Bob Wilson
fed76ffa56
Add missing address register update to t2LDM_RET instruction.
...
Patch by Brian Lucas. PR7636.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108332 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-14 16:02:13 +00:00
Eli Friedman
54cc0e12da
A couple potential optimizations inspired by comment 4 in PR6773.
...
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108328 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-14 06:58:26 +00:00
Evan Cheng
dedd974e7e
Fix for PR7193 was overly conservative. The only case where sibcall callee
...
address cannot be allocated a register is in 32-bit mode where the first
three arguments are marked inreg. In that case EAX, EDX, and ECX will be
used for argument passing.
This fixes PR7610.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108327 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-14 06:44:01 +00:00
Bob Wilson
7e3f0d2690
Add support for NEON VMVN immediate instructions.
...
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108324 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-14 06:31:50 +00:00
Bob Wilson
046afdb50b
The bits in the cmode field of 32-bit VMOV immediate instructions all depend
...
of the value of the immediate.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108323 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-14 06:30:44 +00:00
Chris Lattner
b09a97e565
fix a bug found by a warning I added to clang this morning.
...
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108309 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-14 01:57:17 +00:00
Bob Wilson
9e82bf12a0
Add an ARM-specific DAG combining to avoid redundant VDUPLANE nodes.
...
Radar 7373643.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108303 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-14 01:22:12 +00:00
Dan Gohman
a10b8494a5
Don't propagate debug locations to instructions for materializing
...
constants, since they may not be emited near the other instructions
which get the same line, and this confuses debug info.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108302 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-14 01:07:44 +00:00
Bruno Cardoso Lopes
7dbf7d8b1c
Add AVX 256-bit compare instructions and a bunch of testcases
...
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108286 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-13 22:06:38 +00:00
Bob Wilson
cba270d042
Use a target-specific VMOVIMM DAG node instead of BUILD_VECTOR to represent
...
NEON VMOV-immediate instructions. This simplifies some things.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108275 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-13 21:16:48 +00:00
Bruno Cardoso Lopes
87a85c7ef0
AVX 256-bit conversion instructions
...
Add the x86 VEX_L form to handle special cases where VEX_L must be set.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108274 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-13 21:07:28 +00:00
Kevin Enderby
52a18aedae
Added a check that pusha cannot be encoded in 64-bit mode.
...
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108265 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-13 20:05:41 +00:00
Evan Cheng
218977b53e
Extend the r107852 optimization which turns some fp compare to code sequence using only i32 operations. It now optimize some f64 compares when fp compare is exceptionally slow (e.g. cortex-a8). It also catches comparison against 0.0.
...
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108258 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-13 19:27:42 +00:00
Evan Cheng
7a41599962
Add an ARM "feature". Cortex-a8 fp comparison is very slow (> 20 cycles).
...
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108256 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-13 19:21:50 +00:00
Evan Cheng
ea4cdb7ead
-enable-unsafe-fp-math should not imply -enable-finite-only-fp-math.
...
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108254 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-13 18:46:14 +00:00
Gabor Greif
63d024fc9a
rotate CallInst operands
...
with this commit the callee moves to the end of
the operand array (from the start) and the call
arguments now start at index 0 (formerly 1)
this ordering is now consistent with InvokeInst
this commit only flips the switch,
functionally it is equivalent to
r101465
I intend to commit several cleanups after a few
days of soak period
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108240 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-13 15:31:36 +00:00
Bob Wilson
6dce00ced4
Move NEON "modified immediate" encode/decode into ARMAddressingModes.h to
...
avoid replicated code.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108227 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-13 04:44:34 +00:00
Chris Lattner
37a746bc85
my work on adding segment registers to LEA missed the
...
disassembler. Remove some code from the disassembler to
compensate, unbreaking disassembly of lea's.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108226 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-13 04:23:55 +00:00
Bruno Cardoso Lopes
fd920fa59a
Add AVX 256-bit packed logical forms
...
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108224 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-13 02:38:35 +00:00
Bruno Cardoso Lopes
6991623dd7
Add AVX 256-bit unop arithmetic instructions
...
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108223 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-13 01:53:31 +00:00
Bruno Cardoso Lopes
4344d85769
Since AVX is a superset of all SSE versions, only use HasAVX for AVX instructions
...
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108222 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-13 00:38:47 +00:00
David Greene
8f17bc4fbd
Move some SIMD fragment code into X86InstrFragmentsSIMD so that the
...
utility classes can be used from multiple files. This will aid
transitioning to a new refactored x86 SIMD specification.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108213 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-12 23:41:28 +00:00
Bruno Cardoso Lopes
a0d09a85e2
Add AVX 256 binary arithmetic instructions
...
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108207 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-12 23:04:15 +00:00
Bruno Cardoso Lopes
f428fee70d
More refactoring of basic SSE arith instructions. Open room for 256-bit instructions
...
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108204 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-12 22:41:32 +00:00
Dan Gohman
cfbf0ed8b0
Apply the SSE dependence idiom for SSE unary operations to
...
SD instructions too, in addition to SS instructions. And
add a comment about it.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108191 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-12 20:46:04 +00:00
Bob Wilson
c7a797b82b
Remove some code that doesn't appear to do anything. All the ARM call
...
instructions already have implicit defs of LR. The comment suggests that
this is intended to fix something like pr6111, but it doesn't really do
that either.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108186 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-12 20:22:45 +00:00
Bruno Cardoso Lopes
aa099be71f
Add AVX 256-bit MOVMSK forms
...
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108184 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-12 20:06:32 +00:00
Dan Gohman
ed42f1e58f
Check begin!=end, rather than !begin.
...
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108167 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-12 18:12:35 +00:00
Dan Gohman
be4d10d7fa
Don't fast-isel an x87 comparison opcode, as fast-isel doesn't
...
support branching on x87 comparisons yet. This fixes PR7624.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108149 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-12 15:46:30 +00:00
Duncan Sands
3472766f9e
Convert some tab stops into spaces.
...
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108130 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-12 08:16:59 +00:00
Rafael Espindola
5a717a3ae7
Convert getLoadStoreRegOpcode to use a switch.
...
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108123 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-12 03:43:04 +00:00
Rafael Espindola
7e1b566322
Convert the last use of getPhysicalRegisterRegClass and remove it.
...
AggressiveAntiDepBreaker should not be using getPhysicalRegisterRegClass. An
instruction might be using a register that can only be replaced with one from
a subclass of getPhysicalRegisterRegClass.
With this patch we use getMinimalPhysRegClass. This is correct, but
conservative. We should check the uses of the register and select the
largest register class that can be used in all of them.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108122 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-12 02:55:34 +00:00
Jakob Stoklund Olesen
8b78d4b5bb
A basic block that only uses RFP registers still needs the FP_REG_KILL marker.
...
This fixes PR7375.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108120 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-12 02:12:47 +00:00
Rafael Espindola
0bfd09201e
Convert the last getPhysicalRegisterRegClass in VirtRegRewriter.cpp to
...
getMinimalPhysRegClass. It was used to produce spills, and it is better to
use the most specific class if possible.
Update getLoadStoreRegOpcode to handle GR32_AD.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108115 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-12 00:52:33 +00:00
Jakob Stoklund Olesen
600f171486
RISC architectures get their memory operand folding for free.
...
The only folding these load/store architectures can do is converting COPY into a
load or store, and the target independent part of foldMemoryOperand already
knows how to do that.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108099 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-11 19:19:13 +00:00
Jakob Stoklund Olesen
a66450d227
Use target independent COPY instructions for the fake fextend and fround
...
operations in x87 code.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108098 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-11 18:19:39 +00:00
Jakob Stoklund Olesen
4cae5af54e
Remove redundant branch. Thanks, Anton!
...
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108097 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-11 17:17:35 +00:00
Jakob Stoklund Olesen
aef48d7b36
Remove obsolete README_SSE note.
...
We are generating movaps for all XMM register copies, including scalar
floating point values. This is known to be at least as good as movss and movsd
for all known architectures up to and including Nehalem because it avoids a
partial register stall.
The SSEDomainFix pass will switch movaps to movdqa when appropriate (i.e., when
operands come from the integer unit). We don't now that switching movaps to
movapd has any benefit.
The same applies to andps -> pand.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108096 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-11 17:13:42 +00:00
Rafael Espindola
d6d7abaf4e
Make getPhysicalRegisterRegClass non-virtual. Should be able to remove it soon.
...
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108094 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-11 16:49:10 +00:00
Jakob Stoklund Olesen
f7d55b97f0
Replace copyRegToReg with copyPhysReg for SystemZ.
...
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108092 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-11 16:40:46 +00:00
Jakob Stoklund Olesen
75be45cb2e
Avoid SSE instructions in FastIsel when it is not available.
...
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108091 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-11 16:22:13 +00:00
Chandler Carruth
c5080ba0c8
Remove two other uses of ATTRIBUTE_UNUSED for variables only used within
...
assert()s, switching to void-casts. Removed an unneeded Compiler.h include as
a result. There are two other uses in LLVM, but they're not due to assert()s,
so I've left them alone.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108088 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-11 08:18:12 +00:00
Jakob Stoklund Olesen
a98625cdad
Replace copyRegToReg with copyPhysReg for XCore.
...
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108087 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-11 07:56:13 +00:00
Jakob Stoklund Olesen
8e18a1a5cf
Replace copyRegToReg with copyPhysReg for Sparc.
...
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108086 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-11 07:56:09 +00:00
Jakob Stoklund Olesen
377b7b7ca3
Replace copyRegToReg with copyPhysReg for CellSPU.
...
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108084 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-11 07:31:03 +00:00
Jakob Stoklund Olesen
27689b0aff
Replace copyRegToReg with copyPhysReg for PowerPC.
...
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108083 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-11 07:31:00 +00:00
Jakob Stoklund Olesen
26a99d17de
Fix PIC16 comments referencing copyRegToReg.
...
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108082 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-11 07:30:57 +00:00
Jakob Stoklund Olesen
d86adf3d6e
Replace copyRegToReg with copyPhysReg for PIC16.
...
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108081 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-11 06:53:33 +00:00
Jakob Stoklund Olesen
41ce3cfd1b
Replace copyRegToReg with copyPhysReg for MSP430.
...
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108080 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-11 06:53:30 +00:00
Jakob Stoklund Olesen
e6afcf8da2
Replace copyRegToReg with copyPhysReg for MBlaze.
...
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108079 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-11 06:53:27 +00:00