First half of PR40800, this patch adds DAG undef handling to icmp instructions to match the behaviour in llvm::ConstantFoldCompareInstruction and SimplifyICmpInst, this permits constant folding of vector comparisons where some elements had been reduced to UNDEF (by SimplifyDemandedVectorElts etc.).
This involved a lot of tweaking to reduced tests as bugpoint loves to reduce icmp arguments to undef........
Differential Revision: https://reviews.llvm.org/D59363
llvm-svn: 356938
This is the result of discussions on the list about how to deal with intrinsics
which require codegen to disambiguate them via only the integer/fp overloads.
It causes problems for GlobalISel as some of that information is lost during
translation, while with other operations like IR instructions the information is
encoded into the instruction opcode.
This patch changes clang to emit the new faddp intrinsic if the vector operands
to the builtin have FP element types. LLVM IR AutoUpgrade has been taught to
upgrade existing calls to aarch64.neon.addp with fp vector arguments, and
we remove the workarounds introduced for GlobalISel in r355865.
This is a more permanent solution to PR40968.
Differential Revision: https://reviews.llvm.org/D59655
llvm-svn: 356722
The AArch64 test was broken since the result register already had a
set register class, so this test was a no-op. The mapping verify call
would fail because the result size is not the same as the inputs like
in a copy or phi.
The AMDGPU testcases are half broken and introduce illegal VGPR->SGPR
copies which need much more work to handle correctly (same for phis),
but add them as a baseline.
llvm-svn: 356713
Added subtarget features for AArch64 to use TPIDR_EL[1|2|3] as the TLS base
register, rather than the default TPIDR_EL0.
Patch by Philip Derrin!
Differential revision: https://reviews.llvm.org/D54685
llvm-svn: 356657
This adds pattern matching for the insert+shufflevector sequence so we can
generate dup instructions instead of the current TBL sequence.
Differential Revision: https://reviews.llvm.org/D59558
llvm-svn: 356526
The 2nd loop calculates spill costs but reports free registers as cost
0 anyway, so there is little benefit from having a separate early
loop.
Surprisingly this is not NFC, as many register are marked regDisabled
so the first loop often picks up later registers unnecessarily instead
of the first one available in the allocation order...
Patch by Matthias Braun
llvm-svn: 356499
These changes are related to PR37743 and include:
SelectionDAGBuilder::visitSelect handles the unary SelectPatternFlavor::SPF_ABS case to build ABS node.
Delete the redundant recognizer of the integer ABS pattern from the DAGCombiner.
Add promoting the integer ABS node in the LegalizeIntegerType.
Expand-based legalization of integer result for the ABS nodes.
Expand-based legalization of ABS vector operations.
Add some integer abs testcases for different typesizes for Thumb arch
Add the custom ABS expanding and change the SAD pattern recognizer for X86 arch: The i64 result of the ABS is expanded to:
tmp = (SRA, Hi, 31)
Lo = (UADDO tmp, Lo)
Hi = (XOR tmp, (ADDCARRY tmp, hi, Lo:1))
Lo = (XOR tmp, Lo)
The "detectZextAbsDiff" function is changed for the recognition of pattern with the ABS node. Given a ABS node, detect the following pattern:
(ABS (SUB (ZERO_EXTEND a), (ZERO_EXTEND b))).
Change integer abs testcases for codegen with the ABS node support for AArch64.
Indicate that the ABS is legal for the i64 type when the NEON is supported.
Change the integer abs testcases to show changing of codegen.
Add combine and legalization of ABS nodes for Thumb arch.
Extend 'matchSelectPattern' to recognize the ABS patterns with ICMP_SGE condition.
For discussion, see https://bugs.llvm.org/show_bug.cgi?id=37743
Patch by: @ikulagin (Ivan Kulagin)
Differential Revision: https://reviews.llvm.org/D49837
llvm-svn: 356468
It uses the generic AArch64_IMM::expandMOVImm to get the correct
number of instruction used in immediate materialization.
Reviewers: efriedma
Differential Revision: https://reviews.llvm.org/D58461
llvm-svn: 356391
This patch follows some ideas from r352866 to optimize the floating
point materialization even further. It changes isFPImmLegal to
considere up to 2 mov instruction or up to 5 in case subtarget has
fused literals.
The rationale is the cost is the same for mov+fmov vs. adrp+ldr; but
the mov+fmov sequence is always better because of the reduced d-cache
pressure. The timings are still the same if you consider movw+movk+fmov
vs. adrp+ldr will be fused (although one instruction longer).
Reviewers: efriedma
Differential Revision: https://reviews.llvm.org/D58460
llvm-svn: 356390
Fixes https://bugs.llvm.org/show_bug.cgi?id=35094
The Dead register definition pass should leave alone the atomicrmw
instructions on AArch64 (LTE extension). The reason is the following
statement in the Arm ARM:
"The ST<OP> instructions, and LD<OP> instructions where the destination
register is WZR or XZR, are not regarded as doing a read for the purpose
of a DMB LD barrier."
A good example was given in the gcc thread by Will Deacon (linked in the
bugzilla ticket 35094):
P0 (atomic_int* y,atomic_int* x) {
atomic_store_explicit(x,1,memory_order_relaxed);
atomic_thread_fence(memory_order_release);
atomic_store_explicit(y,1,memory_order_relaxed);
}
P1 (atomic_int* y,atomic_int* x) {
atomic_fetch_add_explicit(y,1,memory_order_relaxed); // STADD
atomic_thread_fence(memory_order_acquire);
int r0 = atomic_load_explicit(x,memory_order_relaxed);
}
P2 (atomic_int* y) {
int r1 = atomic_load_explicit(y,memory_order_relaxed);
}
My understanding is that it is forbidden for r0 == 0 and r1 == 2 after
this test has executed. However, if the relaxed add in P1 compiles to
STADD and the subsequent acquire fence is compiled as DMB LD, then we
don't have any ordering guarantees in P1 and the forbidden result could
be observed.
Change-Id: I419f9f9df947716932038e1100c18d10a96408d0
llvm-svn: 356360
Fold (x & ~y) | y and it's four commuted variants to x | y. This pattern
can in particular appear when a vselect c, x, -1 is expanded to
(x & ~c) | (-1 & c) and combined to (x & ~c) | c.
This change has some overlap with D59066, which avoids creating a
vselect of this form in the first place during uaddsat expansion.
Differential Revision: https://reviews.llvm.org/D59174
llvm-svn: 356333
This is a subset of what was proposed in:
D59006
...and may overlap with test changes from:
D59174
...but it seems like a good general optimization to turn selects
into bitwise-logic when possible because we never know exactly
what can happen at this stage of DAG combining depending on how
the target has defined things.
Differential Revision: https://reviews.llvm.org/D59066
llvm-svn: 356332
Switch BIC immediate creation for vector ANDs from custom lowering
to a DAG combine, which gives generic DAG combines a change to
apply first. In particular this avoids (and x, -1) being turned into
a (bic x, 0) instead of being eliminated entirely.
Differential Revision: https://reviews.llvm.org/D59187
llvm-svn: 356299
Since we can't insert s16 gprs as we don't have 16 bit GPR registers, we need to
teach RBS to assign them to the FPR bank so our selector works.
llvm-svn: 356282
This isn't necessary according to the DWARF standard, but it matches the
.eh_frame sections emitted by other tools in practice, and the Android
libunwindstack rejects .eh_frame sections where an FDE refers to a CIE
other than the closest previous CIE. So match the other tools and also
sort accordingly.
I consider this a bug in libunwindstack, but it's easy enough to emit
a compatible .eh_frame section for compatibility with installed
operating systems.
Differential Revision: https://reviews.llvm.org/D58266
llvm-svn: 356216
This adds instruction selection support for G_UADDO on s32s and s64s.
Also
- Add an instruction selection test
- Update the arm64-xaluo.ll test to show that we generate the correct assembly
Differential Revision: https://reviews.llvm.org/D58734
llvm-svn: 356214
This re-uses the previous support for extract vector elt to extract the
subvectors.
Differential Revision: https://reviews.llvm.org/D59390
llvm-svn: 356213
This adds support for inserting elements into packed vectors. It also adds
two tests: one for selection, and one for regbank select.
Unpacked vectors will come in a follow-up.
Differential Revision: https://reviews.llvm.org/D59325
llvm-svn: 356182
This is consistent with what SelectionDAG does and is much easier to
work with than the extract sequence with an artificial wide register.
For the AMDGPU control flow intrinsics, this was producing an s128 for
the i64, i1 tuple return. Any legalization that should apply to a real
s128 value would badly obscure the direct values that need to be seen.
llvm-svn: 356147
Summary:
A number of optimizations are inhibited by single-use TokenFactors not
being merged into the TokenFactor using it. This makes we consider if
we can do the merge immediately.
Most tests changes here are due to the change in visitation causing
minor reorderings and associated reassociation of paired memory
operations.
CodeGen tests with non-reordering changes:
X86/aligned-variadic.ll -- memory-based add folded into stored leaq
value.
X86/constant-combiners.ll -- Optimizes out overlap between stores.
X86/pr40631_deadstore_elision -- folds constant byte store into
preceding quad word constant store.
Reviewers: RKSimon, craig.topper, spatel, efriedma, courbet
Reviewed By: courbet
Subscribers: dylanmckay, sdardis, nemanjai, jvesely, nhaehnle, javed.absar, eraman, hiraditya, kbarton, jrtc27, atanasyan, jsji, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D59260
llvm-svn: 356068
First step towards PR40800 - I intend to move the float case in a separate future patch.
I had to tweak the (overly reduced) thumb2 test and the x86 widening test change is annoying (no longer rematerializable) but we should address this separately.
Differential Revision: https://reviews.llvm.org/D59244
llvm-svn: 356040
After r355865, we should be able to safely select G_EXTRACT_VECTOR_ELT without
running into any problematic intrinsics.
Also add a fix for lane copies, which don't support index 0.
llvm-svn: 355871
Overloaded intrinsics aren't necessarily safe for instruction selection. One
such intrinsic is aarch64.neon.addp.*.
This is a temporary workaround to ensure that we always fall back on that
intrinsic. Eventually this will be replaced with a proper solution.
https://bugs.llvm.org/show_bug.cgi?id=40968
Differential Revision: https://reviews.llvm.org/D59062
llvm-svn: 355865
Fixes https://bugs.llvm.org/show_bug.cgi?id=36796.
Implement basic legalizations (PromoteIntRes, PromoteIntOp,
ExpandIntRes, ScalarizeVecOp, WidenVecOp) for VECREDUCE opcodes.
There are more legalizations missing (esp float legalizations),
but there's no way to test them right now, so I'm not adding them.
This also includes a few more changes to make this work somewhat
reasonably:
* Add support for expanding VECREDUCE in SDAG. Usually
experimental.vector.reduce is expanded prior to codegen, but if the
target does have native vector reduce, it may of course still be
necessary to expand due to legalization issues. This uses a shuffle
reduction if possible, followed by a naive scalar reduction.
* Allow the result type of integer VECREDUCE to be larger than the
vector element type. For example we need to be able to reduce a v8i8
into an (nominally) i32 result type on AArch64.
* Use the vector operand type rather than the scalar result type to
determine the action, so we can control exactly which vector types are
supported. Also change the legalize vector op code to handle
operations that only have vector operands, but no vector results, as
is the case for VECREDUCE.
* Default VECREDUCE to Expand. On AArch64 (only target using VECREDUCE),
explicitly specify for which vector types the reductions are supported.
This does not handle anything related to VECREDUCE_STRICT_*.
Differential Revision: https://reviews.llvm.org/D58015
llvm-svn: 355860
Summary:
This change change the instrumentation to allow users to view the registers at the point at which tag mismatch occured. Most of the heavy lifting is done in the runtime library, where we save the registers to the stack and emit unwind information. This allows us to reduce the overhead, as very little additional work needs to be done in each __hwasan_check instance.
In this implementation, the fast path of __hwasan_check is unmodified. There are an additional 4 instructions (16B) emitted in the slow path in every __hwasan_check instance. This may increase binary size somewhat, but as most of the work is done in the runtime library, it's manageable.
The failure trace now contains a list of registers at the point of which the failure occured, in a format similar to that of Android's tombstones. It currently has the following format:
Registers where the failure occurred (pc 0x0055555561b4):
x0 0000000000000014 x1 0000007ffffff6c0 x2 1100007ffffff6d0 x3 12000056ffffe025
x4 0000007fff800000 x5 0000000000000014 x6 0000007fff800000 x7 0000000000000001
x8 12000056ffffe020 x9 0200007700000000 x10 0200007700000000 x11 0000000000000000
x12 0000007fffffdde0 x13 0000000000000000 x14 02b65b01f7a97490 x15 0000000000000000
x16 0000007fb77376b8 x17 0000000000000012 x18 0000007fb7ed6000 x19 0000005555556078
x20 0000007ffffff768 x21 0000007ffffff778 x22 0000000000000001 x23 0000000000000000
x24 0000000000000000 x25 0000000000000000 x26 0000000000000000 x27 0000000000000000
x28 0000000000000000 x29 0000007ffffff6f0 x30 00000055555561b4
... and prints after the dump of memory tags around the buggy address.
Every register is saved exactly as it was at the point where the tag mismatch occurs, with the exception of x16/x17. These registers are used in the tag mismatch calculation as scratch registers during __hwasan_check, and cannot be saved without affecting the fast path. As these registers are designated as scratch registers for linking, there should be no important information in them that could aid in debugging.
Reviewers: pcc, eugenis
Reviewed By: pcc, eugenis
Subscribers: srhines, kubamracek, mgorny, javed.absar, krytarowski, kristof.beyls, hiraditya, jdoerfert, llvm-commits, #sanitizers
Tags: #sanitizers, #llvm
Differential Revision: https://reviews.llvm.org/D58857
llvm-svn: 355738
During the lowering of a switch that would result in the generation of a
jump table, a range check is performed before indexing into the jump
table, for the switch value being outside the jump table range and a
conditional branch is inserted to jump to the default block. In case the
default block is unreachable, this conditional jump can be omitted. This
patch implements omitting this conditional branch for unreachable
defaults.
Differential Revision: https://reviews.llvm.org/D52002
Reviewers: Hans Wennborg, Eli Freidman, Roman Lebedev
llvm-svn: 355490
During the lowering of a switch that would result in the generation of a
jump table, a range check is performed before indexing into the jump
table, for the switch value being outside the jump table range and a
conditional branch is inserted to jump to the default block. In case the
default block is unreachable, this conditional jump can be omitted. This
patch implements omitting this conditional branch for unreachable
defaults.
Differential Revision: https://reviews.llvm.org/D52002
Reviewers: Hans Wennborg, Eli Freidman, Roman Lebedev
llvm-svn: 355483
Tests only for integers, not floating point or pointers.
The scalar 8-bit case uses branch instead of CMOV,
because there is no no 8-bit CMOV.
Vector tests are for consistency, since it can be vectorized.
https://bugs.llvm.org/show_bug.cgi?id=40965
llvm-svn: 355436
This adds instruction selection support for G_EXTRACT_VECTOR_ELT for cases
where the index is defined by a G_CONSTANT.
It also factos out the lane copy opcode selection part into its own function,
`getLaneCopyOpcode`. This is used by both `selectUnmergeValues` and
`selectExtractElt`.
Differential Revision: https://reviews.llvm.org/D58469
llvm-svn: 355344
The code to materialize a mask from a constant pool load tried to use a 128 bit
LDR to load a 64 bit constant pool entry, which was 8 byte aligned. This resulted
in a link failure in the NEON tests in the test suite since the LDR address was
unaligned. This change fixes that to instead emit a 64 bit LDR if the entry is
64 bit, before converting back to a 128 bit register for the TBL.
llvm-svn: 355326
This patch enables combining integer bitcasts of integer build vectors when the new scalar type is legal. I've avoided floating point because the implementation bitcasts float to int along the way and we would need to check the intermediate types for legality
Differential Revision: https://reviews.llvm.org/D58884
llvm-svn: 355324
In certain cases, the first non-frame-setup instruction in a function is
a branch. For example, it could be a cbz on an argument. Make sure we
correctly allocate the UnwindHelp, and find an appropriate register to
use to initialize it.
Fixes https://bugs.llvm.org/show_bug.cgi?id=40184
Differential Revision: https://reviews.llvm.org/D58752
llvm-svn: 355136
This extends the existing support for shufflevector to handle cases like
<2 x float>, which we can implement by concating the vectors and using a TBL1.
Differential Revision: https://reviews.llvm.org/D58684
llvm-svn: 355104
Simon Pilgrim