Summary:
1. Enable infrastructure of AVX512_BF16, which is supported for BFLOAT16 in Cooper Lake;
2. Enable VCVTNE2PS2BF16, VCVTNEPS2BF16 and DPBF16PS instructions, which are Vector Neural Network Instructions supporting BFLOAT16 inputs and conversion instructions from IEEE single precision.
VCVTNE2PS2BF16: Convert Two Packed Single Data to One Packed BF16 Data.
VCVTNEPS2BF16: Convert Packed Single Data to Packed BF16 Data.
VDPBF16PS: Dot Product of BF16 Pairs Accumulated into Packed Single Precision.
For more details about BF16 isa, please refer to the latest ISE document: https://software.intel.com/en-us/download/intel-architecture-instruction-set-extensions-programming-reference
Author: LiuTianle
Reviewers: craig.topper, smaslov, LuoYuanke, wxiao3, annita.zhang, RKSimon, spatel
Reviewed By: craig.topper
Subscribers: kristina, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D60550
llvm-svn: 360017
Summary:
Prior to DWARF v5, a directory index of 0 represents DW_AT_comp_dir.
In DWARF v5, the index starts with 0 and Entry.DirIdx is the index into
Prologue.IncludeDirectories.
Reviewed By: labath
Differential Revision: https://reviews.llvm.org/D61253
llvm-svn: 360015
Optimization pass lib/Transforms/IPO/GlobalOpt.cpp needs to insert
DW_OP_deref_size instead of DW_OP_deref to be compatible with big-endian
targets for same reasons as in D59687.
Differential Revision: https://reviews.llvm.org/D60611
llvm-svn: 360013
In C a function declaration with an empty argument list isn't a real
prototype, it will allow calling the function with any number of
arguments. It will also cause warnings when used in C code compiled with
'-Wstrict-prototypes'
Reviewed By: whitequark
Differential Revision: https://reviews.llvm.org/D61568
llvm-svn: 360012
Based on PR41748, not all cases are handled in this function.
llvm_unreachable is treated as an optimization hint than can prune code paths
in a release build. This causes weird behavior when PR41748 is encountered on a
release build. It appears to generate an fp_round instruction from the floating
point code.
Making this a report_fatal_error prevents incorrect optimization of the code
and will instead generate a message to file a bug report.
llvm-svn: 360008
When builing the hermetic static library, the compiler switch
-fvisibility-global-new-delete-hidden is necessary to get the new and
delete operator definitions made correctly. However, when those
definitions are not included in the library, then this switch does harm.
With lld (though not all linkers) setting STV_HIDDEN on SHN_UNDEF
symbols makes it an error to leave them undefined or defined via dynamic
linking that should generate PLTs for -shared linking (lld makes this a
hard error even without -z defs). Though leaving the symbols undefined
would usually work in practice if the linker were to allow it (and the
user didn't pass -z defs), this actually indicates a real problem that
could bite some target configurations more subtly at runtime. For
example, x86-32 ELF -fpic code generation uses hidden visibility on
declarations in the caller's scope as a signal that the call will never
be resolved to a PLT entry and so doesn't have to meet the special ABI
requirements for PLT calls (setting %ebx). Since these functions might
actually be resolved to PLT entries at link time (we don't know what the
user is linking in when the hermetic library doesn't provide all the
symbols itself), it's not safe for the compiler to treat their
declarations at call sites as having hidden visibility.
Differential Revision: https://reviews.llvm.org/D61572
llvm-svn: 360004
When builing the hermetic static library, the compiler switch
-fvisibility-global-new-delete-hidden is necessary to get the new and
delete operator definitions made correctly. However, when those
definitions are not included in the library, then this switch does harm.
With lld (though not all linkers) setting STV_HIDDEN on SHN_UNDEF
symbols makes it an error to leave them undefined or defined via dynamic
linking that should generate PLTs for -shared linking (lld makes this a
hard error even without -z defs). Though leaving the symbols undefined
would usually work in practice if the linker were to allow it (and the
user didn't pass -z defs), this actually indicates a real problem that
could bite some target configurations more subtly at runtime. For
example, x86-32 ELF -fpic code generation uses hidden visibility on
declarations in the caller's scope as a signal that the call will never
be resolved to a PLT entry and so doesn't have to meet the special ABI
requirements for PLT calls (setting %ebx). Since these functions might
actually be resolved to PLT entries at link time (we don't know what the
user is linking in when the hermetic library doesn't provide all the
symbols itself), it's not safe for the compiler to treat their
declarations at call sites as having hidden visibility.
Differential Revision: https://reviews.llvm.org/D61571
llvm-svn: 360003
Matches what we do in other getValueType functions and fixes a null dereference warning in scan-build.
Also cleans up the rest of the function - use auto and standardize the variable names.
llvm-svn: 360000
Thus it does not assume that the old basic block is the basic block
for which we are looking at successors.
Not reviewed, but seems rather trivial, in line with the rest of
previous few patches.
llvm-svn: 359997
Summary:
It is a common thing to loop over every `PHINode` in some `BasicBlock`
and change old `BasicBlock` incoming block to a new `BasicBlock` incoming block.
`replaceSuccessorsPhiUsesWith()` already had code to do that,
it just wasn't a function.
So outline it into a new function, and use it.
Reviewers: chandlerc, craig.topper, spatel, danielcdh
Reviewed By: craig.topper
Subscribers: llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D61013
llvm-svn: 359996
Summary:
There is `PHINode::getBasicBlockIndex()`, `PHINode::setIncomingBlock()`
and `PHINode::getNumOperands()`, but no function to replace every
specified `BasicBlock*` predecessor with some other specified `BasicBlock*`.
Clearly, there are a lot of places that could use that functionality.
Reviewers: chandlerc, craig.topper, spatel, danielcdh
Reviewed By: craig.topper
Subscribers: llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D61011
llvm-svn: 359995
Summary:
There is `Instruction::getNumSuccessors()`, `Instruction::getSuccessor()`
and `Instruction::setSuccessor()`, but no function to replace every
specified `BasicBlock*` successor with some other specified `BasicBlock*`.
I've found one place where it should clearly be used.
Reviewers: chandlerc, craig.topper, spatel, danielcdh
Reviewed By: craig.topper
Subscribers: llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D61010
llvm-svn: 359994
Summary:
If `deleteDeadLoop()` is called on such a loop, that has "bad" exit block,
one that e.g. has no terminator instruction, the `DIBuilder::insertDbgValueIntrinsic()`
will be told to insert the Dbg Value Intrinsic after `nullptr`
(since there is no first non-PHI instruction), which will cause it to not insert
those instructions into any basic block. The instructions will be parent-less,
and IR verifier will complain. It is rather obvious to track down the root cause
when that happens, so let's just assert it never happens.
Reviewers: sanjoy, davide, vsk
Reviewed By: vsk
Subscribers: llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D61008
llvm-svn: 359993
Summary: This check appears to be a leftover from when add/sub/mul could be either integer or fp. The NSW/NUW flags are only set for add/sub/mul/shl earlier. And we check that those operations only have integer types just below this. So it seems unnecessary to explicitly error for NUW/NSW being used on a add/sub/mul that have the wrong type that would later error for that.
Reviewers: spatel, dblaikie, jyknight, arsenm
Reviewed By: spatel
Subscribers: wdng, llvm-commits, hiraditya
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D61562
llvm-svn: 359987
Summary:
These methods previously took a 0, 1, or 2 to indicate what types were allowed, but the 0 encoding which meant both fp and integer types has been unused for years. Its leftover from when add/sub/mul used to be shared between int and fp
Simplify it by changing it to just a bool to distinquish int and fp.
Reviewers: spatel, dblaikie, jyknight, arsenm
Reviewed By: spatel
Subscribers: wdng, hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D61561
llvm-svn: 359986
Summary:
Remove duplicate checks that both operands have the same type. This is checked
before the switch.
Use 'integer' or 'floating-point' instead of 'arithmetic' type. I think this
might be a leftover to the days when floating point and integer operations
shared the same opcodes.
Reviewers: spatel, RKSimon, dblaikie
Reviewed By: RKSimon
Subscribers: hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D61558
llvm-svn: 359985
This is a subset of the original commit from rL359879
which was reverted because it could crash when using the 'RemovedInstructions'
structure that enables delayed deletion of dead instructions. The motivating
compile-time win does not require that change though. We should get most of
that win from this change alone.
Using/updating a dominator tree to match math overflow patterns may be very
expensive in compile-time (because of the way CGP uses a DT), so just handle
the single-block case.
See post-commit thread for rL354298 for more details:
http://lists.llvm.org/pipermail/llvm-commits/Week-of-Mon-20190422/646276.html
Differential Revision: https://reviews.llvm.org/D61075
llvm-svn: 359969
These operations were already used in eh-frame registration, and are likely to
be used in other runtime registrations, so this commit moves them into a header
where they can be re-used.
llvm-svn: 359950
This saves us some unnecessary copies.
If the inputs to a G_SELECT are floating point, we should use fcsel rather than
csel.
Changes here are...
- Teach selectCopy about s1-to-s1 copies across register banks.
- AArch64RegisterBankInfo about G_SELECT in general.
- Teach the instruction selector about the FCSEL instructions.
Also add two tests:
- select-select.mir to show that we get the expected FCSEL
- regbank-select.mir (unfortunately named) to show the register banks on
G_SELECT are properly preserved
And update fast-isel-select.ll to show that we do the same thing as other
instruction selectors in these cases.
llvm-svn: 359940
Not all versions of git support git rev-parse --git-common-dir. Rather than erorr or print any kind of
useful error, they just print back '--git-common-dir' instead of a directory. The git-llvm script
ends up taking this '--git-common-dir' as a diretory name to use.
Not sure exactly what happens after that, but the end result is that the 'git llvm push' ends up
looking like it pushed your commits, but really did nothing.
This patch makes the script detect the bogus directory name for --git-common-dir and falls back to using --git-dir instead.
llvm-svn: 359939
Summary:
This change enables `cl::Grouping` for short options --
options with names of a single character. This is consistent with GNU
getopt behavior.
Reviewers: rnk, MaskRay
Reviewed By: MaskRay
Subscribers: thopre, cfe-commits, MaskRay, rupprecht, hiraditya, llvm-commits
Tags: #llvm, #clang
Differential Revision: https://reviews.llvm.org/D61270
llvm-svn: 359917
