[PATCH] Fix a crash in Retain Count checker error reporting

[Anna Zaks]

> On Jun 13, 2014, at 1:36 PM, Jordan Rose <jordan_rose at apple.com> wrote:
> 
> Why does this happen with this test case? AFAIK the "there are no such calls" note is still correct.
> 

Attached is an updated patch that reinstates the "there are no such calls" note and explains why the BlockEdge is possible in the comment.

> Since this is testing where the allocation statement is located, can you make this a path-notes/plist test, so that we can verify that it's actually picking the right statement?

This is not testing where the allocation statement is located but corrects a case where we were previously crashing.

> 
> Jordan
> 
> 
> On Jun 13, 2014, at 13:33 , Anna Zaks <ganna at apple.com> wrote:
> 
>> Fixes a crash in Retain Count checker error reporting logic by handing the allocation statement retrieval from a BlockEdge program point.
>> 
>> Anna.
>> <fixCrashInRetainCountReporting.diff>
> 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: fixCrashInRetainCountReporting2.diff
Type: application/octet-stream
Size: 4718 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/cfe-commits/attachments/20140613/b76c6adf/attachment.obj>