February 2022 - webkit-dev - lists.webkit.org

Smart Pointer Analysis Tool for WebKit
by Ryosuke Niwa 01 Mar '23

01 Mar '23

Hi all, I’ve been working with Geoff (ggaren) and Jan Korous from Apple's compiler team to create a static analyzer which detects dangerous use of ref counted objects, and we’re looking for folks who are interested in trying out this tool and helping us refine the tool. Please let me know if you’re interested in using this tool & try applying code changes to WebKit. Our goal is to eventually integrate this tool into buildbot and EWS so that we can catch dangerous code before they get committed. *What is Dangerous?* So *what is* a *dangerous* use of ref counted objects you may ask? It’s any use of which we can’t trivially conclude that it doesn’t lead to a use-after-free. For now, we don’t detect dangerous use of non-ref counted objects including ones that can vend WeakPtr. It’s on us, humans, to decide which objects need to be ref counted or need to be CanMakeWeakPtr. Consider the following example. This code most certainly will lead to a use-after-free of “parent” in the third line because the code doesn’t keep the parent alive. Because isContentEditable updates style in some cases, it could lead to arbitrary script execution which could remove the parent from the document. Node* parent = element->parentElement(); if (parent->isContentEditable()) parent->scrollIntoView(); In general, relying on a complex data structure such as DOM tree to keep RefCounted objects alive while we call a non-trivial function is not safe. All it takes for the code to have a use-after-free is for someone to start updating style, layout, etc… inside the function either directly or indirectly. And we don’t want to make WebKit development really hard by forcing anyone who modifies a function to check every caller of the function and their callers, etc… to make sure it’s safe to do so. For this reason, it’s *dangerous* to store a raw pointer or a reference to a ref counted object as a local variable and use it after calling a non-trivial function. We did a similar analysis of a number of other patterns and usage of ref counted objects in WebKit and came up with the following basic rules for using ref counted objects in a safe manner. We’re hoping that these rules would be eventually incorporated into our coding style guideline: https://webkit.org/code-style-guidelines/ *Rules for Using Ref Counted Objects* 1. Every data member to a ref counted object must use either Ref, RefPtr, or WeakPtr. webkit.NoUncountedMemberChecker <https://github.com/llvm/llvm-project/blob/master/clang/docs/analyzer/checke…> 2. Every ref counted base class, if it has a derived class, must define a virtual destructor. webkit.RefCntblBaseVirtualDtor <https://github.com/llvm/llvm-project/blob/master/clang/docs/analyzer/checke…> 3. Every ref counted object passed to a non-trivial function as an argument (including "this" pointer) should be stored as a Ref or RefPtr in the caller’s local scope unless it's an argument to the caller itself by transitive property [1]. alpha.webkit.UncountedCallArgsChecker <https://github.com/llvm/llvm-project/blob/master/clang/docs/analyzer/checke…> 4. Every ref counted object must be captured using Ref or RefPtr for lambda. webkit.UncountedLambdaCapturesChecker <https://github.com/llvm/llvm-project/blob/master/clang/docs/analyzer/checke…> 5. Local variables - we’re still working on this ( https://reviews.llvm.org/D83259) Below, I've dissected each one of these rules with the real warning emitted by the analysis tool in development. Please let me know if you have any comments / concerns. ------------------------------ (1) is pretty trivial. Every ref counted data member should be stored using Ref, RefPtr, or WeakPtr since it would be not trivially obvious that life cycles of two or more objects are correctly tied or managed together. ------------------------------ (2) is also pretty easy to understand. In the following example, if someone destroys an instance of B using Ref<A>, then it would result in an undefined behavior. struct A : public RefCounted<A> { Vector<int> someData; }; struct B : public A { Vector<int> otherData; }; ------------------------------ For (3), passing a ref counted object as a raw pointer or reference to a function as an argument, the tool may generate a warning like this: Source/WebCore/html/FormAssociatedElement.cpp:181:13: warning: [WebKit] call argument is a raw pointers to a ref-countable type [-Wfunction-arg-ptr-to-refcntbl] setForm(findAssociatedForm(&asHTMLElement(), originalForm.get())); ^ This warns that void setForm(HTMLFormElement*) is called with the result of findAssociatedForm, which returns HTMLFormElement* without storing it anywhere. If setForm can somehow cause HTMLFormElement to be deleted, then this can result in the use-after-free in setForm. void FormAssociatedElement::resetFormOwner() { RefPtr<HTMLFormElement> originalForm = m_form.get(); setForm(findAssociatedForm(&asHTMLElement(), originalForm.get())); HTMLElement& element = asHTMLElement(); auto* newForm = m_form.get(); if (newForm && newForm != originalForm && newForm->isConnected()) element.document().didAssociateFormControl(element); } Why, you may ask, we don't put HTMLFormElement* in a Ref or RefPtr in setForm instead? This is because a lot of code has calls to a function with a local variable. Because the callee doesn’t know whether the caller has already stored each argument in Ref / RefPtr or not, we need to be safe and store them again in Ref / RefPtr, resulting in an unnecessary ref-churn. Additionally, if we took the approach of the callee being responsible for keeping every argument alive, then every non-trivial member function of a ref counted object must have "protectedThis" at the beginning of the function, which would be particularly wasteful if those member functions are sometimes called by other member functions. Once this rule is applied everywhere, for example, we can get rid of “protectedThis” from our codebase because it would be redundant. Furthermore, this rule transitively allows a raw pointer or a reference passed in as an argument to be used as arguments to call another function without storing it in Ref or RefPtr. For example, the following function satisfies this rule because both oldDocument and newDocument are raw references passed to this function as arguments, and therefore the caller of this function must have already stored as Ref / RefPtr as local variables. void HTMLMediaElement::didMoveToNewDocument(Document& oldDocument, Document& newDocument) { ASSERT_WITH_SECURITY_IMPLICATION(&document() == &newDocument); if (m_shouldDelayLoadEvent) { oldDocument.decrementLoadEventDelayCount(); newDocument.incrementLoadEventDelayCount(); } unregisterWithDocument(oldDocument); registerWithDocument(newDocument); HTMLElement::didMoveToNewDocument(oldDocument, newDocument); updateShouldAutoplay(); } ------------------------------ For (4), capturing a ref counted object as a raw pointer or a reference, the tool may generate a warning like this: Source/WebKit/UIProcess/API/APIHTTPCookieStore.cpp:115:74: warning: [WebKit] captured a raw pointer to a ref-countable type [-Wlambda-capture-ptr-to-refcntbl] cookieManager->deleteCookie(m_owningDataStore->sessionID(), cookie, [pool = WTFMove(pool), completionHandler = WTFMove(completionHandler)]() mutable { ^ This is warning about webPage object getting captured using a raw reference in the following function: void HTTPCookieStore::deleteCookie(const WebCore::Cookie& cookie, CompletionHandler<void()>&& completionHandler) { auto* pool = m_owningDataStore->processPoolForCookieStorageOperations(); if (!pool) { if (m_owningDataStore->sessionID() == PAL::SessionID::defaultSessionID() && !cookie.session) deleteCookieFromDefaultUIProcessCookieStore(cookie); else m_owningDataStore->removePendingCookie(cookie); RunLoop::main().dispatch([completionHandler = WTFMove(completionHandler)] () mutable { completionHandler(); }); return; } auto* cookieManager = pool->supplement<WebKit::WebCookieManagerProxy>(); cookieManager->deleteCookie(m_owningDataStore->sessionID(), cookie, [pool = WTFMove(pool), completionHandler = WTFMove(completionHandler)]() mutable { completionHandler(); }); } Here, we should have stored WebProcessPool in Ref instead. - R. Niwa

9 17

Request for Position: MSE-in-Workers
by Matt Wolenetz 27 Jul '22

27 Jul '22

Hi webkit-dev, This is a request for Webkit's position on the MSE-in-Workers feature: Explainer: https://github.com/wicg/media-source/blob/mse-in-workers-using-handle/mse-i… Main MSE spec issue tracking this feature: https://github.com/w3c/media-source/issues/175 ChromeStatus entry: https://chromestatus.com/features/5177263249162240 Draft specification: https://github.com/w3c/media-source/pull/282 TAG review: https://github.com/w3ctag/design-reviews/issues/656 Demo page (Chromium has experimental implementation): https://wolenetz.github.io/mse-in-workers-demo/mse-in-workers-demo.html Please let us know if you have any feedback! Thanks, Matt

1 1

Request for Position: User Preference Media Features Client Hints Headers
by Thomas Steiner 29 Jun '22

29 Jun '22

Hi WebKittens, We are interested in hearing your thoughts on User Preference Media Features Client Hints Headers, a set of HTTP Client Hints headers around user preference media features as defined by Media Queries Level 5. Specification Title: User Preference Media Features Client Hints Headers Specification or proposal URL: https://tomayac.github.io/user-preference-media-features-headers/ Explainer URL: https://github.com/tomayac/user-preference-media-features-headers/blob/main… TAG review: https://github.com/w3ctag/design-reviews/issues/632 Mozilla Standards Position: https://github.com/mozilla/standards-positions/issues/526 Thank you very much for your time and consideration! Cheers, Tom

1 5

Request for feedback on CompressionStream and DecompressionStream
by Adam Rice 02 May '22

02 May '22

I am trying to gauge feedback on compression streams with a view to shipping them in Chromium. Very briefly, they are a way to do gzip and gunzip in the browser. Less briefly, the explainer https://github.com/WICG/compression/blob/master/explainer.md goes into some detail of the how and why. The specification https://wicg.github.io/compression/ gives verbose detail of how. You may also find the W3C TAG review https://github.com/w3ctag/design-reviews/issues/410 interesting. Any and all feedback is welcome! FAQ: Q. Will this support brotli, zstd, etc.? A. Not yet. It's on the radar. Thanks, Adam Rice, Chromium

3 3

Request for Position: Multi-Screen Window Placement API
by Ben Morss 12 Apr '22

12 Apr '22

Hi all, We'd love to get WebKit's position on the new Multi-Screen Window Placement API. The proposal is to make incremental improvements to existing screen information and window placement APIs, allowing web applications to be intentional about the experiences they offer to users of multi-screen devices. We'd like to make it simple to: - detect when screen properties changes - discover the presence of extended screen areas - request extended screen information - place a window on a specific screen - show a window fullscreen on a specific screen This is useful for applications like the following: - a slideshow app shows slides full-screen on a projector, while displaying speaker notes on a laptop - a creativity app pops up secondary windows, like palettes, on a separate screen - a videoconferencing app shows video on a large screen, and controls on a smaller screen - a medical app shows images on multiple displays - a financial app shows dashboards on multiple monitors Future work could involve support for foldable devices, showing a window that spans multiple screens, etc. For more details, please see these links: - Explainer: https://github.com/webscreens/window-placement/blob/main/EXPLAINER.md - Draft Community Group Report: https://webscreens.github.io/window-placement/ - Additional ideas: https://github.com/webscreens/window-placement/blob/main/additional_explora… Thanks for your feedback on this! All best, Ben

1 1

Request for position: Capture Handle Identity
by Elad Alon 14 Mar '22

14 Mar '22

This is a request for WebKit's position on introducing an API that allows captured Web-applications to declare their identity to a would-be capturing Web-application. Links: * Explainer: https://github.com/WICG/capture-handle/blob/main/README.md * Spec: https://wicg.github.io/capture-handle/ * ChromeStatus entry: https://chromestatus.com/feature/4854125411958784

2 2

Using C++20 in WebKit
by Alex Christensen 10 Mar '22

10 Mar '22

In April 2019 in https://bugs.webkit.org/show_bug.cgi?id=197131 <https://bugs.webkit.org/show_bug.cgi?id=197131> I increased WebKit’s minimum C++ language requirement from C++14 to C++17. In 2022 I’m planning to increase WebKit’s minimum C++ requirement from C++17 to C++20. Would April 2022 be a good time to do that?

4 14

Proposal to update WebKitGTK dependency policy
by Michael Catanzaro 08 Mar '22

08 Mar '22

Hi, I'd like to selfishly propose updating our dependencies policy https://trac.webkit.org/wiki/WebKitGTK/DependenciesPolicy in order to accommodate RHEL in addition to Debian and Ubuntu. My goal is to provide WebKitGTK updates for the first five years (the "full support" period) of a RHEL major release, not just three years. However, we have some magic I don't understand allows use of newer toolchains, including newer libstdc++. We can actually somehow link newer libstdc++ into the same process as system libstdc++, so we can exempt the entire build toolchain (including CMake) from this policy, so this won't have any effect on discussions like "when can we use C++20 features" or "what version of CMake can we depend on." The primary impact would be on dependencies like ICU, GStreamer, etc. ICU is particularly important because this library bumps its ABI with every major release, so updating ICU to newer major versions is not possible. This would lock us into supporting ICU 67 until spring 2027. If we decide to land https://bugs.webkit.org/show_bug.cgi?id=235367 -- which currently looks unlikely -- then it would additionally lock us into ICU 60 until spring 2024. I think five years' support would benefit Ubuntu as well -- this matches the primary support lifetime of an Ubuntu LTS -- except Ubuntu doesn't seem to have the capability to build with newer toolchains, which means that, in practice, they will stop updating WebKit whenever we require a newer build toolchain. And although I think it would be a good idea to support Ubuntu for longer, I'm not brave enough to propose that we freeze our build toolchain dependencies for five years. So I will not suggest extending the support period for Ubuntu. Specifically, I propose adding the following text to our policy: * "We support the latest minor release of each major version of RHEL until two years after the release of the next major version." (Note: we currently have a three-year time-based release cycle, so that's five years total. If that were to unexpectedly change in the future, then adjusting the text of the policy would be needed.) And: * "For RHEL, WebKit is not expected to remain buildable using the default system libstdc++. The requirement for WebKit to remain buildable may be satisfied using GCC and LLVM toolsets from Application Streams." Michael

2 7

WebKit Style: Whitespace for Objective-C protocols
by Myles Maxfield 03 Mar '22

03 Mar '22

Hello! I was working on a patch recently where I wanted to give an Objective-C variable a type of “id that conforms to protocol Foo.” Should this be spelled like this: id <Foo> myVariable Or like this: id<Foo> myVariable I don’t see anything about this in the WebKit style guide <https://webkit.org/code-style-guidelines/>. The former appears in WebCore, WebKit, and WebKitLegacy 1035 times, whereas the latter appears only 281 times. On the other hand, Apple’s documentation <https://developer.apple.com/documentation/metal/1433401-mtlcreatesystemdefa…> seems to prefer the latter. Another consideration - This question is only about Objective-C protocols, not Objective-C generics. We already have a longstanding practice of spelling generics without the space, e.g. NSArray<NSString *> *. So, we could either intentionally choose for protocols to look visually similar to generics, or alternatively we could intentionally choose for protocols to look visually distinct to generics. Consider seeing something like Foo<Bar <Baz> *> *. Would the inconsistent spacing annoying you? Or would the spacing help you understand that Baz is a protocol but Bar is supplied as a generic type? Maybe the * characters already make that clear? What should the style guide say here? Thanks, Myles

3 3

Removing Python 2 in WebKit
by Jonathan Bedard 28 Feb '22

28 Feb '22

Hey folks, The macOS 12.3 seed remove Python 2. Most of WebKit’s tools are already Python 3 compatible, and in the last few weeks, we changed a number of buildbot invocation and shebangs so that workflows critical to the project are Python 3 by default. There are still a number of scripts which are Python 2 by default, though. Bellow, all of those scripts are listed. If a script important to your workflow is listed, now is a good time to help the project out by converting that script to Python 3. Tools/wpe/jhbuildrc Tools/lldb/dump_class_layout_unittest.py Tools/lldb/lldb_webkit_unittest.py Tools/glib/common.py Tools/glib/api_test_runner.py Tools/glib/generate-inspector-gresource-manifest.py Tools/glib/svn-revision Tools/Scripts/check-for-duplicated-platform-test-results Tools/Scripts/compare-webkit-configurations Tools/Scripts/read-checksum-from-png Tools/Scripts/check-for-invalid-symbols-in-version-script Tools/Scripts/sync-feature-defines Tools/Scripts/dump-webkit-tests-run Tools/Scripts/run-web-platform-tests Tools/Scripts/check-inspector-strings Tools/Scripts/import-w3c-performance-wg-tests Tools/Scripts/export-w3c-performance-wg-tests Tools/Scripts/malloc-tree Tools/Scripts/update-test-expectations-from-bugzilla Tools/Scripts/parse-gc-phase-timings Tools/Scripts/rebase-patch-after-webkit-move Tools/Scripts/merge-result-jsons Tools/Scripts/make-dist Tools/Scripts/webkit-filter-log Tools/Scripts/update-webkit-flatpak Tools/Scripts/export-w3c-test-changes Tools/Scripts/import-w3c-tests Tools/Scripts/compare-results Tools/Scripts/webkit-flatpak Tools/Scripts/check-for-global-bss-symbols-in-webkitgtk-libs Tools/Scripts/browserperfdash-benchmark Tools/Scripts/mark-jsc-stress-test Tools/Scripts/generate-win32-export-forwards Tools/Scripts/check-for-platform-layering-violations Tools/Scripts/find-duplicate-files Tools/Scripts/import-webdriver-tests Tools/Scripts/sampstat Tools/jhbuild/jhbuildrc_common.py Tools/jhbuild/jhbuild-wrapper Tools/WebKitArchiveSupport/run-webkit-archive Tools/CISupport/ews-app/manage.py Tools/gtk/jhbuildrc Tools/gtk/ycm_extra_conf.py Tools/CygwinDownloader/cygwin-downloader.py Thanks for your help, Jonathan Bedard

1 0