[llvm-dev] noalias parameter attribute not currently exploited by alias analysis?

[Alex Bradbury via llvm-dev]

On 9 November 2015 at 14:30, Hal Finkel <hfinkel at anl.gov> wrote:
> Thanks for reporting this issue, and I apologize for the delayed reply. It seems you've run into a limitation of the current implementation, and please do file a bug report.
>
> If I take your IR, and run it through:
>
>   opt -basicaa -scoped-noalias -aa-eval -evaluate-aa-metadata -print-all-alias-modref-info -disable-output 2>&1 < /tmp/your/input.ll
>
> there seems to be a limitation with the way that pointer provenance is determined. Looking at the mod/ref info, things quickly become suboptimal:
>
>   NoAlias:   %4 = load i8, i8* %.04.prol, align 1, !tbaa !1 <->   store i8 %11, i8* %.013.prol, align 1, !tbaa !1
>   NoAlias:   %4 = load i8, i8* %.04.prol, align 1, !tbaa !1 <->   store i8 %14, i8* %18, align 1, !tbaa !1
>   NoAlias:   %4 = load i8, i8* %.04.prol, align 1, !tbaa !1 <->   store i8 %17, i8* %19, align 1, !tbaa !1
>   MayAlias:   %4 = load i8, i8* %.04.prol, align 1, !tbaa !1 <->   store i8 %31, i8* %.013, align 1, !tbaa !1
>   MayAlias:   %4 = load i8, i8* %.04.prol, align 1, !tbaa !1 <->   store i8 %34, i8* %38, align 1, !tbaa !1
>   ...
>
> it seems like looking through one PHI is okay, but as soon as we need to look though both the PHI in the loop and also the PHI in the preheader, we're unable to do so conclusively. This seems to be the underlying problem.

Thanks for confirming I'm not missing anything obvious here. I'll file
an issue later today.

I don't suppose you had any thoughts on the 'Unknown' dependencies
reported in -print-memdeps? Their occurrence seems rather mysterious
to me - even more mysterious that some 'Unknown' dependencies can be
associated with a basic block but not others, despite the code being
perfectly regular.

Cheers,

Alex