[llvm-commits] [llvm] r48438 - /llvm/trunk/lib/Target/README.txt
Chris Lattner
sabre at nondot.org
Sun Mar 16 18:47:51 PDT 2008
Author: lattner
Date: Sun Mar 16 20:47:51 2008
New Revision: 48438
URL: http://llvm.org/viewvc/llvm-project?rev=48438&view=rev
Log:
various rotate fun.
Modified:
llvm/trunk/lib/Target/README.txt
Modified: llvm/trunk/lib/Target/README.txt
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/README.txt?rev=48438&r1=48437&r2=48438&view=diff
==============================================================================
--- llvm/trunk/lib/Target/README.txt (original)
+++ llvm/trunk/lib/Target/README.txt Sun Mar 16 20:47:51 2008
@@ -728,3 +728,47 @@
exit value computation.
//===---------------------------------------------------------------------===//
+
+We miss a bunch of rotate opportunities on various targets, including ppc, x86,
+etc. On X86, we miss a bunch of 'rotate by variable' cases because the rotate
+matching code in dag combine doesn't look through truncates aggressively
+enough. Here are some testcases reduces from GCC PR17886:
+
+unsigned long long f(unsigned long long x, int y) {
+ return (x << y) | (x >> 64-y);
+}
+unsigned f2(unsigned x, int y){
+ return (x << y) | (x >> 32-y);
+}
+unsigned long long f3(unsigned long long x){
+ int y = 9;
+ return (x << y) | (x >> 64-y);
+}
+unsigned f4(unsigned x){
+ int y = 10;
+ return (x << y) | (x >> 32-y);
+}
+unsigned long long f5(unsigned long long x, unsigned long long y) {
+ return (x << 8) | ((y >> 48) & 0xffull);
+}
+unsigned long long f6(unsigned long long x, unsigned long long y, int z) {
+ switch(z) {
+ case 1:
+ return (x << 8) | ((y >> 48) & 0xffull);
+ case 2:
+ return (x << 16) | ((y >> 40) & 0xffffull);
+ case 3:
+ return (x << 24) | ((y >> 32) & 0xffffffull);
+ case 4:
+ return (x << 32) | ((y >> 24) & 0xffffffffull);
+ default:
+ return (x << 40) | ((y >> 16) & 0xffffffffffull);
+ }
+}
+
+On X86-64, we only handle f3/f4 right. On x86-32, several of these
+generate truly horrible code, instead of using shld and friends. On
+ARM, we end up with calls to L___lshrdi3/L___ashldi3 in f, which is
+badness. PPC64 misses f, f5 and f6. CellSPU aborts in isel.
+
+//===---------------------------------------------------------------------===//
More information about the llvm-commits
mailing list