llvm-svn: 115414

Need to specify SSE4 for machines which don't have SSE4. The code checked for is generated by SSE4. Otherwise, we get something else.

llvm-svn: 115352

Now that the profitable bits of EnableFullLoadPRE have been enabled by default, rip out the remainder.
Anyone interested in more general PRE would be better served by implementing it separately, to get real
anticipation calculation, etc.

llvm-svn: 115337

Fix code gen crash reported in PR 8235. We still lose debug info for the unused argument here. This is a known limitation recorded debuginfo-tests/trunk/dbg-declare2.ll function 'f6' test case.

llvm-svn: 115323

llvm-svn: 115309

llvm-svn: 115308

llvm-svn: 115307

Also remove some code that died in the process.
One now non-existant ori is checked for.

llvm-svn: 115306

llvm-svn: 115304

llvm-svn: 115303

llvm-svn: 115296

llvm-svn: 115295

llvm-svn: 115294

llvm-svn: 115293

llvm-svn: 115292

maybe even Linux.

llvm-svn: 115291

llvm-svn: 115257

llvm-svn: 115250

The x86_mmx type is used for MMX intrinsics, parameters and
return values where these use MMX registers, and is also
supported in load, store, and bitcast.

Only the above operations generate MMX instructions, and optimizations
do not operate on or produce MMX intrinsics. 

MMX-sized vectors <2 x i32> etc. are lowered to XMM or split into
smaller pieces.  Optimizations may occur on these forms and the
result casted back to x86_mmx, provided the result feeds into a
previous existing x86_mmx operation.

The point of all this is prevent optimizations from introducing
MMX operations, which is unsafe due to the EMMS problem.

llvm-svn: 115243

llvm-svn: 115215

We do want to allow LoadPRE to perform LICM-like transformations: we already consider PHI nodes to be negligible for
code size (making this transform code size neutral), and it allows us to hoist values out of loops, which is always
a good thing.

llvm-svn: 115205

The bug that broke i386 linux has been fixed in r115191.

llvm-svn: 115204

fixes some cases where we were producing relocations with at symbol that
should use a section instead.

llvm-svn: 115194

edited during emission.

If the basic block ends in a switch that gets lowered to a jump table, any
phis at the default edge were getting updated wrong. The jump table data
structure keeps a pointer to the header blocks that wasn't getting updated
after the MBB is split.

This bug was exposed on 32-bit Linux when disabling critical edge splitting in
codegen prepare.

The fix is to uipdate stale MBB pointers whenever a block is split during
emission.

llvm-svn: 115191

llvm-svn: 115178

llvm-svn: 115177

llvm-svn: 115175

llvm-svn: 115174

llvm-svn: 115172

llvm-svn: 115171

llvm-svn: 115170

llvm-svn: 115168

llvm-svn: 115165

llvm-svn: 115164

llvm-svn: 115163

llvm-svn: 115162

llvm-svn: 115161

llvm-svn: 115156

llvm-svn: 115154

time. That way, the EntrySize field is initialized for other code paths, 
namely, the .ll -> .o code path.

llvm-svn: 115141