Chris Lattner [Tue, 30 Mar 2010 04:58:26 +0000 (04:58 +0000)]
if a timergroup is destroyed before its timers, print times.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99873
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Tue, 30 Mar 2010 04:40:01 +0000 (04:40 +0000)]
change TimerGroup to keep a linked list of active timers
instead of just a count of them, and refactor the guts of
report printing out of removeTimer into its own method.
Refactor addTimerToPrint away.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99872
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Tue, 30 Mar 2010 04:03:22 +0000 (04:03 +0000)]
reapply my timer rewrite with a change for PassManager to store
timers by pointer instead of by-value.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99871
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Tue, 30 Mar 2010 03:57:00 +0000 (03:57 +0000)]
revert r99862 which is causing FNT failures.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99870
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Tue, 30 Mar 2010 02:38:19 +0000 (02:38 +0000)]
fairly major rewrite of various timing related stuff.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99862
91177308-0d34-0410-b5e6-
96231b3b80d8
Eric Christopher [Tue, 30 Mar 2010 01:04:59 +0000 (01:04 +0000)]
Add FIXME for operand promotion.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99859
91177308-0d34-0410-b5e6-
96231b3b80d8
Jakob Stoklund Olesen [Tue, 30 Mar 2010 00:09:32 +0000 (00:09 +0000)]
Be gentle to MSVC. C++ is hard, after all.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99855
91177308-0d34-0410-b5e6-
96231b3b80d8
Daniel Dunbar [Mon, 29 Mar 2010 23:56:40 +0000 (23:56 +0000)]
MC/Mach-O/x86_64: Support @GOTPCREL on symbols, even for non-PCrel relocations!
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99853
91177308-0d34-0410-b5e6-
96231b3b80d8
Bill Wendling [Mon, 29 Mar 2010 23:37:07 +0000 (23:37 +0000)]
Re-add back in the slow way of determining of a clean-up should become a
catch-all. The "dominates" way won't catch all of the selectors which must be
changed.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99850
91177308-0d34-0410-b5e6-
96231b3b80d8
Jakob Stoklund Olesen [Mon, 29 Mar 2010 23:24:21 +0000 (23:24 +0000)]
Basic implementation of SSEDomainFix pass.
Cross-block inference is primitive and wrong, but the pass is working otherwise.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99848
91177308-0d34-0410-b5e6-
96231b3b80d8
Bill Wendling [Mon, 29 Mar 2010 23:02:46 +0000 (23:02 +0000)]
A more general (and simpler!) implementation of r99671. It performs a similar
transform. I.e., if a clean-up eh.selector call dominates the invoke of an
_Unwind_Resume_or_Rethrow, then we convert the eh.selector into a
catch-all. This patch, however, uses the DominatorTree information, and doesn't
go through the whole rigmarole of starting at the eh.exception call, finding the
corresponding URoR and eh.selector calls, and trying to trace through any number
of instruction types to get to them.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99846
91177308-0d34-0410-b5e6-
96231b3b80d8
Devang Patel [Mon, 29 Mar 2010 22:59:58 +0000 (22:59 +0000)]
Encode start location of debug value, communicated through DBG_VALUE machine instruction, in a variable's DIE.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99845
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Mon, 29 Mar 2010 21:34:06 +0000 (21:34 +0000)]
move a function into a more logical place in the file
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99842
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Mon, 29 Mar 2010 21:28:41 +0000 (21:28 +0000)]
remove support for per-time peak memory tracking, this
isn't used by anyone and is better exposed as a non-per-timer
thing. Also, stop including System/Mutex.h in Timer.h
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99841
91177308-0d34-0410-b5e6-
96231b3b80d8
Evan Cheng [Mon, 29 Mar 2010 21:27:30 +0000 (21:27 +0000)]
Fix PR4975. Avoid referencing empty vector.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99840
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Mon, 29 Mar 2010 21:24:52 +0000 (21:24 +0000)]
various timer fixes: move operator= out of line,
eliminate the per-timer lock (timers should be
externally locked if needed), the info-output-stream
can never be dbgs(), so drop the check. Make some
stuff private.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99839
91177308-0d34-0410-b5e6-
96231b3b80d8
Benjamin Kramer [Mon, 29 Mar 2010 21:13:41 +0000 (21:13 +0000)]
Make isInt?? and isUint?? template specializations of the generic versions. This
makes calls a little bit more consistent and allows easy removal of the
specializations in the future. Convert all callers to the templated functions.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99838
91177308-0d34-0410-b5e6-
96231b3b80d8
Evan Cheng [Mon, 29 Mar 2010 20:48:30 +0000 (20:48 +0000)]
Pool allocate SDDbgValue nodes.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99836
91177308-0d34-0410-b5e6-
96231b3b80d8
Eric Christopher [Mon, 29 Mar 2010 20:41:51 +0000 (20:41 +0000)]
We'll never match these as instructions, just as intrinsics so remove
the SDNodes.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99835
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Mon, 29 Mar 2010 20:40:19 +0000 (20:40 +0000)]
s/.../.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99834
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Mon, 29 Mar 2010 20:39:38 +0000 (20:39 +0000)]
use RAII for ExceptionTimer too
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99833
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Mon, 29 Mar 2010 20:38:20 +0000 (20:38 +0000)]
fix a variety of issues were we'd start DebugTimer but
not stop it by using RAII.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99832
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Mon, 29 Mar 2010 20:35:01 +0000 (20:35 +0000)]
move code around and improve indentation, no functionality change.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99831
91177308-0d34-0410-b5e6-
96231b3b80d8
Benjamin Kramer [Mon, 29 Mar 2010 19:07:58 +0000 (19:07 +0000)]
Remove a bunch of integer width predicate functions in favor of MathExtras.
Most of these were unused, some of them were wrong and unused (isS16Constant<short>,
isS10Constant<short>).
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99827
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Mon, 29 Mar 2010 18:42:42 +0000 (18:42 +0000)]
some other notes.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99826
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Mon, 29 Mar 2010 18:39:28 +0000 (18:39 +0000)]
apparently llvm 2.7 now has great armv4 support, from
John Tytgat
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99825
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Mon, 29 Mar 2010 18:34:13 +0000 (18:34 +0000)]
update the vmkit blurb
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99824
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Mon, 29 Mar 2010 17:50:39 +0000 (17:50 +0000)]
add tce
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99820
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Mon, 29 Mar 2010 17:38:47 +0000 (17:38 +0000)]
From Kalle Raiskila:
"the bigstack patch for SPU, with testcase. It is essentially the patch committed as 97091, and reverted as 97099, but with the following additions:
-in vararg handling, registers are marked to be live, to not confuse the register scavenger
-function prologue and epilogue are not emitted, if the stack size is 16. 16 means it is empty - there is only the register scavenger emergency spill slot, which is not used as there is no stack."
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99819
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Mon, 29 Mar 2010 17:36:02 +0000 (17:36 +0000)]
add support for zero initialized unions, patch by Tim Northover!
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99818
91177308-0d34-0410-b5e6-
96231b3b80d8
Devang Patel [Mon, 29 Mar 2010 17:20:31 +0000 (17:20 +0000)]
Refactor code to push DILocation prcessing into DwarfDebug.cpp from AsmPrinter.cpp.
This is same as r99772 (which was reverted) with just one meaningful difference where two source lines exchanged their positions.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99816
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Mon, 29 Mar 2010 17:02:02 +0000 (17:02 +0000)]
add a note.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99815
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Mon, 29 Mar 2010 02:02:45 +0000 (02:02 +0000)]
Switch pattern sorting predicate from stable sort -> sort, it
doesn't need to be stable because the patterns are fully ordered.
Add a first level sort predicate that orders patterns in this
order: 1) scalar integer operations 2) scalar floating point
3) vector int 4) vector float. This is a trivial sort on their
top level pattern type so it is nice and transitive. The
benefit of doing this is that simple integer operations are
much more common than insane vector things and isel was trying
to match the big complex vector patterns before the simple
ones because the complexity of the vector operations was much
higher. Since they can't both match, it is best (for compile
time) to try the simple integer ones first.
This cuts down the # failed match attempts on real code by
quite a bit, for example, this reduces backtracks on crafty
(as a random example) from 228285 -> 188369.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99797
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Mon, 29 Mar 2010 01:58:15 +0000 (01:58 +0000)]
revert 99795, as mentioned, it is disabled anyway.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99796
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Mon, 29 Mar 2010 01:56:19 +0000 (01:56 +0000)]
Check in a (disabled) failed attempt to improve the ordering of
patterns within the generated matcher. This works great except
that the sort fails because the relation defined isn't
transitive. I have a much simpler solution coming next, but want
to archive the code.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99795
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Mon, 29 Mar 2010 01:40:38 +0000 (01:40 +0000)]
print the complexity of the pattern being matched in the
comment in the generated table.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99794
91177308-0d34-0410-b5e6-
96231b3b80d8
Johnny Chen [Mon, 29 Mar 2010 01:14:22 +0000 (01:14 +0000)]
Add NVTBLFrm to represent A8.6.406 VTBL, VTBX Vector Table Lookup Instructions.
These instructions use byte index in a control vector (M:Vm) to lookup byte
values in a table and generate a new vector (D:Vd). The table is specified via
a list of vectors, which can be:
{Dn}
{Dn D<n+1>}
{Dn D<n+1> D<n+2>}
{Dn D<n+1> D<n+2> D<n+3>}
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99789
91177308-0d34-0410-b5e6-
96231b3b80d8
Devang Patel [Sun, 28 Mar 2010 21:23:37 +0000 (21:23 +0000)]
Revert 99772.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99778
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sun, 28 Mar 2010 19:46:56 +0000 (19:46 +0000)]
add a statistic for the # times isel has to backtrack.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99774
91177308-0d34-0410-b5e6-
96231b3b80d8
Devang Patel [Sun, 28 Mar 2010 18:57:09 +0000 (18:57 +0000)]
Refactoring. Push DILocation processing in to DwarfDebug from AsmPrinter.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99772
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sun, 28 Mar 2010 18:52:28 +0000 (18:52 +0000)]
zap an extra line that Eli noticed!
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99770
91177308-0d34-0410-b5e6-
96231b3b80d8
Anton Korobeynikov [Sun, 28 Mar 2010 15:07:02 +0000 (15:07 +0000)]
Properly quote the quotes :) during cmdline construction on Windows.
Otherwise, e.g. in the invocation like clang -DFOO=\"bar\" FOO macro
got the bar value, not "bar".
Patch by Alexander Esilevich!
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99763
91177308-0d34-0410-b5e6-
96231b3b80d8
Torok Edwin [Sun, 28 Mar 2010 11:07:36 +0000 (11:07 +0000)]
Fix use-of-uninitialized value when RWX memory can't be allocated (PR6701).
SELinux doesn't allow 'execmem', returning MAP_FAILED and 'Permission denied'
for mmap or RWX memory. In this case AllocateRWX was returning a MemoryBlock
with uninitialized fields, which sometimes caused crashes.
This patch initializes MemoryBlock fields to 0, so that the RWX-failure check
works.
It doesn't fix the SELinux 'execmem' issues though (the JIT will not work when
SELinux is in enforcing mode).
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99762
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sun, 28 Mar 2010 08:48:47 +0000 (08:48 +0000)]
improve type checking of SDNode operand count. This rejects all cases
where an incorrect number of operands is provided to an sdnode instead
of just a few cases.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99761
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sun, 28 Mar 2010 08:47:39 +0000 (08:47 +0000)]
fix a type contradition: XCoreISD::RETSP has one argument, not zero.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99760
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sun, 28 Mar 2010 08:43:23 +0000 (08:43 +0000)]
finally remove the immAllOnesV_bc/immAllZerosV_bc patterns
and those derived from them. These are obnoxious because
they were written as: PatLeaf<(bitconvert). Not having an
argument was foiling adding better type checking for operand
count matching up with what was required (in this case,
bitconvert always requires an operand!)
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99759
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sun, 28 Mar 2010 08:40:48 +0000 (08:40 +0000)]
remove a pattern with no testcase that doesn't appear to be
matchable: it seems like it would always constant fold.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99758
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sun, 28 Mar 2010 08:39:10 +0000 (08:39 +0000)]
fix integer negates to use the proper type for the zero vectors,
this also depends on the new "bitconvert dropping" behavior just
added to tblgen.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99757
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sun, 28 Mar 2010 08:38:32 +0000 (08:38 +0000)]
teach tblgen to allow patterns like (add (i32 (bitconvert (i32 GPR))), 4),
transforming it into (add (i32 GPR), 4). This allows us to write type
generic multi patterns and have tblgen automatically drop the bitconvert
in the case when the types align. This allows us to fold an extra load
in the changed testcase.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99756
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sun, 28 Mar 2010 08:36:45 +0000 (08:36 +0000)]
fix a typo, bitconvert from node to itself isn't valid.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99755
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sun, 28 Mar 2010 08:08:07 +0000 (08:08 +0000)]
fix vnot matching to explicitly specify the type of the
input to be v8i8 or v16i8, which buildvectors get canonicalized to.
This allows the patterns that were previously using a bare 'vnot' to
match, before they couldn't.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99754
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sun, 28 Mar 2010 08:00:23 +0000 (08:00 +0000)]
fix up vnot matching, eliminating a dead pattern, correcting a couple of
patterns that would never match because of bitcast, and eliminating use
of vnot_conv.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99753
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sun, 28 Mar 2010 07:58:37 +0000 (07:58 +0000)]
add some nounwinds
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99752
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sun, 28 Mar 2010 07:58:09 +0000 (07:58 +0000)]
this takes an insane amount of time to run, disable it for now (PR6727)
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99751
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sun, 28 Mar 2010 07:48:17 +0000 (07:48 +0000)]
stop using vnot_conv
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99750
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sun, 28 Mar 2010 07:38:39 +0000 (07:38 +0000)]
revert r99743, this is saying that the repmovs instructinos have an
*input* of other type, which is the VT.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99749
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sun, 28 Mar 2010 07:38:00 +0000 (07:38 +0000)]
remove a bunch of dead patterns.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99748
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sun, 28 Mar 2010 06:57:56 +0000 (06:57 +0000)]
validate that input sdnodes don't exist in an output pattern.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99747
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sun, 28 Mar 2010 06:50:34 +0000 (06:50 +0000)]
eliminate a bunch of code duplication in ParseTreePattern
by rotating it.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99746
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sun, 28 Mar 2010 06:16:41 +0000 (06:16 +0000)]
add some node definitions.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99745
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sun, 28 Mar 2010 06:04:39 +0000 (06:04 +0000)]
SDTCisVT<0, isVoid> is not valid, reject it.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99744
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sun, 28 Mar 2010 05:57:36 +0000 (05:57 +0000)]
claiming to return other is pointless.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99743
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sun, 28 Mar 2010 05:55:17 +0000 (05:55 +0000)]
comply with the wishes of a fixme.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99742
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sun, 28 Mar 2010 05:54:03 +0000 (05:54 +0000)]
now that (parallel) is gone and a variety of bugs in targets
are cleaned up, we can remove an old fixme.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99741
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sun, 28 Mar 2010 05:50:16 +0000 (05:50 +0000)]
add an optimized form of OPC_EmitMergeInputChains for the 1, 0 and
1, 1 cases which are by-far the most frequent. This shrinks the X86
isel table from 77014 -> 74657 bytes.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99740
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sun, 28 Mar 2010 05:28:31 +0000 (05:28 +0000)]
don't add nodes to the now-dead nodes list multiple times, this
can cause a crash on crazy situations in msp430 when morph-node-to
is disabled.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99739
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sun, 28 Mar 2010 05:21:52 +0000 (05:21 +0000)]
Improve systemz to model cmp and ucmp nodes as returning
their flags correctly.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99738
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sun, 28 Mar 2010 05:12:57 +0000 (05:12 +0000)]
the FPCmp node returns an i32.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99737
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sun, 28 Mar 2010 05:07:17 +0000 (05:07 +0000)]
fix some modelling problems exposed by a patch I'm working on. bsr/bsf/ptest
nodes all have an EFLAGS result when made by isel lowering.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99736
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sun, 28 Mar 2010 04:54:33 +0000 (04:54 +0000)]
don't add flag nodes with chain results to the NowDeadNodes
list multiple times when MorphNodeTo can't be applied.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99735
91177308-0d34-0410-b5e6-
96231b3b80d8
Jeffrey Yasskin [Sun, 28 Mar 2010 01:03:47 +0000 (01:03 +0000)]
Oops, r98447 dropped the reference to ForwardType in the wrong place in
Type::destroy(), so it got skipped for FunctionTypes, StructTypes, and
UnionTypes. This fixes the resulting leaks in test/Feature/opaquetypes.ll and
test/Integer/opaquetypes_bt.ll.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99732
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sat, 27 Mar 2010 20:45:15 +0000 (20:45 +0000)]
fix a long standing fixme, which required fixing a bunch of other
issues to get here. We now trim the result type list of the
CompleteMatch or MorphNodeTo operation to be the same size as the
thing we're matching. this means that if you match (add GPR, GPR)
with an instruction that produces a normal result and a flag that
we now trim the result in tblgen instead of having to do it
dynamically. This exposed a bunch of inconsistencies in result
counting that happened to be getting lucky since the days of the
old isel.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99728
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sat, 27 Mar 2010 20:32:26 +0000 (20:32 +0000)]
fix CodeGenTarget::getRegisterVTs to not return the
same vt multiple times for a register. For example,
ECX is in 5 different i32 reg classes, just return
1 i32 instead of 5.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99727
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sat, 27 Mar 2010 20:09:24 +0000 (20:09 +0000)]
hoist some funky logic into CodeGenInstruction
from two places in CodeGenDAGPatterns.cpp, and
use it in DAGISelMatcherGen.cpp instead of using
an incorrect predicate that happened to get lucky
on our current targets.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99726
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sat, 27 Mar 2010 19:15:02 +0000 (19:15 +0000)]
continue pushing tblgen's support for nodes with multiple
results forward. We can now handle an instruction that
produces one implicit def and one result instead of one or
the other when not at the root of the pattern.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99725
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sat, 27 Mar 2010 18:54:50 +0000 (18:54 +0000)]
improve -debug-only=isel comments for cases when we don't enter a
scope due to obviously false predicate.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99723
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sat, 27 Mar 2010 18:49:33 +0000 (18:49 +0000)]
fix a bug in my recent patch that increased opcode size to 2 bytes:
the index comments nested under OPC_SwitchOpcode were off by one.
This fixes the comments.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99722
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sat, 27 Mar 2010 18:01:10 +0000 (18:01 +0000)]
revert r99719 which is breaking the botz.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99721
91177308-0d34-0410-b5e6-
96231b3b80d8
Dan Gohman [Sat, 27 Mar 2010 16:49:51 +0000 (16:49 +0000)]
Make llvm-ld remove its output files in the event of an error.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99719
91177308-0d34-0410-b5e6-
96231b3b80d8
Dan Gohman [Sat, 27 Mar 2010 16:39:56 +0000 (16:39 +0000)]
Give FileRemover a default constructor to allow FileRemovers to be created
and initialized separately.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99717
91177308-0d34-0410-b5e6-
96231b3b80d8
Dan Gohman [Sat, 27 Mar 2010 16:36:08 +0000 (16:36 +0000)]
No need to check the same condition twice.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99716
91177308-0d34-0410-b5e6-
96231b3b80d8
Dan Gohman [Sat, 27 Mar 2010 16:34:32 +0000 (16:34 +0000)]
Add a comment.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99715
91177308-0d34-0410-b5e6-
96231b3b80d8
Dan Gohman [Sat, 27 Mar 2010 16:33:36 +0000 (16:33 +0000)]
Remove an unnecessary #include.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99714
91177308-0d34-0410-b5e6-
96231b3b80d8
Jeffrey Yasskin [Sat, 27 Mar 2010 09:09:17 +0000 (09:09 +0000)]
Remove another memory leak from ABCD by using Edges by value instead of
pointer. There was also a SmallPtrSet whose settiness wasn't being used, so I
changed it to a SmallVector.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99713
91177308-0d34-0410-b5e6-
96231b3b80d8
Jeffrey Yasskin [Sat, 27 Mar 2010 08:15:46 +0000 (08:15 +0000)]
In ABCD, change the non-null Bound*s to Bound&s.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99711
91177308-0d34-0410-b5e6-
96231b3b80d8
Jeffrey Yasskin [Sat, 27 Mar 2010 08:09:24 +0000 (08:09 +0000)]
Fix a memory leak in ABCD by giving ownership of Bound objects to the
MemoizedResultChart.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99710
91177308-0d34-0410-b5e6-
96231b3b80d8
Jeffrey Yasskin [Sat, 27 Mar 2010 04:59:47 +0000 (04:59 +0000)]
XFAIL a new tblgen test for memory leak checking.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99707
91177308-0d34-0410-b5e6-
96231b3b80d8
Jeffrey Yasskin [Sat, 27 Mar 2010 04:53:56 +0000 (04:53 +0000)]
Avoid leaking the memory allocated for GlobalVariables in the interpreter, by
freeing that memory when the GV is destroyed.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99706
91177308-0d34-0410-b5e6-
96231b3b80d8
Bob Wilson [Sat, 27 Mar 2010 04:01:23 +0000 (04:01 +0000)]
Fix indentation.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99705
91177308-0d34-0410-b5e6-
96231b3b80d8
Bob Wilson [Sat, 27 Mar 2010 03:56:52 +0000 (03:56 +0000)]
Add a format argument to the N3V and N3VX classes, removing the N3Vf class.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99704
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sat, 27 Mar 2010 02:53:27 +0000 (02:53 +0000)]
remove parallel support.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99703
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sat, 27 Mar 2010 02:47:14 +0000 (02:47 +0000)]
eliminate the last of the parallel's!
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99700
91177308-0d34-0410-b5e6-
96231b3b80d8
Eric Christopher [Sat, 27 Mar 2010 01:54:00 +0000 (01:54 +0000)]
When we promote a load of an argument make sure to take the alignment
of the previous load - it's usually important. For example, we don't want
to blindly turn an unaligned load into an aligned one.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99699
91177308-0d34-0410-b5e6-
96231b3b80d8
Bill Wendling [Sat, 27 Mar 2010 01:24:30 +0000 (01:24 +0000)]
Forgot the part where we handle the ".llvm.eh.catch.all.value".
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99697
91177308-0d34-0410-b5e6-
96231b3b80d8
Bill Wendling [Sat, 27 Mar 2010 01:22:38 +0000 (01:22 +0000)]
Return if we changed anything or not.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99695
91177308-0d34-0410-b5e6-
96231b3b80d8
Bill Wendling [Sat, 27 Mar 2010 01:19:12 +0000 (01:19 +0000)]
If a selector has a call to ".llvm.eh.catch.all.value" that we haven't
converted, then use the initializer, since using the name itself won't work.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99692
91177308-0d34-0410-b5e6-
96231b3b80d8
Johnny Chen [Sat, 27 Mar 2010 01:03:13 +0000 (01:03 +0000)]
Add NVMulSLFrm to represent "3-register multiply with scalar" operations and set
it as the format for the appropriate N3V*SL*<> classes. These instructions
require special handling of the M:Vm field which encodes the restricted Dm and
the lane index within Dm.
Examples are A8.6.325 VMLA, VMLAL, VMLS, VMLSL (by scalar):
vmlal.s32 q3, d2, d10[0]
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99690
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sat, 27 Mar 2010 00:45:04 +0000 (00:45 +0000)]
eliminate almost all the rest of the x86-32 parallels.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99686
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sat, 27 Mar 2010 00:29:41 +0000 (00:29 +0000)]
bump FIRST_TARGET_MEMORY_OPCODE back up, there is no reason to make it a tight bound anymore.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99685
91177308-0d34-0410-b5e6-
96231b3b80d8
Jim Grosbach [Sat, 27 Mar 2010 00:09:12 +0000 (00:09 +0000)]
Thumb2 storeFrom/LoadToStackSlot() need to handle tGPR regs directly, not pass
through to the generic version. The generic functions use STR/LDR, but T2
needs the t2STR/t2LDR instead so we get the addressing mode correct.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99678
91177308-0d34-0410-b5e6-
96231b3b80d8