Audit usage of subprocess in run-webkit-tests to discover any issues with threads |
||||||||||
Issue descriptionIn the bug at https://bugs.chromium.org/p/chromium/issues/detail?id=522396 we discovered that the following code will lock up on Windows; ---------------------------- p = subprocess.Popen(...., stdout=subprocess.PIPE) def reader(pipe): while True: data = pipe.read(x) if not data: break t1 = threading.Thread(run=reader, args=(p.stdout)) t1.join() ---------------------------- We should check that nothing in run-webkit-tests uses this model.
,
May 10 2016
I'm starting with the ones under "third_party/WebKit/Tools/Scripts/webkitpy/common/"
,
May 10 2016
,
May 10 2016
,
May 18 2016
,
May 18 2016
Deprecating component:Blink>LayoutTests, to use label Test=Layout instead. Merging these to component:Blink for the Blink rotation to pick up and re-triage as appropriate.
,
May 23 2016
,
Dec 2 2016
Note: In bug 637578 , kbr noted that subprocess42 (https://cs.chromium.org/chromium/infra/luci/client/utils/subprocess42.py) may be a better replacement for subprocess in some places. Also, a note about where to investigate first: Most sub-processes created by run-webkit-tests are created via Executive.run_command (https://cs.chromium.org/chromium/src/third_party/WebKit/Tools/Scripts/webkitpy/common/system/executive.py?l=348)
,
Feb 20 2017
,
Aug 9 2017
,
Oct 19
|
||||||||||
►
Sign in to add a comment |
||||||||||
Comment 1 by tansell@chromium.org
, May 10 2016