maxjhandsome
/
llvm-project
forked from OSchip/llvm-project
1
0
Fork 0
Code Issues Pull Requests Packages Releases Wiki Activity

add a note 

llvm-svn: 47720
This commit is contained in:
Chris Lattner 2008-02-28 17:21:27 +00:00
parent e28886a913
commit 643dce1146
1 changed files with 19 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

19
llvm/lib/Target/README.txt
View File

@ -631,6 +631,25 @@ implementations of ceil/floor/rint.
//===---------------------------------------------------------------------===// //===---------------------------------------------------------------------===//
This GCC bug: http://gcc.gnu.org/bugzilla/show_bug.cgi?id=34043
contains a testcase that compiles down to:
%struct.XMM128 = type { <4 x float> }
..
%src = alloca %struct.XMM128
..
%tmp6263 = bitcast %struct.XMM128* %src to <2 x i64>*
%tmp65 = getelementptr %struct.XMM128* %src, i32 0, i32 0
store <2 x i64> %tmp5899, <2 x i64>* %tmp6263, align 16
%tmp66 = load <4 x float>* %tmp65, align 16
%tmp71 = add <4 x float> %tmp66, %tmp66
If the mid-level optimizer turned the bitcast of pointer + store of tmp5899
into a bitcast of the vector value and a store to the pointer, then the
store->load could be easily removed.
//===---------------------------------------------------------------------===//
Consider: Consider:
int test() { int test() {