[PATCH] Specify the access behaviour of the memcpy, memmove and memset intrinsics

Arnaud A. de Grandmaison arnaud.adegm at gmail.com
Tue May 21 00:52:32 PDT 2013


Hi Chandler,

Now that we are done with the conferences and vacations :), can we
finish implementing this feature ?

Note:

After implementing the new API with this patch, the next steps would be :
 - Upgrade all tests (but auto_upgrade_intrinsics) to use the new format
for memcpy/memmove. A boringly mechanical patch.
 - Kill setVolatile in favor of the new setDestVolatile.
 - Kill uses of isVolatile in favor of the new isDestVolatile /
isSourceVolatile

All those steps (including the patch under review) impact LLVM & Clang
code bases.

Cheers,

-- 
Arnaud A. de Grandmaison


On 05/01/2013 09:58 PM, Arnaud Allard de Grandmaison wrote:
>   Fix ARMFastISel/SelectCall calls.
>
>   I just found out I was not building the ARM target and did not caught this trivial error :(.
>
> Hi chandlerc,
>
> http://llvm-reviews.chandlerc.com/D386
>
> CHANGE SINCE LAST DIFF
>   http://llvm-reviews.chandlerc.com/D386?vs=1793&id=1795#toc
>
> Files:
>   docs/LangRef.rst
>   include/llvm/IR/IntrinsicInst.h
>   include/llvm/IR/Intrinsics.td
>   lib/CodeGen/SelectionDAG/SelectionDAGBuilder.cpp
>   lib/IR/AutoUpgrade.cpp
>   lib/IR/IRBuilder.cpp
>   lib/IR/Verifier.cpp
>   lib/Target/ARM/ARMFastISel.cpp
>   lib/Transforms/Utils/InlineFunction.cpp
>   test/Analysis/TypeBasedAliasAnalysis/memcpyopt.ll
>   test/Assembler/auto_upgrade_intrinsics.ll
>   test/Transforms/DeadStoreElimination/OverwriteStoreEnd.ll
>   test/Transforms/InstCombine/objsize.ll
>   test/Transforms/InstCombine/sprintf-1.ll
>   test/Transforms/LoopIdiom/basic.ll
>   test/Transforms/MemCpyOpt/align.ll
>   test/Transforms/MemCpyOpt/smaller.ll
>   test/Verifier/2006-12-12-IntrinsicDefine.ll




More information about the llvm-commits mailing list