77144 2012-01-26 14:51:20 -0800 run-perf-test should include relative file paths in test names 2012-01-26 19:50:01 -0800 1 1 1 Unclassified WebKit Tools / Tests 528+ (Nightly build) Unspecified Unspecified RESOLVED FIXED P2 Normal --- 77037 1 rniwa rniwa abarth dpranke eric morrita ojan tony webkit.review.bot oldest_to_newest 543508 0 rniwa 2012-01-26 14:51:20 -0800 Right now, run-perf-test only uses the filename as the test name but this won't scale when we add more perf tests. 543566 1 124192 rniwa 2012-01-26 15:29:18 -0800 Created attachment 124192 fixes the bug 543584 2 124192 abarth 2012-01-26 15:39:39 -0800 Comment on attachment 124192 fixes the bug ok. How MacOS 9 of you! 543594 3 124192 dpranke 2012-01-26 15:48:20 -0800 Comment on attachment 124192 fixes the bug Do we have to follow the chromium perf bot convention? It would seem better to follow the webkit conventions for test names and use '/'. 543607 4 rniwa 2012-01-26 15:55:18 -0800 (In reply to comment #3) > (From update of attachment 124192 [details]) > Do we have to follow the chromium perf bot convention? It would seem better to follow the webkit conventions for test names and use '/'. We could but we don't have to match layout tests' convention either though. Is there any reason using '/' as the delimiter is better? 543616 5 dpranke 2012-01-26 16:06:58 -0800 (In reply to comment #4) > (In reply to comment #3) > > (From update of attachment 124192 [details] [details]) > > Do we have to follow the chromium perf bot convention? It would seem better to follow the webkit conventions for test names and use '/'. > > We could but we don't have to match layout tests' convention either though. Is there any reason using '/' as the delimiter is better? It seems to me that having a single convention used by the tools in a single code base would be better. That way I don't have to remember that it's run-webkit-tests fast/html but run-perf-tests fast:html . I type this now realizing that what is in the JSON file is not necessarily what's on the command line, but it seems to me that being uniform is good, and if test names get displayed on a web page I would want them to display as "/" to match the other tools. If the chromium perf test store "foo:bar" but display "foo/bar", then ":" is fine by me. 543619 6 dpranke 2012-01-26 16:07:52 -0800 (In reply to comment #5) > If the chromium perf test store "foo:bar" but display "foo/bar", then ":" is fine by me. Well, "fine" might be a bit strong; I'd probably still grumble a little over it being different :). 543639 7 rniwa 2012-01-26 16:28:53 -0800 I guess being able to copy & paste test name is a strong argument. Will change it to /. 543756 8 rniwa 2012-01-26 19:50:01 -0800 http://trac.webkit.org/changeset/106062 124192 2012-01-26 15:29:18 -0800 2012-01-26 15:48:20 -0800 fixes the bug fix77144 text/plain 3504 rniwa SW5kZXg6IFRvb2xzL1NjcmlwdHMvd2Via2l0cHkvcGVyZm9ybWFuY2VfdGVzdHMvcGVyZnRlc3Rz cnVubmVyX3VuaXR0ZXN0LnB5Cj09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT0KLS0tIFRvb2xzL1NjcmlwdHMvd2Via2l0cHkv cGVyZm9ybWFuY2VfdGVzdHMvcGVyZnRlc3RzcnVubmVyX3VuaXR0ZXN0LnB5CShyZXZpc2lvbiAx MDYwMTApCisrKyBUb29scy9TY3JpcHRzL3dlYmtpdHB5L3BlcmZvcm1hbmNlX3Rlc3RzL3BlcmZ0 ZXN0c3J1bm5lcl91bml0dGVzdC5weQkod29ya2luZyBjb3B5KQpAQCAtMTg5LDcgKzE4OSw3IEBA CiAKICAgICAgICAgc2VsZi5hc3NlcnRFcXVhbChqc29uLmxvYWRzKHJ1bm5lci5faG9zdC5maWxl c3lzdGVtLmZpbGVzWycvbW9jay1jaGVja291dC9vdXRwdXQuanNvbiddKSwgewogICAgICAgICAg ICAgInRpbWVzdGFtcCI6IDEyMzQ1Njc4OSwgInJlc3VsdHMiOgotICAgICAgICAgICAgeyJldmVu dC10YXJnZXQtd3JhcHBlciI6IHsibWF4IjogMTUxMCwgImF2ZyI6IDE0ODkuMDUsICJtZWRpYW4i OiAxNDg3LCAibWluIjogMTQ3MSwgInN0ZGV2IjogMTQuNDZ9LAorICAgICAgICAgICAgeyJCaW5k aW5nczpldmVudC10YXJnZXQtd3JhcHBlciI6IHsibWF4IjogMTUxMCwgImF2ZyI6IDE0ODkuMDUs ICJtZWRpYW4iOiAxNDg3LCAibWluIjogMTQ3MSwgInN0ZGV2IjogMTQuNDZ9LAogICAgICAgICAg ICAgImdyb3VwX25hbWU6dGVzdF9uYW1lIjogNDJ9LAogICAgICAgICAgICAgInJldmlzaW9uIjog MTIzNH0pCiAKQEAgLTIwOSw3ICsyMDksNyBAQAogCiAgICAgICAgIHNlbGYuYXNzZXJ0RXF1YWwo anNvbi5sb2FkcyhydW5uZXIuX2hvc3QuZmlsZXN5c3RlbS5maWxlc1snL21vY2stY2hlY2tvdXQv b3V0cHV0Lmpzb24nXSksIHsKICAgICAgICAgICAgICJ0aW1lc3RhbXAiOiAxMjM0NTY3ODksICJy ZXN1bHRzIjoKLSAgICAgICAgICAgIHsiZXZlbnQtdGFyZ2V0LXdyYXBwZXIiOiB7Im1heCI6IDE1 MTAsICJhdmciOiAxNDg5LjA1LCAibWVkaWFuIjogMTQ4NywgIm1pbiI6IDE0NzEsICJzdGRldiI6 IDE0LjQ2fSwKKyAgICAgICAgICAgIHsiQmluZGluZ3M6ZXZlbnQtdGFyZ2V0LXdyYXBwZXIiOiB7 Im1heCI6IDE1MTAsICJhdmciOiAxNDg5LjA1LCAibWVkaWFuIjogMTQ4NywgIm1pbiI6IDE0NzEs ICJzdGRldiI6IDE0LjQ2fSwKICAgICAgICAgICAgICJncm91cF9uYW1lOnRlc3RfbmFtZSI6IDQy fSwKICAgICAgICAgICAgICJyZXZpc2lvbiI6IDEyMzQsCiAgICAgICAgICAgICAia2V5IjogInZh bHVlIn0pCkluZGV4OiBUb29scy9TY3JpcHRzL3dlYmtpdHB5L3BlcmZvcm1hbmNlX3Rlc3RzL3Bl cmZ0ZXN0c3J1bm5lci5weQo9PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09Ci0tLSBUb29scy9TY3JpcHRzL3dlYmtpdHB5L3Bl cmZvcm1hbmNlX3Rlc3RzL3BlcmZ0ZXN0c3J1bm5lci5weQkocmV2aXNpb24gMTA2MDEwKQorKysg VG9vbHMvU2NyaXB0cy93ZWJraXRweS9wZXJmb3JtYW5jZV90ZXN0cy9wZXJmdGVzdHNydW5uZXIu cHkJKHdvcmtpbmcgY29weSkKQEAgLTI0OSw3ICsyNDksNyBAQAogICAgIF9saW5lc190b19pZ25v cmVfaW5fcGFyc2VyX3Jlc3VsdCA9IFsKICAgICAgICAgcmUuY29tcGlsZShyJ15SdW5uaW5nIFxk KyB0aW1lcyQnKSwKICAgICAgICAgcmUuY29tcGlsZShyJ15JZ25vcmluZyB3YXJtLXVwICcpLAot ICAgICAgICByZS5jb21waWxlKHInXlxkKyQnKSwKKyAgICAgICAgcmUuY29tcGlsZShyJ15cZCso LlxkKyk/JCcpLAogICAgIF0KIAogICAgIGRlZiBfc2hvdWxkX2lnbm9yZV9saW5lX2luX3BhcnNl cl90ZXN0X3Jlc3VsdChzZWxmLCBsaW5lKToKQEAgLTI4MSw3ICsyODEsNyBAQAogCiAgICAgICAg IGlmIHRlc3RfZmFpbGVkIG9yIHNldChrZXlzKSAhPSBzZXQocmVzdWx0cy5rZXlzKCkpOgogICAg ICAgICAgICAgcmV0dXJuIFRydWUKLSAgICAgICAgc2VsZi5fcmVzdWx0c1t0ZXN0X25hbWVdID0g cmVzdWx0cworICAgICAgICBzZWxmLl9yZXN1bHRzW2ZpbGVzeXN0ZW0uam9pbihjYXRlZ29yeSwg dGVzdF9uYW1lKS5yZXBsYWNlKCcvJywgJzonKS5yZXBsYWNlKCdcXCcsICc6JyldID0gcmVzdWx0 cwogICAgICAgICBzZWxmLl9idWlsZGJvdF9vdXRwdXQud3JpdGUoJ1JFU1VMVCAlczogJXM9ICVz IG1zXG4nICUgKGNhdGVnb3J5LCB0ZXN0X25hbWUsIHJlc3VsdHNbJ2F2ZyddKSkKICAgICAgICAg c2VsZi5fYnVpbGRib3Rfb3V0cHV0LndyaXRlKCcsICcuam9pbihbJyVzPSAlcyBtcycgJSAoa2V5 LCByZXN1bHRzW2tleV0pIGZvciBrZXkgaW4ga2V5c1sxOl1dKSArICdcbicpCiAgICAgICAgIHJl dHVybiBGYWxzZQpJbmRleDogVG9vbHMvQ2hhbmdlTG9nCj09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT0KLS0tIFRvb2xzL0No YW5nZUxvZwkocmV2aXNpb24gMTA2MDUzKQorKysgVG9vbHMvQ2hhbmdlTG9nCSh3b3JraW5nIGNv cHkpCkBAIC0xLDMgKzEsMjIgQEAKKzIwMTItMDEtMjYgIFJ5b3N1a2UgTml3YSAgPHJuaXdhQHdl YmtpdC5vcmc+CisKKyAgICAgICAgcnVuLXBlcmYtdGVzdCBzaG91bGQgaW5jbHVkZSByZWxhdGl2 ZSBmaWxlIHBhdGhzIGluIHRlc3QgbmFtZXMKKyAgICAgICAgaHR0cHM6Ly9idWdzLndlYmtpdC5v cmcvc2hvd19idWcuY2dpP2lkPTc3MTQ0CisKKyAgICAgICAgUmV2aWV3ZWQgYnkgTk9CT0RZIChP T1BTISkuCisKKyAgICAgICAgSW5jbHVkZSBhbGwgZGlyZWN0b3J5IG5hbWVzIGluIHRoZSB0ZXN0 IG5hbWUgc2VwYXJhdGVkIGJ5ICI6Ii4gV2UgdXNlICI6IiBpbnN0ZWFkIG9mIC8gb3IgXFwgYXMg dGhlIGRlbGltaXRlcgorICAgICAgICBzaW5jZSB0aGF0J3MgdGhlIGNvbnZlbnRpb24gdXNlZCBi eSBDaHJvbWl1bSBwZXJmIGJvdHMuCisKKyAgICAgICAgZS5nLiBCaW5kaW5ncy9ldmVudC10YXJn ZXQtd3JhcHBlci5odG1sIHRyYW5zbGF0ZXMgdG8gQmluZGluZ3M6ZXZlbnQtdGFyZ2V0LXdyYXBw ZXIKKworICAgICAgICAqIFNjcmlwdHMvd2Via2l0cHkvcGVyZm9ybWFuY2VfdGVzdHMvcGVyZnRl c3RzcnVubmVyLnB5OgorICAgICAgICAoUGVyZlRlc3RzUnVubmVyKToKKyAgICAgICAgKFBlcmZU ZXN0c1J1bm5lci5fcHJvY2Vzc19wYXJzZXJfdGVzdF9yZXN1bHQpOgorICAgICAgICAqIFNjcmlw dHMvd2Via2l0cHkvcGVyZm9ybWFuY2VfdGVzdHMvcGVyZnRlc3RzcnVubmVyX3VuaXR0ZXN0LnB5 OgorICAgICAgICAodGVzdF9ydW5fdGVzdF9zZXRfd2l0aF9qc29uX291dHB1dCk6CisgICAgICAg ICh0ZXN0X3J1bl90ZXN0X3NldF93aXRoX2pzb25fc291cmNlKToKKwogMjAxMi0wMS0yNiAgU2hl cmlmZiBCb3QgIDx3ZWJraXQucmV2aWV3LmJvdEBnbWFpbC5jb20+CiAKICAgICAgICAgVW5yZXZp ZXdlZCwgcm9sbGluZyBvdXQgcjEwNjAzOC4K