Commit Graph

47915 Commits

Author SHA1 Message Date
Devang Patel 7293f0f344 Fix comments, help messages.
llvm-svn: 56282
2008-09-17 16:01:39 +00:00
Dan Gohman 6ab52a8018 Don't worry about clobbering physical register defs that aren't used.
llvm-svn: 56281
2008-09-17 15:25:49 +00:00
Steve Naroff a5629376e8 Fix http://llvm.org/bugs/show_bug.cgi?id=2760.
llvm-svn: 56280
2008-09-17 14:05:40 +00:00
Evan Cheng 7848cfcd77 Fix addrmode1 instruction encodings; fix bx_ret encoding.
llvm-svn: 56277
2008-09-17 07:53:38 +00:00
Evan Cheng 937569afe3 Specify instruction encoding using range list to avoid endianess issues.
llvm-svn: 56276
2008-09-17 07:16:21 +00:00
Evan Cheng 038ca4aa0f Add instruction names as comments to InstBits entries.
llvm-svn: 56275
2008-09-17 06:29:52 +00:00
Ted Kremenek 45a928b956 Updated checker build.
llvm-svn: 56274
2008-09-17 04:22:09 +00:00
Zhongxing Xu 7b7394c8ec Fix comment typo.
llvm-svn: 56271
2008-09-17 02:35:50 +00:00
Dan Gohman 173aa8602d Simplify and generalize X86DAGToDAGISel::CanBeFoldedBy, and draw
up some new ascii art to illustrate what it does. This change
currently has no effect on generated code.

llvm-svn: 56270
2008-09-17 01:39:10 +00:00
Daniel Dunbar 8fc81b02e2 Add support for ABIArgInfo::Expand
- No functionality change.

llvm-svn: 56269
2008-09-17 00:51:38 +00:00
Dan Gohman c24cd015a7 Add a new MachineInstr-level DCE pass. It is very simple, and is intended to
be used with fast-isel.

llvm-svn: 56268
2008-09-17 00:43:24 +00:00
Bill Wendling 95e1af217f Add trampoline support to PPC. GCC simply calls the "__trampoline_setup"
function with appropriate parameters. This allows us to support blocks on PPC.

llvm-svn: 56267
2008-09-17 00:30:57 +00:00
Steve Naroff 2da868e9c2 Add support for rewriting blocks...
llvm-svn: 56266
2008-09-17 00:13:27 +00:00
Devang Patel 11b862aed5 Fix cut-n-pasto.
llvm-svn: 56265
2008-09-17 00:01:04 +00:00
Daniel Dunbar 86169d6d0d Ignore XFAIL tests when checking for make failure in test/
llvm-svn: 56262
2008-09-16 23:43:13 +00:00
Ted Kremenek 025f83534c Fix copy-paste error in test case.
llvm-svn: 56261
2008-09-16 23:25:28 +00:00
Ted Kremenek 8782716c4a Minor pass-sensitivity improvement:
if we know that 'len != 0' and know that 'i == 0' then we know that
  'i < len' must evaluate to true and cannot evaluate to false

llvm-svn: 56260
2008-09-16 23:24:45 +00:00
Ted Kremenek 2b4b3b13a9 Update Xcode project.
llvm-svn: 56259
2008-09-16 23:21:45 +00:00
Evan Cheng a904f466e8 When converting a CopyFromReg to a copy instruction, use the register class of its uses to determine the right destination register class of the copy. This is important for targets where a physical register may belong to multiple register classes.
llvm-svn: 56258
2008-09-16 23:12:11 +00:00
Steve Naroff 7a147c6a3c Remove support for BlockExprExpr. For example...
^(expression) or ^(int arg1, float arg2)(expression)
...is no longer supported. 
All block literals now require a compound statement.

llvm-svn: 56257
2008-09-16 23:11:46 +00:00
Devang Patel 3fccede02a Remove.
llvm-svn: 56256
2008-09-16 22:26:44 +00:00
Devang Patel 9966ccffb4 Add -O1, -O2 and -O3 that matches llvm-gcc's -O1, -O2 and -O3 respectively.
llvm-svn: 56255
2008-09-16 22:25:14 +00:00
Steve Naroff 3b1e172d7e Sema::ActOnBlockReturnStmt(): Need to perform the UsualUnaryConversions on the return type.
Sema::CheckReturnStackAddr(): Make sure we skip over implicit casts.
Added some more test cases...

llvm-svn: 56254
2008-09-16 22:25:10 +00:00
Dan Gohman 64d6c6fe30 Change SelectionDAG::getConstantPool to always set the alignment of the
ConstantPoolSDNode, using the target's preferred alignment for the
constant type.

In LegalizeDAG, when performing loads from the constant pool, the
ConstantPoolSDNode's alignment is used in the calls to getLoad and
getExtLoad.

This change prevents SelectionDAG::getLoad/getExtLoad from incorrectly
choosing the ABI alignment for constant pool loads when Alignment == 0.
The incorrect alignment is only a performance issue when ABI alignment
does not equal preferred alignment (i.e., on x86 it was generating
MOVUPS instead of MOVAPS for v4f32 constant loads when the default ABI
alignment for 128bit vectors is forced to 1 byte.)

Patch by Paul Redmond!

llvm-svn: 56253
2008-09-16 22:05:41 +00:00
Bill Wendling 24c79f28b1 Reverting r56249. On further investigation, this functionality isn't needed.
Apologies for the thrashing.

llvm-svn: 56251
2008-09-16 21:48:12 +00:00
Dan Gohman ab26f20d44 Include the alignment value when displaying ConstantPoolSDNodes.
llvm-svn: 56250
2008-09-16 21:18:22 +00:00
Bill Wendling 8bc392fb1d - Change "ExternalSymbolSDNode" to "SymbolSDNode".
- Add linkage to SymbolSDNode (default to external).
- Change ISD::ExternalSymbol to ISD::Symbol.
- Change ISD::TargetExternalSymbol to ISD::TargetSymbol

These changes pave the way to allowing SymbolSDNodes with non-external linkage.

llvm-svn: 56249
2008-09-16 21:12:30 +00:00
Dan Gohman 5cf6120a7c Fix these comments to reflect current reality. Surprisingly,
MachineConstantPool::getConstantPoolIndex actually expects
a log2-encoded alignment.

llvm-svn: 56248
2008-09-16 20:45:53 +00:00
Dan Gohman dc5f5cbe59 Finally re-apply r46959. This is made feasible by the combination
of r56230, r56232, and r56246.

llvm-svn: 56247
2008-09-16 18:52:57 +00:00
Dan Gohman dafa9c6e85 Improve instcombine's handling of integer min and max in two ways:
- Recognize expressions like "x > -1 ? x : 0" as min/max and turn them
   into expressions like "x < 0 ? 0 : x", which is easily recognizable
   as a min/max operation.
 - Refrain from folding expression like "y/2 < 1" to "y < 2" when the
   comparison is being used as part of a min or max idiom, like
   "y/2 < 1 ? 1 : y/2". In that case, the division has another use, so
   folding doesn't eliminate it, and obfuscates the min/max, making it
   harder to recognize as a min/max operation.

These benefit ScalarEvolution, CodeGen, and anything else that wants to
recognize integer min and max.

llvm-svn: 56246
2008-09-16 18:46:06 +00:00
Ted Kremenek 0ecb53a421 ProgramPoint now takes the space of two pointers instead of one. This change was
motivated because it became clear that the number of subclasses of ProgramPoint
would expand and we ran out of bits to represent a pointer variant. As a plus of
this change, BlockEdge program points can now be represented explicitly without
using a cache of CFGBlock* pairs in CFG.

llvm-svn: 56245
2008-09-16 18:44:52 +00:00
Evan Cheng 7c5dbd95e2 AllocateRWXMemory -> AllocateRWX.
llvm-svn: 56244
2008-09-16 17:28:18 +00:00
Dan Gohman a9c71ce5dc FCmpInst predicates UNO, ORD, FALSE, and TRUE are commutative.
llvm-svn: 56243
2008-09-16 16:44:00 +00:00
Zhongxing Xu 5912c6e6f0 Fixed an offset calculation error.
llvm-svn: 56242
2008-09-16 07:58:21 +00:00
Bill Wendling 3f41b87e8d Fix test to account for no more whitespace and that one of the matches is the
"declare" statement.

llvm-svn: 56241
2008-09-16 06:40:04 +00:00
Bill Wendling 07fe6c8a07 Remove space that was forgotten.`
llvm-svn: 56240
2008-09-16 06:30:47 +00:00
Dan Gohman 050d7835c6 Don't take the time to CheckDAGForTailCallsAndFixThem when tail calls
are not enabled. Instead just omit the tail call flag when calls are
created.

llvm-svn: 56235
2008-09-16 01:42:28 +00:00
Dan Gohman 162568842e Fix spacing in the grep line for this test, following the recent
SCEV-whitespace changes.

llvm-svn: 56234
2008-09-16 01:37:08 +00:00
Dan Gohman c5d2892e62 Re-enables the new vector select in the bitcode reader, by modifying the
bitcode reader/writer as follows:

- add and use new bitcode FUNC_CODE_INST_VSELECT to handle the llvm
select opcode using either i1 or [N x i1] as the selector.
- retain old BITCODE FUNC_CODE_INST_SELECT in the bitcode reader to
handle select on i1 for backwards compatibility with existing bitcode
files.
- re-enable the vector-select.ll test program.

Also, rename the recently added bitcode opcode FUNC_CODE_INST_VCMP to
FUNC_CODE_INST_CMP2 and make the bitcode writer use it to handle
fcmp/icmp on scalars or vectors. In the bitcode writer, use
FUNC_CODE_INST_CMP for vfcmp/vicmp only. In the bitcode reader, have
FUNC_CODE_INST_CMP handle icmp/fcmp returning bool, for backwards
compatibility with existing bitcode files.

Patch by Preston Gurd!

llvm-svn: 56233
2008-09-16 01:01:33 +00:00
Dan Gohman f9081a2cd5 Teach ScalarEvolution to consider loop preheaders in the search for
an if statement that guards a loop, to allow indvars to avoid smax
operations in more situations.

llvm-svn: 56232
2008-09-15 22:18:04 +00:00
Owen Anderson 82ab1e7280 Live intervals for live-in registers should begin at the beginning of a basic block, not at the first
instruction.  Also, their valno's should have an unknown def.  This has no effect currently, but was
causing issues when StrongPHIElimination was enabled.

llvm-svn: 56231
2008-09-15 22:00:38 +00:00
Dan Gohman 68e7735a38 Teach LSR to optimize away SMAX operations for tripcounts in common
cases.  See the comment above OptimizeSMax for the full story, and
the testcase for an example. This cancels out a pessimization
commonly attributed to indvars, and will allow us to lift some of
the artificial throttles in indvars, rather than add new ones.

llvm-svn: 56230
2008-09-15 21:22:06 +00:00
Devang Patel 4089ccb0fa Extract optimization pass selection code from llvm-gcc into a separate routine.
This can be used by other stand alone tools, such as 'opt'.

llvm-svn: 56229
2008-09-15 21:13:42 +00:00
Bruno Cardoso Lopes 233785daae Fixed Bug 2751
http://llvm.org/bugs/show_bug.cgi?id=2751

Abicall was enabled even when static code model was provided 
in the command line.
The correct behavior is to disable abicall when static is
specified.

llvm-svn: 56228
2008-09-15 21:06:55 +00:00
Dan Gohman 3c7b9ba547 Re-enable SelectionDAG CSE for calls. It matters in the case of
libcalls, as in this testcase on ARM.

llvm-svn: 56226
2008-09-15 19:46:03 +00:00
Bruno Cardoso Lopes 595229a15a Added testcase for bswap allegrexel intrinsic
llvm-svn: 56225
2008-09-15 19:38:11 +00:00
Duncan Sands 18da83555f End of the GlobalsModRef experiment.
llvm-svn: 56222
2008-09-15 18:42:38 +00:00
Dan Gohman 46673ad82a Update the LLVM polygen grammar for recent language changes:
x86_ssecallcc, function notes, and some whitespace adjustments.

llvm-svn: 56221
2008-09-15 16:10:51 +00:00
Nick Lewycky 57368a561f Fix documentation for these functions; they do not only modify users within the
block specified by the InstList. Patch by Stefanus Du Toit.

llvm-svn: 56200
2008-09-15 06:31:52 +00:00
Evan Cheng 02acc35abd Correctly update kill infos after extending a live range and merge 2 val#'s; fix 56165 - do not mark val# copy field if the copy does not define the val#.
llvm-svn: 56199
2008-09-15 06:28:41 +00:00