[llvm-dev] One more No-alias case on Alias analysis

[Friedman, Eli via llvm-dev]

On 6/11/2018 10:06 AM, jingu at codeplay.com via llvm-dev wrote:
> Hello All,
>
> I have met one may-alias case from llvm's alias analysis. The code 
> snippet is as following:
>
> char buf[4];
>
> void test (int idx) {
> char *a = &buf[3 - idx];
> char *b = &buf[idx];
> *a = 1;
> *b = 2;
> }
>
> I can see below output from alias set tracker for above code snippet.
>
> Alias sets for function 'test':
> Alias Set Tracker: 1 alias sets for 2 pointer values.
>   AliasSet[0x53d8070, 2] may alias, Mod       Pointers: (i8* 
> %arrayidx, 1), (i8* %arrayidx2, 1)
>
> As you can see on above code snippet, the 'a' and 'b' are not aliased. 
> I think if we have following offset form, we can say No-alias between 
> them.
>
> offset1 = odd_number - index
>
> offset2 = index
>
> I have implemented simple code for it and the output is as following:
>
> Alias sets for function 'test':
> Alias Set Tracker: 2 alias sets for 2 pointer values.
>   AliasSet[0x541a070, 1] must alias, Mod       Pointers: (i8* 
> %arrayidx, 1)
>   AliasSet[0x541cc00, 1] must alias, Mod       Pointers: (i8* 
> %arrayidx2, 1)
>
> How do you think about this? Is it legal for current alias analysis or 
> not? I have attached the diff file as reference. If I missed 
> something, please let me know. 

The concept works. I'm not sure your patch handles all the edge cases 
correctly, at first glance.  (If you want a full review, please post on 
Phabricator.)

-Eli

-- 
Employee of Qualcomm Innovation Center, Inc.
Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum, a Linux Foundation Collaborative Project