Summary:
D30400 has enabled tADC and tSBC instructions to be unglued, thereby allowing CPSR to remain live between Thumb1 scheduling units.
Most Thumb1 instructions have an OptionalDef for CPSR; but the scheduler ignored the OptionalDefs, and could unwittingly insert a flag-setting instruction in between an ADDS and the corresponding ADC.
Reviewers: javed.absar, atrick, MatzeB, t.p.northover, jmolloy, rengolin
Reviewed By: javed.absar
Subscribers: rogfer01, efriedma, aemerson, rengolin, llvm-commits, MatzeB
Differential Revision: https://reviews.llvm.org/D31081
llvm-svn: 301106
canMutate() was returning true when the operands were all in the same order as
the matched instruction. However, it wasn't checking the operands were actually
on that instruction. This worked when we could only match a single instruction
but the addition of nested instruction matching led to cases where the operands
could be split across multiple instructions. canMutate() now returns false if
operands belong to instructions other than the root of the match.
llvm-svn: 301077
Since Split DWARF needs to name the actual .dwo file that is generated,
it can't be known at the time the llvm::Module is produced as it may be
merged with other Modules before the object is generated and that object
may be generated with any name.
By passing the Split DWARF file name when LLVM is producing object code
the .dwo file name in the object file can match correctly.
The support for Split DWARF for implicit modules remains the same -
using metadata to store the dwo name and dwo id so that potentially
multiple skeleton CUs referring to different dwo files can be generated
from one llvm::Module.
llvm-svn: 301062
The code assumed that when saving an additional CSR register
(ExtraCSSpill==true) we would have a free register throughout the
function. This was not true if this CSR register is also used to pass
values as in the swiftself case.
rdar://31451816
llvm-svn: 301057
In addition to the original commit, tighten the condition for when to
pad empty functions to COFF Windows. This avoids running into problems
when targeting e.g. Win32 AMDGPU, which caused test failures when this
was committed initially.
llvm-svn: 301047
Empty functions can lead to duplicate entries in the Guard CF Function
Table of a binary due to multiple functions sharing the same RVA,
causing the kernel to refuse to load that binary.
We had a terrific bug due to this in Chromium.
It turns out we were already doing this for Mach-O in certain
situations. This patch expands the code for that in
AsmPrinter::EmitFunctionBody() and renames
TargetInstrInfo::getNoopForMachoTarget() to simply getNoop() since it
seems it was used for not just Mach-O anyway.
Differential Revision: https://reviews.llvm.org/D32330
llvm-svn: 301040
Otherwise there's some mismatch, and we'll either form an illegal type or an
illegal node.
Thanks to Eli Friedman for pointing out the problem with my original solution.
llvm-svn: 301036
SI_MASKED_UNREACHABLE does not have machine instruction encoding.
It needs special handling in AMDGPUAsmPrinter::EmitInstruction like some
other pseudo instructions.
This patch fixes compilation failure of RadeonRays.
Differential Revision: https://reviews.llvm.org/D32364
llvm-svn: 301025
immediate operands.
This commit adds an AArch64 dag-combine that optimizes code generation
for logical instructions taking immediate operands. The optimization
uses demanded bits to change a logical instruction's immediate operand
so that the immediate can be folded into the immediate field of the
instruction.
This recommits r300932 and r300930, which was causing dag-combine to
loop forever. The problem was that optimizeLogicalImm was returning
true even when there was no change to the immediate node (which happened
when the immediate was all zeros or ones), which caused dag-combine to
push and pop the same node to the work list over and over again without
making any progress.
This commit fixes the bug by returning false early in optimizeLogicalImm
if the immediate is all zeros or ones. Also, it changes the code to
compare the immediate with 0 or Mask rather than calling
countPopulation.
rdar://problem/18231627
Differential Revision: https://reviews.llvm.org/D5591
llvm-svn: 301019
DAG combine was mistakenly assuming that the step-up it was looking at was
always a doubling, but it can sometimes be a larger extension in which case
we'd crash.
llvm-svn: 301002
places based on it.
Existing constant hoisting pass will merge a group of contants in a small range
and hoist the const materialization code to the common dominator of their uses.
However, if the uses are all in cold pathes, existing implementation may hoist
the materialization code from cold pathes to a hot place. This may hurt performance.
The patch introduces BFI to the pass and selects the best insertion places based
on it.
The change is controlled by an option consthoist-with-block-frequency which is
off by default for now.
Differential Revision: https://reviews.llvm.org/D28962
llvm-svn: 300989
Select them as copies. We only select if both the source and the
destination are on the same register bank, so this shouldn't cause any
trouble.
llvm-svn: 300971
The condition in isSupportedType didn't handle struct/array arguments
properly. Fix the check and add a test to make sure we use the fallback
path in this kind of situation. The test deals with some common cases
where the call lowering should error out. There are still some issues
here that need to be addressed (tail calls come to mind), but they can
be addressed in other patches.
llvm-svn: 300967
when the subtarget has fast strings.
This has two advantages:
- Speed is improved. For example, on Haswell thoughput improvements increase
linearly with size from 256 to 512 bytes, after which they plateau:
(e.g. 1% for 260 bytes, 25% for 400 bytes, 40% for 508 bytes).
- Code is much smaller (no need to handle boundaries).
llvm-svn: 300957
It seems that r300930 was creating an infinite loop in dag-combine when
compling the following file:
MultiSource/Benchmarks/MiBench/consumer-typeset/z21.c
llvm-svn: 300940
immediate operands.
This commit adds an AArch64 dag-combine that optimizes code generation
for logical instructions taking immediate operands. The optimization
uses demanded bits to change a logical instruction's immediate operand
so that the immediate can be folded into the immediate field of the
instruction.
This recommits r300913, which broke bots because I didn't fix a call to
ShrinkDemandedConstant in SIISelLowering.cpp after changing the APIs of
TargetLoweringOpt and TargetLowering.
rdar://problem/18231627
Differential Revision: https://reviews.llvm.org/D5591
llvm-svn: 300930
immediate operands.
This commit adds an AArch64 dag-combine that optimizes code generation
for logical instructions taking immediate operands. The optimization
uses demanded bits to change a logical instruction's immediate operand
so that the immediate can be folded into the immediate field of the
instruction.
rdar://problem/18231627
Differential Revision: https://reviews.llvm.org/D5591
llvm-svn: 300913
Single-threaded fences aren't required to provide any synchronization with
other processing elements so there's no need for a DMB. They should still be a
barrier for compiler optimizations though.
llvm-svn: 300904
Before, we assumed that any ConstantInt offset was precisely the access width,
so we could use the "[rN]!" form. ISelLowering only ever created that kind, but
further simplification during combining could lead to unexpected constants and
incorrect codegen.
Should fix PR32658.
llvm-svn: 300878
Recently alloca address space has been added to data layout. Due to this
change, pointer returned by alloca may have different size as pointer in
address space 0.
However, currently the value type of frame index is assumed to be of the
same size as pointer in address space 0.
This patch fixes that.
Most targets assume alloca returning pointer in address space 0, which
is the default alloca address space. Therefore it is NFC for them.
AMDGCN target with amdgiz environment requires this change since it
assumes alloca returning pointer to addr space 5 and its size is 32,
which is different from the size of pointer in addr space 0 which is 64.
Differential Revision: https://reviews.llvm.org/D32021
llvm-svn: 300864
Masked vectors which hold shift amounts when creating the following nodes:
ISD::SHL, ISD::SRL or ISD::SRA.
Instructions that use said nodes, which have had their arguments altered are
sll, srl, sra, bneg, bclr and bset.
For said instructions, the shift amount or the bit position that is
specified in the corresponding vector elements will be interpreted as the
shift amount/bit position modulo the size of the element in bits.
The problem lies in compiling with -O2 enabled, where the instructions for
formats .w and .d are not generated, but are instead optimized away.
In this case, having shift amounts that are either negative or greater than
the element bit size results in generation of incorrect results when
constant folding.
We remedy this by masking the operands for the nodes mentioned above before
actually creating them, so that the final result is correct before placed
into the constant pool.
Patch by Stefan Maksimovic.
Differential Revision: https://reviews.llvm.org/D31331
llvm-svn: 300839
Debug information is calculated with getFrameIndexReference() which was
missing some logic for the fixed object cases (= parameters on the stack).
rdar://24557797
Differential Revision: https://reviews.llvm.org/D32204
llvm-svn: 300781
I've changed one of the tests to not fold away, but we didn't and still don't do the transform
that the comment claims we do (and I don't know why we'd want to do that).
Follow-up to:
https://reviews.llvm.org/rL300725https://reviews.llvm.org/rL300763
llvm-svn: 300772
This allows forming more 'not' ops, so we get improvements for ISAs that have and-not.
Follow-up to:
https://reviews.llvm.org/rL300725
llvm-svn: 300763
Re-commit after revert in r300668. Changed getMaxFPOffset() to a
more conservative heuristic instead of trying to be clever and missing
for some exotic calling conventions.
We need to reserve an emergency spill slot in cases with large argument
types that could overflow immediate offsets for FP relative address
calculations.
rdar://31317893
Differential Revision: https://reviews.llvm.org/D31643
llvm-svn: 300761
This is inserted directly in the text section. The relocation
for the function ends up resolving to the beginning of the
amd_kernel_code_t header rather than the actual function
entry point.
Also skip some of the comments for initialization
that only makes sense for kernels.
llvm-svn: 300736
The most common case for a branch condition is
a single use compare. Directly invert the branch
predicate rather than adding a lot of xor i1 true
which the DAG will have to fold later.
This produces nicer to read structurizer output.
This produces some random changes in codegen
due to the DAG swapping branch conditions itself,
and then does a poor job of dealing with those
inverts.
llvm-svn: 300732
The patch itself is simple: stop discriminating against vectors in visitAnd() and again in
SimplifyDemandedBits().
Some notes for reference:
1. We're not consistent about calls to SimplifyDemandedBits in the various visitXXX functions.
Sometimes, we check if the RHS is a constant first. Other times (like here), we just dive in.
2. I'd like to break the vector shackles in steps for the sake of risk minimization, but we could
make similar simultaneous changes in other places if we think that would be better.
3. I don't know what the intent of the changed tests in this patch was supposed to be, but since
they wiggled in a positive way, I'm just going with that. :)
4. In the rotate tests, note that we can see through non-splat constants. This is a result of D24253.
5. My motivation for being here now is to make D31944 look better, so this is step 1 of N towards
improving the vector codegen in that patch without writing any actual new code.
Differential Revision: https://reviews.llvm.org/D32230
llvm-svn: 300725
Also, make a few changes to allow using the pass in .mir testcases.
Among other things, change the abbreviation from opt-amode to amode-opt,
because otherwise lit would expand the "opt" part to the full path to
the opt binary.
llvm-svn: 300707
A bunch of tests failed because memory operations have been reordered.
I am unsure which commit changed this behaviour as the AVR build was
failing at that point with an unrelated error.
This commit just reoders some of the CHECK lines in some tests to suit
current llc output.
llvm-svn: 300682
Support G_MUL, very similar to G_ADD and G_SUB. The only difference is
in the instruction selector, where we have to select either MUL or MULv5
depending on the target.
llvm-svn: 300665
This fixes PR32471.
As comment 10 on that bug report highlights
(https://bugs.llvm.org//show_bug.cgi?id=32471#c10), there are quite a
few different defendable design tradeoffs that could be made, including
not representing pointers at all in LLT.
I decided to go for representing vector-of-pointer as a concept in LLT,
while keeping the size of the LLT type 64 bits (this is an increase from
48 bits before). My rationale for keeping pointers explicit is that on
some targets probably it's very handy to have the distinction between
pointer and non-pointer (e.g. 68K has a different register bank for
pointers IIRC). If we keep a scalar pointer, it probably is easiest to
also have a vector-of-pointers to keep LLT relatively conceptually clean
and orthogonal, while we don't have a very strong reason to break that
orthogonality. Once we gain more experience on the use of LLT, we can
of course reconsider this direction.
Rejecting vector-of-pointer types in the IRTranslator is also an option
to avoid the crash reported in PR32471, but that is only a very
short-term solution; also needs quite a bit of code tweaks in places,
and is probably fragile. Therefore I didn't consider this the best
option.
llvm-svn: 300664
We need to reserve an emergency spill slot in cases with large argument
types that could overflow immediate offsets for FP relative address
calculations.
rdar://31317893
Differential Revision: https://reviews.llvm.org/D31643
llvm-svn: 300639
Android x86_64 target uses f128 type and stores f128 values in %xmm* registers.
SoftenFloatRes_EXTRACT_VECTOR_ELT should not convert result value
from f128 to i128.
Differential Revision: http://reviews.llvm.org/D32102
llvm-svn: 300583
Remove non-consecutive stores from store merge candidate search as
they cannot be merged and will prevent us from finding subsequent
mergeable store cases.
Reviewers: jyknight, bogner, javed.absar, spatel
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D32086
llvm-svn: 300561
In the assembler, we should emit build attributes based on the target
selected with command-line options. This matches the GNU assembler's
behaviour. We only do this for build attributes which describe the
hardware that is expected to be available, not the ones that describe
ABI compatibility.
This is done by moving some of the attribute emission code to
ARMTargetStreamer, so that it can be shared between the assembly and
code-generation code paths. Since the assembler only creates a
MCSubtargetInfo, not an ARMSubtarget, the code had to be changed to
check raw features, and not use the convenience functions in
ARMSubtarget.
If different attributes are later specified using the .eabi_attribute
directive, then they will take precedence, as happens when the same
.eabi_attribute is specified twice.
This must be enabled by an option, because we don't want to do this when
parsing inline assembly. The attributes would match the ones emitted at
the start of the file, so wouldn't actually change the emitted object
file, but the extra directives would be added to every inline assembly
block when emitting assembly, which we'd like to avoid.
The majority of the changes in the build-attributes.ll test are just
re-ordering the directives, because the hardware attributes are now
emitted before the ABI ones. However, I did fix one bug which I spotted:
Tag_CPU_arch_profile was not being emitted for v6M.
Differential revision: https://reviews.llvm.org/D31812
llvm-svn: 300547
This reverts r300535 and r300537.
The newly added tests in test/CodeGen/AArch64/GlobalISel/arm64-fallback.ll
produces slightly different code between LLVM versions being built with different compilers.
E.g., dependent on the compiler LLVM is built with, either one of the following
can be produced:
remark: <unknown>:0:0: unable to legalize instruction: %vreg0<def>(p0) = G_EXTRACT_VECTOR_ELT %vreg1, %vreg2; (in function: vector_of_pointers_extractelement)
remark: <unknown>:0:0: unable to legalize instruction: %vreg2<def>(p0) = G_EXTRACT_VECTOR_ELT %vreg1, %vreg0; (in function: vector_of_pointers_extractelement)
Non-determinism like this is clearly a bad thing, so reverting this until
I can find and fix the root cause of the non-determinism.
llvm-svn: 300538
For subtargets that use the custom lowering for divmod, e.g. gnueabi,
we used to check if the subtarget has hardware divide and then lower to
a div-mul-sub sequence if true, or to a libcall if false.
However, judging by the usage of hasDivide vs hasDivideInARMMode, it
seems that hasDivide only refers to Thumb. For instance, in the
ARMTargetLowering constructor, the code that specifies whether to use
libcalls for (S|U)DIV looks like this:
bool hasDivide = Subtarget->isThumb() ? Subtarget->hasDivide()
: Subtarget->hasDivideInARMMode();
In the case of divmod for arm-gnueabi, using only hasDivide() to
determine what to do means that instead of lowering to __aeabi_idivmod
to get the remainder, we lower to div-mul-sub and then further lower the
div to __aeabi_idiv. Even worse, if we have hardware divide in ARM but
not in Thumb, we generate a libcall instead of using it (this is not an
issue in practice since AFAICT none of the cores that we support have
hardware divide in ARM but not Thumb).
This patch fixes the code dealing with custom lowering to take into
account the mode (Thumb or ARM) when deciding whether or not hardware
division is available.
Differential Revision: https://reviews.llvm.org/D32005
llvm-svn: 300536
This fixes PR32471.
As comment 10 on that bug report highlights
(https://bugs.llvm.org//show_bug.cgi?id=32471#c10), there are quite a
few different defendable design tradeoffs that could be made, including
not representing pointers at all in LLT.
I decided to go for representing vector-of-pointer as a concept in LLT,
while keeping the size of the LLT type 64 bits (this is an increase from
48 bits before). My rationale for keeping pointers explicit is that on
some targets probably it's very handy to have the distinction between
pointer and non-pointer (e.g. 68K has a different register bank for
pointers IIRC). If we keep a scalar pointer, it probably is easiest to
also have a vector-of-pointers to keep LLT relatively conceptually clean
and orthogonal, while we don't have a very strong reason to break that
orthogonality. Once we gain more experience on the use of LLT, we can
of course reconsider this direction.
Rejecting vector-of-pointer types in the IRTranslator is also an option
to avoid the crash reported in PR32471, but that is only a very
short-term solution; also needs quite a bit of code tweaks in places,
and is probably fragile. Therefore I didn't consider this the best
option.
llvm-svn: 300535
Summary:
Refactoring changed paramHasAttr(1 + i) to paramHasAttr(0), fix that to
paramHasAttr(i).
Add more tests to WebAssemblyOptimizeReturned that catch that
regression.
Reviewers: dschuff
Subscribers: jfb, sbc100, llvm-commits
Differential Revision: https://reviews.llvm.org/D32136
llvm-svn: 300502
Avoid looping through program to determine register counts.
This avoids needing to look at regmask operands.
Also fixes some counting errors with flat_scr when there
are no stack objects.
llvm-svn: 300482
It's basically a terrible idea anyway but objc_msgSend gets emitted like that.
We can decide on a better way to deal with it in the unlikely event that anyone
actually uses it.
llvm-svn: 300474
It's almost certainly not a good idea to actually use it in most cases (there's
a pretty large code size overhead on AArch64), but we can't do those
experiments until it's supported.
llvm-svn: 300462
Our 16 bit support is assembler-only + the terrible hack that is
.code16gcc. Simply using 32 bit registers does the right thing for the
latter.
Fixes PR32681.
llvm-svn: 300429
Summary:
In PR32594, inline assembly using the 'A' constraint on x86_64 causes
llvm to crash with a "Cannot select" stack trace. This is because
`X86TargetLowering::getRegForInlineAsmConstraint` hardcodes that 'A'
means the EAX and EDX registers.
However, on x86_64 it means the RAX and RDX registers, and on 16-bit x86
(ia16?) it means the old AX and DX registers.
Add new register classes in `X86RegisterInfo.td` to support these cases,
and amend the logic in `getRegForInlineAsmConstraint` to cope with
different subtargets. Also add a test case, derived from PR32594.
Reviewers: craig.topper, qcolombet, RKSimon, ab
Reviewed By: ab
Subscribers: ab, emaste, royger, llvm-commits
Differential Revision: https://reviews.llvm.org/D31902
llvm-svn: 300404
If a kernel's pointer argument is known to be readonly
set access qualifier accordingly. This allows RT not to
flush caches before dispatches.
Differential Revision: https://reviews.llvm.org/D32091
llvm-svn: 300362
MOVNTDQA non-temporal aligned vector loads can be correctly represented using generic builtin loads, allowing us to remove the existing x86 intrinsics.
Clang companion patch: D31766.
Differential Revision: https://reviews.llvm.org/D31767
llvm-svn: 300325
This further improves Ahmed's change in rL299482. See the new comment for the
rationale.
The patch recovers most of the regression for bzip2 after D31965. We're down
to +2.68% from +6.97%.
Differential Revision: https://reviews.llvm.org/D32028
llvm-svn: 300276
If the offset cannot fit into the instruction, an addition to the
pointer is emitted before the actual access. However, BPF offsets are
16-bit but LLVM considers them to be, for the matter of this check,
to be 32-bit long.
This causes the following program:
int bpf_prog1(void *ign)
{
volatile unsigned long t = 0x8983984739ull;
return *(unsigned long *)((0xffffffff8fff0002ull) + t);
}
To generate the following (wrong) code:
0: 18 01 00 00 39 47 98 83 00 00 00 00 89 00 00 00
r1 = 590618314553ll
2: 7b 1a f8 ff 00 00 00 00 *(u64 *)(r10 - 8) = r1
3: 79 a1 f8 ff 00 00 00 00 r1 = *(u64 *)(r10 - 8)
4: 79 10 02 00 00 00 00 00 r0 = *(u64 *)(r1 + 2)
5: 95 00 00 00 00 00 00 00 exit
Fix it by changing the offset check to 16-bit.
Patch by Nadav Amit <nadav.amit@gmail.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Differential Revision: https://reviews.llvm.org/D32055
llvm-svn: 300269
In many cases ds operations can be combined even if offsets do not
fit into 8 bit encoding. What it takes is to adjust base address.
Differential Revision: https://reviews.llvm.org/D31993
llvm-svn: 300227
If workgroup size is known inform llvm about range returned by local
id and local size queries.
Differential Revision: https://reviews.llvm.org/D31804
llvm-svn: 300102
Check if the scale operand is identical (doesn't have to be 1) and
do not check the chaain operand.
Differential revision: https://reviews.llvm.org/D31833
llvm-svn: 299986
If you run llc -stop-after=codegenprepare and feed the resulting MIR
to llc -start-after=codegenprepare, you'll have an empty machine
function since we haven't run any isel yet. Of course, this only works
if the MIRParser believes you that this is okay.
This is essentially a revert of r241862 with a fix for the problem it
was papering over.
llvm-svn: 299975
Use the same handling in the generic legalizer code as for the other
libcalls (G_FREM, G_FPOW).
Enable it on ARM for float and double so we can test it.
llvm-svn: 299931
Summary: Legalize only if the type is marked as Legal or Custom. If not, return Unsupported as LegalizerHelper is not able to handle non-power-of-2 types right now.
Reviewers: qcolombet, aditya_nandakumar, dsanders, t.p.northover, kristof.beyls, javed.absar, ab
Reviewed By: kristof.beyls, ab
Subscribers: dberris, rovka, igorb, llvm-commits
Differential Revision: https://reviews.llvm.org/D31711
llvm-svn: 299929
A fix for the bug reported in PR30911.
The issue arises when multiple CALLSEQ_BEGIN nodes are unscheduled as
the last node to be unscheduled will gain access to the CallResource
register. But when a node is being picked, only CALLSEQ_END nodes are
checked against the CallResource and have their chains evaluated.
This then means that other CALLSEQ_BEGIN nodes can be scheduled
before the existing call sequence has been finalised. This patch adds
a check against the FrameSetup nodes in DelayForLiveRegs to prevent
this from happening.
Differential Revision: https://reviews.llvm.org/D31536
llvm-svn: 299926
Check the legality of ISD::[US]MULO to see whether
Intrinsic::[us]mul_with_overflow will legalize into a function call (and, thus,
will use the CTR register). Fixes PR32485.
Patch by Tim Neumann!
Differential Revision: https://reviews.llvm.org/D31790
llvm-svn: 299910
The math works out where it can actually be counter-productive. The probability
calculations correctly handle the case where the alternative is 0 probability,
rely on those calculations.
Includes a test case that demonstrates the problem.
llvm-svn: 299892
Qin may be large, and Succ may be more frequent than BB. Take these both into
account when deciding if tail-duplication is profitable.
llvm-svn: 299891
Merging identical blocks when it doesn't reduce fallthrough. It is common for
the blocks created from critical edge splitting to be identical. We would like
to merge these blocks whenever doing so would not reduce fallthrough.
llvm-svn: 299890
BIC is generally faster, and it can put the output in a different
register from the input.
We already do this in Thumb2 mode; not sure why the equivalent fix
never got applied to ARM mode.
Differential Revision: https://reviews.llvm.org/D31797
llvm-svn: 299803
The original instruction might get legalized and erased and expanded
into intermediate instructions and the intermediate instructions might
fail legalization. This end up in reporting GISelFailure on the erased
instruction.
Instead report GISelFailure on the intermediate instruction which failed
legalization.
Reviewed by: ab
llvm-svn: 299802
When using -ffixed-x18, the x18 (or w18) register can safely be used
with the "global register variable" GCC extension, but the backend
fails to recognize it.
Patch by Roland McGrath.
Differential Revision: https://reviews.llvm.org/D31793
llvm-svn: 299799
This reverts commit r299766. This change appears to have broken the MIPS
buildbots. Reverting while I investigate.
Revert "[mips] Remove usage of debug only variable (NFC)"
This reverts commit r299769. Follow up commit.
llvm-svn: 299788
Increase threshold to unroll a loop which contains an "if" statement
whose condition defined by a PHI belonging to the loop. This may help
to eliminate if region and potentially even PHI itself, saving on
both divergence and registers used for the PHI.
Add a small bonus for each of such "if" statements.
Differential Revision: https://reviews.llvm.org/D31693
llvm-svn: 299779
We have two cases here, the first one being the following instruction
selection from the builtin function:
bm(n)zi builtin -> vselect node -> bins[lr]i machine instruction
In case of bm(n)zi having an immediate which has either its high or low bits
set, a bins[lr] instruction can be selected through the selectVSplatMask[LR]
function. The function counts the number of bits set, and that value is
being passed to the bins[lr]i instruction as its immediate, which in turn
copies immediate modulo the size of the element in bits plus 1 as per specs,
where we get the off-by-one-error.
The other case is:
bins[lr]i -> vselect node -> bsel.v
In this case, a bsel.v instruction gets selected with a mask having one bit
less set than required.
Patch by Stefan Maksimovic.
Differential Revision: https://reviews.llvm.org/D30579
llvm-svn: 299768
By target hookifying getRegisterType, getNumRegisters, getVectorBreakdown,
backends can request that LLVM to scalarize vector types for calls
and returns.
The MIPS vector ABI requires that vector arguments and returns are passed in
integer registers. With SelectionDAG's new hooks, the MIPS backend can now
handle LLVM-IR with vector types in calls and returns. E.g.
'call @foo(<4 x i32> %4)'.
Previously these cases would be scalarized for the MIPS O32/N32/N64 ABI for
calls and returns if vector types were not legal. If vector types were legal,
a single 128bit vector argument would be assigned to a single 32 bit / 64 bit
integer register.
By teaching the MIPS backend to inspect the original types, it can now
implement the MIPS vector ABI which requires a particular method of
scalarizing vectors.
Previously, the MIPS backend relied on clang to scalarize types such as "call
@foo(<4 x float> %a) into "call @foo(i32 inreg %1, i32 inreg %2, i32 inreg %3,
i32 inreg %4)".
This patch enables the MIPS backend to take either form for vector types.
Reviewers: zoran.jovanovic, jaydeep, vkalintiris, slthakur
Differential Revision: https://reviews.llvm.org/D27845
llvm-svn: 299766
A test case was found with llvm-stress that caused DAGCombiner to crash
when compiling for an older subtarget without vector support.
SystemZTargetLowering::combineTruncateExtract() should do nothing for older
subtargets.
This check was placed in canTreatAsByteVector(), which also helps in a few
other places.
Review: Ulrich Weigand
llvm-svn: 299763
It turns out -float-abi=hard doesn't set the hard float calling
convention for libcalls. We need to use a hard float triple instead
(e.g. gnueabihf).
llvm-svn: 299761
Summary:
Difference beetween PreRegAlloc() and MachineSSAOptimization() are that the former is run despite of -O0 optimization level. In my undestanding SiShrinkInstructions and SDWAPeephole shouldn't run when optimizations are disabled.
With this change order of passes will not change.
Reviewers: arsenm, vpykhtin, rampitec
Subscribers: qcolombet, kzhuravl, wdng, nhaehnle, yaxunl, dstuttard, tpr, t-tye
Differential Revision: https://reviews.llvm.org/D31705
llvm-svn: 299757
Legalize to a libcall.
On this occasion, also start allowing soft float subtargets. For the
moment G_FREM is the only legal floating point operation for them.
llvm-svn: 299753
The new codepath has been in the tree for years, and there isn't any
reason to use two codepaths here.
Differential Revision: https://reviews.llvm.org/D30596
llvm-svn: 299723
This is possible in ways that are not compiler bugs,
so stop asserting on them.
This emits an extra error when emitting objects when it
can't encode the new pseudo, but I'm not sure that matters.
llvm-svn: 299712
In LowerMUL, the chain information is not preserved for the new
created Load SDNode.
For example, if a Store alias with one of the operand of Mul.
The Load for that operand need to be scheduled before the Store.
The dependence is recorded in the chain of Store, in TokenFactor.
However, when lowering MUL, the SDNodes for the new Loads for
VMULL are not updated in the TokenFactor for the Store. Thus the
chain is not preserved for the lowered VMULL.
llvm-svn: 299701
Summary:
Host CPU detection now supports Kryo, so we need to recognize it in ARM
target.
Reviewers: mcrosier, t.p.northover, rengolin, echristo, srhines
Reviewed By: t.p.northover, echristo
Subscribers: aemerson
Differential Revision: https://reviews.llvm.org/D31775
llvm-svn: 299674
If a workgroup size is known to be not greater than wavefront size
the s_barrier instruction is not needed since all threads are guarantied
to come to the same point at the same time.
Differential Revision: https://reviews.llvm.org/D31731
llvm-svn: 299659
During the optimisation of jump tables in the constant island pass,
an extra ADD could be left over, now dead but not removed.
Differential Revision: https://reviews.llvm.org/D31389
llvm-svn: 299634
This is a follow-on to r299096 which added support for fmadd.
Subtract does not have the case where with two multiply operands we commute in
order to fuse with the multiply with the fewer uses.
llvm-svn: 299572
Commit r298799 changed code that made the XFAIL on MachineBranchProb.ll
irrelevant, but some configurations still failed. I can't reproduce it
locally, so I'm hoping that enabling this will tell me if some
configurations will really fail or if they were just too slow.
llvm-svn: 299558
This is a generic combine enabled via target hook to reduce icmp logic as discussed in:
https://bugs.llvm.org/show_bug.cgi?id=32401
It's likely that other targets will want to enable this hook for scalar transforms,
and there are probably other patterns that can use bitwise logic to reduce comparisons.
Note that we are missing an IR canonicalization for these patterns, and we will probably
prefer the pair-of-compares form in IR (shorter, more likely to fold).
Differential Revision: https://reviews.llvm.org/D31483
llvm-svn: 299542
When DAGCombiner visits a SIGN_EXTEND_INREG of a BUILD_VECTOR with
constant operands, a new BUILD_VECTOR node will be created transformed
constants.
Llvm-stress found a case where the new BUILD_VECTOR had constant operands
of an illegal type, because the (legal) element type is in fact not a legal
scalar type.
This patch changes this so that the new BUILD_VECTOR has the same operand
type as the old one.
Review: Eli Friedman, Nirav Dave
https://bugs.llvm.org//show_bug.cgi?id=32422
llvm-svn: 299540
Before r294774, there was a problem when lowering broadcasts to use
128-bit subvectors.
When we looked through a bitcast to find the broadcast input, we'd keep
using the original type, so you'd end up with things like:
(v8f32 (broadcast
(v4f32 (extract_subvector
(v8i32 V),
...))
))
r294774 fixed it to always emit subvectors with the scalar type of the
original source.
It also introduced some asserts, to check that we use scalars with
the same size, and vectors with the same number of elements.
The scalar size equality is checked earlier when looking through bitcasts,
and is a useful assert.
However, the number of elements don't have to be identical: we're always
going to extract a 128-bit subvector, and we can have different size
inputs if we looked through a concat_vector to find a 256-bit source.
Relax the overzealous assert.
Replace it with a check of the original source vector being 256 or 512
bits. If it's 128 bits, we can't extract_subvector from it.
Fixes PR32371.
llvm-svn: 299490
This improves upon r246462: that prevented FMOVs from being emitted
for the cross-class INSERT_SUBREGs by disabling the formation of
INSERT_SUBREGs of LOAD. But the ld1.s that we started selecting
caused us to introduce partial dependencies on the vector register.
Avoid that by using SCALAR_TO_VECTOR: it's a first-class citizen that
is folded away by many patterns, including the scalar LDRS that we
want in this case.
Credit goes to Adam for finding the issue!
llvm-svn: 299482
This mode is just like -mcmodel=small except that it moves the
thread pointer from TPIDR_EL0 to TPIDR_EL1.
Patch by Roland McGrath.
Differential Revision: https://reviews.llvm.org/D31624
llvm-svn: 299462
This patch is a part one of two reviews, one for the clang and the other for LLVM.
The patch deletes the back-end intrinsics and adds support for them in the auto upgrade.
Differential Revision: https://reviews.llvm.org/D31393
llvm-svn: 299432
Summary:
Lift the restrictions that prevented the tree walking introduced in the
previous change and add support for patterns like:
(G_ADD (G_MUL (G_SEXT $src1), (G_SEXT $src2)), $src3) -> SMADDWrrr $dst, $src1, $src2, $src3
Also adds support for G_SEXT and G_ZEXT to support these cases.
One particular aspect of this that I should draw attention to is that I've
tried to be overly conservative in determining the safety of matches that
involve non-adjacent instructions and multiple basic blocks. This is intended
to be used as a cheap initial check and we may add a more expensive check in
the future. The current rules are:
* Reject if any instruction may load/store (we'd need to check for intervening
memory operations.
* Reject if any instruction has implicit operands.
* Reject if any instruction has unmodelled side-effects.
See isObviouslySafeToFold().
Reviewers: t.p.northover, javed.absar, qcolombet, aditya_nandakumar, ab, rovka
Reviewed By: ab
Subscribers: igorb, dberris, llvm-commits, kristof.beyls
Differential Revision: https://reviews.llvm.org/D30539
llvm-svn: 299430
This patch teaches the hazard scheduler how to handle empty blocks
when search for the next real instruction when dealing with forbidden
slots.
Reviewers: slthakur
Differential Revision: https://reviews.llvm.org/D31293
llvm-svn: 299427
PSADBW pattern currently supports the 32 bit IR pattern and only GLT (greather than) comparison.
The patch extends the pattern to catch also 64 bit IR pattern and includes all other comparison types (not only GLT).
Differential Revision: https://reviews.llvm.org/D31577
llvm-svn: 299425
It can be costly to transfer from the gprs to the xmm registers and can prevent loads merging.
This patch splits vXi16/vXi32/vXi64 BUILD_VECTORS that use the same operand in multiple elements into a BUILD_VECTOR with only a single insertion of each of those elements and then performs an unary shuffle to duplicate the values.
There are a couple of minor regressions this patch unearths due to some missing MOVDDUP/BROADCAST folds that I will address in a future patch.
Note: Now that vector shuffle lowering and combining is pretty good we should be reusing that instead of duplicating so much in LowerBUILD_VECTOR - this is the first of several patches to address this.
Differential Revision: https://reviews.llvm.org/D31373
llvm-svn: 299387
The x86_64 ABI requires that the stack is 16 byte aligned on function calls. Thus, the 8-byte error code, which is pushed by the CPU for certain exceptions, leads to a misaligned stack. This results in bugs such as Bug 26413, where misaligned movaps instructions are generated.
This commit fixes the misalignment by adjusting the stack pointer in these cases. The adjustment is done at the beginning of the prologue generation by subtracting another 8 bytes from the stack pointer. These additional bytes are popped again in the function epilogue.
Fixes Bug 26413
Patch by Philipp Oppermann.
Differential Revision: https://reviews.llvm.org/D30049
llvm-svn: 299383
Summary:
Move the aarch64-type-promotion pass within the existing type promotion framework in CGP.
This change also support forking sexts when a new sext is required for promotion.
Note that change is based on D27853 and I am submitting this out early to provide a better idea on D27853.
Reviewers: jmolloy, mcrosier, javed.absar, qcolombet
Reviewed By: qcolombet
Subscribers: llvm-commits, aemerson, rengolin, mcrosier
Differential Revision: https://reviews.llvm.org/D28680
llvm-svn: 299379
This reverts commit r299047 which is incorrect because the
simplification may result in incorrect propogation of undefs to users of
the folded shuffle.
Thanks to Andrea Di Biagio for pointing this out.
llvm-svn: 299368
The code already allowed vector types in via "isInteger" (which might want
a more specific name), so use splat-friendly constant predicates to match
those types.
llvm-svn: 299304
This can only happen when we have a mix of zero and undef elements and the two vectors have a different arrangement of zeros/undefs. The shuffle should eventually be constant folded to all zeros.
Fixes PR32484.
llvm-svn: 299291
REG_SEQUENCE falls into the same category as COPY for operands mapping:
- They don't have MCInstrDesc with register constraints
- The input variable could use whatever register classes
- It is possible to have register class already assigned to the operands
In particular, given REG_SEQUENCE are always target specific because of
the subreg indices. Those indices must apply to the register class of
the definition of the REG_SEQUENCE and therefore, the target must set a
register class to that definition. As a result, the generic code can
always use that register class to derive a valid mapping for a
REG_SEQUENCE.
llvm-svn: 299285
This patch refactors the code used in llc such that all the users of the
addPassesToEmitFile API have access to a homogeneous way of handling
start/stop-after/before options right out of the box.
Previously each user would have needed to duplicate this logic and set
up its own options.
NFC
llvm-svn: 299282
(and (setlt X, 0), (setlt Y, 0)) --> (setlt (and X, Y), 0)
We have 7 similar folds, but this one got away. The fact that the
x86 test with a branch didn't change is probably a separate bug. We
may also be missing this and the related folds in instcombine.
llvm-svn: 299252
Add a new node to act as a fancy bitcast from f16 operations to
i32 that implicitly zero the high 16-bits of the result.
Alternatively could try making v2f16 legal and canonicalizing
on build_vectors.
llvm-svn: 299246
These are the same tests added for x86 with r299238,
but PPC doesn't specify all branches as cheap, so we
see different patterns in tests with branches.
llvm-svn: 299244
Summary:
This feature enables folding of logical shift operations of up to 3 places into addressing mode on Kryo and Falkor that have a fastpath LSL.
Reviewers: mcrosier, rengolin, t.p.northover
Subscribers: junbuml, gberry, llvm-commits, aemerson
Differential Revision: https://reviews.llvm.org/D31113
llvm-svn: 299240
Our _MM_HINT_T0/T1 constant values are 3/2 which matches gcc, but not icc or Intel documentation. Interestingly gcc had this same bug on their implementation of the gather/scatter builtins at one point too.
Fixes PR32411.
llvm-svn: 299234
Implementation of TargetInstrInfo::findCommutedOpIndices for MIPS target,
restricting commutativity to second and third operand only for
dpaadd_[su].df instructions therein.
Prior to this change, there were cases where the vector that is to be added
to the dot product of the other two could take a position other than the
first one in the instruction, generating false output in the destination
vector.
Such behavior has been noticed in the two functions generating v2i64 output
values so far. Other ones may exhibit such behavior as well, just not for
the vector operands which are present in the test at the moment.
Tests altered so that the function's first operand is a constant splat so
that it can be loaded with a ldi instruction, since that is the case in
which the erroneous instruction operand placement has occurred. We check
that the register which is present in the ldi instruction is placed as the
first operand in the corresponding dpadd instruction.
Patch by Stefan Maksimovic.
Differential Revision: https://reviews.llvm.org/D30827
llvm-svn: 299223
Since LOCR only accepts GR32 virtual registers, its operands must be copied
into this regclass in insertSelect(), when an LOCR is built. Otherwise, the
case where the source operand was GRX32 will produce invalid IR.
Review: Ulrich Weigand
llvm-svn: 299220
Currently ComputeNumSignBits returns the minimum number of sign bits for all elements of vector data, when we may only be interested in one/some of the elements.
This patch adds a DemandedElts argument that allows us to specify the elements we actually care about. The original ComputeNumSignBits implementation calls with a DemandedElts demanding all elements to match current behaviour. Scalar types set this to 1.
I've only added support for BUILD_VECTOR and EXTRACT_VECTOR_ELT so far, all others will default to demanding all elements but can be updated in due course.
Followup to D25691.
Differential Revision: https://reviews.llvm.org/D31311
llvm-svn: 299219
Even on older subtargets that lack vector support, there may be vector values
with just one element in the input program. These are converted during DAG
legalization to scalar values.
The pre-legalize SystemZ DAGCombiner methods should in this circumstance not
touch these nodes. This patch adds a check for this in
SystemZTargetLowering::combineEXTRACT_VECTOR_ELT().
Review: Ulrich Weigand
llvm-svn: 299213
Previously compiler often extracted common immediates into specific register, e.g.:
```
%vreg0 = S_MOV_B32 0xff;
%vreg2 = V_AND_B32_e32 %vreg0, %vreg1
%vreg4 = V_AND_B32_e32 %vreg0, %vreg3
```
Because of this SDWA peephole failed to find SDWA convertible pattern. E.g. in previous example this could be converted into 2 SDWA src operands:
```
SDWA src: %vreg2 src_sel:BYTE_0
SDWA src: %vreg4 src_sel:BYTE_0
```
With this change peephole check if operand is either immediate or register that is copy of immediate.
llvm-svn: 299202
If set to false it does not remove global aliases. With this parameter
set to false it should be safe to run the pass before link.
Differential Revision: https://reviews.llvm.org/D31489
llvm-svn: 299108
Now alternatively to the TargetOption.AllowFPOpFusion global flag, FMUL->FADD
can also use the per operation FMF to allow fusion.
The idea here is not to port everything to the new scheme (e.g. fused
multiply-and-sub will be ported later) but that this work all the way from
clang.
The transformation is conditionalized on *both* the FADD and the FMUL having
the FMF contract flag.
Differential Revision: https://reviews.llvm.org/D31169
llvm-svn: 299096
In the long-term, we want to replace statistics with something
finer-grained that lets us gather per-function data.
Remarks are that replacement.
Create an ORE instance in SelectionDAGISel, and pass it to
SelectionDAG.
SelectionDAG was used so that we can emit remarks from all
SelectionDAG-related code, including TargetLowering and DAGCombiner.
This isn't used in the current patch but Adam tells me he's interested
for the fp-contract combines.
Use the ORE instance to emit FastISel failures as remarks (instead of
the mix of dbgs() dumps and statistics that we currently have).
Eventually, we want to have an API that tells us whether remarks are
enabled (http://llvm.org/PR32352) so that we don't emit expensive
remarks (in this case, dumping IR) when it's not needed. For now, use
'isEnabled' as a crude replacement.
This does mean that the replacement for '-fast-isel-verbose' is now
'-pass-remarks-missed=isel'. Additionally, clang users also need to
enable remark diagnostics, using '-Rpass-missed=isel'.
This also removes '-fast-isel-verbose2': there are no static statistics
that we want to only enable in asserts builds, so we can always use
the remarks regardless of the build type.
Differential Revision: https://reviews.llvm.org/D31405
llvm-svn: 299093
This will result in a KMOVW or KMOVD being emitted during register allocation. And in at least some cases this might allow the register coalescer to remove the copy all together.
llvm-svn: 298984
In PPCBoolRetToInt bool value is changed to i32 type. On ppc64 it may introduce an extra zero extension for the return value. This patch changes the integer type to i64 to avoid the zero extension on ppc64.
This patch fixed PR32442.
Differential Revision: https://reviews.llvm.org/D31407
llvm-svn: 298955
This is less important than increase threshold for private memory,
but still brings performance improvements in a wide range of tests.
Unrolling more for local memory serves three purposes: it allows
to combine ds operations if offset becomes static, saves registers
used for offsets in case of static offsets, and allows better lds
latency hiding.
Differential Revision: https://reviews.llvm.org/D31412
llvm-svn: 298948
We currently perform the various fp_to_sint XMM conversion and then transfer to the MMX register (on 32-bit via the stack).
This patch improves support for MOVDQ2Q XMM to MMX transfers and adds the XMM->MMX fp_to_sint direct conversion patterns. The SSE2 specifications are the same as for XMM->XMM and XMM->MMX rounding/exceptions/etc.
Differential Revision: https://reviews.llvm.org/D30868
llvm-svn: 298943
Previously it was covered by the internalization. It turns out we cannot
run internalizer in FE, it break separate compilation tests. Thus early
inliner gets its own option.
Differential Revision: https://reviews.llvm.org/D31429
llvm-svn: 298935
Deal with case that initial node is deleted during dag-combine leading
to an assertional failure in promoteIntShiftOp.
Fixes PR32420.
Reviewers: spatel, RKSimon
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D31403
llvm-svn: 298931
We've had several bugs(PR32256, PR32241) recently that resulted from usages of AH/BH/CH/DH either before or after a copy to/from a mask register.
This ultimately occurs because we create COPY_TO_REGCLASS with VK1 and GR8. Then in CopyToFromAsymmetricReg in X86InstrInfo we find a 32-bit super register for the GR8 to emit the KMOV with. But as these tests are demonstrating, its possible for the GR8 register to be a high register and we end up doing an accidental extra or insert from bits 15:8.
I think the best way forward is to stop making copies directly between mask registers and GR8/GR16. Instead I think we should restrict to only copies between mask registers and GR32/GR64 and use EXTRACT_SUBREG/INSERT_SUBREG to handle the conversion from GR32 to GR16/8 or vice versa.
Unfortunately, this complicates fastisel a bit more now to create the subreg extracts where we used to create GR8 copies. We can probably make a helper function to bring down the repitition.
This does result in KMOVD being used for copies when BWI is available because we don't know the original mask register size. This caused a lot of deltas on tests because we have to split the checks for KMOVD vs KMOVW based on BWI.
Differential Revision: https://reviews.llvm.org/D30968
llvm-svn: 298928
We want to check each test on each target, so we need another prefix
when SSE and AVX diverge (as they will if we handle 32-byte and higher).
llvm-svn: 298926
Reorder work in PromoteIntBinOp to prevent stale (deleted) nodes from
being used.
Fixes PR32340 and PR32345.
Reviewers: hfinkel, dbabokin
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D31148
llvm-svn: 298923
Summary:
G_LOAD/G_STORE, add alternative RegisterBank mapping.
For G_LOAD, Fast and Greedy mode choose the same RegisterBank mapping (GprRegBank ) for the G_GLOAD + G_FADD , can't get rid of cross register bank copy GprRegBank->VecRegBank.
Reviewers: zvi, rovka, qcolombet, ab
Reviewed By: zvi
Subscribers: llvm-commits, dberris, kristof.beyls, eladcohen, guyblank
Differential Revision: https://reviews.llvm.org/D30979
llvm-svn: 298907
This patch enables schedulers to specify instructions that
cannot be issued with any other instructions.
It also fixes BeginGroup/EndGroup.
Reviewed by: Andrew Trick
Differential Revision: https://reviews.llvm.org/D30744
llvm-svn: 298885
We're not to the point of supporting the load/store patterns yet
(because they extensively use PatFrags).
But in the meantime, we can implement some of the simplest addressing
modes.
llvm-svn: 298863
CBZ/CBNZ represent a substantial portion of all conditional branches.
Look through G_ICMP to select them.
We can't use tablegen yet because the existing patterns match an
AArch64ISD node.
llvm-svn: 298856
Among other things, this allows Machine LICM to hoist a costly 'mrs'
instruction from within a loop.
Differential Revision: http://reviews.llvm.org/D31151
llvm-svn: 298851
This is a patch for an on-going bugzilla bug 21281 on the generated X86 code for a matrix transpose8x8 subroutine which requires vector interleaving. The generated code in AVX2 is currently non-optimal and requires 60 instructions as opposed to only 40 instructions generated for AVX1.
The patch includes a fix for the AVX2 case where vector unpack instructions use less operations than the vector blend operations available in AVX2.
In this case using vector unpack instructions is more efficient.
Reviewers:
zvi
delena
igorb
craig.topper
guyblank
eladcohen
m_zuckerman
aymanmus
RKSimon
llvm-svn: 298840
Fixed -verify-machineinstrs errors in fast-isel-select-sse.ll (one of many in PR27481)
The VMOVSSZrr/VMOVSSZrrk and VMOVSDZrr/VMOVSDZrrk instructions were assuming both source registers were V128X when the second is actually supposed to be FR32X/FR64X
Differential Revision: https://reviews.llvm.org/D31200
llvm-svn: 298805
This is the payoff for D31156 - if a target has efficient comparison instructions for vector-sized equality,
we can replace memcmp calls with inline code that is both smaller and faster.
Differential Revision: https://reviews.llvm.org/D31290
llvm-svn: 298775
Switch data layout by target triple environment amdgiz and amdgizcl indicating using of an address space mapping in which generic address space is 0.
amdgiz is for non-OpenCL environment where generic address space is 0.
amdgizcl is for OpenCL environment where generic address space is 0.
Differential Revision: https://reviews.llvm.org/D31211
llvm-svn: 298758
If the branch condition for a loop was a phi which itself
was fed from a phi from a loop, it isn't safe to try
to delete the phi until after the loop is handled.
llvm-svn: 298737
StructurizeCFG can't handle cases with multiple
returns creating regions with multiple exits.
Create a copy of UnifyFunctionExitNodes that only
unifies exit nodes that skips exit nodes
with uniform branch sources.
llvm-svn: 298729
Summary:
For the following CFG:
A->B
B->C
A->C
If there is another edge B->D, then ABC should not be considered as triangle.
Reviewers: davidxl, iteratee
Reviewed By: iteratee
Subscribers: nemanjai, llvm-commits
Differential Revision: https://reviews.llvm.org/D31310
llvm-svn: 298661
- Avoid explosive growth of the simplification queue by not queuing
expressions that are alredy in it.
- Add an iteration counter and abort after a sufficiently large number
of iterations (assuming that it's a symptom of an infinite loop).
llvm-svn: 298655
Summary:
The true and false operands for the CMOV are operands 0 and 1.
ARMISelLowering.cpp::computeKnownBits was looking at operands 1 and 2
instead. This can cause CMOV instructions to be incorrectly folded into
BFI if value set by the CMOV is another CMOV, whose known bits are
computed incorrectly.
This patch fixes the issue and adds a test case.
Reviewers: kristof.beyls, jmolloy
Subscribers: llvm-commits, aemerson, srhines, rengolin
Differential Revision: https://reviews.llvm.org/D31265
llvm-svn: 298624
Summary: Add tests for all atomic operations for powerpc64le, so that all changes can be easily examined.
Reviewers: kbarton, hfinkel, echristo
Subscribers: mehdi_amini, nemanjai, llvm-commits
Differential Revision: https://reviews.llvm.org/D31285
llvm-svn: 298614
Add tests for vector lengths that could be handled without a libcall.
There's an explicit test for 'nobuiltin', so there's not much value
in a separate run that checks that same behavior over and over again.
llvm-svn: 298611
Summary:
1. Support pointer type as function argumnet and return value
2. G_STORE/G_LOAD - set legal action for i8/i16/i32/i64/f32/f64/vec128
3. RegisterBank - support typeless operations like G_STORE/G_LOAD, for scalar use GPR bank.
4. Support instruction selection for G_LOAD/G_STORE
Reviewers: zvi, rovka, ab, qcolombet
Reviewed By: rovka
Subscribers: llvm-commits, dberris, kristof.beyls, eladcohen, guyblank
Differential Revision: https://reviews.llvm.org/D30973
llvm-svn: 298609
Up until now, vpmovm2 instruction described its destination operand size
by the source operand size. This patch adds new pattern for the vpmovm2
instruction. The node describes new expansion of the destination (from
{128|256} to 512).
Differential Revision: https://reviews.llvm.org/D30654
llvm-svn: 298586
- Rename runtime metadata -> code object metadata
- Make metadata not flow
- Switch enums to use ScalarEnumerationTraits
- Cleanup and move AMDGPUCodeObjectMetadata.h to AMDGPU/MCTargetDesc
- Introduce in-memory representation for attributes
- Code object metadata streamer
- Create metadata for isa and printf during EmitStartOfAsmFile
- Create metadata for kernel during EmitFunctionBodyStart
- Finalize and emit metadata to .note during EmitEndOfAsmFile
- Other minor improvements/bug fixes
Differential Revision: https://reviews.llvm.org/D29948
llvm-svn: 298552
This patch changes the behavior of IRTranslating intrinsics where we
now create VREG + G_CONSTANT for ConstantInt values. We already do this
for FloatingPoint values. This makes it easier for the backends to
select code and it won't have to de-duplicate creation+selection of
constants.
Reviewed by: ab
llvm-svn: 298473
Quentin points out that r298358 would cause us to emit different code
with debug info. That's a big no-no; also erase the instructions that
only live thanks to DBG_VALUE users.
Adrian explained how this is an existing problem and an OK thing to do:
clang has allocas for all variables so shouldn't be affected at -O0, but
swift uses a bit of inlineasm to explicitly keep values live for the
purpose of debug info quality. I'm not sure there is a better scheme.
llvm-svn: 298460
MI can represent fallthrough to layout successor blocks, and our
post-isel representation uses that extensively.
We might as well use it too, to avoid translating and carrying along
unnecessary branches.
llvm-svn: 298459
This is used for a specific type of return to a shader part's
epilog code. Rename to try avoiding confusion from a true
call's return.
llvm-svn: 298452
Fix two problems related to r298025:
- SplitKit would create duplicate VNIs in some cases leading to crashs
when hoisting copies.
- VirtRegMap could fail expanding copies at the beginning of a basic
block.
This fixes http://llvm.org/PR32353
llvm-svn: 298448
Currently the default C calling convention functions are treated
the same as compute kernels. Make this explicit so the default
calling convention can be changed to a non-kernel.
Converted with perl -pi -e 's/define void/define amdgpu_kernel void/'
on the relevant test directories (and undoing in one place that actually
wanted a non-kernel).
llvm-svn: 298444
A bool is represented by a single byte, which the ARM ABI requires to be either
0 or 1. So we cannot use G_ANYEXT when legalizing the type.
llvm-svn: 298439
This adds a parameter to @llvm.objectsize that makes it return
conservative values if it's given null.
This fixes PR23277.
Differential Revision: https://reviews.llvm.org/D28494
llvm-svn: 298430
including the amended (no UB anymore) fix for adding/subtracting -2147483648.
This reverts r298328 "[ARM] Revert r297443 and r297820."
and partially reverts r297842 "Revert "[Thumb1] Fix the bug when adding/subtracting -2147483648""
llvm-svn: 298417
[Hexagon] Recognize polynomial-modulo loop idiom again
Regain the ability to recognize loops calculating polynomial modulo
operation. This ability has been lost due to some changes in the
preceding optimizations. Add code to preprocess the IR to a form
that the pattern matching code can recognize.
llvm-svn: 298400
We could do better by splitting any oversized type into whatever vector size the target supports,
but I left that for future work if it ever comes up. The motivating case is memcmp() calls on 16-byte
structs, so I think we can wire that up with a TLI hook that feeds into this.
Differential Revision: https://reviews.llvm.org/D31156
llvm-svn: 298376
Summary:
First iteration of SDWA peephole.
This pass tries to combine several instruction into one SDWA instruction. E.g. it converts:
'''
V_LSHRREV_B32_e32 %vreg0, 16, %vreg1
V_ADD_I32_e32 %vreg2, %vreg0, %vreg3
V_LSHLREV_B32_e32 %vreg4, 16, %vreg2
'''
Into:
'''
V_ADD_I32_sdwa %vreg4, %vreg1, %vreg3 dst_sel:WORD_1 dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:DWORD
'''
Pass structure:
1. Iterate over machine instruction in basic block and try to apply "SDWA patterns" to each of them. SDWA patterns match machine instruction into either source or destination SDWA operand. E.g. ''' V_LSHRREV_B32_e32 %vreg0, 16, %vreg1''' is matched to source SDWA operand '''%vreg1 src_sel:WORD_1'''.
2. Iterate over found SDWA operands and find instruction that could be potentially coverted into SDWA. E.g. for source SDWA operand potential instruction are all instruction in this basic block that uses '''%vreg0'''
3. Iterate over all potential instructions and check if they can be converted into SDWA.
4. Convert instructions to SDWA.
This review contains basic implementation of SDWA peephole pass. This pass requires additional testing fot both correctness and performance (no performance testing done).
There are several ways this pass can be improved:
1. Make this pass work on whole function not only basic block. As I can see this can be done right now without changes to pass.
2. Introduce more SDWA patterns
3. Introduce mnemonics to limit when SDWA patterns should apply
Reviewers: vpykhtin, alex-t, arsenm, rampitec
Subscribers: wdng, nhaehnle, mgorny
Differential Revision: https://reviews.llvm.org/D30038
llvm-svn: 298365
This patch fixes an issue in the Optimize LEAs pass where redundant LEAs were
not removed because they were being used by debug values. The debug values are
now ignored when determining whether LEAs are redundant.
For now the debug values for the redundant LEAs are marked as undefined,
effectively lost. The intention is for a follow up patch which will attempt to
preserve the debug values where possible.
Patch by Andrew Ng.
Differential Revision: https://reviews.llvm.org/D30835
llvm-svn: 298360
Previously, PromoteIntRes_TRUNCATE() did not handle the case where
the operand needs widening, which resulted in llvm_unreachable().
This patch adds the needed handling, along with a test case.
Review: Eli Friedman, Simon Pilgrim.
https://reviews.llvm.org/D31077
llvm-svn: 298357
The def operand of the new LG/LD should have the old def operands
flags and subreg index.
New test: test/CodeGen/SystemZ/fold-memory-op-impl.ll
Review: Ulrich Weigand
llvm-svn: 298341
The glueless lowering of addc/adde in Thumb1 has known serious
miscompiles (see https://reviews.llvm.org/D31081), and r297820
causes an infinite loop for certain constructs. It's not
clear when they will be fixed, so let's just take them out
of the tree for now.
(I resolved a small conflict with r297453.)
llvm-svn: 298328
The special case of zero sized values was previously not handled correctly.
This patch handles this by not promoting if the size is zero.
Patch by Tim Neumann.
Differential Revision: https://reviews.llvm.org/D31116
llvm-svn: 298320
Regain the ability to recognize loops calculating polynomial modulo
operation. This ability has been lost due to some changes in the
preceding optimizations. Add code to preprocess the IR to a form
that the pattern matching code can recognize.
llvm-svn: 298282
Move the check for "MF->hasWinCFI()" up into the calculation of the
shouldEmitMoves boolean, rather than putting it in the early returning
if. This ensures that endFunction doesn't try to emit .seh_* directives
for leaf functions.
llvm-svn: 298276
This commit adds a parameter that lets us pass in the calling convention
of the call to CallLowering::lowerCall. This allows us to handle
situations where the calling convetion of the callee is different from
that of the caller.
Differential Revision: https://reviews.llvm.org/D31039
llvm-svn: 298254
Summary:
Currently we handle these intrinsics at isel with special patterns. But as they just map to normal logic operations, we should just handle them at lowering. This will expose them to DAG combine optimizations. Right now the kor-sequence test generates a bunch of regclass copies between GR16 and VK16 that the peephole optimizer and/or register coallescing are removing to keep everything in the mask domain. By handling the logic op intrinsics earlier, these copies become bitcasts in the DAG and get removed by DAG combine which seems more robust.
This should help enable my plan to stop copying between K registers and GR8/GR16. The peephole optimizer can't remove a chain of copies between K and GR32 with insert_subreg/extract_subreg present in the chain so the kor-sequence test break. But this patch should dodge the problem entirely.
Reviewers: zvi, delena, RKSimon, igorb
Reviewed By: igorb
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D31056
llvm-svn: 298228
We make the assumption in most of our constant folding code that a fp2int will target an integer of 128-bits or less, calling the APFloat::convertToInteger with only uint64_t[2] of raw bits for the result.
Fuzz testing (PR24662) showed that we don't handle other cases at all, resulting in stack overflows and all sorts of crashes.
This patch uses the APSInt version of APFloat::convertToInteger instead to better handle such cases.
Differential Revision: https://reviews.llvm.org/D31074
llvm-svn: 298226
Folding instructions when selecting can cause them to become dead.
Don't select these dead instructions (if they don't have other side
effects, and don't define physical registers).
Preserve existing tests by adding COPYs.
In some tests, the G_CONSTANT vregs never get constrained to a class:
the only use of the vreg was folded into another instruction, so the
G_CONSTANT, now dead, never gets selected.
llvm-svn: 298224
And remove some redundant bitcast tests.
Also split the test functions themselves: it makes it obvious to see
what's tested where and what isn't, it makes the tests much easier to
read and manually update, and, most importantly, it makes them almost
trivial to update using tooling. Yes, it's obnoxiously verbose, but
said tooling helps upgrade to better MIR syntax whenever available.
llvm-svn: 298222
The MIR printer dumps a string that describe the register mask of a function.
A static predefined list of register masks matches a static list of strings.
However when the register mask is not from the static predefined list, there is no descriptor string and the printer fails.
This patch adds support to custom register mask printing and dumping.
Also the list of callee saved registers (describing the registers that must be preserved for the caller) might be dynamic.
As such this data needs to be dumped and parsed back to the Machine Register Info.
Differential Revision: https://reviews.llvm.org/D30971
llvm-svn: 298207
This is direct port of HSAILAliasAnalysis pass, just cleaned for
style and renamed.
Differential Revision: https://reviews.llvm.org/D31103
llvm-svn: 298172
This commit adds the necessary target hooks for outlining in AArch64. It also
refactors the switch statement used in `getMemOpBaseRegImmOfsWidth` into a
more general function, `getMemOpInfo`. This allows the outliner to share that
code without copying and pasting it.
The AArch64 outliner can be run using -mllvm -enable-machine-outliner, as with
the X86-64 outliner.
The test for this pass verifies that the outliner does, in fact outline
functions, fixes up the stack accesses properly, and can correctly generate a
tail call. In the future, this test should be replaced with a MIR test, so that
we can properly test immediate offset overflows in fixed-up instructions.
llvm-svn: 298162
This is an ELF-specific thing that adds SHF_LINK_ORDER to the global's section
pointing to the metadata argument's section. The effect of that is a reverse dependency
between sections for the linker GC.
!associated does not change the behavior of global-dce. The global
may also need to be added to llvm.compiler.used.
Since SHF_LINK_ORDER is per-section, !associated effectively enables
fdata-sections for the affected globals, the same as comdats do.
Differential Revision: https://reviews.llvm.org/D29104
llvm-svn: 298157
Handle TokenFactors more aggressively in
SDValue::reachesChainWithoutSideEffects. This isn't really a
very effective change anymore because of other changes to
chain handling, but it's a cheap check, and the expanded
comments are still useful.
It might be possible to loosen the hasOneUse() requirement with a
deeper analysis, but a naive implementation of that check would be
expensive.
Differential Revision: https://reviews.llvm.org/D29845
llvm-svn: 298156
If the loop condition was an i1 phi with a constantexpr input, this
would add a loop intrinsic fed by a phi dependent on a call to
if.break in the same block. Insert the call in the loop header.
llvm-svn: 298121
Summary:
Use this code pattern when RAX is live, instead of emitting up to 2
billion adjustments:
pushq %rax
movabsq +-$Offset+-8, %rax
addq %rsp, %rax
xchg %rax, (%rsp)
movq (%rsp), %rsp
Try to clean this code up a bit while I'm here. In particular, hoist the
logic that handles the entire adjustment with `movabsq $imm, %rax` out
of the loop.
This negates the offset in the prologue and uses ADD because X86 only
has a two operand subtract which always subtracts from the destination
register, which can no longer be RSP.
Fixes PR31962
Reviewers: majnemer, sdardis
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D30052
llvm-svn: 298116
Summary:
Instead of just looking for a load which is mergable with Ext to form ExtLoad, trying to promote Exts as long as the cost is acceptable. This change is not a NFC as it continue promoting Exts even after finding a load during promotions; the change in arm64-codegen-prepare-extload.ll described in 2.b might show the case.
This change was motivated from D26524. Based on this change, I will move the transformation performed in aarch64-type-promotion into CGP.
Reviewers: jmolloy, qcolombet, mcrosier, javed.absar
Reviewed By: qcolombet
Subscribers: rengolin, llvm-commits, aemerson
Differential Revision: https://reviews.llvm.org/D27853
llvm-svn: 298114
As noted in the comment, we might want to account for this case,
but I didn't look at what that would mean for the asm.
I'm also not sure why this only reproduces with avx512, but I'm
putting a conservative fix in for now to avoid the crash.
Also, if both sides of an add are zexted, shouldn't we shrink that add?
https://bugs.llvm.org/show_bug.cgi?id=32316
llvm-svn: 298107
We weren't able to handle isel of the 128/256-bit FMA instructions when AVX512F was enabled but VLX and FMA weren't.
I didn't mask FeatureAVX512 imply FeatureFMA as I wasn't sure I wanted disabling FMA to also disable AVX512. Instead we just can't prevent FMA instructions if AVX512 is enabled.
Another option would be to promote 128/256-bit to 512-bit, do the operation and extract it. But that requires a lot of extra isel patterns. Since no CPUs exist that support AVX512, but not FMA just using the VEX instructions seems better.
llvm-svn: 298051
New SystemZ tests for the improved codegen of vector compare and select,
including cases with a logical combination of two compares.
Review: Ulrich Weigand.
https://reviews.llvm.org/D29489
llvm-svn: 298049
If one of the subregs of the 128 bit reg is undefined when splitMove() splits
a store into two instructions, a use of an undefined physical register
results.
To remedy this, an implicit use of the super register is added onto both new
instructions, along with propagated kill and undef flags.
This was discovered with llvm-stress, and that test case is attached as
test/CodeGen/SystemZ/splitMove_undefReg_mverifier.ll
Thanks to Matthias Braun for helping with a nice explanation.
Review: Ulrich Weigand
llvm-svn: 298047
This bug only exists on the scalar llvm.fma instrinsics. Looks like we don't test the llvm.fma intrinsics very thoroughly. In fact I don't see any tests for the vector versions.
llvm-svn: 298045
- This fixes a bug where subregister incompatible with the vregs register
class where used.
- Implement the case where multiple copies are necessary to cover a
given lanemask.
Differential Revision: https://reviews.llvm.org/D30438
llvm-svn: 298025
This allows the optimization to rearrange loads and stores more
aggressively. This doesn't really affect performance, but it helps
codesize.
Differential Revision: https://reviews.llvm.org/D30839
llvm-svn: 298021
I had ajusted the test case before when testing a chain of length 2, and then
reverted it with rL296845 when I switched to 3 triangles. After running
benchmarks and examining generated code at length 2 I forgot to put the test
back.
llvm-svn: 298000
Earlier stages of GlobalISel always use ConstantInt in G_CONSTANT so that's
what we should check for.
This fixes a crash introduced in r297782.
llvm-svn: 297968
Citing http://bugs.llvm.org/show_bug.cgi?id=32288
The DWARF generated by LLVM includes this location:
0x55 0x93 0x04 DW_OP_reg5 DW_OP_piece(4) When GCC's DWARF is simply
0x55 (DW_OP_reg5) without the DW_OP_piece. I believe it's reasonable
to assume the DWARF consumer knows which part of a register
logically holds the value (low bytes, high bytes, how many bytes,
etc) for a primitive value like an integer.
This patch gets rid of the redundant DW_OP_piece when a subregister is
at offset 0. It also adds previously missing subregister masking when
a subregister is followed by another operation.
(This reapplies r297960 with two additional testcase updates).
rdar://problem/31069390
https://reviews.llvm.org/D31010
llvm-svn: 297965
We can mark functions to always inline early in the opt. Since we do not have
call support this early inlining creates opportunities for inter-procedural
optimizations which would not occur otherwise.
Differential Revision: https://reviews.llvm.org/D31016
llvm-svn: 297958
Don't scalarize VSELECT->SETCC when operands/results needs to be widened,
or when the type of the SETCC operands are different from those of the VSELECT.
(VSELECT SETCC) and (VSELECT (AND/OR/XOR (SETCC,SETCC))) are handled.
The previous splitting of VSELECT->SETCC in DAGCombiner::visitVSELECT() is
no longer needed and has been removed.
Updated tests:
test/CodeGen/ARM/vuzp.ll
test/CodeGen/NVPTX/f16x2-instructions.ll
test/CodeGen/X86/2011-10-19-widen_vselect.ll
test/CodeGen/X86/2011-10-21-widen-cmp.ll
test/CodeGen/X86/psubus.ll
test/CodeGen/X86/vselect-pcmp.ll
Review: Eli Friedman, Simon Pilgrim
https://reviews.llvm.org/D29489
llvm-svn: 297930
This produces a 1% speedup on an important internal Google benchmark
(protocol buffers), with no other regressions in google or in the llvm
test-suite. Only 5 targets in the entire llvm test-suite are affected,
and on those 5 targets the size increase is 0.027%
llvm-svn: 297925
Currently, we create a G_CONSTANT for every "synthetic" integer
constant operand (for instance, for the G_GEP offset).
Instead, share the G_CONSTANTs we might have created by going through
the ValueToVReg machinery.
When we're emitting synthetic constants, we do need to get Constants from
the context. One could argue that we shouldn't modify the context at
all (for instance, this means that we're going to use a tad more memory
if the constant wasn't used elsewhere), but constants are mostly
harmless. We currently do this for extractvalue and all.
For constant fcmp, this does mean we'll emit an extra COPY, which is not
necessarily more optimal than an extra materialized constant.
But that preserves the current intended design of uniqued G_CONSTANTs,
and the rematerialization problem exists elsewhere and should be
resolved with a single coherent solution.
llvm-svn: 297875
computeKnownBits didn't handle fp_to_fp16 to report
the high bits as 0. ARM maps the generic node to an instruction
that does not modify the high bits of the register, so introduce
a target node where the high bits are known 0.
llvm-svn: 297873
If we got unlucky with register allocation and actual constpool placement, we
could end up producing a tTBB_JT with an index that's already been clobbered.
Technically, we might be able to fix this situation up with a MOV, but I think
the constant islands pass is complex enough without having to deal with more
weird edge-cases.
llvm-svn: 297871
Now that we preserve the IR layout, we would end up with all the newly
synthesized switch comparison blocks at the end of the function.
Instead, use a hopefully more reasonable layout, with the comparison
blocks immediately following the switch comparison blocks.
llvm-svn: 297869
It makes the output function layout more predictable; the layout has
an effect on performance, we don't want it to be at the mercy of the
translator's visitation order and such.
The predictable output is also easier to digest.
getOrCreateBB isn't appropriately named anymore, as it never needs to
create anything. Rename it and extract the MBB creation logic out of it.
A couple tests were sensitive to the order. Update them.
llvm-svn: 297868
Using the module ID here is wrong for a couple of reasons:
1) The module ID is not persisted, so we can end up with different
object file contents given the same input file (for example if the same
file is accessed via different paths).
2) With ThinLTO the module ID field may contain the path to a bitcode file,
which is incorrect, as the .file argument is supposed to contain the path to
a source file.
Differential Revision: https://reviews.llvm.org/D30584
llvm-svn: 297853
mfvrd and mffprd are both alias to mfvrsd.
This patch enables correct parsing of the aliases, but we still emit a mfvrsd.
Committing on behalf of brunoalr (Bruno Rosa).
Differential Revision: https://reviews.llvm.org/D29177
llvm-svn: 297849
Reduced from a mixture of PR32273 and David Green's test cases showing SelectionDAG::ComputeNumSignBits not correctly handling BUILD_VECTOR implicit truncation of inputs.
llvm-svn: 297847
This patch adds support for recognizing more patterns to match to DEXT and
CINS instructions.
It finds cases where multiple instructions could be replaced with a single
DEXT or CINS instruction.
For example, for the following:
define i64 @dext_and32(i64 zeroext %a) {
entry:
%and = and i64 %a, 4294967295
ret i64 %and
}
instead of generating:
0000000000000088 <dext_and32>:
88: 64010001 daddiu at,zero,1
8c: 0001083c dsll32 at,at,0x0
90: 6421ffff daddiu at,at,-1
94: 03e00008 jr ra
98: 00811024 and v0,a0,at
9c: 00000000 nop
the following gets generated:
0000000000000068 <dext_and32>:
68: 03e00008 jr ra
6c: 7c82f803 dext v0,a0,0x0,0x20
Cases that are covered:
DEXT:
1. and $src, mask where mask > 0xffff
2. zext $src zero extend from i32 to i64
CINS:
1. and (shl $src, pos), mask
2. shl (and $src, mask), pos
3. zext (shl $src, pos) zero extend from i32 to i64
Patch by Violeta Vukobrat.
Differential Revision: https://reviews.llvm.org/D30464
llvm-svn: 297832
Enable the selection of the 64-bit signed multiply accumulate
instructions which operate on 16-bit operands. These are enabled for
ARMv5TE onwards for ARM and for V6T2 and other DSP enabled Thumb
architectures.
Differential Revision: https://reviews.llvm.org/D30044
llvm-svn: 297809
Summary: D25742 improved the precision of debug locations for PGO by removing debug locations from common tail when tail-merging. However, if identical insturctions that are merged into a common tail have the same debug locations, there's no need to remove them. This patch creates a merged debug location of identical instructions across SameTails and assign it to the instruction in the common tail, so that the debug locations are maintained if they are same across identical instructions.
Reviewers: aprantl, probinson, MatzeB, rob.lougher
Reviewed By: aprantl
Subscribers: andreadb, llvm-commits
Differential Revision: https://reviews.llvm.org/D30226
llvm-svn: 297805
On MachO platforms that use subsections-via-symbols dead code stripping will
drop prefix data. Unfortunately there is no great way to convey the relationship
between a function and its prefix data to the linker. We are forced to use a bit
of a hack: we give the prefix data it’s own symbol, and mark the actual function
entry an .alt_entry.
Patch by Moritz Angermann!
Differential Revision: https://reviews.llvm.org/D30770
llvm-svn: 297804
Summary:
<1 x Ty> is not a legal vector type in LLT, we shouldn’t build G_MERGE_VALUES
instruction for them.
Reviewers: qcolombet, aditya_nandakumar, dsanders, t.p.northover, ab, javed.absar
Reviewed By: qcolombet
Subscribers: dberris, rovka, kristof.beyls, llvm-commits
Differential Revision: https://reviews.llvm.org/D30948
llvm-svn: 297792
Summary:
Adds a new kind of MachineOperand: MO_Placeholder.
This operand must not appear in the MIR and only exists as a way of
creating an 'uninitialized' operand until a matcher function overwrites it.
Depends on D30046, D29712
Reviewers: t.p.northover, ab, rovka, aditya_nandakumar, javed.absar, qcolombet
Reviewed By: qcolombet
Subscribers: dberris, kristof.beyls, llvm-commits
Differential Revision: https://reviews.llvm.org/D30089
llvm-svn: 297782
Reduced version of D26357 - based on the discussion on llvm-dev about canonicalization of UMIN/UMAX/SMIN/SMAX as well as ABS I've reduced that patch to just the ABS ISD node (with x86/sse support) to improve basic combines and lowering.
ARM/AArch64, Hexagon, PowerPC and NVPTX all have similar instructions allowing us to make this a generic opcode and move away from the hard coded tablegen patterns which makes it tricky to match more complex patterns.
At the moment this patch doesn't attempt legalization as we only create an ABS node if its legal/custom.
Differential Revision: https://reviews.llvm.org/D29639
llvm-svn: 297780
Create nodes for smulwb and smulwt and move their selection from
DAGToDAG to DAG combine. smlawb and smlawt can then be selected
using tablegen. Added some helper functions to detect shift patterns
as well as a wrapper around SimplifyDemandBits. Added a couple of
extra tests.
Differential Revision: https://reviews.llvm.org/D30708
llvm-svn: 297716
Each Calling convention (CC) defines a static list of registers that should be preserved by a callee function. All other registers should be saved by the caller.
Some CCs use additional condition: If the register is used for passing/returning arguments – the caller needs to save it - even if it is part of the Callee Saved Registers (CSR) list.
The current LLVM implementation doesn’t support it. It will save a register if it is part of the static CSR list and will not care if the register is passed/returned by the callee.
The solution is to dynamically allocate the CSR lists (Only for these CCs). The lists will be updated with actual registers that should be saved by the callee.
Since we need the allocated lists to live as long as the function exists, the list should reside inside the Machine Register Info (MRI) which is a property of the Machine Function and managed by it (and has the same life span).
The lists should be saved in the MRI and populated upon LowerCall and LowerFormalArguments.
The patch will also assist to implement future no_caller_saved_regsiters attribute intended for interrupt handler CC.
Differential Revision: https://reviews.llvm.org/D28566
llvm-svn: 297715
When checking if chain node is foldable, make sure the intermediate nodes have a single use across all results not just the result that was used to reach the chain node.
This recovers a test case that was severely broken by r296476, my making sure we don't create ADD/ADC that loads and stores when there is also a flag dependency.
llvm-svn: 297698
Recommiting with compiler time improvements
Recommitting after fixup of 32-bit aliasing sign offset bug in DAGCombiner.
* Simplify Consecutive Merge Store Candidate Search
Now that address aliasing is much less conservative, push through
simplified store merging search and chain alias analysis which only
checks for parallel stores through the chain subgraph. This is cleaner
as the separation of non-interfering loads/stores from the
store-merging logic.
When merging stores search up the chain through a single load, and
finds all possible stores by looking down from through a load and a
TokenFactor to all stores visited.
This improves the quality of the output SelectionDAG and the output
Codegen (save perhaps for some ARM cases where we correctly constructs
wider loads, but then promotes them to float operations which appear
but requires more expensive constant generation).
Some minor peephole optimizations to deal with improved SubDAG shapes (listed below)
Additional Minor Changes:
1. Finishes removing unused AliasLoad code
2. Unifies the chain aggregation in the merged stores across code
paths
3. Re-add the Store node to the worklist after calling
SimplifyDemandedBits.
4. Increase GatherAllAliasesMaxDepth from 6 to 18. That number is
arbitrary, but seems sufficient to not cause regressions in
tests.
5. Remove Chain dependencies of Memory operations on CopyfromReg
nodes as these are captured by data dependence
6. Forward loads-store values through tokenfactors containing
{CopyToReg,CopyFromReg} Values.
7. Peephole to convert buildvector of extract_vector_elt to
extract_subvector if possible (see
CodeGen/AArch64/store-merge.ll)
8. Store merging for the ARM target is restricted to 32-bit as
some in some contexts invalid 64-bit operations are being
generated. This can be removed once appropriate checks are
added.
This finishes the change Matt Arsenault started in r246307 and
jyknight's original patch.
Many tests required some changes as memory operations are now
reorderable, improving load-store forwarding. One test in
particular is worth noting:
CodeGen/PowerPC/ppc64-align-long-double.ll - Improved load-store
forwarding converts a load-store pair into a parallel store and
a memory-realized bitcast of the same value. However, because we
lose the sharing of the explicit and implicit store values we
must create another local store. A similar transformation
happens before SelectionDAG as well.
Reviewers: arsenm, hfinkel, tstellarAMD, jyknight, nhaehnle
llvm-svn: 297695
rL230225 made the assumption that only the lower 32-bits of an MMX register load is used as a shift value, when in fact the whole 64-bits are reloaded and treated as a i64 to determine the shift value.
This patch reverts rL230225 to ensure that the whole 64-bits of memory are folded and ensures that the upper 32-bit are zero'd for cases where the shift value has come from a scalar source.
Found during fuzz testing.
Differential Revision: https://reviews.llvm.org/D30833
llvm-svn: 297667
I am leaving the code in clang which filters mxcsr from the clobber list because that is still technically correct and will be useful again when the MXCSR register is reintroduced.
llvm-svn: 297664
This commit adds tail call support to the MachineOutliner pass. This allows
the outliner to insert jumps rather than calls in areas where tail calling is
possible. Outlined tail calls include the return or terminator of the basic
block being outlined from.
Tail call support allows the outliner to take returns and terminators into
consideration while finding candidates to outline. It also allows the outliner
to save more instructions. For example, in the X86-64 outliner, a tail called
outlined function saves one instruction since no return has to be inserted.
llvm-svn: 297653
For AVX-512 we force the input to zero if the input is undef or the mask is all ones to break an execution dependency. This patch brings the same behavior to AVX2.
llvm-svn: 297652
We were already forcing undef inputs to become a zero vector, this now catches an all ones mask too.
Ideally we'd use undef and let execution dep fix handle picking the best register/clearance for the undef, but I don't think it can handle the early clobber today.
llvm-svn: 297651
We used to hit an unreachable in getRegBankFromRegClass when dealing with the
stack pointer. This commit adds support for the GPRsp reg class.
llvm-svn: 297621
This reverts r297596.
There were other issues that were making this not work that have been fixed now. Reverting this results in a more accurate table.
llvm-svn: 297602
This exposed that we have several intrinsic instructions that have identical TSFlags to other instructions. We should merge their patterns and kill of the duplicate. I'll fix that in a follow up patch.
llvm-svn: 297596
The immediate should be 1 or 2, not 0 or 1. This was found while adding bounds checking to clang. In fact the existing clang builtin test failed if we ran it all the way to assembly.
llvm-svn: 297591
I noticed unnecessary 'sbb' instructions in D30472 and while looking at 'ptest' codegen recently.
This happens because we were transforming any 'setb' - even when we only wanted a single-bit result.
This patch moves those transforms under visitAdd/visitSub, so we we're only creating sbb/adc when it
is a win. I don't know why we need a SETCC_CARRY node type, but I'm not proposing to change that
existing behavior in this patch.
Also, I'm skeptical that sbb/adc are a win for all micro-arches, so I added comments to the test files
where this transform still fires.
The test changes here are all cases where we no longer produce sbb/adc. Avoiding partial register
stalls (generating an xor to clear a register) is not handled in some cases, but that's a separate
issue.
Differential Revision: https://reviews.llvm.org/D30611
llvm-svn: 297586
Summary:
A53 scheduler causes an assertion failure on all CRC instructions:
include/llvm/CodeGen/MachineInstr.h:280: const llvm::MachineOperand
&llvm::MachineInstr::getOperand(unsigned int) const: Assertion `i <
getNumOperands() && "getOperand() out of range!"' failed.
The case statements corresponding to CRC instructions are incorrect and should
be removed.
Also adding a testcase while on this.
Reviewers: t.p.northover, javed.absar, apazos, rengolin
Reviewed By: rengolin
Subscribers: evandro, aemerson, llvm-commits, rengolin
Differential Revision: https://reviews.llvm.org/D30274
llvm-svn: 297582
I'm pretty sure there are more problems lurking here. But I think this fixes PR32241.
I've added the test case from that bug and added asserts that will fail if we ever try to copy between high registers and mask registers again.
llvm-svn: 297574
Without SSE41 (pextrb) we currently extract byte elements from a vector by spilling to stack and reloading the byte.
This patch is an initial attempt at using MOVD/PEXTRW to extract the relevant DWORD/WORD from the vector and then shift+truncate to collect the correct byte.
Extraction of multiple bytes this way would result in code bloat, but as explained in the patch we could probably afford to be more aggressive with the supported extractions before again falling back on spilling - possibly through counting the number of extracts and which DWORD/WORD they originate?
Differential Revision: https://reviews.llvm.org/D29841
llvm-svn: 297568
Since v_max_f32_e64/v_max_f16_e64 can be folded if the target
instruction supports the clamp bit, we also need to maintain
modifiers when converting v_mac to v_mad.
This fixes a rendering issue with Dirt Rally because a v_mac
instruction with the clamp bit set was converted to a v_mad
but that bit was lost during the conversion.
Fixes: e184e01dd79 ("AMDGPU: Fold FP clamp as modifier bit")
Patch by Samuel Pitoiset <samuel.pitoiset@gmail.com>
llvm-svn: 297556
This method inverts the Reason field of a scheduling candidate.
It does right comparison between RegCritical and RegExcess, but
everything else is broken. In fact it can prefer less strong reason
such as Weak over RegCritical because Weak > -RegCritical.
The CandReason enum is properly sorted, so just remove artificial
ranking.
Differential Revision: https://reviews.llvm.org/D30557
llvm-svn: 297536
This only requires a 64-bit memory source, not the whole 128-bits. But the 128-bit case is still supported via X86InstrInfo::foldMemoryOperandImpl
llvm-svn: 297523
SelectionDAG::ComputeNumSignBits is poor at build_vector handling, meaning that we can't see that all the vXi64 sources are in fact sign extended i32 or smaller.
llvm-svn: 297486
Summary:
Depends on D30379
This improves the state of things for the sub class of operation.
Reviewers: jyknight, nemanjai, mkuper, spatel, RKSimon, zvi, bkramer
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D30436
llvm-svn: 297482
If we are transferring MMX registers to XMM for conversion we could use the MMX equivalents (CVTPI2PD + CVTPI2PS) without affecting rounding/exceptions etc.
llvm-svn: 297481
Summary: As per title. This is extracted from D29872 and I threw SADDO in.
Reviewers: jyknight, nemanjai, mkuper, spatel, RKSimon, zvi, bkramer
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D30379
llvm-svn: 297479
If we are transferring XMM conversion results to MMX registers we could use the MMX equivalents (CVTPD2PI/CVTTPD2PI + CVTPS2PI/CVTTPS2PI) with affecting rounding/expections etc.
llvm-svn: 297476
This patches teaches the MIPS backend to accept more values for constant
splats. Previously, only 10 bit signed immediates or values that could be
loaded using an ldi.[bhwd] instruction would be acceptted. This patch relaxes
that constraint so that any constant value that be splatted is accepted.
As a result, the constant pool is used less for vector operations, and the
suite of bit manipulation instructions b(clr|set|neg)i can now be used with
the full range of their immediate operand.
Reviewers: slthakur
Differential Revision: https://reviews.llvm.org/D30640
llvm-svn: 297457
same as already done for ARM and Thumb2.
Reviewers: jmolloy, rogfer01, efriedma
Subscribers: aemerson, llvm-commits, rengolin
Differential Revision: https://reviews.llvm.org/D30400
llvm-svn: 297443
Summary: This essentially does the same transform as for ADC.
Reviewers: jyknight, nemanjai, mkuper, spatel, RKSimon, zvi, bkramer
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D30417
llvm-svn: 297416
- Fix the insertion point, which occasionally could have been incorrect.
- Avoid creating multiple bitsplits with the same operands, if an old one
could be reused.
llvm-svn: 297414
The good reason to do this is that static allocas are pretty simple to handle
(especially at -O0) and avoiding tracking DBG_VALUEs throughout the pipeline
should give some kind of performance benefit.
The bad reason is that the debug pipeline is an unholy mess of implicit
contracts, where determining whether "DBG_VALUE %reg, imm" actually implies a
load or not involves the services of at least 3 soothsayers and the sacrifice
of at least one chicken. And it still gets it wrong if the variable is at SP
directly.
llvm-svn: 297410
Summary: This essentially does the same transform as for SUBC.
Reviewers: jyknight, nemanjai, mkuper, spatel, RKSimon, zvi, bkramer
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D30437
llvm-svn: 297404
I previously removed the T2XtPk feature from the ARM backend, but it
looks like I missed some of the tests that were using the feature.
Differential Revision: https://reviews.llvm.org/D30778
llvm-svn: 297386
As discussed in the review thread for rL297026, this is actually 2 changes that
would independently fix all of the test cases in the patch:
1. Return undef in FoldConstantArithmetic for div/rem by 0.
2. Move basic undef simplifications for div/rem (simplifyDivRem()) before
foldBinopIntoSelect() as a matter of efficiency.
I will handle the case of vectors with any zero element as a follow-up. That change
is the DAG sibling for D30665 + adding a check of vector elements to FoldConstantVectorArithmetic().
I'm deleting the test for PR30693 because it does not test for the actual bug any more
(dangers of using bugpoint).
Differential Revision:
https://reviews.llvm.org/D30741
llvm-svn: 297384
The fix introduces segfaults and clobbers the value to be stored when
the atomic sequence loops.
Revert "[Target/MIPS] Kill dead code, no functional change intended."
This reverts commit r296153.
Revert "Recommit "[mips] Fix atomic compare and swap at O0.""
This reverts commit r296134.
llvm-svn: 297380
Fix a machine verifier issue where a instruction was using a invalid
register. The return pseudo is expanded and has the return address
register added to it. The return register may have been spuriously
mark as killed earlier.
This partially resolves PR/27458
Thanks to Quentin Colombet for reporting the issue!
llvm-svn: 297372
When extracting a bitfield from the high register in a register pair,
the final offset should be relative to the high register (for 32-bit
extracts).
llvm-svn: 297288
Summary: By using reg_nodbg_empty() to determine if a function can be
treated as a leaf function or not, we miss the case when the register
pair L0_L1 is used but not L0 by itself. This has the effect that
use_all_i32_regs(), a test in reserved-regs.ll which tries to use all
registers, gets treated as a leaf function.
Reviewers: jyknight, venkatra
Reviewed By: jyknight
Subscribers: davide, RKSimon, sepavloff, llvm-commits
Differential Revision: https://reviews.llvm.org/D27089
llvm-svn: 297285
After inspection, it's an UB in our code base. Someone cast a var-arg
function pointer to a non-var-arg one. :/
Re-commit r296771 to continue testing on the patch.
Sorry for the trouble!
llvm-svn: 297256
If there is only one successor, and that successor only
has one predecessor the wait can obviously be delayed until
uses or the end of the next block. This avoids code quality
regressions when there are trivial fallthrough blocks inserted
for structurization.
llvm-svn: 297251
This helps in cases involving bitfields where an AND is exposed by
legalization.
Differential Revision: https://reviews.llvm.org/D30472
llvm-svn: 297249
We were calculating incorrect extract/insert offsets by trying to be too
tricksy with min/max. It's clearer to just split the logic up into "register
starts before this segment" vs "after".
llvm-svn: 297226
Some intrinsics take metadata parameters. These all need custom
handling of some form, and cannot possibly be lowered generically to
G_INTRINSIC calls with vreg operands.
Reject them, instead of hitting an assert later in getOrCreateVReg.
llvm-svn: 297209
When we translate a no-op (same type) bitcast, we try to be clever and
only emit a COPY if we already assigned a vreg to the defined value.
However, when we didn't, we tried to assign to a reference into the
ValToVReg DenseMap, even though the RHS of the assignment
(getOrCreateVReg) could potentially grow that DenseMap, invalidating the
reference.
Avoid that by getting the source vreg first.
I audited the rest of the translator; this is the only tricky case.
The test is quite unwieldy, as the problem is caused by the DenseMap
growing, which happens after the 47th mapped value.
llvm-svn: 297208
For vector operands, the `select` instruction supports both vector and
non-vector conditions. The MIR builder had an overly restrictive
assertion, that only accepted vector conditions for vector selects
(in effect implementing ISD::VSELECT).
Make it possible to express the full range of G_SELECTs.
llvm-svn: 297207
When computing the mapping for non-generic instructions, we skipped
%noreg operands, because we can't always reason about their banks.
Also skip them when applying the mapping. Otherwise, we could end
up with mappings that we can't apply.
While there, duplicate an assert to distinguish between the two
error conditions.
llvm-svn: 297201
When a dbg_value has a constant operand that isn't representable in MI,
there isn't much we can do. Use %noreg (0) for those situations.
This matches the SelectionDAG behavior.
llvm-svn: 297200
We cannot leave the identity copies 'select true, arg, undef' that this pass
inserts for arguments to simplify handling of values on swifterror arguments.
swifterror arguments have restrictions on their uses.
rdar://30839288
llvm-svn: 297197
Broadcom Vulcan is now Cavium ThunderX2T99.
LLVM Bugzilla: http://bugs.llvm.org/show_bug.cgi?id=32113
Minor fixes for the alignments of loops and functions for
ThunderX T81/T83/T88 (better performance).
Patch was tested with SpecCPU2006.
Patch by Stefan Teleman
Differential Revision: https://reviews.llvm.org/D30510
llvm-svn: 297190
The original patch r296865 was reverted as it broke the chromium builds for
Android https://bugs.llvm.org/show_bug.cgi?id=32134, this patch reapplies
r296865 with a fix to make sure it doesn't cause the build regression.
The problem was that intrinsic selection on int_arm_get_fpscr was failing in
ISel this was because the code to manually select this intrinsic still thought
it was the version with no side-effects (INTRINSIC_WO_CHAIN) which is wrong as
it doesn't semantically match the definition in the tablegen code which says it
does have side-effects, I've fixed this by updating the intrinsic type to
INTRINSIC_W_CHAIN (has side-effects). I've also added a test for this based on
Hans original reproducer.
Differential Revision: https://reviews.llvm.org/D30645
llvm-svn: 297137
Summary: Previously, it had always been materialized as a push/pop sequence.
Reviewers: labrinea, jroelofs
Reviewed By: jroelofs
Subscribers: llvm-commits, rengolin
Differential Revision: https://reviews.llvm.org/D30648
llvm-svn: 297134
evex2vex pass defines 2 tables which maps EVEX instructions to their VEX identical when possible. Adding all missing entries.
Differential Revision: https://reviews.llvm.org/D30501
llvm-svn: 297126
This reverts commit r296771.
We found some wide spread test failures internally. I'm working on a
testcase. Politely revert the patch in the mean time. :)
llvm-svn: 297124
A bit more painful than G_INSERT because it was more widely used, but this
should simplify the handling of extract operations in most locations.
llvm-svn: 297100
Fixed the asan bot failure which led to the last commit of the outliner being reverted.
The change is in lib/CodeGen/MachineOutliner.cpp in the SuffixTree's constructor. LeafVector
is no longer initialized using reserve but just a standard constructor.
llvm-svn: 297081
This patch extends the current functionality of the AArch64 redundant copy
elimination pass to handle CMN instructions as well as a shifted
immediates.
Differential Revision: https://reviews.llvm.org/D30576.
llvm-svn: 297078
If a block has non-analyzable branches, the listed successors don't need
to add up to one. For example, if a block has a conditional tail call,
that tail call will not have a corresponding successor in the successor
list, but will still be a possible branch.
Differential Revision: https://reviews.llvm.org/D30556
llvm-svn: 297054
Before, we were producing G_INSERT instructions that were actually closer to a
cast or even a COPY when both input and output sizes are the same. This doesn't
really make sense and means that everything interpreting a G_INSERT also has to
handle all these kinds of casts.
So now we detect these degenerate cases and emit real casts instead.
llvm-svn: 297051
Use the store size of the argument type, which will be a byte-sized
quantity, rather than dividing the size in bits by 8.
Fixes PR32136 and re-enables copy elision from i64 arguments.
Reverts the workaround in from r296950.
llvm-svn: 297045
Now that G_INSERT instructions can only insert one register, this code was
overly general. In another direction it didn't handle registers that crossed
split boundaries properly, which needed to be fixed.
llvm-svn: 297042
Merge the tail block into the loop in cases where the main loop body
exits early, subject to profitability constraints. This will coalesce
the loop body into fewer blocks.
For example:
loop: loop:
// loop body // loop body
if (...) jump exit --> // more body
more: if (...) jump exit
// more body jump loop
jump loop
llvm-svn: 297033
The code in updateDeadFlags removed unnecessary <dead> flags, but there
can be cases where such a flag is not set, and yet a register has become
dead. For example, if a mux with identical inputs is replaced with a COPY,
the predicate register may no longer be used after that.
llvm-svn: 297032
Refactoring of duplicated code and more fixes to follow.
This is motivated by the post-commit comments for r296699:
http://lists.llvm.org/pipermail/llvm-commits/Week-of-Mon-20170306/435182.html
Ie, we can crash if we're missing obvious simplifications like this that
exist in the IR simplifier or if these occur later than expected.
The x86 change for non-splat division shows a potential opportunity to improve
vector codegen: we assumed that since only one lane had meaningful results, we
should do the math in scalar. But that means moving back and forth from vector
registers.
llvm-svn: 297026
These are not x86-specific, but the problem is not visible for all targets
because it is masked by other transforms. These can lead to compiler crashes.
llvm-svn: 297017
Summary:
Functions with the "xray-log-args" attribute will have a special XRay sled kind
emitted, for compiler-rt to copy any call arguments to your logging handler.
For practical and performance reasons, only the first argument is supported, and
only up to 64 bits.
Reviewers: dberris
Reviewed By: dberris
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D29702
llvm-svn: 296998
As described on PR31712, we miss a variety of legalization combines because we lower these to X86ISD::VSEXT/VZEXT despite them having the same functionality. This patch makes 128-bit (SSE41) SIGN/ZERO_EXTEND_VECTOR_IN_REG ops legal, adds the necessary tablegen plumbing and uses a helper 'getExtendInVec' to decide when to use SIGN/ZERO_EXTEND_VECTOR_IN_REG or VSEXT/VZEXT.
We're missing a couple of shuffle combines that will be added in a future patch for review.
Later patches can then support the AVX2 cases as a mixture of SIGN/ZERO_EXTEND and SIGN/ZERO_EXTEND_VECTOR_IN_REG, and then finally deal with the AVX512 cases.
Differential Revision: https://reviews.llvm.org/D30549
llvm-svn: 296985
The larger goal is to move the ADC/SBB transforms currently in
combineX86SetCC() to combineAddOrSubToADCOrSBB() because we're
creating ADC/SBB in lots of places where we shouldn't.
This was intended to be an NFC change, but avx-512 has something
strange going on. It doesn't seem like any of the affected tests
should really be using SET+TEST or ADC; a simple ADD could replace
several instructions. But that's another bug...
llvm-svn: 296978
select Cond, C +/- 1, C --> add(ext Cond), C -- with a target hook.
This is part of the ongoing process to obsolete D24480. The motivation is to
canonicalize to select IR in InstCombine whenever possible, so we need to have a way to
undo that easily in codegen.
PowerPC is an obvious winner for this kind of transform because it has fast and complete
bit-twiddling abilities but generally lousy conditional execution perf (although this might
have changed in recent implementations).
x86 also sees some wins, but the effect is limited because these transforms already mostly
exist in its target-specific combineSelectOfTwoConstants(). The fact that we see any x86
changes just shows that that code is a mess of special-case holes. We may be able to remove
some of that logic now.
My guess is that other targets will want to enable this hook for most cases. The likely
follow-ups would be to add value type and/or the constants themselves as parameters for the
hook. As the tests in select_const.ll show, we can transform any select-of-constants to
math/logic, but the general transform for any 2 constants needs one more instruction
(multiply or 'and').
ARM is one target that I think may not want this for most cases. I see infinite loops there
because it wants to use selects to enable conditionally executed instructions.
Differential Revision: https://reviews.llvm.org/D30537
llvm-svn: 296977
Long ago (2010 according to svn blame), combineShuffle probably needed to prevent the accidental creation of illegal i64 types but there doesn't appear to be any combines that can cause this any more as they all have their own legality checks.
Differential Revision: https://reviews.llvm.org/D30213
llvm-svn: 296966
Summary:
When replacing a SDValue, we should remove the replaced value from
SoftenedFloats (and possibly the other maps as well?).
When we revisit a Node because it needs analyzing again, we have to
remove all result values from SoftenedFloats (and possibly other maps?).
This fixes the fp128 test failures with expensive checks for X86.
I think we probably should also remove the values from the other maps
(PromotedIntegers and so on), let me know what you think.
Reviewers: baldrick, bogner, davidxl, ab, arsenm, pirama, chh, RKSimon
Reviewed By: chh
Subscribers: danalbert, wdng, srhines, hfinkel, sepavloff, llvm-commits
Differential Revision: https://reviews.llvm.org/D29265
llvm-svn: 296964
This fixes cases where i1 types were not properly legalized yet and lead
to the creating of 0-sized stack slots.
This fixes http://llvm.org/PR32136
llvm-svn: 296950
These are simplified variants of the current G_SEQUENCE and G_EXTRACT, which
assume the individual parts will be contiguous, homogeneous, and occupy the
entirity of the larger register. This makes reasoning about them much easer
since you only have to look at the first register being merged and the result
to know what the instruction is doing.
I intend to gradually replace all uses of the more complicated sequence/extract
with these (or single-element insert/extracts), and then remove the older
variants. For now we start with legalization.
llvm-svn: 296921
The intrinsics __builtin_arm_get_fpscr and __builtin_arm_set_fpscr read and
write to the fpscr (Floating-Point Status and Control Register) register.
A bug exists in the __builtin_arm_get_fpscr intrinsic definition in llvm which
treats this intrinsic as a IntroNoMem which means it's not a memory access and
doesn't have any other side-effects. Having this property on this intrinsic
means that various optimizations can be done on this such as common
sub-expression elimination with other reads. This can cause issues if there has
been write to this register, e.g.
void foo(int *p) {
p[0] = __builtin_arm_get_fpscr();
__builtin_arm_set_fpscr(1);
p[1] = __builtin_arm_get_fpscr();
}
in the above example the second read is currently CSE'd into the first read,
this is because llvm isn't aware that the write done by __builtin_arm_set_fpscr
effects the same register that __builtin_arm_get_fpscr reads from, to fix this
problem I've removed the property IntrNoMem so that __builtin_arm_get_fpscr is
treated as a memory access.
Differential Revision: https://reviews.llvm.org/D30542
llvm-svn: 296865
This patch causes compile times for some patterns to explode. I have
a (large, unreduced) test case that slows down by more than 20x and
several test cases slow down by 2x. I'm sending some of the test cases
directly to Nirav and following up with more details in the review log,
but this should unblock anyone else hitting this.
llvm-svn: 296862
VZEROUPPER should not be issued on Knights Landing (KNL), but on Skylake-avx512 it should be.
Differential Revision: https://reviews.llvm.org/D29874
llvm-svn: 296859
For chains of triangles with small join blocks that can be tail duplicated, a
simple calculation of probabilities is insufficient. Tail duplication
can be profitable in 3 different ways for these cases:
1) The post-dominators marked 50% are actually taken 56% (This shrinks with
longer chains)
2) The chains are statically correlated. Branch probabilities have a very
U-shaped distribution.
[http://nrs.harvard.edu/urn-3:HUL.InstRepos:24015805]
If the branches in a chain are likely to be from the same side of the
distribution as their predecessor, but are independent at runtime, this
transformation is profitable. (Because the cost of being wrong is a small
fixed cost, unlike the standard triangle layout where the cost of being
wrong scales with the # of triangles.)
3) The chains are dynamically correlated. If the probability that a previous
branch was taken positively influences whether the next branch will be
taken
We believe that 2 and 3 are common enough to justify the small margin in 1.
The code pre-scans a function's CFG to identify this pattern and marks the edges
so that the standard layout algorithm can use the computed results.
llvm-svn: 296845
Summary:
Currently, when 't1: i1 = setcc t2, t3, cc' followed by 't4: i1 = xor t1, Constant:i1<-1>' is folded into 't5: i1 = setcc t2, t3 !cc', SDLoc of newly created SDValue 't5' follows SDLoc of 't4', not 't1'. However, as the opcode of newly created SDValue is 'setcc', it make more sense to take DebugLoc from 't1' than 't4'. For the code below
```
extern int bar();
extern int baz();
int foo(int x, int y) {
if (x != y)
return bar();
else
return baz();
}
```
, following is the bitcode representation of 'foo' at the end of llvm-ir level optimization:
```
define i32 @foo(i32 %x, i32 %y) !dbg !4 {
entry:
tail call void @llvm.dbg.value(metadata i32 %x, i64 0, metadata !9, metadata !11), !dbg !12
tail call void @llvm.dbg.value(metadata i32 %y, i64 0, metadata !10, metadata !11), !dbg !13
%cmp = icmp ne i32 %x, %y, !dbg !14
br i1 %cmp, label %if.then, label %if.else, !dbg !16
if.then: ; preds = %entry
%call = tail call i32 (...) @bar() #3, !dbg !17
br label %return, !dbg !18
if.else: ; preds = %entry
%call1 = tail call i32 (...) @baz() #3, !dbg !19
br label %return, !dbg !20
return: ; preds = %if.else, %if.then
%retval.0 = phi i32 [ %call, %if.then ], [ %call1, %if.else ]
ret i32 %retval.0, !dbg !21
}
!14 = !DILocation(line: 5, column: 9, scope: !15)
!16 = !DILocation(line: 5, column: 7, scope: !4)
```
As you can see, in 'entry' block, 'icmp' instruction and 'br' instruction have different debug locations. However, with current implementation, there's no distinction between debug locations of these two when they are lowered to asm instructions. This is because 'icmp' and 'br' become 'setcc' 'xor' and 'brcond' in SelectionDAG, where SDLoc of 'setcc' follows the debug location of 'icmp' but SDLOC of 'xor' and 'brcond' follows the debug location of 'br' instruction, and SDLoc of 'xor' overwrites SDLoc of 'setcc' when they are folded. This patch addresses this issue.
Reviewers: atrick, bogner, andreadb, craig.topper, aprantl
Reviewed By: andreadb
Subscribers: jlebar, mkuper, jholewinski, andreadb, llvm-commits
Differential Revision: https://reviews.llvm.org/D29813
llvm-svn: 296825
This commit also relied on r296812, which I just reverted. We should probably
apply it again, after the r296812 has been discussed and been reapplied in some
variant.
llvm-svn: 296820
In ARMPreAllocLoadStoreOpt::RescheduleOps, LastOp should be the last
operation which we want to merge. If we break out of the loop because
an operation has the wrong offset, we shouldn't use that operation
as LastOp.
This patch fixes some cases where we would move stores to the wrong
insert point.
Re-commit with a fix to increment NumMove in the right place.
Differential Revision: https://reviews.llvm.org/D30124
llvm-svn: 296815
This patch fixes pr32063.
Current code in PPCTargetLowering::PerformDAGCombine can transform
bswap
store
into a single PPCISD::STBRX instruction. but it doesn't consider the case that the operand size of bswap may be larger than store size. When it occurs, we need 2 modifications,
1 For the last operand of PPCISD::STBRX, we should not use DAG.getValueType(N->getOperand(1).getValueType()), instead we should use cast<StoreSDNode>(N)->getMemoryVT().
2 Before PPCISD::STBRX, we need to shift the original operand of bswap to the right side.
Differential Revision: https://reviews.llvm.org/D30362
llvm-svn: 296811
This patch extends the current functionality of the AArch64 redundant copy
elimination pass to handle non-zero cases such as:
BB#0:
cmp x0, #1
b.eq .LBB0_1
.LBB0_1:
orr x0, xzr, #0x1 ; <-- redundant copy; x0 known to hold #1.
Differential Revision: https://reviews.llvm.org/D29344
llvm-svn: 296809
This patch adds support for struct return values to the MSP430
target backend. It also reverses the order of argument and return
registers in the calling convention to bring it into closer
alignment with the published EABI from TI.
Patch by Andrew Wygle (awygle).
Differential Revision: https://reviews.llvm.org/D29069
llvm-svn: 296807
MMX extraction often ends up as extract_i32(bitcast_v2i32(extract_i64(bitcast_v1i64(x86mmx v), 0)), 0) which fails to simplify on 32-bit targets as i64 isn't legal
llvm-svn: 296782
This patch reduces the stack frame size by not allocating the parameter area if
it is not required. In the current implementation LowerFormalArguments_64SVR4
already handles the parameter area, but LowerCall_64SVR4 does not
(when calculating the stack frame size). What this patch does is make
LowerCall_64SVR4 consistent with LowerFormalArguments_64SVR4.
Committing on behalf of Hiroshi Inoue.
Differential Revision: https://reviews.llvm.org/D29881
llvm-svn: 296771
This bug was introduced with:
https://reviews.llvm.org/rL296699
There may be a way to loosen the restriction, but for now just bail out
on any opaque constant.
The tests show that opacity is target-specific. This goes back to cost
calculations in ConstantHoisting based on TTI->getIntImmCost().
llvm-svn: 296768
The CallingConv.td rules allocate 8 bytes for these kinds of arguments
on AAPCS targets, but we were only recording the smaller amount. The
difference is theoretical on AArch64 because we don't actually store
more than the smaller amount, but it's still much better to have these
two components in agreement.
Based on Diana Picus's ARM equivalent patch (where it matters a lot
more).
llvm-svn: 296754
If dominator tree is not calculated or is invalidated, set corresponding
pointer in the pass state to nullptr. Such pointer value will indicate
that operations with dominator tree are not allowed. In particular, it
allows to skip verification for such pass state. The dominator tree is
not calculated if the machine dominator pass was skipped, it occures in
the case of entities with linkage available_externally.
The change fixes some test fails observed when expensive checks
are enabled.
Differential Revision: https://reviews.llvm.org/D29280
llvm-svn: 296742
Surprisingly, one of the three interference checks in LiveRegMatrix was
using the main live range instead of the apropriate subregister range
resulting in unnecessarily conservative results.
llvm-svn: 296722
Original commit message:
[ARM] Fix insert point for store rescheduling.
In ARMPreAllocLoadStoreOpt::RescheduleOps, LastOp should be the last
operation which we want to merge. If we break out of the loop because
an operation has the wrong offset, we shouldn't use that operation as
LastOp.
This patch fixes some cases where we would sink stores for no reason.
llvm-svn: 296718
Summary:
This can be used to optimize large multiplications after legalization.
Depends on D29565
Reviewers: mkuper, spatel, RKSimon, zvi, bkramer, aaboud, craig.topper
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D29587
llvm-svn: 296711
Until now, we've had to use -global-isel to enable GISel. But using
that on other targets that don't support it will result in an abort, as we
can't build a full pipeline.
Additionally, we want to experiment with enabling GISel by default for
some targets: we can't just enable GISel by default, even among those
target that do have some support, because the level of support varies.
This first step adds an override for the target to explicitly define its
level of support. For AArch64, do that using
a new command-line option (I know..):
-aarch64-enable-global-isel-at-O=<N>
Where N is the opt-level below which GISel should be used.
Default that to -1, so that we still don't enable GISel anywhere.
We're not there yet!
While there, remove a couple LLVM_UNLIKELYs. Building the pipeline is
such a cold path that in practice that shouldn't matter at all.
llvm-svn: 296710
In ARMPreAllocLoadStoreOpt::RescheduleOps, LastOp should be the last
operation which we want to merge. If we break out of the loop because
an operation has the wrong offset, we shouldn't use that operation as
LastOp.
This patch fixes some cases where we would sink stores for no reason.
Differential Revision: https://reviews.llvm.org/D30124
llvm-svn: 296708
This code starts from the high end of the sorted vector of offsets, and
works backwards: it tries to find contiguous offsets, process them, then
pops them from the end of the vector. Most of the code agrees with this
order of processing, but one loop doesn't: it instead processes elements
from the low end of the vector (which are nodes with unrelated offsets).
Fix that loop to process the correct elements.
This has a few implications. One, we don't incorrectly return early when
processing multiple groups of offsets in the same block (which allows
rescheduling prera-ldst-insertpt.mir). Two, we pick the correct insert
point for loads, so they're correctly sorted (which affects the
scheduling of vldm-liveness.ll). I think it might also impact some of
the heuristics slightly.
Differential Revision: https://reviews.llvm.org/D30368
llvm-svn: 296701
This is part of the ongoing attempt to improve select codegen for all targets and select
canonicalization in IR (see D24480 for more background). The transform is a subset of what
is done in InstCombine's FoldOpIntoSelect().
I first noticed a regression in the x86 avx512-insert-extract.ll tests with a patch that
hopes to convert more selects to basic math ops. This appears to be a general missing DAG
transform though, so I added tests for all standard binops in rL296621
(PowerPC was chosen semi-randomly; it has scripted FileCheck support, but so do ARM and x86).
The poor output for "sel_constants_shl_constant" is tracked with:
https://bugs.llvm.org/show_bug.cgi?id=32105
Differential Revision: https://reviews.llvm.org/D30502
llvm-svn: 296699
Summary:
Avoids tons of prologue boilerplate when arguments are passed in memory
and left in memory. This can happen in a debug build or in a release
build when an argument alloca is escaped. This will dramatically affect
the code size of x86 debug builds, because X86 fast isel doesn't handle
arguments passed in memory at all. It only handles the x86_64 case of up
to 6 basic register parameters.
This is implemented by analyzing the entry block before ISel to identify
copy elision candidates. A copy elision candidate is an argument that is
used to fully initialize an alloca before any other possibly escaping
uses of that alloca. If an argument is a copy elision candidate, we set
a flag on the InputArg. If the the target generates loads from a fixed
stack object that matches the size and alignment requirements of the
alloca, the SelectionDAG builder will delete the stack object created
for the alloca and replace it with the fixed stack object. The load is
left behind to satisfy any remaining uses of the argument value. The
store is now dead and is therefore elided. The fixed stack object is
also marked as mutable, as it may now be modified by the user, and it
would be invalid to rematerialize the initial load from it.
Supersedes D28388
Fixes PR26328
Reviewers: chandlerc, MatzeB, qcolombet, inglorion, hans
Subscribers: igorb, llvm-commits
Differential Revision: https://reviews.llvm.org/D29668
llvm-svn: 296683
This patch adds a MachineSSA pass that coalesces blocks that branch
on the same condition.
Committing on behalf of Lei Huang.
Differential Revision: https://reviews.llvm.org/D28249
llvm-svn: 296670
Add check that deleted nodes do not get added to worklist. This can
occur when a node's operand is simplified to an existing node.
This fixes PR32108.
Reviewers: jyknight, hfinkel, chandlerc
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D30506
llvm-svn: 296668
Resubmit r295336 after the bug with non-zero offset patterns on BE targets is fixed (r296336).
Support {a|s}ext, {a|z|s}ext load nodes as a part of load combine patters.
Reviewed By: filcab
Differential Revision: https://reviews.llvm.org/D29591
llvm-svn: 296651
On Hexagon, values of type i1 are passed in registers of type i32,
even though i1 is not a legal value for these registers. This is a
special case and needs special handling to maintain consistency of
the lowering information.
This fixes PR32089.
llvm-svn: 296645
Lower i1, i8 and i16 call parameters by extending them before storing them on
the stack. Also make sure we encode the correct, extended size in the
corresponding memory operand, and that we compute the correct stack size in the
end.
The latter is a bit more complicated because we used to compute the stack size
in the getStackAddress method, based on the Size and Offset of the parameters.
However, if the last parameter is sign extended, we'd be using the wrong,
non-extended size, and we'd end up with a smaller stack than we need to hold the
extended value. Instead of hacking this up based on the value of Size in
getStackAddress, we move our stack size handling logic to assignArg, where we
have access to the CCState which knows everything we could possibly want to know
about the stack. This way we don't need to duplicate any knowledge or resort to
any ugly hacks.
On this same occasion, update the IRTranslator test to check the sizes of the
stores everywhere, not just for sign extended paramteres.
llvm-svn: 296631
Modify the test so that it is still testing something
closer to what it was intended to originally.
I think the original intent was to test the situation where
there was a branch on execz and then unconditional branch
required relaxing.With the change in r296539,
there was no longer and execz branch.
Change the test so that there is now an execz branch inserted.
There is no longer an unconditional branch after the execz branch,
so this might need to be tricked in some other way to keep that
there.
llvm-svn: 296574
When SDAGISel (top-down) selects a tail-call, it skips the remainder
of the block.
If, before that, FastISel (bottom-up) selected some of the (no-op) next
few instructions, we can end up with dead instructions following the
terminator (selected by SDAGISel).
We need to erase them, as we know they aren't necessary (in addition to
being incorrect).
We already do this when FastISel falls back on the tail-call itself.
Also remove the FastISel-emitted code if we fallback on the
instructions between the tail-call and the return.
llvm-svn: 296552
Iterating on the use-list we're modifying doesn't work: after the first
iteration, the use-list iterator will point to a MachineOperand
referencing the new register. This caused us to skip the other uses to
replace.
Instead, use MRI.replaceRegWith(), which accounts for this behavior.
llvm-svn: 296551
To facilitate this, add a new hidden command-line option to disable
the explicit-locals pass. That causes llc to emit invalid code that doesn't
have all locals converted to get_local/set_local, however it simplifies
testwriting in many cases.
llvm-svn: 296540
This prevents generating stm r1!, {r0, r1} on Thumb1, where value
stored for r1 is UNKONWN.
Patch by Zhaoshi Zheng.
Differential Revision: https://reviews.llvm.org/D27910
llvm-svn: 296538
Requesting DWARF v5 will now get you the new compile-unit and
type-unit headers. llvm-dwarfdump will also recognize them.
Differential Revision: http://reviews.llvm.org/D30206
llvm-svn: 296514
This recovers a test case that was severely broken by r296476, my making sure we don't create ADD/ADC that loads and stores when there is also a flag dependency.
llvm-svn: 296486
Stack Smash Protection is not completely free, so in hot code, the overhead it causes can cause performance issues. By adding diagnostic information for which functions have SSP and why, a user can quickly determine what they can do to stop SSP being applied to a specific hot function.
This change adds a remark that is reported by the stack protection code when an instruction or attribute is encountered that causes SSP to be applied.
Patch by: James Henderson
Differential Revision: https://reviews.llvm.org/D29023
llvm-svn: 296483
Recommiting after fixup of 32-bit aliasing sign offset bug in DAGCombiner.
* Simplify Consecutive Merge Store Candidate Search
Now that address aliasing is much less conservative, push through
simplified store merging search and chain alias analysis which only
checks for parallel stores through the chain subgraph. This is cleaner
as the separation of non-interfering loads/stores from the
store-merging logic.
When merging stores search up the chain through a single load, and
finds all possible stores by looking down from through a load and a
TokenFactor to all stores visited.
This improves the quality of the output SelectionDAG and the output
Codegen (save perhaps for some ARM cases where we correctly constructs
wider loads, but then promotes them to float operations which appear
but requires more expensive constant generation).
Some minor peephole optimizations to deal with improved SubDAG shapes (listed below)
Additional Minor Changes:
1. Finishes removing unused AliasLoad code
2. Unifies the chain aggregation in the merged stores across code
paths
3. Re-add the Store node to the worklist after calling
SimplifyDemandedBits.
4. Increase GatherAllAliasesMaxDepth from 6 to 18. That number is
arbitrary, but seems sufficient to not cause regressions in
tests.
5. Remove Chain dependencies of Memory operations on CopyfromReg
nodes as these are captured by data dependence
6. Forward loads-store values through tokenfactors containing
{CopyToReg,CopyFromReg} Values.
7. Peephole to convert buildvector of extract_vector_elt to
extract_subvector if possible (see
CodeGen/AArch64/store-merge.ll)
8. Store merging for the ARM target is restricted to 32-bit as
some in some contexts invalid 64-bit operations are being
generated. This can be removed once appropriate checks are
added.
This finishes the change Matt Arsenault started in r246307 and
jyknight's original patch.
Many tests required some changes as memory operations are now
reorderable, improving load-store forwarding. One test in
particular is worth noting:
CodeGen/PowerPC/ppc64-align-long-double.ll - Improved load-store
forwarding converts a load-store pair into a parallel store and
a memory-realized bitcast of the same value. However, because we
lose the sharing of the explicit and implicit store values we
must create another local store. A similar transformation
happens before SelectionDAG as well.
Reviewers: arsenm, hfinkel, tstellarAMD, jyknight, nhaehnle
llvm-svn: 296476
Lower i32, float and double parameters that need to live on the stack. This
boils down to creating some G_GEPs starting from the stack pointer and storing
the values there. During the process we also keep track of the stack size and
use the final value in the ADJCALLSTACKDOWN/UP instructions.
We currently assert for smaller types, since they usually require extensions.
They will be handled in a separate patch.
llvm-svn: 296473
Summary:
With this change ImplicitNullCheck optimization uses alias analysis
and can use load/store memory access for implicit null check if there
are other load/store before but memory accesses do not alias.
Patch by Serguei Katkov!
Reviewers: sanjoy
Reviewed By: sanjoy
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D30331
llvm-svn: 296440
This is a patch for the outliner described in the RFC at:
http://lists.llvm.org/pipermail/llvm-dev/2016-August/104170.html
The outliner is a code-size reduction pass which works by finding
repeated sequences of instructions in a program, and replacing them with
calls to functions. This is useful to people working in low-memory
environments, where sacrificing performance for space is acceptable.
This adds an interprocedural outliner directly before printing assembly.
For reference on how this would work, this patch also includes X86
target hooks and an X86 test.
The outliner is run like so:
clang -mno-red-zone -mllvm -enable-machine-outliner file.c
Patch by Jessica Paquette<jpaquette@apple.com>!
rdar://29166825
Differential Revision: https://reviews.llvm.org/D26872
llvm-svn: 296418
Splitting critical edges when one of the source edges is an indirectbr
is hard in general (because it requires changing the memory the indirectbr
reads). But if a block only has a single indirectbr predecessor (which is
the common case), we can simulate splitting that edge by splitting
the destination block, and retargeting the *direct* branches.
This is motivated by the use of computed gotos in python 2.7: PyEval_EvalFrame()
ends up using an indirect branch with ~100 successors, and passing a constant to
each of those. Since MachineSink can't break indirect critical edges on demand
(and doing this in MIR doesn't look feasible), this causes us to emit about ~100
defs of registers containing constants, which we in the predecessor block, where
only one of those constants is used in each successor. So, at each computed goto,
we needlessly spill about a 100 constants to stack. The end result is that a
clang-compiled python interpreter can be about ~2.5x slower on a simple python
reduction loop than a gcc-compiled interpreter.
Differential Revision: https://reviews.llvm.org/D29916
llvm-svn: 296416
The transform in question claims to be doing:
// fold (add (select cc, 0, c), x) -> (select cc, x, (add, x, c))
...starting in PerformADDCombineWithOperands(), but it wasn't actually checking for a setcc node
for the sext/zext patterns.
This is exactly the opposite of a transform I'd like to add to DAGCombiner's foldSelectOfConstants(),
so I was seeing infinite loops with my draft of a patch applied.
The changes in select_const.ll look positive (less instructions). The change in arm-and-tst-peephole.ll
is unrelated. We're changing the input IR in that test to preserve the intent of the test, but that's
not affected by this code change.
Differential Revision:
https://reviews.llvm.org/D30355
llvm-svn: 296389
DAGCombiner already supports peeking thorough shuffles to improve vector element extraction, but legalization often leaves us in situations where we need to extract vector elements after shuffles have already been lowered.
This patch adds support for VECTOR_EXTRACT_ELEMENT/PEXTRW/PEXTRB instructions to attempt to handle target shuffles as well. I've covered some basic scenarios including handling shuffle mask scaling and the implicit zero-extension of PEXTRW/PEXTRB, there is more that could be done here (that I've mentioned in TODOs) but I haven't found many cases where its worth it.
Differential Revision: https://reviews.llvm.org/D30176
llvm-svn: 296381
Summary: Existing implementation of duplicateSimpleBB function drops DebugLoc metadata of branch instructions during the transformation. This patch addresses this issue by making newly created branch instructions to keep the metadata of replaced branch instructions.
Reviewers: qcolombet, craig.topper, aprantl, MatzeB, sanjoy, dblaikie
Reviewed By: dblaikie
Subscribers: dblaikie, llvm-commits
Differential Revision: https://reviews.llvm.org/D30026
llvm-svn: 296371
This pattern is essentially a i16 load from p+1 address:
%p1.i16 = bitcast i8* %p to i16*
%p2.i8 = getelementptr i8, i8* %p, i64 2
%v1 = load i16, i16* %p1.i16
%v2.i8 = load i8, i8* %p2.i8
%v2 = zext i8 %v2.i8 to i16
%v1.shl = shl i16 %v1, 8
%res = or i16 %v1.shl, %v2
Current implementation would identify %v1 load as the first byte load and would mistakenly emit a i16 load from %p1.i16 address. This patch adds a check that the first byte is loaded from a non-zero offset of the first load address. This way this address can be used as the base address for the combined value. Otherwise just give up combining.
llvm-svn: 296336
Summary:
While collecting operands we make copies of the LiveReg objects which are stored in the LiveRegs array. If the instruction uses the same register multiple times we end up with multiple copies. Later we iterate through the collected list of LiveReg objects and merge DomainValues. In the process of doing this the merge function can change the contents of the original LiveReg object in the LiveRegs array, but not the copies that have been made. So when we get to the second usage of the register we end up seeing a stale copy of the LiveReg object.
To fix this I've stopped copying and now just store a pointer to the original LiveReg object. Another option might be to avoid adding the same register to the Regs array twice, but this approach seemed simpler.
The included test case exposes this bug due to an AVX-512 masked OR instruction using the same register for the passthru operand and one of the inputs to the OR operation.
Fixes PR30284.
Reviewers: RKSimon, stoklund, MatzeB, spatel, myatsina
Reviewed By: RKSimon
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D30242
llvm-svn: 296260
Recommiting after fixup of 32-bit aliasing sign offset bug in DAGCombiner.
* Simplify Consecutive Merge Store Candidate Search
Now that address aliasing is much less conservative, push through
simplified store merging search and chain alias analysis which only
checks for parallel stores through the chain subgraph. This is cleaner
as the separation of non-interfering loads/stores from the
store-merging logic.
When merging stores search up the chain through a single load, and
finds all possible stores by looking down from through a load and a
TokenFactor to all stores visited.
This improves the quality of the output SelectionDAG and the output
Codegen (save perhaps for some ARM cases where we correctly constructs
wider loads, but then promotes them to float operations which appear
but requires more expensive constant generation).
Some minor peephole optimizations to deal with improved SubDAG shapes (listed below)
Additional Minor Changes:
1. Finishes removing unused AliasLoad code
2. Unifies the chain aggregation in the merged stores across code
paths
3. Re-add the Store node to the worklist after calling
SimplifyDemandedBits.
4. Increase GatherAllAliasesMaxDepth from 6 to 18. That number is
arbitrary, but seems sufficient to not cause regressions in
tests.
5. Remove Chain dependencies of Memory operations on CopyfromReg
nodes as these are captured by data dependence
6. Forward loads-store values through tokenfactors containing
{CopyToReg,CopyFromReg} Values.
7. Peephole to convert buildvector of extract_vector_elt to
extract_subvector if possible (see
CodeGen/AArch64/store-merge.ll)
8. Store merging for the ARM target is restricted to 32-bit as
some in some contexts invalid 64-bit operations are being
generated. This can be removed once appropriate checks are
added.
This finishes the change Matt Arsenault started in r246307 and
jyknight's original patch.
Many tests required some changes as memory operations are now
reorderable, improving load-store forwarding. One test in
particular is worth noting:
CodeGen/PowerPC/ppc64-align-long-double.ll - Improved load-store
forwarding converts a load-store pair into a parallel store and
a memory-realized bitcast of the same value. However, because we
lose the sharing of the explicit and implicit store values we
must create another local store. A similar transformation
happens before SelectionDAG as well.
Reviewers: arsenm, hfinkel, tstellarAMD, jyknight, nhaehnle
llvm-svn: 296252
With the "wasm32-unknown-unknown-wasm" triple, this allows writing out
simple wasm object files, and is another step in a larger series toward
migrating from ELF to general wasm object support. Note that this code
and the binary format itself is still experimental.
llvm-svn: 296190
This reverts commit r296009. It broke one out of tree target and also
does not account for all partial lines added or removed when calculating
PressureDiff.
llvm-svn: 296182
All G_CONSTANTS created by the MachineIRBuilder have an operand of type CImm
(i.e. a ConstantInt), so that's what the selector needs to look for.
llvm-svn: 296176
Splitting critical edges when one of the source edges is an indirectbr
is hard in general (because it requires changing the memory the indirectbr
reads). But if a block only has a single indirectbr predecessor (which is
the common case), we can simulate splitting that edge by splitting
the destination block, and retargeting the *direct* branches.
This is motivated by the use of computed gotos in python 2.7: PyEval_EvalFrame()
ends up using an indirect branch with ~100 successors, and passing a constant to
each of those. Since MachineSink can't break indirect critical edges on demand
(and doing this in MIR doesn't look feasible), this causes us to emit about ~100
defs of registers containing constants, which we in the predecessor block, where
only one of those constants is used in each successor. So, at each computed goto,
we needlessly spill about a 100 constants to stack. The end result is that a
clang-compiled python interpreter can be about ~2.5x slower on a simple python
reduction loop than a gcc-compiled interpreter.
Differential Revision: https://reviews.llvm.org/D29916
llvm-svn: 296149
Provide a 64-bit pattern to use SUBFIC for subtracting from a 16-bit immediate.
The corresponding pattern already exists for 32-bit integers.
Committing on behalf of Hiroshi Inoue.
Differential Revision: https://reviews.llvm.org/D29387
llvm-svn: 296144
Emit clrrdi (extended mnemonic for rldicr) for AND-ing with masks that
clear bits from the right hand size.
Committing on behalf of Hiroshi Inoue.
Differential Revision: https://reviews.llvm.org/D29388
llvm-svn: 296143
The motivation for filling out these select-of-constants cases goes back to D24480,
where we discussed removing an IR fold from add(zext) --> select. And that goes back to:
https://reviews.llvm.org/rL75531https://reviews.llvm.org/rL159230
The idea is that we should always canonicalize patterns like this to a select-of-constants
in IR because that's the smallest IR and the best for value tracking. Note that we currently
do the opposite in some cases (like the cases in *this* patch). Ie, the proposed folds in
this patch already exist in InstCombine today:
https://github.com/llvm-mirror/llvm/blob/master/lib/Transforms/InstCombine/InstCombineSelect.cpp#L1151
As this patch shows, most targets generate better machine code for simple ext/add/not ops
rather than a select of constants. So the follow-up steps to make this less of a patchwork
of special-case folds and missing IR canonicalization:
1. Have DAGCombiner convert any select of constants into ext/add/not ops.
2 Have InstCombine canonicalize in the other direction (create more selects).
Differential Revision: https://reviews.llvm.org/D30180
llvm-svn: 296137
This time with the missing files.
Similar to PR/25526, fast-regalloc introduces spills at the end of basic
blocks. When this occurs in between an ll and sc, the store can cause the
atomic sequence to fail.
This patch fixes the issue by introducing more pseudos to represent atomic
operations and moving their lowering to after the expansion of postRA
pseudos.
This resolves PR/32020.
Thanks to James Cowgill for reporting the issue!
Reviewers: slthakur
Differential Revision: https://reviews.llvm.org/D30257
llvm-svn: 296134
Similar to PR/25526, fast-regalloc introduces spills at the end of basic
blocks. When this occurs in between an ll and sc, the store can cause the
atomic sequence to fail.
This patch fixes the issue by introducing more pseudos to represent atomic
operations and moving their lowering to after the expansion of postRA
pseudos.
This resolves PR/32020.
Thanks to James Cowgill for reporting the issue!
Reviewers: slthakur
Differential Revision: https://reviews.llvm.org/D30257
llvm-svn: 296132
Summary:
This isn't testable for AArch64 by itself so this patch also adds
support for constant immediates in the pattern and physical
register uses in the result.
The new IntOperandMatcher matches the constant in patterns such as
'(set $rd:GPR32, (G_XOR $rs:GPR32, -1))'. It's always safe to fold
immediates into an instruction so this is the first rule that will match
across multiple BB's.
The Renderer hierarchy is responsible for adding operands to the result
instruction. Renderers can copy operands (CopyRenderer) or add physical
registers (in particular %wzr and %xzr) to the result instruction
in any order (OperandMatchers now import the operand names from
SelectionDAG to allow renderers to access any operand). This allows us to
emit the result instruction for:
%1 = G_XOR %0, -1 --> %1 = ORNWrr %wzr, %0
%1 = G_XOR -1, %0 --> %1 = ORNWrr %wzr, %0
although the latter is untested since the matcher/importer has not been
taught about commutativity yet.
Added BuildMIAction which can build new instructions and mutate them where
possible. W.r.t the mutation aspect, MatchActions are now told the name of
an instruction they can recycle and BuildMIAction will emit mutation code
when the renderers are appropriate. They are appropriate when all operands
are rendered using CopyRenderer and the indices are the same as the matcher.
This currently assumes that all operands have at least one matcher.
Finally, this change also fixes a crash in
AArch64InstructionSelector::select() caused by an immediate operand
passing isImm() rather than isCImm(). This was uncovered by the other
changes and was detected by existing tests.
Depends on D29711
Reviewers: t.p.northover, ab, qcolombet, rovka, aditya_nandakumar, javed.absar
Reviewed By: rovka
Subscribers: aemerson, dberris, kristof.beyls, llvm-commits
Differential Revision: https://reviews.llvm.org/D29712
llvm-svn: 296131
The Fuchsia ABI defines slots from the thread pointer where the
stack-guard value for stack-protector, and the unsafe stack pointer
for safe-stack, are stored. This parallels the Android ABI support.
Patch by Roland McGrath
Differential Revision: https://reviews.llvm.org/D30237
llvm-svn: 296081
clang will generate IR like this for input using packed bitfields;
very simple semantically, but it's a bit tricky to actually
generate good code.
llvm-svn: 296080
Splitting critical edges when one of the source edges is an indirectbr
is hard in general (because it requires changing the memory the indirectbr
reads). But if a block only has a single indirectbr predecessor (which is
the common case), we can simulate splitting that edge by splitting
the destination block, and retargeting the *direct* branches.
This is motivated by the use of computed gotos in python 2.7: PyEval_EvalFrame()
ends up using an indirect branch with ~100 successors, and passing a constant to
each of those. Since MachineSink can't break indirect critical edges on demand
(and doing this in MIR doesn't look feasible), this causes us to emit about ~100
defs of registers containing constants, which we in the predecessor block, where
only one of those constants is used in each successor. So, at each computed goto,
we needlessly spill about a 100 constants to stack. The end result is that a
clang-compiled python interpreter can be about ~2.5x slower on a simple python
reduction loop than a gcc-compiled interpreter.
Differential Revision: https://reviews.llvm.org/D29916
llvm-svn: 296060
This patch enables support for .f16x2 operations.
Added new register type Float16x2.
Added support for .f16x2 instructions.
Added handling of vectorized loads/stores of v2f16 values.
Differential Revision: https://reviews.llvm.org/D30057
Differential Revision: https://reviews.llvm.org/D30310
llvm-svn: 296032
FastISel wasn't checking the isFPOnlySP subtarget feature before emitting
double-precision operations, so it got completely invalid CodeGen for doubles
on Cortex-M4F.
The normal ISel testing wasn't spectacular either so I added a second RUN line
to improve that while I was in the area.
llvm-svn: 296031
Having more fine-grained information on the specific construct that
caused us to fallback is valuable for large-scale data collection.
We still have the fallback warning, that's also used for FastISel.
We still need to remove the fallback warning, and teach FastISel to also
emit remarks (it currently has a combination of the warning, stats, and
debug prints: the remarks could unify all three).
The abort-on-fallback path could also be better handled using remarks:
one could imagine a "-Rpass-error", analoguous to "-Werror", which would
promote missed/failed remarks to errors. It's not clear whether that
would be useful for other remarks though, so we're not there yet.
llvm-svn: 296013
If a subreg is used in an instruction it counts as a whole superreg
for the purpose of register pressure calculation. This patch corrects
improper register pressure calculation by examining operand's lane mask.
Differential Revision: https://reviews.llvm.org/D29835
llvm-svn: 296009
Introduce a common ValueHandler for call returns and formal arguments, and
inherit two different versions for handling the differences (at the moment the
only difference is the way physical registers are marked as used).
llvm-svn: 295973
Add support for lowering calls with parameters than can fit into regs. Use the
same ValueHandler that we used for function returns, but rename it to match its
new, extended purpose.
llvm-svn: 295971
The ARMConstantIslandPass didn't have support for handling accesses to
constant island objects through ARM::t2LDRBpci instructions. This adds
support for that.
This fixes PR31997.
llvm-svn: 295964
AVX versions of the converts work on f32/f64 types, while AVX512 version work on vectors.
Differential Revision: https://reviews.llvm.org/D29988
llvm-svn: 295940
The manual is unclear on the details of this. It's not
clear to me if denormals are not allowed with clamp,
or if that is only omod. Not allowing denorms for
fp16 or fp64 isn't useful so I also question if that
is really a restriction. Same with whether this is valid
without IEEE mode enabled.
llvm-svn: 295905
This should avoid reporting any stack needs to be allocated in the
case where no stack is truly used. An unused stack slot is still
left around in other cases where there are real stack objects
but no spilling occurs.
llvm-svn: 295891
This allows us to ensure that 0 is never a valid pointer
to a user object, and ensures that the offset is always legal
without needing a register to access it. This comes at the cost
of usable offsets and wasted stack space.
llvm-svn: 295877
Summary:
Extend AArch64RedundantCopyElimination to catch cases where the register
that is known to be zero is COPY'd in the predecessor block. Before
this change, this pass would catch cases like:
CBZW %W0, <BB#1>
BB#1:
%W0 = COPY %WZR // removed
After this change, cases like the one below are also caught:
%W0 = COPY %W1
CBZW %W1, <BB#1>
BB#1:
%W0 = COPY %WZR // removed
This change results in a 4% increase in static copies removed by this
pass when compiling the llvm test-suite. It also fixes regressions
caused by doing post-RA copy propagation (a separate change to be put up
for review shortly).
Reviewers: junbuml, mcrosier, t.p.northover, qcolombet, MatzeB
Subscribers: aemerson, rengolin, llvm-commits
Differential Revision: https://reviews.llvm.org/D30113
llvm-svn: 295863
The pass tries to fix a spill of LR that turns out to be unnecessary.
So it removes the tPOP but forgets to remove tPUSH.
This causes the stack be misaligned upon returning the function.
Thus, remove the tPUSH as well in this case.
Differential Revision: https://reviews.llvm.org/D30207
llvm-svn: 295816
This patch adds missing sched classes for Thumb2 instructions.
This has been missing so far, and as a consequence, machine
scheduler models for individual sub-targets have tended to
be larger than they needed to be. These patches should help
write schedulers better and faster in the future
for ARM sub-targets.
Reviewer: Diana Picus
Differential Revision: https://reviews.llvm.org/D29953
llvm-svn: 295811
This patch introduces new X86ISD::FMAXS and X86ISD::FMINS opcodes. The legacy intrinsics now lower to this node. As do the AVX-512 masked intrinsics when the rounding mode is CUR_DIRECTION.
I've merged a copy of the tablegen multiclass avx512_fp_scalar into avx512_fp_scalar_sae. avx512_fp_scalar still needs to support CUR_DIRECTION appearing as a rounding mode for X86ISD::FADD_ROUND and others.
Differential revision: https://reviews.llvm.org/D30186
llvm-svn: 295810
Change implementation to use max instead of add.
min/max/med3 do not flush denormals regardless of the mode,
so it is OK to use it whether or not they are enabled.
Also allow using clamp with f16, and use knowledge
of dx10_clamp.
llvm-svn: 295788
Original code only used vector loads/stores for explicit vector arguments.
It could also do more loads/stores than necessary (e.g v5f32 would
touch 8 f32 values). Aggregate types were loaded one element at a time,
even the vectors contained within.
This change attempts to generalize (and simplify) parameter space
loads/stores so that vector loads/stores can be used more broadly.
Functionality of the patch has been verified by compiling thrust
test suite and manually checking the differences between PTX
generated by llvm with and without the patch.
General algorithm:
* ComputePTXValueVTs() flattens input/output argument into a flat list
of scalars to load/store and returns their types and offsets.
* VectorizePTXValueVTs() uses that data to create vectorization plan
which returns an array of flags marking boundaries of vectorized
load/stores. Scalars are represented as 1-element vectors.
* Code that generates loads/stores implements a simple state machine
that constructs a vector according to the plan.
Differential Revision: https://reviews.llvm.org/D30011
llvm-svn: 295784
Before frame offsets are calculated, try to eliminate the
frame indexes used by SGPR spills. Then we can delete them
after.
I think for now we can be sure that no other instruction
will be re-using the same frame indexes. It should be easy
to notice if this assumption ever breaks since everything
asserts if it tries to use a dead frame index later.
The unused emergency stack slot seems to still be left behind,
so an additional 4 bytes is still wasted.
llvm-svn: 295753
Summary:
Rework the code that was sinking/duplicating (icmp and, 0) sequences
into blocks where they were being used by conditional branches to form
more tbz instructions on AArch64. The new code is more general in that
it just looks for 'and's that have all icmp 0's as users, with a target
hook used to select which subset of 'and' instructions to consider.
This change also enables 'and' sinking for X86, where it is more widely
beneficial than on AArch64.
The 'and' sinking/duplicating code is moved into the optimizeInst phase
of CodeGenPrepare, where it can take advantage of the fact the
OptimizeCmpExpression has already sunk/duplicated any icmps into the
blocks where they are used. One minor complication from this change is
that optimizeLoadExt needed to be updated to always mark 'and's it has
determined should be in the same block as their feeding load in the
InsertedInsts set to avoid an infinite loop of hoisting and sinking the
same 'and'.
This change fixes a regression on X86 in the tsan runtime caused by
moving GVNHoist to a later place in the optimization pipeline (see
PR31382).
Reviewers: t.p.northover, qcolombet, MatzeB
Subscribers: aemerson, mcrosier, sebpop, llvm-commits
Differential Revision: https://reviews.llvm.org/D28813
llvm-svn: 295746
As i64 isn't a value type on 32-bit targets, we fail to fold the VZEXT_LOAD into VPBROADCASTQ.
Also shows that we're not decoding VPERMIV3 shuffles very well....
llvm-svn: 295729
This matches what is already done during shuffle lowering and helps prevent the need for a zero-vector in cases where shuffles match both patterns.
llvm-svn: 295723
Currently just contains one case where we combine to VZEXT_MOVL instead of VZEXT which would avoid the need for a zero vector to be generated
llvm-svn: 295721
They are all covered by the SSE4.2 intrinsics test with SSE4.2, AVX, and AVX512 command lines.
Merge sse42.ll into the other intrinsics test. Rename sse42_64.ll to be named like other intrinsic tests.
llvm-svn: 295707
They are all covered by the SSE2 intrinsics test with SSE2, AVX, and AVX512 command lines.
Also remove an unneeded lfence intrinsic test since it was already covered.
llvm-svn: 295700
They are all covered by the SSE intrinsics test with SSE, AVX, and AVX512 command lines.
Also remove an unneeded sfence intrinsic test since it was already covered.
llvm-svn: 295699
Summary:
Sandy Bridge and later CPUs have better throughput using a SHLD to implement rotate versus the normal rotate instructions. Additionally it saves one uop and avoids a partial flag update dependency.
This patch implements this change on any Sandy Bridge or later processor without BMI2 instructions. With BMI2 we will use RORX as we currently do.
Reviewers: zvi
Reviewed By: zvi
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D30181
llvm-svn: 295697
Summary:
Currently, BranchFolder drops DebugLoc for branch instructions in some places. For example, for the test code attached, the branch instruction of 'entry' block has a DILocation of
```
!12 = !DILocation(line: 6, column: 3, scope: !11)
```
, but this information is gone when then block is lowered because BranchFolder misses it. This patch is a fix for this issue.
Reviewers: qcolombet, aprantl, craig.topper, MatzeB
Reviewed By: aprantl
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D29902
llvm-svn: 295684
Its more profitable to go through memory (1 cycles throughput)
than using VMOVD + VPERMV/PSHUFB sequence ( 2/3 cycles throughput) to implement EXTRACT_VECTOR_ELT with variable index.
IACA tool was used to get performace estimation (https://software.intel.com/en-us/articles/intel-architecture-code-analyzer)
For example for var_shuffle_v16i8_v16i8_xxxxxxxxxxxxxxxx_i8 test from vector-shuffle-variable-128.ll I get 26 cycles vs 79 cycles.
Removing the VINSERT node, we don't need it any more.
Differential Revision: https://reviews.llvm.org/D29690
llvm-svn: 295660
Summary:
This file was missed in the commit for Cortex-M23 and Cortex-M33
support. See https://reviews.llvm.org/D29073?id=85814 .
Reviewers: rengolin, javed.absar, samparker
Reviewed By: samparker
Subscribers: llvm-commits, aemerson
Differential Revision: https://reviews.llvm.org/D30162
llvm-svn: 295655
Replaces existing approach that could only search BUILD_VECTOR nodes.
Requires getTargetConstantBitsFromNode to discriminate cases with all/partial UNDEF bits in each element - this should also be useful when we get around to supporting getTargetShuffleMaskIndices with UNDEF elements.
llvm-svn: 295613
As discussed on D27692, this permits another domain to be used to combine a shuffle at high depths.
We currently set the required depth at 4 or more combined shuffles, this is probably too high for most targets but is a good starting point and already helps avoid a number of costly variable shuffles.
llvm-svn: 295608
The instructions are marked commutable, but without special handling we don't get the immediate correct.
While here also remove the masked memory forms that aren't commutable.
llvm-svn: 295602
It seems we were already upgrading 128-bit VPCMOV, but the intrinsic was still defined and being used in isel patterns. While I was here I also simplified the tablegen multiclasses.
llvm-svn: 295564
- Adapt MachineBasicBlock::getName() to have the same behavior as the IR
BasicBlock (Value::getName()).
- Add it to lib/CodeGen/CodeGen.cpp::initializeCodeGen so that it is linked in
the CodeGen library.
- MachineRegionInfoPass's name conflicts with RegionInfoPass's name ("region").
- MachineRegionInfo should depend on MachineDominatorTree,
MachinePostDominatorTree and MachineDominanceFrontier instead of their
respective IR versions.
- Since there were no tests for this, add a X86 MIR test.
Patch by Francis Visoiu Mistrih<fvisoiumistrih@apple.com>
llvm-svn: 295518
A line number doesn't make much sense if you don't say where it's
from. Add a verifier check for this and update some tests that had
bogus debug info.
llvm-svn: 295516
When promoting the Load of a Store-Load pair to a COPY all kill flags
between the store and the load need to be cleared.
rdar://30402435
Differential Revision: https://reviews.llvm.org/D30110
llvm-svn: 295512
Removed the HasT2ExtractPack feature and replaced its references
with HasDSP. This then allows the Thumb2 extend instructions to be
selected for ARMv8M +dsp. These instruction descriptions have also
been refactored and more target tests have been added for their isel.
Differential Revision: https://reviews.llvm.org/D29623
llvm-svn: 295452
During legalization we are often creating shuffles (via a build_vector scalarization stage) that are "any_extend_vector_inreg" style masks, and also other masks that are the equivalent of "truncate_vector_inreg" (if we had such a thing).
This patch is an attempt to match these cases to help undo the effects of just leaving shuffle lowering to handle it - which typically means we lose track of the undefined elements of the shuffles resulting in an unnecessary extension+truncation stage for widened illegal types.
The 2011-10-21-widen-cmp.ll regression will be fixed by making SIGN_EXTEND_VECTOR_IN_REG legal in SSE instead of lowering them to X86ISD::VSEXT (PR31712).
Differential Revision: https://reviews.llvm.org/D29454
llvm-svn: 295451
Start using the Subtarget to make decisions about what's legal. In particular,
we only mark floating point operations as legal if we have VFP2, which is
something we should've done from the very start.
llvm-svn: 295439
The original commit was reverted in r283329 due to a miscompile in
Chromium. That turned out to be the same issue as PR31257, which was
fixed in r295262.
llvm-svn: 295357
The existing code always saves the xmm registers for 64-bit targets even if the
target doesn't support SSE (which is common for kernels). Thus, the compiler
inserts movaps instructions which lead to CPU exceptions when an interrupt
handler is invoked.
This commit fixes this bug by returning a register set without xmm registers
from getCalleeSavedRegs and getCallPreservedMask for such targets.
Patch by Philipp Oppermann.
Differential Revision: https://reviews.llvm.org/D29959
llvm-svn: 295347
Resubmit -r295314 with PowerPC and AMDGPU tests updated.
Support {a|s}ext, {a|z|s}ext load nodes as a part of load combine patters.
Reviewed By: filcab
Differential Revision: https://reviews.llvm.org/D29591
llvm-svn: 295336
Support {a|s}ext, {a|z|s}ext load nodes as a part of load combine patters.
Reviewed By: filcab
Differential Revision: https://reviews.llvm.org/D29591
llvm-svn: 295314
Since they're only used for passing around double precision floating point
values into the general purpose registers, we'll lower them to VMOVDRR and
VMOVRRD.
llvm-svn: 295310
For now we just mark them as legal all the time and let the other passes bail
out if they can't handle it. In the future, we'll want to move more of the
brains into the legalizer.
llvm-svn: 295300
For the hard float calling convention, we just use the D registers.
For the soft-fp calling convention, we use the R registers and move values
to/from the D registers by means of G_SEQUENCE/G_EXTRACT. While doing so, we
make sure to honor the endianness of the target, since the CCAssignFn doesn't do
that for us.
For pure soft float targets, we still bail out because we don't support the
libcalls yet.
llvm-svn: 295295
The new 512-bit unmasked intrinsics will make it easy to handle these with the SSE/AVX intrinsics in InstCombine where we currently have a TODO.
llvm-svn: 295290
This reverts r294348, which removed support for conditional tail calls
due to the PR above. It fixes the PR by marking live registers as
implicitly used and defined by the now predicated tailcall. This is
similar to how IfConversion predicates instructions.
Differential Revision: https://reviews.llvm.org/D29856
llvm-svn: 295262
Uses a Custom implementation because the slot sizes being a multiple of the
pointer size isn't really universal, even for the architectures that do have a
simple "void *" va_list.
llvm-svn: 295255
Fixes PR 31921
Summary:
Predicateinfo requires an ugly workaround to try to avoid literal
struct types due to the intrinsic mangling not being implemented.
This workaround actually does not work in all cases (you can hit the
assert by bootstrapping with -print-predicateinfo), and can't be made
to work without DFS'ing the type (IE copying getMangledStr and using a
version that detects if it would crash).
Rather than do that, i just implemented the mangling. It seems
simple, since they are unified structurally.
Looking at the overloaded-mangling testcase we have, it actually turns
out the gc intrinsics will *also* crash if you try to use a literal
struct. Thus, the testcase added fails before this patch, and works
after, without needing to resort to predicateinfo.
Reviewers: chandlerc, davide
Subscribers: llvm-commits, sanjoy
Differential Revision: https://reviews.llvm.org/D29925
llvm-svn: 295253
Lay out trellis-shaped CFGs optimally.
A trellis of the shape below:
A B
|\ /|
| \ / |
| X |
| / \ |
|/ \|
C D
would be laid out A; B->C ; D by the current layout algorithm. Now we identify
trellises and lay them out either A->C; B->D or A->D; B->C. This scales with an
increasing number of predecessors. A trellis is a a group of 2 or more
predecessor blocks that all have the same successors.
because of this we can tail duplicate to extend existing trellises.
As an example consider the following CFG:
B D F H
/ \ / \ / \ / \
A---C---E---G---Ret
Where A,C,E,G are all small (Currently 2 instructions).
The CFG preserving layout is then A,B,C,D,E,F,G,H,Ret.
The current code will copy C into B, E into D and G into F and yield the layout
A,C,B(C),E,D(E),F(G),G,H,ret
define void @straight_test(i32 %tag) {
entry:
br label %test1
test1: ; A
%tagbit1 = and i32 %tag, 1
%tagbit1eq0 = icmp eq i32 %tagbit1, 0
br i1 %tagbit1eq0, label %test2, label %optional1
optional1: ; B
call void @a()
br label %test2
test2: ; C
%tagbit2 = and i32 %tag, 2
%tagbit2eq0 = icmp eq i32 %tagbit2, 0
br i1 %tagbit2eq0, label %test3, label %optional2
optional2: ; D
call void @b()
br label %test3
test3: ; E
%tagbit3 = and i32 %tag, 4
%tagbit3eq0 = icmp eq i32 %tagbit3, 0
br i1 %tagbit3eq0, label %test4, label %optional3
optional3: ; F
call void @c()
br label %test4
test4: ; G
%tagbit4 = and i32 %tag, 8
%tagbit4eq0 = icmp eq i32 %tagbit4, 0
br i1 %tagbit4eq0, label %exit, label %optional4
optional4: ; H
call void @d()
br label %exit
exit:
ret void
}
here is the layout after D27742:
straight_test: # @straight_test
; ... Prologue elided
; BB#0: # %entry ; A (merged with test1)
; ... More prologue elided
mr 30, 3
andi. 3, 30, 1
bc 12, 1, .LBB0_2
; BB#1: # %test2 ; C
rlwinm. 3, 30, 0, 30, 30
beq 0, .LBB0_3
b .LBB0_4
.LBB0_2: # %optional1 ; B (copy of C)
bl a
nop
rlwinm. 3, 30, 0, 30, 30
bne 0, .LBB0_4
.LBB0_3: # %test3 ; E
rlwinm. 3, 30, 0, 29, 29
beq 0, .LBB0_5
b .LBB0_6
.LBB0_4: # %optional2 ; D (copy of E)
bl b
nop
rlwinm. 3, 30, 0, 29, 29
bne 0, .LBB0_6
.LBB0_5: # %test4 ; G
rlwinm. 3, 30, 0, 28, 28
beq 0, .LBB0_8
b .LBB0_7
.LBB0_6: # %optional3 ; F (copy of G)
bl c
nop
rlwinm. 3, 30, 0, 28, 28
beq 0, .LBB0_8
.LBB0_7: # %optional4 ; H
bl d
nop
.LBB0_8: # %exit ; Ret
ld 30, 96(1) # 8-byte Folded Reload
addi 1, 1, 112
ld 0, 16(1)
mtlr 0
blr
The tail-duplication has produced some benefit, but it has also produced a
trellis which is not laid out optimally. With this patch, we improve the layouts
of such trellises, and decrease the cost calculation for tail-duplication
accordingly.
This patch produces the layout A,C,E,G,B,D,F,H,Ret. This layout does have
back edges, which is a negative, but it has a bigger compensating
positive, which is that it handles the case where there are long strings
of skipped blocks much better than the original layout. Both layouts
handle runs of executed blocks equally well. Branch prediction also
improves if there is any correlation between subsequent optional blocks.
Here is the resulting concrete layout:
straight_test: # @straight_test
; BB#0: # %entry ; A (merged with test1)
mr 30, 3
andi. 3, 30, 1
bc 12, 1, .LBB0_4
; BB#1: # %test2 ; C
rlwinm. 3, 30, 0, 30, 30
bne 0, .LBB0_5
.LBB0_2: # %test3 ; E
rlwinm. 3, 30, 0, 29, 29
bne 0, .LBB0_6
.LBB0_3: # %test4 ; G
rlwinm. 3, 30, 0, 28, 28
bne 0, .LBB0_7
b .LBB0_8
.LBB0_4: # %optional1 ; B (Copy of C)
bl a
nop
rlwinm. 3, 30, 0, 30, 30
beq 0, .LBB0_2
.LBB0_5: # %optional2 ; D (Copy of E)
bl b
nop
rlwinm. 3, 30, 0, 29, 29
beq 0, .LBB0_3
.LBB0_6: # %optional3 ; F (Copy of G)
bl c
nop
rlwinm. 3, 30, 0, 28, 28
beq 0, .LBB0_8
.LBB0_7: # %optional4 ; H
bl d
nop
.LBB0_8: # %exit
Differential Revision: https://reviews.llvm.org/D28522
llvm-svn: 295223
We currently can't legalize those, but we should really not be creating
them in the first place, since legalization would probably look similar to the
way we legalize CONCAT_VECTORS - basically replace the INSERT with a BUILD.
This fixes PR311956.
Differential Revision: https://reviews.llvm.org/D29961
llvm-svn: 295213
This patch reverts region's scheduling to the original untouched state
in case if we have have decreased occupancy.
In addition it switches to use TargetRegisterInfo occupancy callback
for pressure limits instead of gradually increasing limits which were
just passed by. We are going to stay with the best schedule so we do
not need to tolerate worsened scheduling anymore.
Differential Revision: https://reviews.llvm.org/D29971
llvm-svn: 295206
Summary:
We don't seem to have great rules on what a valid VBROADCAST node looks like. And as a consequence we end up with a lot of patterns to try to catch everything. We have patterns with scalar inputs, 128-bit vector inputs, 256-bit vector inputs, and 512-bit vector inputs.
As you can see from the things improved here we are currently missing patterns for 128-bit loads being extended to 256-bit before the vbroadcast.
I'd like to propose that VBROADCAST should always take a 128-bit vector type as input. As a first step towards that this patch adds an EXTRACT_SUBVECTOR in front of VBROADCAST when the input is 256 or 512-bits. In the future I would like to add scalar_to_vector around all the scalar operations. And maybe we should consider adding a VBROADCAST+load node to avoid separating loads from the broadcasting operation when the load itself isn't foldable.
This requires an additional change in target shuffle combining to look for the extract subvector and look through it to find the original operand. I'm sure this change isn't perfect but was enough to fix a few test failures that were being caused.
Another interesting thing I noticed is that the changes in masked_gather_scatter.ll show cases were we don't remove a useless insert into element 1 before broadcasting element 0.
Reviewers: delena, RKSimon, zvi
Reviewed By: zvi
Subscribers: igorb, llvm-commits
Differential Revision: https://reviews.llvm.org/D28747
llvm-svn: 295155
This patch corrects the maximum workgroups per CU if we have big
workgroups (more than 128). This calculation contributes to the
occupancy calculation in respect to LDS size.
Differential Revision: https://reviews.llvm.org/D29974
llvm-svn: 295134
Summary:
Blocks ending in unreachable are typically cold because they end the
program or throw an exception, so merging them with other identical
blocks is usually profitable because it reduces the size of cold code.
MachineBlockPlacement generally does not arrange to fall through to such
blocks, so commoning these blocks will not introduce additional
unconditional branches.
Reviewers: hans, iteratee, haicheng
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D29153
llvm-svn: 295105
This instruction clears the low bits of a pointer without requiring (possibly
dodgy if pointers aren't ints) conversions to and from an integer. Since (as
far as I'm aware) all masks are statically known, the instruction takes an
immediate operand rather than a register to specify the mask.
llvm-svn: 295103
This adds MXCSR to the set of recognized registers for X86 targets and updates the instructions that read or write it. I do not intend for all of the various floating point instructions that implicitly use the control bits or update the status bits of this register to ever have that usage modeled by default. However, when constrained floating point modes (such as strict FP exception status modeling or dynamic rounding modes) are enabled, implicit use/def information for MXCSR will be added to those instructions.
Until those additional updates are made this should cause (almost?) no functional changes. Theoretically, this will prevent instructions like LDMXCSR and STMXCSR from being moved past one another, but that should be prevented anyway and I haven't found a case where it is happening now.
Differential Revision: https://reviews.llvm.org/D29903
llvm-svn: 295004
Simon Pilgrim