Go to file
Tim Northover 85d6a16c46 ARM64: use regalloc-friendly COPY_TO_REGCLASS for bitcasts
The previous patterns directly inserted FMOV or INS instructions into
the DAG for scalar_to_vector & bitconvert patterns. This is horribly
inefficient and can generated lots more GPR <-> FPR register traffic
than necessary.

It's much better to emit instructions the register allocator
understands so it can coalesce the copies when appropriate.

It led to at least one ISelLowering hack to avoid the problems, which
was incorrect for v1i64 (FPR64 has no dsub). It can now be removed
entirely.

This should also fix PR19331.

llvm-svn: 205616
2014-04-04 09:03:09 +00:00
clang clang-format: Don't merge simple blocks in case statements. 2014-04-04 06:46:23 +00:00
clang-tools-extra Fix regex bug in clang-tidy. 2014-04-02 08:52:06 +00:00
compiler-rt Revert r205613. 2014-04-04 08:58:16 +00:00
debuginfo-tests relax testcase for LLDB output format compatibility. 2014-03-19 23:06:18 +00:00
libclc Introduce M_LOG2E_F and M_LOG2E 2014-03-28 21:19:03 +00:00
libcxx RTTI Uniqueness: remove __name_for_load function. 2014-04-03 09:12:38 +00:00
libcxxabi Include stdlib.h for getenv when !NDEBUG. 2014-04-03 22:00:08 +00:00
lld SymbolTable::size() returns an unsigned int. 2014-04-04 01:22:51 +00:00
lldb sweep up -Wformat warnings from gcc 2014-04-04 04:06:10 +00:00
llvm ARM64: use regalloc-friendly COPY_TO_REGCLASS for bitcasts 2014-04-04 09:03:09 +00:00
openmp Revert commit (testing commit hook). 2014-03-04 18:06:04 +00:00
polly Add another hint for fixing check-polly errors to get_started.html 2014-04-03 19:38:38 +00:00