April 2018 - webkit-dev - lists.webkit.org

PSA: Bit fields won't be packed on Windows if you mix types
by Ryosuke Niwa 04 Sep '20

04 Sep '20

Hi, Unlike gcc and clang, MSVC pads each consecutive member variables of the same type in bitfields. e.g. if you have: struct AB { unsigned m_1 : 31; bool m_2 : 1; } then *MSVC pads m_1 and allocates sizeof(unsigned) * 2 for AB* whereas gcc and clang only allocate sizeof(unsigned) * 1 for AB. This is *not a compiler bug*. It's a spec. compliant behavior, and may in fact have a better run-time performance in some situations. However, for core objects like RenderObject and InlineBox, allocating extra 4-8 bytes per each object is prohibitory expensive. In such cases, please *use the same POD type for all bitfield member variables*. (Storage class classifiers and variable qualifiers seem to have no effect on how variables are packed; e.g. mutable, const, etc...). For example, MSVC will allocate sizeof(unsigned) * 1 for AB if we rewrite the above code as: struct AB { unsigned m_1 : 31; unsigned m_2 : 1; } When you're making this change, *be sure to audit all code that assigns a non-boolean value to m_2* because implicit type coercion into boolean is no longer in effect. For example, AB ab; ab.m_2 = 2; puts(ab.m_2 ? "true" : "false"); will print "true" before the change and will print "false" after the change. An easy way to ensure you've audited all such code is to add getters and setters for all bitfield member variables or wrap them in a special structure or class as done in http://trac.webkit.org/changeset/103353 and http://trac.webkit.org/changeset/104254. Best regards, Ryosuke Niwa Software Engineer Google Inc.

4 8

Support "registerProtocolHandler" in WebKit2
by Gyuyoung Kim 20 Apr '20

20 Apr '20

Hello, I would like to listen what do you think to support 'registerProtocolHandler' in WebKit2. This feature is to execute web content through registered custom protocol. - For example, web content can register "mailto" custom protocol using this feature, <script> navigator.registerProtocolHandler("mailto", "https://mail.naver.com/", "Web Mail"); </script> <html> <head> <title>Web Protocol Handler Sample - Test</title> </head> <body> <p>Send an email : <a href="mailto://">this !</a></p> </body> </html> Besides this feature has been supported by Mozilla and Chromium (From Mozilla 3.0, Chromium 13). - https://developer.mozilla.org/en-US/docs/Web/API/Navigator/registerProtocol… - https://developer.mozilla.org/en/docs/Web-based_protocol_handlers The feature is included in the W3C recommendation 28 released on Oct. 2014. - W3C spec : http://www.w3.org/TR/html5/webappapis.html#custom-handlers IIRC, some WebKit1 ports supported this feature though, those ports were deprecated. Now WebKit port supports this feature no more. There is a very old bug to support this feature though, it wasn't maintained so far. Recently I updated it based on latest WebKit. - https://bugs.webkit.org/show_bug.cgi?id=92749 Feel free to give me any feedback about this feature. Gyuyoung.

10 26

WebKit project in Coverity
by Carlos Alberto Lopez Perez 09 Jan '20

09 Jan '20

Hi, Coverity is an static analysis tool that allows to find bugs and vulnerabilities on the source code via static analysis. For open source projects, they offer free usage of their platform. The WebKit project is already registered there since a while. [1] To read the reports in detail or run new scans you have to be member of the WebKit project in Coverity. I happen to be one of the admins there, and I will happily grant you access to this platform if you are a WebKit committer (listed in contributors.json). So if you are interested in this, just send me an email requesting access. Regards ------- [1] https://scan.coverity.com/projects/webkit

2 1

C++17 is here. Should we use it?
by JF Bastien 26 Jun '19

26 Jun '19

Hello WebKilters, Our Chrome-y friends are considering the use of C++14 <https://groups.google.com/a/chromium.org/forum/?utm_medium=email&utm_source…>. I have to say that C++14 in WebKit has been quite amazing, and we should consider using C++17: it has many wonderful new things <https://github.com/tvaneerd/cpp17_in_TTs/blob/master/ALL_IN_ONE.md>, some of which we already use through WTF’s re-implementation of library features. By now (table as witness <http://en.cppreference.com/w/cpp/compiler_support>) most C++17 languages features are in clang and GCC, and MSVC isn’t doing too bad either. Language things can just come through WTF if we really want them. So how about it? JF

8 23

For your consideration: Naming scheme for fooIfExists/ensureFoo
by Darin Adler 22 Feb '19

22 Feb '19

Lets bike shed! For some time, functions with names like fooIfExists and ensureFoo have been bothering me. I find both names kind of opaque and unpleasant. Here’s an example: StyleResolver* styleResolverIfExists(); StyleResolver* ensureStyleResolver() What do you think of these names instead? StyleResolver* optionalStyleResolver(); StyleResolver& requiredStyleResolver(); I like them better. Note also that I think the requiredStyleResolver function should return a reference so nobody is tempted to do null checks. It seems like the C++ community likes the name optional for this concept; isn’t there some kind of std::optional template? -- Darin

15 25

CSSOM View behavior for scrollLeft, scrollTop, scrollWidth etc
by Frédéric Wang 24 Sep '18

24 Sep '18

Hi everybody, I'd like to announce that I started to implement the behavior for scrollLeft, scrollTop, scrollWidth, scrollHeight, scrollTo, scrollBy and scrollingElement [1] as specified by the CSSOM View specification [2]. To summarize, the main difference is which element to use in order to set or get the scroll properties of the viewport in standard (non-Quirks) mode: * The CSSOM View spec says it should be document.documentElement and most browsers implement it that way. * Chromium used to use document.body but this has recently been changed to follow the CSSOM View spec [3]. * WebKit still uses document.body and I'm proposing to change this behavior. Note that this is likely to break existing content. The change was attempted in WebKit in the past but broke some non-regression tests [4]. Chromium people also had some issues when trying to ship the change but things went well at the end [5]. For now I'm developing it under a CSSOMViewScrollingAPI developer flag. I was able to run the tests with that flag enabled, modulo a few test adjustments. [1] https://bugs.webkit.org/show_bug.cgi?id=5991 [2] https://drafts.csswg.org/cssom-view [3] https://groups.google.com/a/chromium.org/forum/#!topic/blink-dev/X64Sg16RhT4 [4] https://bugs.webkit.org/show_bug.cgi?id=106133, https://bugs.webkit.org/show_bug.cgi?id=121876 [5] https://groups.google.com/a/chromium.org/forum/#!msg/blink-dev/2iOz5-fgD8Y/… ; https://bugs.webkit.org/show_bug.cgi?id=5991#c20 -- Frédéric Wang - frederic-wang.fr

1 3

bmalloc design question about relation with std malloc
by Yusuke SUZUKI 03 May '18

03 May '18

Hi, WebKittens, IIRC, bmalloc uses mmap based page allocator for internal memory use. For example, bmalloc::Vector uses it instead of calling malloc. But recent changes start using std::vector, which means it uses std malloc under the hood. So my question is, if we want some internal memory allocation in bmalloc, shoud we use std::malloc? Or should we use mmap based allocator? Best regards, Yusuke Suzuki -- Regards, Yusuke Suzuki

3 3

cant get libvpx repo here?
by tugouxp 30 Apr '18

30 Apr '18

HI folks: i am compiling the latest webkit on ubuntu18.04. when i follow the manul to do the steps: ./Tools/Scripts/update-webkitgtk-libs i meet the problem of cant get the libvpx package. such as follows. this problem cant be fixed because of gov policy, you know. so is there any other way to dealt with this problem? thanks for your kindly reply. git clone https://chromium.googlesource.com/webm/libvpx libvpx Cloning into 'libvpx'... fatal: unable to access 'https://chromium.googlesource.com/webm/libvpx/': Failed to connect to chromium.googlesource.com port 443: Connection timed out

2 1

Hosting precompiled `jsc` binaries for Linux
by Mathias Bynens 29 Apr '18

29 Apr '18

Dear WebKit friends, Please consider hosting precompiled `jsc` binaries for Linux somewhere official. Perhaps https://webkit.org/build-archives/? The build bots already generate those binaries anyway — all that’s missing is an additional build step that archives and uploads them. Doing so would make it easier for developers to test JSC directly, or to include JSC when running benchmarks such as the Web Tooling Benchmark (ref. https://github.com/v8/web-tooling-benchmark/pull/29#discussion_r155587697) I recently built a tool that automates the process of installing precompiled binaries for various JS engines: https://github.com/GoogleChromeLabs/jsvu It’d be great to be able to support JavaScriptCore on linux32 and linux64 as well! Prior discussion: https://bugs.webkit.org/show_bug.cgi?id=179945#c11 Regards, Mathias

9 24

Re: [webkit-dev] New iOS versions sending bogus User-Agent build data
by Michael Catanzaro 27 Apr '18

27 Apr '18

On Thu, Apr 26, 2018 at 7:15 AM, Ricky Young <ricky(a)gmx.com> wrote: > I guess that soon it won't be possible to determine iOS version > distribution using User-Agent sniffing. Right? Is there an > alternative for the browser to tell the iOS version?  Hi, The very point of no longer sending the verison in the user agent is to make it difficult for websites to determine the operating system version. We already have to deal with enough serious web compatibility problems caused by the rest of the information in the user agent. I'm pleased that the version number is now stable -- that's a huge improvement -- but I won't be happy until Safari and other major browsers start sending fully-randomized UAs, or else a stub or blank UA ("Mozilla/5.0"). Michael

7 12