136587 2014-09-05 15:46:10 -0700 Make Dashboard metrics time calculation work with out of order builds 2014-09-07 23:18:16 -0700 1 1 1 Unclassified WebKit Tools / Tests 528+ (Nightly build) Unspecified Unspecified RESOLVED WONTFIX P2 Normal --- 1 ap ap thorton oldest_to_newest 1033570 0 ap 2014-09-05 15:46:10 -0700 We have multiple bots on some queues, so results can come out of order. 1033573 1 237716 ap 2014-09-05 15:51:05 -0700 Created attachment 237716 proposed patch 1033591 2 237716 ap 2014-09-05 16:41:39 -0700 Comment on attachment 237716 proposed patch I need to think more about this. Getting a result for a newer build doesn't mean that you have a result for an older build - e.g. the new build could fail due to a new mistaken commit, so you don't know if your commit was the one that introduced the problem. 1033779 3 ap 2014-09-07 23:18:16 -0700 Considering how we use the dashboard, I think that this patch was right. We do not wait for all results for the bots to tell us which commit caused the problem - we just see the red, and analyze the breakage ourselves. However, I'm going to include this change into a different patch. 237716 2014-09-05 15:51:05 -0700 2014-09-05 16:41:39 -0700 proposed patch OutOfOrder.txt text/plain 3195 ap SW5kZXg6IFRvb2xzL0NoYW5nZUxvZwo9PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09Ci0tLSBUb29scy9DaGFuZ2VMb2cJKHJl dmlzaW9uIDE3MzMzNykKKysrIFRvb2xzL0NoYW5nZUxvZwkod29ya2luZyBjb3B5KQpAQCAtMSwz ICsxLDEzIEBACisyMDE0LTA5LTA1ICBBbGV4ZXkgUHJvc2t1cnlha292ICA8YXBAYXBwbGUuY29t PgorCisgICAgICAgIE1ha2UgRGFzaGJvYXJkIG1ldHJpY3MgdGltZSBjYWxjdWxhdGlvbiB3b3Jr IHdpdGggb3V0IG9mIG9yZGVyIGJ1aWxkcworICAgICAgICBodHRwczovL2J1Z3Mud2Via2l0Lm9y Zy9zaG93X2J1Zy5jZ2k/aWQ9MTM2NTg3CisKKyAgICAgICAgUmV2aWV3ZWQgYnkgTk9CT0RZIChP T1BTISkuCisKKyAgICAgICAgKiBCdWlsZFNsYXZlU3VwcG9ydC9idWlsZC53ZWJraXQub3JnLWNv bmZpZy9wdWJsaWNfaHRtbC9kYXNoYm9hcmQvU2NyaXB0cy9NZXRyaWNzQW5hbHl6ZXIuanM6Cisg ICAgICAgIChBbmFseXplci5wcm90b3R5cGUuX2NvdW50VGltZXMpOgorCiAyMDE0LTA5LTA1ICBB bGV4ZXkgUHJvc2t1cnlha292ICA8YXBAYXBwbGUuY29tPgogCiAgICAgICAgIEJ1aWxkYm90IG1l dHJpY3MgcGFnZSBnaXZlcyB3cm9uZyByZXN1bHRzIGFmdGVyIGEgbmV3IGJvdCBnZXRzIGFkZGVk CkluZGV4OiBUb29scy9CdWlsZFNsYXZlU3VwcG9ydC9idWlsZC53ZWJraXQub3JnLWNvbmZpZy9w dWJsaWNfaHRtbC9kYXNoYm9hcmQvU2NyaXB0cy9NZXRyaWNzQW5hbHl6ZXIuanMKPT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09 PQotLS0gVG9vbHMvQnVpbGRTbGF2ZVN1cHBvcnQvYnVpbGQud2Via2l0Lm9yZy1jb25maWcvcHVi bGljX2h0bWwvZGFzaGJvYXJkL1NjcmlwdHMvTWV0cmljc0FuYWx5emVyLmpzCShyZXZpc2lvbiAx NzMzMzMpCisrKyBUb29scy9CdWlsZFNsYXZlU3VwcG9ydC9idWlsZC53ZWJraXQub3JnLWNvbmZp Zy9wdWJsaWNfaHRtbC9kYXNoYm9hcmQvU2NyaXB0cy9NZXRyaWNzQW5hbHl6ZXIuanMJKHdvcmtp bmcgY29weSkKQEAgLTIyMSwxNCArMjIxLDEzIEBAIEFuYWx5emVyLnByb3RvdHlwZSA9IHsKICAg ICB7CiAgICAgICAgIC8vIENvbWJpbmUgYWxsIGl0ZXJhdGlvbnMgdGhhdCBzdGFydGVkIGFuZCBm aW5pc2hlZCB3aXRoaW4gdGhlIHJhbmdlIGludG8gb25lIGFycmF5LgogICAgICAgICAvLyBUaGVz ZSBhcmUgdGhlIGl0ZXJhdGlvbnMgdGhhdCBjYW4gaGF2ZSByZXN1bHRzIGZvciByZXZpc2lvbnMg aW4gdGhlIHJhbmdlLgotICAgICAgICB2YXIgaXRlcmF0aW9uc0J5UmV2aXNpb24gPSBbXTsKKyAg ICAgICAgdmFyIGl0ZXJhdGlvbnMgPSBbXTsKICAgICAgICAgcXVldWVzLmZvckVhY2goZnVuY3Rp b24ocXVldWUpIHsKLSAgICAgICAgICAgIGl0ZXJhdGlvbnNCeVJldmlzaW9uID0gaXRlcmF0aW9u c0J5UmV2aXNpb24uY29uY2F0KHF1ZXVlLml0ZXJhdGlvbnMuZmlsdGVyKGZ1bmN0aW9uKGl0ZXJh dGlvbikgeworICAgICAgICAgICAgaXRlcmF0aW9ucyA9IGl0ZXJhdGlvbnMuY29uY2F0KHF1ZXVl Lml0ZXJhdGlvbnMuZmlsdGVyKGZ1bmN0aW9uKGl0ZXJhdGlvbikgewogICAgICAgICAgICAgICAg IHJldHVybiBpdGVyYXRpb24ucHJvZHVjdGl2ZSAmJiBpdGVyYXRpb24uc3RhcnRUaW1lID4gdGhp cy5fcmFuZ2VTdGFydFRpbWUgJiYgaXRlcmF0aW9uLmVuZFRpbWUgPCB0aGlzLl9yYW5nZUVuZFRp bWU7CiAgICAgICAgICAgICB9LmJpbmQodGhpcykpKTsKICAgICAgICAgfS5iaW5kKHRoaXMpKTsK LSAgICAgICAgaXRlcmF0aW9uc0J5UmV2aXNpb24uc29ydChmdW5jdGlvbihhLCBiKSB7IHJldHVy biBhLmVuZFRpbWUgLSBiLmVuZFRpbWU7IH0pOyAvLyBXaGVuIHRoZXJlIGFyZSBtdWx0aXBsZSBp dGVyYXRpb25zIGJ1aWxkaW5nIHRoZSBzYW1lIHJldmlzaW9uLCB0aGUgZmlyc3Qgb25lIHdpbnMg KGFzIHRoZSBvdGhlciBvbmVzIHdlcmUgcHJvYmFibHkgbWFudWFsIHJldHJpZXMpLgotICAgICAg ICBpdGVyYXRpb25zQnlSZXZpc2lvbi5zb3J0KGZ1bmN0aW9uKGEsIGIpIHsgcmV0dXJuIGEub3Bl blNvdXJjZVJldmlzaW9uIC0gYi5vcGVuU291cmNlUmV2aXNpb247IH0pOworICAgICAgICBpdGVy YXRpb25zLnNvcnQoZnVuY3Rpb24oYSwgYikgeyByZXR1cm4gYS5lbmRUaW1lIC0gYi5lbmRUaW1l OyB9KTsKIAogICAgICAgICAvLyBSZXZpc2lvbnMgdGhhdCBsYW5kZWQgd2l0aGluIHRoZSB0aW1l IHJhbmdlLgogICAgICAgICB2YXIgcmV2aXNpb25zID0gd2Via2l0VHJhYy5yZWNvcmRlZENvbW1p dHMucmVkdWNlKGZ1bmN0aW9uKHJlc3VsdCwgcmV2aXNpb24pIHsKQEAgLTI1MCwxMCArMjQ5LDEy IEBAIEFuYWx5emVyLnByb3RvdHlwZSA9IHsKIAogICAgICAgICB2YXIgcHJldmlvdXNGdWxseVRl c3RlZFJldmlzaW9uTnVtYmVyID0gLTE7CiAKLSAgICAgICAgZm9yICh2YXIgaSA9IDA7IGkgPCBp dGVyYXRpb25zQnlSZXZpc2lvbi5sZW5ndGg7ICsraSkgewotICAgICAgICAgICAgdmFyIGl0ZXJh dGlvbiA9IGl0ZXJhdGlvbnNCeVJldmlzaW9uW2ldOworICAgICAgICBmb3IgKHZhciBpID0gMDsg aSA8IGl0ZXJhdGlvbnMubGVuZ3RoOyArK2kpIHsKKyAgICAgICAgICAgIHZhciBpdGVyYXRpb24g PSBpdGVyYXRpb25zW2ldOworCisgICAgICAgICAgICBpZiAobGFzdFRlc3RlZFJldmlzaW9uQnlR dWV1ZVtpdGVyYXRpb24ucXVldWUuaWRdID49IGl0ZXJhdGlvbi5vcGVuU291cmNlUmV2aXNpb24p CisgICAgICAgICAgICAgICAgY29udGludWU7CiAKLSAgICAgICAgICAgIGNvbnNvbGUuYXNzZXJ0 KGxhc3RUZXN0ZWRSZXZpc2lvbkJ5UXVldWVbaXRlcmF0aW9uLnF1ZXVlLmlkXSA9PT0gdW5kZWZp bmVkIHx8IGxhc3RUZXN0ZWRSZXZpc2lvbkJ5UXVldWVbaXRlcmF0aW9uLnF1ZXVlLmlkXSA8PSBp dGVyYXRpb24ub3BlblNvdXJjZVJldmlzaW9uKTsKICAgICAgICAgICAgIGxhc3RUZXN0ZWRSZXZp c2lvbkJ5UXVldWVbaXRlcmF0aW9uLnF1ZXVlLmlkXSA9IGl0ZXJhdGlvbi5vcGVuU291cmNlUmV2 aXNpb247CiAKICAgICAgICAgICAgIHZhciBuZXdGdWxseVRlc3RlZFJldmlzaW9uTnVtYmVyID0g dGhpcy5fZnVsbHlUZXN0ZWRSZXZpc2lvbk51bWJlcihsYXN0VGVzdGVkUmV2aXNpb25CeVF1ZXVl KTsK