Matt Arsenault [Mon, 3 Jun 2019 21:33:22 +0000 (21:33 +0000)]
AMDGPU: Disable stack realignment for kernels
This is something of a workaround, and the state of stack realignment
controls is kind of a mess. Ideally, we would be able to specify the
stack is infinitely aligned on entry to a kernel.
TargetFrameLowering provides multiple controls which apply at
different points. The StackRealignable field is used during
SelectionDAG, and for some reason distinct from this
hook. StackAlignment is a single field not dependent on the
function. It would probably be better to make that dependent on the
calling convention, and the maximum value for kernels.
Currently this doesn't really change anything, since the frame
lowering mostly does its own thing. This helps avoid regressions in a
future change which will rely more heavily on hasFP.
[AArch64][GlobalISel] Optimize G_FCMP + G_SELECT pairs when G_SELECT is fp
Instead of emitting all of the test stuff for a compare when it's only used by
a select, instead, just emit the compare + select. The select will use the
value of NZCV correctly, so we don't need to emit all of the test instructions
etc.
For now, only support fp selects which use G_FCMP. Also only support condition
codes which will only require one select to represent.
Craig Topper [Mon, 3 Jun 2019 19:29:14 +0000 (19:29 +0000)]
[X86] Fix the pattern for merge masked vcvtps2pd.
r362199 fixed it for zero masking, but not zero masking. The load
folding in the peephole pass hid the bug. This patch turns off
the peephole pass on the relevant test to ensure coverage.
Michael Berg [Mon, 3 Jun 2019 19:12:15 +0000 (19:12 +0000)]
Propagate fmf for setcc/select folds
Summary: This change facilitates propagating fmf which was placed on setcc from fcmp through folds with selects so that back ends can model this path for arithmetic folds on selects in SDAG.
[PowerPC] Look through copies for compare elimination
We currently miss the opportunities for optmizing comparisons in the peephole
optimizer if the input is the result of a COPY since we look for record-form
versions of the producing instruction.
This patch simply lets the optimization peek through copies.
Nikita Popov [Mon, 3 Jun 2019 18:19:54 +0000 (18:19 +0000)]
[ConstantRange] Add sdiv() support
The implementation is conceptually simple: We separate the LHS and
RHS into positive and negative components and then also compute the
positive and negative components of the result, taking into account
that e.g. only pos/pos and neg/neg will give a positive result.
However, there's one significant complication: SignedMin / -1 is UB
for sdiv, and we can't just ignore it, because the APInt result of
SignedMin would break the sign segregation. Instead we drop SignedMin
or -1 from the corresponding ranges, taking into account some edge
cases with wrapped ranges.
Because of the sign segregation, the implementation ends up being
nearly fully precise even for wrapped ranges (the remaining
imprecision is due to ranges that are both signed and unsigned
wrapping and are divided by a trivial divisor like 1). This means
that the testing cannot just check the signed envelope as we
usually do. Instead we collect all possible results in a bitvector
and construct a better sign wrapped range (than the full envelope).
Reid Kleckner [Mon, 3 Jun 2019 18:15:38 +0000 (18:15 +0000)]
[PDB] Copy inlinee lines records into the PDB
Summary:
- Fixes inline call frame line table display in windbg.
- Improve llvm-pdbutil to dump extra file ids.
- Warn on unknown subsections so we don't have this kind of bug in the
future.
Andrew Kaylor [Mon, 3 Jun 2019 17:54:15 +0000 (17:54 +0000)]
Fix a crash when the default of a switch is removed
This patch fixes a problem that occurs in LowerSwitch when a switch statement has a PHI node as its condition, and the PHI node only has two incoming blocks, and one of those incoming blocks is through an unreachable default in the switch statement. When this condition occurs, LowerSwitch holds a pointer to the condition value, but removes the switch block as a predecessor of the PHI block, causing the PHI node to be replaced. LowerSwitch then tries to use its stale pointer to the original condition value, causing a crash.
Add `nm` and `objcopy` to the default value for the tools that we install now
that they are sufficiently feature complete to replace bintuils' implementation.
[PowerPC] Set PROT_READ flag for MF_EXEC to prevent segfaults on PPC machines
The big endian PPC buildbots are all failing now due to calls to cache
invalidation in unit tests on data that has only the PROT_EXEC flag set.
This has been an issue all along on FreeBSD but it can affect Linux machines
depending on configuration.
This patch mitigates the issue the same way it is mitigated on FreeBSD.
Since this is needed to bring the buildbots back to green, I plan to commit this
and allow for post-commit review, but I thought I would also post it here for
ease of access/readability.
Owen Reynolds [Mon, 3 Jun 2019 15:26:07 +0000 (15:26 +0000)]
[llvm-ar] Fix relative thin archive path handling
This fixes some thin archive relative path issues, paths are shortened where possible and paths are output correctly when using the display table command.
Michal Gorny [Mon, 3 Jun 2019 14:50:03 +0000 (14:50 +0000)]
[llvm] [test] Remove non-portable EISDIR test from macho-disassemble-g-dsym.test
Remove the test checking error message for 'is a directory'. It does
not seem to serve any real purpose, and it relies on matching platform
error strings which are unpredictable and makes the test fragile.
Furthermore, it fails on NetBSD where read() works on directories,
and therefore does not return EISDIR at all.
Dmitri Gribenko [Mon, 3 Jun 2019 12:42:48 +0000 (12:42 +0000)]
Include what you use in LanaiMCCodeEmitter.cpp
LanaiMCCodeEmitter.cpp was not using any APIs from Lanai.h, and was only
including it for transitive dependencies. Doing so is problematic from
include-what-you-use perspective, but it is also a layering issue (it
creates a dependency cycle between the primary Lanai target library and
the MCTargetDesc library).
Dmitri Gribenko [Mon, 3 Jun 2019 11:41:22 +0000 (11:41 +0000)]
Include what you use in HexagonInstPrinter.cpp
HexagonInstPrinter.cpp was not using any APIs from HexagonAsmPrinter.h.
Doing so is problematic from include-what-you-use perspective, but it is
also a layering issue (it creates a dependency cycle between the primary
Hexagon target library and the MCTargetDesc library).
Dmitri Gribenko [Mon, 3 Jun 2019 11:25:37 +0000 (11:25 +0000)]
Include what you use in HexagonMCInstrInfo.cpp
HexagonMCInstrInfo.cpp was not using any APIs from Hexagon.h. Doing so
is problematic from include-what-you-use perspective, but it is also a
layering issue (it creates a dependency cycle between the primary
Hexagon target library and the MCTargetDesc library).
Dmitri Gribenko [Mon, 3 Jun 2019 11:20:53 +0000 (11:20 +0000)]
Include what you use in HexagonMCCodeEmitter.cpp
HexagonMCCodeEmitter.cpp was not using any APIs from Hexagon.h. Doing
so is problematic from include-what-you-use perspective, but it is also
a layering issue (it creates a dependency cycle between the primary
Hexagon target library and the MCTargetDesc library).
Dmitri Gribenko [Mon, 3 Jun 2019 11:20:48 +0000 (11:20 +0000)]
Include what you use in HexagonMCCompound.cpp
HexagonMCCompound.cpp was not using any APIs from Hexagon.h. Doing so
is problematic from include-what-you-use perspective, but it is also a
layering issue (it creates a dependency cycle between the primary
Hexagon target library and the MCTargetDesc library).
Dmitri Gribenko [Mon, 3 Jun 2019 11:14:20 +0000 (11:14 +0000)]
Include what you use in HexagonShuffler.cpp
HexagonShuffler.cpp was not using any APIs from Hexagon.h, and was only
including it for transitive dependencies. Doing so is problematic from
include-what-you-use perspective, but it is also a layering issue (it
creates a dependency cycle between the primary Hexagon target library
and the MCTargetDesc library).
Dmitri Gribenko [Mon, 3 Jun 2019 11:14:15 +0000 (11:14 +0000)]
Include what you use in HexagonMCChecker.cpp
HexagonMCChecker.cpp was not using any APIs from Hexagon.h. Doing so is
problematic from include-what-you-use perspective, but it is also a
layering issue (it creates a dependency cycle between the primary
Hexagon target library and the MCTargetDesc library).
Dmitri Gribenko [Mon, 3 Jun 2019 11:14:10 +0000 (11:14 +0000)]
Include what you use in HexagonMCTargetDesc.cpp
HexagonMCTargetDesc.cpp was not using any APIs from Hexagon.h. Doing so
is problematic from include-what-you-use perspective, but it is also a
layering issue (it creates a dependency cycle between the primary
Hexagon target library and the MCTargetDesc library).
Dmitri Gribenko [Mon, 3 Jun 2019 11:14:05 +0000 (11:14 +0000)]
Include what you use in HexagonMCShuffler.cpp
HexagonMCShuffler.cpp was not using any APIs from Hexagon.h. Doing so
is problematic from include-what-you-use perspective, but it is also a
layering issue (it creates a dependency cycle between the primary
Hexagon target library and the MCTargetDesc library).
Simon Tatham [Mon, 3 Jun 2019 11:02:53 +0000 (11:02 +0000)]
[ARM] Fix recent breakage of -mfpu=none.
The recent change D60691 introduced a bug in clang when handling
option combinations such as `-mcpu=cortex-m4 -mfpu=none`. Those
options together should select Cortex-M4 but disable all use of
hardware FP, but in fact, now hardware FP instructions can still be
generated in that mode.
The reason is because the handling of FPUVersion::NONE disables all
the same feature names it used to, of which the base one is `vfp2`.
But now there are further features below that, like `vfp2d16fp` and
(following D60694) `fpregs`, which also need to be turned off to
disable hardware FP completely.
Added a tiny test which double-checks that compiling a simple FP
function doesn't access the FP registers.
George Rimar [Mon, 3 Jun 2019 09:58:41 +0000 (09:58 +0000)]
[llvm-readobj] - Convert gnu-sections.test to use YAML.
gnu-sections.test currently use relocs.obj.elf-x86_64 and
relocs.obj.elf-i386 precompiled objects as an inputs.
These inputs actually initially were introduced to test the
dump of relocations and have almost nothing common with dumping
sections.
Patch converts the test to use yaml2obj. That allows to remove
relocs.obj.elf-i386 binary.
(relocs.obj.elf-x86_64 is still used by another test and can't be removed atm).
Dmitri Gribenko [Mon, 3 Jun 2019 09:56:40 +0000 (09:56 +0000)]
Include what you use in HexagonELFObjectWriter.cpp
HexagonELFObjectWriter.cpp was not using any APIs from Hexagon.h, and
was only including it for transitive dependencies. Doing so is
problematic from include-what-you-use perspective, but it is also a
layering issue (it creates a dependency cycle between the primary
Hexagon target library and the MCTargetDesc library).
rL362089 introduced a set of yaml based reloc-types-*.test test cases
(instead of huge reloc-types.test that used a lot of precompiled binaries)
These test cases checks LLVM-styled dumping of the relocations.
gnu-relocations.test was a test case to check GNU styled relocations dumping.
It did that only for elf-x86 and elf-x86_64 targets. It did not test all of the
relocations though.
Now, after rL362089, it does not make sence to keep it.
This patch updates reloc-types-elf-i386.test and reloc-types-elf-x64.test tests
with llvm-readelf calls to check GNU styled output in one place.
It removes gnu-relocations.test completely.
One of intentions of doing this is also to get rid of relocs.obj.elf-i386 and
relocs.obj.elf-x86_64 precompiled objects completely (they are used in other tests still).
Nikola Prica [Mon, 3 Jun 2019 09:48:29 +0000 (09:48 +0000)]
[LiveDebugValues] Close range for previous variable's location when adding newly deduced location
When LiveDebugValues deduces new variable's location from spill, restore or
register copy instruction it should close old variable's location. Otherwise
we can have multiple block output locations for same variable. That could lead
to inserting two DBG_VALUEs for same variable to the beginning of the successor
block which results to ignoring of first DBG_VALUE.
Dmitri Gribenko [Mon, 3 Jun 2019 09:43:05 +0000 (09:43 +0000)]
Include what you use in HexagonAsmBackend.cpp
HexagonAsmBackend.cpp was not using any APIs from Hexagon.h. Doing so
is problematic from include-what-you-use perspective, but it is also a
layering issue (it creates a dependency cycle between the primary
Hexagon target library and the MCTargetDesc library).
Dmitri Gribenko [Mon, 3 Jun 2019 09:38:48 +0000 (09:38 +0000)]
Include what you use in HexagonAsmParser.cpp
HexagonAsmParser.cpp was not using any APIs from Hexagon.h. Doing so is
problematic from include-what-you-use perspective, but it is also a
layering issue (it creates a dependency cycle between the primary
Hexagon target library and the AsmParser library).
Dmitri Gribenko [Mon, 3 Jun 2019 09:33:48 +0000 (09:33 +0000)]
Include what you use in HexagonShuffler.h
HexagonShuffler.h was not using any APIs from Hexagon.h, and was only
including it for transitive dependencies. Doing so is problematic from
include-what-you-use perspective, but it is also a layering issue (it
creates a dependency cycle between the primary Hexagon target library
and the MCTargetDesc library).
Dmitri Gribenko [Mon, 3 Jun 2019 09:29:51 +0000 (09:29 +0000)]
Include what you use in BPFMCTargetDesc.cpp
BPFMCTargetDesc.cpp was not using any APIs from BPF.h. Doing so is
problematic from include-what-you-use perspective, but it is also a
layering issue (it creates a dependency cycle between the primary
BPF target library and the MCTargetDesc library).
[ARM][FIX] Ran out of registers due tail recursion
Summary:
- pr42062
When compiling for MinSize,
ARMTargetLowering::LowerCall decides to indirect
multiple calls to a same function. However,
it disconsiders the limitation that thumb1
indirect calls require the callee to be in a
register from r0 to r3 (llvm limiation).
If all those registers are used by arguments, the
compiler dies with "error: run out of registers
during register allocation".
This patch tells the function
IsEligibleForTailCallOptimization if we intend to
perform indirect calls, as to avoid tail call
optimization.
Mikael Holmen [Mon, 3 Jun 2019 06:38:01 +0000 (06:38 +0000)]
[TableGen] Fix std::array initializer to avoid warnings with older tool chains. NFC
A std::array is implemented as a template with an array inside a struct.
Older versions of clang, like 3.6, require an extra set of curly braces
around std::array initializations to avoid warnings.
The C++ language was changed regarding this by CWG 1270. So more modern
tool chains does not complain even if leaving out one level of braces.
Jim Lin [Mon, 3 Jun 2019 02:31:07 +0000 (02:31 +0000)]
[AVR] Fix incorrect source regclass of LDWRdPtr
Summary:
LDWRdPtr would be expanded to ld+ldd. ldd only accepts the pointer register is Y or Z.
So the register class of pointer of LDWRdPtr should be PTRDISPREGS instead of PTRREGS.
Florian Hahn [Mon, 3 Jun 2019 01:30:19 +0000 (01:30 +0000)]
Recommit r360171: [DAGCombiner] Avoid creating large tokenfactors in visitTokenFactor.
If we hit the limit, we do expand the outstanding tokenfactors.
Otherwise, we might drop nodes with users in the unexpanded
tokenfactors. This fixes the crashes reported by Jordan Rupprecht.
Mike Spertus [Sun, 2 Jun 2019 23:33:32 +0000 (23:33 +0000)]
Update MSVC Visualizer to reflect new variadic PointerUnion
This changed updates the MSVC Visualizer to work with the recent change
of PointerUnion into a variadic template. As an extra bonus, we
fix some bit rot in the SmallPtrSet visualizer as well
Nico Weber [Sun, 2 Jun 2019 23:26:57 +0000 (23:26 +0000)]
llvm-undname; Add more test coverage for demangleFunctionClass()
Also add two FC_Far that seem to be missing, by symmetry from
the public and protected cases. (But FC_Far isn't really a thing
anymore, so this doesn't really have an observable effect.)
A mixture of internal tests and review of the scheduler models indicates we're overestimating the cost of a masked load, which we're estimating at 4x regular memory ops - more realistic values indicates that its closer to 2x. Masked stores costs are a lot more diverse but 8x is roughly in the middle of the range.
Don Hinton [Sun, 2 Jun 2019 15:53:43 +0000 (15:53 +0000)]
[test] Fix plugin tests
Recommit of r361790 that was temporarily reverted in r361793 due to bot breakage.
Summary:
The following changes were required to fix these tests:
1) Change LLVM_ENABLE_PLUGINS to an option and move it to
llvm/CMakeLists.txt with an appropriate default -- which matches
the original default behavior.
2) Move the plugins directory from clang/test/Analysis
clang/lib/Analysis. It's not enough to add an exclude to the
lit.local.cfg file because add_lit_testsuites recurses the tree and
automatically adds the appropriate `check-` targets, which don't
make sense for the plugins because they aren't tests and don't
have `RUN` statements.
Here's a list of the `clang-check-anlysis*` targets with this
change:
Simon Pilgrim [Sun, 2 Jun 2019 11:56:39 +0000 (11:56 +0000)]
[DAG] isBitwiseNot / isConstOrConstSplat - add support for build vector undefs + truncation (PR41020)
Add (opt-in) support for implicit truncation to isConstOrConstSplat, which allows us to match truncated 'all ones' cases in isBitwiseNot.
PR41020 compares against using ISD::isBuildVectorAllOnes() instead, but that predicate silently accepts any UNDEF elements in the build vector which might not be what we want in isBitwiseNot - so I've added an opt-in 'AllowUndefs' flag that is set to false by default but will allow us to enable it on individual cases where its safe.
Petr Hosek [Sun, 2 Jun 2019 02:05:01 +0000 (02:05 +0000)]
[CMake] Use libtool for runtimes when building for Apple platform
LLVM CMake build already uses libtool instead of ar when building
for Apple platform and we should be using the same when building
runtimes. To do so, this change extracts the logic for finding
libtool into a separate file and then uses it from both the LLVM
build as well as the LLVM runtimes build.
Craig Topper [Sun, 2 Jun 2019 01:36:48 +0000 (01:36 +0000)]
[X86] Fix several places that weren't passing what they though they were to MachineInstr::print
Over a year ago, MachineInstr gained a fourth boolean parameter that occurs
before the TII pointer. When this happened, several places started accidentally
passing TII into this boolean parameter instead of the TII parameter.
Extract a willNotOverflow() helper function that is shared between
eliminateOverflowIntrinsic() and strengthenOverflowingOperation().
Use WithOverflowInst for the former.
We'll be able to reuse the same code for saturating intrinsics as
well.
projects / llvm / log