On Sun, 2007-01-28 at 20:31 -0800, Chris Lattner wrote: > On Jan 28, 2007, at 7:52 PM, Reid Spencer wrote: > > > On Sun, 2007-01-28 at 16:04 -0800, Chris Lattner wrote: > >>> For PR1138: http://llvm.org/PR1138 : > >>> Force memcpy to be the 32-bit variant. Since this is only used with > >>> CBE and lli which both target 32-bit machines, this should be okay. > >> > >> What do you mean, "which both target 32-bit machines"? CBE supports > >> Alpha, sparcv9, x86-64, and ppc64. > > > > Just going on statements you made in IRC about CBE targeting only > > 32-bit > > machines. > > I don't recall making that the statement, but even if I did, it is > certainly not true :)
I gathered that :) I'll use TD to get the right size and trunc or zext up to deal with 32/64 mismatches. However, this will take a bit. IntrinsicLowering doesn't have TargetData so I have to pass it in from the callers. If I don't get to it tonight, I'll do it tomorrow. > > -Chris _______________________________________________ llvm-commits mailing list llvm-commits@cs.uiuc.edu http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits