oota-llvm.git
16 years agoMoved method call within a conditional branch because its effects will
Ted Kremenek [Mon, 21 Jan 2008 22:51:35 +0000 (22:51 +0000)]
Moved method call within a conditional branch because its effects will
be ignored on the false branch.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46228 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoAdjusted ImutAVLTree::ComputeHash to compute a hash value that is based on a
Ted Kremenek [Mon, 21 Jan 2008 22:50:37 +0000 (22:50 +0000)]
Adjusted ImutAVLTree::ComputeHash to compute a hash value that is based on a
clearer sequence of hashing compositions.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46227 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoAdded "clear" method to FoldingSetNodeID to allow a FoldingSetNodeID object
Ted Kremenek [Mon, 21 Jan 2008 22:49:22 +0000 (22:49 +0000)]
Added "clear" method to FoldingSetNodeID to allow a FoldingSetNodeID object
to be reused to compute multiple object profiles.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46226 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoReplaced (FoldingSet) profiling of ImutAVLTree with a hashing based scheme. The
Ted Kremenek [Mon, 21 Jan 2008 22:33:30 +0000 (22:33 +0000)]
Replaced (FoldingSet) profiling of ImutAVLTree with a hashing based scheme. The
problem was that we previously hashed based on the pointers of the left and
right children, but this is bogus: we can easily have different trees that
represent the same set. Now we use a hashing based scheme that compares the
*contents* of the trees, but not without having to do a full scan of a tree. The
only caveat is that with hashing is that we may have collisions, which result in
two different trees being falsely labeled as equivalent. If this becomes a
problem, we can add extra data to the profile to hopefully resolve most
collisions.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46224 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoProvide iterator access to ListInit contents
Anton Korobeynikov [Mon, 21 Jan 2008 22:30:26 +0000 (22:30 +0000)]
Provide iterator access to ListInit contents

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46223 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoFix the failures in the PPC JIT by marking relocation entries for
Chris Lattner [Mon, 21 Jan 2008 22:27:27 +0000 (22:27 +0000)]
Fix the failures in the PPC JIT by marking relocation entries for
external symbols (e.g. 'fmod') as needing a stub.  This regression
was introduced by Evan's jit patch here:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20071231/056749.html

With this fixed, the two ExecutionEngine failures are passing on ppc,
and the ppc jit works on freebench and olden.

This should be pulled into the 2.2 release branch.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46222 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoNew test.
Devang Patel [Mon, 21 Jan 2008 22:15:58 +0000 (22:15 +0000)]
New test.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46220 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoClarify a deviation from the original algorithm.
Owen Anderson [Mon, 21 Jan 2008 22:03:00 +0000 (22:03 +0000)]
Clarify a deviation from the original algorithm.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46218 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoImprove a few comments.
Owen Anderson [Mon, 21 Jan 2008 22:01:01 +0000 (22:01 +0000)]
Improve a few comments.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46217 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoIt turns out that in C++ it is legal to declare functions
Duncan Sands [Mon, 21 Jan 2008 21:37:41 +0000 (21:37 +0000)]
It turns out that in C++ it is legal to declare functions
that return an opaque type by value, as long as you don't
call it or provide a body (you can take the address of it).
So it is wrong to insist that sret parameters not be an
opaque*.  And I guess it is really up to codegen to complain
if someone tries to call such a function.  I'm also removing
the analogous check from byval parameters, since I don't
see why we shouldn't allow them as long as no-one tries to
call the function or give it a body.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46216 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoNew test.
Devang Patel [Mon, 21 Jan 2008 19:28:13 +0000 (19:28 +0000)]
New test.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46209 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoCheck that sret is only used on pointers to types
Duncan Sands [Mon, 21 Jan 2008 11:28:49 +0000 (11:28 +0000)]
Check that sret is only used on pointers to types
with a size, like byval.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46207 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoBe consistent with other attribute methods, and
Duncan Sands [Mon, 21 Jan 2008 11:27:55 +0000 (11:27 +0000)]
Be consistent with other attribute methods, and
check the callee also if it is known.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46206 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoMove DAG-changing code out of #ifndef NDEBUG.
Dale Johannesen [Mon, 21 Jan 2008 01:00:34 +0000 (01:00 +0000)]
Move DAG-changing code out of #ifndef NDEBUG.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46204 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoMake sure the caller doesn't use freed memory.
Duncan Sands [Sun, 20 Jan 2008 16:51:46 +0000 (16:51 +0000)]
Make sure the caller doesn't use freed memory.
Fixes PR1935.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46203 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoHonour ByVal parameter attribute for name decoration
Anton Korobeynikov [Sun, 20 Jan 2008 14:00:07 +0000 (14:00 +0000)]
Honour ByVal parameter attribute for name decoration

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46200 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoRemove Darwin'ism
Anton Korobeynikov [Sun, 20 Jan 2008 13:59:37 +0000 (13:59 +0000)]
Remove Darwin'ism

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46199 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoEnable PIC codegen on x86-64/linux
Anton Korobeynikov [Sun, 20 Jan 2008 13:58:16 +0000 (13:58 +0000)]
Enable PIC codegen on x86-64/linux

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46198 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoInitializing an unsigned with ~0UL causes the compiler
Duncan Sands [Sun, 20 Jan 2008 10:49:23 +0000 (10:49 +0000)]
Initializing an unsigned with ~0UL causes the compiler
to complain on x86-64 (gcc 4.1).  Use ~0U instead.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46197 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoDo not generate a FP_ROUND of f64 to f64.
Dale Johannesen [Sun, 20 Jan 2008 01:18:38 +0000 (01:18 +0000)]
Do not generate a FP_ROUND of f64 to f64.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46195 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoChanged argument name for 'Profile' method to potentially fix a name conflict
Ted Kremenek [Sat, 19 Jan 2008 17:21:43 +0000 (17:21 +0000)]
Changed argument name for 'Profile' method to potentially fix a name conflict
reported in pr1929 (http://llvm.org/PR1929).

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46193 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoNeed to handle any 'nest' parameter before integer
Duncan Sands [Sat, 19 Jan 2008 16:42:10 +0000 (16:42 +0000)]
Need to handle any 'nest' parameter before integer
parameters, since otherwise it won't be passed in
the right register.  With this change trampolines
work on x86-64 (thanks to Luke Guest for providing
access to an x86-64 box).

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46192 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoModified ImmutableSet/ImmutableMap to use FoldingSet profiling using
Ted Kremenek [Sat, 19 Jan 2008 04:51:55 +0000 (04:51 +0000)]
Modified ImmutableSet/ImmutableMap to use FoldingSet profiling using
FoldingSetTrait instead of directly calling a 'Profile' method.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46190 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoAdded FoldingSet style 'profiling' support for APSInt.
Ted Kremenek [Sat, 19 Jan 2008 04:31:12 +0000 (04:31 +0000)]
Added FoldingSet style 'profiling' support for APSInt.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46189 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoAdded FoldingSet style 'profiling' support for APInt.
Ted Kremenek [Sat, 19 Jan 2008 04:23:33 +0000 (04:23 +0000)]
Added FoldingSet style 'profiling' support for APInt.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46188 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoMade 'FoldingSetNodeID' a proper class instead of a nested class in
Ted Kremenek [Sat, 19 Jan 2008 04:22:50 +0000 (04:22 +0000)]
Made 'FoldingSetNodeID' a proper class instead of a nested class in
'FoldingSetNodeImpl' (previously 'FoldingSetNodeID' was a typedef of
'FoldingSetNodeImpl::NodeID').

Why?  Clients can now easily forward declare 'FoldingSetNodeID' without having
to include FoldingSet.h.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46187 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoMade 'profiling' of objects in a FoldingSet trait-based using FoldingSetTrait
Ted Kremenek [Sat, 19 Jan 2008 03:58:00 +0000 (03:58 +0000)]
Made 'profiling' of objects in a FoldingSet trait-based using FoldingSetTrait
instead of always assuming that the stored objects had a method called
'Profile'. The default behavior is to dispatch to a 'Profile' method (as
before), but via template specialization this behavior can now be overridden by
clients.

Added templated class 'FoldingSetNodeWrapper', a generic wrapper class that
allows one to insert objects into a FoldingSet that do not directly inherit from
FoldingSetNode. This is useful for inserting objects that do not always need to
pay the overhead of inheriting from FoldingSetNode, or were designed with that
behavior in mind.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46186 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoFix makefiles to enable Apply style debug build.
Evan Cheng [Fri, 18 Jan 2008 21:01:00 +0000 (21:01 +0000)]
Fix makefiles to enable Apply style debug build.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46180 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoImplement flt_rounds for PowerPC.
Dale Johannesen [Fri, 18 Jan 2008 19:55:37 +0000 (19:55 +0000)]
Implement flt_rounds for PowerPC.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46174 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoremove extraneous &&'s from tests, as Scott is apparently not going to.
Chris Lattner [Fri, 18 Jan 2008 19:53:43 +0000 (19:53 +0000)]
remove extraneous &&'s from tests, as Scott is apparently not going to.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46173 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoTest is correct again for the moment.
Dale Johannesen [Fri, 18 Jan 2008 19:53:31 +0000 (19:53 +0000)]
Test is correct again for the moment.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46172 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoremove extraneous &'s.
Chris Lattner [Fri, 18 Jan 2008 19:36:20 +0000 (19:36 +0000)]
remove extraneous &'s.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46171 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agodon't form an std::string with a null pointer, it aborts.
Chris Lattner [Fri, 18 Jan 2008 18:54:31 +0000 (18:54 +0000)]
don't form an std::string with a null pointer, it aborts.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46166 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoget symbolic information for ppc ldbl nodes.
Chris Lattner [Fri, 18 Jan 2008 18:51:16 +0000 (18:51 +0000)]
get symbolic information for ppc ldbl nodes.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46165 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoremove magic numbers.
Chris Lattner [Fri, 18 Jan 2008 17:13:03 +0000 (17:13 +0000)]
remove magic numbers.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46162 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoFix a latent bug exposed by my truncstore patch. We compiled stfiwx-2.ll to:
Chris Lattner [Fri, 18 Jan 2008 16:54:56 +0000 (16:54 +0000)]
Fix a latent bug exposed by my truncstore patch.  We compiled stfiwx-2.ll to:

_test:
fctiwz f0, f1
stfiwx f0, 0, r4
blr

instead of:

_test:
fctiwz f0, f1
stfd f0, -8(r1)
nop
nop
lwz r2, -4(r1)
stb r2, 0(r4)
blr

The former is not correct (stores 4 bytes, not 1).

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46161 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agomake a method public
Chris Lattner [Fri, 18 Jan 2008 06:52:41 +0000 (06:52 +0000)]
make a method public

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46159 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoReverted implementation of ImmutableMap::find() to return a TreeTy* instead of
Ted Kremenek [Fri, 18 Jan 2008 00:38:04 +0000 (00:38 +0000)]
Reverted implementation of ImmutableMap::find() to return a TreeTy* instead of
an iterator, since the implementation returned an iterator that pointed to a
different node! Renamed this implementation to SlimFind() so that users do not
expect it to return an iterator (it is a more efficient implementation than
returning an iterator if the user just wants to find the value of a key).

Added a FIXME to implement ImmutableMap::find() that returns an iterator.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46150 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoRevert the part of 45849 that treated weak globals
Dale Johannesen [Thu, 17 Jan 2008 23:36:04 +0000 (23:36 +0000)]
Revert the part of 45849 that treated weak globals
as weak globals rather than commons.  While not wrong,
this change tickled a latent bug in Darwin's strip,
so revert it for now as a workaround.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46147 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoRevert the part of 45848 that treated weak globals
Dale Johannesen [Thu, 17 Jan 2008 23:04:07 +0000 (23:04 +0000)]
Revert the part of 45848 that treated weak globals
as weak globals rather than commons.  While not wrong,
this change tickled a latent bug in Darwin's strip,
so revert it for now as a workaround.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46144 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoForward progress: crtbegin.c now compiles successfully!
Scott Michel [Thu, 17 Jan 2008 20:38:41 +0000 (20:38 +0000)]
Forward progress: crtbegin.c now compiles successfully!

Fixed CellSPU's A-form (local store) address mode, so that all globals,
externals, constant pool and jump table symbols are now wrapped within
a SPUISD::AFormAddr pseudo-instruction. This now identifies all local
store memory addresses, although it requires a bit of legerdemain during
instruction selection to properly select loads to and stores from local
store, properly generating "LQA" instructions.

Also added mul_ops.ll test harness for exercising integer multiplication.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46142 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoThis commit changes:
Chris Lattner [Thu, 17 Jan 2008 19:59:44 +0000 (19:59 +0000)]
This commit changes:

1. Legalize now always promotes truncstore of i1 to i8.
2. Remove patterns and gunk related to truncstore i1 from targets.
3. Rename the StoreXAction stuff to TruncStoreAction in TLI.
4. Make the TLI TruncStoreAction table a 2d table to handle from/to conversions.
5. Mark a wide variety of invalid truncstores as such in various targets, e.g.
   X86 currently doesn't support truncstore of any of its integer types.
6. Add legalize support for truncstores with invalid value input types.
7. Add a dag combine transform to turn store(truncate) into truncstore when
   safe.

The later allows us to compile CodeGen/X86/storetrunc-fp.ll to:

_foo:
fldt 20(%esp)
fldt 4(%esp)
faddp %st(1)
movl 36(%esp), %eax
fstps (%eax)
ret

instead of:

_foo:
subl $4, %esp
fldt 24(%esp)
fldt 8(%esp)
faddp %st(1)
fstps (%esp)
movl 40(%esp), %eax
movss (%esp), %xmm0
movss %xmm0, (%eax)
addl $4, %esp
ret

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46140 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agonew testcase.
Chris Lattner [Thu, 17 Jan 2008 19:47:23 +0000 (19:47 +0000)]
new testcase.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46139 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoImplemented "FIXME" in ImutAVLTree: isEqual() now also compares the *data* value
Ted Kremenek [Thu, 17 Jan 2008 17:36:49 +0000 (17:36 +0000)]
Implemented "FIXME" in ImutAVLTree: isEqual() now also compares the *data* value
and not just the key value when comparing trees. To do this we added data_type
and data_type_ref to the ImutContainerInfo trait classes. For values stored in
the tree that do not have separate key and data components, data_type is simply
a typedef of bool, and isDataEqual() always evaluates to true. This allows us to
support both ImmutableSet and ImmutableMap using the same underlying logic.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46130 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoadd some helper methods.
Chris Lattner [Thu, 17 Jan 2008 07:30:38 +0000 (07:30 +0000)]
add some helper methods.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46128 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoTest case for varargs parameter attribute issue I just fixed.
Evan Cheng [Thu, 17 Jan 2008 07:26:31 +0000 (07:26 +0000)]
Test case for varargs parameter attribute issue I just fixed.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46127 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agocode cleanups, no functionality change.
Chris Lattner [Thu, 17 Jan 2008 07:20:38 +0000 (07:20 +0000)]
code cleanups, no functionality change.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46126 91177308-0d34-0410-b5e6-96231b3b80d8

16 years ago* Introduce a new SelectionDAG::getIntPtrConstant method
Chris Lattner [Thu, 17 Jan 2008 07:00:52 +0000 (07:00 +0000)]
* Introduce a new SelectionDAG::getIntPtrConstant method
  and switch various codegen pieces and the X86 backend over
  to using it.

* Add some comments to SelectionDAGNodes.h

* Introduce a second argument to FP_ROUND, which indicates
  whether the FP_ROUND changes the value of its input. If
  not it is safe to xform things like fp_extend(fp_round(x)) -> x.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46125 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoadd testcase that has been sitting in my tree for awhile.
Chris Lattner [Thu, 17 Jan 2008 06:54:09 +0000 (06:54 +0000)]
add testcase that has been sitting in my tree for awhile.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46124 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoUpdate license for current year.
Tanya Lattner [Thu, 17 Jan 2008 05:57:59 +0000 (05:57 +0000)]
Update license for current year.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46120 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoUpdate version to 2.3svn
Tanya Lattner [Thu, 17 Jan 2008 05:57:22 +0000 (05:57 +0000)]
Update version to 2.3svn
Regenerate configure with 2.60.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46119 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoDAE bug fix. Don't lose parameter attributes on vararg arguments.
Evan Cheng [Thu, 17 Jan 2008 04:18:54 +0000 (04:18 +0000)]
DAE bug fix. Don't lose parameter attributes on vararg arguments.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46113 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoEnable CBE
Devang Patel [Thu, 17 Jan 2008 02:10:08 +0000 (02:10 +0000)]
Enable CBE

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46112 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoWhen a live virtual register is being clobbered by an implicit def, it is spilled
Evan Cheng [Thu, 17 Jan 2008 02:08:17 +0000 (02:08 +0000)]
When a live virtual register is being clobbered by an implicit def, it is spilled
and the spill is its kill. However, if the local allocator has determined the
register has not been modified (possible when its value was reloaded), it would
not issue a restore. In that case, mark the last use of the virtual register as
kill.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46111 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoFix arg promotion to propagate the correct attrs on the calls to
Chris Lattner [Thu, 17 Jan 2008 01:17:03 +0000 (01:17 +0000)]
Fix arg promotion to propagate the correct attrs on the calls to
promoted functions.  This is important for varargs calls in
particular.  Thanks to duncan for providing a great testcase.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46108 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoReplace std::vector<bool> with BitVector.
Evan Cheng [Thu, 17 Jan 2008 00:35:26 +0000 (00:35 +0000)]
Replace std::vector<bool> with BitVector.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46104 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoFixes a nasty dag combiner bug that causes a bunch of tests to fail at -O0.
Evan Cheng [Wed, 16 Jan 2008 23:11:54 +0000 (23:11 +0000)]
Fixes a nasty dag combiner bug that causes a bunch of tests to fail at -O0.

It's not safe to use the two value CombineTo variant to combine away a dead load.
e.g.
v1, chain2 = load chain1, loc
v2, chain3 = load chain2, loc
v3         = add v2, c
Now we replace use of v1 with undef, use of chain2 with chain1.
ReplaceAllUsesWith() will iterate through uses of the first load and update operands:
v1, chain2 = load chain1, loc
v2, chain3 = load chain1, loc
v3         = add v2, c
Now the second load is the same as the first load, SelectionDAG cse will ensure
the use of second load is replaced with the first load.
v1, chain2 = load chain1, loc
v3         = add v1, c
Then v1 is replaced with undef and bad things happen.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46099 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoTrampoline support for x86-64. This looks like
Duncan Sands [Wed, 16 Jan 2008 22:55:25 +0000 (22:55 +0000)]
Trampoline support for x86-64.  This looks like
it should work, but I have no machine to test
it on.  Committed because it will at least
cause no harm, and maybe someone can test it
for me!

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46098 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoHandle attribute(used) global variables that are i8.
Chris Lattner [Wed, 16 Jan 2008 21:35:43 +0000 (21:35 +0000)]
Handle attribute(used) global variables that are i8.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46090 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoDo not mark EH tables no-dead-strip unless the
Dale Johannesen [Wed, 16 Jan 2008 19:59:28 +0000 (19:59 +0000)]
Do not mark EH tables no-dead-strip unless the
associated function is so marked.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46088 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoadd testcase for regression
Chris Lattner [Wed, 16 Jan 2008 18:03:52 +0000 (18:03 +0000)]
add testcase for regression

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46073 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoFix a ppc long double regression I introduced yesterday due to a
Chris Lattner [Wed, 16 Jan 2008 17:59:31 +0000 (17:59 +0000)]
Fix a ppc long double regression I introduced yesterday due to a
simplification.  This fixes automotive-basicmath on PPC.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46072 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agomerge a few pieces of code that do the store/load to stack
Chris Lattner [Wed, 16 Jan 2008 07:51:34 +0000 (07:51 +0000)]
merge a few pieces of code that do the store/load to stack
pattern to use EmitStackConvert now.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46066 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agorename ExpandBIT_CONVERT to EmitStackConvert, generalizing
Chris Lattner [Wed, 16 Jan 2008 07:45:30 +0000 (07:45 +0000)]
rename ExpandBIT_CONVERT to EmitStackConvert, generalizing
it to allow it to emit different load and store kinds.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46065 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agosimplify a bunch of code by using SelectionDAG::CreateStackTemporary
Chris Lattner [Wed, 16 Jan 2008 07:03:22 +0000 (07:03 +0000)]
simplify a bunch of code by using SelectionDAG::CreateStackTemporary
instead of inlining its body.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46062 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoChange legalizeop of FP_ROUND and FP_EXTEND to not fall through
Chris Lattner [Wed, 16 Jan 2008 06:57:07 +0000 (06:57 +0000)]
Change legalizeop of FP_ROUND and FP_EXTEND  to not fall through
into the ANY_EXTEND/ZERO_EXTEND/SIGN_EXTEND code to simplify it.

Unmerge the code for FP_ROUND and FP_EXTEND from each other to
make each one simpler.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46061 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agomake sure to use a cpu that has sse.
Chris Lattner [Wed, 16 Jan 2008 06:32:02 +0000 (06:32 +0000)]
make sure to use a cpu that has sse.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46060 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agomake it more clear that this predicate only applies to scalar FP types.
Chris Lattner [Wed, 16 Jan 2008 06:24:21 +0000 (06:24 +0000)]
make it more clear that this predicate only applies to scalar FP types.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46058 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agointroduce a isTypeInSSEReg predicate, which allows us to simplify
Chris Lattner [Wed, 16 Jan 2008 06:19:45 +0000 (06:19 +0000)]
introduce a isTypeInSSEReg predicate, which allows us to simplify
some code.  No functionality change.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46055 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoMy previous commit had an incomplete message, it should have been:
Chris Lattner [Wed, 16 Jan 2008 05:56:59 +0000 (05:56 +0000)]
My previous commit had an incomplete message, it should have been:

make the 'fp return in ST(0)' optimization smart enough to
look through token factor nodes.  THis allows us to compile
testcases like CodeGen/X86/fp-stack-retcopy.ll into:

_carg:
subl $12, %esp
call L_foo$stub
fstpl (%esp)
fldl (%esp)
addl $12, %esp
ret

instead of:

_carg:
subl $28, %esp
call L_foo$stub
fstpl 16(%esp)
movsd 16(%esp), %xmm0
movsd %xmm0, 8(%esp)
fldl 8(%esp)
addl $28, %esp
ret

Still not optimal, but much better and this is a trivial patch.  Fixing
the rest requires invasive surgery that is is not llvm 2.2 material.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46054 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agomake the 'fp return in ST(0)' optimization smart enough to
Chris Lattner [Wed, 16 Jan 2008 05:53:06 +0000 (05:53 +0000)]
make the 'fp return in ST(0)' optimization smart enough to
look through token factor

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46053 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agovarious whitespace cleanups, no functionality change.
Chris Lattner [Wed, 16 Jan 2008 05:52:18 +0000 (05:52 +0000)]
various whitespace cleanups, no functionality change.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46052 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoFactor the ReachesChainWithoutSideEffects out of dag combiner into
Chris Lattner [Wed, 16 Jan 2008 05:49:24 +0000 (05:49 +0000)]
Factor the ReachesChainWithoutSideEffects out of dag combiner into
a public SDOperand::reachesChainWithoutSideEffects method.  No
functionality change.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46050 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoDo not strip llvm.used values.
Devang Patel [Wed, 16 Jan 2008 03:33:05 +0000 (03:33 +0000)]
Do not strip llvm.used values.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46045 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoChanged ImmutableMap::find to return an iterator instead of a pointer
Ted Kremenek [Tue, 15 Jan 2008 23:53:53 +0000 (23:53 +0000)]
Changed ImmutableMap::find to return an iterator instead of a pointer
to the tree node.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46034 91177308-0d34-0410-b5e6-96231b3b80d8

16 years ago- Introduces versioning macro LLVM_LTO_VERSION
Devang Patel [Tue, 15 Jan 2008 23:52:34 +0000 (23:52 +0000)]
- Introduces versioning macro LLVM_LTO_VERSION
- Communicate symbol visibility
- Communicate code generation model

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46033 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoCommit a piece that I missed before, patch by Alain Frisch
Chris Lattner [Tue, 15 Jan 2008 23:27:40 +0000 (23:27 +0000)]
Commit a piece that I missed before, patch by Alain Frisch

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46032 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoMissed file from previous checkin.
Dale Johannesen [Tue, 15 Jan 2008 23:25:27 +0000 (23:25 +0000)]
Missed file from previous checkin.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46030 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoFix and enable EH for x86-64 Darwin. Adds
Dale Johannesen [Tue, 15 Jan 2008 23:24:56 +0000 (23:24 +0000)]
Fix and enable EH for x86-64 Darwin.  Adds
ShortenEHDataFor64Bits as a not-very-accurate
abstraction to cover all the changes in DwarfWriter.
Some cosmetic changes to Darwin assembly code for
gcc testsuite compatibility.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46029 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoMove some calls to getVRegDef higher in the callgraph, so they don't get executed...
Owen Anderson [Tue, 15 Jan 2008 22:58:11 +0000 (22:58 +0000)]
Move some calls to getVRegDef higher in the callgraph, so they don't get executed as frequently in performance sensitive code.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46027 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoimprove compatibility with mingw, patch by Alain Frisch
Chris Lattner [Tue, 15 Jan 2008 22:50:50 +0000 (22:50 +0000)]
improve compatibility with mingw, patch by Alain Frisch

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46026 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoadd a test to ensure that argpromote of one argument doesn't
Chris Lattner [Tue, 15 Jan 2008 22:38:12 +0000 (22:38 +0000)]
add a test to ensure that argpromote of one argument doesn't
break the byval attr on some other argument.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46025 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoAdd description of 'llvm.trap' intrinsic. Also, minor fixes in formatting.
Anton Korobeynikov [Tue, 15 Jan 2008 22:31:34 +0000 (22:31 +0000)]
Add description of 'llvm.trap' intrinsic. Also, minor fixes in formatting.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46024 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoverify x86 generates ud2 for llvm.trap
Chris Lattner [Tue, 15 Jan 2008 22:22:02 +0000 (22:22 +0000)]
verify x86 generates ud2 for llvm.trap

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46023 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoUnbreak the build.
Evan Cheng [Tue, 15 Jan 2008 22:19:35 +0000 (22:19 +0000)]
Unbreak the build.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46021 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agonew testcase for llvm.trap.
Chris Lattner [Tue, 15 Jan 2008 22:17:26 +0000 (22:17 +0000)]
new testcase for llvm.trap.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46020 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoIf someone wants to implement ppc TRAP, they can go for it :)
Chris Lattner [Tue, 15 Jan 2008 22:15:02 +0000 (22:15 +0000)]
If someone wants to implement ppc TRAP, they can go for it :)

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46019 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoThe type of the 'abort' node should be pointer type (because
Chris Lattner [Tue, 15 Jan 2008 22:09:33 +0000 (22:09 +0000)]
The type of the 'abort' node should be pointer type (because
it's a function pointer) not MVT::Other.  This fixes builtin_trap
lowering on ppc, alpha, ia64

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46018 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agorename SDTRet -> SDTNone.
Chris Lattner [Tue, 15 Jan 2008 22:02:54 +0000 (22:02 +0000)]
rename SDTRet -> SDTNone.
Move definition of 'trap' sdnode up from x86 instrinfo to targetselectiondag.td.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46017 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoRemove DefInst from LiveVariables::VarInfo. Use the facilities on MachineRegisterInf...
Owen Anderson [Tue, 15 Jan 2008 22:02:46 +0000 (22:02 +0000)]
Remove DefInst from LiveVariables::VarInfo.  Use the facilities on MachineRegisterInfo instead.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46016 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agono need to expand ISD::TRAP to X86ISD::TRAP, just match ISD::TRAP.
Chris Lattner [Tue, 15 Jan 2008 21:58:22 +0000 (21:58 +0000)]
no need to expand ISD::TRAP to X86ISD::TRAP, just match ISD::TRAP.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46015 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoAdd support for targets that have a legal ISD::TRAP.
Chris Lattner [Tue, 15 Jan 2008 21:58:08 +0000 (21:58 +0000)]
Add support for targets that have a legal ISD::TRAP.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46014 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoAdd files to windows project files. Also include <algorithm> explicitly so that...
Chuck Rose III [Tue, 15 Jan 2008 21:43:17 +0000 (21:43 +0000)]
Add files to windows project files.  Also include <algorithm> explicitly so that vstudio build works

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46013 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoFix JIT encoding of trap/ud2 instruction
Anton Korobeynikov [Tue, 15 Jan 2008 21:40:02 +0000 (21:40 +0000)]
Fix JIT encoding of trap/ud2 instruction

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46012 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoReformatted. It was confusing the other way. No functionality change.
Bill Wendling [Tue, 15 Jan 2008 21:16:32 +0000 (21:16 +0000)]
Reformatted. It was confusing the other way. No functionality change.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46009 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoFix a memory correctness error noticed by valgrind (harmless in practice).
Gordon Henriksen [Tue, 15 Jan 2008 20:02:11 +0000 (20:02 +0000)]
Fix a memory correctness error noticed by valgrind (harmless in practice).

Thanks to Duncan Sands for noticing it.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46007 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoTestcase for gimplify_expr crash caused by an
Duncan Sands [Tue, 15 Jan 2008 19:55:41 +0000 (19:55 +0000)]
Testcase for gimplify_expr crash caused by an
unexpected placeholder_expr.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46006 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoOops. Forgot to commit this.
Evan Cheng [Tue, 15 Jan 2008 07:49:54 +0000 (07:49 +0000)]
Oops. Forgot to commit this.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46003 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoOops. Forgot to commit this.
Evan Cheng [Tue, 15 Jan 2008 07:49:36 +0000 (07:49 +0000)]
Oops. Forgot to commit this.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46002 91177308-0d34-0410-b5e6-96231b3b80d8

16 years agoFor PR1839: add initial support for __builtin_trap. llvm-gcc part is missed
Anton Korobeynikov [Tue, 15 Jan 2008 07:02:33 +0000 (07:02 +0000)]
For PR1839: add initial support for __builtin_trap. llvm-gcc part is missed
as well as PPC codegen

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46001 91177308-0d34-0410-b5e6-96231b3b80d8