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

Optimized stores to the constant pool, while cool, are unnecessary. 

llvm-svn: 27948
This commit is contained in:
Nate Begeman 2006-04-22 22:31:45 +00:00
parent 4ca2ea5b43
commit 9f0b13c885
1 changed files with 0 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
llvm/lib/Target/X86/X86InstrInfo.td
View File

@ -2323,9 +2323,6 @@ def : Pat<(add R32:$src1, (X86Wrapper tglobaladdr :$src2)),
def : Pat<(add R32:$src1, (X86Wrapper texternalsym:$src2)),
(ADD32ri R32:$src1, texternalsym:$src2)>;
// FIXME: can you really ever store to a constant pool?
def : Pat<(store (X86Wrapper tconstpool:$src), addr:$dst),
(MOV32mi addr:$dst, tconstpool:$src)>;
def : Pat<(store (X86Wrapper tglobaladdr:$src), addr:$dst),
(MOV32mi addr:$dst, tglobaladdr:$src)>;
def : Pat<(store (X86Wrapper texternalsym:$src), addr:$dst),