Zachary Turner [Thu, 18 May 2017 23:03:06 +0000 (23:03 +0000)]
[CodeView] Provide a common interface for type collections.
Right now we have multiple notions of things that represent collections of
types. Most commonly used are TypeDatabase, which is supposed to keep
mappings from TypeIndex to type name when reading a type stream, which
happens when reading PDBs. And also TypeTableBuilder, which is used to
build up a collection of types dynamically which we will later serialize
(i.e. when writing PDBs).
But often you just want to do some operation on a collection of types, and
you may want to do the same operation on any kind of collection. For
example, you might want to merge two TypeTableBuilders or you might want
to merge two type streams that you loaded from various files.
This dichotomy between reading and writing is responsible for a lot of the
existing code duplication and overlapping responsibilities in the existing
CodeView library classes. For example, after building up a
TypeTableBuilder with a bunch of type records, if we want to dump it we
have to re-invent a bunch of extra glue because our dumper takes a
TypeDatabase or a CVTypeArray, which are both incompatible with
TypeTableBuilder.
This patch introduces an abstract base class called TypeCollection which
is shared between the various type collection like things. Wherever we
previously stored a TypeDatabase& in some common class, we now store a
TypeCollection&.
The advantage of this is that all the details of how the collection are
implemented, such as lazy deserialization of partial type streams, is
completely transparent and you can just treat any collection of types the
same regardless of where it came from.
Sanjay Patel [Thu, 18 May 2017 20:53:16 +0000 (20:53 +0000)]
[InstCombine] add helper to foldXorOfICmps(); NFCI
Also, fix the old-style capitalization of the related functions
and move them to the 'private' section of the class since they
are just helpers of the visit* functions.
As shown in the post-commit comments for D32143, we are missing
folds for xor-of-icmps.
Reid Kleckner [Thu, 18 May 2017 17:24:10 +0000 (17:24 +0000)]
[IR] De-virtualize ~Value to save a vptr
Summary:
Implements PR889
Removing the virtual table pointer from Value saves 1% of RSS when doing
LTO of llc on Linux. The impact on time was positive, but too noisy to
conclusively say that performance improved. Here is a link to the
spreadsheet with the original data:
This change makes it invalid to directly delete a Value, User, or
Instruction pointer. Instead, such code can be rewritten to a null check
and a call Value::deleteValue(). Value objects tend to have their
lifetimes managed through iplist, so for the most part, this isn't a big
deal. However, there are some places where LLVM deletes values, and
those places had to be migrated to deleteValue. I have also created
llvm::unique_value, which has a custom deleter, so it can be used in
place of std::unique_ptr<Value>.
I had to add the "DerivedUser" Deleter escape hatch for MemorySSA, which
derives from User outside of lib/IR. Code in IR cannot include MemorySSA
headers or call the MemoryAccess object destructors without introducing
a circular dependency, so we need some level of indirection.
Unfortunately, no class derived from User may have any virtual methods,
because adding a virtual method would break User::getHungOffOperands(),
which assumes that it can find the use list immediately prior to the
User object. I've added a static_assert to the appropriate OperandTraits
templates to help people avoid this trap.
Wei Mi [Thu, 18 May 2017 17:21:22 +0000 (17:21 +0000)]
[LSR] Call canonicalize after we generate a new Formula in GenerateTruncates. Fix PR33077.
The testcase in PR33077 generates a LSR Use Formula with two SCEVAddRecExprs for the same
loop. Such uncommon formula will become non-canonical after GenerateTruncates adds sign
extension to the ScaledReg of the Formula, and it will break the assertion that every
Formula to be inserted is canonical.
The fix is to call canonicalize for the raw Formula generated by GenerateTruncates
before inserting it.
This provides a new way to access the TargetMachine through
TargetPassConfig, as a dependency.
The patterns replaced here are:
* Passes handling a null TargetMachine call
`getAnalysisIfAvailable<TargetPassConfig>`.
* Passes not handling a null TargetMachine
`addRequired<TargetPassConfig>` and call
`getAnalysis<TargetPassConfig>`.
* MachineFunctionPasses now use MF.getTarget().
* Remove all the TargetMachine constructors.
* Remove INITIALIZE_TM_PASS.
This fixes a crash when running `llc -start-before prologepilog`.
PEI needs StackProtector, which gets constructed without a TargetMachine
by the pass manager. The StackProtector pass doesn't handle the case
where there is no TargetMachine, so it segfaults.
Zachary Turner [Thu, 18 May 2017 15:14:44 +0000 (15:14 +0000)]
Fix some minor issues in PDB parsing library.
1) Until now I'd never seen a valid PDB where the DBI stream and
the PDB Stream disagreed on the "Age" field. Because of that,
we had code to assert that they matched. Recently though I was
given a PDB where they disagreed, so this assumption has proven
to be incorrect. Remove this check.
2) We were walking the entire list of hash values for types up front
and then throwing away the values. For large PDBs this was a
significant slow down. Remove this.
With this patch, I can dump the list of all compilands from a
1.5GB PDB file in just a few seconds.
Anna Thomas [Thu, 18 May 2017 13:12:18 +0000 (13:12 +0000)]
[JumpThreading] Dont RAUW condition incorrectly
Summary:
We have a bug when RAUWing the condition if experimental.guard or assumes is a use of that
condition. This is because LazyValueInfo may have used the guards/assumes to identify the
value of the condition at the end of the block. RAUW replaces the uses
at the guard/assume as well as uses before the guard/assume. Both of
these are incorrect.
For now, disable RAUW for conditions and fix the logic as a next
step: https://reviews.llvm.org/D33257
Sam Kolton [Thu, 18 May 2017 12:12:03 +0000 (12:12 +0000)]
[AMDGPU] SDWA operands should not intersect with potential MIs
Summary:
There should be no intesection between SDWA operands and potential MIs. E.g.:
```
v_and_b32 v0, 0xff, v1 -> src:v1 sel:BYTE_0
v_and_b32 v2, 0xff, v0 -> src:v0 sel:BYTE_0
v_add_u32 v3, v4, v2
```
In that example it is possible that we would fold 2nd instruction into 3rd (v_add_u32_sdwa) and then try to fold 1st instruction into 2nd (that was already destroyed). So if SDWAOperand is also a potential MI then do not apply it.
The previous commit failed on test-suite/Bitcode/simd_ops/AArch64_halide_runtime.bc
because isImmOperandEqual() assumed MO was a register operand and that's not
always true.
Lama Saba [Thu, 18 May 2017 08:11:50 +0000 (08:11 +0000)]
[X86] Replace slow LEA instructions in X86
According to Intel's Optimization Reference Manual for SNB+:
" For LEA instructions with three source operands and some specific situations, instruction latency has increased to 3 cycles, and must
dispatch via port 1:
- LEA that has all three source operands: base, index, and offset
- LEA that uses base and index registers where the base is EBP, RBP,or R13
- LEA that uses RIP relative addressing mode
- LEA that uses 16-bit addressing mode "
This patch currently handles the first 2 cases only.
George Rimar [Thu, 18 May 2017 08:00:01 +0000 (08:00 +0000)]
[lib/Object] - Minor API update for llvm::Decompressor.
I revisited Decompressor API (issue with it was triggered during D32865 review)
and found it is probably provides more then we really need.
Issue was about next method's signature:
Error decompress(SmallString<32> &Out);
It is too strict. At first I wanted to change it to decompress(SmallVectorImpl<char> &Out),
but then found it is still not flexible because sticks to SmallVector.
During reviews was suggested to use templating to simplify code. Patch do that.
Serguei Katkov [Thu, 18 May 2017 06:11:56 +0000 (06:11 +0000)]
[BPI] Reduce the probability of unreachable edge to minimal value greater than 0
The probability of edge coming to unreachable block should be as low as possible.
The change reduces the probability to minimal value greater than zero.
The bug https://bugs.llvm.org/show_bug.cgi?id=32214 show the example when
the probability of edge coming to unreachable block is greater than for edge
coming to out of the loop and it causes incorrect loop rotation.
Please note that with this change the behavior of unreachable heuristic is a bit different
than others. Specifically, before this change the sum of probabilities
coming to unreachable blocks have the same weight for all branches
(it was just split over all edges of this block coming to unreachable blocks).
With this change it might be slightly different but not to much due to probability of
taken branch to unreachable block is really small.
Akira Hatanaka [Thu, 18 May 2017 03:52:29 +0000 (03:52 +0000)]
[ThinLTO] Do not assert when adding a module with a different but
compatible target triple
Currently, an assertion fails in ThinLTOCodeGenerator::addModule when
the target triple of the module being added doesn't match that of the
one stored in TMBuilder. This patch relaxes the constraint and makes
changes to allow target triples that only differ in their version
numbers on Apple platforms, similarly to what r228999 did.
Craig Topper [Thu, 18 May 2017 00:51:39 +0000 (00:51 +0000)]
[Statistics] Add a method to atomically update a statistic that contains a maximum
Summary:
There are several places in the codebase that try to calculate a maximum value in a Statistic object. We currently do this in one of two ways:
The first version reads from MaxNumFoo one time and uncontionally rwrites to it. The second version possibly reads it twice depending on the result of the first compare. But we have no way of knowing if the value was changed by another thread between the reads and the writes.
This patch adds a method to the Statistic object that can ensure that we only store if our value is the max and the previous max didn't change after we read it. If it changed we'll recheck if our value should still be the max or not and try again.
This spawned from an audit I'm trying to do of all places we uses the implicit conversion to unsigned on the Statistics objects. See my previous thread on llvm-dev https://groups.google.com/forum/#!topic/llvm-dev/yfvxiorKrDQ
Craig Topper [Wed, 17 May 2017 23:22:10 +0000 (23:22 +0000)]
[Statistics] Use Statistic::operator+= instead of adding and assigning separately.
I believe this technically fixes a multithreaded race condition in this code. But my primary concern was as part of looking at removing the ability to treat Statistics like a plain unsigned. There are many weird operations on Statistics in the codebase.
This breaks the GISel bot:
http://lab.llvm.org:8080/green/job/Compiler_Verifiers_GlobalISEL/5163/consoleFull#-134276167849ba4694-19c4-4d7e-bec5-911270d8a58c
Sanjay Patel [Wed, 17 May 2017 22:29:40 +0000 (22:29 +0000)]
[InstCombine] handle icmp i1 X, C early to avoid creating an unknown pattern
The missing optimization for xor-of-icmps still needs to be added, but by
being more efficient (not generating unnecessary logic ops with constants)
we avoid the bug.
See discussion in post-commit comments:
https://reviews.llvm.org/D32143
Sanjay Patel [Wed, 17 May 2017 22:15:07 +0000 (22:15 +0000)]
[InstCombine] move icmp bool canonicalizations to helper; NFC
As noted in the post-commit comments in D32143, we should be
catching the constant operand cases sooner to be more efficient
and less likely to expose a missing fold.
Matt Arsenault [Wed, 17 May 2017 21:56:25 +0000 (21:56 +0000)]
AMDGPU: Start defining a calling convention
Partially implement callee-side for arguments and return values.
byval doesn't work properly, and most likely sret or other on-stack
return values most as well.
Kyle Butt [Wed, 17 May 2017 21:54:41 +0000 (21:54 +0000)]
CodeGen: Power: Add lowering for shifts of v1i128.
When legalizing vector operations on vNi128, they will be split to v1i128
because that is a legal type on ppc64, but then the compiler will crash in
selection dag because it fails to select for these operations. This patch fixes
shift operations. Logical shift right and left shift can be performed in the
vector unit, but algebraic shift right requires being split.
Michael Liao [Wed, 17 May 2017 21:48:00 +0000 (21:48 +0000)]
Fix PR33028
- '-verify-mahcineinstrs' starts to complain allocatable live-in physical
registers on non-entry or non-landing-pad basic blocks.
- Refactor the XBEGIN translation to define EAX on a dedicated fallback code
path due to XABORT. Add a pseudo instruction to define EAX explicitly to
avoid add physical register live-in.
Matt Arsenault [Wed, 17 May 2017 21:23:14 +0000 (21:23 +0000)]
AMDGPU: Expand frame indexes to be relative to scratch wave offset
In order for an arbitrary callee to access an object
in a caller's stack frame, the 32-bit offset used as
the private pointer needs to be relative to the kernel's
scratch wave offset register.
Convert to this by finding the difference from the current
stack frame and scaling by the wavefront size.
Bob Haarman [Wed, 17 May 2017 20:46:48 +0000 (20:46 +0000)]
[llvm-pdbdump] in yaml2pdb, generate default output filename if none given
Summary:
llvm-pdbdump yaml2pdb used to fail with a misleading error
message ("An I/O error occurred on the file system") if no output file
was specified. This change adds an assert to PDBFileBuilder to check
that an output file name is specified, and makes llvm-pdbdump generate
an output file name based on the input file name if no output file
name is explicitly specified.
Sanjay Patel [Wed, 17 May 2017 20:27:55 +0000 (20:27 +0000)]
[InstSimplify] handle all icmp i1 X, C in one place; NFCI
We already handled all of the new tests identically, but several
of those went through a lot of unnecessary processing before
getting folded.
Another motivation for grouping these cases together is that
InstCombine needs a similar fold. Currently, it handles the
'not' cases inefficiently which can lead to bugs as described
in the post-commit comments of:
https://reviews.llvm.org/D32143
Zachary Turner [Wed, 17 May 2017 20:23:31 +0000 (20:23 +0000)]
[BinaryStream] Reduce the amount of boiler plate needed to use.
Often you have an array and you just want to use it. With the current
design, you have to first construct a `BinaryByteStream`, and then create
a `BinaryStreamRef` from it. Worse, the `BinaryStreamRef` holds a pointer
to the `BinaryByteStream`, so you can't just create a temporary one to
appease the compiler, you have to actually hold onto both the `ArrayRef`
as well as the `BinaryByteStream` *AND* the `BinaryStreamReader` on top of
that. This makes for very cumbersome code, often requiring one to store a
`BinaryByteStream` in a class just to circumvent this.
At the cost of some added complexity (not exposed to users, but internal
to the library), we can do better than this. This patch allows us to
construct `BinaryStreamReaders` and `BinaryStreamWriters` directly from
source data (e.g. `StringRef`, `MutableArrayRef<uint8_t>`, etc). Not only
does this reduce the amount of code you have to type and make it more
obvious how to use it, but it solves real lifetime issues when it's
inconvenient to hold onto a `BinaryByteStream` for a long time.
The additional complexity is in the form of an added layer of indirection.
Whereas before we simply stored a `BinaryStream*` in the ref, we now store
both a `BinaryStream*` **and** a `std::shared_ptr<BinaryStream>`. When
the user wants to construct a `BinaryStreamRef` directly from an
`ArrayRef` etc, we allocate an internal object that holds ownership over a
`BinaryByteStream` and forwards all calls, and store this in the
`shared_ptr<>`. This also maintains the ref semantics, as you can copy it
by value and references refer to the same underlying stream -- the one
being held in the object stored in the `shared_ptr`.
Dimitry Andric [Wed, 17 May 2017 19:46:49 +0000 (19:46 +0000)]
Reapply part of rL303015, fixing just the DynamicLibaryTest. Add
retrieval of the original argv[0] from the GoogleTest framework, so it
is more likely the correct main executable path is found.
Dimitry Andric [Wed, 17 May 2017 19:33:10 +0000 (19:33 +0000)]
Revert r303015, because it has the unintended side effect of breaking
driver-mode recognition in clang (this is because the sysctl method
always returns one and only one executable path, even for an executable
with multiple links):
Fix DynamicLibraryTest.cpp on FreeBSD and NetBSD
Summary:
After rL301562, on FreeBSD the DynamicLibrary unittests fail, because
the test uses getMainExecutable("DynamicLibraryTests", Ptr), and since
the path does not contain any slashes, retrieving the main executable
will not work.
Reimplement getMainExecutable() for FreeBSD and NetBSD using sysctl(3),
which is more reliable than fiddling with relative or absolute paths.
Also add retrieval of the original argv[] from the GoogleTest framework,
to use as a fallback for other OSes.
Reid Kleckner [Wed, 17 May 2017 18:16:17 +0000 (18:16 +0000)]
Re-land r303274: "[CrashRecovery] Use SEH __try instead of VEH when available"
We have to check gCrashRecoveryEnabled before using __try.
In other words, SEH works too well and we ended up recovering from
crashes in implicit module builds that we weren't supposed to. Only
libclang is supposed to enable CrashRecoveryContext to allow implicit
module builds to crash.
Reid Kleckner [Wed, 17 May 2017 17:02:16 +0000 (17:02 +0000)]
[CrashRecovery] Use SEH __try instead of VEH when available
Summary:
It avoids problems when other libraries raise exceptions. In particular,
OutputDebugString raises an exception that the debugger is supposed to
catch and suppress. VEH kicks in first right now, and that is entirely
incorrect.
Unfortunately, GCC does not support SEH, so I've kept the old buggy VEH
codepath around. We could fix it with SetUnhandledExceptionFilter, but
that is not per-thread, so a well-behaved library shouldn't set it.
Zachary Turner [Wed, 17 May 2017 16:39:06 +0000 (16:39 +0000)]
[CodeView] Simplify the use of visiting type records & streams.
There is often a lot of boilerplate code required to visit a type
record or type stream. The #1 use case is that you have a sequence
of bytes that represent one or more records, and you want to
deserialize each one, switch on it, and call a callback with the
deserialized record that the user can examine. Currently this
requires at least 6 lines of code:
This is done by having the deserialization happen internally inside
of the visitTypeRecord function. Since this is occasionally not
desirable, the function provides a 3rd parameter that can be used
to change this behavior.
Hopefully this can significantly reduce the barrier to entry
to using the visitation infrastructure.
Zachary Turner [Wed, 17 May 2017 15:49:45 +0000 (15:49 +0000)]
[BitVector] Add find_[first,last]_[set,unset]_in.
A lot of code is duplicated between the first_last and the
next / prev methods. All of this code can be shared if they
are implemented in terms of find_first_in(Begin, End) etc,
in which case find_first = find_first_in(0, Size) and find_next
is find_first_in(Prev+1, Size), with similar reductions for
the other methods.
Sanjay Patel [Wed, 17 May 2017 14:21:19 +0000 (14:21 +0000)]
[InstCombine] add isCanonicalPredicate() helper function and use it; NFCI
There should be a slight efficiency improvement from handling icmp/fcmp with one matcher and reducing duplicated code.
The larger motivation is that there are questions about how predicate canonicalization is handled, and the refactoring
should make it easier if we want to change any of that behavior.
1. As noted in the code comment, we've chosen 3 of the 16 FCMP preds as not canonical. Why those 3? It goes back to
rL32751 from what I can tell, but I'm not sure if there's a justification for that rule.
2. We currently do not canonicalize integer select conditions. Should we use the same rule that applies to branches
for selects?
3. We currently do canonicalize some FP select conditions, and those rules would conflict with the rule shown here.
Should one or both be changed?
No-functional-change-intended, but adding tests anyway because there's no coverage for most of the predicates.
The variables MinGPR/MinG8R were not updated properly when resetting the
offsets, which in the included testcase lead to saving the CR register
in the same location as R30.
George Rimar [Wed, 17 May 2017 12:10:51 +0000 (12:10 +0000)]
[DWARF] - Cleanup relocations proccessing.
RelocAddrMap was a pair of <width, address>, where width is relocation size (4/8/x, x < 8),
and width field was never used in code.
Relocations proccessing loop had checks for width field. Does not look like DWARF parser
should do that. There is probably no much sense to validate relocations during proccessing
them in parser.
Patch removes relocation's width relative code from DWARFContext.
Diana Picus [Wed, 17 May 2017 08:57:28 +0000 (08:57 +0000)]
[GlobalISel][TableGen] Fix handling of default operands
When looping through a destination pattern's operands to decide how many
default operands we need to introduce, we used to count the "expanded"
number of operands. So if one default operand would be rendered as 2
values, we'd count it as 2 operands, when in fact it needs to count as
only 1 operand regardless of how many values it expands to.
This turns out to be a problem only in some very specific cases, e.g.
when we have one operand with multiple default values followed by more
operands with default values (see the new test). In such a situation
we'd stop looping before looking at all the operands, and then error out
assuming that we don't have enough default operands to make up the
shortfall.
At the moment this only affects ARM.
The patch removes the loop counting default operands entirely and
assumes that we'll have to introduce values for any default operand that
we find (i.e. we're assuming it cannot be given as a child at all). It
also extracts the code for adding renderers for default operands into a
helper method.
Summary:
Debug info sections, (or non-SHF_ALLOC sections in general) should be
linked as if their load address was zero to emulate the behavior of the
static linker.
This bug was discovered because it was breaking lldb expression evaluation on
linux.
Jonas Paulsson [Wed, 17 May 2017 07:36:03 +0000 (07:36 +0000)]
Make sure -optimize-regalloc=false is used correctly by user.
Don't allow -optimize-regalloc=false with -regalloc given for anything other
than 'fast'. The other register allocators depend on the supporting passes
added by addOptimizedRegAlloc().
Max Kazantsev [Wed, 17 May 2017 04:09:14 +0000 (04:09 +0000)]
[SCEV] Always sort AddRecExprs from different loops by dominance
Sorting of AddRecExprs by loop nesting does not make sense since we only invoke
the CompareSCEVComplexity for AddRecExprs that are used by one SCEV. This
guarantees that there is always a dominance relationship between them. This
patch removes the sorting by nesting which is a dead code in current usage of
this function.
Gor Nishanov [Wed, 17 May 2017 03:09:22 +0000 (03:09 +0000)]
[coroutines] Handle spills before catchswitch
If we need to spill the result of the PHI instruction, we insert the spill after
all of the PHIs and EHPads, however, in a catchswitch block there is no
room to insert the spill. Make room by splitting away catchswitch into a separate
block.
projects / llvm / log