There is an assertion/crash that occurs in XPCShell when setting SORT_BY_ANNOTATION_* on a query. It happens on both ascending and descending sorts. I will attach a testcase. Here is the output from XPCShell: <<<<<<< ../../../../_tests/xpcshell-simple/test_places/queries/test_simple-anno-crash.js: FAIL ../../../../_tests/xpcshell-simple/test_places/queries/test_simple-anno-crash.js.log: >>>>>>> +++ JavaScript debugging hooks installed. *** test pending ###!!! ASSERTION: Invalid sorting mode: 'Not Reached', file /Users/clint/code/fx-trunk/mozilla/toolkit/components/places/src/nsNavHistory.cpp, line 3193 UNKNOWN [/Users/clint/code/fx-trunk/mozilla/dbg/i386/dist/bin/components/libplaces.dylib +0x00013623] UNKNOWN [/Users/clint/code/fx-trunk/mozilla/dbg/i386/dist/bin/components/libplaces.dylib +0x000177A6] UNKNOWN [/Users/clint/code/fx-trunk/mozilla/dbg/i386/dist/bin/components/libplaces.dylib +0x00017D9C] UNKNOWN [/Users/clint/code/fx-trunk/mozilla/dbg/i386/dist/bin/components/libplaces.dylib +0x0002C84C] UNKNOWN [/Users/clint/code/fx-trunk/mozilla/dbg/i386/dist/bin/components/libplaces.dylib +0x00046718] UNKNOWN [/Users/clint/code/fx-trunk/mozilla/dbg/i386/dist/bin/components/libplaces.dylib +0x00046B0E] UNKNOWN [/Users/clint/code/fx-trunk/mozilla/dbg/i386/dist/bin/components/libplaces.dylib +0x000424AD] nsCOMArray_base::Sort(int (*)(void const*, void const*, void*), void*)+0x00002BD8 [/Users/clint/code/fx-trunk/mozilla/dbg/i386/dist/bin/components/libplaces.dylib +0x000850D4] NS_InvokeByIndex_P+0x00000062 [../../../../dist/bin/libxpcom_core.dylib +0x0007C2C2] NSGetModule+0x000130A3 [/Users/clint/code/fx-trunk/mozilla/dbg/i386/dist/bin/components/libxpconnect.dylib +0x00040227] nsCRT::IsUpper(char)+0x00001515 [/Users/clint/code/fx-trunk/mozilla/dbg/i386/dist/bin/components/libxpconnect.dylib +0x0007DCD9] NSGetModule+0x0001ADEB [/Users/clint/code/fx-trunk/mozilla/dbg/i386/dist/bin/components/libxpconnect.dylib +0x00047F6F] js_Invoke+0x00000A22 [../../../../dist/bin/libmozjs.dylib +0x00077ADC] js_Invoke+0x00000D5A [../../../../dist/bin/libmozjs.dylib +0x00077E14] js_Invoke+0x00000FA5 [../../../../dist/bin/libmozjs.dylib +0x0007805F] js_FindProperty+0x000008F4 [../../../../dist/bin/libmozjs.dylib +0x0008CC11] js_FindProperty+0x00002024 [../../../../dist/bin/libmozjs.dylib +0x0008E341] JS_CompareValues+0x0000DB56 [../../../../dist/bin/libmozjs.dylib +0x000697B7] js_Invoke+0x0000130F [../../../../dist/bin/libmozjs.dylib +0x000783C9] JS_ExecuteScript+0x00000082 [../../../../dist/bin/libmozjs.dylib +0x00017777] tart+0x00001799 [/Users/clint/code/fx-trunk/mozilla/dbg/i386/toolkit/components/places/tests/../../../../dist/bin/xpcshell +0x00002039] tart+0x00001AD6 [/Users/clint/code/fx-trunk/mozilla/dbg/i386/toolkit/components/places/tests/../../../../dist/bin/xpcshell +0x00002376] tart+0x0000212C [/Users/clint/code/fx-trunk/mozilla/dbg/i386/toolkit/components/places/tests/../../../../dist/bin/xpcshell +0x000029CC] tart+0x000033D0 [/Users/clint/code/fx-trunk/mozilla/dbg/i386/toolkit/components/places/tests/../../../../dist/bin/xpcshell +0x00003C70] tart+0x00000102 [/Users/clint/code/fx-trunk/mozilla/dbg/i386/toolkit/components/places/tests/../../../../dist/bin/xpcshell +0x000009A2] tart+0x00000029 [/Users/clint/code/fx-trunk/mozilla/dbg/i386/toolkit/components/places/tests/../../../../dist/bin/xpcshell +0x000008C9] It looks like we are falling through the switch statement here: http://mxr.mozilla.org/mozilla/source/toolkit/components/places/src/nsNavHistoryResult.cpp#662 = Expected Behavior = Don't assert/crash.

Also happens when using SORT_BY_TAG_* sorting methods. I think we just need some kind of smart handling for TAG and ANNOTATION sorting methods in this switch statement so we don't assert. http://mxr.mozilla.org/mozilla/source/toolkit/components/places/src/nsNavHistory.cpp#3198

drivers: busted sorting options in the api, fallout from bug 385245.

Is there an ETA on a fix?

(In reply to comment #4) > drivers: busted sorting options in the api, fallout from bug 385245. It was missing at this place already before the bug 385245, and the test was not yet checked in, so the question is whether the test is not simply wrong or in better words expecting too much from the existing framework. However, I will check if it is possible to implement it.

The testing script is incomplete, when you use this options.sortingMode = options.SORT_BY_ANNOTATION_ASCENDING; you should say what annotation do we use for sorting: options.sortingAnnotation = annoN; According to IDL it should be called before setting sortingMode. Setting annotation on query influences just filtering. The simple patch that I'm submitting seems to solve the problem. I have tested sorting by both ANNOTATION and by TAGS and it works.

Comment on attachment 313912 [details] [diff] [review] Do not assert on ANNO and TAGS sorting r=me, thanks.

test can be added once bug 384226 lands

Checking in toolkit/components/places/src/nsNavHistory.cpp; /cvsroot/mozilla/toolkit/components/places/src/nsNavHistory.cpp,v <-- nsNavHistory.cpp new revision: 1.285; previous revision: 1.284 done

backed out due to unit test orange

(In reply to comment #11) > backed out due to unit test orange I doubt it could be my patch. Tests in toolkit are ok, and I can see test in browser places failing too with latest build. Nothing in them looks like they would expect an assertion on the now added sorting modes.

yes, likely not. i'll re-check-in tonight.

Checking in toolkit/components/places/src/nsNavHistory.cpp; /cvsroot/mozilla/toolkit/components/places/src/nsNavHistory.cpp,v <-- nsNavHistory.cpp new revision: 1.287; previous revision: 1.286 done

Bug 451915 - move Firefox/Places bugs to Firefox/Bookmarks and History. Remove all bugspam from this move by filtering for the string "places-to-b-and-h". In Thunderbird 3.0b, you do that as follows: Tools | Message Filters Make sure the correct account is selected. Click "New" Conditions: Body contains places-to-b-and-h Change the action to "Delete Message". Select "Manually Run" from the dropdown at the top. Click OK. Select the filter in the list, make sure "Inbox" is selected at the bottom, and click "Run Now". This should delete all the bugspam. You can then delete the filter. Gerv