On Darwin / Linux x86-32, v8i8, v4i16, v2i32 values are passed in MM[0-2].                                                                                                                                      
On Darwin / Linux x86-32, v1i64 values are passed in memory.                                                                                                                                                    
On Darwin x86-64, v8i8, v4i16, v2i32 values are passed in XMM[0-7].                                                                                                                                     
On Darwin x86-64, v1i64 values are passed in 64-bit GPRs.

llvm-svn: 50257

idea what this code (findNonImmUse) does, so I'm only guessing 
that this is the right thing.  It would be really really nice
if this had comments and perhaps switched to SmallPtrSet
(hint hint) :)

This fixes rdar://5886601, a crash on gcc.target/i386/sse4_1-pblendw.c

llvm-svn: 50252

Fix bug in x86 memcpy / memset lowering. If there are trailing bytes not handled by rep instructions, a new memcpy / memset is introduced for them. However, since source / destination addresses are already adjusted, their offsets should be zero.

llvm-svn: 50239

llvm-svn: 50169

llvm-svn: 50168

Be over-conservative: scan for all used virtual registers and calculate maximal stack alignment in assumption, that there will be spill of vector register.

llvm-svn: 50167

llvm-svn: 50166

llvm-svn: 50163

llvm-svn: 50162

Eastimate required stack alignment early, so we can decide, whether we will need frame pointer or not

llvm-svn: 50161

llvm-svn: 50159

llvm-svn: 50158

llvm-svn: 50157

llvm-svn: 50156

llvm-svn: 50155

llvm-svn: 50154

llvm-svn: 50153

llvm-svn: 50152

argument. The x86-64 ABI requires the incoming value of %rdi to
be copied to %rax on exit from a function that is returning a
large C struct.

Also, add a README-X86-64 entry detailing the missed optimization
opportunity and proposing an alternative approach.

llvm-svn: 50075

llvm-svn: 50053

llvm-svn: 50029

llvm-svn: 50007

llvm-svn: 49986

llvm-svn: 49977

llvm-svn: 49948

llvm-svn: 49946

llvm-gcc is now lowering va_copy on x86-64, so this completes
the fix for PR2230.

llvm-svn: 49922

- Remove unused instructions.

llvm-svn: 49921

llvm-svn: 49878

llvm-svn: 49869

llvm-svn: 49830

llvm-svn: 49822

Correlate stubs with functions in JIT: when emitting a stub, the JIT tells the memory manager which function
the stub will resolve.

llvm-svn: 49814

Infrastructure for getting the machine code size of a function and an instruction. X86, PowerPC and ARM are implemented

llvm-svn: 49809

llvm-svn: 49808

Rename SDOperandImpl back to SDOperand.
Introduce the SDUse class that represents a use of the SDNode referred by
an SDOperand. Now it is more similar to Use/Value classes.

Patch is approved by Dan Gohman.

llvm-svn: 49795

puts its result in a 32-bit GPR.

llvm-svn: 49762

memcpy lowering code; this ensures that the size node has the desired
result type. This fixes a regression from r49572 with @llvm.memcpy.i64
on x86-32.

llvm-svn: 49761

llvm-svn: 49759

to 64-bit GPR registers on x86-64.

llvm-svn: 49757