projects / oota-llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: ae218de) | patch
In the ARM global merging pass, allow extraneous alignment specifiers. This pass
authorCameron Zwarich <zwarich@apple.com>
Wed, 29 Jun 2011 22:24:25 +0000 (22:24 +0000)
committerCameron Zwarich <zwarich@apple.com>
Wed, 29 Jun 2011 22:24:25 +0000 (22:24 +0000)
commitfaff12731968c2d3f1ff56a43749d27e1696aafa
tree3f8acbcb39deb74a8e73625364f6d8bbfdd8fdf6tree | snapshot
parentae218dee5e3a079becd6b9b8d47e67cf814b9b70commit | diff
In the ARM global merging pass, allow extraneous alignment specifiers. This pass
already makes the assumption, which is correct on ARM, that a type's alignment is
less than its alloc size. This improves codegen with Clang (which inserts a lot of
extraneous alignment specifiers) and fixes <rdar://problem/9695089>.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@134106 91177308-0d34-0410-b5e6-96231b3b80d8
lib/Target/ARM/ARMGlobalMerge.cpp diff | blob | history
test/CodeGen/ARM/2011-06-29-MergeGlobalsAlign.ll [new file with mode: 0644] blob
C/C++ LLVM-based compilers that forbids OOTA behaviors