August 2018 - webkit-dev - lists.webkit.org

Introducing the GC Heap Inspector
by Simon Fraser 24 Aug '18

24 Aug '18

Yesterday I landed support for a way to dump and inspect a GC heap snapshot. This work was motivated by the need to have easier ways to debug Document leaks, especially those involving GC. The basic idea is that you do some steps that trigger a document leak, then get a heap dump. Typically this involves loading the test page, then navigating away, clearing caches, triggering a GC and capturing a snapshot. The snapshot is dumped as JSON file in /tmp; you can then drag this snapshot into the view which lives in Tools/GCHeapInspector, and see why objects are referencing the leaked document. This is documented on https://trac.webkit.org/wiki/Inspecting%20the%20GC%20heap <https://trac.webkit.org/wiki/Inspecting%20the%20GC%20heap> Let me know if you have questions. Improvements to the inspector page are welcomed. Simon

1 0

Bugzilla component for CMake
by Michael Catanzaro 24 Aug '18

24 Aug '18

Hi, I think it would be useful to have a Bugzilla component for CMake build system changes. We have a lot of such bugs, and normally use the Tools/Tests component for them currently, which is not a great fit (and causes Lucas to be CCed unnecessarily on a bunch of bugs ;) Michael

2 2

Re: [webkit-dev] CSS Logical Properties and Values
by Oriol Brufau 22 Aug '18

22 Aug '18

In Blink, properties are defined in CSSProperties.json5, there you can specify a "runtime_flag" property whose value is the name of a runtime flag. This overrides the IsEnabled method with a check for the runtime flag: https://cs.chromium.org/chromium/src/third_party/blink/renderer/build/scrip… And then this method is called in various places like CSSStyleDeclaration::NamedPropertyEnumerator https://cs.chromium.org/chromium/src/third_party/blink/renderer/core/css/cs… Runtime flags can be defined like https://cs.chromium.org/chromium/src/third_party/blink/renderer/platform/ru… I can run chrome or content_shell with --enable-experimental-web-platform-features in order to enable experimental flags. Tests are run this way. El 22/8/18 a les 03:26, Simon Fraser ha escrit: > Right, sorry. I don't think we have a way to do that. > > How does Blink solve this problem? > > Simon > >

1 0

Having trouble using webkitpy-based scripts?
by dean_johnson＠apple.com 17 Aug '18

17 Aug '18

Hi all, Some of the webkitpy dependencies hosted on files.pythonhosted.org <http://files.pythonhosted.org/> are taking ridiculous amounts of time to be fetched/downloaded, which is resulting in long stalls after using WebKit tools that invoke the `autoinstalled` codepath if you do not have webkitpy dependencies already downloaded/installed. Example: ➜ time pip install -i https://pypi.python.org/simple requests Looking in indexes: https://pypi.python.org/simple Collecting requests Downloading https://files.pythonhosted.org/packages/65/47/7e02164a2a3db50ed6d8a6ab1d6d6… (91kB) 100% |████████████████████████████████| 92kB 3.8MB/s Collecting idna<2.8,>=2.5 (from requests) Downloading https://files.pythonhosted.org/packages/4b/2a/0276479a4b3caeb8a8c1af2f8e435… (58kB) 100% |████████████████████████████████| 61kB 2.7MB/s Collecting certifi>=2017.4.17 (from requests) Downloading https://files.pythonhosted.org/packages/16/1f/50d729c104b21c1042aa51560da61… (146kB) 100% |████████████████████████████████| 153kB 784kB/s Collecting chardet<3.1.0,>=3.0.2 (from requests) Downloading https://files.pythonhosted.org/packages/bc/a9/01ffebfb562e4274b6487b4bb1dde… (133kB) 100% |████████████████████████████████| 143kB 4.1MB/s Collecting urllib3<1.24,>=1.21.1 (from requests) Downloading https://files.pythonhosted.org/packages/bd/c9/6fdd990019071a4a32a5e7cb78a1d… (133kB) 100% |████████████████████████████████| 143kB 3.9MB/s Installing collected packages: idna, certifi, chardet, urllib3, requests Successfully installed certifi-2018.8.13 chardet-3.0.4 idna-2.7 requests-2.19.1 urllib3-1.23 sudo pip install -i https://pypi.python.org/simple requests 1.17s user 0.44s system 1% cpu 2:36.42 total It’s my understanding that letting the script hang for multiple minutes may allow you to eventually get the content requested. Cheers, Dean

1 0

2018 WebKit Contributors Meeting - Registration is Open!
by Jonathan Davis 16 Aug '18

16 Aug '18

Hello WebKit Contributors, You are invited to attend the annual WebKit Contributors Meeting to be held at the Hyatt House San Jose/Silicon Valley hotel on Friday, October 12th, 2018 from 8 AM to 6 PM Pacific. Breakfast and sign-in will begin at 8 AM. Presentations will run between 9 AM to 6 PM. Lunch will be provided with all day coffee, tea and a mid-afternoon snack break. A reception is planned at the venue after the meeting with drinks and snacks from 6 PM to 9 PM. To attend you must be an active WebKit contributor. The meeting will be free of charge, and registration is now open. Sign up at https://webkit.org/meeting/ to reserve your spot. The registration deadline is Friday, September 14th, 2018. If you are traveling internationally and require a formal invitation letter, please email me. The event format will follow last year’s mix of proposed talks of around 40 minutes long and spontaneous lightning talks about 5-10 minutes long. CALL FOR SPEAKERS This year we have 6 openings for 30-40 minute talks with 10 minutes of Q&A. If you’d like to present at the meeting, please email me directly. Have something to share, but it won’t take that long? Consider emailing me about reserving a spot for a 5-10 minute Lightning Talk. Email me soon to reserve your spot before they fill up. We look forward to seeing you there! Thank you, Jon Davis

1 0

Megan Gardner is a reviewer!
by Beth Dakin 15 Aug '18

15 Aug '18

Hi all, I’m pleased to announce that Megan Gardner is now a reviewer! Megan specializes in iOS selection, editing, and gestures. Please send her patches and congratulate her! 🎉 -Beth

1 0

Freenode spam counter-measure
by Philippe Normand 06 Aug '18

06 Aug '18

Hi, Can one of the #webkit admin please set the +r mode on? This would help reducing spam. Only messages from registered nicks would come through. Philippe

6 9

Proposal: Not supporting x86 w/o SSE2
by Yusuke SUZUKI 04 Aug '18

04 Aug '18

Hello WebKittens, To mitigate recent side-channel attacks, we added `lfence` in x86 WebKit. But this is supported in x86 SSE2. So the ToT WebKit is not usable with very old x86 CPUs which do not have SSE2[1]. According to [2], Mozilla Firefox 49 no longer supports x86 w/o SSE2. Since Firefox 45 ESR is already End of Life, all the supported Firefox require SSE2 on x86. Chromium also does not support x86 if it does not have SSE2 IIRC[3]. So, no major browsers support x86 w/o SSE2 right now. I think our stakeholders do not maintain non-SSE2 x86. Even if it works, it is potentially vulnerable, and I think it is not acceptable for the browser engine. My proposal here is clarifying that WebKit does not support x86 not having SSE2. [1]: https://bugs.webkit.org/show_bug.cgi?id=188145 [2]: https://support.mozilla.org/en-US/kb/your-hardware-no-longer-supported [3]: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=763290

3 3