingURLWrapper instead
    of a NSURL when decoding the baseURL of a NSURL.

    I am not sure how we ended up in this situation but I made the
    bug go away by simplifying the code. The coder used to encode
    the URL in two parts:
    1. The baseURL
    2. The bytes from the URL's relative string

    Then it would decode the URL in 2 parts:
    1. The baseURL
    2. The bytes from the URL's relative string

    It would then call CFURLCreateAbsoluteURLWithBytes() with those
    2 parts, which would result in an *absolute* URL. The information
    about baseURL / relative string would be lost.

    As a result, I have decided to simply encode the URL in one part,
    the absolute URL bytes. The decoding results ends up being the
    same (an absolute URL). It simplifies both coding and decoding
    and makes the bug go away since it was about decoding baseURLs.

    * Source/WebKit/Shared/Cocoa/ArgumentCodersCocoa.mm:
    (-[WKSecureCodingURLWrapper encodeWithCoder:]):
    (-[WKSecureCodingURLWrapper initWithCoder:]):
    * Tools/TestWebKitAPI/Tests/WebKitCocoa/IPCTestingAPI.mm:

    Canonical link: https://commits.webkit.org/265529@main

Canonical link: https://commits.webkit.org/265500.2@safari-7616.1.21-branch


  Commit: eb558de512214f96fc2687c5b07fd6b49970262f
      https://github.com/WebKit/WebKit/commit/eb558de512214f96fc2687c5b07fd6b49970262f
  Author: Per Arne Vollan <pvollan at apple.com>
  Date:   2023-06-27 (Tue, 27 Jun 2023)

  Changed paths:
    M Source/WebKit/UIProcess/mac/WebViewImpl.mm

  Log Message:
  -----------
  Cherry-pick b2a27f1fe286. rdar://problem/111347166

    Fix null pointer dereference
    https://bugs.webkit.org/show_bug.cgi?id=258529
    rdar://111347166

    Reviewed by Brent Fulgham.

    The 'configuration' parameter to WebViewImpl::WebViewImpl has been moved at this point. Use the configuration from the page instead.

    * Source/WebKit/UIProcess/mac/WebViewImpl.mm:
    (WebKit::WebViewImpl::WebViewImpl):

    Canonical link: https://commits.webkit.org/265528@main
Identifier: 265423.80 at safari-7616.1.21-branch


  Commit: 69354feb53560e265a9f10d859232363e2014eae
      https://github.com/WebKit/WebKit/commit/69354feb53560e265a9f10d859232363e2014eae
  Author: Alex Christensen <achristensen at apple.com>
  Date:   2023-06-28 (Wed, 28 Jun 2023)

  Changed paths:
    M Source/WTF/wtf/ThreadSafeWeakPtr.h

  Log Message:
  -----------
  Cherry-pick 883a699fc0cf. rdar://problem/111409234

    Revert unnecessary part of 265344 at main
    https://bugs.webkit.org/show_bug.cgi?id=258636
    rdar://111409234

    Reviewed by Chris Dumez.

    265344 at main made it so we can make weak references during the destructor safely.
    I also tried to do the same thing we do in RefCounted and ThreadSafeRefCounted
    and make it so we can safely make strong references, however m_strongReferenceCount
    is also used in weakDeref so it caused the control block to be leaked if there
    are weak references when the destructor is called.  As a minimum risk fix,
    revert that unneeded part of last week's PR.

    * Source/WTF/wtf/ThreadSafeWeakPtr.h:
    (WTF::ThreadSafeWeakPtrControlBlock::strongDeref const):

    Canonical link: https://commits.webkit.org/265598@main
Identifier: 265423.81 at safari-7616.1.21-branch


  Commit: d0c88b64e7472c7a8df0f600a95c0f73fbba2b06
      https://github.com/WebKit/WebKit/commit/d0c88b64e7472c7a8df0f600a95c0f73fbba2b06
  Author: Myah Cobbs <mcobbs at apple.com>
  Date:   2023-06-29 (Thu, 29 Jun 2023)

  Changed paths:
    M Configurations/Version.xcconfig

  Log Message:
  -----------
  Versioning.

WebKit-7616.1.21.1

Identifier: 265423.82 at safari-7616.1.21-branch


  Commit: b689b631e5ff6586f9d4f80ff08e3c2bff4332a4
      https://github.com/WebKit/WebKit/commit/b689b631e5ff6586f9d4f80ff08e3c2bff4332a4
  Author: Yusuke Suzuki <ysuzuki at apple.com>
  Date:   2023-06-30 (Fri, 30 Jun 2023)

  Changed paths:
    A JSTests/stress/int32-isnan-fold.js
    A JSTests/stress/int32-number-isnan-fold.js

  Log Message:
  -----------
  Cherry-pick eee2dfc2c470. rdar://problem/111223515

    Unreviewed, add regression tests for 265570 at main
    https://bugs.webkit.org/show_bug.cgi?id=258596
    rdar://111223515

    * JSTests/stress/int32-isnan-fold.js: Added.
    (shouldBe):
    (test):
    * JSTests/stress/int32-number-isnan-fold.js: Added.
    (shouldBe):
    (test):

    Canonical link: https://commits.webkit.org/265571@main

Identifier: 263823.1683 at safari-7616.1.21-branch


  Commit: 1c3381df9a1ebf6677285719b84d2efa15fd871f
      https://github.com/WebKit/WebKit/commit/1c3381df9a1ebf6677285719b84d2efa15fd871f
  Author: Aditya Keerthi <akeerthi at apple.com>
  Date:   2023-06-30 (Fri, 30 Jun 2023)

  Changed paths:
    M Source/WebKit/UIProcess/WebContextMenuProxy.cpp
    M Tools/TestWebKitAPI/Tests/mac/ContextMenuTests.mm

  Log Message:
  -----------
  Cherry-pick 16e3d68d600a. rdar://problem/110708014

    REGRESSION (260897 at main): Occasional crash when right clicking an image
    https://bugs.webkit.org/show_bug.cgi?id=258599
    rdar://110708014

    Reviewed by Wenson Hsieh.

    260897 at main introduced additional asynchronicity when generating context menu
    information, to perform QR code detection. Consequently, it is possible for
    the `WebPageProxy` to be destroyed prior to finishing detection. Add a
    null-check for `WebContextMenuProxy::page` to avoid crashes in this scenario.

    * Source/WebKit/UIProcess/WebContextMenuProxy.cpp:
    (WebKit::WebContextMenuProxy::show):
    * Tools/TestWebKitAPI/Tests/mac/ContextMenuTests.mm:
    (TestWebKitAPI::TEST):

    Canonical link: https://commits.webkit.org/265582@main

Identifier: 263823.1684 at safari-7616.1.21-branch


  Commit: 181c4671c750f014193e4b91e5136518cb7904e9
      https://github.com/WebKit/WebKit/commit/181c4671c750f014193e4b91e5136518cb7904e9
  Author: Chris Dumez <cdumez at apple.com>
  Date:   2023-06-30 (Fri, 30 Jun 2023)

  Changed paths:
    M Source/WebKit/UIProcess/WebAuthentication/Cocoa/WebAuthenticatorCoordinatorProxy.mm

  Log Message:
  -----------
  Cherry-pick de109a5e7580. rdar://problem/110384303

    Crash under WebAuthenticatorCoordinatorProxy::isUserVerifyingPlatformAuthenticatorAvailable(WebCore::SecurityOriginData const&, WTF::CompletionHandler<void (bool)>&&)
    https://bugs.webkit.org/show_bug.cgi?id=258646
    rdar://110384303

    Reviewed by J Pascoe.

    Add missing early return in this function. Otherwise, we may end up calling
    the handler CompletionHandler twice and we'll crash the second time. This
    is because CompletionHandler null themselves out when they get called, since
    they're meant to only be called once.

    * Source/WebKit/UIProcess/WebAuthentication/Cocoa/WebAuthenticatorCoordinatorProxy.mm:

    Canonical link: https://commits.webkit.org/265600@main

Identifier: 263823.1685 at safari-7616.1.21-branch


  Commit: d4e3f05f57684faa66c5e4165dee2eb661147138
      https://github.com/WebKit/WebKit/commit/d4e3f05f57684faa66c5e4165dee2eb661147138
  Author: Ryosuke Niwa <rniwa at webkit.org>
  Date:   2023-06-30 (Fri, 30 Jun 2023)

  Changed paths:
    M Source/WebCore/editing/TextManipulationController.cpp

  Log Message:
  -----------
  Cherry-pick 3f0a2a47e444. rdar://problem/85597139

    Crash in TextManipulationController::observeParagraphs on macrumors.com
    https://bugs.webkit.org/show_bug.cgi?id=258649

    Reviewed by Wenson Hsieh.

    In some cases, we can reach the end of shadow root without TextIterator hitting the end.
    We need to stop the iteration in such cases to avoid dereferencing nullptr.

    Unfortunately no new tests since I cannot create a reduction.

    * Source/WebCore/editing/TextManipulationController.cpp:
    (WebCore::ParagraphContentIterator::atEnd const):

    Canonical link: https://commits.webkit.org/265612@main

Identifier: 263823.1686 at safari-7616.1.21-branch


  Commit: 9aeecb6554f3c0b7a87f446293981c87c32d69be
      https://github.com/WebKit/WebKit/commit/9aeecb6554f3c0b7a87f446293981c87c32d69be
  Author: Simon Fraser <simon.fraser at apple.com>
  Date:   2023-06-30 (Fri, 30 Jun 2023)

  Changed paths:
    M Source/WebKit/Shared/WebWheelEventCoalescer.cpp
    M Source/WebKit/Shared/WebWheelEventCoalescer.h
    M Source/WebKit/UIProcess/WebPageProxy.cpp

  Log Message:
  -----------
  Cherry-pick c09328b452c7. rdar://problem/111271905

    Occasional crashes under WebWheelEventCoalescer::takeOldestEventBeingProcessed()
    https://bugs.webkit.org/show_bug.cgi?id=258653
    rdar://111271905

    Reviewed by Jer Noble.

    Crash data suggest that we can reach WebWheelEventCoalescer::takeOldestEventBeingProcessed() with
    m_eventsBeingProcessed being empty. We get here after one or more trips to the web process for wheel
    event handling, so it's possible there's some code path where we can get here with an empty m_eventsBeingProcessed,
    but I have not figured out how, so do a defensive fix of returning a std::optional<NativeWebWheelEvent>.

    * Source/WebKit/Shared/WebWheelEventCoalescer.cpp:
    (WebKit::WebWheelEventCoalescer::takeOldestEventBeingProcessed):
    * Source/WebKit/Shared/WebWheelEventCoalescer.h:
    * Source/WebKit/UIProcess/WebPageProxy.cpp:
    (WebKit::WebPageProxy::wheelEventHandlingCompleted):

    Canonical link: https://commits.webkit.org/265625@main

Identifier: 263823.1687 at safari-7616.1.21-branch


  Commit: 0061c6a0d373b206bafc2d255a954a99e0142592
      https://github.com/WebKit/WebKit/commit/0061c6a0d373b206bafc2d255a954a99e0142592
  Author: Myah Cobbs <mcobbs at apple.com>
  Date:   2023-07-02 (Sun, 02 Jul 2023)

  Changed paths:
    M Configurations/Version.xcconfig

  Log Message:
  -----------
  Versioning.

WebKit-7616.1.21.2

Identifier: 263823.1688 at safari-7616.1.21-branch


Compare: https://github.com/WebKit/WebKit/compare/531333346b57%5E...0061c6a0d373


More information about the webkit-changes mailing list