Alina Sbirlea [Fri, 22 Dec 2017 19:54:03 +0000 (19:54 +0000)]
[MemorySSA] Allow reordering of loads that alias in the presence of volatile loads.
Summary:
Make MemorySSA allow reordering of two loads that may alias, when one is volatile.
This makes MemorySSA less conservative and behaving the same as the AliasSetTracker.
For more context, see D16875.
LLVM language reference: "The optimizers must not change the number of volatile operations or change their order of execution relative to other volatile operations. The optimizers may change the order of volatile operations relative to non-volatile operations. This is not Java’s “volatile” and has no cross-thread synchronization behavior."
Guozhi Wei [Fri, 22 Dec 2017 18:54:04 +0000 (18:54 +0000)]
[SimplifyCFG] Don't do if-conversion if there is a long dependence chain
If after if-conversion, most of the instructions in this new BB construct a long and slow dependence chain, it may be slower than cmp/branch, even if the branch has a high miss rate, because the control dependence is transformed into data dependence, and control dependence can be speculated, and thus, the second part can execute in parallel with the first part on modern OOO processor.
This patch checks for the long dependence chain, and give up if-conversion if find one.
In https://reviews.llvm.org/rL321077 and https://reviews.llvm.org/D41231 I fixed a regression in the c-api which prevented the pruning from being *effectively* disabled.
However this approach, helpfully recommended by @labath, is cleaner.
It is also nice to remove the weasel words about effectively disabling from the api comments.
Sanjoy Das [Fri, 22 Dec 2017 18:21:59 +0000 (18:21 +0000)]
(Re-landing) Expose a TargetMachine::getTargetTransformInfo function
Re-land r321234. It had to be reverted because it broke the shared
library build. The shared library build broke because there was a
missing LLVMBuild dependency from lib/Passes (which calls
TargetMachine::getTargetIRAnalysis) to lib/Target. As far as I can
tell, this problem was always there but was somehow masked
before (perhaps because TargetMachine::getTargetIRAnalysis was a
virtual function).
Original commit message:
This makes the TargetMachine interface a bit simpler. We still need
the std::function in TargetIRAnalysis to avoid having to add a
dependency from Analysis to Target.
See discussion:
http://lists.llvm.org/pipermail/llvm-dev/2017-December/119749.html
I avoided adding all of the backend owners to this review since the
change is simple, but let me know if you feel differently about this.
Craig Topper [Fri, 22 Dec 2017 17:18:13 +0000 (17:18 +0000)]
[SelectionDAG] Reverse the order of operands in the ISD::ADD created by TargetLowering::getVectorElementPointer so that the FrameIndex is on the left.
This seems to improve X86's ability to match this into an address computation. Otherwise the other operand gets assigned to the base register and the stack pointer + frame index ends up in the index register. But index registers can't encode ESP/RSP so we end up having to move it into another register to meet the constraint.
I could try to improve the address matcher in X86, but swapping the producer seemed easier. Several other places already have the operands in this order so this is at least consistent.
Craig Topper [Fri, 22 Dec 2017 17:18:11 +0000 (17:18 +0000)]
[X86] When lowering insert_vector_elt/extract_vector_elt of vXi1 with a non-constant index just use either a 128-bit type or the vXi8 type with the correct number of elements.
Despite what the comment said there isn't better codegen for 512-bit vectors. The 128/256/512 bit implementation jus stores to memory and loads an element. There's no advantage to doing that with a larger size. In fact in many cases it causes a stack realignment and generates worse code.
Haicheng Wu [Fri, 22 Dec 2017 17:09:09 +0000 (17:09 +0000)]
[InlineCost] Find more free binary operations
Currently, inline cost model considers a binary operator as free only if both
its operands are constants. Some simple cases are missing such as a + 0, a - a,
etc. This patch modifies visitBinaryOperator() to call SimplifyBinOp() without
going through simplifyInstruction() to get rid of the constant restriction.
Thus, visitAnd() and visitOr() are not needed.
Diana Picus [Fri, 22 Dec 2017 11:09:18 +0000 (11:09 +0000)]
[ARM GlobalISel] Support pointer constants
Pointer constants are pretty rare, since we usually represent them as
integer constants and then cast to pointer. One notable exception is the
null pointer constant, which is represented directly as a G_CONSTANT 0
with pointer type. Mark it as legal and make sure it is selected like
any other integer constant.
Chandler Carruth [Fri, 22 Dec 2017 06:41:23 +0000 (06:41 +0000)]
Rewrite the cached map used for locating the most precise DIE among
inlined subroutines for a given address.
This is essentially the hot path of llvm-symbolizer when extracting
inlined frames during symbolization. Previously, we would read every
subprogram and every inlined subroutine, building a std::map across the
entire PC space to the best DIE, and then do only a handful of queries
as we symbolized a backtrace. A huge fraction of the time was spent
building the map itself.
This patch changes it two a two-level system. First, we just build a map
from PC-interval to DWARF subprograms. These are required to be disjoint
and so constructing this is pretty easy. Second, we build a map *just*
for the inlined subroutines within the subprogram containing the query
address. This allows us to look at far fewer DIEs and build a *much*
smaller set of cached maps in the llvm-symbolizer case where only a few
address get symbolized during the entire run.
It also builds both interval maps in a very different way. It constructs
a single flat vector of pairs that maps from offset -> index. The
indices point into collections of DIE objects, but can also be
"tombstones" (-1) to mark gaps. In the case of subprograms, this mostly
just simplifies the data structure a bit. For inlined subroutines,
because we carefully split them as we build the map, we end up in many
cases having no holes and not having to store both start and stop
offsets.
Finally, the PC ranges for the inlined subroutines are compressed into
32-bits by making them relative to the base PC of the outer subprogram.
This means that if you have a single function body with over 2gb of
executable code in it, we will stop mapping address past the first 2gb
of that function into inlined subroutines and just give you the
subprogram. This doesn't seem like a problem. ;]
All of this combines to make llvm-symbolizer *well* over 2x faster for
symbolizing backtraces out of LLVM's unittests. Death-test heavy unit
tests are running >2x faster. I'm still going to look at completely
disabling symbolization there, but figured while I had a good benchmark
we should make symbolization a bit better.
Sadly, the logic to build the flat interval map for the inlined
subroutines is fairly complex. I'm not super happy about this and
welcome any simplifying suggestions.
Huge thanks to Dave Blaikie who helped walk me through what the various
things I needed to do in DWARF to make this work.
Craig Topper [Fri, 22 Dec 2017 02:30:30 +0000 (02:30 +0000)]
[X86] Add prefetchwt1 instruction and overhaul priorities and isel enabling for prefetch instructions.
Previously prefetch was only considered legal if sse was enabled, but it should be supported with 3dnow as well.
The prfchw flag now imply at least some form of prefetch without the write hint is available, either the sse or 3dnow version. This is true even if 3dnow and sse are explicitly disabled.
Similarly prefetchwt1 feature implies availability of prefetchw and the the prefetcht0/1/2/nta instructions. This way we can support _MM_HINT_ET0 using prefetchw and _MM_HINT_ET1 with prefetchwt1. And its assumed that if we have levels for the write hint we would have levels for the non-write hint, thus why we enable the sse prefetch instructions.
I believe this behavior is consistent with gcc. I've updated the prefetch.ll to test all of these combinations.
Eli Friedman [Fri, 22 Dec 2017 02:08:08 +0000 (02:08 +0000)]
[Inliner] Restrict soft-float inlining penalty.
The penalty is currently getting applied in a bunch of places where it
doesn't make sense, like bitcasts (which are free) and calls (which
were getting the call penalty applied twice). Instead, just apply the
penalty to binary operators and floating-point casts.
While I'm here, also fix getFPOpCost() to do the right thing in more
cases, so we don't have to dig into function attributes.
Easwaran Raman [Fri, 22 Dec 2017 01:33:52 +0000 (01:33 +0000)]
Add hasProfileData() to check if a function has profile data. NFC.
Summary:
This replaces calls to getEntryCount().hasValue() with hasProfileData
that does the same thing. This refactoring is useful to do before adding
synthetic function entry counts but also a useful cleanup IMO even
otherwise. I have used hasProfileData instead of hasRealProfileData as
David had earlier suggested since I think profile implies "real" and I
use the phrase "synthetic entry count" and not "synthetic profile count"
but I am fine calling it hasRealProfileData if you prefer.
Alina Sbirlea [Thu, 21 Dec 2017 21:41:53 +0000 (21:41 +0000)]
[ModRefInfo] Add must alias info to ModRefInfo.
Summary:
Add an additional bit to ModRefInfo, ModRefInfo::Must, to be cleared for known must aliases.
Shift existing Mod/Ref/ModRef values to include an additional most
significant bit. Update wrappers that modify ModRefInfo values to
reflect the change.
Notes:
* ModRefInfo::Must is almost entirely cleared in the AAResults methods, the remaining changes are trying to preserve it.
* Only some small changes to make custom AA passes set ModRefInfo::Must (BasicAA).
* GlobalsModRef already declares a bit, who's meaning overlaps with the most significant bit in ModRefInfo (MayReadAnyGlobal). No changes to shift the value of MayReadAnyGlobal (see AlignedMap). FunctionInfo.getModRef() ajusts most significant bit so correctness is preserved, but the Must info is lost.
* There are cases where the ModRefInfo::Must is not set, e.g. 2 calls that only read will return ModRefInfo::NoModRef, though they may read from exactly the same location.
Craig Topper [Thu, 21 Dec 2017 20:45:13 +0000 (20:45 +0000)]
[X86] When lowering truncates to vXi1, don't sign extend i16/i8 types to 512-bit if we have VLX.
This should only affect what we do for v8i16. Previously we went to v8i64, but if we have VLX we only need v8i32. This prevents an unnecessary zmm usage.
Wolfgang Pieb [Thu, 21 Dec 2017 19:38:13 +0000 (19:38 +0000)]
[DWARF v5] Rework of string offsets table reader
Reorganizes the DWARF consumer to derive the string offsets table
contribution's format from the contribution header instead of
(incorrectly) from the unit's format.
Craig Topper [Thu, 21 Dec 2017 18:44:06 +0000 (18:44 +0000)]
[X86] Promote v8i1 shuffles to v8i32 instead of v8i64 if we have VLX.
We should have equally good shuffle options for v8i32 with VLX. This was spotted during my attempts to remove 512-bit vectors from SKX.
We still use 512-bits for v16i1, v32i1, and v64i1. I'm less sure we can handle those well with narrower vectors. i32 and i64 element sizes get the best shuffle support.
Simon Pilgrim [Thu, 21 Dec 2017 16:34:46 +0000 (16:34 +0000)]
[DAGCombiner] Generalize (or (and X, c1), c2) -> (and (or X, c2), c1|c2) combine to work on non-splat vectors
The knownbits_mask_or_shuffle_uitofp change is interesting - shuffle combines manage to kick in, removing the AND constant mask load. For targets with fast-variable-shuffle this should reduce further to VPOR+VPSHUFB+VCVTDQ2PS.
which is clearly wrong since ppcf128 are definitely different type with f32 and
we cannot change the node value type when do DAGCombine. The fix is don't
handle ppc_fp128 or i1 conversions in PPCTargetLowering::combineFPToIntToFP and
leave it to downstream to legalize it and expand it to small legal types.
Simon Dardis [Thu, 21 Dec 2017 15:14:07 +0000 (15:14 +0000)]
[mips] Fix the invalid EVA test
During the review of D40362 I spotted that this test wasn't actually
testing the eva instructions due to '-mattr==eva', rather than '-mattr=+eva',
which resulted in test having no effect.
Sam Parker [Thu, 21 Dec 2017 12:55:04 +0000 (12:55 +0000)]
[DAGCombine] Improve ReduceLoadWidth for SRL
If the SRL node is only used by an AND, we may be able to set the
ExtVT to the width of the mask, making the AND redundant. To support
this, another check has been added in isLegalNarrowLoad which queries
whether the load is valid.
There is nothing useful that can be done with a read-only uninitialized
buffer without const_casting its contents to initialize it. A better
solution is to obtain a writable buffer
(WritableMemoryBuffer::getNewUninitMemBuffer), and then convert it to a
read-only buffer after initialization. All callers of this function have
already been updated to do this, so this function is now unused.
Simon Atanasyan [Thu, 21 Dec 2017 10:26:02 +0000 (10:26 +0000)]
[llvm-readobj] Support 'GNU' style for MIPS GOT/PLT dumping
This change adds `printMipsGOT` and `printMipsPLT` methods to the
`DumpStyle` class and overrides them in the `GNUStyle` and `LLVMStyle`
descendants. To pass information about GOT/PLT layout into these
methods, the `MipsGOTParser` class has been extended to hold all
necessary data.
Sam Clegg [Thu, 21 Dec 2017 02:30:38 +0000 (02:30 +0000)]
[WebAssembly] Fix local references to weak aliases
When weak aliases are used with in same translation
unit we need to be able to directly reference to alias
and not just the thing it is aliases. We do this by
defining both a wasm import and a wasm export in this
case that result in a single Symbol. This change is
a partial revert of rL314245. A corresponding lld
change address the previous issues we had with this.
[SimplifyCFG] Avoid quadratic on a predecessors number behavior in instruction sinking.
If a block has N predecessors, then the current algorithm will try to
sink common code to this block N times (whenever we visit a
predecessor). Every attempt to sink the common code includes going
through all predecessors, so the complexity of the algorithm becomes
O(N^2).
With this patch we try to sink common code only when we visit the block
itself. With this, the complexity goes down to O(N).
As a side effect, the moment the code is sunk is slightly different than
before (the order of simplifications has been changed), that's why I had
to adjust two tests (note that neither of the tests is supposed to test
SimplifyCFG):
* test/CodeGen/AArch64/arm64-jumptable.ll - changes in this test mimic
the changes that previous implementation of SimplifyCFG would do.
* test/CodeGen/ARM/avoid-cpsr-rmw.ll - in this test I disabled common
code sinking by a command line flag.
Sanjoy Das [Thu, 21 Dec 2017 01:06:58 +0000 (01:06 +0000)]
Expose a TargetMachine::getTargetTransformInfo function
Summary:
This makes the TargetMachine interface a bit simpler. We still need
the std::function in TargetIRAnalysis to avoid having to add a
dependency from Analysis to Target.
See discussion:
http://lists.llvm.org/pipermail/llvm-dev/2017-December/119749.html
I avoided adding all of the backend owners to this review since the
change is simple, but let me know if you feel differently about this.
Simon Dardis [Wed, 20 Dec 2017 22:26:41 +0000 (22:26 +0000)]
[orc][cmake] Check if 8 byte atomics require libatomic for unittest
rL319838 introduced SymbolStringPool which uses 8 byte atomics for
reference counters. On systems which do not support such atomics
natively such as MIPS32, explicitly add libatomic as one of the
libraries for SymbolStringPool's unittest.
Yonghong Song [Wed, 20 Dec 2017 19:39:58 +0000 (19:39 +0000)]
bpf: add support for objdump -print-imm-hex
Add support for 'objdump -print-imm-hex' for imm64, operand imm
and branch target. If user programs encode immediate values
as hex numbers, such an option will make it easy to correlate
asm insns with source code. This option also makes it easy
to correlate imm values with insn encoding.
There is one changed behavior in this patch. In old way, we
print the 64bit imm as u64:
O << (uint64_t)Op.getImm();
and the new way is:
O << formatImm(Op.getImm());
The formatImm is defined in llvm/MC/MCInstPrinter.h as
format_object<int64_t> formatImm(int64_t Value)
So the new way to print 64bit imm is i64 type.
If a 64bit value has the highest bit set, the old way
will print the value as a positive value and the
new way will print as a negative value. The new way
is consistent with x86_64.
For the code (see the test program):
...
if (a == 0xABCDABCDabcdabcdULL)
...
x86_64 objdump, with and without -print-imm-hex, looks like:
48 b8 cd ab cd ab cd ab cd ab movabsq $-6067004223159161907, %rax
48 b8 cd ab cd ab cd ab cd ab movabsq $-0x5432543254325433, %rax
Signed-off-by: Yonghong Song <yhs@fb.com>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@321215 91177308-0d34-0410-b5e6-96231b3b80d8
Craig Topper [Wed, 20 Dec 2017 19:36:43 +0000 (19:36 +0000)]
[X86] Refactor DomainReassignment pass to make the Closure class not stores references to the main data structures of the pass itself
Multiple Closure objects can be created and stored for a single function. It's not a good idea to devote so many fields of it to storing pointers and references to global data structures of the pass. The closure class should only store the things needed to represent the closure itself.
This patch refactors many of the methods of Closure to belong to the pass object and to pass around a reference to the current Closure. The Closure class gains a few simple methods to add instructions and edges, and to return iterators to edges and instructions
Matt Arsenault [Wed, 20 Dec 2017 19:36:28 +0000 (19:36 +0000)]
TableGen: Allow setting SDNodeProperties on intrinsics
Allows preserving MachineMemOperands on intrinsics
through selection. For reasons I don't understand, this
is a static property of the pattern and the selector
deliberately goes out of its way to drop if not present.
Intrinsics already inherit from SDPatternOperator allowing
them to be used directly in instruction patterns. SDPatternOperator
has a list of SDNodeProperty, but you currently can't set them on
the intrinsic. Without SDNPMemOperand, when the node is selected
any memory operands are always dropped. Allowing setting this
on the intrinsics avoids needing to introduce another equivalent
target node just to have SDNPMemOperand set.
This patch modifies the indirect call promotion utilities by exposing and using
an unconditional call promotion interface. The unconditional promotion
interface (i.e., call promotion without creating an if-then-else) can be used
if it's known that an indirect call has only one possible callee. The existing
conditional promotion interface uses this unconditional interface to promote an
indirect call after it has been versioned and placed within the "then" block.
A consequence of unconditional promotion is that the fix-up operations for phi
nodes in the normal destination of invoke instructions are changed. This is
necessary because the existing implementation assumed that an invoke had been
versioned, creating a "merge" block where a return value bitcast could be
placed. In the new implementation, the edge between a promoted invoke's parent
block and its normal destination is split if needed to add a bitcast for the
return value. If the invoke is also versioned, the phi node merging the return
value of the promoted and original invoke instructions is placed in the "merge"
block.
Craig Topper [Wed, 20 Dec 2017 19:25:33 +0000 (19:25 +0000)]
[X86] Remove zext from vXi32 to vXi64 on indices of gather/scatter instructions if we can prove the pre-extended value is positive.
Gather/scatter can implicitly sign extend from i32->i64 on indices. So if we know the sign bit of the input to a zext is 0 we can use the implicit extension.
Matt Arsenault [Wed, 20 Dec 2017 19:11:59 +0000 (19:11 +0000)]
DAG: Tolerate non-MemSDNodes for OPC_RecordMemRef
When intrinsics are allowed to have mem operands, there
are two ways this can happen. First is an intrinsic
that is marked has having a mem operand, but is not handled
by getTgtMemIntrinsic.
The second way can occur even for intrinsics which do not
have a mem operand. It seems the selector table does
some kind of sorting based on the opcode, and the
mem ref recording can happen in the same scope for
intrinsics that both do and do not have mem refs.
I haven't been able to figure out exactly why this happens
(although it happens even with the matcher optimizations disabled).
I'm not sure if it's worth trying to avoid hitting this for
these nodes since I think it's still reasonable to handle
this in case getTgtMemIntrinic is not implemented.
Stefan Pintilie [Wed, 20 Dec 2017 19:07:44 +0000 (19:07 +0000)]
[PowerPC] Added an assert to make sure that the MBBI iterator is valid.
The function createTailCallBranchInstr assumes that the iterator MBBI is valid.
However, only one use of MBBI is guarded in the function.
Fix this by adding an assert.
Teresa Johnson [Wed, 20 Dec 2017 17:53:10 +0000 (17:53 +0000)]
[PGO] Function section hotness prefix should look at all blocks
Summary:
The function section prefix for PGO based layout (e.g. hot/unlikely)
should look at the hotness of all blocks not just the entry BB.
A function with a cold entry but a very hot loop should be placed in the
hot section, for example, so that it is located close to other hot
functions it may call. For SamplePGO it was already looking at the
branch weights on calls, and I made that code conditional on whether
this is SamplePGO since it was essentially a noop for instrumentation
PGO anyway.
Nemanja Ivanovic [Wed, 20 Dec 2017 15:44:32 +0000 (15:44 +0000)]
[JumpTables] Let targets decide which switch instructions are suitable
This commits the non-controversial part of https://reviews.llvm.org/D41029
(making the queries virtual). The PPC-specific portion of this will be
committed in a subsequent patch once some of the finer points are ironed out.
Add optional SelectionDAG* parameter to SValue::dump and SDValue::dumpr
These functions simply call their counterparts in the associated SDNode,
which do take an optional SelectionDAG. This change makes the legalization
debug trace a little easier to read, since target-specific nodes will
now have their names shown instead of "Unknown node #123".
Alexey Bataev [Wed, 20 Dec 2017 14:55:10 +0000 (14:55 +0000)]
[NVPTX] Initial adaptation of MCAsmStreamer/MCTargetStreamer for debug info in Cuda.
Summary:
Initial changes in interfaces of MCAsmStreamer/MCTargetStreamer for
correct debug info emission for Cuda.
1. PTX foramt does not support `.ascii` directives. Added the ability to
nullify it.
2. The initial function label must follow the first debug `.loc`
directive, not be followed by.
3. DWARF sections must be enclosed in braces.
canVectorize is only checking if the loop has a normalized pre-header if DoExtraAnalysis is true.
This doesn't make sense to me because reporting analysis information shouldn't alter legality
checks. This is probably the result of a last minute minor change before committing (?).
Sander de Smalen [Wed, 20 Dec 2017 12:45:40 +0000 (12:45 +0000)]
Trivial commit to force LLVM to run TableGen for Mips target after
a change to the AsmMatcherEmitter, and should fix the buildbot
failure on llvm-clang-x86_64-expensive-checks-win.
The issue is also described here:
http://lists.llvm.org/pipermail/llvm-dev/2017-December/119617.html
Diana Picus [Wed, 20 Dec 2017 11:27:10 +0000 (11:27 +0000)]
[ARM GlobalISel] Fix assertion in RegBankSelect
We get an assertion in RegBankSelect for code along the lines of
my_32_bit_int = my_64_bit_int, which tends to translate into a 64-bit
load, followed by a G_TRUNC, followed by a 32-bit store. This appears in
a couple of places in the test-suite.
At the moment, the legalizer doesn't distinguish between integer and
floating point scalars, so a 64-bit load will be marked as legal for
targets with VFP, and so will the rest of the sequence, leading to a
slightly bizarre G_TRUNC reaching RegBankSelect.
Since the current support for 64-bit integers is rather immature, this
patch works around the issue by explicitly handling this case in
RegBankSelect and InstructionSelect. In the future, we may want to
revisit this decision and make sure 64-bit integer loads are narrowed
before reaching RegBankSelect.
Florian Hahn [Wed, 20 Dec 2017 11:13:57 +0000 (11:13 +0000)]
[ARM] Lower unsigned saturation to USAT
Summary:
Implement lower of unsigned saturation on an interval [0, k] where k + 1 is a power of two using USAT instruction in a similar way to how [~k, k] is lowered using SSAT on ARM models that supports it.
Sander de Smalen [Wed, 20 Dec 2017 11:02:42 +0000 (11:02 +0000)]
[AArch64][SVE] Re-submit patch series for ZIP1/ZIP2
This patch resubmits the SVE ZIP1/ZIP2 patch series consisting of
of r320992, r320986, r320973, and r320970 by reverting
https://reviews.llvm.org/rL321024.
The issue that caused r321024 has been addressed in https://reviews.llvm.org/rL321158,
so this patch-series should be safe to resubmit.
Bjorn Steinbrink [Wed, 20 Dec 2017 10:01:30 +0000 (10:01 +0000)]
Give up on array allocas in getPointerDereferenceableBytes
Summary:
As suggested by Eli Friedman, don't try to handle array allocas here,
because of possible overflows, instead rely on instcombine converting
them to allocations of array types.
Sam Parker [Wed, 20 Dec 2017 08:56:41 +0000 (08:56 +0000)]
[AArch64] CCSIDR2 system register
Implement the 'Current Cache Size' register that has been introduced
as part of the Armv8.3 architecture. I originally missed this, and
(hopefully) should be the final patch for assembler support.
projects / llvm / log