Julian Lettner [Wed, 16 Oct 2019 21:53:20 +0000 (21:53 +0000)]
[lit] Small refactoring and cleanups in main.py
* Remove outdated precautions for Python versions < 2.7
* Remove dead code related to `maxIndividualTestTime` option
* Move printing of test and result summary out of main into its own
function
[dsymutil] Print warning/error for unknown/missing arguments.
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.
Shoaib Meenai [Wed, 16 Oct 2019 21:41:05 +0000 (21:41 +0000)]
[AArch64] Fix offset calculation
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!
Philip Reames [Wed, 16 Oct 2019 19:58:26 +0000 (19:58 +0000)]
[IndVars] Fix a miscompile in off-by-default loop predication implementation
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.
Jordan Rupprecht [Wed, 16 Oct 2019 18:39:52 +0000 (18:39 +0000)]
[llvm-ar] Implement the V modifier as an alias for --version
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`.
Sanjay Patel [Wed, 16 Oct 2019 18:06:24 +0000 (18:06 +0000)]
[SLP] avoid reduction transform on patterns that the backend can load-combine (2nd try)
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=42708
https://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:
Jason Liu [Wed, 16 Oct 2019 17:36:31 +0000 (17:36 +0000)]
[NFC][XCOFF][AIX] Rename ControlSections to CsectGroup
The name of ControlSections is not expressive enough to convey what they really are.
CsectGroup can better communicate the concept of grouping csects together since they have similar property.
Joel E. Denny [Wed, 16 Oct 2019 17:21:57 +0000 (17:21 +0000)]
[lit] Fix internal diff's --strip-trailing-cr and use it
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.
Joel E. Denny [Wed, 16 Oct 2019 17:21:24 +0000 (17:21 +0000)]
[lit] Clean up internal diff's encoding handling
As suggested by rnk at D67643#1673043, instead of reading files
multiple times until an appropriate encoding is found, read them once
as binary, and then try to decode what was read.
For Python >= 3.5, don't fail when attempting to decode the
`diff_bytes` output in order to print it.
Avoid failures for Python 2.7 used on some Windows bots by
transforming diff output with `lit.util.to_string` before writing it
to stdout.
Jiong Wang [Wed, 16 Oct 2019 15:27:59 +0000 (15:27 +0000)]
bpf: fix wrong truncation elimination when there is back-edge/loop
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>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@375007 91177308-0d34-0410-b5e6-96231b3b80d8
Luis Marques [Wed, 16 Oct 2019 15:06:02 +0000 (15:06 +0000)]
[RISCV] Add MachineInstr immediate verification
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.
David Stuttard [Wed, 16 Oct 2019 14:37:39 +0000 (14:37 +0000)]
[AMDGPU] Fix-up cases where writelane has 2 SGPR operands
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).
Piotr Sobczak [Wed, 16 Oct 2019 11:14:01 +0000 (11:14 +0000)]
[InstCombine][AMDGPU] Fix crash with v3i16/v3f16 buffer intrinsics
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.
Mikhail Maltsev [Wed, 16 Oct 2019 10:40:57 +0000 (10:40 +0000)]
[ARM] Add a register class for GPR pairs without SP and use it. NFCI
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.
Sam Parker [Wed, 16 Oct 2019 09:37:03 +0000 (09:37 +0000)]
[ARM][ParallelDSP] Change smlad insertion order
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.
[NFC] Replace a linked list in LiveDebugVariables pass with a DenseMap
In LiveDebugVariables.cpp:
Prior to this patch, UserValues were grouped into linked list chains. Each
chain was the union of two sets: { A: Matching Source variable } or
{ B: Matching virtual register }. A ptr to the heads (or 'leaders')
of each of these chains were kept in a map with the { Source variable } used
as the key (set A predicate) and another with { Virtual register } as key
(set B predicate).
There was a search through the chains in the function getUserValue looking for
UserValues with matching { Source variable, Complex expression, Inlined-at
location }. Essentially searching for a subset of A through two interleaved
linked lists of set A and B. Importantly, by design, the subset will only
contain one or zero elements here. That is to say a UserValue can be uniquely
identified by the tuple { Source variable, Complex expression, Inlined-at
location } if it exists.
This patch removes the linked list and instead uses a DenseMap to map
the tuple { Source variable, Complex expression, Inlined-at location }
to UserValue ptrs so that the getUserValue search predicate is this map key.
The virtual register map now maps a vreg to a SmallVector<UserVal *> so that
set B is still available for quick searches.
Craig Topper [Wed, 16 Oct 2019 03:29:24 +0000 (03:29 +0000)]
[LegalizeTypes] Don't use PromoteTargetBoolean in WidenVecOp_SETCC.
Similar to r374970, but I don't have a test for this.
PromoteTargetBoolean is intended to be use for legalizing an
operand that needs to be promoted. It picks its type based on
the return from getSetccResultType and is intended to be used
when we have freedom to pick the new type. But the return type
we need for WidenVecOp_SETCC is completely determined by the
type of the input node.
Craig Topper [Wed, 16 Oct 2019 02:50:04 +0000 (02:50 +0000)]
[LegalizeTypes] Don't call PromoteTargetBoolean from SplitVecOp_VSETCC.
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.
[dsymutil] Support and relocate base address selection entries for debug_loc
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.
[Reland][VirtualFileSystem] Support virtual working directory in the RedirectingFS
Before this patch, changing the working directory of the RedirectingFS
would just forward to its external file system. This prevented us from
having a working directory that only existed in the VFS mapping.
This patch adds support for a virtual working directory in the
RedirectingFileSystem. It now keeps track of its own WD in addition to
updating the WD of the external file system. This ensures that we can
still fall through for relative paths.
This change was originally motivated by the reproducer infrastructure in
LLDB where we want to deal transparently with relative paths.
Lang Hames [Tue, 15 Oct 2019 21:41:12 +0000 (21:41 +0000)]
[Orc] Add a method for ObjectLinkingLayer to return ownership of object buffers.
RTDyldObjectLinkingLayer allowed clients to register a NotifyEmitted function to
reclaim ownership of object buffers once they had been linked. This patch adds
similar functionality to ObjectLinkingLayer: Clients can now optionally call the
ObjectLinkingLayer::setReturnObjectBuffer method to register a function that
will be called when discarding object buffers. If set, this function will be
called to return ownership of the object regardless of whether the link
succeeded or failed.
Use cases for this function include debug dumping (it provides a way to dump
all objects linked into JIT'd code) and object re-use (e.g. storing an
object in a cache).
Lang Hames [Tue, 15 Oct 2019 21:06:57 +0000 (21:06 +0000)]
[JITLink] Switch to slab allocation for InProcessMemoryManager, re-enable test.
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.
Julian Lettner [Tue, 15 Oct 2019 20:57:20 +0000 (20:57 +0000)]
[lit] Add back LitTestCase
This essentially reverts a commit [1] that removed the adaptor for
Python unittests. The code has been slightly refactored to make it more
additive: all code is contained in LitTestCase.py.
Usage sites will require a small adaption:
```
[old]
import lit.discovery
...
test_suite = lit.discovery.load_test_suite(...)
Austin Kerbow [Tue, 15 Oct 2019 19:59:45 +0000 (19:59 +0000)]
AMDGPU: Fix infinite searches in SIFixSGPRCopies
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.
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.
Thomas Lively [Tue, 15 Oct 2019 18:28:22 +0000 (18:28 +0000)]
[WebAssembly] Allow multivalue types in block signature operands
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.
Jordan Rupprecht [Tue, 15 Oct 2019 18:13:20 +0000 (18:13 +0000)]
[llvm-objdump] Use a counter for llvm-objdump -h instead of the section index.
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.
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
Digger Lin [Tue, 15 Oct 2019 17:40:41 +0000 (17:40 +0000)]
[XCOFF] Output object text section header and symbol entry for program code.
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.
Alina Sbirlea [Tue, 15 Oct 2019 17:25:36 +0000 (17:25 +0000)]
[NewGVN] Check that call has an access.
Check that a call has an attached MemoryAccess before calling
getClobbering on the instruction.
If no access is attached, the instruction does not access memory.
[VirtualFileSystem] Support virtual working directory in the RedirectingFS
Before this patch, changing the working directory of the RedirectingFS
would just forward to its external file system. This prevented us from
having a working directory that only existed in the VFS mapping.
This patch adds support for a virtual working directory in the
RedirectingFileSystem. It now keeps track of its own WD in addition to
updating the WD of the external file system. This ensures that we can
still fall through for relative paths.
This change was originally motivated by the reproducer infrastructure in
LLDB where we want to deal transparently with relative paths.
Digger Lin [Tue, 15 Oct 2019 17:09:54 +0000 (17:09 +0000)]
[AIX][XCOFF] Output XCOFF object text section header and symbol entry for program code.
SUMMARY
Original form of this patch is provided by Stefan Pintillie.
The patch try to output program code section header , symbol entry for program code (PR) and Instruction into the raw text section.
The patch include how to alignment and layout the CSection in the text section.
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.
This is a small generalization of a fold requested in PR43650:
https://bugs.llvm.org/show_bug.cgi?id=43650
The sign-bit of the condition operand can be used as a mask for the true operand:
https://rise4fun.com/Alive/paT
Note that we already handle some of the patterns (isNegative + scalar) because
there's an over-specialized, yet over-reaching fold for that in foldSelectCCToShiftAnd().
It doesn't use any TLI hooks, so I can't easily rip out that code even though we're
duplicating part of it here. This fold is guarded by TLI.convertSelectOfConstantsToMath(),
so it should not cause problems for targets that prefer select over shift.
Also worth noting: I thought we could generalize this further to include the case where
the true operand of the select is not constant, but Alive says that may allow poison to
pass through where it does not in the original select form of the code.
Summary:
This is patch is part of a series to introduce an Alignment type.
See this thread for context: http://lists.llvm.org/pipermail/llvm-dev/2019-July/133851.html
See this patch for the introduction of the type: https://reviews.llvm.org/D64790
Sam Parker [Tue, 15 Oct 2019 13:12:51 +0000 (13:12 +0000)]
[ARM][MVE] validForTailPredication insts
Reverse the logic for valid tail predication instructions and create
a whitelist instead. Added other instruction groups that aren't
obviously safe:
- instructions that 'narrow' their result.
- lane moves.
- byte swapping instructions.
- interleaving loads and stores.
- cross-beat carries.
- top/bottom instructions.
- complex operations.
Hopefully we should be able to add more of these instructions to the
whitelist, once we have a more concrete idea of the transform.
Sanjay Patel [Tue, 15 Oct 2019 13:12:44 +0000 (13:12 +0000)]
[InstCombine] fold a shifted bool zext to a select (2nd try)
The 1st attempt at rL374828 inserted the code
at the wrong position (outside of the constant-shift-amount
block). Trying again with an additional test to verify
const-ness.
For a constant shift amount, add the following fold.
shl (zext (i1 X)), ShAmt --> select (X, 1 << ShAmt, 0)
Summary:
This is patch is part of a series to introduce an Alignment type.
See this thread for context: http://lists.llvm.org/pipermail/llvm-dev/2019-July/133851.html
See this patch for the introduction of the type: https://reviews.llvm.org/D64790
David Stenberg [Tue, 15 Oct 2019 11:31:21 +0000 (11:31 +0000)]
[DebugInfo] Add a DW_OP_LLVM_entry_value operation
Summary:
Internally in LLVM's metadata we use DW_OP_entry_value operations with
the same semantics as DWARF; that is, its operand specifies the number
of bytes that the entry value covers.
At the time of emitting entry values we don't know the emitted size of
the DWARF expression that the entry value will cover. Currently the size
is hardcoded to 1 in DIExpression, and other values causes the verifier
to fail. As the size is 1, that effectively means that we can only have
valid entry values for registers that can be encoded in one byte, which
are the registers with DWARF numbers 0 to 31 (as they can be encoded as
single-byte DW_OP_reg0..DW_OP_reg31 rather than a multi-byte
DW_OP_regx). It is a bit confusing, but it seems like llvm-dwarfdump
will print an operation "correctly", even if the byte size is less than
that, which may make it seem that we emit correct DWARF for registers
with DWARF numbers > 31. If you instead use readelf for such cases, it
will interpret the number of specified bytes as a DWARF expression. This
seems like a limitation in llvm-dwarfdump.
As suggested in D66746, a way forward would be to add an internal
variant of DW_OP_entry_value, DW_OP_LLVM_entry_value, whose operand
instead specifies the number of operations that the entry value covers,
and we then translate that into the byte size at the time of emission.
In this patch that internal operation is added. This patch keeps the
limitation that a entry value can only be applied to simple register
locations, but it will fix the issue with the size operand being
incorrect for DWARF numbers > 31.
[Alignment][NFC] Remove dependency on GlobalObject::setAlignment(unsigned)
Summary:
This is patch is part of a series to introduce an Alignment type.
See this thread for context: http://lists.llvm.org/pipermail/llvm-dev/2019-July/133851.html
See this patch for the introduction of the type: https://reviews.llvm.org/D64790
David Stenberg [Tue, 15 Oct 2019 11:14:35 +0000 (11:14 +0000)]
[DebugInfo] Add interface for pre-calculating the size of emitted DWARF
Summary:
DWARF's DW_OP_entry_value operation has two operands; the first is a
ULEB128 operand that specifies the size of the second operand, which is
a DWARF block. This means that we need to be able to pre-calculate and
emit the size of DWARF expressions before emitting them. There is
currently no interface for doing this in DwarfExpression, so this patch
introduces that.
When implementing this I initially thought about running through
DwarfExpression's emission two times; first with a temporary buffer to
emit the expression, in order to being able to calculate the size of
that emitted data. However, DwarfExpression is a quite complex state
machine, so I decided against that, as it seemed like the two runs could
get out of sync, resulting in incorrect size operands. Therefore I have
implemented this in a way that we only have to run DwarfExpression once.
The idea is to emit DWARF to a temporary buffer, for which it is
possible to query the size. The data in the temporary buffer can then be
emitted to DwarfExpression's main output.
In the case of DIEDwarfExpression, a temporary DIE is used. The values
are all allocated using the same BumpPtrAllocator as for all other DIEs,
and the values are then transferred to the real value list. In the case
of DebugLocDwarfExpression, the temporary buffer is implemented using a
BufferByteStreamer which emits to a buffer in the DwarfExpression
object.
Jeremy Morse [Tue, 15 Oct 2019 10:46:24 +0000 (10:46 +0000)]
[DebugInfo] Remove some users of DBG_VALUEs IsIndirect field
This patch kills off a significant user of the "IsIndirect" field of
DBG_VALUE machine insts. Brought up in in PR41675, IsIndirect is
techncally redundant as it can be expressed by the DIExpression of a
DBG_VALUE inst, and it isn't helpful to have two ways of expressing
things.
Rather than setting IsIndirect, have DBG_VALUE creators add an extra deref
to the insts DIExpression. There should now be no appearences of
IsIndirect=True from isel down to LiveDebugVariables / VirtRegRewriter,
which is ensured by an assertion in LDVImpl::handleDebugValue. This means
we also get to delete the IsIndirect handling in LiveDebugVariables. Tests
can be upgraded by for example swapping the following IsIndirect=True
DBG_VALUE:
Petar Avramovic [Tue, 15 Oct 2019 09:30:08 +0000 (09:30 +0000)]
[MIPS GlobalISel] Add MSA registers to fprb. Select vector load, store
Add vector MSA register classes to fprb, they are 128 bit wide.
MSA instructions use the same registers for both integer and floating
point operations. Therefore we only need to check for vector element
size during legalization or instruction selection.
Add helper function in MipsLegalizerInfo and switch to legalIf
LegalizeRuleSet to keep legalization rules compact since they depend
on MipsSubtarget and presence of MSA.
fprb is assigned to all vector operands.
Move selectLoadStoreOpCode to MipsInstructionSelector in order to
reduce number of arguments.
David Stenberg [Tue, 15 Oct 2019 09:21:09 +0000 (09:21 +0000)]
Change Comments SmallVector to std::vector in DebugLocStream [NFC]
This changes the 32-element SmallVector to a std::vector. When building
a RelWithDebInfo clang-8 binary, the average size of the vector was
~10000, so it does not seem very beneficial or practical to use a small
vector for that.
The DWARFBytes SmallVector grows in the same way as Comments, so perhaps
that also should be changed to a purely dynamically allocated structure,
but that requires some more code changes, so I let that remain as a
SmallVector for now.
Check if size of operand LLT matches sizes of available register banks
before inspecting the opcode in order to reduce number of checks.
Factor commonly used pieces of code into functions.
Martin Storsjo [Tue, 15 Oct 2019 08:29:56 +0000 (08:29 +0000)]
[Demangle] Add a few more options to the microsoft demangler
This corresponds to commonly used options to UnDecorateSymbolName
within llvm.
Add them as hidden options in llvm-undname. MS undname.exe takes
numeric flags, corresponding to the UNDNAME_* constants, but instead
of hardcoding in mappings for those numbers, just add textual
options instead, as it the use of them here is primarily intended
for testing.
Craig Topper [Tue, 15 Oct 2019 06:10:11 +0000 (06:10 +0000)]
[X86] Don't check for VBROADCAST_LOAD being a user of the source of a VBROADCAST when trying to share broadcasts.
The only things VBROADCAST_LOAD uses is an address and a chain
node. It has no vector inputs.
So if its a user of the source of another broadcast that could
only mean one of two things. The other broadcast is broadcasting
the address of the broadcast_load. Or the source is a load and
the use we're seeing is the chain result from that load. Neither
of these cases make sense to combine here.
This issue was reported post-commit r373871. Test case has not
been reduced yet.
Shiva Chen [Tue, 15 Oct 2019 02:04:29 +0000 (02:04 +0000)]
[RISCV] Support fast calling convention
LLVM may annotate the function with fastcc if there has only one caller
and there're no other caller out of the module and the function is not
naked or contain variable arguments.
The fastcc functions could pass the arguments by the caller saved registers.
Thomas Lively [Tue, 15 Oct 2019 01:11:51 +0000 (01:11 +0000)]
[WebAssembly] Trapping fptoint builtins and intrinsics
Summary:
The WebAssembly backend lowers fptoint instructions to a code sequence
that checks for overflow to avoid traps because fptoint is supposed to
be speculatable. These new builtins and intrinsics give users a way to
depend on the trapping semantics of the underlying instructions and
avoid the extra code generated normally.