Duncan Sands [Wed, 22 Apr 2009 10:39:51 +0000 (10:39 +0000)]
These tests are x86 specific.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69798
91177308-0d34-0410-b5e6-
96231b3b80d8
Owen Anderson [Wed, 22 Apr 2009 08:50:12 +0000 (08:50 +0000)]
Real fix for PR3549, by using caching for predecessor counts in addition to the predecessors themselves. This halves the time
to optimize the testcase, beyond what my previous patch did.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69792
91177308-0d34-0410-b5e6-
96231b3b80d8
Owen Anderson [Wed, 22 Apr 2009 08:46:33 +0000 (08:46 +0000)]
Add caching of predecessor counts as well as predecessors themselves.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69791
91177308-0d34-0410-b5e6-
96231b3b80d8
Owen Anderson [Wed, 22 Apr 2009 08:09:13 +0000 (08:09 +0000)]
Use PredIteratorCache in LCSSA, which gives a 37% overall speedup on the testcase from PR3549. More improvements to come.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69788
91177308-0d34-0410-b5e6-
96231b3b80d8
Owen Anderson [Wed, 22 Apr 2009 07:16:54 +0000 (07:16 +0000)]
Fix for PR3946, in which SmallVector could behave strangely in corner cases.
Patch by Peter Johnson.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69785
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Tue, 21 Apr 2009 23:37:18 +0000 (23:37 +0000)]
use predicate instead of hand-rolled loop
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69752
91177308-0d34-0410-b5e6-
96231b3b80d8
Dan Gohman [Tue, 21 Apr 2009 23:15:49 +0000 (23:15 +0000)]
De-pImpl-ify ScalarEvolution. The pImpl pattern doesn't provide much
practical benefit in the case of ScalarEvolution, and it's otherwise
a nuisance.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69749
91177308-0d34-0410-b5e6-
96231b3b80d8
Evan Cheng [Tue, 21 Apr 2009 22:46:52 +0000 (22:46 +0000)]
It has finally happened. Spiller is now using live interval info.
This fixes a very subtle bug. vr defined by an implicit_def is allowed overlap with any register since it doesn't actually modify anything. However, if it's used as a two-address use, its live range can be extended and it can be spilled. The spiller must take care not to emit a reload for the vn number that's defined by the implicit_def. This is both a correctness and performance issue.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69743
91177308-0d34-0410-b5e6-
96231b3b80d8
Dan Gohman [Tue, 21 Apr 2009 20:18:36 +0000 (20:18 +0000)]
When turning (ashr(shl(x, n), n)) into sext(trunc(x)), the width of the
type to truncate to should be the number of bits of the value that are
preserved, not the number that are clobbered with sign-extension.
This fixes regressions in ldecod.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69704
91177308-0d34-0410-b5e6-
96231b3b80d8
Mikhail Glushenkov [Tue, 21 Apr 2009 19:46:10 +0000 (19:46 +0000)]
Support --with-llvmgccdir and friends in llvmc, take 2.
Should now work when building with objdir != srcdir and when llvm-gcc is not
available.
Thanks to Duncan Sands for testing and advice!
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69700
91177308-0d34-0410-b5e6-
96231b3b80d8
Devang Patel [Tue, 21 Apr 2009 17:21:01 +0000 (17:21 +0000)]
Test case for revision 69683.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69684
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Tue, 21 Apr 2009 16:52:12 +0000 (16:52 +0000)]
fix a crash on a pointless but valid zero-length memset, rdar://
6808691
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69680
91177308-0d34-0410-b5e6-
96231b3b80d8
Anton Korobeynikov [Tue, 21 Apr 2009 16:04:56 +0000 (16:04 +0000)]
Silence warnings.
Patch by Jay Foad!
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69679
91177308-0d34-0410-b5e6-
96231b3b80d8
Anton Korobeynikov [Tue, 21 Apr 2009 16:04:41 +0000 (16:04 +0000)]
Drop obsolete reference to __eprintf.
Patch by Jay Foad!
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69678
91177308-0d34-0410-b5e6-
96231b3b80d8
Anton Korobeynikov [Tue, 21 Apr 2009 16:04:14 +0000 (16:04 +0000)]
'The "or die" is intended to catch the case where nm returned a
non-zero exit status, so nm will already have printed some error
messages.'
Patch by Jay Foad!
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69677
91177308-0d34-0410-b5e6-
96231b3b80d8
Duncan Sands [Tue, 21 Apr 2009 09:44:39 +0000 (09:44 +0000)]
Get rid of what looks like a copy-and-pasted typo.
Spotted by gcc-4.5.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69673
91177308-0d34-0410-b5e6-
96231b3b80d8
Rafael Espindola [Tue, 21 Apr 2009 08:22:09 +0000 (08:22 +0000)]
TLS_addr64 and TLS_addr32 define RDI and EAX. They don't use them.
This fixes PR4002.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69672
91177308-0d34-0410-b5e6-
96231b3b80d8
Sanjiv Gupta [Tue, 21 Apr 2009 05:54:51 +0000 (05:54 +0000)]
Handle direct aggregate type arguments.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69665
91177308-0d34-0410-b5e6-
96231b3b80d8
Dan Gohman [Tue, 21 Apr 2009 02:26:00 +0000 (02:26 +0000)]
Teach ScalarEvolution how to recognize zext-inreg and sext-inreg,
as they appear in LLVM IR. This isn't particularly interesting
on its own; this is just setting up some infrastructure.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69655
91177308-0d34-0410-b5e6-
96231b3b80d8
Dan Gohman [Tue, 21 Apr 2009 01:41:18 +0000 (01:41 +0000)]
This FIXME is fixed, now that SCEV understands pointers.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69651
91177308-0d34-0410-b5e6-
96231b3b80d8
Dan Gohman [Tue, 21 Apr 2009 01:25:57 +0000 (01:25 +0000)]
Factor out a common base class from SCEVTruncateExpr, SCEVZeroExtendExpr,
and SCEVSignExtendExpr.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69649
91177308-0d34-0410-b5e6-
96231b3b80d8
Dan Gohman [Tue, 21 Apr 2009 01:11:19 +0000 (01:11 +0000)]
Usage getAnalysisToUpdate for TargetData, per PR760.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69645
91177308-0d34-0410-b5e6-
96231b3b80d8
Dan Gohman [Tue, 21 Apr 2009 01:07:12 +0000 (01:07 +0000)]
Introduce encapsulation for ScalarEvolution's TargetData object, and refactor
the code to minimize dependencies on TargetData.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69644
91177308-0d34-0410-b5e6-
96231b3b80d8
Dan Gohman [Tue, 21 Apr 2009 00:55:22 +0000 (00:55 +0000)]
Move some assertion checks so they can do more complete checking.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69643
91177308-0d34-0410-b5e6-
96231b3b80d8
Dan Gohman [Tue, 21 Apr 2009 00:47:46 +0000 (00:47 +0000)]
Convert ScalarEvolution to use raw_ostream instead of OStream.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69640
91177308-0d34-0410-b5e6-
96231b3b80d8
Devang Patel [Tue, 21 Apr 2009 00:08:56 +0000 (00:08 +0000)]
Fix Visual Studio 2008 build failure.
Patch by Marius Wachtler
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69637
91177308-0d34-0410-b5e6-
96231b3b80d8
Dan Gohman [Mon, 20 Apr 2009 22:54:34 +0000 (22:54 +0000)]
Make X86's copyRegToReg able to handle copies to and from subclasses.
This makes the extra copyRegToReg calls in ScheduleDAGSDNodesEmit.cpp
unnecessary. Derived from a patch by Jakob Stoklund Olesen.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69635
91177308-0d34-0410-b5e6-
96231b3b80d8
Dan Gohman [Mon, 20 Apr 2009 22:51:43 +0000 (22:51 +0000)]
Simplify this code. getConstant knows how to make
broadcasted vector constants.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69634
91177308-0d34-0410-b5e6-
96231b3b80d8
Evan Cheng [Mon, 20 Apr 2009 22:49:59 +0000 (22:49 +0000)]
No, we are not avoiding -O3, just -fstrict-aliasing.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69633
91177308-0d34-0410-b5e6-
96231b3b80d8
Dale Johannesen [Mon, 20 Apr 2009 22:19:33 +0000 (22:19 +0000)]
Adjust loop size estimate for full unrolling;
GEP's don't usually become instructions.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69631
91177308-0d34-0410-b5e6-
96231b3b80d8
Evan Cheng [Mon, 20 Apr 2009 22:16:40 +0000 (22:16 +0000)]
One Mac OS X, just build with -O3 but without -fstrict-aliasing (which is kinda broken).
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69630
91177308-0d34-0410-b5e6-
96231b3b80d8
Daniel Dunbar [Mon, 20 Apr 2009 20:50:13 +0000 (20:50 +0000)]
Make Unix.h:MakeErrMsg separate the prefix and errno string, so we get:
clang: error: unable to make temporary file: /etc/cc: can't make
unique filename: Permission denied
instead of
clang: error: unable to make temporary file: /etc/cc: can't make
unique filenamePermission denied
for example.
Also, audited the uses of MakeErrMsg to make the prefix strings
consistent (not end with newline/punctuation/space/": ").
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69626
91177308-0d34-0410-b5e6-
96231b3b80d8
Daniel Dunbar [Mon, 20 Apr 2009 20:34:38 +0000 (20:34 +0000)]
Remove unused variable.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69624
91177308-0d34-0410-b5e6-
96231b3b80d8
Bob Wilson [Mon, 20 Apr 2009 18:36:57 +0000 (18:36 +0000)]
Move duplicated AddLiveIn function from X86 and ARM backends to be a method
in the MachineFunction class, renaming it to addLiveIn for consistency with
the same method in MachineBasicBlock. Thanks for Anton for suggesting this.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69615
91177308-0d34-0410-b5e6-
96231b3b80d8
Devang Patel [Mon, 20 Apr 2009 18:07:22 +0000 (18:07 +0000)]
Match C backend only if it explicitly requested.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69613
91177308-0d34-0410-b5e6-
96231b3b80d8
Tanya Lattner [Mon, 20 Apr 2009 17:48:16 +0000 (17:48 +0000)]
Remove clang since its conditionally there already.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69610
91177308-0d34-0410-b5e6-
96231b3b80d8
Bob Wilson [Mon, 20 Apr 2009 17:27:09 +0000 (17:27 +0000)]
Revise my previous change 68996 as suggested by Duncan.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69607
91177308-0d34-0410-b5e6-
96231b3b80d8
Evan Cheng [Mon, 20 Apr 2009 17:23:48 +0000 (17:23 +0000)]
- Remove an arbitrary spill weight tweak that should not have been there.
- Find more reloads from SS.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69606
91177308-0d34-0410-b5e6-
96231b3b80d8
Sanjiv Gupta [Mon, 20 Apr 2009 16:59:35 +0000 (16:59 +0000)]
Emit the auto variables of a function into a different section than parameters.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69605
91177308-0d34-0410-b5e6-
96231b3b80d8
Dan Gohman [Mon, 20 Apr 2009 16:26:25 +0000 (16:26 +0000)]
It's not necessary for PrintModulePass to flush the output streams
now that errs() is properly non-buffered.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69602
91177308-0d34-0410-b5e6-
96231b3b80d8
Dan Gohman [Mon, 20 Apr 2009 16:19:02 +0000 (16:19 +0000)]
Use .empty() instead of .size().
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69599
91177308-0d34-0410-b5e6-
96231b3b80d8
Dan Gohman [Mon, 20 Apr 2009 16:10:33 +0000 (16:10 +0000)]
Don't discard an AssemblyAnnotationWriter when writing GlobalValues,
which include Functions, where it can be quite useful to use an
AssemblyAnnotationWriter.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69598
91177308-0d34-0410-b5e6-
96231b3b80d8
Duncan Sands [Mon, 20 Apr 2009 16:03:21 +0000 (16:03 +0000)]
These bitfields were being miscompiled on some
64 bit platforms when building with optimization.
So replace them by a hand-coded implementation.
This fixes PR3822.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69597
91177308-0d34-0410-b5e6-
96231b3b80d8
Dan Gohman [Mon, 20 Apr 2009 15:55:38 +0000 (15:55 +0000)]
Implement operator<<(raw_ostream &OS, const Type &T).
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69596
91177308-0d34-0410-b5e6-
96231b3b80d8
Evan Cheng [Mon, 20 Apr 2009 08:01:12 +0000 (08:01 +0000)]
Added a linearscan register allocation optimization. When the register allocator spill an interval with multiple uses in the same basic block, it creates a different virtual register for each of the reloads. e.g.
%reg1498<def> = MOV32rm %reg1024, 1, %reg0, 12, %reg0, Mem:LD(4,4) [sunkaddr39 + 0]
%reg1506<def> = MOV32rm %reg1024, 1, %reg0, 8, %reg0, Mem:LD(4,4) [sunkaddr42 + 0]
%reg1486<def> = MOV32rr %reg1506
%reg1486<def> = XOR32rr %reg1486, %reg1498, %EFLAGS<imp-def,dead>
%reg1510<def> = MOV32rm %reg1024, 1, %reg0, 4, %reg0, Mem:LD(4,4) [sunkaddr45 + 0]
=>
%reg1498<def> = MOV32rm %reg2036, 1, %reg0, 12, %reg0, Mem:LD(4,4) [sunkaddr39 + 0]
%reg1506<def> = MOV32rm %reg2037, 1, %reg0, 8, %reg0, Mem:LD(4,4) [sunkaddr42 + 0]
%reg1486<def> = MOV32rr %reg1506
%reg1486<def> = XOR32rr %reg1486, %reg1498, %EFLAGS<imp-def,dead>
%reg1510<def> = MOV32rm %reg2038, 1, %reg0, 4, %reg0, Mem:LD(4,4) [sunkaddr45 + 0]
From linearscan's point of view, each of reg2036, 2037, and 2038 are separate registers, each is "killed" after a single use. The reloaded register is available and it's often clobbered right away. e.g. In thise case reg1498 is allocated EAX while reg2036 is allocated RAX. This means we end up with multiple reloads from the same stack slot in the same basic block.
Now linearscan recognize there are other reloads from same SS in the same BB. So it'll "downgrade" RAX (and its aliases) after reg2036 is allocated until the next reload (reg2037) is done. This greatly increase the likihood reloads from SS are reused.
This speeds up sha1 from OpenSSL by 5.8%. It is also an across the board win for SPEC2000 and 2006.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69585
91177308-0d34-0410-b5e6-
96231b3b80d8
Douglas Gregor [Mon, 20 Apr 2009 07:34:17 +0000 (07:34 +0000)]
Make all raw_ostreams support the tell() function.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69583
91177308-0d34-0410-b5e6-
96231b3b80d8
Sanjiv Gupta [Mon, 20 Apr 2009 06:05:54 +0000 (06:05 +0000)]
Before trying to introduce/eliminate cast/ext/trunc to make indices type as
pointer type, make sure that the pointer size is a valid sequential index type.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69574
91177308-0d34-0410-b5e6-
96231b3b80d8
Nick Lewycky [Sun, 19 Apr 2009 18:32:03 +0000 (18:32 +0000)]
Use an AssertingVH to detect the case where the Function was deleted but
freeMachineCodeForFunction was never called.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69531
91177308-0d34-0410-b5e6-
96231b3b80d8
Nick Lewycky [Sun, 19 Apr 2009 18:20:21 +0000 (18:20 +0000)]
Fix missing text in doxygen documentation.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69529
91177308-0d34-0410-b5e6-
96231b3b80d8
Duncan Sands [Sun, 19 Apr 2009 06:40:30 +0000 (06:40 +0000)]
Now that BUILD_VECTOR operands are allowed to be
bigger than the vector element type, turn checking
of the operand type back on again, appropriately
adjusted.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69516
91177308-0d34-0410-b5e6-
96231b3b80d8
Duncan Sands [Sun, 19 Apr 2009 06:23:05 +0000 (06:23 +0000)]
Remove the SimpleTy enumerated type field from the MVT
value type union: this field was causing problems for
some compilers on 64 bit systems, presumably because
SimpleTy is 32 bits wide while the other fields are
64 bits wide.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69515
91177308-0d34-0410-b5e6-
96231b3b80d8
Mikhail Glushenkov [Sun, 19 Apr 2009 00:22:35 +0000 (00:22 +0000)]
Add some assertions.
Fixes segfaults in some corner cases.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69494
91177308-0d34-0410-b5e6-
96231b3b80d8
Bill Wendling [Sat, 18 Apr 2009 21:45:27 +0000 (21:45 +0000)]
Revert 69474 and 69475. They are causing failures during a bootstrap on Darwin.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69478
91177308-0d34-0410-b5e6-
96231b3b80d8
Mikhail Glushenkov [Sat, 18 Apr 2009 20:55:55 +0000 (20:55 +0000)]
Reconfigure.
Turns out that doing this by hand is easier than using autoreconf:-).
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69475
91177308-0d34-0410-b5e6-
96231b3b80d8
Mikhail Glushenkov [Sat, 18 Apr 2009 20:55:28 +0000 (20:55 +0000)]
Add a configure check for llvm-gcc (reapply).
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69474
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sat, 18 Apr 2009 20:49:22 +0000 (20:49 +0000)]
testcase for PR3898
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69473
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sat, 18 Apr 2009 20:48:07 +0000 (20:48 +0000)]
Fix PR3898, which manifests as failures on are an Xcore,
patch by Jakob Stoklund Olesen!
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69472
91177308-0d34-0410-b5e6-
96231b3b80d8
Duncan Sands [Sat, 18 Apr 2009 20:16:54 +0000 (20:16 +0000)]
Don't try to make BUILD_VECTOR operands have the same
type as the vector element type: allow them to be of
a wider integer type than the element type all the way
through the system, and not just as far as LegalizeDAG.
This should be safe because it used to be this way
(the old type legalizer would produce such nodes), so
backends should be able to handle it. In fact only
targets which have legal vector types with an illegal
promoted element type will ever see this (eg: <4 x i16>
on ppc). This fixes a regression with the new type
legalizer (vec_splat.ll). Also, treat SCALAR_TO_VECTOR
the same as BUILD_VECTOR. After all, it is just a
special case of BUILD_VECTOR.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69467
91177308-0d34-0410-b5e6-
96231b3b80d8
Nick Lewycky [Sat, 18 Apr 2009 18:11:26 +0000 (18:11 +0000)]
Generalize to support more ARM types.
Configure was not actually regenerated, but the change last time only touched
this one line, so I'm being lazy and cheating by fixing it manually.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69453
91177308-0d34-0410-b5e6-
96231b3b80d8
Dan Gohman [Sat, 18 Apr 2009 17:58:19 +0000 (17:58 +0000)]
Add a ScalarEvolution::getCouldNotCompute() function, and use it
instead of allocating and leaking new SCEVCouldNotCompute objects.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69452
91177308-0d34-0410-b5e6-
96231b3b80d8
Dan Gohman [Sat, 18 Apr 2009 17:57:20 +0000 (17:57 +0000)]
More const qualifiers.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69451
91177308-0d34-0410-b5e6-
96231b3b80d8
Dan Gohman [Sat, 18 Apr 2009 17:56:28 +0000 (17:56 +0000)]
Use more const qualifiers with SCEV interfaces.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69450
91177308-0d34-0410-b5e6-
96231b3b80d8
Nick Lewycky [Sat, 18 Apr 2009 15:41:38 +0000 (15:41 +0000)]
Regenerate.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69447
91177308-0d34-0410-b5e6-
96231b3b80d8
Nick Lewycky [Sat, 18 Apr 2009 15:34:25 +0000 (15:34 +0000)]
Detect beagleboard as ARM. The $target is "armv7l".
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69446
91177308-0d34-0410-b5e6-
96231b3b80d8
Duncan Sands [Sat, 18 Apr 2009 12:40:19 +0000 (12:40 +0000)]
Make it clearer that llvm-gcc is not needed for
building llvm.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69441
91177308-0d34-0410-b5e6-
96231b3b80d8
Bill Wendling [Sat, 18 Apr 2009 11:20:33 +0000 (11:20 +0000)]
Temporarily revert r69438 and r69439. These were causing failures during a
release build of llvm.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69440
91177308-0d34-0410-b5e6-
96231b3b80d8
Mikhail Glushenkov [Sat, 18 Apr 2009 09:59:26 +0000 (09:59 +0000)]
Regenerate.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69439
91177308-0d34-0410-b5e6-
96231b3b80d8
Mikhail Glushenkov [Sat, 18 Apr 2009 09:57:58 +0000 (09:57 +0000)]
Add a configure check for llvm-gcc.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69438
91177308-0d34-0410-b5e6-
96231b3b80d8
Evan Cheng [Sat, 18 Apr 2009 08:52:15 +0000 (08:52 +0000)]
Add a new LiveInterval::overlaps(). It checks if the live interval overlaps a range specified by [Start, End).
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69434
91177308-0d34-0410-b5e6-
96231b3b80d8
Mon P Wang [Sat, 18 Apr 2009 05:16:01 +0000 (05:16 +0000)]
Fixed a few 64 bit cases in X86InstrInfo::commuteInstruction
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69417
91177308-0d34-0410-b5e6-
96231b3b80d8
Dale Johannesen [Sat, 18 Apr 2009 02:01:23 +0000 (02:01 +0000)]
Adjust XFAIL syntax, maybe that will help. The other
way worked for me...
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69414
91177308-0d34-0410-b5e6-
96231b3b80d8
Dale Johannesen [Sat, 18 Apr 2009 00:11:50 +0000 (00:11 +0000)]
patch 69408 breaks this by removing the opportunity
for the optimization it's testing to kick in (although
it improves the code, getting rid of all spills).
I don't understand the optimization well enough to
rescue the test, so XFAILing.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69409
91177308-0d34-0410-b5e6-
96231b3b80d8
Dale Johannesen [Sat, 18 Apr 2009 00:09:40 +0000 (00:09 +0000)]
Inline asm's were still introducing bogus dependencies;
my earlier patch to this code only fixed half of it.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69408
91177308-0d34-0410-b5e6-
96231b3b80d8
Jim Grosbach [Fri, 17 Apr 2009 23:30:55 +0000 (23:30 +0000)]
remove trailing whitespace
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69402
91177308-0d34-0410-b5e6-
96231b3b80d8
Bill Wendling [Fri, 17 Apr 2009 22:40:38 +0000 (22:40 +0000)]
Recommit r69335 and r69336. These were not causing problems.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69394
91177308-0d34-0410-b5e6-
96231b3b80d8
Bob Wilson [Fri, 17 Apr 2009 20:42:34 +0000 (20:42 +0000)]
Move the AddLiveIn function definition closer to its uses.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69382
91177308-0d34-0410-b5e6-
96231b3b80d8
Bob Wilson [Fri, 17 Apr 2009 20:40:45 +0000 (20:40 +0000)]
Rearrange code to reduce indentation.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69381
91177308-0d34-0410-b5e6-
96231b3b80d8
Bob Wilson [Fri, 17 Apr 2009 20:40:20 +0000 (20:40 +0000)]
Rename file to have the correct suffix.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69380
91177308-0d34-0410-b5e6-
96231b3b80d8
Bob Wilson [Fri, 17 Apr 2009 20:35:10 +0000 (20:35 +0000)]
Clean up formatting, remove trailing whitespace, fix comment typos and
punctuation. No functional changes.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69378
91177308-0d34-0410-b5e6-
96231b3b80d8
Dan Gohman [Fri, 17 Apr 2009 20:03:31 +0000 (20:03 +0000)]
Delete an unused field.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69375
91177308-0d34-0410-b5e6-
96231b3b80d8
Bob Wilson [Fri, 17 Apr 2009 19:07:39 +0000 (19:07 +0000)]
Use CallConvLower.h and TableGen descriptions of the calling conventions
for ARM. Patch by Sandeep Patel.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69371
91177308-0d34-0410-b5e6-
96231b3b80d8
Jim Grosbach [Fri, 17 Apr 2009 17:25:16 +0000 (17:25 +0000)]
Update information on canadian cross builds
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69358
91177308-0d34-0410-b5e6-
96231b3b80d8
David Greene [Fri, 17 Apr 2009 14:56:18 +0000 (14:56 +0000)]
Use a safer iterator interface and get rid of std C++ library misuse.
This fixes a --enable-expensive-checks problem.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69353
91177308-0d34-0410-b5e6-
96231b3b80d8
David Greene [Fri, 17 Apr 2009 14:50:39 +0000 (14:50 +0000)]
Regenerate configure.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69352
91177308-0d34-0410-b5e6-
96231b3b80d8
David Greene [Fri, 17 Apr 2009 14:49:22 +0000 (14:49 +0000)]
Add a --enable-profiling option to configure to build Debug+Profile and
Opt+Profile tools. Now we can profile any kind of flavor we build.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69351
91177308-0d34-0410-b5e6-
96231b3b80d8
Rafael Espindola [Fri, 17 Apr 2009 14:35:58 +0000 (14:35 +0000)]
For general dynamic TLS access we must use
leaq foo@TLSGD(%rip), %rdi
as part of the instruction sequence. Using a register other than %rdi and then
copying it to %rdi is not valid.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69350
91177308-0d34-0410-b5e6-
96231b3b80d8
Bill Wendling [Fri, 17 Apr 2009 04:19:22 +0000 (04:19 +0000)]
Revert r69335 and r69336. They were causing build failures.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69347
91177308-0d34-0410-b5e6-
96231b3b80d8
Dan Gohman [Fri, 17 Apr 2009 02:02:52 +0000 (02:02 +0000)]
Don't create ConstantInts with pointer type. This fixes a
regression in 403.gcc in PIC_CODEGEN=1 and DISABLE_LTO=1
mode.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69344
91177308-0d34-0410-b5e6-
96231b3b80d8
Evan Cheng [Fri, 17 Apr 2009 01:29:40 +0000 (01:29 +0000)]
Teach spiller to unfold instructions which modref spill slot when a scratch
register is available and when it's profitable.
e.g.
xorq %r12<kill>, %r13
addq %rax, -184(%rbp)
addq %r13, -184(%rbp)
==>
xorq %r12<kill>, %r13
movq -184(%rbp), %r12
addq %rax, %r12
addq %r13, %r12
movq %r12, -184(%rbp)
Two more instructions, but fewer memory accesses. It can also open up
opportunities for more optimizations.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69341
91177308-0d34-0410-b5e6-
96231b3b80d8
Dan Gohman [Fri, 17 Apr 2009 00:45:17 +0000 (00:45 +0000)]
MOV8rr_NOREX is a "Move" instruction. This doesn't currently
matter, because this instruction isn't generated until after
things that care.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69336
91177308-0d34-0410-b5e6-
96231b3b80d8
Dan Gohman [Fri, 17 Apr 2009 00:43:09 +0000 (00:43 +0000)]
Don't use MOV8rr_NOREX on x86-32. It doesn't actually hurt anything at
present, but it's inconsistent.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69335
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Fri, 17 Apr 2009 00:26:12 +0000 (00:26 +0000)]
Fix some failures in targets on available_externally functions,
this fixes a crash on CodeGen/Generic/externally_available.ll
on ppc hosts. Thanks to Nicholas L for pointing this out.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69333
91177308-0d34-0410-b5e6-
96231b3b80d8
Dan Gohman [Thu, 16 Apr 2009 22:35:57 +0000 (22:35 +0000)]
Use TargetData::getTypeSizeInBits instead of getPrimitiveSizeInBits()
to get the correct answer for pointer types.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69321
91177308-0d34-0410-b5e6-
96231b3b80d8
Bob Wilson [Thu, 16 Apr 2009 21:51:05 +0000 (21:51 +0000)]
Fix PR3994: LLVMMatchType arguments do not refer to absolute return value
and argument positions but only to the overloaded intrinsic parameters.
Keep a separate list of these overloaded parameters in CodeGenTarget.cpp
so they can be resolved easily. Remove assertions from IntrinsicEmitter.cpp:
they were harmless but confusing, and the assertions elsewhere in TableGen
will catch any incorrect values.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69316
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Thu, 16 Apr 2009 21:47:59 +0000 (21:47 +0000)]
make sure to unlock keymgr if the JIT is created and destroyed, all
locks must be matched with unlocks. Also, use calloc to allocate the
block so that it is properly zero'd. Thanks to Nick Kledzik for
tracking this down.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69314
91177308-0d34-0410-b5e6-
96231b3b80d8
Bob Wilson [Thu, 16 Apr 2009 21:46:42 +0000 (21:46 +0000)]
Add a comment to describe LLVMMatchType.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69313
91177308-0d34-0410-b5e6-
96231b3b80d8
Eli Friedman [Thu, 16 Apr 2009 21:40:28 +0000 (21:40 +0000)]
Fix for PR3944: make mem2reg O(N) instead of O(N^2) in the number of
incoming edges for a block with many predecessors.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69312
91177308-0d34-0410-b5e6-
96231b3b80d8
Dan Gohman [Thu, 16 Apr 2009 21:34:54 +0000 (21:34 +0000)]
Handle a pointer type correctly in SCEVExpander::visitAddRecExpr.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69310
91177308-0d34-0410-b5e6-
96231b3b80d8
Dan Gohman [Thu, 16 Apr 2009 20:59:02 +0000 (20:59 +0000)]
In the list-burr's pseudo two-addr dependency heuristics, don't
add dependencies on nodes with exactly one successor which is a
COPY_TO_REGCLASS node. In the case that the copy is coalesced
away, the dependence should be on the user of the copy, rather
than the copy itself.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69309
91177308-0d34-0410-b5e6-
96231b3b80d8
Dan Gohman [Thu, 16 Apr 2009 20:57:10 +0000 (20:57 +0000)]
Handle SUBREG_TO_REG instructions with the same heuristics
as INSERT_SUBREG instructions in the list-burr scheduler.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@69308
91177308-0d34-0410-b5e6-
96231b3b80d8