The special case here is really G_UNMERGE_VALUES, not G_EXTRACT. The
other opcodes can hardcode index 1 like G_EXTRACT.

Summary:
This patch adds --source flag to indicate the source file. Then it tries to find insert
points in the source file and insert corresponding checks at those places.

Example output from Tensorflow XLA:

// -----

// CHECK-LABEL:   func @main.3(
// CHECK-SAME:                 %[[VAL_0:.*]]: memref<2x2xf32> {xla_lhlo.params = 0 : index},
// CHECK-SAME:                 %[[VAL_1:.*]]: memref<16xi8> {xla_lhlo.alloc = 0 : index, xla_lhlo.liveout = true}) {
// CHECK:           %[[VAL_2:.*]] = constant 0 : index
// CHECK:           %[[VAL_3:.*]] = constant 0 : index
// CHECK:           %[[VAL_4:.*]] = std.view %[[VAL_1]]{{\[}}%[[VAL_3]]][] : memref<16xi8> to memref<2x2xf32>
// CHECK:           "xla_lhlo.tanh"(%[[VAL_0]], %[[VAL_4]]) : (memref<2x2xf32>, memref<2x2xf32>) -> ()
// CHECK:           return
// CHECK:         }
func @main(%value0: tensor<2x2xf32>) -> tensor<2x2xf32> {
  %res = "xla_hlo.tanh"(%value0) : (tensor<2x2xf32>) -> tensor<2x2xf32>
  return %res : tensor<2x2xf32>
}

Differential Revision: https://reviews.llvm.org/D81903

UBSAN complains when tblgen performs SHL of a negative
value.

Differential Revision: https://reviews.llvm.org/D81952

Summary:
delete(void*, unsigned int, align_val_t)
delete(void*, unsigned long, align_val_t)
delete[](void*, unsigned int, align_val_t)
delete[](void*, unsigned long, align_val_t)

Differential Revision: https://reviews.llvm.org/D81853

Summary: we use the alias attribute, similar to what is done for ELF.

Reviewers: ZarkoCA, jasonliu, hubert.reinterpretcast, sfertile

Reviewed By: jasonliu

Subscribers: dberris, aheejin, mstorsjo, #sanitizers

Tags: #sanitizers

Differential Revision: https://reviews.llvm.org/D81120

Executing commands below will get you bombarded by a wall of Python
command prompts (>>> ).

$ echo 'foo' | ./bin/lldb -o script
$ cat /tmp/script
script
print("foo")
$ lldb --source /tmp/script

The issue is that our custom input reader doesn't handle EOF. According
to the Python documentation, file.readline always includes a trailing
newline character unless the file ends with an incomplete line. An empty
string signals EOF. This patch raises an EOFError when that happens.

[1] https://docs.python.org/2/library/stdtypes.html#file.readline

Differential revision: https://reviews.llvm.org/D81898

Generalize scalarization (recently enhanced with D80885)
to allow compares as well as binops.
Similar to binops, we are avoiding scalarization of a loaded
value because that could avoid a register transfer in codegen.
This requires 1 extra predicate that I am aware of: we do not
want to scalarize the condition value of a vector select. That
might also invert a transform that we do in instcombine that
prefers a vector condition operand for a vector select.

I think this is the final step in solving PR37463:
https://bugs.llvm.org/show_bug.cgi?id=37463

Differential Revision: https://reviews.llvm.org/D81661

The Standard documents the signature of std::advance as

    template <class Iter, class Distance>
    constexpr void advance(Iter& i, Distance n);

Furthermore, it does not appear to put any restriction on what the type
of Distance should be. While it is understood that it should usually
be std::iterator_traits::difference_type, I couldn't find any wording
that mandates that. Similarly, I couldn't find wording that forces the
distance to be a signed type.

This patch changes std::advance to accept any type in the second argument,
which appears to be what the Standard mandates. We then coerce it to the
iterator's difference type, but that's an implementation detail.

Differential Revision: https://reviews.llvm.org/D81425

the range start is already set

The range start could be set already in some invalid cases. Fixes
PR46336.

When selecting 32 b -> 64 b G_ZEXTs, we don't have to always emit the extend.

If the instruction feeding into the G_ZEXT implicitly zero extends the high
half of the register, we can just emit a SUBREG_TO_REG instead.

Differential Revision: https://reviews.llvm.org/D81897

The type test use this method to store the golden output. This currently
fails if the reproducer directory hasn't yet been created.

Differential Revision: https://reviews.llvm.org/D81933

https://bugs.llvm.org/show_bug.cgi?id=46253

This is an obvious hack because realloc isn't any more affected than other
functions modeled by MallocChecker (or any user of CallDescription really),
but the nice solution will take some time to implement.

Differential Revision: https://reviews.llvm.org/D81745

Compiling assembly files when newlines are reduced to line markers within a `.macro` context will generate wrong information in `.debug_line` section.
This patch fixes this issue by evaluating line markers within the macro scope but not when they are used and evaluated.

Reviewed By: probinson

Differential Revision: https://reviews.llvm.org/D80381

This is currently different from the IR rules.

Adds the callbacks for ordered with source/sink dependencies.

The test for task dependencies changed, because callbach.h now actually prints
the passed dependencies and the test also checks for the address.

Reviewed by: hbae

Differential Revision: https://reviews.llvm.org/D81807

Summary:
This revision replaces MatmulOp, now that DRR rules have been dropped.
This revision also fixes minor parsing bugs and a plugs a few holes to get e2e paths working (e.g. library call emission).

During the replacement the i32 version had to be dropped because only the EDSC operators +, *, etc support type inference.

Deciding on a type-polymorphic behavior, and implementing it, is left for future work.

Reviewers: aartbik

Subscribers: mehdi_amini, rriddle, jpienaar, shauheen, antiagainst, arpith-jacob, mgester, lucyrfox, aartbik, liufengdb, stephenneuendorffer, Joonsoo, grosul1, frgossen, Kayjukh, jurahul, msifontes

Tags: #mlir

Differential Revision: https://reviews.llvm.org/D81935

Some tests were missing alignment info. Subsequent changes properly
preserve the set alignment. Set it properly beforehand, to avoid
unnecessary test changes.

It also updates cases where an alignment of 16 was specified, instead of
the vector element type alignment.

gcc 9.1/9.2 has a bug (https://gcc.gnu.org/bugzilla/show_bug.cgi?id=90538)
which leads to an incorrect error when expanding parameter packs multiple
times in a lambda. Inlining this lambda to work around this issue.

Reviewed By: rriddle, CarolineConcatto

Differential Revision: https://reviews.llvm.org/D81828

This patch upstreams support for BFloat Matrix Multiplication Intrinsics
and Code Generation from __bf16 to AArch64. This includes IR intrinsics. Unittests are
provided as needed. AArch32 Intrinsics + CodeGen will come after this
patch.

This patch is part of a series implementing the Bfloat16 extension of
the
Armv8.6-a architecture, as detailed here:

https://community.arm.com/developer/ip-products/processors/b/processors-ip-blog/posts/arm-architecture-developments-armv8-6-a

The bfloat type, and its properties are specified in the Arm
Architecture
Reference Manual:

https://developer.arm.com/docs/ddi0487/latest/arm-architecture-reference-manual-armv8-for-armv8-a-architecture-profile

The following people contributed to this patch:

Luke Geeson
 - Momchil Velikov
 - Mikhail Maltsev
 - Luke Cheeseman

Reviewers: SjoerdMeijer, t.p.northover, sdesmalen, labrinea, miyuki,
stuij

Reviewed By: miyuki, stuij

Subscribers: kristof.beyls, hiraditya, danielkiss, cfe-commits,
llvm-commits, miyuki, chill, pbarrio, stuij

Tags: #clang, #llvm

Differential Revision: https://reviews.llvm.org/D80752

Change-Id: I174f0fd0f600d04e3799b06a7da88973c6c0703f

This patch upstreams support for ld / st variants of BFloat intrinsics
in from __bf16 to AArch64. This includes IR intrinsics. Unittests are
provided as needed.

This patch is part of a series implementing the Bfloat16 extension of
the
Armv8.6-a architecture, as detailed here:

https://community.arm.com/developer/ip-products/processors/b/processors-ip-blog/posts/arm-architecture-developments-armv8-6-a

The bfloat type, and its properties are specified in the Arm
Architecture
Reference Manual:

https://developer.arm.com/docs/ddi0487/latest/arm-architecture-reference-manual-armv8-for-armv8-a-architecture-profile

The following people contributed to this patch:

 - Luke Geeson
 - Momchil Velikov
 - Luke Cheeseman

Reviewers: fpetrogalli, SjoerdMeijer, sdesmalen, t.p.northover, stuij

Reviewed By: stuij

Subscribers: arsenm, pratlucas, simon_tatham, labrinea, kristof.beyls,
hiraditya, danielkiss, cfe-commits, llvm-commits, pbarrio, stuij

Tags: #clang, #llvm

Differential Revision: https://reviews.llvm.org/D80716

Change-Id: I22e1dca2a8a9ec25d1e4f4b200cb50ea493d2575

Reviewers: yaxunl, arsenm

Subscribers: kzhuravl, jvesely, wdng, nhaehnle, dstuttard, tpr, t-tye, kerbowa, cfe-commits

Tags: #clang

Differential Revision: https://reviews.llvm.org/D81849

Summary:
EOF macro token coming from a PCH file on macOS while marked as literal,
doesn't contain any literal data.  This causes crash on every project
using PCHs.

This commit doesn't resolve the problem with PCH (maybe it was
designed like this for a purpose) or with `tryExpandAsInteger`, but
rather simply shoots off a crash itself.

Differential Revision: https://reviews.llvm.org/D81916

Ignore template instantiations in the matchers, Addresses [[ https://bugs.llvm.org/show_bug.cgi?id=46226 | readability-simplify-boolean-expr false-positive for bool from template. ]]

Reviewed By: aaron.ballman, lebedev.ri

Differential Revision: https://reviews.llvm.org/D81336

Specification (https://refspecs.linuxbase.org/LSB_5.0.0/LSB-Core-generic/LSB-Core-generic/ehframechpt.html#AEN1349)
says that the value of Version field for .eh_frame should be 1.

Though we accept other values and might perform an attempt to read
it as a .debug_frame because of that, what is wrong.

This patch adds a version check.

Differential revision: https://reviews.llvm.org/D81469

Previously we only printed a symbol value when it has a non-empty name
or non-zero value.

This patch changes the behavior. Now we only omit a symbols value when
a relocation does not reference a symbol (i.e. symbol index == 0).

Seems it is what GNU readelf does, looking on its output.

Differential revision: https://reviews.llvm.org/D81842

Reviewers: jdoerfert

Subscribers: yaxunl, guansong, mehdi_amini, rriddle, jpienaar, shauheen, antiagainst, nicolasvasilache, arpith-jacob, mgester, lucyrfox, aartbik, liufengdb, stephenneuendorffer, Joonsoo, grosul1, frgossen, Kayjukh, jurahul, sstefan1, msifontes

Tags: #mlir

Differential Revision: https://reviews.llvm.org/D81264

This reverts commit 90c50cad.

Currently, llvm-readelf crashes when there is a STT_SECTION symbol for the null section
and this symbol is used in a relocation.

Differential revision: https://reviews.llvm.org/D81840

An instruction like this will need to allocate some stack space for the
last parameter:

  %x = call addrspace(1) i16 @bar(i64 undef, i64 undef, i16 undef, i16 0)

This worked fine when passing an actual value (in this case 0). However,
when passing undef, no value was pushed to the stack and therefore no
push instructions were created. This caused an unbalanced stack leading
to interesting results.

This commit fixes that by replacing the push logic with a regular stack
adjustment and stack-relative load/stores. This is less efficient but at
least it correctly compiles the code.

I can think of a few improvements in the future:

  * The stack should have been adjusted in the function prologue when
    there are no allocas in the function.
  * Many (if not most) stack adjustments can be replaced by
    pushing/popping the values directly. Exactly like the previous code
    attempted but didn't do correctly.
  * Small stack adjustments can be done more efficiently with a few
    push/pop instructions (pushing/popping bogus values), both for code
    size and for speed.

All in all, as long as there are no allocas in the function I think that
it is almost always more efficient to emit regular push/pop
instructions. This is however left for future optimizations.

Differential Revision: https://reviews.llvm.org/D78581

This patch fixes a bug in stack save/restore code. Because the frame
pointer was saved/restored manually (not by marking it as clobbered) the
StackSize variable was not updated accordingly. Most code still worked,
but code that tried to load a parameter passed on the stack did not.

This commit fixes this by marking the frame pointer as a
callee-clobbered register. This will let it be saved without any effort
in prolog/epilog code and will make sure the correct address is
calculated for loading parameters that are passed on the stack.

This approach is used by most other targets (such as X86, AArch64 and
RISC-V).

Differential Revision: https://reviews.llvm.org/D78579

Summary: Add runtime support, adjust the tests and enable LSan.

Reviewers: vitalybuka, eugenis, uweigand, jonpa

Reviewed By: uweigand

Subscribers: mgorny, cfe-commits, #sanitizers

Tags: #clang, #sanitizers

Differential Revision: https://reviews.llvm.org/D78644