Jeff Walden [:Waldo] Reporter
	Description • 17 years ago

xpcshell tests in debug builds fail because they hit an assertion in netwerk, on three different tests; see bug 372581 comment 10 for the stack.  Binary search to within a few minutes (rather more painful than you might think given bug 394647 was a very tempting red herring) -- shows it's fallout from bug 403885.

This, among other issues, is preventing debug tinderboxen with fatal assertions from happening.

	Florian Quèze [:florian] Assignee
	Comment 1 • 17 years ago

While running the failing tests, before the assertion there was a JS exception:
[Exception... "Component returned failure code: 0x80004005 (NS_ERROR_FAILURE) [nsIProperties.get]"  nsresult: "0x80004005 (NS_ERROR_FAILURE)"  location: "JS frame :: file:///Users/fqueze/moz/mozilla/obj/dist/bin/components/nsHandlerService.js :: anonymous :: line 800"  data: no]

The attached patch restores the previous behavior of _updateDB: all exceptions are caught.

Note that I don't really know why this code is executed in these tests.

The test I used while debugging is http://mxr.mozilla.org/seamonkey/source/netwerk/test/unit/test_reopen.js (we get the assertion when testing test_file_channel).

	Dan Mosedale (:dmosedale, :dmose)
	Comment 2 • 17 years ago

Comment on attachment 294991 [details] [diff] [review]
patch v1

Looks good to me; be sure to add your name to the contributors list in the license header.

	Christian :Biesinger (don't email me, ping me on IRC)
	Comment 3 • 17 years ago

Comment on attachment 294991 [details] [diff] [review]
patch v1

not really a fan of JS scoping, but sr=biesi

	Florian Quèze [:florian] Assignee
	Comment 4 • 17 years ago

Checking in uriloader/exthandler/nsHandlerService.js;
/cvsroot/mozilla/uriloader/exthandler/nsHandlerService.js,v  <--  nsHandlerService.js
new revision: 1.19; previous revision: 1.18
done