Dan Liew [Thu, 3 Sep 2015 18:43:56 +0000 (18:43 +0000)]
Try to clarify the semantics of fptrunc
* ``the value cannot fit within the destination type`` is ambiguous.
It could mean overflow, underflow (not in the IEEE-754 sense) or a
result that cannot be exactly represented and requires rounding or it
could mean some combination of these. The semantics now state it means
overflow **only**.
* Using "truncation" in the semantics is very misleading given that it
doesn't necessarily truncate (i.e. round to zero). For example on
x86_64 with SSE2 this is currently mapped to cvtsd2ss instruction
who's rounding behaviour is dependent on the MXCSR register which
is usually set to round to nearest even by default. The semantics
now state that the rounding mode is undefined.
Karl Schimpf [Thu, 3 Sep 2015 18:06:44 +0000 (18:06 +0000)]
Allow global address space forward decls using IDs in .ll files.
Summary:
This fixes bugzilla bug 24656. Fixes the case where there is a forward
reference to a global variable using an ID (i.e. @0). It does this by
passing the address space of the initializer pointer for which the
forward referenced global is used.
This prevents MC clients from getting COFF.h, which conflicts with
winnt.h macros. Also a minor IWYU cleanup. Now the only public headers
including COFF.h are in Object, and they actually need it.
Karl Schimpf [Thu, 3 Sep 2015 15:41:37 +0000 (15:41 +0000)]
Fix SEGV in InlineAsm::ConstraintInfo::Parse.
Summary:
Fixes bug 24646. Previous code was not checking if an index into a vector
was valid, resulting in a SEGV. Fixed by assuming the construct can't
be parsed when given this input.
Karl Schimpf [Thu, 3 Sep 2015 15:41:34 +0000 (15:41 +0000)]
Fix SEGV in InlineAsm::ConstraintInfo::Parse.
Fixes bug 24646. Previous code was not checking if an index into a vector
was valid, resulting in a SEGV. Fixed by assuming the construct can't
be parsed when given this input.
check for fastness before merging in DAGCombiner::MergeConsecutiveStores()
Use and check the 'IsFast' optional parameter to TLI.allowsMemoryAccess() any time
we have a merged access candidate. Without this patch, we were generating unaligned
16-byte (SSE) memops for x86 targets where those accesses are slow.
This change was mentioned in:
http://reviews.llvm.org/D10662 and
http://reviews.llvm.org/D10905
Summary:
This intrinsic can be used to extract a pointer to the exception caught by
a given catchpad. Its argument has token type and must be a `catchpad`.
Also clarify ExtendingLLVM documentation regarding overloaded intrinsics.
Summary:
Add a `cleanupendpad` instruction, used to mark exceptional exits out of
cleanups (for languages/targets that can abort a cleanup with another
exception). The `cleanupendpad` instruction is similar to the `catchendpad`
instruction in that it is an EH pad which is the target of unwind edges in
the handler and which itself has an unwind edge to the next EH action.
The `cleanupendpad` instruction, similar to `cleanupret` has a `cleanuppad`
argument indicating which cleanup it exits. The unwind successors of a
`cleanuppad`'s `cleanupendpad`s must agree with each other and with its
`cleanupret`s.
Update WinEHPrepare (and docs/tests) to accomodate `cleanupendpad`.
Summary:
This patch introduces a side table in Merge Functions to
efficiently remove functions from the function set when functions
they refer to are merged. Previously these functions would need to
be compared lg(N) times to find the appropriate FunctionNode in the
tree to defer. With the recent determinism changes, this comparison
is more expensive. In addition, the removal function would not always
actually remove the function from the set (i.e. after remove(F),
there would sometimes still be a node in the tree which contains F).
With these changes, these functions are properly deferred, and so more
functions can be merged. In addition, when there are many merged
functions (and thus more deferred functions), there is a speedup:
[libFuzzer] deprecate the -tokens flag. This was a bad idea because the corpus with this flag contains encrypted inputs, not the real inputs, which complicates interoperation with other fuzzers. Instead we'll need to implement AFL dictionary support
Ahmed Bougacha [Wed, 2 Sep 2015 23:25:39 +0000 (23:25 +0000)]
[X86] Require 32-byte alignment for 32-byte VMOVNTs.
We used to accept (and even test, and generate) 16-byte alignment
for 32-byte nontemporal stores, but they require 32-byte alignment,
per SDM. Found by inspection.
Instead of hardcoding 16 in the patfrag, check for natural alignment.
Also fix the autoupgrade and the various tests.
Also, use explicit -mattr instead of -mcpu: I stared at the output
several minutes wondering why I get 2x movntps for the unaligned
case (which is the ideal output, but needs some work: see FIXME),
until I remembered corei7-avx implies +slow-unaligned-mem-32.
Ahmed Bougacha [Wed, 2 Sep 2015 22:27:38 +0000 (22:27 +0000)]
[X86] Cleanup nontemporal fragments. NFCI.
We can chain other fragments to avoid repeating conditions.
This also fixes a potential bug (that realistically can't happen),
where we would match indexed nontemporal stores for i32/i64.
Philip Reames [Wed, 2 Sep 2015 22:25:07 +0000 (22:25 +0000)]
[RewriteStatepointsForGC] Bugfix for change 246133
Fix a bug in change 246133. I didn't handle the case where we had a cycle in the use graph and could add an instruction we were about to erase back on to the worklist. Oddly, I have not been able to write a small test case for this, even with the AssertingVH added. I have confirmed the basic theory for the fix on a large failing example, but all attempts to reduce that to something appropriate for a test case have failed.
Hal Finkel [Wed, 2 Sep 2015 21:03:28 +0000 (21:03 +0000)]
[PowerPC] Cleanup cost model for unaligned vector loads/stores
I'm adding a regression test to better cover code generation for unaligned
vector loads and stores, but there's no functional change to the code
generation here. There is an improvement to the cost model for unaligned vector
loads and stores, mostly for QPX (for which we were not previously accounting
for the permutation-based loads), and the cost model implementation is cleaner.
Piotr Padlewski [Wed, 2 Sep 2015 19:59:53 +0000 (19:59 +0000)]
Constant propagation after hiting llvm.assume
After hitting @llvm.assume(X) we can:
- propagate equality that X == true
- if X is icmp/fcmp (with eq operation), and one of operand
is constant we can change all variables with constants in the same BasicBlock
Benjamin Kramer [Wed, 2 Sep 2015 19:52:23 +0000 (19:52 +0000)]
[RemoveDuplicatePHINodes] Start over after removing a PHI.
This makes RemoveDuplicatePHINodes more effective and fixes an assertion
failure. Triggering the assertions requires a DenseSet reallocation
so this change only contains a constructive test.
I'll explain the issue with a small example. In the following function
there's a duplicate PHI, %4 and %5 are identical. When this is found
the DenseSet in RemoveDuplicatePHINodes contains %2, %3 and %4.
after RemoveDuplicatePHINodes runs the function looks like this. %3 has
changed and is now identical to %2, but RemoveDuplicatePHINodes never
saw this.
If the DenseSet does a reallocation now it will reinsert all
keys and stumble over %3 now having a different hash value than it had
when inserted into the map for the first time. This change clears the
set whenever a PHI is deleted and starts the progress from the
beginning, allowing %3 to be deleted and avoiding inconsistent DenseSet
state. This potentially has a negative performance impact because
it rescans all PHIs, but I don't think that this ever makes a difference
in practice.
This patch defines 'unpredictable' metadata. This metadata can be used to signal to the optimizer
or backend that a branch or switch is unpredictable, and therefore, it's probably better to not
split a compound predicate into multiple branches such as in CodeGenPrepare::splitBranchCondition().
This was discussed in:
https://llvm.org/bugs/show_bug.cgi?id=23827
Dependent patches to alter codegen and expose this in clang to follow.
test: Only warn about missing substitutions for required tools
Every time lit is invoked, I get warnings like so:
lit.py: lit.cfg:286: note: Did not find llvm-go in /Users/bogner/build/llvm/./bin
lit.py: lit.cfg:286: note: Did not find Kaleidoscope-Ch3 in /Users/bogner/build/llvm/./bin
Since these tools are only built in certain configs, these warnings
are superfluous. Change it so that we only warn about tools that are
built in all configs.
James Molloy [Wed, 2 Sep 2015 17:25:25 +0000 (17:25 +0000)]
[ValueTracking] Look through casts when both operands are casts.
We only looked through casts when one operand was a constant. We can also look through casts when both operands are non-constant, but both are in fact the same cast type. For example:
Hal Finkel [Wed, 2 Sep 2015 16:52:37 +0000 (16:52 +0000)]
[PowerPC] Don't always consider P8Altivec-only masks in LowerVECTOR_SHUFFLE
LowerVECTOR_SHUFFLE needs to decide whether to pass a vector shuffle off to the
TableGen-generated matching code, and it does this by testing the same
predicates used by the TableGen files. Unfortunately, when we added new
P8Altivec-only predicates, we started universally testing them in
LowerVECTOR_SHUFFLE, and if then matched when targeting a system prior to a P8,
we'd end up with a selection failure.
Reapply r246012 [dsymutil] Emit real dSYM companion binaries.
With a fix for big endian machines. Thanks to Daniel Sanders for the debugging!
Original commit message:
The binaries containing the linked DWARF generated by dsymutil are not
standard relocatable object files like emitted did previsously. They should be
dSYM companion files, which means they have a different file type in the
header, but also a couple other peculiarities:
- they contain the segments and sections from the original binary in their
load commands, but not the actual contents. This means they get an address
and a size, but their offset is always 0 (but these are not virtual sections)
- they also conatin all the defined symbols from the original binary
This makes MC a really bad fit to emit these kind of binaries. The approach
that was used in this patch is to leverage MC's section layout for the
debug sections, but to use a replacement for MachObjectWriter that lives
in MachOUtils.cpp. Some of the low-level helpers from MachObjectWriter
were reused too.
[x86] fix allowsMisalignedMemoryAccesses() for 8-byte and smaller accesses
This is a continuation of the fix from:
http://reviews.llvm.org/D10662
and discussion in:
http://reviews.llvm.org/D12154
Here, we distinguish slow unaligned SSE (128-bit) accesses from slow unaligned
scalar (64-bit and under) accesses. Other lowering (eg, getOptimalMemOpType)
assumes that unaligned scalar accesses are always ok, so this changes
allowsMisalignedMemoryAccesses() to match that behavior.
Summary:
Add the necessary plumbing so that llvm_token_ty can be used as an
argument/return type in intrinsic definitions and correspondingly require
TokenTy in function types. TokenTy is an opaque type that has no target
lowering, but can be used in machine-independent intrinsics. It is
required for the upcoming llvm.eh.padparam intrinsic.
James Molloy [Wed, 2 Sep 2015 10:15:39 +0000 (10:15 +0000)]
[LV] Don't bail to MiddleBlock if a runtime check fails, bail to ScalarPH instead
We were bailing to two places if our runtime checks failed. If the initial overflow check failed, we'd go to ScalarPH. If any other check failed, we'd go to MiddleBlock. This caused us to have to have an extra PHI per induction and reduction as the vector loop's exit block was not dominated by its latch.
There's no need to have this behavior - if we just always go to ScalarPH we can get rid of a bunch of complexity.
James Molloy [Wed, 2 Sep 2015 10:15:05 +0000 (10:15 +0000)]
[LV] Never widen an induction variable.
There's no need to widen canonical induction variables. It's just as efficient to create a *new*, wide, induction variable.
Consider, if we widen an indvar, then we'll have to truncate it before its uses anyway (1 trunc). If we create a new indvar instead, we'll have to truncate that instead (1 trunc) [besides which IndVars should go and clean up our mess after us anyway on principle].
James Molloy [Wed, 2 Sep 2015 10:14:54 +0000 (10:14 +0000)]
[LV] Switch to using canonical induction variables.
Vectorized loops only ever have one induction variable. All induction PHIs from the scalar loop are rewritten to be in terms of this single indvar.
We were trying very hard to pick an indvar that already existed, even if that indvar wasn't canonical (didn't start at zero). But trying so hard is really fruitless - creating a new, canonical, indvar only results in one extra add in the worst case and that add is trivially easy to push through the PHI out of the loop by instcombine.
If we try and be less clever here and instead let instcombine clean up our mess (as we do in many other places in LV), we can remove unneeded complexity.
Optimization for Gather/Scatter with uniform base
Vector 'getelementptr' with scalar base is an opportunity for gather/scatter intrinsic to generate a better sequence.
While looking for uniform base, we want to use the scalar base pointer of GEP, if exists.
Move createEliminateAvailableExternallyPass earlier in the pass pipeline
to save running many ModulePasses on available external functions that
are thrown away anyhow.
David Majnemer [Tue, 1 Sep 2015 23:46:11 +0000 (23:46 +0000)]
[MC] Generate a timestamp for COFF object files
The MS incremental linker seems to inspect the timestamp written into
the object file to determine whether or not it's contents need to be
considered. Failing to set the timestamp to a date newer than the
executable will result in the object file not participating in
subsequent links. To ameliorate this, write the current time into the
object file's TimeDateStamp field.
Ahmed Bougacha [Tue, 1 Sep 2015 21:56:00 +0000 (21:56 +0000)]
[ARM] Don't abort on variable-idx extractelt in ReconstructShuffle.
The code introduced in r244314 assumed that EXTRACT_VECTOR_ELT only
takes constant indices, but it does accept variables.
Bail out for those: we can't use them, as the shuffles we want to
reconstruct do require constant masks.
David Majnemer [Tue, 1 Sep 2015 21:23:58 +0000 (21:23 +0000)]
[MC] Add support for generating COFF CRCs
COFF sections are accompanied with an auxiliary symbol which includes a
checksum. This checksum used to be filled with just zero but this seems
to upset LINK.exe when it is processing a /INCREMENTAL link job.
Instead, fill the CheckSum field with the JamCRC of the section
contents. This matches MSVC's behavior.
This fixes PR19666.
N.B. A rather simple implementation of JamCRC is given. It implements
a byte-wise calculation using the method given by Sarwate. There are
implementations with higher throughput like slice-by-eight and making
use of PCLMULQDQ. We can switch to one of those techniques if it turns
out to be a significant use of time.
rename "slow-unaligned-mem-under-32" to slow-unaligned-mem-16" (NFCI)
This is a follow-on suggested by:
http://reviews.llvm.org/D12154 ( http://reviews.llvm.org/rL245729 )
http://reviews.llvm.org/D10662 ( http://reviews.llvm.org/rL245075 )
This makes the attribute name match most of the existing lowering logic
and regression test expectations.
But the current use of this attribute is inconsistent; see the FIXME
comment for "allowsMisalignedMemoryAccesses()". That change will
result in functional changes and should be coming soon.
Ahmed Bougacha [Tue, 1 Sep 2015 16:23:45 +0000 (16:23 +0000)]
[AArch64] Lower READCYCLECOUNTER using MRS PMCCTNR_EL0.
This matches the ARM behavior. In both cases, the register is part
of the optional Performance Monitors extension, so, add the feature,
and enable it for the A-class processors we support.
David Majnemer [Tue, 1 Sep 2015 16:19:03 +0000 (16:19 +0000)]
[MC] Allow MCObjectWriter's output stream to be swapped out
There are occasions where it is useful to consider the entirety of the
contents of a section. For example, compressed debug info needs the
entire section available before it can compress it and write it out.
The compressed debug info scenario was previously implemented by
mirroring the implementation of writeSectionData in the ELFObjectWriter.
Instead, allow the output stream to be swapped on demand. This lets
callers redirect the output stream to a more convenient location before
it hits the object file.
projects / llvm / log