Commit Graph

518 Commits

Author SHA1 Message Date
Anton Korobeynikov
b1a392e7c5 Make sure FABS on v2f32 and v4f32 is legal on ARM NEON
This fixes PR14359


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@168200 91177308-0d34-0410-b5e6-96231b3b80d8
2012-11-16 21:15:20 +00:00
Jakob Stoklund Olesen
17f42e02a1 Revert r163298 "Optimize codegen for VSETLNi{8,16,32} operating on Q registers."
Keep the integer_insertelement test case, the new coalescer can handle
this kind of lane insertion without help from pseudo-instructions.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@166835 91177308-0d34-0410-b5e6-96231b3b80d8
2012-10-26 23:39:46 +00:00
Jim Grosbach
64ba635209 ARM: v1i64 and v2i64 VBSL intrinsic support.
rdar://12502028

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@165981 91177308-0d34-0410-b5e6-96231b3b80d8
2012-10-15 21:23:40 +00:00
Evan Cheng
6b61491de3 Add isel patterns for v2f32 / v4f32 neon.vbsl intrinsics. rdar://12471808
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@165673 91177308-0d34-0410-b5e6-96231b3b80d8
2012-10-10 23:06:34 +00:00
Bob Wilson
eb1641d54a Add LLVM support for Swift.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@164899 91177308-0d34-0410-b5e6-96231b3b80d8
2012-09-29 21:43:49 +00:00
Jim Grosbach
ced674e470 ARM: Use a dedicated intrinsic for vector bitwise select.
The expression based expansion too often results in IR level optimizations
splitting the intermediate values into separate basic blocks, preventing
the formation of the VBSL instruction as the code author intended. In
particular, LICM would often hoist part of the computation out of a loop.

rdar://11011471

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@164340 91177308-0d34-0410-b5e6-96231b3b80d8
2012-09-21 00:18:20 +00:00
Evan Cheng
d10eab0a95 Use vld1 / vst2 for unaligned v2f64 load / store. e.g. Use vld1.16 for 2-byte
aligned address. Based on patch by David Peixotto.

Also use vld1.64 / vst1.64 with 128-bit alignment to take advantage of alignment
hints. rdar://12090772, rdar://12238782


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@164089 91177308-0d34-0410-b5e6-96231b3b80d8
2012-09-18 01:42:45 +00:00
Tim Northover
64eacd9136 Use correct part of complex operand to encode VST1 alignment.
Patch by Chris Lidbury.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@163318 91177308-0d34-0410-b5e6-96231b3b80d8
2012-09-06 14:36:55 +00:00
James Molloy
6c822eea47 Optimize codegen for VSETLNi{8,16,32} operating on Q registers. Degenerate to a VSETLN on D registers, instead of an (INSERT_SUBREG (VSETLN (EXTRACT_SUBREG ))) sequence to help the register coalescer.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@163298 91177308-0d34-0410-b5e6-96231b3b80d8
2012-09-06 09:16:01 +00:00
Evan Cheng
a99c508c8d Use vld1/vst1 to load/store f64 if alignment is < 4 and the target allows unaligned access. rdar://12091029
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@161962 91177308-0d34-0410-b5e6-96231b3b80d8
2012-08-15 17:44:53 +00:00
Tim Northover
e0b464f291 Use correct loads for vector types during extending-load operations.
Previously, we used VLD1.32 in all cases, however there are both 16 and 64-bit
accesses being selected, so we need to use an appropriate width load in those
cases.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@161748 91177308-0d34-0410-b5e6-96231b3b80d8
2012-08-13 09:06:31 +00:00
Joel Jones
7c82e6a32a More replacing of target-dependent intrinsics with target-indepdent
intrinsics.  The second instruction(s) to be handled are the vector versions 
of count set bits (ctpop).

The changes here are to clang so that it generates a target independent 
vector ctpop when it sees an ARM dependent vector bits set count.  The changes 
in llvm are to match the target independent vector ctpop and in 
VMCore/AutoUpgrade.cpp to update any existing bc files containing ARM 
dependent vector pop counts with target-independent ctpops.  There are also 
changes to an existing test case in llvm for ARM vector count instructions and 
to a test for the bitcode upgrade.

<rdar://problem/11892519>

There is deliberately no test for the change to clang, as so far as I know, no
consensus has been reached regarding how to test neon instructions in clang;
q.v. <rdar://problem/8762292>


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@160410 91177308-0d34-0410-b5e6-96231b3b80d8
2012-07-18 00:02:16 +00:00
Joel Jones
06a6a300c5 This is one of the first steps at moving to replace target-dependent
intrinsics with target-indepdent intrinsics.  The first instruction(s) to be 
handled are the vector versions of count leading zeros (ctlz).

The changes here are to clang so that it generates a target independent 
vector ctlz when it sees an ARM dependent vector ctlz.  The changes in llvm 
are to match the target independent vector ctlz and in VMCore/AutoUpgrade.cpp 
to update any existing bc files containing ARM dependent vector ctlzs with 
target-independent ctlzs.  There are also changes to an existing test case in 
llvm for ARM vector count instructions and a new test for the bitcode upgrade.

<rdar://problem/11831778>

There is deliberately no test for the change to clang, as so far as I know, no
consensus has been reached regarding how to test neon instructions in clang;
q.v. <rdar://problem/8762292>


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@160200 91177308-0d34-0410-b5e6-96231b3b80d8
2012-07-13 23:25:25 +00:00
Jim Grosbach
a4fba5eaf8 ARM: Allow more flexible patterns in NEON formats.
Some NEON instructions want to match against normal SDNodes for some
operand types and Intrinsics for others. For example, CTLZ. To enable this,
switch from explicitly requiring Intrinsic on the class templates to using
SDPatternOperator instead.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@159974 91177308-0d34-0410-b5e6-96231b3b80d8
2012-07-10 00:51:13 +00:00
Jim Grosbach
2727930ab4 ARM: Add missing two-operand VBIC aliases.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@156019 91177308-0d34-0410-b5e6-96231b3b80d8
2012-05-02 21:11:56 +00:00
Lang Hames
7787800481 Fix the order of the operands in the llvm.fma intrinsic patterns for ARM,
<rdar://problem/11325085>.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155724 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-27 18:51:24 +00:00
Tim Northover
37abe8df4a Use VLD1 in NEON extenting-load patterns instead of VLDR.
On some cores it's a bad idea for performance to mix VFP and NEON instructions
and since these patterns are NEON anyway, the NEON load should be used.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155630 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-26 08:46:29 +00:00
Jim Grosbach
c92ba4e905 Tidy up. 80 columns, whitespace, et. al.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155399 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-23 22:04:10 +00:00
Jim Grosbach
1d52184cd1 ARM: VSLI two-operand assmebly aliases are tblgen'erated.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155393 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-23 21:22:04 +00:00
Jim Grosbach
e1d866e3c3 ARM: tblgen'erate VSRA/VRSRA/VSRI assembly two-operand aliases.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155392 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-23 21:00:49 +00:00
Jim Grosbach
13b7352fe4 ARM: vqdmulh two-operand aliases are tblgen'erated now.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155387 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-23 20:37:20 +00:00
Jim Grosbach
8e3c17aabf ARM: tblgen'erate more NEON two-operand aliases.
VMUL and VEXT.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155258 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-20 23:46:33 +00:00
Jim Grosbach
d83c9ea7d1 ARM: tblgen'erate more NEON two-operand aliases.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155254 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-20 23:30:14 +00:00
Jim Grosbach
d8b3ed8f25 ARM: Update NEON assembly two-operand aliases.
Use the new TwoOperandAliasConstraint to handle lots of the two-operand aliases
for NEON instructions. There's still more to go, but this is a good chunk of
them.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155210 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-20 18:12:54 +00:00
James Molloy
72aadc057c Fix bad EXTRACT_SUBREG in instruction selection for extending-loads on NEON.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@154915 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-17 08:18:00 +00:00
Jim Grosbach
bf42f24e6e ARM two-operand forms for vhadd and vhsub instructions.
rdar://11252521

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@154875 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-16 23:00:25 +00:00
Jim Grosbach
199366a6a6 ARM assembly two-operand forms for VRSHL.
rdar://11252521

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@154840 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-16 18:03:16 +00:00
Jim Grosbach
705e2572b4 ARM two-operand aliases for VRHADD instructions.
rdar://11252521

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@154832 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-16 17:14:11 +00:00
Jim Grosbach
1835547ec1 ARM 'vuzp.32 Dd, Dm' is a pseudo-instruction.
While there is an encoding for it in VUZP, the result of that is undefined,
so we should avoid it. Define the instruction as a pseudo for VTRN.32
instead, as the ARM ARM indicates.

rdar://11222366

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@154511 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-11 17:40:18 +00:00
Jim Grosbach
6073b30b05 ARM 'vzip.32 Dd, Dm' is a pseudo-instruction.
While there is an encoding for it in VZIP, the result of that is undefined,
so we should avoid it. Define the instruction as a pseudo for VTRN.32
instead, as the ARM ARM indicates.

rdar://11221911

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@154505 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-11 16:53:25 +00:00
Evan Cheng
14b4c03580 Add more fused mul+add/sub patterns. rdar://10139676
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@154484 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-11 06:59:47 +00:00
Evan Cheng
bee78fe5fc Clean up ARM fused multiply + add/sub support some more: rename some isel
predicates.
Also remove NEON2 since it's not really useful and it is confusing. If
NEON + VFP4 implies NEON2 but NEON2 doesn't imply NEON + VFP4, what does it
really mean?

rdar://10139676


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@154480 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-11 05:33:07 +00:00
Evan Cheng
82509e5c62 Fix a number of problems with ARM fused multiply add/subtract instructions.
1. The new instruction itinerary entries are not properly described.
2. The asm parser can't handle vfms and vfnms.
3. There were no assembler, disassembler test cases.
4. HasNEON2 has the wrong assembler predicate.
rdar://10139676


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@154456 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-11 00:13:00 +00:00
Evan Cheng
3aef2ff514 Handle llvm.fma.* intrinsics. rdar://10914096
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@154439 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-10 21:40:28 +00:00
Jim Grosbach
22378fd664 ARM assembly aliases for two-operand V[R]SHR instructions.
rdar://11189467

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@154087 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-05 07:23:53 +00:00
Jim Grosbach
a45e3747e6 ARM encoding for VSWP got the second operand incorrect.
Make the non-tied register operand names line up with what the base
class encoding handler expects.

rdar://11157236

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@153766 91177308-0d34-0410-b5e6-96231b3b80d8
2012-03-30 18:53:01 +00:00
Jakob Stoklund Olesen
5b2f913664 Spill DPair registers, not just QPR.
The arm_neon intrinsics can create virtual registers from the DPair
register class which allows both even-odd and odd-even D-register pairs.

This fixes PR12389.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@153603 91177308-0d34-0410-b5e6-96231b3b80d8
2012-03-28 21:20:32 +00:00
Richard Barton
6e9d66c756 Fixup VST1.32 with writeback instruction. Also re-factor non-writeback version.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@153573 91177308-0d34-0410-b5e6-96231b3b80d8
2012-03-28 10:18:11 +00:00
Jim Grosbach
4d0983a4d7 ARM more NEON VLD/VST composite physical register refactoring.
Register pair, all lanes subscripting.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@152157 91177308-0d34-0410-b5e6-96231b3b80d8
2012-03-06 23:10:38 +00:00
Jim Grosbach
c0fc450f07 ARM refactor more NEON VLD/VST instructions to use composite physregs
Register pair VLD1/VLD2 all-lanes instructions. Kill off more of the
pseudos as a result.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@152150 91177308-0d34-0410-b5e6-96231b3b80d8
2012-03-06 22:01:44 +00:00
Jim Grosbach
c3384c93c0 ARM Refactor VLD/VST spaced pair instructions.
Use the new composite physical registers.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@152063 91177308-0d34-0410-b5e6-96231b3b80d8
2012-03-05 21:43:40 +00:00
Jim Grosbach
af9f4bc752 ARM Remove a bit of dead code.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@152061 91177308-0d34-0410-b5e6-96231b3b80d8
2012-03-05 21:09:58 +00:00
Jim Grosbach
28f08c93e7 ARM refactor away a bunch of VLD/VST pseudo instructions.
With the new composite physical registers to represent arbitrary pairs
of DPR registers, we don't need the pseudo-registers anymore. Get rid of
a bunch of them that use DPR register pairs and just use the real
instructions directly instead.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@152045 91177308-0d34-0410-b5e6-96231b3b80d8
2012-03-05 19:33:30 +00:00
Sebastian Pop
74bebde7c4 updated patch for the ARM fused multiply add/sub
In this update:
- I assumed neon2 does not imply vfpv4, but neon and vfpv4 imply neon2.
- I kept setting .fpu=neon-vfpv4 code attribute because that is what the
assembler understands.

Patch by Ana Pazos <apazos@codeaurora.org>

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@152036 91177308-0d34-0410-b5e6-96231b3b80d8
2012-03-05 17:39:52 +00:00
Jim Grosbach
c94206e552 ARM vbit/vbif/vbsl assembly optional size suffix.
These instructions accept but do not require a size suffix.

rdar://10947225

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@151646 91177308-0d34-0410-b5e6-96231b3b80d8
2012-02-28 19:11:07 +00:00
James Molloy
873fd5f753 Improve generated code for extending loads and some trunc stores on ARM.
Teach TargetSelectionDAG about lengthening loads for vector types and set v4i8 as legal. Allow FP_TO_UINT for v4i16 from v4i32.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@150956 91177308-0d34-0410-b5e6-96231b3b80d8
2012-02-20 09:24:05 +00:00
Jia Liu
31d157ae1a Emacs-tag and some comment fix for all ARM, CellSPU, Hexagon, MBlaze, MSP430, PPC, PTX, Sparc, X86, XCore.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@150878 91177308-0d34-0410-b5e6-96231b3b80d8
2012-02-18 12:03:15 +00:00
Lang Hames
1a4cb1caf0 Third time's the charm...?
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@150447 91177308-0d34-0410-b5e6-96231b3b80d8
2012-02-14 00:34:30 +00:00
Lang Hames
6b0e9d97fa Unswap swap operands, partially reducing confusion.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@150444 91177308-0d34-0410-b5e6-96231b3b80d8
2012-02-14 00:17:12 +00:00
Lang Hames
2cc494b963 Make operands for VSWP read-modify-write.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@150433 91177308-0d34-0410-b5e6-96231b3b80d8
2012-02-13 23:37:19 +00:00