[llvm-commits] [llvm] r71033 - /llvm/trunk/lib/Transforms/Scalar/LoopStrengthReduce.cpp

Dan Gohman gohman at apple.com
Tue May 5 15:59:59 PDT 2009


Author: djg
Date: Tue May  5 17:59:55 2009
New Revision: 71033

URL: http://llvm.org/viewvc/llvm-project?rev=71033&view=rev
Log:
Delete a FIXME which is no longer relevant, and add a FIXME that is.

Modified:
    llvm/trunk/lib/Transforms/Scalar/LoopStrengthReduce.cpp

Modified: llvm/trunk/lib/Transforms/Scalar/LoopStrengthReduce.cpp
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Transforms/Scalar/LoopStrengthReduce.cpp?rev=71033&r1=71032&r2=71033&view=diff

==============================================================================
--- llvm/trunk/lib/Transforms/Scalar/LoopStrengthReduce.cpp (original)
+++ llvm/trunk/lib/Transforms/Scalar/LoopStrengthReduce.cpp Tue May  5 17:59:55 2009
@@ -2464,8 +2464,9 @@
     // computation of some other indvar to decide when to terminate the loop.
     OptimizeIndvars(L);
 
-    // FIXME: We can widen subreg IV's here for RISC targets.  e.g. instead of
-    // doing computation in byte values, promote to 32-bit values if safe.
+    // FIXME: We can widen overlarge IV's here.  e.g. if the code has
+    // computation in i64 values and the target doesn't support i64, demote
+    // the computation to 32-bit if safe.
 
     // FIXME: Attempt to reuse values across multiple IV's.  In particular, we
     // could have something like "for(i) { foo(i*8); bar(i*16) }", which should





More information about the llvm-commits mailing list