If the input vector to INSERT_SUBVECTOR is another INSERT_SUBVECTOR, and this inserted subvector replaces the last insertion, then insert into the common source vector.
i.e.
INSERT_SUBVECTOR( INSERT_SUBVECTOR( Vec, SubOld, Idx ), SubNew, Idx ) --> INSERT_SUBVECTOR( Vec, SubNew, Idx )
Differential Revision: https://reviews.llvm.org/D23330
llvm-svn: 278211
Created a Thumb2 predicated pattern matcher that uses Thumb2 and
HasT2ExtractPack and used it to redefine the patterns for sxta{b|h}
and uxta{b|h}. Also used the similar patterns to fill in isel pattern
gaps for the corresponding instructions in the ARM backend.
The patch is mainly changes to tests since most of this functionality
appears not to have been tested.
Differential Revision: https://reviews.llvm.org/D23273
llvm-svn: 278207
This patch adds -emscripten-cxx-exceptions-whitelist option to
WebAssemblyLowerEmscriptenExceptions pass. This options is the list of
function names in which Emscripten-style exception handling is enabled.
This is to support emscripten's EXCEPTION_CATCHING_WHITELIST which
exists because of the performance impact of emscripten's non-zero-cost
EH method.
Patch by Heejin Ahn
Differential Revision: https://reviews.llvm.org/D23292
llvm-svn: 278171
For now put them all in the entry block. This should be correct but may give
poor runtime performance. Hopefully MachineSinking combined with
isReMaterializable can solve those issues, but if not the interface is sound
enough to support alternatives.
llvm-svn: 278168
A UD2 might make its way into the program via a call to @llvm.trap.
Obviously, calls are not terminators. However, we modeled the X86
instruction, UD2, as a terminator. Later on, this confuses the epilogue
insertion machinery which results in the epilogue getting inserted
before the UD2. For some platforms, like x64, the result is a
violation of the ABI.
Instead, model UD2/UD2B as a side effecting instruction which may
observe memory.
llvm-svn: 278144
As detailed on D22726, much of the shift combining code assume constant values will fit into a uint64_t value and calls ConstantSDNode::getZExtValue where it probably shouldn't (leading to asserts). Using APInt directly avoids this problem but we encounter other assertions if we attempt to compare/operate on 2 APInt of different bitwidths.
This patch adds a helper function to ensure that 2 APInt values are zero extended as required so that they can be safely used together. I've only added an initial example use for this to the '(SHIFT (SHIFT x, c1), c2) --> (SHIFT x, (ADD c1, c2))' combines. Further cases can easily be added as required.
Differential Revision: https://reviews.llvm.org/D23007
llvm-svn: 278141
This makes a trivial change in the emission of the per-function XRay
tables, and makes sure that the xray_instr_map section does show up in
the object file.
llvm-svn: 278113
We only had partial memory folding support for the intrinsic definitions, and (as noted on PR27481) was causing FR32/FR64/VR128 mismatch errors with the machine verifier.
This patch adds missing memory folding support for both intrinsics and the ffloor/fnearbyint/fceil/frint/ftrunc patterns and in doing so fixes the failing machine verifier stack folding tests from PR27481.
Differential Revision: https://reviews.llvm.org/D23276
llvm-svn: 278106
* Delete extra '_' prefixes from JS library function names. fixImports()
function in JS glue code deals with this for wasm.
* Change command-line option names in order to be consistent with
asm.js.
* Add missing lowering code for llvm.eh.typeid.for intrinsics
* Delete commas in mangled function names
* Fix a function argument attributes bug. Because we add the pointer to
the original callee as the first argument of invoke wrapper, all
argument attribute indices have to be incremented by one.
Patch by Heejin Ahn
Differential Revision: https://reviews.llvm.org/D23258
llvm-svn: 278081
This reverts commit r278048. Something changed between the last time I
built this--it takes awhile on my ridiculously slow and ancient
computer--and now that broke this.
llvm-svn: 278053
Summary:
Based on two patches by Michael Mueller.
This is a target attribute that causes a function marked with it to be
emitted as "hotpatchable". This particular mechanism was originally
devised by Microsoft for patching their binaries (which they are
constantly updating to stay ahead of crackers, script kiddies, and other
ne'er-do-wells on the Internet), but is now commonly abused by Windows
programs to hook API functions.
This mechanism is target-specific. For x86, a two-byte no-op instruction
is emitted at the function's entry point; the entry point must be
immediately preceded by 64 (32-bit) or 128 (64-bit) bytes of padding.
This padding is where the patch code is written. The two byte no-op is
then overwritten with a short jump into this code. The no-op is usually
a `movl %edi, %edi` instruction; this is used as a magic value
indicating that this is a hotpatchable function.
Reviewers: majnemer, sanjoy, rnk
Subscribers: dberris, llvm-commits
Differential Revision: https://reviews.llvm.org/D19908
llvm-svn: 278048
This patch adds support for some new relocation models to the ARM
backend:
* Read-only position independence (ROPI): Code and read-only data is accessed
PC-relative. The offsets between all code and RO data sections are known at
static link time. This does not affect read-write data.
* Read-write position independence (RWPI): Read-write data is accessed relative
to the static base register (r9). The offsets between all writeable data
sections are known at static link time. This does not affect read-only data.
These two modes are independent (they specify how different objects
should be addressed), so they can be used individually or together. They
are otherwise the same as the "static" relocation model, and are not
compatible with SysV-style PIC using a global offset table.
These modes are normally used by bare-metal systems or systems with
small real-time operating systems. They are designed to avoid the need
for a dynamic linker, the only initialisation required is setting r9 to
an appropriate value for RWPI code.
I have only added support to SelectionDAG, not FastISel, because
FastISel is currently disabled for bare-metal targets where these modes
would be used.
Differential Revision: https://reviews.llvm.org/D23195
llvm-svn: 278015
Summary:
The DAG combine transformation that was generating the
aarch64_neon_vcvtfp2fxs node was assuming that all
inputs where legal and wasn't accounting that the input
could be a v4f64 if we're trying to do the transformation
before legalization. We now bail out in this case.
All illegal types besides v4f64 were already rejected.
Fixes https://llvm.org/bugs/show_bug.cgi?id=28877.
Reviewers: jmolloy
Subscribers: aemerson, rengolin, llvm-commits
Differential Revision: https://reviews.llvm.org/D23261
llvm-svn: 278002
Optimized lowering of BITCAST node. The BITCAST node can be replaced with COPY_TO_REG instead of KMOV.
It allows to suppress two opposite BITCAST operations and avoid redundant "movs".
Differential Revision: https://reviews.llvm.org/D23247
llvm-svn: 277958
This is a new test that should explore a current suboptimal sequence in passing values between cmp and kor intrinsics.
The code will be optimized in an upcoming patch.
Submitted bug here:
https://llvm.org/bugs/show_bug.cgi?id=28839
llvm-svn: 277954
Assuming SSE2 is available then we can safely commute between these, removing some unnecessary register moves and improving memory folding opportunities.
VEX encoded versions don't benefit so I haven't added support to them.
llvm-svn: 277930
Summary:
This is the setting of the Vulkan closed source driver.
It decreases the max wave count from 10 to 8.
26010 shaders in 14650 tests
Totals:
VGPRS: 829593 -> 808440 (-2.55 %)
Spilled SGPRs: 81878 -> 42226 (-48.43 %)
Spilled VGPRs: 367 -> 358 (-2.45 %)
Scratch VGPRs: 1764 -> 1748 (-0.91 %) dwords per thread
Code Size: 36677864 -> 35923932 (-2.06 %) bytes
There is a massive decrease in SGPR spilling in general and -7.4% spilled
VGPRs for DiRT Showdown (= SGPRs spilled to scratch?)
Reviewers: arsenm, tstellarAMD, nhaehnle
Subscribers: arsenm, llvm-commits, kzhuravl
Differential Revision: https://reviews.llvm.org/D23034
llvm-svn: 277867
Summary: Thumb2 supports encoding immediates with specific patterns into mov.w by splatting the low 8 bits into other bytes.
I'm resubmitting this patch. The test case in the original commit
r277610 does not specify triple, so builds with differnt default triple
will have different output.
This patch fixed trile as thumb-darwin-apple.
Reviewers: john.brawn, jmolloy, bruno
Subscribers: jmolloy, aemerson, rengolin, samparker, llvm-commits
Differential Revision: https://reviews.llvm.org/D23090
llvm-svn: 277865
There were two locations where fast-isel would generate a LFD instruction
with a target register class VSFRC instead of F8RC when VSX was enabled.
This can ccause invalid registers to be used in certain cases, like:
lfd 36, ...
instead of using a VSX load instruction. The wrong register number gets
silently truncated, causing invalid code to be generated.
The first place is PPCFastISel::PPCEmitLoad, which had multiple problems:
1.) The IsVSSRC and IsVSFRC flags are not initialized correctly, since they
are computed from resultReg, which is still zero at this point in many cases.
Fixed by changing the helper routines to operate on a register class instead
of a register and passing in UseRC.
2.) Even with this fixed, Is64VSXLoad is still wrong due to a typo:
bool Is32VSXLoad = IsVSSRC && Opc == PPC::LFS;
bool Is64VSXLoad = IsVSSRC && Opc == PPC::LFD;
The second line needs to use isVSFRC (like PPCEmitStore does).
3.) Once both the above are fixed, we're now generating a VSX instruction --
but an incorrect one, since generation of an indexed instruction with null
index is wrong. Fixed by copying the code handling the same issue in
PPCEmitStore.
The second place is PPCFastISel::PPCMaterializeFP, where we would emit an
LFD to load a constant from the literal pool, and use the wrong result
register class. Fixed by hardcoding a F8RC class even on systems
supporting VSX.
Fixes: https://llvm.org/bugs/show_bug.cgi?id=28630
Differential Revision: https://reviews.llvm.org/D22632
llvm-svn: 277823
This patch fixes passing long double type arguments to function in
soft float mode. If there is less than 4 argument registers free
(long double type is mapped in 4 gpr registers in soft float mode)
long double type argument must be passed through stack.
Differential Revision: https://reviews.llvm.org/D20114.
llvm-svn: 277804
These are the operations that are trivially identical. Division is omitted for
now because you need to use the correct sign/zero extension.
llvm-svn: 277775
Adding missing tests for OCL type names for half, float, double, char, short, long, and unknown.
Patch by Aaron En Ye Shi.
Differential Revision: https://reviews.llvm.org/D22964
llvm-svn: 277759
Previously, FastISel for WebAssembly wasn't checking the return value of
`getRegForValue` in certain cases, which would generate instructions
referencing NoReg. This patch fixes this behavior.
Patch by Dominic Chen
Differential Revision: https://reviews.llvm.org/D23100
llvm-svn: 277742
On modern Intel processors hardware SQRT in many cases is faster than RSQRT
followed by Newton-Raphson refinement. The patch introduces a simple heuristic
to choose between hardware SQRT instruction and Newton-Raphson software
estimation.
The patch treats scalars and vectors differently. The heuristic is that for
scalars the compiler should optimize for latency while for vectors it should
optimize for throughput. It is based on the assumption that throughput bound
code is likely to be vectorized.
Basically, the patch disables scalar NR for big cores and disables NR completely
for Skylake. Firstly, scalar SQRT has shorter latency than NR code in big cores.
Secondly, vector SQRT has been greatly improved in Skylake and has better
throughput compared to NR.
Differential Revision: https://reviews.llvm.org/D21379
llvm-svn: 277725
Enable tail calls by default for (micro)MIPS(64).
microMIPS is slightly more tricky than doing it for MIPS(R6) or microMIPSR6.
microMIPS has two instruction encodings: 16bit and 32bit along with some
restrictions on the size of the instruction that can fill the delay slot.
For safe tail calls for microMIPS, the delay slot filler attempts to find
a correct size instruction for the delay slot of TAILCALL pseudos.
Reviewers: dsanders, vkalintris
Subscribers: jfb, dsanders, sdardis, llvm-commits
Differential Revision: https://reviews.llvm.org/D21138
llvm-svn: 277708
This should ensure that we can atomically write two bytes (on top of the
retq and the one past it) and have those two bytes not straddle cache
lines.
We also move the label past the alignment instruction so that we can refer
to the actual first instruction, as opposed to potential padding before the
aligned instruction.
Update the tests to allow us to reflect the new order of assembly.
Reviewers: rSerge, echristo, majnemer
Subscribers: llvm-commits, mehdi_amini
Differential Revision: https://reviews.llvm.org/D23101
llvm-svn: 277701
This just tests that the register limit isn't exceeded,
so the regisetr allocation doesn't need to be great.'
The critically slow part is all in greedy RA, so
switch to basic.
llvm-svn: 277700
rewriteOperands() always performed liveness queries at the base index
rather than the RegSlot/Base as apropriate for the machine operand. This
could lead to illegal rewriting in some cases.
llvm-svn: 277661
This patch fixes pr25548.
Current implementation of PPCBoolRetToInt doesn't handle CallInst correctly, so it failed to do the intended optimization when there is a CallInst with parameters. This patch fixed that.
llvm-svn: 277655
We currently only support combining target shuffles that consist of a single source input (plus elements known to be undef/zero).
This patch generalizes the recursive combining of the target shuffle to collect all the inputs, merging any duplicates along the way, into a full set of src ops and its shuffle mask.
We uncover a number of cases where we have failed to combine a unary shuffle because the input has been duplicated and separated during lowering.
This will allow us to combine to 2-input shuffles in a future patch.
Differential Revision: https://reviews.llvm.org/D22859
llvm-svn: 277631
Currently we have a number of tests that fail with -verify-machineinstrs.
To detect this cases earlier we add the option to the testcases with the
exception of tests that will currently fail with this option. PR 27456 keeps
track of this failures.
No code review, as discussed with Hal Finkel.
llvm-svn: 277624
Summary: Thumb2 supports encoding immediates with specific patterns into mov.w by splatting the low 8 bits into other bytes.
Reviewers: john.brawn, jmolloy
Subscribers: jmolloy, aemerson, rengolin, samparker, llvm-commits
Differential Revision: https://reviews.llvm.org/D23090
llvm-svn: 277610
When expanding FP constants, we attempt to shrink doubles to floats and perform an extending load.
However, on SystemZ, and possibly on other targets (I've only confirmed the problem on SystemZ), the FP extending load instruction may convert SNaN into QNaN, or may cause an exception. So in the general case, we would still like to shrink FP constants, but SNaNs should be left as doubles.
Differential Revision: https://reviews.llvm.org/D22685
llvm-svn: 277602
When the same base address is used to load two different data types, LSR
would assume a memory type of "void". This type is not sized and has no
alignment information. Checking for it causes a crash.
llvm-svn: 277601
Summary:
We also add a test to show what currently happens when we create a
section per function and emit an xray_instr_map. This illustrates the
relationship (or lack thereof) between the per-function section and the
xray_instr_map section.
We also change the code generation slightly so that we don't always
create group sections, but rather only do so if a function where the
table is associated with is in a group.
Also in this change:
- Remove the "merge" flag on the xray_instr_map section.
- Test that we're generating the right table for comdat and non-comdat functions.
Reviewers: echristo, majnemer
Subscribers: llvm-commits, mehdi_amini
Differential Revision: https://reviews.llvm.org/D23104
llvm-svn: 277580
In this particular example we wouldn't want the smmls anyway (the value is
actually unused), but in general smmls does not provide the required flags
register so if that SUBE result is used we can't replace it.
llvm-svn: 277541
We were relying on the misleadingly-names $status result to actually be the
status. Actually it's just a scratch register that may or may not be valid (and
is the inverse of the real ststus anyway). Success can be determined by
comparing the value loaded against the one we wanted to see for "cmpxchg
strong" loops like this.
Should fix PR28819.
llvm-svn: 277513
Summary:
Two types of stores are possible in pixel shaders: stores to memory that are
explicitly requested at the API level, and stores that are an implementation
detail of register spilling or lowering of arrays.
For the first kind of store, we must ensure that helper pixels have no effect
and hence WQM must be disabled. The second kind of store must always be
executed, because the written value may be loaded again in a way that is
relevant for helper pixels as well -- and there are no externally visible
effects anyway.
This is a candidate for the 3.9 release branch.
Reviewers: arsenm, tstellarAMD, mareko
Subscribers: arsenm, kzhuravl, llvm-commits
Differential Revision: https://reviews.llvm.org/D22675
llvm-svn: 277504
Summary:
There are cases where uniform branch conditions are computed in VGPRs, and
we didn't correctly mark those as WQM.
The stray change in basic-branch.ll is because invoking the LiveIntervals
analysis leads to the detection of a dead register that would otherwise not
be seen at -O0.
This is a candidate for the 3.9 branch, as it fixes a possible hang.
Reviewers: arsenm, tstellarAMD, mareko
Subscribers: arsenm, llvm-commits, kzhuravl
Differential Revision: https://reviews.llvm.org/D22673
llvm-svn: 277500
Identify patterns where the address is aligned to an 8-byte boundary,
but both the base address and the constant offset are both proper
multiples of 4. In such cases, extract Base+4 into a separate instruc-
tion, and use S2_storerd_io, instead of using S4_storerd_rr.
llvm-svn: 277497
I thought the directory had a lit.local.cfg, but it doesn't.
I'll add one, but for now, add the REQUIRES line. While there,
move the triple into the IR and add a datalayout.
llvm-svn: 277486
None of GlobalISel requires the property, but this lets us use the
verifier instead of rolling our own "all instructions selected" check.
llvm-svn: 277484
After instruction selection, there should be no pre-isel generic
instructions remaining, nor should generic virtual registers be
used. Verify that.
llvm-svn: 277483
Selected: the InstructionSelect pass ran and all pre-isel generic
instructions have been eliminated; i.e., all instructions are now
target-specific or non-pre-isel generic instructions (e.g., COPY).
Since only pre-isel generic instructions can have generic virtual register
operands, this also means that all generic virtual registers have been
constrained to virtual registers (assigned to register classes) and that
all sizes attached to them have been eliminated.
This lets us enforce certain invariants across passes.
This property is GlobalISel-specific, but is always available.
llvm-svn: 277482
The InstructionSelect pass assumes that RegBankSelect ran; set the
property on all tests (thereby verifying the test inputs) and require
it in the pass.
llvm-svn: 277477
RegBankSelected: the RegBankSelect pass ran and all generic virtual
registers have been assigned to a register bank.
This lets us enforce certain invariants across passes.
This property is GlobalISel-specific, but is always available.
llvm-svn: 277475
RegBankSelect and InstructionSelect run after the legalizer and
require a Legalized function: check that all instructions are legal.
Note that this should be in the MachineVerifier, but it can't use the
MachineLegalizer as it's currently in the separate GlobalISel library.
Note that the RegBankSelect verifier checks have the same layering
problem, but we only use inline methods so end up not needing to link
against the GlobalISel library.
llvm-svn: 277472
Legalized: The MachineLegalizer ran; all pre-isel generic instructions
have been legalized, i.e., all instructions are now one of:
- generic and always legal (e.g., COPY)
- target-specific
- legal pre-isel generic instructions.
This lets us enforce certain invariants across passes.
This property is GlobalISel-specific, but is always available.
llvm-svn: 277470
Added (sra (shl x, 16), 16) to the sext_16_node PatLeaf for ARM to
simplify some pattern matching. This has allowed several patterns
for smul* and smla* to be removed as well as making it easier to add
the matching for the corresponding instructions for Thumb2 targets.
Also added two Pat classes that are predicated on Thumb2 with the
hasDSP flag and UseMulOps flags. Updated the smul codegen test with
the wider range of patterns plus the ThumbV6 and ThumbV6T2 targets.
Differential Revision: https://reviews.llvm.org/D22908
llvm-svn: 277450
The branch relaxation pass has the worst test coverage
of any pass in AArch64. Add a few tests that hit some
large pieces of code in the pass.
llvm-svn: 277428
Summary: This patch implements CFI for WebAssembly. It modifies the
LowerTypeTest pass to pre-assign table indexes to functions that are
called indirectly, and lowers type checks to test against the
appropriate table indexes. It also modifies the WebAssembly backend to
support a special ".indidx" assembly directive that propagates the table
index assignments out to the linker.
Patch by Dominic Chen
Differential Revision: https://reviews.llvm.org/D21768
llvm-svn: 277398
Summary: This patch includes asm.js-style exception handling support for
WebAssembly. The WebAssembly MVP does not have any support for
unwinding or non-local control flow. In order to support C++ exceptions,
emscripten currently uses JavaScript exceptions along with some support
code (written in JavaScript) that is bundled by emscripten with the
generated code.
This scheme lowers exception-related instructions for wasm such that
wasm modules can be compatible with emscripten's existing scheme and
share the support code.
Patch by Heejin Ahn
Differential Revision: https://reviews.llvm.org/D22958
llvm-svn: 277391
We used to combine "sext(setcc x, y, cc) -> (select (setcc x, y, cc), -1, 0)"
Instead, we should combine to (select (setcc x, y, cc), T, 0) where the value
of T is 1 or -1, depending on the type of the setcc, and getBooleanContents()
for the type if it is not i1.
This fixes PR28504.
llvm-svn: 277371
Scavenging slots were only reserved when pseudo-instruction expansion in
frame lowering created new virtual registers. It is possible to still
need a scavenging slot even if no virtual registers were created, in cases
where the stack is large enough to overflow instruction offsets.
llvm-svn: 277355
Summary:
Allocating an AFGR64 shadows two GPR32's instead of just one.
This fixes an LNT regression detected by our internal buildbots.
Reviewers: sdardis
Subscribers: dsanders, sdardis, llvm-commits
Differential Revision: https://reviews.llvm.org/D23012
llvm-svn: 277348
The branch relaxation pass is computing the wrong offsets because it assumes
TLSDESC_CALLSEQ eats up 4 bytes, when in fact it is lowered to an instruction
sequence taking up 16 bytes. This can become a problem in huge files with lots
of TLS accesses, as it may slowly move branch targets out of the range computed
by the branch relaxation pass.
Fixes PR24234 https://llvm.org/bugs/show_bug.cgi?id=24234
Differential Revision: https://reviews.llvm.org/D22870
llvm-svn: 277331
Initialize all AArch64-specific passes in the TargetMachine so they can be run
by llc. This can lead to conflicts in opt with some command line options that
share the same name as the pass, so I took this opportunity to do some cleanups:
* rename all relevant command line options from "aarch64-blah" to
"aarch64-enable-blah" and update the tests accordingly
* run clang-format on their declarations
* move all these declarations to a common place (the TargetMachine) as opposed
to having them scattered around (AArch64BranchRelaxation and
AArch64AddressTypePromotion were the only offenders)
llvm-svn: 277322
As discussed on PR14593, this patch adds support for lowering to SHLD/SHRD from the patterns generated by DAGTypeLegalizer::ExpandShiftWithKnownAmountBit.
Differential Revision: https://reviews.llvm.org/D23000
llvm-svn: 277299
Summary:
When performing cmp for EQ/NE and the operand is sign extended, we can
avoid the truncaton if the bits to be tested are no less than origianl
bits.
Reviewers: eli.friedman
Subscribers: eli.friedman, aemerson, nemanjai, t.p.northover, llvm-commits
Differential Revision: https://reviews.llvm.org/D22933
llvm-svn: 277252
These come in two variants for now: G_INTRINSIC and G_INTRINSIC_W_SIDE_EFFECTS.
We may decide to split the latter up with finer-grained restrictions later, if
necessary.
llvm-svn: 277224
Up until now, we only had code to match PSADBW patterns that look like what
comes out of the loop vectorizer - a partial reduction inside the loop body
that gets fed into a horizontal operation in a different basic block.
This adds support for straight-line patterns, like those generated by the
SLP vectorizer.
Differential Revision: https://reviews.llvm.org/D22889
llvm-svn: 277219
Support for lowering to VBROADCASTF128 etc. in D22460 was not correctly ensuring that the only users of the 128-bit vector load were the insertions of the vector into the lower/upper subvectors.
llvm-svn: 277214
Patch by Sunita Marathe
Third try, now following fixes to MSan to handle mempcy in such a way that this commit won't break the MSan buildbots. (Thanks, Evegenii!)
llvm-svn: 277189
The following pattern was being layed out poorly:
A
/ \
B C
/ \ / \
D E ? (Doesn't matter)
Where A->B is far more likely than A->C, and prob(B->D) = prob(B->E)
The current algorithm gives:
A,B,C,E (D goes on worklist)
It does this even if C has a frequency count of 0. This patch
adjusts the layout calculation so that if freq(B->E) >> freq(C->E)
then we go ahead and layout E rather than C. Fallthrough half the time
is better than fallthrough never, or fallthrough very rarely. The
resulting layout is:
A,B,E, (C and D are in a worklist)
llvm-svn: 277187
The DAG combiner will try to merge consecutive stores into a bigger
store, unless the resulting store is not fast. Misaligned vector stores
are allowed on Hexagon, but are not fast. Add a testcase to make sure
this type of merging does not occur.
Patch by Pranav Bhandarkar.
llvm-svn: 277182
The DAG combiner tries to merge stores to adjacent vector wide memory
locations by creating stores which are integral multiples of the vector
width. Discourage this by informing it that this is slow. This should
not affect legalization passes, because all of them ignore the "Fast"
argument.
Patch by Pranav Bhandarkar.
llvm-svn: 277178
Mostly straightforward as we ignore addressing modes and just
use the base + unsigned immediate offset (always 0) variants.
This currently fails to select extloads because we have yet to
agree on a representation.
llvm-svn: 277171
Software pipelining is an optimization for improving ILP by
overlapping loop iterations. Swing Modulo Scheduling (SMS) is
an implementation of software pipelining that attempts to
reduce register pressure and generate efficient pipelines with
a low compile-time cost.
This implementaion of SMS is a target-independent back-end pass.
When enabled, the pass should run just prior to the register
allocation pass, while the machine IR is in SSA form. If the pass
is successful, then the original loop is replaced by the optimized
loop. The optimized loop contains one or more prolog blocks, the
pipelined kernel, and one or more epilog blocks.
This pass is enabled for Hexagon only. To enable for other targets,
a couple of target specific hooks must be implemented, and the
pass needs to be called from the target's TargetMachine
implementation.
Differential Review: http://reviews.llvm.org/D16829
llvm-svn: 277169
If the mask of a vector shuffle has alternating odd or even numbers
starting with 1 or 0 respectively up to the largest possible index
for the given type in the given HVX mode (single of double) we can
generate vpacko or vpacke instruction respectively.
E.g.
%42 = shufflevector <32 x i16> %37, <32 x i16> %41,
<32 x i32> <i32 1, i32 3, ..., i32 63>
is %42.h = vpacko(%41.w, %37.w)
Patch by Pranav Bhandarkar.
llvm-svn: 277168
Rebalances address calculation trees and applies Hexagon-specific
optimizations to the trees to improve instruction selection.
Patch by Tobias Edler von Koch.
llvm-svn: 277151
Summary:
Implements fastLowerArguments() to avoid the need to fall back on
SelectionDAG for 0-4 argument functions that don't do tricky things like
passing double in a pair of i32's.
This allows us to move all except one test to -fast-isel-abort=3. The
remaining one has function prototypes of the form 'i32 (i32, double, double)'
which requires floats to be passed in GPR's.
The previous commit had an uninitialized variable that caused the incoming
argument region to have undefined size. This has been fixed.
Reviewers: sdardis
Subscribers: dsanders, llvm-commits, sdardis
Differential Revision: https://reviews.llvm.org/D22680
llvm-svn: 277136
We currently default to using either generic shuffles or MASK+PACKUS/PACKSS to truncate all integer vectors. For vector comparisons, we know that the result will be either all or zero bits in every element, which can be efficiently truncated by directly using PACKSS to repeatedly halve the size of each element.
Due to the limited input values (-1 or 0) we don't need to account for vector element size, so for simplicity we just use the PACKSS(vXi16,vXi16) implementation in all cases. Additionally for AVX2 PACKSS of 256bit data we must perform a PERMQ shuffle to reorder the data into the correct order. I did investigate performing a single shuffle after all the PACKSS calls but the need to cross 128bit lanes makes this difficult to achieve efficiently.
We avoid performing this on AVX512 as it should have better alternative truncation instructions.
Differential Revision: https://reviews.llvm.org/D22814
llvm-svn: 277132
Summary:
The MOV/MOVT instructions being chosen for struct_byval predicates was
conditional only on Thumb2, resulting in an ARM MOV/MOVT instruction
being incorrectly emitted in Thumb1 mode. This is especially apparent
with v8-m.base targets. This patch ensures that Thumb instructions are
emitted in both Thumb modes.
Reviewers: rengolin, t.p.northover
Subscribers: llvm-commits, aemerson, rengolin
Differential Revision: https://reviews.llvm.org/D22865
llvm-svn: 277128
I'm not convinced the patterns for the rm_Int was correct anyway. It had a tied source that should't exist for the unmasked version. The load form of MOVSS always zeros the most significant bits. I've left the patterns off the masked load instructions as I'm not sure what the correct pattern should be and we don't have any tests currently. Nor do we implement masked scalar load intrinsics in clang currently.
llvm-svn: 277098
Normally, CFI instructions should be inserted after allocframe, but
if allocframe is in the same packet with a call, the CFI instructions
should be inserted before that packet.
llvm-svn: 277020
LLT() has a particular meaning: it's one invalid type. But we really
want selected instructions to have no type whatsoever.
Also verify that types don't linger after ISel, and enable the verifier
on the AArch64 select test.
llvm-svn: 277001
Summary:
Implements fastLowerArguments() to avoid the need to fall back on
SelectionDAG for 0-4 argument functions that don't do tricky things like
passing double in a pair of i32's.
This allows us to move all except one test to -fast-isel-abort=3. The
remaining one has function prototypes of the form 'i32 (i32, double, double)'
which requires floats to be passed in GPR's.
Reviewers: sdardis
Subscribers: dsanders, llvm-commits, sdardis
Differential Revision: https://reviews.llvm.org/D22680
llvm-svn: 276982
Summary:
SI_ELSE is lowered into two parts:
s_or_saveexec_b64 dst, src (at the start of the basic block)
s_xor_b64 exec, exec, dst (at the end of the basic block)
The idea is that dst contains the exec mask of the preceding IF block. It can
happen that SIWholeQuadMode decides to switch from WQM to Exact mode inside
the basic block that contains SI_ELSE, in which case it introduces an instruction
s_and_b64 exec, exec, s[...]
which masks out bits that can correspond to both the IF and the ELSE paths.
So the resulting sequence must be:
s_or_savexec_b64 dst, src
s_and_b64 exec, exec, s[...] <-- added by SIWholeQuadMode
s_and_b64 dst, dst, exec <-- added by SILowerControlFlow
s_xor_b64 exec, exec, dst
Whether to add the additional s_and_b64 dst, dst, exec is currently determined
via the ExecModified tracking. With this change, it is instead determined by
an additional flag on SI_ELSE which is set by SIWholeQuadMode.
Finally: It also occured to me that an alternative approach for the long run
is for SILowerControlFlow to unconditionally emit
s_or_saveexec_b64 dst, src
...
s_and_b64 dst, dst, exec
s_xor_b64 exec, exec, dst
and have a pass that detects and cleans up the "redundant AND with exec"
pattern where possible. This could be useful anyway, because we also add
instructions
s_and_b64 vcc, exec, vcc
before s_cbranch_scc (in moveToALU), and those are often redundant. I have
some pending changes to how KILL is lowered that could also benefit from
such a cleanup pass.
In any case, this current patch could help in the short term with the whole
ExecModified business.
Reviewers: tstellarAMD, arsenm
Subscribers: arsenm, llvm-commits, kzhuravl
Differential Revision: https://reviews.llvm.org/D22846
llvm-svn: 276972
Before adding a new preheader block, check if there is a candidate block
where the loop setup could be placed speculatively. This will be off by
default.
llvm-svn: 276919
Using getZExtValue() will assert if the value doesn't fit into uint64_t - SHL was already doing this, I've just updated ASHR/LSHR to match
As mentioned on D22726
llvm-svn: 276855
Change the syntax to use `%0.sub8` to denote a subregister.
This seems like a more natural fit to denote subregisters; I also plan
to introduce a new ":classname" syntax in upcoming patches to denote the
register class of a vreg.
Note that this commit disallows plain identifiers to start with a '.'
character. This shouldn't affect anything as external names/IR
references are all prefixed with '$'/'%', plain identifiers are only
used for instruction names, register mask names and subreg indexes.
Differential Revision: https://reviews.llvm.org/D22390
llvm-svn: 276815
Consider this case:
vreg1 = A2_zxth vreg0 (1)
...
vreg2 = A2_zxth vreg1 (2)
Redundant instruction elimination could delete the instruction (1)
because the user (2) only cares about the low 16 bits. Then it could
delete (2) because the input is already zero-extended. The problem
is that the properties allowing each individual instruction to be
deleted depend on the existence of the other instruction, so either
one can be deleted, but not both.
The existing check for this situation in RIE was insufficient. The
fix is to update all dependent cells when an instruction is removed
(replaced via COPY) in RIE.
llvm-svn: 276792
In an instruction like:
CFI_INSTRUCTION .cfi_def_cfa ...
we can drop the '.cfi_' prefix since that should be obvious by the
context:
CFI_INSTRUCTION def_cfa ...
While being a terser and cleaner syntax this also prepares to dropping
support for identifiers starting with a dot character so we can use it
for expressions.
Differential Revision: http://reviews.llvm.org/D22388
llvm-svn: 276785
Fixed typo in the intrinsic definitions of (v)cvtsd2ss with memory folding.
This was only unearthed when rL276102 started using the intrinsic again.....
llvm-svn: 276740
MIPS64R6 compact branch support. As the MIPS LLVM backend uses distinct
MachineInstrs for certain 32 and 64 bit instructions (e.g. BEQ & BEQ64) that
map to the same instruction, extend compact branch support for the
corresponding 64bit branches.
Reviewers: dsanders
Differential Revision: https://reviews.llvm.org/D20164
llvm-svn: 276739
The saturation instructions appeared in v6T2, with DSP extensions, but they
were being accepted / generated on any, with the new introduction of the
saturation detection in the back-end. This commit restricts the usage to
DSP-enable only cores.
Fixes PR28607.
llvm-svn: 276701
This adds LLVM's 3 main cast instructions (inttoptr, ptrtoint, bitcast) to the
IRTranslator. The first two are direct translations (with 2 MachineInstr types
each). Since LLT discards information, a bitcast might become trivial and we
emit a COPY in those cases instead.
llvm-svn: 276690
They're basically i64 for AArch64, but we'll leave them intact for stranger
targets. Also add some tests for the (very few) other cases we can handle right
now.
llvm-svn: 276689
Avoid MipsAnalyzeImmediate usage if the constant fits in an 32-bit
integer. This allows us to generate the same instructions for the
materialization of the same constants regardless the width of their
type.
Patch by: Vasileios Kalintiris
Contributions by: Simon Dardis
Reviewers: Daniel Sanders
Differential Review: https://reviews.llvm.org/D21689
llvm-svn: 276628
This places the 132/213/231 form number in front of the SS/SD/PS/PD. Move the Y for 256-bit versions to be after the PS/PD. Change the AVX512 scalar forms to include a Z in the their name. This new format should be consistent with the general naming of instructions.
llvm-svn: 276559
This reverts commit r276298.
Data stored in .rodata can have a negative offset from .text, but we
don't support negative values in relocations yet.
This caused a regression in one of the amp conformance tests:
5_Data_Cont/5_2_a_v/5_2_3_m/Assignment/Test.02.01
llvm-svn: 276498
This adds the actual MachineLegalizeHelper to do the work and a trivial pass
wrapper that legalizes all instructions in a MachineFunction. Currently the
only transformation supported is splitting up a vector G_ADD into one acting on
smaller vectors.
llvm-svn: 276461
Summary:
The llvm.invariant.start and llvm.invariant.end intrinsics currently
support specifying invariant memory objects only in the default address
space.
With this change, these intrinsics are overloaded for any adddress space
for memory objects
and we can use these llvm invariant intrinsics in non-default address
spaces.
Example: llvm.invariant.start.p1i8(i64 4, i8 addrspace(1)* %ptr)
This overloaded intrinsic is needed for representing final or invariant
memory in managed languages.
Reviewers: apilipenko, reames
Subscribers: llvm-commits
llvm-svn: 276447
The size can exceed s_movk_i32's limit, and we don't
want to use it this early since it inhibits optimizations.
This should probably be merged to the release branch.
llvm-svn: 276438
An extension of D19978, this patch replaces the default BITREVERSE evaluation of individual bit masks+shifts with block mask+shifts when we have integer elements of power-of-2 bits in size.
After calling BSWAP to reverse the order of the constituent bytes (which typically follows a similar approach), every neighbouring 4-bits, 2-bits and finally 1-bit pairs are masked off and swapped over with shifts.
In doing so we can significantly reduce the number of operations required.
Differential Revision: https://reviews.llvm.org/D21578
llvm-svn: 276432
As reported on PR26235, we don't currently make use of the VBROADCASTF128/VBROADCASTI128 instructions (or the AVX512 equivalents) to load+splat a 128-bit vector to both lanes of a 256-bit vector.
This patch enables lowering from subvector insertion/concatenation patterns and auto-upgrades the llvm.x86.avx.vbroadcastf128.pd.256 / llvm.x86.avx.vbroadcastf128.ps.256 intrinsics to match.
We could possibly investigate using VBROADCASTF128/VBROADCASTI128 to load repeated constants as well (similar to how we already do for scalar broadcasts).
Reapplied with fix for PR28657 - removed intrinsic definitions (clang companion patch to be be submitted shortly).
Differential Revision: https://reviews.llvm.org/D22460
llvm-svn: 276416
When we failed to parse a function in the mir parser, we should abort
the whole compilation instead of continuing in a weird state. Indeed,
this was creating strange machine function passes failures that were
hard to understand, until we notice that the function actually did not
get parsed correctly!
llvm-svn: 276348
This variant is (as documented in the TD) for disassembler use only, and should
not be used in patterns - it is longer, and is broken on 64-bit.
llvm-svn: 276347
when constraint "w" is used on a 32-bit operand.
This enables compiling the following code, which used to error out in
the backend:
void foo1(int a) {
asm volatile ("sqxtn h0, %s0\n" : : "w"(a):);
}
Fixes PR28633.
llvm-svn: 276344
Summary:
The llvm.invariant.start and llvm.invariant.end intrinsics currently
support specifying invariant memory objects only in the default address space.
With this change, these intrinsics are overloaded for any adddress space for memory objects
and we can use these llvm invariant intrinsics in non-default address spaces.
Example: llvm.invariant.start.p1i8(i64 4, i8 addrspace(1)* %ptr)
This overloaded intrinsic is needed for representing final or invariant memory in managed languages.
Reviewers: tstellarAMD, reames, apilipenko
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D22519
llvm-svn: 276316
Summary:
This change also changes findMatchingInsn and
findMatchingUpdateInsnForward to take DBG_VALUE opcodes into account
when tracking register defs and uses, which could potentially inhibit
these optimizations in the presence of debug information.
Reviewers: mcrosier
Subscribers: aemerson, rengolin, mcrosier, llvm-commits
Differential Revision: https://reviews.llvm.org/D22582
llvm-svn: 276293
Under normal circumstances we prefer the higher performance MOVD to extract the 0'th element of a v8i16 vector instead of PEXTRW.
But as detailed on PR27265, this prevents the SSE41 implementation of PEXTRW from folding the store of the 0'th element. Additionally it prevents us from making use of the fact that the (SSE2) reg-reg version of PEXTRW implicitly zero-extends the i16 element to the i32/i64 destination register.
This patch only preferentially lowers to MOVD if we will not be zero-extending the extracted i16, nor prevent a store from being folded (on SSSE41).
Fix for PR27265.
Differential Revision: https://reviews.llvm.org/D22509
llvm-svn: 276289
As reported on PR26235, we don't currently make use of the VBROADCASTF128/VBROADCASTI128 instructions (or the AVX512 equivalents) to load+splat a 128-bit vector to both lanes of a 256-bit vector.
This patch enables lowering from subvector insertion/concatenation patterns and auto-upgrades the llvm.x86.avx.vbroadcastf128.pd.256 / llvm.x86.avx.vbroadcastf128.ps.256 intrinsics to match.
We could possibly investigate using VBROADCASTF128/VBROADCASTI128 to load repeated constants as well (similar to how we already do for scalar broadcasts).
Differential Revision: https://reviews.llvm.org/D22460
llvm-svn: 276281
The clearance calculation did not take into account registers defined as outputs or clobbers in inline assembly machine instructions because these register defs are implicit.
Differential Revision: http://reviews.llvm.org/D22580
llvm-svn: 276266
This patch fixes a very subtle bug in regmask calculation. Thanks to zan
jyu Wong <zyfwong@gmail.com> for bringing this to notice.
For example if CL is only clobbered than CH should not be marked
clobbered but CX, RCX and ECX should be mark clobbered. Previously for
each modified register all of its aliases are marked clobbered by
markRegClobbred() in RegUsageInfoCollector.cpp but that is wrong because
when CL is clobbered then MRI::isPhysRegModified() will return true for
CL, CX, ECX, RCX which is correct behavior but then for CX, EXC, RCX we
mark CH also clobbered as CH is aliased to CX,ECX,RCX so
markRegClobbred() is not required because isPhysRegModified already take
cares of proper aliasing register. A very simple test case has been
added to verify this change.
Please find relevant bug report here :
http://llvm.org/PR28567
Patch by Vivek Pandya <vivekvpandya@gmail.com>
Differential Revision: https://reviews.llvm.org/D22400
llvm-svn: 276235
At -O0, cmpxchg survives AtomicExpand: it's mostly straightforward
to select it in fast-isel, and let the pseudo be expanded later.
extractvalues on the result are the tricky part: the generic logic
only works for legal types (and it would be painful to make it
support illegal types), so we can only support i32/i64 cmpxchg.
llvm-svn: 276183
This should be all the low-level instruction selection needs to determine how
to implement an operation, with the remaining context taken from the opcode
(e.g. G_ADD vs G_FADD) or other flags not based on type (e.g. fast-math).
llvm-svn: 276158
Avoid unnecessary spills of byval arguments of device functions to
local space on SASS level and subsequent pointer conversion to generic
address space that follows. Instead, make a local copy in IR, provide
a way to access arguments directly, and let LLVM optimize the copy away
when possible.
Differential Review: https://reviews.llvm.org/D21421
llvm-svn: 276153
Retry r275776 (no changes, we suspect the issue was with another commit).
The current logic for handling inline asm operands in DAGToDAGISel interprets
the operands by looking for constants, which should represent the flags
describing the kind of operand we're dealing with (immediate, memory, register
def etc). The operands representing actual data are skipped only if they are
non-const, with the exception of immediate operands which are skipped explicitly
when a flag describing an immediate is found.
The oversight is that memory operands may be const too (e.g. for device drivers
reading a fixed address), so we should explicitly skip the operand following a
flag describing a memory operand. If we don't, we risk interpreting that
constant as a flag, which is definitely not intended.
Fixes PR26038
Differential Revision: https://reviews.llvm.org/D22103
llvm-svn: 276101
Inference of the 'returned' attribute was fixed in r276008, lets try
turning the backend support back on.
This reverts commit r275677.
llvm-svn: 276081
Reverting this commit for now as it seems to be causing failures on
test-suite tests on the clang-ppc64le-linux-lnt bot.
This reverts commit r276044.
llvm-svn: 276068
If 2.5 ulp is acceptable, denormals are not required, and
isn't a reciprocal which will already be handled, replace
with a faster fdiv.
Simplify the lowering tests by using per function
subtarget features.
llvm-svn: 276051
This is a variant of scavengeRegister() that works for
enterBasicBlockEnd()/backward(). The benefit of the backward mode is
that it is not affected by incomplete kill flags.
This patch also changes
PrologEpilogInserter::doScavengeFrameVirtualRegs() to use the register
scavenger in backwards mode.
Differential Revision: http://reviews.llvm.org/D21885
llvm-svn: 276044
Also verify that we never try to set the size of a vreg associated
to a register class.
Report an error when we encounter that in MIR. Fix a testcase that
hit that error and had a size for no reason.
llvm-svn: 276012
D20859 and D20860 attempted to replace the SSE (V)CVTTPS2DQ and VCVTTPD2DQ truncating conversions with generic IR instead.
It turns out that the behaviour of these intrinsics is different enough from generic IR that this will cause problems, INF/NAN/out of range values are guaranteed to result in a 0x80000000 value - which plays havoc with constant folding which converts them to either zero or UNDEF. This is also an issue with the scalar implementations (which were already generic IR and what I was trying to match).
This patch changes both scalar and packed versions back to using x86-specific builtins.
It also deals with the other scalar conversion cases that are runtime rounding mode dependent and can have similar issues with constant folding.
A companion clang patch is at D22105
Differential Revision: https://reviews.llvm.org/D22106
llvm-svn: 275981
Recommitting after r274347 was reverted. This patch introduces some
classes to refactor the 3 and 4 register Thumb2 multiplication
instruction descriptions, plus improved tests for some of those
instructions.
Differential Revision: https://reviews.llvm.org/D21929
llvm-svn: 275979
As discussed on PR27654, this patch fixes the triples of a lot of aarch64 tests and enables lit tests on windows
This will hopefully help stop cases where windows developers break the aarch64 target
Differential Revision: https://reviews.llvm.org/D22191
llvm-svn: 275973
Summary:
N32 and N64 follow the standard ELF conventions (.L) whereas O32 uses its own
($).
This fixes the majority of object differences between -fintegrated-as and
-fno-integrated-as.
Reviewers: sdardis
Subscribers: dsanders, sdardis, llvm-commits
Differential Revision: https://reviews.llvm.org/D22412
llvm-svn: 275967
The following condition expression ( a >> n) & 1 is converted to "bt a, n" instruction. It works on all intel targets.
But on AVX-512 it was broken because the expression is modified to (truncate (a >>n) to i1).
I added the new sequence (truncate (a >>n) to i1) to the BT pattern.
Differential Revision: https://reviews.llvm.org/D22354
llvm-svn: 275950
This is to help moveSILowerControlFlow to before regalloc.
There are a couple of tradeoffs with this. The complete CFG
is visible to more passes, the loop body avoids an extra copy of m0,
vcc isn't required, and immediate offsets can be shrunk into s_movk_i32.
The disadvantage is the register allocator doesn't understand that
the single lane's vector is dead within the loop body, so an extra
register is used to outlive the loop block when expanding the
VGPR -> m0 loop. This also now results in worse waitcnt insertion
before the loop instead of after for pending operations at the point
of the indexing, but that should be fixed by future improvements to
cross block waitcnt insertion.
v_movreld_b32's operands are now modeled more correctly since vdst
is not a true output. This is kind of a hack to treat vdst as a
use operand. Extra checking is required in the verifier since
I can't seem to get tablegen to emit an implicit operand for a
virtual register.
llvm-svn: 275934
Taking address of a byval variable in PTX is legal, but currently runs
into miscompilation by ptxas on sm_50+ (NVIDIA issue 1789042).
Work around the issue by enforcing minimum alignment on byval arguments
of device functions.
The change is a no-op on SASS level for sm_3x where ptxas already aligns
local copy by at least 4.
Differential Revision: https://reviews.llvm.org/D22428
llvm-svn: 275893
This is currently only called with GEP users. A direct
alloca would only happen with current typed pointers
for arrays which are a perverse case.
Also fix crashes on 0 x and 1 x arrays.
llvm-svn: 275869
Elsewhere (particularly computeKnownBits) we assume that a global will be
aligned to the value returned by Value::getPointerAlignment. This is used to
boost the alignment on memcpy/memset, so any target-specific request can only
increase that value.
llvm-svn: 275866
DAGTypeLegalizer::CanSkipSoftenFloatOperand should allow
SELECT op code for x86_64 fp128 type for MME targets,
so SoftenFloatOperand does not abort on SELECT op code.
Differential Revision: http://reviews.llvm.org/D21758
llvm-svn: 275818
Schedule a load and its use in the same packet in MISched. Previously,
isResourceAvailable was returning false for dependences in the same
packet, which prevented MISched from packetizing a load and its use in
the same packet for v60.
Patch by Ikhlas Ajbar.
llvm-svn: 275804
This patch corresponds to review:
https://reviews.llvm.org/D21354
We use direct moves for extracting integer elements from vectors. We also use
direct moves when converting integers to FP. When these operations are chained,
we get a direct move out of a VSR followed by a direct move back into a VSR.
These are redundant - all we need to do is line up the element and convert.
llvm-svn: 275796
The machine scheduler needs to account for available resources
more accurately in order to avoid scheduling an instruction that
forces a new packet to be created.
This occurs in two ways: First, an instruction without an available
resource may have a large priority due to other metrics and be
scheduled when there are other instructions with available resources.
Second, an instruction with a non-zero latency may become available
prematurely. In both these cases, we attempt change the priority
in order to allow a better instruction to be scheduled.
Patch by Brendon Cahoon.
llvm-svn: 275793
An instruction may have multiple predecessors that are candidates
for using .cur. However, only one of them can use .cur in the
packet. When this case occurs, we need to make sure that only
one of the dependences gets a 0 latency value.
Patch by Brendon Cahoon.
llvm-svn: 275790
When SelectionDAGISel transforms a node representing an inline asm
block, memory constraint information is not preserved. This can cause
constraints to be broken when a memory offset is of the form:
offset + frame index
when the frame is resolved.
By propagating the constraints all the way to the backend, targets can
enforce memory operands of inline assembly to conform to their constraints.
For MIPSR6, some instructions had their offsets reduced to 9 bits from
16 bits such as ll/sc. This becomes problematic when using inline assembly
to perform atomic operations, as an offset can generated that is too big to
encode in the instruction.
Reviewers: dsanders, vkalintris
Differential Review: https://reviews.llvm.org/D21615
llvm-svn: 275786
Summary:
The work item intrinsics are not available for the shader
calling conventions. And even if we did hook them up most
shader stages haves some extra restrictions on the amount
of available LDS.
Reviewers: tstellarAMD, arsenm
Subscribers: nhaehnle, arsenm, llvm-commits, kzhuravl
Differential Revision: https://reviews.llvm.org/D20728
llvm-svn: 275779
The current logic for handling inline asm operands in DAGToDAGISel interprets
the operands by looking for constants, which should represent the flags
describing the kind of operand we're dealing with (immediate, memory, register
def etc). The operands representing actual data are skipped only if they are
non-const, with the exception of immediate operands which are skipped explicitly
when a flag describing an immediate is found.
The oversight is that memory operands may be const too (e.g. for device drivers
reading a fixed address), so we should explicitly skip the operand following a
flag describing a memory operand. If we don't, we risk interpreting that
constant as a flag, which is definitely not intended.
Fixes PR26038
Differential Revision: https://reviews.llvm.org/D22103
llvm-svn: 275776
At higher optimization levels, we generate the libcall for DIVREM_Ix, which is
fine: aeabi_{u|i}divmod. At -O0 we generate the one for REM_Ix, which is the
default {u}mod{q|h|s|d}i3.
This commit makes sure that we don't generate REM_Ix calls for ABIs that
don't support them (i.e. where we need to use DIVREM_Ix instead). This is
achieved by bailing out of FastISel, which can't handle non-double multi-reg
returns, and letting the legalization infrastructure expand the REM_Ix calls.
It also updates the divmod-eabi.ll test to run under -O0 as well, and adds some
Windows checks to it to make sure we don't break things for it.
Fixes PR27068
Differential Revision: https://reviews.llvm.org/D21926
llvm-svn: 275773
r275042 reverted function-attribute inference for the 'returned' attribute
because the feature triggered self-hosting failures on ARM and AArch64. James
Molloy determined that the this-return argument forwarding feature, which
directly ties the returned input argument to the returned value, was the cause.
It seems likely that this forwarding code contains, or triggers, a subtle bug.
Disabling for now until we can track that down.
llvm-svn: 275677
This does not schedule any passes besides the ones necessary to
construct and print the machine function. This is useful to test .mir
file reading and printing.
Differential Revision: http://reviews.llvm.org/D22432
llvm-svn: 275664
test/CodeGen/MIR/ARM/ARMLoadStoreDBG.mir is an actual test for the ARM
load store optimization pass and not a test of the mir parser/printer.
It belongs to test/CodeGen/ARM; This also updates the test to use the
new -run-pass llc syntax.
llvm-svn: 275662
In this situation:
%VGPR2<def> = BUFFER_LOAD_DWORD_OFFSET %SGPR8_SGPR9_SGPR10_SGPR11,
%VGPR7<def,tied3> = V_MAC_F32_e32 %VGPR0<undef>, %VGPR1<kill>, %VGPR7<kill,tied0>, %EXEC<imp-use>
%VGPR3_VGPR4_VGPR5_VGPR6<def> = COPY %VGPR0_VGPR1_VGPR2_VGPR3
%VGPR4<def> = COPY %VGPR2
The copy for VGPR1 -> VGPR4 was an error from reading undefined VGPR1,
but VGPR4 is defined immediately after this copy.
llvm-svn: 275635
Previously, we would expand:
%BL<def> = COPY %DL<kill>, %EBX<imp-use,kill>, %EBX<imp-def>
Into:
%BL<def> = MOV8rr %DL<kill>, %EBX<imp-def>
Dropping the imp-use on the floor.
That confused CriticalAntiDepBreaker, which (correctly) assumes that if an
instruction defs but doesn't use a register, that register is dead immediately
before the instruction - while in this case, the high lanes of EBX can be very
much alive.
This fixes PR28560.
Differential Revision: https://reviews.llvm.org/D22425
llvm-svn: 275634
Remove unnecessary clutter in assembly output. When using SjLj EH, the CFI is
not actually used for anything. Do not emit the CFI needlessly. The minor test
adjustments are interesting. The prologue test was just overzealous matcching.
The interesting case is the LSDA change. It was originally added to ensure that
various compilations did not mangle the name (it explicitly checked the name!).
However, subsequent cleanups made it more reliant on the CFI to find the name.
Parse the generated code flow to generically find the label still.
llvm-svn: 275614
This mostly just works.
Vectorcall rets are still not supported.
The win64_eh test change is because fast isel doesn't use rsi for temporary
computations, so it doesn't need to be pushed. The test case I'm changing was
originally added to test pushes, but by now there are other test cases in that
file exercising that code path.
https://reviews.llvm.org/D22422
llvm-svn: 275607
- Treat bitwise OR with a frame index as an ADD wherever possible, fold it
into addressing mode.
- Extend patterns for memops to allow memops with frame indexes as address
operands.
llvm-svn: 275569
The test used to rely on targeting win64 to disable fast isel,
but I'd like to teach fast isel about win64 rets. Change the
test to use varargs to disable fast isel.
llvm-svn: 275568
Added emitting metadata to elf for runtime.
Runtime requires certain information (metadata) about kernels to be able to execute and query them. Such information is emitted to an elf section as a key-value pair stream.
Differential Revision: https://reviews.llvm.org/D21849
llvm-svn: 275566
As discussed on PR28136, lowerShuffleAsRepeatedMaskAndLanePermute was attempting to match repeated masks at the 128-bit level and then permute the resultant lanes at the 128-bit (AVX1) or 64-bit (AVX2) sub-lane level.
This change allows us to create the repeated masks at the sub-lane level (and then concat them together to create a 128-bit repeated mask) and then select which sub-lane to permute. This has no effect on the AVX1 codegen.
Fixes PR28136.
llvm-svn: 275543
Thumb-1 doesn't have post-inc or pre-inc load or store instructions. However the LDM/STM instructions with writeback can function as post-inc load/store:
ldm r0!, {r1} @ load from r0 into r1 and increment r0 by 4
Obviously, this only works if the post increment is 4.
llvm-svn: 275540
... When we emit several calls to the same function in the same basic block.
An indirect call uses a "BLX r0" instruction which has a 16-bit encoding. If many calls are made to the same target, this can enable significant code size reductions.
llvm-svn: 275537
Also stop trying to insert skip blocks at end_cf. This
was inserting them at the end of the block which doesn't make
sense. The skip should be inserted at the beginning of the block
right after the end cf. Just remove this for now since no tests
seem to stress this and I think this can be handled more generally
later.
Fixes bug 28550
llvm-svn: 275510
This improves the situation discussed in D19228 where we were forcing VPERMPD/VPERMQ where VPERM2F128/VPERM2I128 would have been better.
This was incorrectly reverted in rL275421 during triage of PR28552.
llvm-svn: 275497
On Hexagon is it legal to packetize the instructions setting up call
arguments with the call instruction itself. This was already done,
except for tail calls. Make sure tail calls are handled as well.
llvm-svn: 275458
Note: I removed the checks after each jump because that's noise, but we apparently
need branches rather than returning i1 to see the bt codegen in some cases.
llvm-svn: 275439
stdcall is callee-pop like thiscall, so the thiscall changes already did most
of the work for this. This change only opts stdcall in and adds tests.
llvm-svn: 275414
This improves the situation discussed in D19228 where we were forcing VPERMPD/VPERMQ where VPERM2F128/VPERM2I128 would have been better.
llvm-svn: 275411
Summary:
It was recently discovered that, for Mips's SelectionDAGISel subclasses,
all optimization levels caused SelectionDAGISel to behave like -O2.
This change adds the necessary plumbing to initialize the optimization level.
Reviewers: andrew.w.kaylor
Subscribers: andrew.w.kaylor, sdardis, dean, llvm-commits, vradosavljevic, petarj, qcolombet, probinson, dsanders
Differential Revision: https://reviews.llvm.org/D14900
llvm-svn: 275410
Primarily this is to allow blend with zero instead of having to use vperm2f128, but we can use this in the future to deal with AVX512 cases where we need to keep the original element size to correctly fold masked operations.
llvm-svn: 275406
It wasn't actually running the pass, and since it is
missing the llvm prefix, the eh intrinsic was not
really an IntrinsicInst.
Also add missing test for lifetime markers.
llvm-svn: 275370
Summary:
In this patch we implement the following parts of XRay:
- Supporting a function attribute named 'function-instrument' which currently only supports 'xray-always'. We should be able to use this attribute for other instrumentation approaches.
- Supporting a function attribute named 'xray-instruction-threshold' used to determine whether a function is instrumented with a minimum number of instructions (IR instruction counts).
- X86-specific nop sleds as described in the white paper.
- A machine function pass that adds the different instrumentation marker instructions at a very late stage.
- A way of identifying which return opcode is considered "normal" for each architecture.
There are some caveats here:
1) We don't handle PATCHABLE_RET in platforms other than x86_64 yet -- this means if IR used PATCHABLE_RET directly instead of a normal ret, instruction lowering for that platform might do the wrong thing. We think this should be handled at instruction selection time to by default be unpacked for platforms where XRay is not availble yet.
2) The generated section for X86 is different from what is described from the white paper for the sole reason that LLVM allows us to do this neatly. We're taking the opportunity to deviate from the white paper from this perspective to allow us to get richer information from the runtime library.
Reviewers: sanjoy, eugenis, kcc, pcc, echristo, rnk
Subscribers: niravd, majnemer, atrick, rnk, emaste, bmakam, mcrosier, mehdi_amini, llvm-commits
Differential Revision: http://reviews.llvm.org/D19904
llvm-svn: 275367
This happens to make X86CallFrameOptimization in -O0 / FastISel builds as well,
but it's not clear if the pass should run in that setup.
http://reviews.llvm.org/D22314
llvm-svn: 275320
Currently the MIR framework prints all its outputs (errors and actual
representation) on stderr.
This patch fixes that by printing the regular output in the output
specified with -o.
Differential Revision: http://reviews.llvm.org/D22251
llvm-svn: 275314
We know that pcmp produces all-ones/all-zeros bitmasks, so we can use that behavior to avoid unnecessary constant loading.
One could argue that load+and is actually a better solution for some CPUs (Intel big cores) because shifts don't have the
same throughput potential as load+and on those cores, but that should be handled as a CPU-specific later transformation if
it ever comes up. Removing the load is the more general x86 optimization. Note that the uneven usage of vpbroadcast in the
test cases is filed as PR28505:
https://llvm.org/bugs/show_bug.cgi?id=28505
Differential Revision: http://reviews.llvm.org/D22225
llvm-svn: 275276
We can freeze the registers after the MachineFrameInfo has been configured (by
telling it about calls, inline asm, ...). This doesn't happen at all yet, but
will be part of IR translation.
Fixes -verify-machineinstrs assertion.
llvm-svn: 275221
This patch corresponds to review:
http://reviews.llvm.org/D20239
It adds exploitation of XXINSERTW and XXEXTRACTUW instructions that
are useful in some cases for inserting and extracting vector elements of
v4[if]32 vectors.
llvm-svn: 275215
With r274952 and r275201 in place there are no cases left where a
forward liveness analysis yields different results than a backward one.
So we can remove the forward stepping logic.
Differential Revision: http://reviews.llvm.org/D22083
llvm-svn: 275204
If a subtarget has both ZCZeroing and CustomCheapAsMoveHandling features (now
only Kryo has both), set FMOVS0 and FMOVD0 isAsCheapAsAMove.
Differential Revision: http://reviews.llvm.org/D22256
llvm-svn: 275178
This patch corresponds to review:
http://reviews.llvm.org/D21358
Vector shifts that have the same semantics as a vector swap are cannonicalized
as such to provide additional opportunities for swap removal optimization to
remove unnecessary swaps.
llvm-svn: 275168
Summary:
Previously, constant index insertelements would be turned into SI_INDIRECT_DST,
which is bound to prevent some optimization opportunities. Worse, it mislead
the heuristic that decides whether immediates should be lowered to S_MOV_B32
or V_MOV_B32 in a way that resulted in unnecessary v_readfirstlanes.
Reviewers: arsenm, tstellarAMD
Subscribers: arsenm, kzhuravl, llvm-commits
Differential Revision: http://reviews.llvm.org/D22217
llvm-svn: 275160
Summary:
Setting MIMG to 0 has a bunch of unexpected side effects, including that
isVMEM returns false which leads to incorrect treatment in the hazard
recognizer. The reason I noticed it is that it also leads to incorrect
treatment in VGPR-to-SGPR copies, which is one cause of the referenced bug.
The only reason why MIMG was set to 0 is to signal the special handling of
dmasks, but that can be checked differently.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=96877
Reviewers: arsenm, tstellarAMD
Subscribers: arsenm, kzhuravl, llvm-commits
Differential Revision: http://reviews.llvm.org/D22210
llvm-svn: 275113
Summary:
The main bug fix here is using the 32-bit encoding of V_ADD_I32 in
materializeFrameBaseRegister and resolveFrameIndex, so that arbitrary
immediates work.
The second part is that we may now require the SegmentWaveByteOffset
even when there are initially no stack objects and VGPR spilling isn't
enabled, for stack slots that are allocated later. This means that some
bits become effectively dead and can be cleaned up.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=96602
Tested-by: Kai Wasserbäch <kai@dev.carbon-project.org>
Reviewers: arsenm, tstellarAMD
Subscribers: arsenm, llvm-commits, kzhuravl
Differential Revision: http://reviews.llvm.org/D21551
llvm-svn: 275108
This bug (llvm.org/PR28124) was introduced by r237977, which refactored
the tail call sequence to be generated in two passes instead of one.
Unfortunately, the stack adjustment produced by the first pass was not
recognized by X86FrameLowering::mergeSPUpdates() in all cases, causing
code such as the following, which clobbers the return address, to be
generated:
popl %edi
popl %edi
pushl %eax
jmp tailcallee # TAILCALL
To fix the problem, the entire stack adjustment is performed in
X86ExpandPseudo::ExpandMI() for tail calls.
Patch by Magnus Lång <margnus1@gmail.com>
Differential Revision: http://reviews.llvm.org/D21325
llvm-svn: 275103
It is an optimization pass, and should not run at -O0. Especially since Fast RA
will not do the required register coalescing anyway, so it's a loss even from
the optimization standpoint.
This also works around (but doesn't quite fix) PR28489.
llvm-svn: 275099
Summary: Add support for the z13 instructions LOCHI and LOCGHI which
conditionally load immediate values. Add target instruction info hooks so
that if conversion will allow predication of LHI/LGHI.
Author: RolandF
Reviewers: uweigand
Subscribers: zhanjunl
Commiting on behalf of Roland.
Differential Revision: http://reviews.llvm.org/D22117
llvm-svn: 275086
This adds a new SystemZ-specific intrinsic, llvm.s390.tdc.f(32|64|128),
which maps straight to the test data class instructions. A new IR pass
is added to recognize instructions that can be converted to TDC and
perform the necessary replacements.
Differential Revision: http://reviews.llvm.org/D21949
llvm-svn: 275016
An identity COPY like this:
%AL = COPY %AL, %EAX<imp-def>
has no semantic effect, but encodes liveness information: Further users
of %EAX only depend on this instruction even though it does not define
the full register.
Replace the COPY with a KILL instruction in those cases to maintain this
liveness information. (This reverts a small part of r238588 but this
time adds a comment explaining why a KILL instruction is useful).
llvm-svn: 274952
Errata fixes for various errata in different versions of the Leon variants of the Sparc 32 bit processor.
The nature of the errata are listed in the comments preceding the errata fix passes. Relevant unit tests are implemented for each of these.
Note: Running clang-format has changed a few other lines too, unrelated to the implemented errata fixes. These have been left in as this keeps the code formatting consistent.
Differential Revision: http://reviews.llvm.org/D21960
llvm-svn: 274856
As a result, the urem instruction will not be expanded to a sequence of umull,
lsrs, muls and sub instructions, but just a call to __aeabi_uidivmod.
Differential Revision: http://reviews.llvm.org/D22131
llvm-svn: 274843
Until we have a better way to extract constants through bitcasted build vectors (and how to handle undefs of partial lanes etc.) at least accept build vectors that are all zeroes.
llvm-svn: 274833
Windows on ARM uses a pure thumb-2 environment. This means that it can select a
high register when doing a __builtin_longjmp. We would use a tLDRi which would
truncate the register to a low register. Use a t2LDRi12 to get the full
register file access. Tweak the code to just load into PC, as that is an
interworking branch on all supported cores anyways.
llvm-svn: 274815
Summary:
* Similiar to the ARM backend yse the peephole optimizer to generate more conditional ALU operations;
* Add predicated type with default always true to RR instructions in LanaiInstrInfo.td;
* Move LanaiSetflagAluCombiner into optimizeCompare;
* The ASM parser can currently only handle explicitly specified CC, so specify ".t" (true) where needed in the ASM test;
* Remove unused MachineOperand flags;
Reviewers: eliben
Subscribers: aemerson
Differential Revision: http://reviews.llvm.org/D22072
llvm-svn: 274807
xorl + setcc is generally the preferred sequence due to the partial register
stall setcc + movzbl suffers from. As a bonus, it also encodes one byte smaller.
This fixes PR28146.
The original commit tried inserting an 8bit-subreg into a GR32 (not GR32_ABCD)
which was not appreciated by fast regalloc on 32-bit.
llvm-svn: 274802
The commit reinstates r273279, which was informally approved.
Original Review: http://reviews.llvm.org/D21414
This reverts commit ca632c91aaa7cafc50942f890c49f727a046ace1.
llvm-svn: 274790
These tests don't actually care about the internal opcode number, but have to
be updated whenever we add a new one for GlobalISel. That's bad.
llvm-svn: 274774
- Rename the ptx.read.* intrinsics to nvvm.read.ptx.sreg.* - some but
not all of these registers were already accessible via the nvvm
name.
- Rename ptx.bar.sync nvvm.bar.sync, to match nvvm.bar0.
There's a fair amount of code motion here, but it's all very
mechanical.
llvm-svn: 274769
xorl + setcc is generally the preferred sequence due to the partial register
stall setcc + movzbl suffers from. As a bonus, it also encodes one byte smaller.
This fixes PR28146.
Differential Revision: http://reviews.llvm.org/D21774
llvm-svn: 274692
On CPUs with the zero cycle zeroing feature enabled "movi v.2d" should
be used to zero a vector register. This was previously done at
instruction selection time, however the register coalescer sometimes
widened multiple vregs to the Q width because of that leading to extra
spills. This patch leaves the decision on how to zero a register to the
AsmPrinter phase where it doesn't affect register allocation anymore.
This patch also sets isAsCheapAsAMove=1 on FMOVS0, FMOVD0.
This fixes http://llvm.org/PR27454, rdar://25866262
Differential Revision: http://reviews.llvm.org/D21826
llvm-svn: 274686
On SystemZ, shift and rotate instructions only use the bottom 6 bits of the shift/rotate amount.
Therefore, if the amount is ANDed with an immediate mask that has all of the bottom 6 bits set, we
can remove the AND operation entirely.
Differential Revision: http://reviews.llvm.org/D21854
llvm-svn: 274650
There is a problem in VSXSwapRemoval where it is incorrectly removing permute instructions.
In this case, the permute is feeding both a vector store and also a non-store instruction. In this case, the permute cannot be removed.
The fix is to simply look at all the uses of the vector register defined by the permute and ensure that all the uses are vector store instructions.
This problem was reported in PR 27735 (https://llvm.org/bugs/show_bug.cgi?id=27735).
Test case based on the original problem reported.
Phabricator Review: http://reviews.llvm.org/D21802
llvm-svn: 274645
Summary:
findBetterNeighborChains may or may not find a better chain for each node it finds, which include the node ("St") that visitSTORE is currently processing. If no better chain is found for St, visitSTORE should continue instead of return SDValue(St, 0), as if it's CombinedTo'ed.
This fixes bug 28130. There might be other ways to make the test pass (see D21409). I think both of the patches are fixing actual bugs revealed by the same testcase.
Reviewers: echristo, wschmidt, hfinkel, kbarton, amehsan, arsenm, nemanjai, bogner
Subscribers: mehdi_amini, nemanjai, llvm-commits
Differential Revision: http://reviews.llvm.org/D21692
llvm-svn: 274644
This is a follow-up for r273544.
The end goal is to get rid of the isSwift / isCortexXY / isWhatever methods.
This commit also removes two command-line flags that weren't used in any of the
tests: widen-vmovs and swift-partial-update-clearance. The former may be easily
replaced with the mattr mechanism, but the latter may not (as it is a subtarget
property, and not a proper feature).
Differential Revision: http://reviews.llvm.org/D21797
llvm-svn: 274620
The patch removes redundant kmov instructions (not all, we still have a lot of work here) and redundant "and" instructions after "setcc".
I use "AssertZero" marker between X86ISD::SETCC node and "truncate" to eliminate extra "and $1" instruction.
I also changed zext, aext and trunc patterns in the .td file. It allows to remove extra "kmov" instruictions.
This patch fixes https://llvm.org/bugs/show_bug.cgi?id=28173.
Fast ISEL mode is not supported correctly for AVX-512. ICMP/FCMP scalar instruction should return result in k-reg. It will be fixed in one of the next patches. I redirected handling of "cmp" to the DAG builder mode. (The code looks worse in one specific test case, but without this fix the new patch fails).
Differential revision: http://reviews.llvm.org/D21956
llvm-svn: 274613
Summary:
Since "AMDGPU: Fix verifier errors in SILowerControlFlow", the logic that
ensures that a non-void-returning shader falls off the end of the last
basic block was effectively disabled, since SI_RETURN is now used.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=96731
Reviewers: arsenm, tstellarAMD
Subscribers: arsenm, kzhuravl, llvm-commits
Differential Revision: http://reviews.llvm.org/D21975
llvm-svn: 274612
The way the named arguments for various system instructions are handled at the
moment has a few problems:
- Large-scale duplication between AArch64BaseInfo.h and AArch64BaseInfo.cpp
- That weird Mapping class that I have no idea what I was on when I thought
it was a good idea.
- Searches are performed linearly through the entire list.
- We print absolutely all registers in upper-case, even though some are
canonically mixed case (SPSel for example).
- The ARM ARM specifies sysregs in terms of 5 fields, but those are relegated
to comments in our implementation, with a slightly opaque hex value
indicating the canonical encoding LLVM will use.
This adds a new TableGen backend to produce efficiently searchable tables, and
switches AArch64 over to using that infrastructure.
llvm-svn: 274576
This reverts commit r259387 because it inserts illegal code after legalization
in some backends where i64 OR type is illegal for example.
llvm-svn: 274573
Not all code-paths set the relocation model to static for Windows. This
currently breaks on Windows ARM with `-mlong-calls` when built with clang.
Loosen the assertion to what it was previously. We would ideally ensure that
all the configuration sets Windows to static relocation model.
llvm-svn: 274570
The other use really does only care about the SDNode (it checks the
opcode against a whitelist), but bitFieldPlacement can be misled if
the node produces multiple results.
Patch by Ismail Badawi.
llvm-svn: 274567
Because of the special immediate operand, the constant
bus is already used so SGPRs are never useful.
r263212 changed the name of the immediate operand, which
broke the verifier check for the restriction.
llvm-svn: 274564
The important thing I was missing was ensuring newly added constants were kept in topological order. Repositioning the node is correct if the constant is newly added (so it has no topological ordering) but wrong if it already existed - positioning it next in the worklist would break the topological ordering.
Original commit message:
[Thumb] Select a BIC instead of AND if the immediate can be encoded more optimally negated
If an immediate is only used in an AND node, it is possible that the immediate can be more optimally materialized when negated. If this is the case, we can negate the immediate and use a BIC instead;
int i(int a) {
return a & 0xfffffeec;
}
Used to produce:
ldr r1, [CONSTPOOL]
ands r0, r1
CONSTPOOL: 0xfffffeec
And now produces:
movs r1, #255
adds r1, #20 ; Less costly immediate generation
bics r0, r1
llvm-svn: 274543
This patch corresponds to review:
http://reviews.llvm.org/D20443
It changes the legalization strategy for illegal vector types from integer
promotion to widening. This only applies for vectors with elements of width
that is a multiple of a byte since we have hardware support for vectors with
1, 2, 3, 8 and 16 byte elements.
Integer promotion for vectors is quite expensive on PPC due to the sequence
of breaking apart the vector, extending the elements and reconstituting the
vector. Two of these operations are expensive.
This patch causes between minor and major improvements in performance on most
benchmarks. There are very few benchmarks whose performance regresses. These
regressions can be handled in a subsequent patch with a DAG combine (similar
to how this patch handles int -> fp conversions of illegal vector types).
llvm-svn: 274535
We were using DAG->getConstant instead of DAG->getTargetConstant. This meant that we could inadvertently increase the use count of a constant if stars aligned, which it did in this testcase. Increasing the use count of the constant could cause ISel to fall over (because DAGToDAG lowering assumed the constant had only one use!)
Original commit message:
[Thumb] Select a BIC instead of AND if the immediate can be encoded more optimally negated
If an immediate is only used in an AND node, it is possible that the immediate can be more optimally materialized when negated. If this is the case, we can negate the immediate and use a BIC instead;
int i(int a) {
return a & 0xfffffeec;
}
Used to produce:
ldr r1, [CONSTPOOL]
ands r0, r1
CONSTPOOL: 0xfffffeec
And now produces:
movs r1, #255
adds r1, #20 ; Less costly immediate generation
bics r0, r1
llvm-svn: 274510
We can now handle concatenation of each source multiple times. The previous code just checked for each source to appear once in either order.
This also now handles an entire source vector sized piece having undef indices correctly. We now concat with UNDEF instead of using one of the sources. This is responsible for the test case change.
llvm-svn: 274483
After the block placement, if a block ends with a conditional branch, but the
next block is not its successor. The conditional branch should be changed to
unconditional branch. This patch fixes PR28307, PR28297, PR28402.
Differential Revision: http://reviews.llvm.org/D21811
llvm-svn: 274470
This patch adds support for including the avx512 mask register information in the mask/maskz versions of shuffle instruction comments.
This initial version just adds support for MOVDDUP/MOVSHDUP/MOVSLDUP to reduce the mass of test regenerations, other shuffle instructions can be added in due course.
Differential Revision: http://reviews.llvm.org/D21953
llvm-svn: 274459
Its not worth trying to write out tests for all the avx512f builtins yet, just adding tests for lowering of generic IR as we transition to it (shuffles mainly right now).
llvm-svn: 274434
Due to visit order problems, in the case of an unaligned copy
the legalized DAG fails to eliminate extra instructions introduced
by the expansion of both unaligned parts.
llvm-svn: 274397
There was a combine before to handle the simple copy case.
Split this into handling loads and stores separately.
We might want to change how this handles some of the vector
extloads, since this can result in large code size increases.
llvm-svn: 274394
Summary: original test may have different bahavior on different bot, specifically it broke llvm-clang-lld-x86_64-scei-ps4-ubuntu-fast
Reviewers: majnemer
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D21931
llvm-svn: 274368
Summary: The code generation should be independent of the debug info.
Reviewers: zansari, davidxl, mkuper, majnemer
Subscribers: majnemer, llvm-commits
Differential Revision: http://reviews.llvm.org/D21911
llvm-svn: 274357
This was reverted in r268740 because of problems with corresponding Clang change.
Clang change was updated and resubmitted in r274220.
Check calling convention in AMDGPUMachineFunction::isKernel
This will be used for AMDGPU_HSA_KERNEL symbol type in output ELF.
Also, in the future unused non-kernels may be optimized.
Reviewers: tstellarAMD, arsenm
Subscribers: arsenm, joker.eph, llvm-commits
Differential Revision: http://reviews.llvm.org/D19917
llvm-svn: 274341
Summary:
MSVC provide exception handlers with enhanced information to deal with security buffer feature (/GS).
To be more secure, the security cookies (GS and SEH) are validated when unwinding the stack.
The following code:
```
void f() {}
void foo() {
__try {
f();
} __except(1) {
f();
}
}
```
Reviewers: majnemer, rnk
Subscribers: thakis, llvm-commits, chrisha
Differential Revision: http://reviews.llvm.org/D21101
llvm-svn: 274239
This processor feature had been left out by mistake from the z13
ProcessorModel.
This time with updated test case. Thanks, Hans.
Reviewed by Ulrich Weigand.
llvm-svn: 274216
When lowering two blended PACKUS, we used to disregard the types
of the PACKUS inputs, indiscriminately generating a v16i8 PACKUS.
This leads to non-selectable things like:
(v16i8 (PACKUS (v4i32 v0), (v4i32 v1)))
Instead, check that the PACKUSes have the same type, and use that
as the final result type.
llvm-svn: 274138
Summary: LLVM assumes that large clearance will hide the partial register spill penalty. But in our experiment, 16 clearance is too small. As the inserted XOR is normally fairly cheap, we should have a higher clearance threshold to aggressively insert XORs that is necessary to break partial register dependency.
Reviewers: wmi, davidxl, stoklund, zansari, myatsina, RKSimon, DavidKreitzer, mkuper, joerg, spatel
Subscribers: davidxl, llvm-commits
Differential Revision: http://reviews.llvm.org/D21560
llvm-svn: 274068
Summary: SystemZ shift instructions only use the last 6 bits of the shift
amount. When the result of an AND operation is used as a shift amount, this
means that we can use the NILL instruction (which operates on the last 16 bits)
rather than NILF (which operates on the last 32 bits) for a 16-bit savings in
instruction size.
Reviewers: uweigand
Subscribers: llvm-commits
Author: colpell
Committing on behalf of Elliot.
Differential Revision: http://reviews.llvm.org/D21686
llvm-svn: 274066
This is a resubmittion of 263158 change after fixing the existing problem with intrinsics mangling (see LTO and intrinsics mangling llvm-dev thread for details).
This patch fixes the problem which occurs when loop-vectorize tries to use @llvm.masked.load/store intrinsic for a non-default addrspace pointer. It fails with "Calling a function with a bad signature!" assertion in CallInst constructor because it tries to pass a non-default addrspace pointer to the pointer argument which has default addrspace.
The fix is to add pointer type as another overloaded type to @llvm.masked.load/store intrinsics.
Reviewed By: reames
Differential Revision: http://reviews.llvm.org/D17270
llvm-svn: 274043
The original implementation attempted to zero registers using
XOR %foo, %foo. This is problematic because it constitutes a
read-modify-write of a register which might not be defined.
Instead, use MOV32r0 to avoid these problems; expandPostRAPseudo does
the right thing here.
llvm-svn: 274024
AVX1 can only broadcast vectors as floats/doubles, so for 256-bit vectors we insert bitcasts if we are shuffling v8i32/v4i64 types. Unfortunately the presence of these bitcasts prevents the current broadcast lowering code from peeking through cases where we have concatenated / extracted vectors to create the 256-bit vectors.
This patch allows us to peek through bitcasts as long as the number of elements doesn't change (i.e. element bitwidth is the same) so the broadcast index is not affected.
Note this bitcast peek is different from the stage later on which doesn't care about the type and is just trying to find a load node.
As we're being more aggressive with bitcasts, we also need to ensure that the broadcast type is correctly bitcasted
Differential Revision: http://reviews.llvm.org/D21660
llvm-svn: 274013
This patch allows target shuffles to be combined to single input immediate permute instructions - (V)PSHUFD/VPERMILPD/VPERMILPS - allowing more general pattern matching than what we current do and improves the likelihood of memory folding compared to existing patterns which tend to reuse the input in multiple arguments.
Further permute instructions (V)PSHUFLW/(V)PSHUFHW/(V)PERMQ/(V)PERMPD may be added in the future but its proven tricky to create tests cases for them so far. (V)PSHUFLW/(V)PSHUFHW is already handled quite well in combineTargetShuffle so it may be that removing some of that code may allow us to perform more of the combining in one place without duplication.
Differential Revision: http://reviews.llvm.org/D21148
llvm-svn: 273999