Commit Graph

64523 Commits

Author SHA1 Message Date
NAKAMURA Takumi 7a23aa081a ARM/Disassembler: Fix definitions incompatible(unsigned and uint32_t) to Cygwin-1.5, following up to r113255.
llvm-svn: 113345
2010-09-08 04:48:17 +00:00
Chris Lattner 4703cb4a96 fix the encoding of the "jump on *cx" family of instructions,
rdar://8061602

llvm-svn: 113343
2010-09-08 04:30:51 +00:00
Jim Grosbach 535d3b4e09 remove trailing whitespace
llvm-svn: 113338
2010-09-08 03:54:02 +00:00
Jim Grosbach 19cb2f4c67 remove obsolete comment
llvm-svn: 113337
2010-09-08 03:51:44 +00:00
Jim Grosbach 261df12f64 disable for the moment while tracking down a few Thumb2-O0 failure that look
related. (attempt deux, complete w/ test update this time)

llvm-svn: 113333
2010-09-08 02:00:34 +00:00
Jim Grosbach b2c950187e woops. need to update a test along with this.
llvm-svn: 113332
2010-09-08 01:49:09 +00:00
Jim Grosbach 7cda56ea6a disable temporarily while sorting out a few test failures in Thumb2-O0 tests.
llvm-svn: 113331
2010-09-08 01:47:49 +00:00
Dan Gohman 64d842ec72 Add a new experimental generalized dependence query interface to
AliasAnalysis, and some code for implementing the new query on top of
existing implementations by making standard alias and getModRefInfo
queries.

llvm-svn: 113329
2010-09-08 01:32:20 +00:00
Jim Grosbach 136d035e45 correct spill code to properly determine if dynamic stack realignment is
present in the function and thus whether aligned load/store instructions can
be used.

llvm-svn: 113323
2010-09-08 00:26:59 +00:00
Jim Grosbach abcbe2474d VFP/NEON load/store multiple instructions are addrmode4, not 5.
llvm-svn: 113322
2010-09-08 00:25:50 +00:00
Owen Anderson a4d9c78aa1 Add a separate unrolling threshold when the current function is being optimized for size.
The threshold value of 50 is arbitrary, and I chose it simply by analogy to the inlining thresholds, where
the baseline unrolling threshold is slightly smaller than the baseline inlining threshold.  This could
undoubtedly use some tuning.

llvm-svn: 113306
2010-09-07 23:15:30 +00:00
John McCall 978f007a80 Add documentation for llvm-diff.
llvm-svn: 113303
2010-09-07 23:10:21 +00:00
Jakob Stoklund Olesen 4d19d2651d Don't add <imp-def> operands during register rewriting.
LiveIntervals already adds <imp-def> operands for super-registers when a subreg
def defines the whole register. Thus, it is not necessary to do it again when
rewriting.

In fact, the super-register imp-defs caused miscompilations because the late
scheduler couldn't see that the super-register was read.

We still add super-reg <imp-use,kill> operands when rewriting virtuals to
physicals.

llvm-svn: 113299
2010-09-07 22:38:45 +00:00
Jim Grosbach 88628e9738 To shrink a t2LDM instruction to the 16-bit wide tLDM instruction, the base
register must be one of the destination registers for the load. Otherwise,
the tLDM instruction will write-back to the base register, which isn't what's
desired (otherwise, we'd have a t2LDM_UPD instead).

rdar://8394087

llvm-svn: 113297
2010-09-07 22:30:53 +00:00
Devang Patel 3f4abf397c remove these tests for now.
llvm-svn: 113293
2010-09-07 22:03:44 +00:00
Jim Grosbach 9877af3b46 grammar tweak
llvm-svn: 113289
2010-09-07 21:30:25 +00:00
Bill Wendling 9040eecb49 Remove untrue comments.
llvm-svn: 113287
2010-09-07 21:07:59 +00:00
Bruno Cardoso Lopes 6b1d62c529 Factor out some x86 vector shuffle rewriting and add comments about the direction the shuffle lowering is heading to
llvm-svn: 113286
2010-09-07 21:03:14 +00:00
Devang Patel b0af23a1f6 There is no need to force target if the test is going to run on other x86 platforms.
llvm-svn: 113285
2010-09-07 20:59:09 +00:00
Owen Anderson 866e516ead Remove dead code. ManagedCleanup is unused, and contained a serious bug in that
the provided cleanup function is never actually called.

llvm-svn: 113284
2010-09-07 20:53:39 +00:00
Owen Anderson d12ea002b8 Fix PR7972, in which the PassRegistry was being leaked. As part of this,
switch to using a ManagedStatic for the global PassRegistry instead of a
ManagedCleanup, and fix a destruction ordering bug this exposed.

llvm-svn: 113283
2010-09-07 20:48:10 +00:00
Stuart Hastings 420c8a604f Typo. Thanks to BillW for pointing it out!
llvm-svn: 113281
2010-09-07 20:39:07 +00:00
Dan Gohman 9afd666faf Tidy up the getModRefInfo declarations.
llvm-svn: 113275
2010-09-07 20:37:47 +00:00
Bruno Cardoso Lopes 7c483028fb Move code around to prepare for moving some of the logic together to another function
llvm-svn: 113267
2010-09-07 20:20:27 +00:00
Owen Anderson 29ed9459d0 Properly initialize the pImpl member of PassRegistry to zero.
llvm-svn: 113264
2010-09-07 20:13:48 +00:00
Owen Anderson 3c8019c94d Add doxygen comments for PassRegistry.
llvm-svn: 113262
2010-09-07 20:04:26 +00:00
Bill Wendling 353802114f Add an MVT::x86mmx type. It will take the place of all current MMX vector types.
llvm-svn: 113261
2010-09-07 20:03:56 +00:00
Chris Lattner 6e27b3e004 Fix a serious performance regression introduced by r108687 on linux:
turning (fptrunc (sqrt (fpext x))) -> (sqrtf x)  is great, but we have
to delete the original sqrt as well.  Not doing so causes us to do 
two sqrt's when building with -fmath-errno (the default on linux).

llvm-svn: 113260
2010-09-07 20:01:38 +00:00
Evan Cheng 5444b36e01 Remove a dead comment.
llvm-svn: 113259
2010-09-07 20:01:10 +00:00
Chris Lattner 29570bd695 rename test.
llvm-svn: 113257
2010-09-07 19:57:06 +00:00
Chris Lattner 091012d5d5 hopefully fix a problem building on cygwin-1.5
llvm-svn: 113255
2010-09-07 19:50:53 +00:00
Owen Anderson a74fa15f32 Clean up some of the PassRegistry implementation, and pImpl-ize it to reduce #include clutter
and exposing internal details.

llvm-svn: 113252
2010-09-07 19:16:25 +00:00
Bill Wendling 02b701f558 Fix whitespace, because I'm OCD.
llvm-svn: 113250
2010-09-07 18:49:14 +00:00
Stuart Hastings a3188a81c0 Test case for r113248. Raar 8361341.
llvm-svn: 113249
2010-09-07 18:43:57 +00:00
Bruno Cardoso Lopes 5a45db3e6c decouple MMX check from regular splat checks. Some refactoring is coming, and MMX should be left alone to be easily removed after moving to intrinsics
llvm-svn: 113247
2010-09-07 18:41:45 +00:00
Bruno Cardoso Lopes 4f5d4b4a6e Remove now useless check, because the code can be matched below, no need to leave it for isel
llvm-svn: 113242
2010-09-07 18:29:03 +00:00
Dan Gohman d51b43181e Issue a #error if the host doesn't have an implementation for
GetMainExecutable yet.

llvm-svn: 113240
2010-09-07 18:26:49 +00:00
Bruno Cardoso Lopes c9b3316fea Minor change. Since the checks are equivalent, use isMMX
llvm-svn: 113239
2010-09-07 18:24:00 +00:00
Devang Patel e50b23e223 Fix command line used to link these test cases.
llvm-svn: 113237
2010-09-07 18:17:56 +00:00
Nick Lewycky 1facc1db8f Update configure for change to autoconf/configure.ac to add PTX backend.
llvm-svn: 113236
2010-09-07 18:14:42 +00:00
Nick Lewycky 4c82c6c6fe Create PTX backend. Patch by Che-Liang Chiou!
llvm-svn: 113235
2010-09-07 18:14:24 +00:00
Dale Johannesen 605acfe533 Add patterns for MMX that use the new intrinsics.
Enable palignr intrinsic.
These may need adjustment for a new VT in due course.

llvm-svn: 113233
2010-09-07 18:10:56 +00:00
Devang Patel 9dc0e5be58 Reintroduce dbg-declare tests.
llvm-svn: 113232
2010-09-07 18:01:49 +00:00
Bruno Cardoso Lopes f0ea222255 Remove unused target specific node
llvm-svn: 113224
2010-09-07 17:38:55 +00:00
Devang Patel 688338eec3 Remove last three tests. I need to make them independent of my setup.
llvm-svn: 113213
2010-09-07 17:08:57 +00:00
Devang Patel 55a3bab0d2 Add a test case to check handling of dbg-declare during hybrid mode where we begin using fast-isel but switch back to DAG building at some point.
llvm-svn: 113210
2010-09-07 17:03:44 +00:00
Devang Patel 29a775adf1 Add a test case to check handling of dbg-declare by selection DAG builder.
llvm-svn: 113209
2010-09-07 16:56:35 +00:00
Devang Patel 184c81c3e2 Add a test case to check handling of dbg-declare by fast-isel.
llvm-svn: 113208
2010-09-07 16:40:53 +00:00
Benjamin Kramer 1ecb978214 Don't leak the old operand when transforming "sldt" into "sldtw".
llvm-svn: 113200
2010-09-07 14:40:58 +00:00
Chris Lattner 6282336772 attempt to appease msvc
llvm-svn: 113198
2010-09-07 06:10:48 +00:00
Gabor Greif 7f3ce25e6e fix comment typos
llvm-svn: 113197
2010-09-07 06:06:06 +00:00
Nick Lewycky ad48e01eef Add completely hokey binary-and and binary-or operations to ConstantRange and
teach LazyValueInfo to use them.

llvm-svn: 113196
2010-09-07 05:39:02 +00:00
Nick Lewycky 71972d45dc Fix major bug in thunk detection. Also verify the calling convention.
Switch from isWeakForLinker to mayBeOverridden which is more accurate.

Add more statistics and debugging info. Add comments. Move static function
outside anonymous namespace.

llvm-svn: 113190
2010-09-07 01:42:10 +00:00
Chris Lattner 30bb384944 add missing cmov aliases, this resolves rdar://8208499
llvm-svn: 113189
2010-09-07 00:05:45 +00:00
Chris Lattner 3ae9398d5f remove duplicated entry
llvm-svn: 113188
2010-09-06 23:57:24 +00:00
Nick Lewycky a35462da3d Add a new isSignWrappedSet() method to ConstantRange.
Fix zeroExtend and signExtend to support empty sets, and to return the smallest
possible result set which contains the extension of each element in their
inputs. For example zext i8 [100, 10) to i16 is now [0, 256), not i16 [100, 10)
which contains 63446 members.

llvm-svn: 113187
2010-09-06 23:52:49 +00:00
Chris Lattner 7ece716da2 "sldt <mem>" is ambiguous in 64-bit mode, but should
always be disambiguated as sldtw.  sldtw and sldtq with
a mem operands have the same effect, but sldtw is more
compact.  Force it to sldtw, resolving rdar://8017530

llvm-svn: 113186
2010-09-06 23:51:44 +00:00
Chris Lattner 415e04fad2 fix rdar://8017621 - llvm-mc can't guess encoding for "push $(1000)"
llvm-svn: 113184
2010-09-06 23:40:56 +00:00
Chris Lattner abfe4223c2 generalize my previous operand loc info hack. If the same operand
is busted for all variants, report it as the location.  This allows
us to get the operand right for bugs like:

t.s:3:12: error: invalid operand for instruction
	outb %al, %gs
	          ^

Even though there are reg/imm and reg/reg forms of this instruction.

llvm-svn: 113183
2010-09-06 23:37:39 +00:00
Chris Lattner 34e366b45c fix the operand constraints of the immediate form of in/out,
allowing unsigned 8-bit operands.  This fixes rdar://8208481

llvm-svn: 113182
2010-09-06 23:29:05 +00:00
Chris Lattner 339cc7bfef in the case where an instruction only has one implementation
of a mneumonic, report operand errors with better location
info.  For example, we now report:

t.s:6:14: error: invalid operand for instruction
        cwtl $1
             ^

but we fail for common cases like:

t.s:11:4: error: invalid operand for instruction
   addl $1, $1
   ^

because we don't know if this is supposed to be the reg/imm or imm/reg
form.

llvm-svn: 113178
2010-09-06 22:11:18 +00:00
Chris Lattner 628fbecf4f Now that we know if we had a total fail on the instruction mnemonic,
give a more detailed error.  Before:

t.s:11:4: error: unrecognized instruction
   addl $1, $1
   ^
t.s:12:4: error: unrecognized instruction
   f2efqefa $1
   ^

After:

t.s:11:4: error: invalid operand for instruction
   addl $1, $1
   ^
t.s:12:4: error: invalid instruction mnemonic 'f2efqefa'
   f2efqefa $1
   ^

This fixes rdar://8017912 - llvm-mc says "unrecognized instruction" when it means "invalid operands"

llvm-svn: 113176
2010-09-06 21:54:15 +00:00
Chris Lattner c0658cbd16 simplify DEBUG_WITH_TYPE usage
llvm-svn: 113174
2010-09-06 21:28:52 +00:00
Chris Lattner c4521d1b5f this if can now be an assert.
llvm-svn: 113173
2010-09-06 21:25:43 +00:00
Chris Lattner 9026ac0edd ;
llvm-svn: 113172
2010-09-06 21:23:43 +00:00
Chris Lattner 8130197937 now that the opcode is trivially exposed, start matching instructions
by doing a binary search over the mnemonic instead of doing a linear
search through all possible instructions.  This implements rdar://7785064

llvm-svn: 113171
2010-09-06 21:22:45 +00:00
Chris Lattner 6b6f3dd994 emit the match table at global scope instead of within the
MatchInstructionImpl. This makes it easier to read/understand
MatchInstructionImpl.

llvm-svn: 113170
2010-09-06 21:08:38 +00:00
Chris Lattner 82d88ced92 special case the mnemonic operand of the instruction in the
generated matcher, emiting it as a column in the MatchEntry
table instead of forcing it to go through classification and
everything else.  Making it be classified caused tblgen to
produce a ton of one-off classes for each mneumonic.  This
should reduce the size of the generated matcher significantly
while paving the way for future improvements.

llvm-svn: 113169
2010-09-06 21:01:37 +00:00
Chris Lattner fdb7decfaf The "ambiguous instructions" check only produces anything with -debug,
so only do the N^2 loop with debug mode.

llvm-svn: 113168
2010-09-06 20:21:47 +00:00
Chris Lattner 31c63fb518 simplify the hacks around jrcxz.
llvm-svn: 113167
2010-09-06 20:10:12 +00:00
Chris Lattner b4be28f33d have tblgen detect when an instruction would have matched, but
failed because a subtarget feature was not enabled.  Use this to
remove a bunch of hacks from the X86AsmParser for rejecting things
like popfl in 64-bit mode.  Previously these hacks weren't needed,
but were important to get a message better than "invalid instruction"
when used in the wrong mode.

This also fixes bugs where pushal would not be rejected correctly in
32-bit mode (just pusha).

llvm-svn: 113166
2010-09-06 20:08:02 +00:00
Chris Lattner a22a368e7c change MatchInstructionImpl to return an enum instead of bool.
llvm-svn: 113165
2010-09-06 19:22:17 +00:00
Chris Lattner 8128ca1c2c add note
llvm-svn: 113164
2010-09-06 19:14:40 +00:00
Chris Lattner 3e4582ada5 have AsmMatcherEmitter.cpp produce the hunk of code that gets included
into the middle of the class, and rework how the different sections of
the generated file are conditionally included for simplicity.

llvm-svn: 113163
2010-09-06 19:11:01 +00:00
Eric Christopher b33877b710 Regenerate.
llvm-svn: 113160
2010-09-06 18:47:10 +00:00
Eric Christopher 1243b9df84 Update to 2.9 post-2.8 branch.
llvm-svn: 113159
2010-09-06 18:47:00 +00:00
Roman Divacky e1278b57f9 Redefine LOOP* instructions from I to Ii8PCRel as they take an i8 argument.
llvm-svn: 113158
2010-09-06 18:43:14 +00:00
Chris Lattner 4cfbcdc7b6 random cleanups
llvm-svn: 113157
2010-09-06 18:32:06 +00:00
Benjamin Kramer 3723ba1fe0 MCELF: Align symtab, relocation sections and section headers properly. Patch by Krister Wombell.
llvm-svn: 113155
2010-09-06 16:11:52 +00:00
Chris Lattner 963debc109 compute the HasSSE3 bit correctly, patch by Nikolai Saoukh.
llvm-svn: 113147
2010-09-06 05:19:44 +00:00
Chris Lattner be9019090e fix PR8067, an over-aggressive assertion in LICM.
llvm-svn: 113146
2010-09-06 05:11:24 +00:00
Chris Lattner 07b332f0a0 emit the LLVM intrinsic name -> intrinsic number mapping table with
StringMatcher instead of a linear sequence of memcmps.

llvm-svn: 113145
2010-09-06 03:58:45 +00:00
Chris Lattner a58edd1df3 cleanup some of the lifetime/invariant marker stuff, add a big fixme.
llvm-svn: 113144
2010-09-06 03:58:04 +00:00
Chris Lattner 96fe532c67 allow specifying an indentation level for the string matcher.
llvm-svn: 113143
2010-09-06 03:50:59 +00:00
Chris Lattner 497d13e82b emit the __builtin -> intrinsic map with StringMatcher instead of a
copy of a close relative of it.

llvm-svn: 113142
2010-09-06 03:14:45 +00:00
Chris Lattner 6dcaa42540 same bug, another place.
llvm-svn: 113141
2010-09-06 03:12:27 +00:00
Chris Lattner 9bb3bf1b59 fix a critical bug where the generated table would say
"1 strings to match" in a comment, which isn't gramatic.

llvm-svn: 113140
2010-09-06 03:11:10 +00:00
Chris Lattner 69b5913500 update cmake
llvm-svn: 113139
2010-09-06 02:58:25 +00:00
Chris Lattner ca5a3554b5 factor the snazzy string matcher code that Daniel hates
out of AsmMatcherEmitter.cpp into its own class.

llvm-svn: 113137
2010-09-06 02:01:51 +00:00
Chris Lattner fb43da66b5 slightly improve the runtime and code size of the Intrinsics info table by not
comparing the "llvm." prefix in the memcmp, and not storing it in the string literal.

llvm-svn: 113136
2010-09-06 01:44:44 +00:00
Chris Lattner e34c835bde speed up -gvn 3.4% on the testcase in PR7023
llvm-svn: 113135
2010-09-06 01:26:29 +00:00
Chris Lattner b01c24a945 Teach loop rotate to hoist trivially invariant instructions
in the duplicated block instead of duplicating them.  

Duplicating them into the end of the loop and the preheader 
means that we got a phi node in the header of the loop, 
which prevented LICM from hoisting them.  GVN would
usually come around later and merge the duplicated 
instructions so we'd get reasonable output... except that
anything dependent on the shoulda-been-hoisted value can't
be hoisted.  In PR5319 (which this fixes), a memory value
didn't get promoted.

llvm-svn: 113134
2010-09-06 01:10:22 +00:00
Chris Lattner da24b9a49a pull a simple method out of LICM into a new
Loop::hasLoopInvariantOperands method. Remove
a useless and confusing Loop::isLoopInvariant(Instruction)
method, which didn't do what you thought it did.

No functionality change.

llvm-svn: 113133
2010-09-06 01:05:37 +00:00
Chris Lattner 7b7768a269 fit in 80 columns and don't crash on exit, fixes PR8080
llvm-svn: 113123
2010-09-05 23:09:30 +00:00
Chris Lattner f43cb302ca remove some dead code. t2addrmode_imm8s4 is never used in a
pattern, so there is no need to define a matching function.

llvm-svn: 113122
2010-09-05 22:51:11 +00:00
Chris Lattner 15bd17e572 fix inconsistent formatting.
llvm-svn: 113121
2010-09-05 22:43:56 +00:00
Chris Lattner accb015f7c cleanups: mark stuff static, only tagdecls should be in anon namespaces.
llvm-svn: 113120
2010-09-05 21:25:43 +00:00
Chris Lattner e40007a71b cleanups.
llvm-svn: 113119
2010-09-05 21:18:45 +00:00
Chris Lattner 8c26cee3f3 some random notes.
llvm-svn: 113118
2010-09-05 20:49:45 +00:00
Chris Lattner 419d0aa0ed add a comment about where this should eventually move.
llvm-svn: 113117
2010-09-05 20:33:40 +00:00
Chris Lattner 5cac0f71ca update this.
llvm-svn: 113116
2010-09-05 20:22:09 +00:00
Chris Lattner 1edf7434cf more cleanups
llvm-svn: 113115
2010-09-05 20:13:07 +00:00
Chris Lattner e6214557e7 Change lower atomic pass to use IntrinsicInst to simplify it a bit.
llvm-svn: 113114
2010-09-05 20:10:47 +00:00
Chris Lattner 05ef361b5e eliminate some non-obvious casts. UndefValue isa Constant.
llvm-svn: 113113
2010-09-05 20:03:09 +00:00
Chris Lattner 72d283c826 fix PR8063, a crash in globalopt in the malloc analysis code.
llvm-svn: 113109
2010-09-05 17:20:46 +00:00
Lang Hames 64a4a13617 Added initialisers for reduction rule counters.
llvm-svn: 113108
2010-09-05 13:42:32 +00:00
Nick Lewycky e3ac69eca3 Fix warning reported by MSVC++ builder.
llvm-svn: 113106
2010-09-05 09:11:38 +00:00
Nick Lewycky f3a07ec394 Switch FnSet to containing the ComparableFunction instead of a pointer to one.
This reduces malloc traffic (yay!) and removes MergeFunctionsEqualityInfo.

llvm-svn: 113105
2010-09-05 09:00:32 +00:00
Nick Lewycky 0095937b13 Fix many bugs when merging weak-strong and weak-weak pairs. We now merge all
strong functions first to make sure they're the canonical definitions and then
do a second pass looking only for weak functions.

llvm-svn: 113104
2010-09-05 08:22:49 +00:00
Chris Lattner eeba0c73e5 implement rdar://6653118 - fastisel should fold loads where possible.
Since mem2reg isn't run at -O0, we get a ton of reloads from the stack,
for example, before, this code:

int foo(int x, int y, int z) {
  return x+y+z;
}

used to compile into:

_foo:                                   ## @foo
	subq	$12, %rsp
	movl	%edi, 8(%rsp)
	movl	%esi, 4(%rsp)
	movl	%edx, (%rsp)
	movl	8(%rsp), %edx
	movl	4(%rsp), %esi
	addl	%edx, %esi
	movl	(%rsp), %edx
	addl	%esi, %edx
	movl	%edx, %eax
	addq	$12, %rsp
	ret

Now we produce:

_foo:                                   ## @foo
	subq	$12, %rsp
	movl	%edi, 8(%rsp)
	movl	%esi, 4(%rsp)
	movl	%edx, (%rsp)
	movl	8(%rsp), %edx
	addl	4(%rsp), %edx    ## Folded load
	addl	(%rsp), %edx     ## Folded load
	movl	%edx, %eax
	addq	$12, %rsp
	ret

Fewer instructions and less register use = faster compiles.

llvm-svn: 113102
2010-09-05 02:18:34 +00:00
Jakob Stoklund Olesen 313358fef9 Remove dead code.
Clobber ranges are no longer used when joining physical registers.
Instead, all aliases are checked for interference.

llvm-svn: 113084
2010-09-04 21:09:33 +00:00
Chris Lattner f20ca8028d fix this to work with allocators that have reference type with compilers
that diagnose invalid references to references.

llvm-svn: 113078
2010-09-04 18:45:02 +00:00
Chris Lattner 0963048185 dead method.
llvm-svn: 113077
2010-09-04 18:19:16 +00:00
Chris Lattner 2f095c636a delete dead code.
llvm-svn: 113075
2010-09-04 18:17:16 +00:00
Chris Lattner 65b48b5dfc zap dead code.
llvm-svn: 113073
2010-09-04 18:12:00 +00:00
Chris Lattner 55118e7224 remove dead code, mblaze uses SelectAddrRegImm/SelectAddrRegReg,
not SelectAddr

llvm-svn: 113072
2010-09-04 18:02:47 +00:00
Chris Lattner b1951ea1fe zap dead code.
llvm-svn: 113071
2010-09-04 17:45:58 +00:00
Bruno Cardoso Lopes c6accda78e Remove the last bit of isShuffleMaskLegal checks and improve the comment regarding mmx shuffles
llvm-svn: 113059
2010-09-04 02:58:56 +00:00
Bruno Cardoso Lopes 731bcc1abf make explicit that we not handle several mmx shuffles
llvm-svn: 113058
2010-09-04 02:50:13 +00:00
Dan Gohman 487e250109 Fix LoopSimplify to notify ScalarEvolution when splitting a loop backedge
into an inner loop, as the new loop iteration may differ substantially.
This fixes PR8078.

llvm-svn: 113057
2010-09-04 02:42:48 +00:00
Bruno Cardoso Lopes 20779ee157 Emit target specific nodes to handle palignr. Do not touch it for MMX versions yet.
llvm-svn: 113056
2010-09-04 02:36:07 +00:00
Bruno Cardoso Lopes cff7cd18ab Emit target specific nodes to handle splats starting at zero indicies
llvm-svn: 113055
2010-09-04 02:02:14 +00:00
Bruno Cardoso Lopes 95759917eb Emit target specific nodes for isPSHUFHWMask and isPSHUFLWMask
llvm-svn: 113050
2010-09-04 01:36:45 +00:00
Bruno Cardoso Lopes 2b57008c72 Emit target specific nodes for isSHUFPMask
llvm-svn: 113048
2010-09-04 01:22:57 +00:00
Bruno Cardoso Lopes 2f7af36134 Previous isMOVLMask matching already emits targets nodes, remove check
llvm-svn: 113047
2010-09-04 00:50:08 +00:00
Bruno Cardoso Lopes 9f8e704151 One more check from the original isShuffleMaskLegal goes away
llvm-svn: 113045
2010-09-04 00:46:16 +00:00
Bruno Cardoso Lopes 16959372bb Remove a duplicated but useless check that i've inserted in the previous commit.
llvm-svn: 113044
2010-09-04 00:43:12 +00:00
Bruno Cardoso Lopes 44578d38d3 Refactor some code and remove the extra checks for unpckl_undef and unpckh_undef
llvm-svn: 113043
2010-09-04 00:39:43 +00:00
Chris Lattner 50506787d1 fix a bug in my licm rewrite when a load from the promoted memory
location is being re-stored to the memory location.  We would get
a dangling pointer from the SSAUpdate data structure and miss a 
use.  This fixes PR8068

llvm-svn: 113042
2010-09-04 00:12:30 +00:00
Bruno Cardoso Lopes 7829d0e74b Remove check for unpckh mask
llvm-svn: 113035
2010-09-03 23:32:47 +00:00
Bruno Cardoso Lopes d1dacc57aa Remove check for unpckl mask
llvm-svn: 113034
2010-09-03 23:31:50 +00:00
Bruno Cardoso Lopes 207b9d6218 Inline isShuffleMaskLegal into LowerVECTOR_SHUFFLE, so we can start
checking each standalone condition and decide whether emit target
specific nodes or remove the condition if it's already matched before.

llvm-svn: 113031
2010-09-03 23:24:06 +00:00
Owen Anderson c91c1a205a Propagate non-local comparisons. Fixes PR1757.
llvm-svn: 113025
2010-09-03 22:47:08 +00:00
Eric Christopher 0acd994f90 Reapply this increase to the number of virtual registers. All of the
various breakages appear to be dealt with.

Patch by Pekka Jääskeläinen.

llvm-svn: 113024
2010-09-03 22:46:31 +00:00
Dan Gohman 47bec3cb57 Disable the asserts that check that normalization is perfectly
invertible. ScalarEvolution's folding routines don't always succeed
in canonicalizing equal expressions to a single canonical form, and
this can cause these asserts to fail, even though there's no actual
correctness problem. This fixes PR8066.

llvm-svn: 113021
2010-09-03 22:12:56 +00:00
Bruno Cardoso Lopes 2bef20eda7 Reapply considered harmfull part of rr112934 and r112942.
"Use target specific nodes instead of relying in unpckl and
unpckh pattern fragments during isel time. Also place a
depth limit in getShuffleScalarElt.

llvm-svn: 113020
2010-09-03 22:09:41 +00:00
Jim Grosbach 005155e236 previous patch was a little too tricky for its own good. Don't try to
overload UserInInstr. Explicitly check Allocatable. The early exit in the
condition will mean the performance impact of the extra test should be
minimal.

llvm-svn: 113016
2010-09-03 21:45:15 +00:00
Dale Johannesen 367afb5a00 Remove the rest of the nonexistent 64-bit AVX instructions.
Bruno, please review.

llvm-svn: 113014
2010-09-03 21:23:00 +00:00
David Greene 2a9de4d828 Generalize getFieldType to work on all TypedInits. Add a couple of testcases from
Amaury Pouly.

llvm-svn: 113010
2010-09-03 21:00:49 +00:00
Bruno Cardoso Lopes a750d994fe Reapply last harmless part of r112934, the pattern fragment to match X86Unpcklpd
llvm-svn: 113009
2010-09-03 20:44:26 +00:00
Bruno Cardoso Lopes fe8717c573 Reintroduce a simple function refactoring done in r112934, also without any functionality changes
llvm-svn: 113008
2010-09-03 20:20:02 +00:00
Bruno Cardoso Lopes 48e589b122 Reapply piecies of r112942 and r112934 which don't do
functional changes

llvm-svn: 113007
2010-09-03 20:10:35 +00:00
Bruno Cardoso Lopes 6979cf0808 Reapply Fix comment
llvm-svn: 113006
2010-09-03 19:55:05 +00:00
Daniel Dunbar 6f3da24d70 Revert r112934, "- Use specific nodes to match unpckl masks.", which introduced
some infinite loop and select failures.
 - Apologies for eager reverting, but its branch day.

llvm-svn: 113000
2010-09-03 19:38:11 +00:00
Daniel Dunbar f1aacd55c0 Revert r112938 "Fix comment", which depends on r112934, which introduced some
infinite loop and select failures.

llvm-svn: 112999
2010-09-03 19:38:08 +00:00
Daniel Dunbar 0ffe4db45c Revert r112942, "Use punpckh and unpckh family of nodes instead of using unpckh
mask pattern fragment", which depends on r112934, which introduced some infinite
loop and select failures.

llvm-svn: 112998
2010-09-03 19:38:05 +00:00
Bob Wilson 3626a8c136 Add a missing check when legalizing a vector extending load. This doesn't
solve the root problem, but it corrects the bug in the code I added to
support legalizing in the case where the non-extended type is also legal.

llvm-svn: 112997
2010-09-03 19:20:37 +00:00
Bill Wendling 78395e4b8a Get rid of "passing signed into unsigned parameter" warning on PPC.
llvm-svn: 112995
2010-09-03 19:09:46 +00:00
Owen Anderson c725462245 Add support for simplifying a load from a computed value to a load from a global when it
is provable that they're equivalent.  This fixes PR4855.

llvm-svn: 112994
2010-09-03 19:08:37 +00:00
Bill Wendling cef985bc7e Quiesce warning about non-virtual d'tor in virtual class.
llvm-svn: 112991
2010-09-03 18:41:20 +00:00