Commit Graph

60859 Commits

Author SHA1 Message Date
Evan Cheng
15a16def6e Add a hybrid bottom up scheduler that reduce register usage while avoiding
pipeline stall. It's useful for targets like ARM cortex-a8. NEON has a lot
of long latency instructions so a strict register pressure reduction
scheduler does not work well.
Early experiments show this speeds up some NEON loops by over 30%.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104216 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-20 06:13:19 +00:00
Nick Lewycky
761fd4c1d9 Fix typo in comment.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104209 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-20 03:30:09 +00:00
Dan Gohman
14aaeac5cf Define the x86 pause instruction.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104204 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-20 01:35:50 +00:00
Dan Gohman
ee5673b622 Fix the sfence instruction to use MRM_F8 instead of MRM7r, since it
doesn't have a register operand. Also, use I instead of PSI, for
consistency with mfence and lfence.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104203 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-20 01:23:41 +00:00
Eric Christopher
51f5d6af8c Fix build by actually declaring the variable.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104201 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-20 00:59:30 +00:00
Eric Christopher
c1a887d76d Partial code for emitting thread local bss data.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104197 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-20 00:49:07 +00:00
Bill Wendling
ff9244a1f1 Match "4" or "8" depending upon if it's 32- or 64-bit.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104196 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-20 00:27:10 +00:00
Eric Christopher
8db52ef4b0 Once more, with feeling.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104190 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-20 00:07:13 +00:00
Daniel Dunbar
b0ac8677fd lit: Add another place to look for bash.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104189 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-19 23:56:09 +00:00
Dan Gohman
a2086b3483 Teach LSR how to cope better with unrolled loops on targets where
the addressing modes don't make this trivially easy. This allows
it to avoid falling into the less precise heuristics in more
cases.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104186 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-19 23:43:12 +00:00
Bob Wilson
492fd454ca Optimize away insertelement of an undef value. This shows up in
test/Codegen/ARM/reg_sequence.ll but it doesn't affect the generated code
because the coalescer cleans it up.  Radar 7998853.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104185 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-19 23:42:58 +00:00
Chris Lattner
a7f1354eb5 fix rdar://7986634 - match instruction opcodes case insensitively.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104183 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-19 23:34:33 +00:00
Bill Wendling
a008750aa9 Testcase for r104181.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104182 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-19 23:33:26 +00:00
Jim Grosbach
de70b1f9be Enable preserving debug information through post-RA scheduling
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104175 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-19 22:57:47 +00:00
Jim Grosbach
309d20c89c Fix the post-RA instruction scheduler to handle instructions referenced by
more than one dbg_value instruction. rdar://7759363


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104174 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-19 22:57:06 +00:00
Evan Cheng
e163168aab Code clean up.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104173 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-19 22:42:23 +00:00
Devang Patel
461a646a03 Revert r104165.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104172 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-19 21:58:28 +00:00
Jakob Stoklund Olesen
d32e735ae6 Add support for partial redefs to the fast register allocator.
A partial redef now triggers a reload if required. Also don't add
<imp-def,dead> operands for physical superregisters.

Kill flags are still treated as full register kills, and <imp-use,kill> operands
are added for physical superregisters as before.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104167 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-19 21:36:05 +00:00
Devang Patel
f4ccaeaef9 There is no need to maintain InsnsBeginScopeSet separately.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104165 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-19 21:26:53 +00:00
Eric Christopher
591466baff A more combo tls testcase.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104163 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-19 21:19:42 +00:00
Jakob Stoklund Olesen
7ebc4d63db Add MachineInstr::readsVirtualRegister() in preparation for proper handling of
partial redefines.

We are going to treat a partial redefine of a virtual register as a
read-modify-write:

  %reg1024:6 = OP

Unless the register is fully clobbered:

  %reg1024:6 = OP, %reg1024<imp-def>

MachineInstr::readsVirtualRegister() knows the difference. The first case is a
read, the second isn't.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104149 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-19 20:36:22 +00:00
Eric Christopher
aa6c72ec95 Few more simple tls testcases.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104148 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-19 20:35:15 +00:00
Evan Cheng
211ffa1351 Code refactoring: pull SchedPreference enum from TargetLowering.h to TargetMachine.h and put it in its own namespace.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104147 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-19 20:19:50 +00:00
Jakob Stoklund Olesen
3437352887 TwoAddressInstructionPass doesn't really know how to merge live intervals when
lowering REG_SEQUENCE instructions.

Insert copies for REG_SEQUENCE sources not killed to avoid breaking later passes.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104146 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-19 20:08:00 +00:00
Mikhail Glushenkov
139c9e1f85 llvmc: report an error if a child process segfaults.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104145 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-19 19:24:32 +00:00
Eric Christopher
d38bbfadfd Attempt to run this test on x86 only.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104143 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-19 18:59:37 +00:00
Bob Wilson
29e7e32e08 Testcase to go with 104141.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104142 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-19 18:58:37 +00:00
Bob Wilson
14b2141497 When expanding a vector_shuffle, the element type may not be legal and may
need to be promoted.  The BUILD_VECTOR and EXTRACT_VECTOR_ELT nodes generated
here already allow the promoted type to be used without further changes, so
just do the promotion.  This fixes part of pr7167.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104141 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-19 18:48:32 +00:00
Daniel Dunbar
d94406a940 MC/X86: Add missing entry for TAILJMP_1 to getRelaxedOpcode().
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104122 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-19 17:20:58 +00:00
Daniel Dunbar
52322e7b8b MC/X86: Lower TAILCALLd[64] to JMP_1, to allow relaxation and to avoid same
prefix byte problem as in r104062.
 - As a total hack to keep the TAILCALL markers in the output, which some tests depend on, this invents a new TAILJMP_1 instruction.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104120 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-19 15:26:43 +00:00
Daniel Dunbar
7d4bd203be MC/X86: Strip spurious operands from TAILJMPr64 as we do for CALL64r and
CALL64pcrel32, for the same reason.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104116 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-19 08:07:12 +00:00
Evan Cheng
9085f98b32 t2LEApcrel and tLEApcrel are re-materializable. This makes it possible to hoist more loads during machine LICM.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104115 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-19 07:28:01 +00:00
Evan Cheng
27fa722311 Use 'adr' for LEApcrel and LEApcrel. Mark LEApcrel re-materializable.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104114 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-19 07:26:50 +00:00
Tobias Grosser
0560ce471d Update autoconf/automake versions in the documentation to match the versions used in Autogen.sh
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104113 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-19 07:00:17 +00:00
Daniel Dunbar
597f17d5f1 MC/X86: Lower MOV{8,16,32,64}{rm,mr} to fixed-register forms, as appropriate.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104112 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-19 06:20:44 +00:00
Evan Cheng
5fd1c9be2d Mark pattern-less mayLoad / mayStore instructions neverHasSideEffects. These do not have other un-modeled side effects.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104111 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-19 06:07:03 +00:00
Evan Cheng
3c3195cbf1 Target instruction selection should copy memoperands.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104110 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-19 06:06:09 +00:00
Daniel Dunbar
9248b3208a MC/X86: Strip spurious operands from CALL64r as we do for CALL64pcrel32, to
avoid same prefix byte problem as in r104062.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104108 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-19 04:31:36 +00:00
Evan Cheng
ea420b20d4 Mark a few more pattern-less instructions with neverHasSideEffects. This is especially important on instructions like t2LEApcreal which are prime candidate for machine LICM.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104102 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-19 01:52:25 +00:00
Dan Gohman
86026cd24e Add a comment explaining why this code uses Append mode.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104095 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-19 01:21:34 +00:00
Evan Cheng
0a942dbb1e Intrinsics which do a vector compare (results are all zero or all ones) are modeled as icmp / fcmp + sext. This is turned into a vsetcc by dag combine (yes, not a good long term solution). The targets can then isel the vsetcc to the appropriate instruction.
The trouble arises when the result of a vector cmp + sext is then and'ed with all ones. Instcombine will turn it into a vector cmp + zext, dag combiner will miss turning it into a vsetcc and hell breaks loose after that.

Teach dag combine to turn a vector cpm + zest into a vsetcc + and 1. This fixes rdar://7923010.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104094 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-19 01:08:17 +00:00
Dan Gohman
7c2e03916c Factor out the code for picking integer arithmetic with immediate
opcodes into a helper function. This fixes a few places in the code
which were not properly selecting the 8-bit-immediate opcodes.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104091 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-19 00:53:19 +00:00
Eric Christopher
cc6b6b9348 Add a test to make sure that we're lowering the shift amount correctly.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104090 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-19 00:22:04 +00:00
Dan Gohman
57aaa0b264 Add a comment.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104089 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-18 23:55:57 +00:00
Dan Gohman
360026f07f Fix the predicate which checks for non-sensical formulae which have
constants in registers which partially cancel out their immediate fields.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104088 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-18 23:48:08 +00:00
Dan Gohman
b2df433f71 Factor out the code for recomputing an LSRUse's Regs set after some
of its formulae have been removed into a helper function, and also
teach it how to update the RegUseTracker.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104087 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-18 23:42:37 +00:00
Bob Wilson
417d1705bb Fix a crash when debugging the coalescer. DebugValue instructions are not
in the coalescer's instruction map.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104086 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-18 23:19:42 +00:00
Dan Gohman
d079c300ed Factor out code for estimating search space complexity into a helper
function.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104082 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-18 22:51:59 +00:00
Dan Gohman
0da751baf7 Add some more debug output.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104080 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-18 22:41:32 +00:00
Dan Gohman
d69d62833a Factor out the code for deleting a formula from an LSRUse into
a helper function.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104079 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-18 22:39:15 +00:00