Summary:
Terminology introduced by [[#]] blocks is confusing and does not
integrate well with existing terminology.
First, variables referred by [[]] blocks are called "pattern variables"
while the text a CHECK directive needs to match is called a "CHECK
pattern". This is inconsistent with variables in [[#]] blocks since
[[#]] blocks are also found in CHECK pattern yet those variables are
called "numeric variable".
Second, the replacing of both [[]] and [[#]] blocks by the value of the
variable or expression they contain is represented by a
FileCheckPatternSubstitution class. The naming refers to being a
substitution in a CHECK pattern but could be wrongly understood as being
a substitution of a pattern variable.
Third and lastly, comments use "numeric expression" to refer both to the
[[#]] blocks as well as to the numeric expressions these blocks contain
which get evaluated at match time.
This patch solves these confusions by
- calling variables in [[]] and [[#]] blocks as string and numeric
variables respectively;
- referring to [[]] and [[#]] as substitution *blocks*, with the former
being a string substitution block and the latter a numeric
substitution block;
- calling [[]] and [[#]] blocks to be replaced by the value of a
variable or expression they contain a substitution (as opposed to
definition when these blocks are used to defined a variable), with the
former being a string substitution and the latter a numeric
substitution;
- renaming the FileCheckPatternSubstitution as a FileCheckSubstitution
class with FileCheckStringSubstitution and
FileCheckNumericSubstitution subclasses;
- restricting the use of "numeric expression" to refer to the expression
that is evaluated in a numeric substitution.
While numeric substitution blocks only support numeric substitutions of
numeric expressions at the moment there are plans to augment numeric
substitution blocks to support numeric definitions as well as both a
numeric definition and numeric substitution in the same numeric
substitution block.
Chris Bieneman [Wed, 22 May 2019 21:42:06 +0000 (21:42 +0000)]
[Runtimes] If LLVM_INCLUDE_TESTS=On depend on gtest
Summary: If we are building the tests for the runtimes we should make them depend on gtest so that gtest is built and ready before we run any of the check-* targets.
Matt Arsenault [Wed, 22 May 2019 21:28:20 +0000 (21:28 +0000)]
TableGen: Handle nontrivial foreach range bounds
This allows using anything that isn't a literal integer as the bounds
for a foreach. Some of the diagnostics aren't perfect, but nobody ever
accused tablegen of having good errors. For example, the existing
wording suggests a bitrange is valid, but as far as I can tell this
has never worked.
Petr Hosek [Wed, 22 May 2019 21:08:33 +0000 (21:08 +0000)]
[runtimes] Move libunwind, libc++abi and libc++ to lib/$target/c++ and include/c++
This change is a consequence of the discussion in "RFC: Place libs in
Clang-dedicated directories", specifically the suggestion that
libunwind, libc++abi and libc++ shouldn't be using Clang resource
directory. Tools like clangd make this assumption, but this is
currently not true for the LLVM_ENABLE_PER_TARGET_RUNTIME_DIR build.
This change addresses that by moving the output of these libraries to
lib/$target/c++ and include/c++ directories, leaving resource directory
only for compiler-rt runtimes and Clang builtin headers.
Craig Topper [Wed, 22 May 2019 21:00:18 +0000 (21:00 +0000)]
[X86] Explcitly disable VEXTRACT instruction matching for an immediate of 0. Remove a bunch of isel patterns that become unnecessary.
We effectively had a second set of isel patterns that tried to use a
regular store instruction and an extract_subreg instruction. Or a masked move
and an extract_subreg. These patterns were intended to override the
matching of VEXTRACT instructions by taking advantage of the priority
of the explicit immediate 0 for the index.
This patch instaed just disables the immediate 0 matchin the VEXTRACT
patterns. This each of the component pieces of the larger patterns will
match by themselves.
This found a bug of sorts were we didn't use 128-bit store for 512->128
extract on KNL. Its unclear what the right thing here should be.
Using the vextract avoids constraining the register allocator to use
xmm0-15. But it always results in a longer encoding if the register
allocator ends up choosing xmm0-15 anyway.
Craig Topper [Wed, 22 May 2019 20:04:55 +0000 (20:04 +0000)]
[X86][InstCombine] Remove InstCombine code that turns X86 round intrinsics into llvm.ceil/floor. Remove some isel patterns that existed because that was happening.
We were turning roundss/sd/ps/pd intrinsics with immediates of 1 or 2 into
llvm.floor/ceil. The llvm.ceil/floor intrinsics are supposed to correspond
to the libm functions. For the libm functions we need to disable the
precision exception so the llvm.floor/ceil functions should always map to
encodings 0x9 and 0xA.
We had a mix of isel patterns where some used 0x9 and 0xA and others used
0x1 and 0x2. We need to be consistent and always use 0x9 and 0xA.
Since we have no way in isel of knowing where the llvm.ceil/floor came
from, we can't map X86 specific intrinsics with encodings 1 or 2 to it.
We could map 0x9 and 0xA to llvm.ceil/floor instead, but I'd really like
to see a use case and optimization advantage first.
I've left the backend test cases to show the blend we now emit without
the extra isel patterns. But I've removed the InstCombine tests completely.
It copies 32-bit value from gpr32 into gpr64. But Live DEBUG_VALUE analysis
is not able to create debug location record for that instruction. So the problem
is in that debug info for argc variable is incorrect. The fix is
to write custom isCopyInstrImpl() which would recognize the ORRWrs instr.
Xing Xue [Wed, 22 May 2019 17:41:27 +0000 (17:41 +0000)]
Disable EHFrameSupport in JITLink/RuntimeDyld on AIX
Summary:
EH Frames aren't supported on AIX with the system compiler, but the definition of HAVE_EHTABLE_SUPPORT misses this which causes linking problems on AIX. This patch updates the definition of HAVE_EHTABLE_SUPPORT in both JITLink and RuntimeDyld.
Matt Arsenault [Wed, 22 May 2019 16:28:41 +0000 (16:28 +0000)]
MC: Allow getMaxInstLength to depend on the subtarget
Keep it optional in cases this is ever needed in some global
context. Currently it's only used for getting an upper bound inline
asm code size.
For AMDGPU, gfx10 increases the maximum instruction size to
20-bytes. This avoids penalizing older subtargets when estimating code
size, and making some annoying branch relaxation test adjustments.
Nico Weber [Wed, 22 May 2019 15:53:23 +0000 (15:53 +0000)]
llvm-undname: Fix an assert-on-invalid, found by oss-fuzz
If a template parameter refers to a pointer to member, but the mangling
of that was a string literal instead of a real symbol, llvm-undname used
to crash instead of rejecting the input.
Sanjay Patel [Wed, 22 May 2019 15:50:46 +0000 (15:50 +0000)]
[IR] allow fast-math-flags on select of FP values
This is a minimal start to correcting a problem most directly discussed in PR38086:
https://bugs.llvm.org/show_bug.cgi?id=38086
We have been hacking around a limitation for FP select patterns by using the
fast-math-flags on the condition of the select rather than the select itself.
This patch just allows FMF to appear with the 'select' opcode. No changes are
needed to "FPMathOperator" because it already includes select-of-FP because
that definition is based on the (return) value type.
Once we have this ability, we can start correcting and adding IR transforms
to use the FMF on a 'select' instruction. The instcombine and vectorizer test
diffs only show that the IRBuilder change is behaving as expected by applying
an FMF guard value to 'select'.
For reference:
rL241901 - allowed FMF with fcmp
rL255555 - allowed FMF with FP calls
Fangrui Song [Wed, 22 May 2019 15:12:51 +0000 (15:12 +0000)]
[llvm-objdump] Dump inline relocations if the relocated section is specified with --section
This fixes PR41886: llvm-objdump -d -r -j .text doesn't show inline relocations of .text
While here, switch to stable_sort() because we don't want to change the order of relocations applied to the same location. gABI says consecutive relocation records are composed together and their order matters. In practise it is difficult to see relocations applied to the same location not consecutive, we just have to keep the relative order of relocations with the same offset.
James Henderson [Wed, 22 May 2019 13:23:26 +0000 (13:23 +0000)]
[llvm-objcopy] Tidy up error messages
This patch brings various error messages into line with each other, by
removing trailing full stops, and making the first letter lower-case.
This addresses https://bugs.llvm.org/show_bug.cgi?id=40859.
Roman Lebedev [Wed, 22 May 2019 13:04:34 +0000 (13:04 +0000)]
UpdateTestChecks: sparc march handling
Summary:
Another target that prefers to use `-march` in tests
```
llvm/test/CodeGen/SPARC$ grep -ri mtriple | wc -l
25
llvm/test/CodeGen/SPARC$ grep -ri march | wc -l
165
```
This test is being affected by a further patch,
so regenerate it to better visualize the changes
Sjoerd Meijer [Wed, 22 May 2019 10:40:26 +0000 (10:40 +0000)]
[TargetMachine] error message unsupported code model
When the tiny code model is requested for a target machine that does not
support this, we get an error message (which is nice) but also this diagnostic
and request to submit a bug report:
fatal error: error in backend: Target does not support the tiny CodeModel
[Inferior 2 (process 31509) exited with code 0106]
clang-9: error: clang frontend command failed with exit code 70 (use -v to see invocation)
(gdb) clang version 9.0.0 (http://llvm.org/git/clang.git 29994b0c63a40f9c97c664170244a7bba5ecc15e) (http://llvm.org/git/llvm.git 95606fdf91c2d63a931e865f4b78b2e9828ddc74)
Target: arm-arm-none-eabi
Thread model: posix
clang-9: note: diagnostic msg: PLEASE submit a bug report to https://bugs.llvm.org/ and include the crash backtrace, preprocessed source, and associated run script.
clang-9: note: diagnostic msg:
********************
PLEASE ATTACH THE FOLLOWING FILES TO THE BUG REPORT:
Preprocessed source(s) and associated run script(s) are located at:
clang-9: note: diagnostic msg: /tmp/tiny-dfe1a2.c
clang-9: note: diagnostic msg: /tmp/tiny-dfe1a2.sh
clang-9: note: diagnostic msg:
But this is not a bug, this is a feature. :-) Not only is this not a bug, this
is also pretty confusing. This patch causes just to print the fatal error and
not the diagnostic:
fatal error: error in backend: Target does not support the tiny CodeModel
Anton Afanasyev [Wed, 22 May 2019 07:41:34 +0000 (07:41 +0000)]
[MIR] Add simple PRE pass to MachineCSE
This is the second part of the commit fixing PR38917 (hoisting
partitially redundant machine instruction). Most of PRE (partitial
redundancy elimination) and CSE work is done on LLVM IR, but some of
redundancy arises during DAG legalization. Machine CSE is not enough
to deal with it. This simple PRE implementation works a little bit
intricately: it passes before CSE, looking for partitial redundancy
and transforming it to fully redundancy, anticipating that the next
CSE step will eliminate this created redundancy. If CSE doesn't
eliminate this, than created instruction will remain dead and eliminated
later by Remove Dead Machine Instructions pass.
The third part of the commit is supposed to refactor MachineCSE,
to make it more clear and to merge MachinePRE with MachineCSE,
so one need no rely on further Remove Dead pass to clear instrs
not eliminated by CSE.
Fangrui Song [Wed, 22 May 2019 07:29:59 +0000 (07:29 +0000)]
[PPC64] Parse -elfv1 -elfv2 when specified on target triple
Summary:
For big-endian powerpc64, the default ABI is ELFv1. OpenPower ABI ELFv2 is supported when -mabi=elfv2 is specified. FreeBSD support for PowerPC64 ELFv2 ABI with LLVM is in progress[1]. This patch adds an alternative way to specify ELFv2 ABI on target triple [2].
The following results are expected:
ELFv1 when using:
-target powerpc64-unknown-freebsd12.0
-target powerpc64-unknown-freebsd12.0 -mabi=elfv1
-target powerpc64-unknown-freebsd12.0-elfv1
ELFv2 when using:
-target powerpc64-unknown-freebsd12.0 -mabi=elfv2
-target powerpc64-unknown-freebsd12.0-elfv2
Nikita Popov [Wed, 22 May 2019 06:47:06 +0000 (06:47 +0000)]
[X86] Don't compare i128 through vector if construction not cheap (PR41971)
Fix for https://bugs.llvm.org/show_bug.cgi?id=41971. Make the
combineVectorSizedSetCCEquality() transform more conservative by
checking that the bitcast to the vector type will be cheap/free
for both operands. I'm considering it cheap if it's a constant,
a load or already a vector. I've dropped the explicit check for
f128 because it should fall out naturally (in the cases where
it'd be detrimental).
Don Hinton [Wed, 22 May 2019 00:56:42 +0000 (00:56 +0000)]
[Docs] Increase Doxygen cache size
Summary:
When building Doxygen docs for llvm and clang, it helpfully prints a warning at
the end noting that the `LOOKUP_CACHE_SIZE` value was too small to keep all
symbols in memory.
By increasing to the size it recommends, Doxygen builds have greatly improved
performance. On my machine, time to run `doxygen-llvm` changes from 34 minutes
to 22 minutes, which is a decent amount of time saved by changing a single
number.
Pengfei Wang [Wed, 22 May 2019 00:50:21 +0000 (00:50 +0000)]
[X86] [CET] Deal with return-twice function such as vfork, setjmp when
CET-IBT enabled
Return-twice functions will indirectly jump after the caller's position.
So when CET-IBT is enable, we should make sure these is endbr*
instructions follow these Return-twice function caller. Like GCC does.
Sanjay Patel [Wed, 22 May 2019 00:32:25 +0000 (00:32 +0000)]
[InstCombine] fold shuffles of insert_subvectors
This should be a valid exception to the general rule of not creating new shuffle masks in IR...
because we already do it. :)
Also, DAG combining/legalization will undo this by widening the shuffle back out if needed.
Explanation for how we already do this: SLP or vector source can create chains of insert/extract
as shown in 1 of the examples from PR16739:
https://godbolt.org/z/NlK7rA
https://bugs.llvm.org/show_bug.cgi?id=16739
And we expect instcombine or DAGCombine to clean that up by creating relatively simple shuffles.
Lang Hames [Tue, 21 May 2019 22:07:53 +0000 (22:07 +0000)]
[ORC] Guarantee unique JITDylib names in lli, add usage notes to createJITDylib.
JITDylibs should have unique names. This patch adds code to lli to respect this
invariant (by refering to the exist JITDylib if a -jd <name> option is specified
more than once). It also adds usage notes to the doxygen comment for
createJITDylib method in ExecutionSession and LLJIT.
Sanjay Patel [Tue, 21 May 2019 21:45:24 +0000 (21:45 +0000)]
[InstCombine] add more tests for shuffle folding; NFC
As discussed in D62024, we want to limit any potential IR
transforms of shuffles to cases where we know the SDAG
conversion would result in equivalent patterns for these
IR variants.
Enable CMake policy 77. This alters the behavior of option. The old behavior
would remove the value of the option from the cache and create a new one. The
new behavior does not create the variable if it is defined already. This ensures
that subsequent reconfigures will behave identically. This seems better than the
setting of OLD - the desire is to ensure that it is set to OLD or NEW.
Register coalescer fails for the test in the patch with the assertion in
JoinVals::ConflictResolution `DefMI != nullptr'. It attempts to join
live intervals for two adjacent instructions and erase the copy:
The LI needs to be adjusted to kill subrange for the erased instruction
and extend the subrange of the original def. That was done for the main
interval only but not for the subrange. As a result subrange had a VNI
pointing to the erased slot resulting in the above failure.
Leonard Chan [Tue, 21 May 2019 19:17:19 +0000 (19:17 +0000)]
[Intrinsic] Signed Fixed Point Saturation Multiplication Intrinsic
Add an intrinsic that takes 2 signed integers with the scale of them provided
as the third argument and performs fixed point multiplication on them. The
result is saturated and clamped between the largest and smallest representable
values of the first 2 operands.
This is a part of implementing fixed point arithmetic in clang where some of
the more complex operations will be implemented as intrinsics.
Sanjay Patel [Tue, 21 May 2019 18:28:22 +0000 (18:28 +0000)]
[SelectionDAG] remove redundant code; NFCI
getNode() squashes concatenation of undefs via FoldCONCAT_VECTORS():
// Concat of UNDEFs is UNDEF.
if (llvm::all_of(Ops, [](SDValue Op) { return Op.isUndef(); }))
return DAG.getUNDEF(VT);
Don Hinton [Tue, 21 May 2019 17:56:45 +0000 (17:56 +0000)]
[cmake] Add custom command to touch archives on Darwin so ninja won't rebuild them.
Summary:
clang and newer versions of ninja use high-resolutions timestamps, but
older versions of libtool on Darwin don't, so the archive will often
get an older timestamp than the last object that was added or updated.
To fix this, we add a custom command to touch the archive after it's
been built so that ninja won't rebuild it unnecessarily the next time
it's run.
Chris Bieneman [Tue, 21 May 2019 16:29:31 +0000 (16:29 +0000)]
[docs] Add new document on building distributions
Summary:
This document is an attempt to provide a guide for best practices for using the LLVM build system to generate distributable LLVM-based tools.
Most of the document is geared toward distributions of LLVM-based toolchains, but much of it also applies to distributing other LLVM-based tools and libraries.
Reviewers: tstellar, phosek, jroelofs, hans, sylvestre.ledru
Clement Courbet [Tue, 21 May 2019 13:34:12 +0000 (13:34 +0000)]
[MergeICmps][NFC] Make BCEAtom move-only.
And handle for self-move. This is required so that llvm::sort can work
with EXPENSIVE_CHECKS, as it will do a random shuffle of the input
which can result in self-moves.
Florian Hahn [Tue, 21 May 2019 13:04:53 +0000 (13:04 +0000)]
[ScheduleDAGInstrs] Compute topological ordering on demand.
In most cases, the topological ordering does not get changed in
ScheduleDAGInstrs. We can compute the ordering on demand, similar to
D60125.
This drastically cuts down the number of times we need to compute the
topological ordering, e.g. for SPEC2006, SPEC2k and MultiSource, we get
the following stats for -O3 -flto on X86 (showing the top reductions,
with small absolute values filtered). The smallest reduction is -50%.
Slightly positive impact on compile-time (-0.1 % geomean speedup for
test-suite + SPEC & co, with -O1 on X86)
Paul Robinson [Tue, 21 May 2019 11:59:03 +0000 (11:59 +0000)]
[DebugInfo] Handle '# line "file"' correctly for asm source.
This provides the correct file path for the original source, rather
than the preprocessed source.
Bob Haarman [Tue, 21 May 2019 11:53:41 +0000 (11:53 +0000)]
Revert r360902 "Resubmit: [Salvage] Change salvage debug info ..."
This reverts commit rr360902. It caused an assertion failure in
lib/IR/DebugInfoMetadata.cpp: Assertion `(OffsetInBits + SizeInBits <=
FragmentSizeInBits) && "new fragment outside of original fragment"'
failed.
projects / llvm / log