Commit Graph

6021 Commits

Author SHA1 Message Date
Craig Topper
28692044db Add AVX2 VPMOVMASK instructions and intrinsics.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@143904 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-07 03:20:35 +00:00
Craig Topper
69f5df7778 Add AVX2 VEXTRACTI128 and VINSERTI128 instructions. Fix VPERM2I128 to be qualified with HasAVX2 instead of HasAVX. Mark VINSERTF128 and VEXTRACTF128 as never having side effects.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@143902 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-07 02:00:04 +00:00
Craig Topper
c8eb880a7f More AVX2 instructions and their intrinsics.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@143895 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-06 23:04:08 +00:00
Craig Topper
27e5d0c72a Add more AVX2 instructions and intrinsics.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@143861 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-06 06:12:20 +00:00
Chad Rosier
42536af5ce Add support for passing i1, i8, and i16 call parameters. Also, be sure to
zero-extend the constant integer encoding.  Test case provides testing for
both call parameters and materialization of i1, i8, and i16 types.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@143821 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-05 20:16:15 +00:00
Benjamin Kramer
c25c908977 Add an option to pad an uleb128 to MCObjectWriter and remove the uleb128 encoding from the DWARF asm printer.
As a side effect we now print dwarf ulebs with .ascii directives.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@143809 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-05 11:52:44 +00:00
Eli Friedman
bd00a934c6 Enhanced vzeroupper insertion pass that avoids inserting vzeroupper where it is unnecessary through local analysis. Patch from Bruno Cardoso Lopes, with some additional changes.
I'm going to wait for any review comments and perform some additional testing before turning this on by default.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@143750 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-04 23:46:11 +00:00
Craig Topper
517497cce0 Add intrinsics for X86 vcvtps2ph and vcvtph2ps instructions
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@143682 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-04 06:59:21 +00:00
Chad Rosier
f470cbbad2 Add fast-isel support for returning i1, i8, and i16.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@143669 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-04 00:50:21 +00:00
Dan Gohman
65fd6564b8 Reapply r143206, with fixes. Disallow physical register lifetimes
across calls, and only check for nested dependences on the special
call-sequence-resource register.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@143660 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-03 21:49:52 +00:00
Pete Cooper
71fccadbed Reverted r143600 - selector reference change
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@143646 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-03 20:47:50 +00:00
Dan Bailey
e1f38f2ce1 fixed global array handling for ptx to use the correct bit widths
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@143640 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-03 19:24:46 +00:00
Craig Topper
98e0b9c86d Add new X86 AVX2 VBROADCAST instructions.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@143612 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-03 07:35:53 +00:00
Chad Rosier
463fe24f1d Add support for sign-extending non-legal types in SelectSIToFP().
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@143603 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-03 02:04:59 +00:00
Pete Cooper
d1ffc739c1 Treat objc selector reference globals as invariant so that MachineLICM can hoist them out of loops. Fixes <rdar://problem/6027699>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@143600 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-03 00:56:36 +00:00
Lang Hames
1a1d1fcc0b Try to lower memset/memcpy/memmove to vector instructions on ARM where the alignment permits.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@143582 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-02 22:52:45 +00:00
Nick Lewycky
6c1a703e54 Don't emit a directory entry for the value in DW_AT_comp_dir, that is always
implied by directory index zero.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@143570 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-02 20:55:33 +00:00
Chad Rosier
e07cd5e40a Add support for comparing integer non-legal types.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@143559 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-02 18:08:25 +00:00
Craig Topper
205e3378fd More AVX2 instructions and intrinsics.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@143536 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-02 06:54:17 +00:00
Craig Topper
3f2b2c218f Add a bunch more X86 AVX2 instructions and their corresponding intrinsics.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@143529 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-02 04:42:13 +00:00
Eli Friedman
f6aa6b12f1 Teach the x86 backend a couple tricks for dealing with v16i8 sra by a constant splat value. Fixes PR11289.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@143498 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-01 21:18:39 +00:00
Richard Osborne
0e6c1c536b Don't fold negative offsets into cp / dp accesses to avoid relocation errors.
This can happen if the address + addend is less than the start of the cp / dp.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@143459 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-01 11:31:53 +00:00
Richard Osborne
441ed4ac23 Combine various XCore tests for floating point intrinsic support into a single test.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@143458 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-01 10:51:48 +00:00
Richard Osborne
70ad3951e4 Move various XCore tests to FileCheck
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@143457 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-01 10:41:28 +00:00
Craig Topper
ce7de9f36d Fix operand type for x86 pmadd_ub_sw intrinsic.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@143455 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-01 07:25:22 +00:00
Craig Topper
782c8fbd6e Fix operand type for int_x86_ssse3_phadd_sw_128 intrinsic
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@143336 91177308-0d34-0410-b5e6-96231b3b80d8
2011-10-31 07:16:37 +00:00
Craig Topper
593c1d9761 Test case for X86 FS/GS Base intrinsics
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@143332 91177308-0d34-0410-b5e6-96231b3b80d8
2011-10-31 02:15:47 +00:00
Craig Topper
6b1c5fc02a Begin adding AVX2 instructions. No selection support yet other than intrinsics.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@143331 91177308-0d34-0410-b5e6-96231b3b80d8
2011-10-31 02:15:10 +00:00
Nick Lewycky
4e478fed1b Switch new .file directive emission off by default, change llc's flag for it to
-enable-dwarf-directory.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@143326 91177308-0d34-0410-b5e6-96231b3b80d8
2011-10-31 01:06:02 +00:00
Benjamin Kramer
dade3c1448 X86: Emit logical shift by constant splat of <16 x i8> as a <8 x i16> shift and zero out the bits where zeros should've been shifted in.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@143315 91177308-0d34-0410-b5e6-96231b3b80d8
2011-10-30 17:31:21 +00:00
Craig Topper
6762427e8e Fix return type for X86 mpsadbw instrinsic. The instruction takes in a vector of 8-bit integers, but produces a vector of 16-bit integers.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@143313 91177308-0d34-0410-b5e6-96231b3b80d8
2011-10-30 17:22:45 +00:00
Nadav Rotem
fb0dfbbff7 Fix pr11266.
On x86: (shl V, 1) -> add V,V

Hardware support for vector-shift is sparse and in many cases we scalarize the
result. Additionally, on sandybridge padd is faster than shl.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@143311 91177308-0d34-0410-b5e6-96231b3b80d8
2011-10-30 13:24:22 +00:00
Nadav Rotem
5157588840 Stabilize the test by specifying an exact cpu target
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@143307 91177308-0d34-0410-b5e6-96231b3b80d8
2011-10-30 08:07:50 +00:00
Nadav Rotem
b00418af67 Add a new DAGCombine optimization for BUILD_VECTOR.
If all of the inputs are zero/any_extended, create a new simple BV
which can be further optimized by other BV optimizations.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@143297 91177308-0d34-0410-b5e6-96231b3b80d8
2011-10-29 21:23:04 +00:00
Benjamin Kramer
f86545ecfd Force SSE for this test.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@143291 91177308-0d34-0410-b5e6-96231b3b80d8
2011-10-29 19:43:44 +00:00
Dan Gohman
6f3ddef7c5 Revert r143206, as there are still some failing tests.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@143262 91177308-0d34-0410-b5e6-96231b3b80d8
2011-10-29 00:41:52 +00:00
NAKAMURA Takumi
29ceb7c104 test/CodeGen/PowerPC/2008-10-17-AsmMatchingOperands.ll: [PR11218] Mark "REQUIRES: asserts" for now.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@143247 91177308-0d34-0410-b5e6-96231b3b80d8
2011-10-28 23:11:03 +00:00
Dan Gohman
bf923b815d Reapply r143177 and r143179 (reverting r143188), with scheduler
fixes: Use a separate register, instead of SP, as the
calling-convention resource, to avoid spurious conflicts with
actual uses of SP. Also, fix unscheduling of calling sequences,
which can be triggered by pseudo-two-address dependencies.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@143206 91177308-0d34-0410-b5e6-96231b3b80d8
2011-10-28 17:55:38 +00:00
NAKAMURA Takumi
c3e48c38bf Dwarf: [PR11022] Fix emitting DW_AT_const_value(>i64), to be host-endian-neutral.
Don't assume APInt::getRawData() would hold target-aware endianness nor host-compliant endianness. rawdata[0] holds most lower i64, even on big endian host.

FIXME: Add a testcase for big endian target.

FIXME: Ditto on CompileUnit::addConstantFPValue() ?

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@143194 91177308-0d34-0410-b5e6-96231b3b80d8
2011-10-28 14:12:22 +00:00
NAKAMURA Takumi
5c56f0b589 test/CodeGen/X86/2010-08-10-DbgConstant.ll: Add explicit -mtriple=i686-linux. It must be for elf!
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@143189 91177308-0d34-0410-b5e6-96231b3b80d8
2011-10-28 10:50:52 +00:00
Duncan Sands
62c1d00dfd Speculatively disable Dan's commits 143177 and 143179 to see if
it fixes the dragonegg self-host (it looks like gcc is miscompiled).
Original commit messages:
Eliminate LegalizeOps' LegalizedNodes map and have it just call RAUW
on every node as it legalizes them. This makes it easier to use
hasOneUse() heuristics, since unneeded nodes can be removed from the
DAG earlier.

Make LegalizeOps visit the DAG in an operands-last order. It previously
used operands-first, because LegalizeTypes has to go operands-first, and
LegalizeTypes used to be part of LegalizeOps, but they're now split.
The operands-last order is more natural for several legalization tasks.
For example, it allows lowering code for nodes with floating-point or
vector constants to see those constants directly instead of seeing the
lowered form (often constant-pool loads). This makes some things
somewhat more complicated today, though it ought to allow things to be
simpler in the future. It also fixes some bugs exposed by Legalizing
using RAUW aggressively.

Remove the part of LegalizeOps that attempted to patch up invalid chain
operands on libcalls generated by LegalizeTypes, since it doesn't work
with the new LegalizeOps traversal order. Instead, define what
LegalizeTypes is doing to be correct, and transfer the responsibility
of keeping calls from having overlapping calling sequences into the
scheduler.

Teach the scheduler to model callseq_begin/end pairs as having a
physical register definition/use to prevent calls from having
overlapping calling sequences. This is also somewhat complicated, though
there are ways it might be simplified in the future.

This addresses rdar://9816668, rdar://10043614, rdar://8434668, and others.
Please direct high-level questions about this patch to management.

Delete #if 0 code accidentally left in.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@143188 91177308-0d34-0410-b5e6-96231b3b80d8
2011-10-28 09:55:57 +00:00
Nick Lewycky
6a7efcfc02 Always use the string pool, even when it makes the .o larger. This may help
tools that read the debug info in the .o files by making the DIE sizes more
consistent.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@143186 91177308-0d34-0410-b5e6-96231b3b80d8
2011-10-28 05:29:47 +00:00
Dan Gohman
2ba60e5930 Eliminate LegalizeOps' LegalizedNodes map and have it just call RAUW
on every node as it legalizes them. This makes it easier to use
hasOneUse() heuristics, since unneeded nodes can be removed from the
DAG earlier.

Make LegalizeOps visit the DAG in an operands-last order. It previously
used operands-first, because LegalizeTypes has to go operands-first, and
LegalizeTypes used to be part of LegalizeOps, but they're now split.
The operands-last order is more natural for several legalization tasks.
For example, it allows lowering code for nodes with floating-point or
vector constants to see those constants directly instead of seeing the
lowered form (often constant-pool loads). This makes some things
somewhat more complicated today, though it ought to allow things to be
simpler in the future. It also fixes some bugs exposed by Legalizing
using RAUW aggressively.

Remove the part of LegalizeOps that attempted to patch up invalid chain
operands on libcalls generated by LegalizeTypes, since it doesn't work
with the new LegalizeOps traversal order. Instead, define what
LegalizeTypes is doing to be correct, and transfer the responsibility
of keeping calls from having overlapping calling sequences into the
scheduler.

Teach the scheduler to model callseq_begin/end pairs as having a
physical register definition/use to prevent calls from having
overlapping calling sequences. This is also somewhat complicated, though
there are ways it might be simplified in the future.

This addresses rdar://9816668, rdar://10043614, rdar://8434668, and others.
Please direct high-level questions about this patch to management.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@143177 91177308-0d34-0410-b5e6-96231b3b80d8
2011-10-28 01:29:32 +00:00
Dan Gohman
33ba8b0e96 Remove the Alpha backend.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@143164 91177308-0d34-0410-b5e6-96231b3b80d8
2011-10-27 22:56:32 +00:00
Jakob Stoklund Olesen
b0117eed84 Also set addrmode6 alignment when align==size.
Previously, we were only setting the alignment bits on over-aligned
loads and stores.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@143160 91177308-0d34-0410-b5e6-96231b3b80d8
2011-10-27 22:39:16 +00:00
Evan Cheng
824a70a384 Avoid partial CPSR dependency from loop backedges. rdar://10357570
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@143145 91177308-0d34-0410-b5e6-96231b3b80d8
2011-10-27 21:21:05 +00:00
Pete Cooper
cbe35f2147 Changed test to check for correct load size instead of shift as the shift might change if optimised
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@143116 91177308-0d34-0410-b5e6-96231b3b80d8
2011-10-27 18:15:58 +00:00
Nick Lewycky
390c40d96a Teach our Dwarf emission to use the string pool.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@143097 91177308-0d34-0410-b5e6-96231b3b80d8
2011-10-27 06:44:11 +00:00
Eli Friedman
fd58cd7563 Don't crash on 128-bit sdiv by constant. Found by inspection.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@143095 91177308-0d34-0410-b5e6-96231b3b80d8
2011-10-27 02:06:39 +00:00
Chad Rosier
6d64b3adab A branch predicated on a constant can just FastEmit an unconditional branch.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@143086 91177308-0d34-0410-b5e6-96231b3b80d8
2011-10-27 00:21:16 +00:00