Bug 22950 - XMLHttpRequest preflight cache tests are failing on Tiger.
Summary: XMLHttpRequest preflight cache tests are failing on Tiger.
Status: RESOLVED FIXED
Alias: None
Product: WebKit
Classification: Unclassified
Component: Tools / Tests (show other bugs)
Version: 528+ (Nightly build)
Hardware: Mac (PowerPC) OS X 10.4
: P2 Normal
Assignee: David Levin
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2008-12-20 17:43 PST by David Levin
Modified: 2008-12-27 00:30 PST (History)
1 user (show)

See Also:


Attachments
Fix for bug. (4.15 KB, patch)
2008-12-20 17:58 PST, David Levin
no flags Details | Formatted Diff | Diff
Add logging to detect why failures are happening on the Tiger buildbot. (2.23 KB, patch)
2008-12-26 02:49 PST, David Levin
no flags Details | Formatted Diff | Diff
Adds a default temp location of "/tmp" to our implementation of sys_get_temp_dir for when no other tmp dir could be determined. (4.81 KB, patch)
2008-12-26 11:56 PST, David Levin
ap: review+
Details | Formatted Diff | Diff
Remove ensureTrailingSlash from one more place. (1.12 KB, patch)
2008-12-26 13:27 PST, David Levin
ap: review+
Details | Formatted Diff | Diff

Note You need to log in before you can comment on or make changes to this bug.
Description David Levin 2008-12-20 17:43:43 PST
They should be fixed.
Comment 1 David Levin 2008-12-20 17:58:43 PST
Created attachment 26176 [details]
Fix for bug.

I'm attempting a fix for the leopard tests run failure, but I don't have access to a leopard machine to verify that this fixes it.

Here's why I think this will fix it.  By scanning the build bot, I found that the tests were passing at r39188 and started before failing at r39217.

I looked through all of these changes and r39202 seemed like the mostly likely candidate to have caused these failures, so I've addressed what seemed like a bug in the change (which will hopefully fix the test run on leopard).
Comment 2 David Levin 2008-12-20 20:59:36 PST
In my previous comment s/leopard/tiger/.


Comment 3 Darin Adler 2008-12-21 09:55:24 PST
Comment on attachment 26176 [details]
Fix for bug.

r=me
Comment 4 Alexey Proskuryakov 2008-12-21 23:31:48 PST
Committed revision 39434.

Comment 5 David Levin 2008-12-22 08:54:30 PST
Looking at the buildbot, it appears that this fix didn't help.  I think I have found a tiger machine, so I'll be able to debug it there soon.


Comment 6 David Levin 2008-12-26 02:49:26 PST
Created attachment 26252 [details]
Add logging to detect why failures are happening on the Tiger buildbot.
Comment 7 Alexey Proskuryakov 2008-12-26 03:02:54 PST
Comment on attachment 26252 [details]
Add logging to detect why failures are happening on the Tiger buildbot.

r=me
Comment 8 Alexey Proskuryakov 2008-12-26 03:05:27 PST
Landed the patch for additional logging in r39476. Not closing the bug, as it couldn't resolve it (I'll also clear review flags from landed patches).
Comment 9 David Levin 2008-12-26 11:56:55 PST
Created attachment 26258 [details]
Adds a default temp location of "/tmp" to our implementation of sys_get_temp_dir for when no other tmp dir could be determined.
Comment 10 Alexey Proskuryakov 2008-12-26 11:59:08 PST
Comment on attachment 26258 [details]
Adds a default temp location of "/tmp" to our implementation of sys_get_temp_dir for when no other tmp dir could be determined.

r=me
Comment 11 Alexey Proskuryakov 2008-12-26 12:01:13 PST
Committed revision 39482.

Comment 12 David Levin 2008-12-26 13:27:29 PST
Created attachment 26259 [details]
Remove ensureTrailingSlash from one more place.

The tests are still failing on Tiger so there is still more to be done to finish fixing this but my last fix missed remove this function call from the appcache php file.
Comment 13 Alexey Proskuryakov 2008-12-26 13:32:02 PST
Committed revision 39485.

Comment 14 Alexey Proskuryakov 2008-12-27 00:30:57 PST
One more fix suggested by Dave committed revision 39486.