[cfe-commits] Add PointerEscapeKind to checkPointerEscape callback

[Jordan Rose]

On Jan 18, 2013, at 18:55 , Anna Zaks <ganna at apple.com> wrote:

> 
> On Jan 18, 2013, at 6:43 PM, Jordan Rose <jordan_rose at apple.com> wrote:
> 
>> 
>> 
>> - if (Call && doesNotFreeMemory(Call, State))
>> +  if ((Kind == PSK_DirectEscapeOnCall ||
>> +      Kind == PSK_IndirectEscapeOnCall) &&
>> +      doesNotFreeMemory(Call, State)) {
>> 
>> This is not correct. Before, this branch was only taken if the Kind is PSK_DirectEscapeOnCall. Indirect escapes can still possibly free memory (although it's unlikely).
>> 
> 
> Jordan,
> 
> I think the condition is correct. In fact, this is the whole point of this commit.
> For example, previously, we would assume that a call to foo_that_does_not_free_memory(p+1), could free memory.
> 
> Branden, can you add a test case that shows this? I know the test cases in the second patch rely on this, but a self contained example would make the change explicit. (Ex: you could call a function from a system header and pass it p+1.)

...ah. Okay, right. Carry on.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/cfe-commits/attachments/20130118/96dbed88/attachment.html>