Bill Wendling 37b52ee6d9 Micro-optimization: ... This code: float floatingPointComparison(float x, float y) { double product = (double)x * y; if (product == 0.0) return product; return product - 1.0; } produces this: _floatingPointComparison: 0000000000000000 cvtss2sd %xmm1,%xmm1 0000000000000004 cvtss2sd %xmm0,%xmm0 0000000000000008 mulsd %xmm1,%xmm0 000000000000000c pxor %xmm1,%xmm1 0000000000000010 ucomisd %xmm1,%xmm0 0000000000000014 jne 0x00000004 0000000000000016 jp 0x00000002 0000000000000018 jmp 0x00000008 000000000000001a addsd 0x00000006(%rip),%xmm0 0000000000000022 cvtsd2ss %xmm0,%xmm0 0000000000000026 ret The "jne/jp/jmp" sequence can be reduced to this instead: _floatingPointComparison: 0000000000000000 cvtss2sd %xmm1,%xmm1 0000000000000004 cvtss2sd %xmm0,%xmm0 0000000000000008 mulsd %xmm1,%xmm0 000000000000000c pxor %xmm1,%xmm1 0000000000000010 ucomisd %xmm1,%xmm0 0000000000000014 jp 0x00000002 0000000000000016 je 0x00000008 0000000000000018 addsd 0x00000006(%rip),%xmm0 0000000000000020 cvtsd2ss %xmm0,%xmm0 0000000000000024 ret for a savings of 2 bytes. This xform can happen when we recognize that jne and jp jump to the same "true" MBB, the unconditional jump would jump to the "false" MBB, and the "true" branch is the fall-through MBB. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@97766 91177308-0d34-0410-b5e6-96231b3b80d8		2010-03-05 00:24:26 +00:00
..
Analysis	fix incorrect folding of icmp with undef, PR6481.	2010-03-03 19:46:03 +00:00
Archive	From PR6228:	2010-02-04 06:19:43 +00:00
AsmParser	Introduce isOpaqueTy and use it rather than isa<OpaqueType>. Also, move some	2010-02-16 14:50:09 +00:00
Bitcode	Add Module functions in place of module providers.	2010-03-02 23:58:54 +00:00
CodeGen	Fix some more places where dbg_value affected codegen.	2010-03-05 00:02:59 +00:00
CompilerDriver	Input files with empty suffixes must be passed to linker.	2010-02-23 09:05:21 +00:00
ExecutionEngine	Fix PR6360. It's easy for a stub's address to escape to user code, so we can't	2010-03-04 19:45:09 +00:00
Linker	Introduce isOpaqueTy and use it rather than isa<OpaqueType>. Also, move some	2010-02-16 14:50:09 +00:00
MC	pass in more section kinds, enough to get the .align 0x90	2010-02-26 18:32:26 +00:00
Support	Don't potentially read past the end of the fill data when making a NaN from	2010-03-01 18:38:45 +00:00
System	follow-on to PR6280	2010-02-14 18:20:09 +00:00
Target	Micro-optimization:	2010-03-05 00:24:26 +00:00
Transforms	Make SCEVExpander and LSR more aggressive about hoisting expressions out	2010-03-03 05:29:13 +00:00
VMCore	Stop leaking MDStrings.	2010-03-04 23:24:19 +00:00
Makefile	remove llvm-db: it is completely broken and if anyone wants to do a debugger,	2009-10-05 02:29:51 +00:00