Joseph Tremoulet [Fri, 28 Aug 2015 01:12:35 +0000 (01:12 +0000)]
[WinEH] Update coloring to handle nested cases cleanly
Summary:
Change the coloring algorithm in WinEHPrepare to visit a funclet's exits
in its parents' contexts and so properly classify the continuations of
nested funclets.
Also change the placement of cloned blocks to be deterministic and to
maintain the relative order of each funclet's blocks.
Add a lit test showing various patterns that require cloning, the last
several of which don't have CHECKs yet because they require cloning
entire funclets which is NYI.
Piotr Padlewski [Fri, 28 Aug 2015 01:01:57 +0000 (01:01 +0000)]
Constant propagation after hiting llvm.assume
After hitting @llvm.assume(X) we can:
- propagate equality that X == true
- if X is icmp/fcmp (with eq operation), and one of operand
is constant we can change all variables with constants in the same BasicBlock
Prior to this patch, we hadn't been marking StratifiedSets with the
appropriate StratifiedAttrs when handling the result of no-args call
instructions. This caused us to report NoAlias when handed, for
example, an escaped alloca and a result from an opaque function. Now we
properly mark the return value of said functions.
Thanks again to Chandler, Richard, and Nick for pinging me about this.
CodeGen: Introduce splitCodeGen and teach LTOCodeGenerator to use it.
llvm::splitCodeGen is a function that implements the core of parallel LTO
code generation. It uses llvm::SplitModule to split the module into linkable
partitions and spawning one code generation thread per partition. The function
produces multiple object files which can be linked in the usual way.
This has been threaded through to LTOCodeGenerator (and llvm-lto for testing
purposes). Separate patches will add parallel LTO support to the gold plugin
and lld.
David Majnemer [Thu, 27 Aug 2015 23:03:01 +0000 (23:03 +0000)]
[ValueTracking] readnone CallInsts are fair game for speculation
Any call which is side effect free is trivially OK to speculate. We
already had similar logic in EarlyCSE and GVN but we were missing it
from isSafeToSpeculativelyExecute.
Ahmed Bougacha [Thu, 27 Aug 2015 21:46:04 +0000 (21:46 +0000)]
[CodeGen] Check FoldConstantArithmetic result before using it.
Fixes PR24602: r245689 introduced an unguarded use of
SelectionDAG::FoldConstantArithmetic, which returns 0 when it fails
because of opaque (hoisted) constants.
Ahmed Bougacha [Thu, 27 Aug 2015 20:43:34 +0000 (20:43 +0000)]
[TableGen][DAGISel] Dedup predicates with same code to run. NFCI.
I locally hit the 255 limit, but a lot of these are redundant: each
predicate coming from a different record was allocated a new number,
even when we already emitted the same code for another predicate.
Instead, re-use numbers and emit the predicate code only once.
This reduces the total text size of *DAGISel.cpp.o by ~1%.
Lang Hames [Thu, 27 Aug 2015 20:31:44 +0000 (20:31 +0000)]
Recommit r246175 - Add Kaleidoscope regression tests, with a fix to make sure
the kaleidoscope 'library' functions aren't dead-stripped in release builds.
Erik Schnetter [Thu, 27 Aug 2015 19:56:57 +0000 (19:56 +0000)]
Enable constant propagation for more math functions
Constant propagation for single precision math functions (such as
tanf) is already working, but was not enabled. This patch enables
these for many single-precision functions, and adds respective test
cases.
Tyler Nowicki [Thu, 27 Aug 2015 18:56:49 +0000 (18:56 +0000)]
Improve vectorization diagnostic messages and extend vectorize(enable) pragma.
This patch changes the analysis diagnostics produced when loops with
floating-point recurrences or memory operations are identified. The new messages
say "cannot prove it is safe to reorder * operations; allow reordering by
specifying #pragma clang loop vectorize(enable)". Depending on the type of
diagnostic the message will include additional options such as ffast-math or
__restrict__.
This patch also allows the vectorize(enable) pragma to override the low pointer
memory check threshold. When the hint is given a higher threshold is used.
See the clang patch for the options produced for each diagnostic.
Erik Schnetter [Thu, 27 Aug 2015 18:56:23 +0000 (18:56 +0000)]
Enable constant propagation for more math functions
Constant propagation for single precision math functions (such as
tanf) is already working, but was not enabled. This patch enables
these for many single-precision functions, and adds respective test
cases.
Erik Schnetter [Thu, 27 Aug 2015 16:36:37 +0000 (16:36 +0000)]
Enable constant propagation for more math functions
Constant propagation for single precision math functions (such as
tanf) is already working, but was not enabled. This patch enables
these for many single-precision functions, and adds respective test
cases.
Chad Rosier [Thu, 27 Aug 2015 14:12:17 +0000 (14:12 +0000)]
[LoopVectorize] Add Support for Small Size Reductions.
Unlike scalar operations, we can perform vector operations on element types that
are smaller than the native integer types. We type-promote scalar operations if
they are smaller than a native type (e.g., i8 arithmetic is promoted to i32
arithmetic on Arm targets). This patch detects and removes type-promotions
within the reduction detection framework, enabling the vectorization of small
size reductions.
In the legality phase, we look through the ANDs and extensions that InstCombine
creates during promotion, keeping track of the smaller type. In the
profitability phase, we use the smaller type and ignore the ANDs and extensions
in the cost model. Finally, in the code generation phase, we truncate the result
of the reduction to allow InstCombine to rewrite the entire expression in the
smaller type.
This fixes PR21369.
http://reviews.llvm.org/D12202
James Molloy [Thu, 27 Aug 2015 09:53:00 +0000 (09:53 +0000)]
[LoopVectorize] Extract InductionInfo into a helper class...
... and move it into LoopUtils where it can be used by other passes, just like ReductionDescriptor. The API is very similar to ReductionDescriptor - that is, not very nice at all. Sorting these both out will come in a followup.
Philip Reames [Thu, 27 Aug 2015 01:32:33 +0000 (01:32 +0000)]
Allow value forwarding past release fences in EarlyCSE
A release fence acts as a publication barrier for stores within the current thread to become visible to other threads which might observe the release fence. It does not require the current thread to observe stores performed on other threads. As a result, we can allow store-load and load-store forwarding across a release fence.
We do need to make sure that stores before the fence can't be eliminated even if there's another store to the same location after the fence. In theory, we could reorder the second store above the fence and *then* eliminate the former, but we can't do this if the stores are on opposite sides of the fence.
Note: While more aggressive then what's there, this patch is still implementing a really conservative ordering. In particular, I'm not trying to exploit undefined behavior via races, or the fact that the LangRef says only 'atomic' accesses are ordered w.r.t. fences.
Philip Reames [Thu, 27 Aug 2015 01:02:28 +0000 (01:02 +0000)]
[RewriteStatepointsForGC] Reduce the number of new instructions for base pointers
When computing base pointers, we introduce new instructions to propagate the base of existing instructions which might not be bases. However, the algorithm doesn't make any effort to recognize when the new instruction to be inserted is the same as an existing one already in the IR. Since this is happening immediately before rewriting, we don't really have a chance to fix it after the pass runs without teaching loop passes about statepoints.
I'm really not thrilled with this patch. I've rewritten it 4 different ways now, but this is the best I've come up with. The case where the new instruction is just the original base defining value could be merged into the existing algorithm with some complexity. The problem is that we might have something like an extractelement from a phi of two vectors. It may be trivially obvious that the base of the 0th element is an existing instruction, but I can't see how to make the algorithm itself figure that out. Thus, I resort to the call to SimplifyInstruction instead.
Note that we can only adjust the instructions we've inserted ourselves. The live sets are still being tracked in side structures at this point in the code. We can't easily muck with instructions which might be in them. Long term, I'm really thinking we need to materialize the live pointer sets explicitly in the IR somehow rather than using side structures to track them.
Tyler Nowicki [Thu, 27 Aug 2015 01:02:04 +0000 (01:02 +0000)]
Improved printing of analysis diagnostics in the loop vectorizer.
This patch ensures that every analysis diagnostic produced by the vectorizer
will be printed if the loop has a vectorization hint on it. The condition has
also been improved to prevent printing when a disabling hint is specified.
Cong Hou [Thu, 27 Aug 2015 00:37:40 +0000 (00:37 +0000)]
Fixed a bug that edge weights are not assigned correctly when lowering switch statement.
This is a one-line-change patch that moves the update to UnhandledWeights to the correct position: it should be updated for all clusters instead of just range clusters.
Philip Reames [Wed, 26 Aug 2015 23:56:46 +0000 (23:56 +0000)]
[SimplifyCFG] Prune code from a provably unreachable switch default
As Sanjoy pointed out over in http://reviews.llvm.org/D11819, a switch on an icmp should always be able to become a branch instruction. This patch generalizes that notion slightly to prove that the default case of a switch is unreachable if the cases completely cover all possible bit patterns in the condition. Once that's done, the switch to branch conversion kicks in just fine.
Note: Duplicate case values are disallowed by the LangRef and verifier.
Cong Hou [Wed, 26 Aug 2015 23:17:52 +0000 (23:17 +0000)]
[ARM] Use BranchProbability::scale() to scale an integer with a probability in ARMBaseInstrInfo.cpp,
Previously in isProfitableToIfCvt() in ARMBaseInstrInfo.cpp, the multiplication between an integer and a branch probability is done manually in an unsafe way that may lead to overflow. This patch corrects those cases by using BranchProbability's member function scale() to avoid overflow (which stores the intermediate result in int64).
Cong Hou [Wed, 26 Aug 2015 23:15:32 +0000 (23:15 +0000)]
Assign weights to edges to jump table / bit test header when lowering switch statement.
Currently, when lowering switch statement and a new basic block is built for jump table / bit test header, the edge to this new block is not assigned with a correct weight. This patch collects the edge weight from all its successors and assign this sum of weights to the edge (and also the other fall-through edge). Test cases are adjusted accordingly.
Philip Reames [Wed, 26 Aug 2015 23:13:35 +0000 (23:13 +0000)]
[docs][Statepoints] More on base pointers
Expand the information on base pointers to include an example, the assumptions a collector is allowed to make, legal optimizations over gc.relocates, and the assumptions made by RewriteStatepointsForGC. This is the result of a recent conversation with folks from LLIC and the confusions that came to light therein.
Change `DIBuilder` always to produce 'distinct' nodes when creating
`DISubprogram` definitions. I measured a ~5% memory improvement in the
link step (of ld64) when using `-flto -g`.
`DISubprogram`s are used in two ways in the debug info graph.
Some are definitions, point at actual functions, and can't really be
shared between compile units. With full debug info, these point down at
their variables, forming uniquing cycles. These uniquing cycles are
expensive to link between modules, since all unique nodes that reference
them transitively need to be duplicated (see commit message for r244181
for more details).
Others are declarations, primarily used for member functions in the type
hierarchy. Definitions never show up there; instead, a definition
points at its corresponding declaration node.
I started by making all subprograms 'distinct'. However, that was too
big a hammer: memory usage *increased* ~5% (net increase vs. this patch
of ~10%) because the 'distinct' declarations undermine LTO type
uniquing. This is a targeted fix for the definitions (where uniquing is
an observable problem).
A couple of notes:
- There's an accompanying commit to update IRGen testcases in clang.
- ^ That's what I'm using to test this commit.
- In a follow-up, I'll change the verifier to require 'distinct' on
definitions and add an upgrade to `BitcodeReader`.
JF Bastien [Wed, 26 Aug 2015 22:09:54 +0000 (22:09 +0000)]
WebAssembly: handle private/internal globals.
Things of note:
- Other linkage types aren't handled yet. We'll figure it out with dynamic linking.
- Special LLVM globals are either ignored, or error out for now.
- TLS isn't supported yet (WebAssembly will have threads later).
- There currently isn't a syntax for alignment, I left it in a comment so it's easy to hook up.
- Undef is convereted to whatever the type's appropriate null value is.
- assert versus report_fatal_error: follow what other AsmPrinters do, and assert only on what should have been caught elsewhere.
Reid Kleckner [Wed, 26 Aug 2015 21:57:25 +0000 (21:57 +0000)]
[ms-inline-asm] Relax assertion around funky identifiers slightly
A corresponding clang change will make it so that clang can consume part
of an assembler token. The assembler treats '.' as an identifier
character while clang does not, so it's view of the token stream is a
little different.
Mehdi Amini [Wed, 26 Aug 2015 21:16:29 +0000 (21:16 +0000)]
Fix LLVM C API for DataLayout
We removed access to the DataLayout on the TargetMachine and
deprecated the C API function LLVMGetTargetMachineData() in r243114.
However the way I tried to be backward compatible was broken: I
changed the wrapper of the TargetMachine to be a structure that
includes the DataLayout as well. However the TargetMachine is also
wrapped by the ExecutionEngine, in the more classic way. A client
using the TargetMachine wrapped by the ExecutionEngine and trying
to get the DataLayout would break.
It seems tricky to solve the problem completely in the C API
implementation. This patch tries to address this backward
compatibility in a more lighter way in the C++ API. The C API is
restored in its original state and the removed C++ API is
reintroduced, but privately. The C API is friended to the
TargetMachine and should be the only consumer for this API.
Matt Arsenault [Wed, 26 Aug 2015 20:48:08 +0000 (20:48 +0000)]
AMDGPU: Delete dead code
There is no context where s_mov_b64 is emitted
and could potentially be moved to the VALU.
It is currently only emitted for materializing
immediates, which can't be dependent on vector sources.
The immediate splitting is already done when selecting
constants. I'm not sure what contexts if any the register
splitting would have been used before.
Also clean up using s_mov_b64 in place of v_mov_b64_pseudo,
although this isn't required and just skips the extra step
of eliminating the copy from the SReg_64.
Matt Arsenault [Wed, 26 Aug 2015 20:47:50 +0000 (20:47 +0000)]
AMDGPU: Don't create intermediate SALU instructions
When splitting 64-bit operations, create the correct
VALU instructions immediately.
This was splitting things like s_or_b64 into the two
s_or_b32s and then pushing the new instructions
onto the worklist. There's no reason we need
to do this intermediate step.
Andrew Kaylor [Wed, 26 Aug 2015 20:36:52 +0000 (20:36 +0000)]
Expose hasLiveCondCodeDef as a member function of the X86InstrInfo class. NFC
This takes the existing static function hasLiveCondCodeDef and makes it a member function of the X86InstrInfo class. This is a useful utility function that an upcoming change would like to use. NFC.
Patch by: Kevin B. Smith
Differential Revision: http://reviews.llvm.org/D12371
Diego Novillo [Wed, 26 Aug 2015 20:00:27 +0000 (20:00 +0000)]
Fix memory leak in sample profile pass.
The problem here were the function analyses invoked by the function pass
manager from the new IPO pass. I looked at other IPO passes needing
dominance information and the only one that requires it (partial
inliner) does not use the standard dependency mechanism.
This patch mimics what the partial inliner does to compute dominance,
post-dominance and loop info. One thing I like about this approach is
that I can delay the computation of all this until I actually need it.
This should bring the ASAN buildbot back to green. If there's a better
way to fix this, I'll do it in a follow-up patch.
Mehdi Amini [Wed, 26 Aug 2015 18:56:01 +0000 (18:56 +0000)]
Fix LLVM C API for DataLayout
We removed access to the DataLayout on the TargetMachine and
deprecated the C API function LLVMGetTargetMachineData() in r243114.
However the way I tried to be backward compatible was broken: I
changed the wrapper of the TargetMachine to be a structure that
includes the DataLayout as well. However the TargetMachine is also
wrapped by the ExecutionEngine, in the more classic way. A client
using the TargetMachine wrapped by the ExecutionEngine and trying
to get the DataLayout would break.
It seems tricky to solve the problem completely in the C API
implementation. This patch tries to address this backward
compatibility in a more lighter way in the C++ API. The C API is
restored in its original state and the removed C++ API is
reintroduced, but privately. The C API is friended to the
TargetMachine and should be the only consumer for this API.
Mehdi Amini [Wed, 26 Aug 2015 18:22:34 +0000 (18:22 +0000)]
Fix LLVM C API for DataLayout
We removed access to the DataLayout on the TargetMachine and
deprecated the C API function LLVMGetTargetMachineData() in r243114.
However the way I tried to be backward compatible was broken: I
changed the wrapper of the TargetMachine to be a structure that
includes the DataLayout as well. However the TargetMachine is also
wrapped by the ExecutionEngine, in the more classic way. A client
using the TargetMachine wrapped by the ExecutionEngine and trying
to get the DataLayout would break.
It seems tricky to solve the problem completely in the C API
implementation. This patch tries to address this backward
compatibility in a more lighter way in the C++ API. The C API is
restored in its original state and the removed C++ API is
reintroduced, but privately. The C API is friended to the
TargetMachine and should be the only consumer for this API.
James Y Knight [Wed, 26 Aug 2015 17:57:51 +0000 (17:57 +0000)]
[SPARC] Fix stupid oversight in stack realignment support.
If you're going to realign %sp to get object alignment properly (which
the code does), and stack offsets and alignments are calculated going
down from %fp (which they are), then the total stack size had better
be a multiple of the alignment. LLVM did indeed ensure that.
And then, after aligning, the sparc frame code added 96 (for sparcv8)
to the frame size, making any requested alignment of 64-bytes or
higher *guaranteed* to be misaligned. The test case added with r245668
even tests this exact scenario, and asserted the incorrect behavior,
which I somehow failed to notice. D'oh.
This change fixes the frame lowering code to align the stack size
*after* adding the spill area, instead.
Vedant Kumar [Wed, 26 Aug 2015 16:20:29 +0000 (16:20 +0000)]
[llvm-mc] Ignore opcode size prefix in 64-bit CALL disassembly
This is a fix for disassembling unusual instruction sequences in 64-bit
mode w.r.t the CALL rel16 instruction. It might be desirable to move the
check somewhere else, but it essentially mimics the special case
handling with JCXZ in 16-bit mode.
The current behavior accepts the opcode size prefix and causes the
call's immediate to stop disassembling after 2 bytes. When debugging
sequences of instructions with this pattern, the disassembler output
becomes extremely unreliable and essentially useless (if you jump midway
into what lldb thinks is a unified instruction, you'll lose %rip). So we
ignore the prefix and consume all 4 bytes when disassembling a 64-bit
mode binary.
Note: in Vol. 2A 3-99 the Intel spec states that CALL rel16 is N.S. N.S.
is defined as:
Indicates an instruction syntax that requires an address override
prefix in 64-bit mode and is not supported. Using an address
override prefix in 64-bit mode may result in model-specific
execution behavior. (Vol. 2A 3-7)
Since 0x66 is an operand override prefix we should be OK (although we
may want to warn about 0x67 prefixes to 0xe8). On the CPUs I tested
with, they all ignore the 0x66 prefix in 64-bit mode.
Silviu Baranga [Wed, 26 Aug 2015 11:11:14 +0000 (11:11 +0000)]
[AArch64] Unify the integer min/max vector selection patterns with the intrinsic ones
Summary:
This change lowers the aarch64 integer vector min/max intrinsic nodes to
generic min/max nodes and replaces the intrinsic selection patterns with
the generic ones.
There should already be testing in place for this, so no further tests
were added.
Chandler Carruth [Wed, 26 Aug 2015 09:09:29 +0000 (09:09 +0000)]
[SROA] Rip out all support for SSAUpdater in SROA.
This was only added to preserve the old ScalarRepl's use of SSAUpdater
which was originally to avoid use of dominance frontiers. Now, we only
need a domtree, and we'll need a domtree right after this pass as well
and so it makes perfect sense to always and only use the dom-tree
powered mem2reg. This was flag-flipper earlier and has stuck reasonably
so I wanted to gut the now-dead code out of SROA before we waste more
time with it. Among other things, this will make passmanager porting
easier.
Frederic Riss [Wed, 26 Aug 2015 05:10:04 +0000 (05:10 +0000)]
[dsymutil] Emit real dSYM companion binaries.
The binaries containing the linked DWARF generated by dsymutil are not
standard relocatable object files like emitted did previsously. They should be
dSYM companion files, which means they have a different file type in the
header, but also a couple other peculiarities:
- they contain the segments and sections from the original binary in their
load commands, but not the actual contents. This means they get an address
and a size, but their offset is always 0 (but these are not virtual sections)
- they also conatin all the defined symbols from the original binary
This makes MC a really bad fit to emit these kind of binaries. The approach
that was used in this patch is to leverage MC's section layout for the
debug sections, but to use a replacement for MachObjectWriter that lives
in MachOUtils.cpp. Some of the low-level helpers from MachObjectWriter
were reused too.
Frederic Riss [Wed, 26 Aug 2015 05:09:59 +0000 (05:09 +0000)]
[dsymutil] Store an optional BinaryPath in the debug map.
llvm-dsymutil needs to emit dSYM companion bundles. These are binary files
that replicate some of the orignal binary file properties (sections and
symbols). To get acces to these properties, pass the binary path in the
debug map.
Frederic Riss [Wed, 26 Aug 2015 05:09:49 +0000 (05:09 +0000)]
[MC] Split the layout part of MCAssembler::finish() into its own method. NFC.
Split a MCAssembler::layout() method out of MCAssembler::finish(). This allows
running the MCSections layout separately from the streaming of the output
file. This way if a client wants to use MC to generate section contents, but
emit something different than the standard relocatable object files it is
possible (llvm-dsymutil is such a client).
Lang Hames [Wed, 26 Aug 2015 03:07:41 +0000 (03:07 +0000)]
Big Kaleidoscope tutorial update.
This commit switches the underlying JIT for the Kaleidoscope tutorials from
MCJIT to a custom ORC-based JIT, KaleidoscopeJIT. This fixes a lot of the bugs
in Kaleidoscope that were introduced when we deleted the legacy JIT. The
documentation for Chapter 4, which introduces the JIT APIs, is updated to
reflect the change.
Also included are a number of C++11 modernizations and general cleanup. Where
appropriate, the docs have been updated to reflect these changes too.
JF Bastien [Wed, 26 Aug 2015 03:02:58 +0000 (03:02 +0000)]
Comparing operands should not require the same ValueID
Summary: When comparing basic blocks, there is an additional check that two Value*'s should have the same ID, which interferes with merging equivalent constants of different kinds (such as a ConstantInt and a ConstantPointerNull in the included testcase). The cmpValues function already ensures that the two values in each function are the same, so removing this check should not cause incorrect merging.
Also, the type comparison is redundant, based on reviewing the code and testing on the test suite and several large LTO bitcodes.
JF Bastien [Wed, 26 Aug 2015 02:32:45 +0000 (02:32 +0000)]
Expose more properties of llvm::fltSemantics
Summary: Adds accessor functions for all the fields in llvm::fltSemantics. This will be used in MergeFunctions to order two APFloats with different semanatics.
Matthias Braun [Wed, 26 Aug 2015 01:38:00 +0000 (01:38 +0000)]
FastISel: Factor out common code; NFC intended
This should be no functional change but for the record: For three cases
in X86FastISel this will change the order in which the FalseMBB and
TrueMBB of a conditional branch is addedd to the successor/predecessor
lists.
Charles Davis [Tue, 25 Aug 2015 23:27:41 +0000 (23:27 +0000)]
Make variable argument intrinsics behave correctly in a Win64 CC function.
Summary:
This change makes the variable argument intrinsics, `llvm.va_start` and
`llvm.va_copy`, and the `va_arg` instruction behave as they do on Windows
inside a `CallingConv::X86_64_Win64` function. It's needed for a Clang patch
I have to add support for GCC's `__builtin_ms_va_list` constructs.
Frederic Riss [Tue, 25 Aug 2015 23:15:26 +0000 (23:15 +0000)]
[dsymutil] Reapply r245960.
There was an issue in the test setup because the test requires an arch that
wasn't filtered by the lit.local.cfg, but given the set of bots that failed,
I'm not confident this is the (only) issue. So this commit also adds more
output to the test to help me track down the failure if it happens again.
Original commit message:
[dsymutil] Rewrite thumb triple names in user visible messages.
We autodetect triples from the input file(s) while reading the mach-o debug map.
As we need to create a Target from those triples, we always chose the thumb
variant (because the arm variant might not be 'instantiable' eg armv7m). The
user visible architecture names should still be 'arm' and not 'thumb' variants
though.
projects / llvm / log