Cong Hou 20b64d0452 Allow X86::COND_NE_OR_P and X86::COND_NP_OR_E to be reversed. ... Currently, AnalyzeBranch() fails non-equality comparison between floating points on X86 (see https://llvm.org/bugs/show_bug.cgi?id=23875). This is because this function can modify the branch by reversing the conditional jump and removing unconditional jump if there is a proper fall-through. However, in the case of non-equality comparison between floating points, this can turn the branch "unanalyzable". Consider the following case: jne.BB1 jp.BB1 jmp.BB2 .BB1: ... .BB2: ... AnalyzeBranch() will reverse "jp .BB1" to "jnp .BB2" and then "jmp .BB2" will be removed: jne.BB1 jnp.BB2 .BB1: ... .BB2: ... However, AnalyzeBranch() cannot analyze this branch anymore as there are two conditional jumps with different targets. This may disable some optimizations like block-placement: in this case the fall-through behavior is enforced even if the fall-through block is very cold, which is suboptimal. Actually this optimization is also done in block-placement pass, which means we can remove this optimization from AnalyzeBranch(). However, currently X86::COND_NE_OR_P and X86::COND_NP_OR_E are not reversible: there is no defined negation conditions for them. In order to reverse them, this patch defines two new CondCode X86::COND_E_AND_NP and X86::COND_P_AND_NE. It also defines how to synthesize instructions for them. Here only the second conditional jump is reversed. This is valid as we only need them to do this "unconditional jump removal" optimization. Differential Revision: http://reviews.llvm.org/D11393 llvm-svn: 258847		2016-01-26 20:08:01 +00:00
..
Analysis	[LIR] Add support for structs and hand unrolled loops	2016-01-26 02:27:47 +00:00
AsmParser	Implemented Support of IA interrupt and exception handlers:	2015-12-21 14:07:14 +00:00
Bitcode	Fix Clang-tidy modernize-use-nullptr and modernize-use-override warnings; other minor fixes.	2016-01-26 18:48:36 +00:00
CodeGen	[ScheduleDAGInstrs] Simplify logic to improve readability. NFC.	2016-01-26 19:33:57 +00:00
DebugInfo	Re-submit r256008 "Improve DWARFDebugFrame::parse to also handle __eh_frame."	2016-01-26 15:09:42 +00:00
ExecutionEngine	Reflect the MC/MCDisassembler split on the include/ level.	2016-01-26 16:44:37 +00:00
Fuzzer	Temporary disable broken fuzzer/timeout tests.	2016-01-25 19:05:45 +00:00
IR	Use binary search for intrinsic ID lookups	2016-01-26 02:06:41 +00:00
IRReader	[ThinLTO] Metadata linking for imported functions	2015-12-17 17:14:09 +00:00
LibDriver	[Option] Use an ArrayRef to store the Option Infos in OptTable. NFC	2015-10-21 16:30:42 +00:00
LineEditor
Linker	[ThinLTO] Find all needed metadata when linking metadata as postpass	2016-01-25 22:04:56 +00:00
LTO	[LTO] Fix error reporting when a file passed to libLTO is invalid or non-existent	2016-01-20 09:03:42 +00:00
MC	Reflect the MC/MCDisassembler split on the include/ level.	2016-01-26 16:44:37 +00:00
Object	[Object][COFF] Revert r258665 - It doesn't do what I had intended.	2016-01-25 01:21:45 +00:00
Option	Convert Arg, ArgList, and Option to dump() to dbgs() rather than errs().	2015-12-18 18:55:26 +00:00
Passes	[attrs] Extract the pure inference of function attributes into	2015-12-27 08:41:34 +00:00
ProfileData	Fix Clang-tidy modernize-use-nullptr and modernize-use-override warnings; other minor fixes.	2016-01-26 18:48:36 +00:00
Support	Fix Clang-tidy modernize-use-nullptr and modernize-use-override warnings; other minor fixes.	2016-01-26 18:48:36 +00:00
TableGen	[TableGen] Use FoldingSets instead of DenseMaps to unique UnOpInit, BinOpInit and TernOpInit. This remove the memory needed to store the key for the DenseMap. NFC	2016-01-18 20:36:06 +00:00
Target	Allow X86::COND_NE_OR_P and X86::COND_NP_OR_E to be reversed.	2016-01-26 20:08:01 +00:00
Transforms	Fix Clang-tidy modernize-use-nullptr and modernize-use-override warnings; other minor fixes.	2016-01-26 18:48:36 +00:00
CMakeLists.txt
LLVMBuild.txt
Makefile