Commit Graph

78057 Commits

Author SHA1 Message Date
Evan Cheng
41e00172c5 At -O0, multiple uses of a virtual registers in the same BB are being marked
"kill". This looks like a bug upstream. Since that's going to take some time
to understand, loosen the assertion and disable the optimization when
multiple kills are seen.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144568 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-14 21:02:09 +00:00
Nick Lewycky
4d0a9ff365 Add support for tsan annotations (thread sanitizer, a valgrind-based tool).
These annotations are disabled entirely when either ENABLE_THREADS is off, or
building a release build. When enabled, they add calls to functions with no
statements to ManagedStatic's getters.

Use these annotations to inform tsan that the race used inside ManagedStatic
initialization is actually benign. Thanks to Kostya Serebryany for helping
write this patch!


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144567 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-14 20:50:16 +00:00
Evan Cheng
76c8f08567 Add a missing pattern for X86ISD::MOVLPD. rdar://10436044
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144566 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-14 20:35:52 +00:00
Chad Rosier
57b2997966 Add support for Thumb load/stores with negative offsets.
rdar://10412592



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144565 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-14 20:22:27 +00:00
Benjamin Kramer
6296ee3ee2 Unbreak Release builds.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144560 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-14 19:51:48 +00:00
Evan Cheng
2a4410df44 Teach two-address pass to re-schedule two-address instructions (or the kill
instructions of the two-address operands) in order to avoid inserting copies.
This fixes the few regressions introduced when the two-address hack was
disabled (without regressing the improvements).
rdar://10422688


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144559 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-14 19:48:55 +00:00
Pete Cooper
a77214a4c4 Changed SSE4/AVX <2 x i64> extract and insert ops to be Custom lowered
Constant idx case is still done in tablegen but other cases are then expanded

Fixes <rdar://problem/10435460>


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144557 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-14 19:38:42 +00:00
Benjamin Kramer
b518cae015 Fold ConstantVector::isAllOnesValue into Constant::isAllOnesValue and simplify it.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144555 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-14 19:12:20 +00:00
Akira Hatanaka
788dc0f4e5 32-to-64-bit extended load.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144554 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-14 19:06:14 +00:00
Akira Hatanaka
4961709688 AnalyzeCallOperands function for N32/64.
N32/64 places all variable arguments in integer registers (or on stack),
regardless of their types, but follows calling convention of non-vaarg function
when it handles fixed arguments.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144553 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-14 19:02:54 +00:00
Akira Hatanaka
bad53f41c2 Modify LowerFormalArguments to correctly handle vaarg arguments for Mips64.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144552 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-14 19:01:09 +00:00
Justin Holewinski
a3f7e229b6 PTX: Let LLVM use loads/stores for all mem* intrinsics, instead of relying on custom implementations.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144551 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-14 18:58:20 +00:00
Wesley Peck
3ff16dbe2c Add release notes for the MicroBlaze backend.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144550 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-14 18:56:41 +00:00
Akira Hatanaka
47a4ab84fe Remove variable that keeps the size of area used to save byval or variable
argument registers on the callee's stack frame, along with functions that set
and get it.
    
It is not necessary to add the size of this area when computing stack size in
emitPrologue, since it has already been accounted for in
PEI::calculateFrameObjectOffsets.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144549 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-14 18:56:20 +00:00
Jakob Stoklund Olesen
f054e19819 Fix early-clobber handling in shrinkToUses.
I broke this in r144515, it affected most ARM testers.

<rdar://problem/10441389>

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144547 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-14 18:45:38 +00:00
Bob Wilson
96b685b4aa Disable generation of compact unwind encodings. <rdar://problem/10441578>
This still seems to be causing some failures.  It needs more testing before
it gets enabled again.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144543 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-14 18:21:07 +00:00
Jakob Stoklund Olesen
4a9b615f3e Delete stale comment.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144542 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-14 18:03:05 +00:00
Jim Grosbach
430052b084 Tidy up. 80 column.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144538 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-14 17:52:47 +00:00
Benjamin Kramer
066b5d8403 Make headers standalone.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144537 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-14 17:45:03 +00:00
Benjamin Kramer
aa5354c3ba Make headers standalone, move a virtual method out of line.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144536 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-14 17:22:45 +00:00
Daniel Dunbar
7c67d32a32 build/Make: Switch over to using llvm-config-2 for dependencies one more (hopefully last) time, now that it also builds as a build tool.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144535 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-14 17:17:45 +00:00
Chandler Carruth
f5e47ac596 It helps to deallocate memory as well as allocate it. =] This actually
cleans up all the chains allocated during the processing of each
function so that for very large inputs we don't just grow memory usage
without bound.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144533 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-14 10:57:23 +00:00
Chandler Carruth
bc83fcd9bd Remove an over-eager assert that was firing on one of the ARM regression
tests when I forcibly enabled block placement.

It is apparantly possible for an unanalyzable block to fallthrough to
a non-loop block. I don't actually beleive this is correct, I believe
that 'canFallThrough' is returning true needlessly for the code
construct, and I've left a bit of a FIXME on the verification code to
try to track down why this is coming up.

Anyways, removing the assert doesn't degrade the correctness of the algorithm.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144532 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-14 10:55:53 +00:00
Chandler Carruth
fa97658b1c Begin chipping away at one of the biggest quadratic-ish behaviors in
this pass. We're leaving already merged blocks on the worklist, and
scanning them again and again only to determine each time through that
indeed they aren't viable. We can instead remove them once we're going
to have to scan the worklist. This is the easy way to implement removing
them. If this remains on the profile (as I somewhat suspect it will), we
can get a lot more clever here, as the worklist's order is essentially
irrelevant. We can use swapping and fold the two loops to reduce
overhead even when there are many blocks on the worklist but only a few
of them are removed.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144531 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-14 09:46:33 +00:00
Chandler Carruth
340d596509 Under the hood, MBPI is doing a linear scan of every successor every
time it is queried to compute the probability of a single successor.
This makes computing the probability of every successor of a block in
sequence... really really slow. ;] This switches to a linear walk of the
successors rather than a quadratic one. One of several quadratic
behaviors slowing this pass down.

I'm not really thrilled with moving the sum code into the public
interface of MBPI, but I don't (at the moment) have ideas for a better
interface. My direction I'm thinking in for a better interface is to
have MBPI actually retain much more state and make *all* of these
queries cheap. That's a lot of work, and would require invasive changes.
Until then, this seems like the least bad (ie, least quadratic)
solution. Suggestions welcome.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144530 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-14 09:12:57 +00:00
Tobias Grosser
ae5a6fd315 Add clang_complete to release notes
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144529 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-14 09:09:26 +00:00
Tobias Grosser
093cb7ed30 Add Polly to release notes
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144528 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-14 09:09:23 +00:00
Chandler Carruth
c4e1562825 Reuse the logic in getEdgeProbability within getHotSucc in order to
correctly handle blocks whose successor weights sum to more than
UINT32_MAX. This is slightly less efficient, but the entire thing is
already linear on the number of successors. Calling it within any hot
routine is a mistake, and indeed no one is calling it. It also
simplifies the code.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144527 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-14 08:55:59 +00:00
Chandler Carruth
2770c14185 Fix an overflow bug in MachineBranchProbabilityInfo. This pass relied on
the sum of the edge weights not overflowing uint32, and crashed when
they did. This is generally safe as BranchProbabilityInfo tries to
provide this guarantee. However, the CFG can get modified during codegen
in a way that grows the *sum* of the edge weights. This doesn't seem
unreasonable (imagine just adding more blocks all with the default
weight of 16), but it is hard to come up with a case that actually
triggers 32-bit overflow. Fortuately, the single-source GCC build is
good at this. The solution isn't very pretty, but its no worse than the
previous code. We're already summing all of the edge weights on each
query, we can sum them, check for an overflow, compute a scale, and sum
them again.

I've included a *greatly* reduced test case out of the GCC source that
triggers it. It's a pretty lame test, as it clearly is just barely
triggering the overflow. I'd like to have something that is much more
definitive, but I don't understand the fundamental pattern that triggers
an explosion in the edge weight sums.

The buggy code is duplicated within this file. I'll colapse them into
a single implementation in a subsequent commit.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144526 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-14 08:50:16 +00:00
Craig Topper
dcce244dd8 Add AVX2 version of instructions to load folding tables. Also add a bunch of missing SSE/AVX instructions.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144525 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-14 08:07:55 +00:00
Chandler Carruth
883f2fab47 Add a cautionary note to this API. It was not at all obvious to me how
expensive the most useful interface to this analysis is.

Fun story -- it's also not correct. That's getting fixed in another
patch.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144523 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-14 06:51:49 +00:00
Craig Topper
3426a3efef Add neverHasSideEffects, mayLoad, and mayStore to many patternless SSE/AVX instructions. Remove MMX check from LowerVECTOR_SHUFFLE since MMX vector types won't go through it anyway.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144522 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-14 06:46:21 +00:00
Chad Rosier
dc9205d9c2 Add support for ARM halfword load/stores and signed byte loads with negative
offsets.
rdar://10412592



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144518 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-14 04:09:28 +00:00
Jakob Stoklund Olesen
194eb71a11 Use getVNInfoBefore() when it makes sense.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144517 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-14 01:39:36 +00:00
Chandler Carruth
b5856c83ff Teach machine block placement to cope with unnatural loops. These don't
get loop info structures associated with them, and so we need some way
to make forward progress selecting and placing basic blocks. The
technique used here is pretty brutal -- it just scans the list of blocks
looking for the first unplaced candidate. It keeps placing blocks like
this until the CFG becomes tractable.

The cost is somewhat unfortunate, it requires allocating a vector of all
basic block pointers eagerly. I have some ideas about how to simplify
and optimize this, but I'm trying to get the logic correct first.

Thanks to Benjamin Kramer for the reduced test case out of GCC. Sadly
there are other bugs that GCC is tickling that I'm reducing and working
on now.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144516 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-14 00:00:35 +00:00
Jakob Stoklund Olesen
6c9cc21d85 Use kill slots instead of the previous slot in shrinkToUses.
It's more natural to use the actual end points.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144515 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-13 23:53:25 +00:00
Chandler Carruth
c0f05b3c3f Cleanup some 80-columns violations and poor formatting. These snuck by
when I was reading through the code for style.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144513 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-13 22:50:09 +00:00
Jakob Stoklund Olesen
1f81e316b0 Terminate all dead defs at the dead slot instead of the 'next' slot.
This makes no difference for normal defs, but early clobber dead defs
now look like:

  [Slot_EarlyClobber; Slot_Dead)

instead of:

  [Slot_EarlyClobber; Slot_Register).

Live ranges for normal dead defs look like:

  [Slot_Register; Slot_Dead)

as before.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144512 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-13 22:42:13 +00:00
Craig Topper
f9fd29ea34 Fix comment for LegalizeTypeAction enum.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144511 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-13 22:11:24 +00:00
Jakob Stoklund Olesen
d14614e677 Simplify early clobber slots a bit.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144507 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-13 22:05:42 +00:00
Chandler Carruth
10252db69b Enhance the assertion mechanisms in place to make it easier to catch
when we fail to place all the blocks of a loop. Currently this is
happening for unnatural loops, and this logic helps more immediately
point to the problem.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144504 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-13 21:39:51 +00:00
Jakob Stoklund Olesen
2debd48ca7 Rename SlotIndexes to match how they are used.
The old naming scheme (load/use/def/store) can be traced back to an old
linear scan article, but the names don't match how slots are actually
used.

The load and store slots are not needed after the deferred spill code
insertion framework was deleted.

The use and def slots don't make any sense because we are using
half-open intervals as is customary in C code, but the names suggest
closed intervals.  In reality, these slots were used to distinguish
early-clobber defs from normal defs.

The new naming scheme also has 4 slots, but the names match how the
slots are really used.  This is a purely mechanical renaming, but some
of the code makes a lot more sense now.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144503 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-13 20:45:27 +00:00
Craig Topper
e6a6277d36 Add BLSI, BLSMSK, and BLSR to getTargetNodeName.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144502 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-13 17:31:07 +00:00
Chandler Carruth
6527ecc918 Teach MBP to force-merge layout successors for blocks with unanalyzable
branches that also may involve fallthrough. In the case of blocks with
no fallthrough, we can still re-order the blocks profitably. For example
instruction decoding will in some cases continue past an indirect jump,
making laying out its most likely successor there profitable.

Note, no test case. I don't know how to write a test case that exercises
this logic, but it matches the described desired semantics in
discussions with Jakob and others. If anyone has a nice example of IR
that will trigger this, that would be lovely.

Also note, there are still assertion failures in real world code with
this. I'm digging into those next, now that I know this isn't the cause.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144499 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-13 12:17:28 +00:00
Chandler Carruth
f3fc0050ab Hoist another gross nested loop into a helper method.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144498 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-13 11:42:26 +00:00
Chandler Carruth
729bec89bd Add a missing doxygen comment for a helper method.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144497 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-13 11:34:55 +00:00
Chandler Carruth
9fd4e056e4 Hoist a nested loop into its own method.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144496 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-13 11:34:53 +00:00
Chandler Carruth
df234353fb Rewrite #3 of machine block placement. This is based somewhat on the
second algorithm, but only loosely. It is more heavily based on the last
discussion I had with Andy. It continues to walk from the inner-most
loop outward, but there is a key difference. With this algorithm we
ensure that as we visit each loop, the entire loop is merged into
a single chain. At the end, the entire function is treated as a "loop",
and merged into a single chain. This chain forms the desired sequence of
blocks within the function. Switching to a single algorithm removes my
biggest problem with the previous approaches -- they had different
behavior depending on which system triggered the layout. Now there is
exactly one algorithm and one basis for the decision making.

The other key difference is how the chain is formed. This is based
heavily on the idea Andy mentioned of keeping a worklist of blocks that
are viable layout successors based on the CFG. Having this set allows us
to consistently select the best layout successor for each block. It is
expensive though.

The code here remains very rough. There is a lot that needs to be done
to clean up the code, and to make the runtime cost of this pass much
lower. Very much WIP, but this was a giant chunk of code and I'd rather
folks see it sooner than later. Everything remains behind a flag of
course.

I've added a couple of tests to exercise the issues that this iteration
was motivated by: loop structure preservation. I've also fixed one test
that was exhibiting the broken behavior of the previous version.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144495 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-13 11:20:44 +00:00
Chad Rosier
9eb674880b The order in which the predicate is added differs between Thumb and ARM mode. Fix predicate when in ARM mode and restore SelectIntrinsicCall.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144494 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-13 09:44:21 +00:00
Chad Rosier
a517ab155b Temporarily disable SelectIntrinsicCall when in ARM mode. This is causing failures.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144492 91177308-0d34-0410-b5e6-96231b3b80d8
2011-11-13 05:14:43 +00:00