WebKit Bugzilla
New
Browse
Search+
Log In
×
Sign in with GitHub
or
Remember my login
Create Account
·
Forgot Password
Forgotten password account recovery
RESOLVED FIXED
175216
Separate jsc stress test script writer from run-jsc-stress-tests
https://bugs.webkit.org/show_bug.cgi?id=175216
Summary
Separate jsc stress test script writer from run-jsc-stress-tests
Stephan Szabo
Reported
2017-08-04 15:28:39 PDT
For running JSC tests on windows without a unix shell, in
https://lists.webkit.org/pipermail/webkit-dev/2017-August/029370.html
we proposed a multiple step process for that with the first being splitting off the test script writing and related code from the other code in run-jsc-stress-tests to eventually allow an alternate implementation that doesn't require shell.
Attachments
Separate out script building/running from other parts of run-jsc-stress-tests
(31.66 KB, patch)
2017-08-07 16:46 PDT
,
Stephan Szabo
mark.lam
: review-
Details
Formatted Diff
Diff
Separate out script building/running from other parts of run-jsc-stress-tests
(30.57 KB, patch)
2017-08-08 15:21 PDT
,
Stephan Szabo
no flags
Details
Formatted Diff
Diff
Show Obsolete
(1)
View All
Add attachment
proposed patch, testcase, etc.
Stephan Szabo
Comment 1
2017-08-07 16:46:00 PDT
Created
attachment 317504
[details]
Separate out script building/running from other parts of run-jsc-stress-tests
Mark Lam
Comment 2
2017-08-08 14:50:09 PDT
Comment on
attachment 317504
[details]
Separate out script building/running from other parts of run-jsc-stress-tests View in context:
https://bugs.webkit.org/attachment.cgi?id=317504&action=review
r- because parseRunCommands is duplicated.
> Tools/Scripts/webkitruby/jsc-stress-test-writer-default.rb:217 > +#$runCommandOptions = {}
Why have this here?n This looks like a copy-paste that hasn't been cleaned up.
> Tools/Scripts/webkitruby/jsc-stress-test-writer-default.rb:320 > +# Returns true if there were run commands found in the file ($benchmarkDirectory + > +# $benchmark), in which case those run commands have already been executed. Otherwise > +# returns false, in which case you're supposed to add your own run commands. > +def parseRunCommands > + oldDidAddRunCommand = $didAddRunCommand > + $didAddRunCommand = false > + > + Dir.chdir($outputDir) { > + File.open($benchmarkDirectory + $benchmark) { > + | inp | > + inp.each_line { > + | line | > + begin > + doesMatch = line =~ /^\/\/@/ > + rescue Exception => e > + # Apparently this happens in the case of some UTF8 stuff in some files, where > + # Ruby tries to be strict and throw exceptions. > + next > + end > + next unless doesMatch > + eval $~.post_match > + } > + } > + } > + > + result = $didAddRunCommand > + $didAddRunCommand = result or oldDidAddRunCommand > + result > +end
There's still a copy of this function in run-jsc-stress-tests. Is there a reason for this?
Stephan Szabo
Comment 3
2017-08-08 15:21:33 PDT
Created
attachment 317628
[details]
Separate out script building/running from other parts of run-jsc-stress-tests Thanks, for parseRunCommands I hadn't intended to move it and missed clearing it out of the writer file with it back in run-jsc-stress-tests. The other line was inside a block of things that was moving so I went back and forth but thought it made more sense in the main script. I was swapping which was commented out and forgot to remove the commented out one at the end.
Mark Lam
Comment 4
2017-08-08 15:46:21 PDT
Comment on
attachment 317628
[details]
Separate out script building/running from other parts of run-jsc-stress-tests r=me if EWS is green.
WebKit Commit Bot
Comment 5
2017-08-08 17:18:11 PDT
Comment on
attachment 317628
[details]
Separate out script building/running from other parts of run-jsc-stress-tests Clearing flags on attachment: 317628 Committed
r220431
: <
http://trac.webkit.org/changeset/220431
>
WebKit Commit Bot
Comment 6
2017-08-08 17:18:13 PDT
All reviewed patches have been landed. Closing bug.
Radar WebKit Bug Importer
Comment 7
2017-08-08 17:18:55 PDT
<
rdar://problem/33788991
>
Note
You need to
log in
before you can comment on or make changes to this bug.
Top of Page
Format For Printing
XML
Clone This Bug