Summary:
This problem consists of several parts:
* Basic sign bit extraction - `trunc? (?shr %x, (bitwidth(x)-1))`.
This is trivial, and easy to do, we have a fold for it.
* Shift amount reassociation - if we have two identical shifts,
and we can simplify-add their shift amounts together,
then we likely can just perform them as a single shift.
But this is finicky, has one-use restrictions,
and shift opcodes must be identical.
But there is a super-pattern where both of these work together.
to produce sign bit test from two shifts + comparison.
We do indeed already handle this in most cases.
But since we get that fold transitively, it has one-use restrictions.
And what's worse, in this case the right-shifts aren't required to be
identical, and we can't handle that transitively:
If the total shift amount is bitwidth-1, only a sign bit will remain
in the output value. But if we look at this from the perspective of
two shifts, we can't fold - we can't possibly know what bit pattern
we'd produce via two shifts, it will be *some* kind of a mask
produced from original sign bit, but we just can't tell it's shape:
https://rise4fun.com/Alive/cM0https://rise4fun.com/Alive/9IN
But it will *only* contain sign bit and zeros.
So from the perspective of sign bit test, we're good:
https://rise4fun.com/Alive/FRzhttps://rise4fun.com/Alive/qBU
Superb!
So the simplest solution is to extend `reassociateShiftAmtsOfTwoSameDirectionShifts()` to also have a
sudo-analysis mode that will ignore extra-uses, and will only check
whether a) those are two right shifts and b) they end up with bitwidth(x)-1
shift amount and return either the original value that we sign-checking,
or null.
This does not have any functionality change for
the existing `reassociateShiftAmtsOfTwoSameDirectionShifts()`.
All that being said, as disscussed in the review, this yet again
increases usage of instsimplify in instcombine as utility.
Some day that may need to be reevaluated.
https://bugs.llvm.org/show_bug.cgi?id=43595
Reviewers: spatel, efriedma, vsk
Reviewed By: spatel
Subscribers: xbolva00, hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D68930
llvm-svn: 375371
We handle it this way for some other address spaces.
Since r349196, SILoadStoreOptimizer has been trying to do this. This
is after SIFoldOperands runs, which can change the addressing
patterns. It's simpler to just split this earlier.
llvm-svn: 375366
This patch tries to resolve problems faced in D68943
and uses some of the code written by Konrad Wilhelm Kleine
in that patch.
Previously, yaml2obj tool always created a .symtab section.
This patch changes that. With it we only create it when
have a "Symbols:" tag in the YAML document or when
we need to create it because it is used by another section(s).
obj2yaml follows the new behavior and does not print "Symbols:"
anymore when there is no symbol table.
Differential revision: https://reviews.llvm.org/D69041
llvm-svn: 375361
This is a common idiom which arises after induction variables are widened, and we have two or more exit conditions. Interestingly, we don't have instcombine or instsimplify support for this either.
Differential Revision: https://reviews.llvm.org/D69006
llvm-svn: 375349
Provides a TLI hook to allow targets to relax the emission of shifts, thus enabling
codegen improvements on targets with no multiple shift instructions and cheap selects
or branches.
Contributes to a Fix for PR43559:
https://bugs.llvm.org/show_bug.cgi?id=43559
Patch by: @joanlluch (Joan LLuch)
Differential Revision: https://reviews.llvm.org/D69116
llvm-svn: 375347
tryToWidenViaDuplication lowers using the shuffle_v8i16(unpack_v16i8(shuffle_v8i16(x),shuffle_v8i16(x))) pattern, but the unpack only needs the even/odd 16i8 args if the original v16i8 shuffle mask references the even/odd elements - which isn't true for many extension style shuffles.
llvm-svn: 375342
We were always generating a single source HADDPD, but really we should only do this if shouldUseHorizontalOp says its a good idea.
Differential Revision: https://reviews.llvm.org/D69175
llvm-svn: 375341
It's not clear why the test had this. I'm unable to break the original
case with the original patch reverted with or without optnone.
This avoids a failure in a future commit.
llvm-svn: 375321
If a subregister def was moved across another subregister def and
another use, the main range was not correctly updated. The end point
of the moved interval ended too early and missed the use from theh
other lanes in the subreg def.
llvm-svn: 375300
If all uses of a PHI are in AGPR register class we should
avoid unneeded copies via VGPRs.
Differential Revision: https://reviews.llvm.org/D69200
llvm-svn: 375297
by ExtBinary format profile
Profile on-demand loading was added for ExtBinary format profile in rL374233,
but currently profile on-demand loading doesn't work well with profile
remapping. The patch adds the support.
Suppose a function in the current module has outline instance in the profile.
The function name in the module is different from the name of the outline
instance, but remapper knows the two names are equal. When loading profile
on-demand, the outline instance has to be loaded with remapper's help.
At the same time SampleProfileReaderItaniumRemapper is changed from a proxy
of SampleProfileReader to a helper member in SampleProfileReader.
Differential Revision: https://reviews.llvm.org/D68901
llvm-svn: 375295
Summary: The implementation was never completed and never used except in tests.
Reviewers: arsenm, mareko
Subscribers: qcolombet, kzhuravl, jvesely, wdng, nhaehnle, yaxunl, dstuttard, tpr, t-tye, hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D69163
llvm-svn: 375293
Previously, the parser checked for a '%' prefix to indicate a register.
In Intel syntax mode, LLVM does not print a '%' prefix on registers, so
LLVM could not parse its own assembly output. Instead, require that
register numbers be integer literals, or at least start with an integer
literal, which is consistent with .cfi_* directive register parsing.
llvm-svn: 375287
Summary:
Also changes the wasm YAML format to reflect the possibility of having
multiple return types and to put the returns after the params for
consistency with the binary encoding.
Reviewers: aheejin, sbc100
Subscribers: dschuff, jgravelle-google, hiraditya, sunfish, arphaman, rupprecht, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D69156
llvm-svn: 375283
Summary:
CVP, unlike InstCombine, does not run till exaustion.
It only does a single pass.
When dealing with those special binops, if we prove that they can
safely be demoted into their usual binop form,
we do set the no-wrap we deduced. But when dealing with usual binops,
we try to deduce both no-wraps.
So if we convert e.g. @llvm.uadd.with.overflow() to `add nuw`,
we won't attempt to check whether it can be `add nuw nsw`.
This patch proposes to call `processBinOp()` on newly-created binop,
which is identical to what we do for div/rem already.
Reviewers: nikic, spatel, reames
Reviewed By: nikic
Subscribers: hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D69183
llvm-svn: 375273
Mostly use SReg_32 instead of SReg_32_XM0 for arbitrary values. This
will allow the register coalescer to do a better job eliminating
copies to m0.
For GlobalISel, as a terrible hack, use SGPR_32 for things that should
use SCC until booleans are solved.
llvm-svn: 375267
Summary:
The PMMIR_EL1 register is present in Armv8.4 with PMU extension.
This patch adds support for it.
Reviewers: t.p.northover, dnsampaio
Reviewed By: dnsampaio
Subscribers: kristof.beyls, hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D68940
llvm-svn: 375228
Adds a new ISD node to replicate a scalar value across all elements of
a vector. This is needed for scalable vectors, since BUILD_VECTOR cannot
be used.
Fixes up default type legalization for scalable vectors after the
new MVT type ranges were introduced.
At present I only use this node for scalable vectors. A DAGCombine has
been added to transform a BUILD_VECTOR into a SPLAT_VECTOR if all
elements are the same, but only if the default operation action of
Expand has been overridden by the target.
I've only added result promotion legalization for scalable vector
i8/i16/i32/i64 types in AArch64 for now.
Reviewers: t.p.northover, javed.absar, greened, cameron.mcinally, jmolloy
Reviewed By: jmolloy
Differential Revision: https://reviews.llvm.org/D47775
llvm-svn: 375222
For arm64, D18619 introduced the ability to combine bumping the stack pointer
upfront in case it needs to be bumped for both the callee-save area as well as
the local stack area.
That diff already remarks that "This change can cause an increase in
instructions", but argues that even when that happens, it should be still be a
performance benefit because the number of micro-ops is reduced.
We have observed that this code-size increase can be significant in practice.
This diff disables combining stack bumping for methods that are marked as
optimize-for-size.
Example of a prologue with the behavior before this diff (combining stack bumping when possible):
sub sp, sp, #0x40
stp d9, d8, [sp, #0x10]
stp x20, x19, [sp, #0x20]
stp x29, x30, [sp, #0x30]
add x29, sp, #0x30
[... compute x8 somehow ...]
stp x0, x8, [sp]
And after this diff, if the method is marked as optimize-for-size:
stp d9, d8, [sp, #-0x30]!
stp x20, x19, [sp, #0x10]
stp x29, x30, [sp, #0x20]
add x29, sp, #0x20
[... compute x8 somehow ...]
stp x0, x8, [sp, #-0x10]!
Note that without combining the stack bump there are two auto-decrements,
nicely folded into the stp instructions, whereas otherwise there is a single
sub sp, ... instruction, but not folded.
Patch by Nikolai Tillmann!
Differential Revision: https://reviews.llvm.org/D68530
llvm-svn: 375217
The default promotion for the add_sat/sub_sat nodes currently does:
ANY_EXTEND iN to iM
SHL by M-N
[US][ADD|SUB]SAT
L/ASHR by M-N
If the promoted add_sat or sub_sat node is not legal, this can produce code
that effectively does a lot of shifting (and requiring large constants to be
materialised) just to use the overflow flag. It is simpler to just do the
saturation manually, using the higher bitwidth addition and a min/max against
the saturating bounds. That is what this patch attempts to do.
Differential Revision: https://reviews.llvm.org/D68926
llvm-svn: 375211
Summary:
Implements the following intrinsics:
- int_aarch64_sve_sunpkhi
- int_aarch64_sve_sunpklo
- int_aarch64_sve_uunpkhi
- int_aarch64_sve_uunpklo
This patch also adds AArch64ISD nodes for UNPK instead of implementing
the intrinsics directly, as they are required for a future patch which
implements the sign/zero extension of legal vectors.
This patch includes tests for the Subdivide2Argument type added by D67549
Reviewers: sdesmalen, SjoerdMeijer, greened, rengolin, rovka
Reviewed By: greened
Subscribers: tschuett, kristof.beyls, rkruppe, psnobl, cfe-commits, llvm-commits
Differential Revision: https://reviews.llvm.org/D67550
llvm-svn: 375210
Adding the reproducer from https://bugs.llvm.org/show_bug.cgi?id=43689,
showing that instcombine is doing a bad transform. It transforms
%0 = insertelement <2 x i16> undef, i16 %a, i32 0
%1 = srem <2 x i16> %0, <i16 2, i16 1>
%2 = shufflevector <2 x i16> %1, <2 x i16> undef, <2 x i32> <i32 undef, i32 0>
into
%1 = insertelement <2 x i16> undef, i16 %a, i32 1
%2 = srem <2 x i16> %1, <i16 undef, i16 2>
The undef denominator makes the whole srem undefined.
llvm-svn: 375207
This relands r374931 (reverted in r375088). It fixes 32-bit builds by using the right format string specifier for uint64_t (PRIu64) instead of `%d`.
Original description:
When listing the index in `llvm-objdump -h`, use a zero-based counter instead of the actual section index (e.g. shdr->sh_index for ELF).
While this is effectively a noop for now (except one unit test for XCOFF), the index values will change in a future patch that filters certain sections out (e.g. symbol tables). See D68669 for more context. Note: the test case in `test/tools/llvm-objdump/X86/section-index.s` already covers the case of incrementing the section index counter when sections are skipped.
Reviewers: grimar, jhenderson, espindola
Reviewed By: grimar
Subscribers: emaste, sbc100, arichardson, aheejin, arphaman, seiya, llvm-commits, MaskRay
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D68848
llvm-svn: 375178
Summary: GNU objcopy accepts the --wildcard flag to allow wildcard matching on symbol-related flags. (Note: it's implicitly true for section flags).
The basic syntax is to allow *, ?, \, and [] which work similarly to how they work in a shell. Additionally, starting a wildcard with ! causes that wildcard to prevent it from matching a flag.
Use an updated GlobPattern in libSupport to handle these patterns. It does not fully match the `fnmatch` used by GNU objcopy since named character classes (e.g. `[[:digit:]]`) are not supported, but this should support most existing use cases (mostly just `*` is what's used anyway).
Reviewers: jhenderson, MaskRay, evgeny777, espindola, alexshap
Reviewed By: MaskRay
Subscribers: nickdesaulniers, emaste, arichardson, hiraditya, jakehehrlich, abrachet, seiya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D66613
llvm-svn: 375169
Quite a while ago, we implemented a pass that will reduce the number of
CR-logical operations we emit. It does so by converting a CR-logical operation
into a branch. We have kept this off by default because it seemed to cause a
significant regression with one benchmark.
However, that regression turned out to be due to a completely unrelated
reason - AADB introducing a self-copy that is a priority-setting nop and it was
just exacerbated by this pass.
Now that we understand the reason for the only degradation, we can turn this
pass on by default. We have long since fixed the cause for the degradation.
Differential revision: https://reviews.llvm.org/D52431
llvm-svn: 375152
Summary:
In the long run we should come up with another mechanism for marking
call instructions as heap allocation sites, and remove this workaround.
For now, we've had two bug reports about this, so let's apply this
workaround. SLH (the other client of instruction labels) probably has
the same bug, but the solution there is more likely to be to mark the
call instruction as not duplicatable, which doesn't work for debug info.
Reviewers: akhuang
Subscribers: aprantl, hiraditya, aganea, chandlerc, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D69068
llvm-svn: 375137
This patch provides support for peudo ops including ADDIStocHA8, ADDIStocHA, LWZtocL,
LDtoc, LDtocL for AIX, lowering them from MIR to assembly.
Differential Revision: https://reviews.llvm.org/D68341
llvm-svn: 375113
Allow us to generate truncating masked store which take v4i32 and
v8i16 vectors and can store to v4i8, v4i16 and v8i8 and memory.
Removed support for unaligned masked stores.
Differential Revision: https://reviews.llvm.org/D68461
llvm-svn: 375108
Since GNU ar 2.31, the 't' operation prints member offsets beside file
names if the 'O' modifier is specified. 'O' is ignored for thin
archives.
Reviewed By: gbreynoo, ruiu
Differential Revision: https://reviews.llvm.org/D69087
llvm-svn: 375106
Exposed by D69041. If SHT_SYMTAB does not exist, ELFObjcopy.cpp:handleArgs will crash due
to a null pointer dereference.
for (const NewSymbolInfo &SI : Config.ELF->SymbolsToAdd) {
...
Obj.SymbolTable->addSymbol(
Fix this by creating .symtab and .strtab on demand in ELFBuilder<ELFT>::readSections,
if --add-symbol is specified.
Reviewed By: grimar
Differential Revision: https://reviews.llvm.org/D69093
llvm-svn: 375105
We can't normally stumble into that assertion because a tautological
*conditional* `br` in loop body is required, one that always
branches to loop latch. But that should have been always folded
to an unconditional branch before we get it.
But that is not guaranteed if the pass is run standalone.
So let's just promote the assertion into a proper check.
Fixes https://bugs.llvm.org/show_bug.cgi?id=43687
llvm-svn: 375100
Remove dead virtual functions from vtables with
replaceNonMetadataUsesWith, so that CGProfile metadata gets cleaned up
correctly.
Original commit message:
Currently, it is hard for the compiler to remove unused C++ virtual
functions, because they are all referenced from vtables, which are referenced
by constructors. This means that if the constructor is called from any live
code, then we keep every virtual function in the final link, even if there
are no call sites which can use it.
This patch allows unused virtual functions to be removed during LTO (and
regular compilation in limited circumstances) by using type metadata to match
virtual function call sites to the vtable slots they might load from. This
information can then be used in the global dead code elimination pass instead
of the references from vtables to virtual functions, to more accurately
determine which functions are reachable.
To make this transformation safe, I have changed clang's code-generation to
always load virtual function pointers using the llvm.type.checked.load
intrinsic, instead of regular load instructions. I originally tried writing
this using clang's existing code-generation, which uses the llvm.type.test
and llvm.assume intrinsics after doing a normal load. However, it is possible
for optimisations to obscure the relationship between the GEP, load and
llvm.type.test, causing GlobalDCE to fail to find virtual function call
sites.
The existing linkage and visibility types don't accurately describe the scope
in which a virtual call could be made which uses a given vtable. This is
wider than the visibility of the type itself, because a virtual function call
could be made using a more-visible base class. I've added a new
!vcall_visibility metadata type to represent this, described in
TypeMetadata.rst. The internalization pass and libLTO have been updated to
change this metadata when linking is performed.
This doesn't currently work with ThinLTO, because it needs to see every call
to llvm.type.checked.load in the linkage unit. It might be possible to
extend this optimisation to be able to use the ThinLTO summary, as was done
for devirtualization, but until then that combination is rejected in the
clang driver.
To test this, I've written a fuzzer which generates random C++ programs with
complex class inheritance graphs, and virtual functions called through object
and function pointers of different types. The programs are spread across
multiple translation units and DSOs to test the different visibility
restrictions.
I've also tried doing bootstrap builds of LLVM to test this. This isn't
ideal, because only classes in anonymous namespaces can be optimised with
-fvisibility=default, and some parts of LLVM (plugins and bugpoint) do not
work correctly with -fvisibility=hidden. However, there are only 12 test
failures when building with -fvisibility=hidden (and an unmodified compiler),
and this change does not cause any new failures for either value of
-fvisibility.
On the 7 C++ sub-benchmarks of SPEC2006, this gives a geomean code-size
reduction of ~6%, over a baseline compiled with "-O2 -flto
-fvisibility=hidden -fwhole-program-vtables". The best cases are reductions
of ~14% in 450.soplex and 483.xalancbmk, and there are no code size
increases.
I've also run this on a set of 8 mbed-os examples compiled for Armv7M, which
show a geomean size reduction of ~3%, again with no size increases.
I had hoped that this would have no effect on performance, which would allow
it to awlays be enabled (when using -fwhole-program-vtables). However, the
changes in clang to use the llvm.type.checked.load intrinsic are causing ~1%
performance regression in the C++ parts of SPEC2006. It should be possible to
recover some of this perf loss by teaching optimisations about the
llvm.type.checked.load intrinsic, which would make it worth turning this on
by default (though it's still dependent on -fwhole-program-vtables).
Differential revision: https://reviews.llvm.org/D63932
llvm-svn: 375094
Summary:
Currently when computing a GEP offset using the function EmitGEPOffset
for the following instruction
getelementptr inbounds i32, i32* %p, i64 %offs
we get
mul nuw i64 %offs, 4
Unfortunately we cannot assume that unsigned wrapping won't happen
here because %offs is allowed to be negative.
Making such assumptions can lead to miscompilations: see the new test
test24_neg_offs in InstCombine/icmp.ll. Without the patch InstCombine
would generate the following comparison:
icmp eq i64 %offs, 4611686018427387902; 0x3ffffffffffffffe
Whereas the correct value to compare with is -2.
This patch replaces the NUW flag with NSW in the multiplication
instructions generated by EmitGEPOffset and adjusts the test suite.
https://bugs.llvm.org/show_bug.cgi?id=42699
Reviewers: chandlerc, craig.topper, ostannard, lebedev.ri, spatel, efriedma, nlopes, aqjune
Reviewed By: lebedev.ri
Subscribers: reames, lebedev.ri, hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D68342
llvm-svn: 375089
This broke llvm-objdump in 32-bit builds, see e.g.
http://lab.llvm.org:8011/builders/clang-cmake-armv7-quick/builds/10925
> Summary:
> When listing the index in `llvm-objdump -h`, use a zero-based counter instead of the actual section index (e.g. shdr->sh_index for ELF).
>
> While this is effectively a noop for now (except one unit test for XCOFF), the index values will change in a future patch that filters certain sections out (e.g. symbol tables). See D68669 for more context. Note: the test case in `test/tools/llvm-objdump/X86/section-index.s` already covers the case of incrementing the section index counter when sections are skipped.
>
> Reviewers: grimar, jhenderson, espindola
>
> Reviewed By: grimar
>
> Subscribers: emaste, sbc100, arichardson, aheejin, arphaman, seiya, llvm-commits, MaskRay
>
> Tags: #llvm
>
> Differential Revision: https://reviews.llvm.org/D68848
llvm-svn: 375088
Summary:
This is a NFC change that removes the NFA->DFA construction and emission logic from DFAPacketizerEmitter and instead uses the generic DFAEmitter logic. This allows DFAPacketizer to use the Automaton class from Support and remove a bunch of logic there too.
After this patch, DFAPacketizer is mostly logic for grepping Itineraries and collecting functional units, with no state machine logic. This will allow us to modernize by removing the 16-functional-unit limit and supporting non-itinerary functional units. This is all for followup patches.
Subscribers: hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D68992
llvm-svn: 375086
Add generic DAG combine for extending masked loads.
Allow us to generate sext/zext masked loads which can access v4i8,
v8i8 and v4i16 memory to produce v4i32, v8i16 and v4i32 respectively.
Differential Revision: https://reviews.llvm.org/D68337
llvm-svn: 375085
Summary:
Each generated helper can be configured to generate an option that disables
rules in that helper. This can be used to bisect rulesets.
The disable bits are stored in a SparseVector as this is very cheap for the
common case where nothing is disabled. It gets more expensive the more rules
are disabled but you're generally doing that for debug purposes where
performance is less of a concern.
Depends on D68426
Reviewers: volkan, bogner
Reviewed By: volkan
Subscribers: hiraditya, Petar.Avramovic, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D68438
llvm-svn: 375067
Teach the combiner helper how to flatten concat_vectors of build_vectors
into a build_vector.
Add this combine as part of AArch64 pre-legalizer combiner.
Differential Revision: https://reviews.llvm.org/D69071
llvm-svn: 375066
After changing dsymutil to use libOption, we lost error reporting for
missing required arguments (input files). Additionally, we stopped
complaining about unknown arguments. This patch fixes both and adds a
test.
llvm-svn: 375044
r374772 changed Offset to be an int64_t but left NewOffset as an int.
Scale is unsigned, so in the calculation `Offset - NewOffset * Scale`,
`NewOffset * Scale` was promoted to unsigned and was then zero-extended
to 64 bits, leading to an incorrect computation which manifested as an
out-of-memory when building the Swift standard library for Android
aarch64. Promote NewOffset to int64_t to fix this, and promote
EmittableOffset as well, since its one user passes it to a function
which takes an int64_t anyway.
Test case based on a suggestion by Sander de Smalen!
Differential Revision: https://reviews.llvm.org/D69018
llvm-svn: 375043
The problem is that we can have two loop exits, 'a' and 'b', where 'a' and 'b' would exit at the same iteration, 'a' precedes 'b' along some path, and 'b' is predicated while 'a' is not. In this case (see the previously submitted test case), we causing the loop to exit through 'b' whereas it should have exited through 'a'.
This only applies to loop exits where the exit counts are not provably inequal, but that isn't as much of a restriction as it appears. If we could order the exit counts, we'd have already removed one of the two exits. In theory, we might be able to prove inequality w/o ordering, but I didn't really explore that piece. Instead, I went for the obvious restriction and ensured we didn't predicate exits following non-predicateable exits.
Credit goes to Evgeny Brevnov for figuring out the problematic case. Fuzzing probably also found it (failures seen), but due to some silly infrastructure problems I hadn't gotten to the results before Evgeny hand reduced it from a benchmark (he manually enabled the transform). Once this is fixed, I'll try to filter through the fuzzer failures to see if there's anything additional lurking.
Differential Revision https://reviews.llvm.org/D68956
llvm-svn: 375038
Summary: Also update the help modifier (h) so that it works as a modifier and not just as a standalone `h`. For example, `llvm-ar h` prints the help message, but `llvm-ar xh` currently prints `unknown option h`.
Reviewers: MaskRay, gbreynoo
Subscribers: llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D69007
llvm-svn: 375028
The 1st attempt at this modified the cost model in a bad way to avoid the vectorization,
but that caused problems for other users (the loop vectorizer) of the cost model.
I don't see an ideal solution to these 2 related, potentially large, perf regressions:
https://bugs.llvm.org/show_bug.cgi?id=42708https://bugs.llvm.org/show_bug.cgi?id=43146
We decided that load combining was unsuitable for IR because it could obscure other
optimizations in IR. So we removed the LoadCombiner pass and deferred to the backend.
Therefore, preventing SLP from destroying load combine opportunities requires that it
recognizes patterns that could be combined later, but not do the optimization itself (
it's not a vector combine anyway, so it's probably out-of-scope for SLP).
Here, we add a cost-independent bailout with a conservative pattern match for a
multi-instruction sequence that can probably be reduced later.
In the x86 tests shown (and discussed in more detail in the bug reports), SDAG combining
will produce a single instruction on these tests like:
movbe rax, qword ptr [rdi]
or:
mov rax, qword ptr [rdi]
Not some (half) vector monstrosity as we currently do using SLP:
vpmovzxbq ymm0, dword ptr [rdi + 1] # ymm0 = mem[0],zero,zero,..
vpsllvq ymm0, ymm0, ymmword ptr [rip + .LCPI0_0]
movzx eax, byte ptr [rdi]
movzx ecx, byte ptr [rdi + 5]
shl rcx, 40
movzx edx, byte ptr [rdi + 6]
shl rdx, 48
or rdx, rcx
movzx ecx, byte ptr [rdi + 7]
shl rcx, 56
or rcx, rdx
or rcx, rax
vextracti128 xmm1, ymm0, 1
vpor xmm0, xmm0, xmm1
vpshufd xmm1, xmm0, 78 # xmm1 = xmm0[2,3,0,1]
vpor xmm0, xmm0, xmm1
vmovq rax, xmm0
or rax, rcx
vzeroupper
ret
Differential Revision: https://reviews.llvm.org/D67841
llvm-svn: 375025
Using GNU diff, `--strip-trailing-cr` removes a `\r` appearing before
a `\n` at the end of a line. Without this patch, lit's internal diff
only removes `\r` if it appears as the last character. That seems
useless. This patch fixes that.
This patch also adds `--strip-trailing-cr` to some tests that fail on
Windows bots when D68664 is applied. Based on what I see in the bot
logs, I think the following is happening. In each test there, lit
diff is comparing a file with `\r\n` line endings to a file with `\n`
line endings. Without D68664, lit diff reads those files in text
mode, which in Windows causes `\r\n` to be replaced with `\n`.
However, with D68664, lit diff reads the files in binary mode instead
and thus reports that every line is different, just as GNU diff does
(at least under Ubuntu). Adding `--strip-trailing-cr` to those tests
restores the previous behavior while permitting the behavior of lit
diff to be more like GNU diff.
Reviewed By: rnk
Differential Revision: https://reviews.llvm.org/D68839
llvm-svn: 375020
Currently, BPF backend is doing truncation elimination. If one truncation
is performed on a value defined by narrow loads, then it could be redundant
given BPF loads zero extend the destination register implicitly.
When the definition of the truncated value is a merging value (PHI node)
that could come from different code paths, then checks need to be done on
all possible code paths.
Above described optimization was introduced as r306685, however it doesn't
work when there is back-edge, for example when loop is used inside BPF
code.
For example for the following code, a zero-extended value should be stored
into b[i], but the "and reg, 0xffff" is wrongly eliminated which then
generates corrupted data.
void cal1(unsigned short *a, unsigned long *b, unsigned int k)
{
unsigned short e;
e = *a;
for (unsigned int i = 0; i < k; i++) {
b[i] = e;
e = ~e;
}
}
The reason is r306685 was trying to do the PHI node checks inside isel
DAG2DAG phase, and the checks are done on MachineInstr. This is actually
wrong, because MachineInstr is being built during isel phase and the
associated information is not completed yet. A quick search shows none
target other than BPF is access MachineInstr info during isel phase.
For an PHI node, when you reached it during isel phase, it may have all
predecessors linked, but not successors. It seems successors are linked to
PHI node only when doing SelectionDAGISel::FinishBasicBlock and this
happens later than PreprocessISelDAG hook.
Previously, BPF program doesn't allow loop, there is probably the reason
why this bug was not exposed.
This patch therefore fixes the bug by the following approach:
- The existing truncation elimination code and the associated
"load_to_vreg_" records are removed.
- Instead, implement truncation elimination using MachineSSA pass, this
is where all information are built, and keep the pass together with other
similar peephole optimizations inside BPFMIPeephole.cpp. Redundant move
elimination logic is updated accordingly.
- Unit testcase included + no compilation errors for kernel BPF selftest.
Patch Review
===
Patch was sent to and reviewed by BPF community at:
https://lore.kernel.org/bpf
Reported-by: David Beckett <david.beckett@netronome.com>
Reviewed-by: Yonghong Song <yhs@fb.com>
Signed-off-by: Jiong Wang <jiong.wang@netronome.com>
llvm-svn: 375007
Summary:
This patch implements the `TargetInstrInfo::verifyInstruction` hook for RISC-V. Currently the hook verifies the machine instruction's immediate operands, to check if the immediates are within the expected bounds. Without the hook invalid immediates are not detected except when doing assembly parsing, so they are silently emitted (including being truncated when emitting object code).
The bounds information is specified in tablegen by using the `OperandType` definition, which sets the `MCOperandInfo`'s `OperandType` field. Several RISC-V-specific immediate operand types were created, which extend the `MCInstrDesc`'s `OperandType` `enum`.
To have the hook called with `llc` pass it the `-verify-machineinstrs` option. For Clang add the cmake build config `-DLLVM_ENABLE_EXPENSIVE_CHECKS=True`, or temporarily patch `TargetPassConfig::addVerifyPass`.
Review concerns:
- The patch adds immediate operand type checks that cover at least the base ISA. There are several other operand types for the C extension and one type for the F/D extensions that were left out of this initial patch because they introduced further design concerns that I felt were best evaluated separately.
- Invalid register classes (e.g. passing a GPR register where a GPRC is expected) are already caught, so were not included.
- This design makes the more abstract `MachineInstr` verification depend on MC layer definitions, which arguably is not the cleanest design, but is in line with how things are done in other parts of the target and LLVM in general.
- There is some duplication of logic already present in the `MCOperandPredicate`s. Since the `MachineInstr` and `MCInstr` notions of immediates are fundamentally different, this is currently necessary.
Reviewers: asb, lenary
Reviewed By: lenary
Subscribers: hiraditya, rbar, johnrusso, simoncook, apazos, sabuasal, niosHD, kito-cheng, shiva0217, jrtc27, MaskRay, zzheng, edward-jones, rogfer01, MartinMosbeck, brucehoult, the_o, rkruppe, PkmX, jocewei, psnobl, benna, Jim, s.egerton, pzheng, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D67397
llvm-svn: 375006
Summary:
Even though writelane doesn't have the same constraints as other valu
instructions it still can't violate the >1 SGPR operand constraint
Due to later register propagation (e.g. fixing up vgpr operands via
readfirstlane) changing writelane to only have a single SGPR is tricky.
This implementation puts a new check after SIFixSGPRCopies that prevents
multiple SGPRs being used in any writelane instructions.
The algorithm used is to check for trivial copy prop of suitable constants into
one of the SGPR operands and perform that if possible. If this isn't possible
put an explicit copy of Src1 SGPR into M0 and use that instead (this is
allowable for writelane as the constraint is for SGPR read-port and not
constant-bus access).
Reviewers: rampitec, tpr, arsenm, nhaehnle
Reviewed By: rampitec, arsenm, nhaehnle
Subscribers: arsenm, kzhuravl, jvesely, wdng, nhaehnle, mgorny, yaxunl, tpr, t-tye, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D51932
Change-Id: Ic7553fa57440f208d4dbc4794fc24345d7e0e9ea
llvm-svn: 375004
When on windows gnu-ar treats member names as case insensitive. This
commit implements the same behaviour.
Differential Revision: https://reviews.llvm.org/D68033
llvm-svn: 375002
Summary:
This is something of a workaround to avoid a crash later on in type
legalizer (WidenVectorResult()).
Also added some f16 tests, including a non-working v3f16 case with
a FIXME.
Reviewers: arsenm, tpr, nhaehnle
Reviewed By: arsenm
Subscribers: kzhuravl, jvesely, wdng, nhaehnle, yaxunl, dstuttard, tpr, t-tye, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D68865
llvm-svn: 374993
Summary:
Currently Thumb2InstrInfo.cpp uses a register class which is
auto-generated by tablegen. Such approach is fragile because
auto-generated classes might change when other register classes are
added. For example, before https://reviews.llvm.org/D62667
we were using GPRPair_with_gsub_1_in_rGPRRegClass, but had to
change it to GPRPair_with_gsub_1_in_GPRwithAPSRnospRegClass
because the former class stopped being generated (this did not change
the functionality though).
This patch adds a register class consisting of even-odd GPR register
pairs from (R0, R1) to (R10, R11), which excludes (R12, SP) and uses
it in Thumb2InstrInfo.cpp instead of
GPRPair_with_gsub_1_in_GPRwithAPSRnospRegClass.
Reviewers: ostannard, simon_tatham, dmgreen, efriedma
Reviewed By: simon_tatham
Subscribers: kristof.beyls, hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D69026
llvm-svn: 374990
Summary:
Extend the SI Load/Store optimizer to merge MIMG load instructions. Handle
different flavours of image_load and image_sample instructions.
When the instructions of the same subclass differ only in dmask, merge
them and update dmask accordingly.
Reviewers: nhaehnle
Reviewed By: nhaehnle
Subscribers: arsenm, kzhuravl, jvesely, wdng, nhaehnle, yaxunl, dstuttard, tpr, t-tye, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D64911
llvm-svn: 374984
Instead of inserting everything after the 'root' of the reduction,
insert all instructions as close to their operands as possible. This
can help reduce register pressure.
Differential Revision: https://reviews.llvm.org/D67392
llvm-svn: 374981
This adds the initial plumbing to support optimisation remarks in
the IR hardware-loop pass.
I have left a todo in a comment where we can improve the reporting,
and will iterate on that now that we have this initial support in.
Differential Revision: https://reviews.llvm.org/D68579
llvm-svn: 374980
PromoteTargetBoolean calls getSetccResultType to get the return
type. But we were passing it the setcc result type rather than the
setcc input type. This causes an issue on X86 with avx512vl where
the setcc result type for vXf16 vectors is vXi16 while the
result type for vXi16 vectors is vXi1.
There's really no guarantee that getSetccResultType is the type
we need here. So now we just grab the extend type from
getExtendForContent and extend to the original result VT of the
node we're splitting.
llvm-svn: 374970
Since r374600 clang emits base address selection entries. Currently
dsymutil does not support these entries and incorrectly interprets them
as location list entries.
This patch adds support for base address selection entries in dsymutil
and makes sure they are relocated correctly.
Thanks to Dave for coming up with the test case!
Differential revision: https://reviews.llvm.org/D69005
llvm-svn: 374957
InProcessMemoryManager used to make separate memory allocation calls for each
permission level (RW, RX, RO), which could lead to target-out-of-range errors
if data and code were placed too far apart (this was the source of failures in
the JITLink/AArch64 testcase when it was first landed).
This patch updates InProcessMemoryManager to allocate a single slab which is
subdivided between text and data. This should guarantee that accesses remain
in-range provided that individual object files do not exceed 1Mb in size.
This patch also re-enables the JITLink/AArch64 testcase.
llvm-svn: 374948
Summary:
Two conditions could lead to infinite loops when processing PHI nodes in
SIFixSGPRCopies.
The first condition involves a REG_SEQUENCE that uses registers defined by both
a PHI and a COPY.
The second condition arises when a physical register is copied to a virtual
register which is then used in a PHI node. If the same virtual register is
copied to the same physical register, the result is an endless loop.
%0:sgpr_64 = COPY $sgpr0_sgpr1
%2 = PHI %0, %bb.0, %1, %bb.1
$sgpr0_sgpr1 = COPY %0
Reviewers: alex-t, rampitec, arsenm
Reviewed By: rampitec
Subscribers: kzhuravl, jvesely, wdng, nhaehnle, yaxunl, dstuttard, tpr, t-tye, hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D68970
llvm-svn: 374944
SUMMARY:
in the xcoff, if the number of relocation entries or line number entries is
overflow(large than or equal 65535) , there will be overflow section for it.
The interpret of overflow section is different with generic section header,
the patch implement parsing the overflow section.
Reviewers: hubert.reinterpretcast,sfertile,jasonliu
Subscribers: rupprecht, seiya
Differential Revision: https://reviews.llvm.org/D68575
llvm-svn: 374941
Summary:
Renames `ExprType` to the more apt `BlockType` and adds a variant for
multivalue blocks. Currently non-void blocks are only generated at the
end of functions where the block return type needs to agree with the
function return type, and that remains true for multivalue
blocks. That invariant means that the actual signature does not need
to be stored in the block signature `MachineOperand` because it can be
inferred by `WebAssemblyMCInstLower` from the return type of the
parent function. `WebAssemblyMCInstLower` continues to lower block
signature operands to immediates when possible but lowers multivalue
signatures to function type symbols. The AsmParser and Disassembler
are updated to handle multivalue block types as well.
Reviewers: aheejin, dschuff, aardappel
Subscribers: sbc100, jgravelle-google, hiraditya, sunfish, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D68889
llvm-svn: 374933
Summary:
When listing the index in `llvm-objdump -h`, use a zero-based counter instead of the actual section index (e.g. shdr->sh_index for ELF).
While this is effectively a noop for now (except one unit test for XCOFF), the index values will change in a future patch that filters certain sections out (e.g. symbol tables). See D68669 for more context. Note: the test case in `test/tools/llvm-objdump/X86/section-index.s` already covers the case of incrementing the section index counter when sections are skipped.
Reviewers: grimar, jhenderson, espindola
Reviewed By: grimar
Subscribers: emaste, sbc100, arichardson, aheejin, arphaman, seiya, llvm-commits, MaskRay
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D68848
llvm-svn: 374931
I removed this test to unblock the ARM bots while looking into failures
(r374915), and am reinstating it now with a fix.
I believe the problem was that counter ptr address I used,
'\0\0\6\0\1\0\0\1', set the high bits of the pointer, not the low bits
like I wanted. On x86_64 this superficially looks like it tests r370826,
but it doesn't, as it would have been caught before r370826. However, on
ARM (or, 32-bit hosts more generally), I suspect the high bits were
cleared, and you get a 'valid' profile.
I verified that setting the *low* bits of the pointer does trigger the
new condition:
-// Note: The CounterPtr here is off-by-one. This should trigger a malformed profile error.
-RUN: printf '\0\0\6\0\1\0\0\1' >> %t.profraw
+// Note: The CounterPtr here is off-by-one.
+//
+// Octal '\11' is 9 in decimal: this should push CounterOffset to 1. As there are two counters,
+// the profile reader should error out.
+RUN: printf '\11\0\6\0\1\0\0\0' >> %t.profraw
This reverts commit c7cf5b3e4b918c9769fd760f28485b8d943ed968.
llvm-svn: 374927
This is remaining part of rG41ca91f2995b: [AIX][XCOFF] Output XCOFF
object text section header and symbol entry for rogram code.
SUMMARY:
Original form of this patch is provided by Stefan Pintillie.
1. The patch try to output program code section header , symbol entry for
program code (PR) and Instruction into the raw text section.
2. The patch include how to alignment and layout the CSection in the text
section.
3. The patch also reorganize the code , put some codes into a function.
(XCOFFObjectWriter::writeSymbolTableEntryForControlSection)
Additional: We can not add raw data of text section test in the patch, If want
to output raw text section data,it need a function description patch first.
Reviewers: hubert.reinterpretcast, sfertile, jasonliu, xingxue.
Subscribers: wuzish, nemanjai, hiraditya, MaskRay, jsjji.
Differential Revision: https://reviews.llvm.org/D66969
llvm-svn: 374923
Nikita Popov