Masked Load / Store Intrinsics - the CodeGen part.
I'm recommiting the codegen part of the patch.
The vectorizer part will be send to review again.
Masked Vector Load and Store Intrinsics.
Introduced new target-independent intrinsics in order to support masked vector loads and stores. The loop vectorizer optimizes loops containing conditional memory accesses by generating these intrinsics for existing targets AVX2 and AVX-512. The vectorizer asks the target about availability of masked vector loads and stores.
Added SDNodes for masked operations and lowering patterns for X86 code generator.
Examples:
<16 x i32> @llvm.masked.load.v16i32(i8* %addr, <16 x i32> %passthru, i32 4 /* align */, <16 x i1> %mask)
declare void @llvm.masked.store.v8f64(i8* %addr, <8 x double> %value, i32 4, <8 x i1> %mask)
Scalarizer for other targets (not AVX2/AVX-512) will be done in a separate patch.
Hal Finkel [Thu, 4 Dec 2014 09:22:28 +0000 (09:22 +0000)]
Restrict somewhat the memory-allocation pointer cmp opt from r223093
Based on review comments from Richard Smith, restrict this optimization from
applying to globals that might resolve lazily to other dynamically-loaded
modules, and also from dynamic allocas (which might be transformed into malloc
calls). In short, take extra care that the compared-to pointer is really
simultaneously live with the memory allocation.
Michael Liao [Thu, 4 Dec 2014 00:56:38 +0000 (00:56 +0000)]
[X86] Restore X86 base pointer after call to llvm.eh.sjlj.setjmp
Commit on
- This patch fixes the bug described in
http://lists.cs.uiuc.edu/pipermail/llvmdev/2013-May/062343.html
The fix allocates an extra slot just below the GPRs and stores the base pointer
there. This is done only for functions containing llvm.eh.sjlj.setjmp that also
need a base pointer. Because code containing llvm.eh.sjlj.setjmp saves all of
the callee-save GPRs in the prologue, the offset to the extra slot can be
computed before prologue generation runs.
Impact at run-time on affected functions is::
- One extra store in the prologue, The store saves the base pointer.
- One extra load after a llvm.eh.sjlj.setjmp. The load restores the base pointer.
Because the extra slot is just above a gap between frame-pointer-relative and
base-pointer-relative chunks of memory, there is no impact on other offset
calculations other than ensuring there is room for the extra slot.
Hal Finkel [Thu, 4 Dec 2014 00:46:20 +0000 (00:46 +0000)]
[PowerPC] 'cc' should be an alias only to 'cr0'
We had mistakenly believed that GCC's 'cc' referred to the entire
condition-code register (cr0 through cr7) -- and implemented this in r205630 to
fix PR19326, but 'cc' is actually an alias only to 'cr0'. This is causing LLVM
to clobber too much with legacy code with inline asm using the 'cc' clobber.
Matt Arsenault [Thu, 4 Dec 2014 00:06:57 +0000 (00:06 +0000)]
Allow target to specify prefix for labels
Use the MCAsmInfo instead of the DataLayout, and allow
specifying a custom prefix for labels specifically. HSAIL
requires that labels begin with @, but global symbols with &.
Philip Reames [Thu, 4 Dec 2014 00:01:48 +0000 (00:01 +0000)]
A few more checks for gc.statepoints in the Verifier
This is simply a grab bag of unrelated checks:
- A statepoint call can't be marked readonly or readnone
- We don't currently support inline asm or varadic target functions. Both could be supported, but don't currently work.
- I forgot to check that the number of call arguments actually matched the wrapped callee in my previous change. Included here.
Hal Finkel [Wed, 3 Dec 2014 23:40:13 +0000 (23:40 +0000)]
[PowerPC] Fix inline asm memory operands not to use r0
On PowerPC, inline asm memory operands might be expanded as 0($r), where $r is
a register containing the address. As a result, this register cannot be r0, and
we need to enforce this register subclass constraint to prevent miscompiling
the code (we'd get this constraint for free with the usual instruction
definitions, but that scheme has no knowledge of how we end up printing inline
asm memory operands, and so here we need to do it 'by hand'). We can accomplish
this within the current address-mode selection framework by introducing an
explicit COPY_TO_REGCLASS node.
Prior to this commit, physical registers defined implicitly were considered free
right after their definition, i.e.. like dead definitions. Therefore, their uses
had to immediately follow their definitions, otherwise the related register may
be reused to allocate a virtual register.
This commit fixes this assumption by keeping implicit definitions alive until
they are actually used. The downside is that if the implicit definition was dead
(and not marked at such), we block an otherwise available register. This is
however conservatively correct and makes the fast register allocator much more
robust in particular regarding the scheduling of the instructions.
Split the set of identified struct types into opaque and non-opaque ones.
The non-opaque part can be structurally uniqued. To keep this to just
a hash lookup, we don't try to unique cyclic types.
Also change the type mapping algorithm to be optimistic about a type
not being recursive and only create a new type when proven to be wrong.
This is not as strong as trying to speculate that we can keep the source
type, but is simpler (no speculation to revert) and more powerfull
than what we had before (we don't copy non-recursive types at least).
I initially wrote this to try to replace the name based type merging.
It is not strong enough to replace it, but is is a useful addition.
With this patch the number of named struct types is a clang lto bootstrap goes
from 49674 to 15986.
Philip Reames [Wed, 3 Dec 2014 19:53:15 +0000 (19:53 +0000)]
Make the Verifier more strict about gc.statepoints
The recently added documentation for statepoints claimed that we checked the parameters of the various intrinsics for validity. This patch adds the code to actually do so. I also removed a couple of redundant checks for conditions which are checked elsewhere in the Verifier and simplified the logic using the helper functions from Statepoint.h.
Erik Eckstein [Wed, 3 Dec 2014 10:39:15 +0000 (10:39 +0000)]
InstCombine: simplify signed range checks
Try to convert two compares of a signed range check into a single unsigned compare.
Examples:
(icmp sge x, 0) & (icmp slt x, n) --> icmp ult x, n
(icmp slt x, 0) | (icmp sgt x, n) --> icmp ugt x, n
Hal Finkel [Wed, 3 Dec 2014 09:37:50 +0000 (09:37 +0000)]
[PowerPC] Print all inline-asm consts as signed numbers
Almost all immediates in PowerPC assembly (both 32-bit and 64-bit) are signed
numbers, and it is important that we print them as such. To make sure that
happens, we change PPCTargetLowering::LowerAsmOperandForConstraint so that it
does all intermediate checks on a signed-extended int64_t value, and then
creates the resulting target constant using MVT::i64. This will ensure that all
negative values are printed as negative values (mirroring what is done in other
backends to achieve the same sign-extension effect).
This came up in the context of inline assembly like this:
"add%I2 %0,%0,%2", ..., "Ir"(-1ll)
where we used to print:
addi 3,3,4294967295
and gcc would print:
addi 3,3,-1
and gas accepts both forms, but our builtin assembler (correctly) does not. Now
we print -1 like gcc does.
While here, I replaced a bunch of custom integer checks with isInt<16> and
friends from MathExtras.h.
Charlie Turner [Wed, 3 Dec 2014 08:12:26 +0000 (08:12 +0000)]
Emit ABI_FP_rounding attribute.
LLVM understands a -enable-sign-dependent-rounding-fp-math codegen option. When
the user has specified this option, the Tag_ABI_FP_rounding attribute should be
emitted with value 1. This option currently does not appear to disable
transformations and optimizations that assume default floating point rounding
behavior, AFAICT, but the intention should be recorded in the build attributes,
regardless of what the compiler actually does with the intention.
When lazy reading a module, the types used in a function will not be visible to
a TypeFinder until the body is read.
This patch fixes that by asking the module for its identified struct types.
If a materializer is present, the module asks it. If not, it uses a TypeFinder.
This fixes pr21374.
I will be the first to say that this is ugly, but it was the best I could find.
Some of the options I looked at:
* Asking the LLVMContext. This could be made to work for gold, but not currently
for ld64. ld64 will load multiple modules into a single context before merging
them. This causes us to see types from future merges. Unfortunately,
MappedTypes is not just a cache when it comes to opaque types. Once the
mapping has been made, we have to remember it for as long as the key may
be used. This would mean moving MappedTypes to the Linker class and having
to drop the Linker::LinkModules static methods, which are visible from C.
* Adding an option to ignore function bodies in the TypeFinder. This would
fix the PR by picking the worst result. It would work, but unfortunately
we are currently quite dependent on the upfront type merging. I will
try to reduce our dependency, but it is not clear that we will be able
to get rid of it for now.
The only clean solution I could think of is making the Module own the types.
This would have other advantages, but it is a much bigger change. I will
propose it, but it is nice to have this fixed while that is discussed.
With the gold plugin, this patch takes the number of types in the LTO clang
binary from 52817 to 49669.
Remove an unnecessary `MDNode::replaceAllUsesWith()`. In the preceding
line, `TheLoop->setLoopID()` visits all backedges and sets the new loop
ID. This sufficiently updates the loop metadata.
Matt Arsenault [Wed, 3 Dec 2014 05:22:35 +0000 (05:22 +0000)]
R600/SI: Remove i1 pseudo VALU ops
Select i1 logical ops directly to 64-bit SALU instructions.
Vector i1 values are always really in SGPRs, with each
bit for each item in the wave. This saves about 4 instructions
when and/or/xoring any condition, and also helps write conditions
that need to be passed in vcc.
This should work correctly now that the SGPR live range
fixing pass works. More work is needed to eliminate the VReg_1
pseudo regclass and possibly the entire SILowerI1Copies pass.
Matt Arsenault [Wed, 3 Dec 2014 05:22:32 +0000 (05:22 +0000)]
R600/SI: Fix suspicious indexing
The loop is over the operands of an instruction, and checks the
register with the sub reg index of the dest register. This probably
meant to be checking the sub reg index of the same operand.
Matt Arsenault [Wed, 3 Dec 2014 05:22:29 +0000 (05:22 +0000)]
R600/SI: Fix live range error hidden by SIFoldOperands
m0 is treated as a virtual register class with a single register
rather than the physical register it really is. This was updating
the live range of the used virtual copy of m0 from the first ds_read
instruction, and leaving the unused copy unchanged. This resulted in a
"Live segment doesn't end at a valid instruction" verifier error because
the erased instructions. Update the live range of the second copy (which
should be dead).
No test since I'm not sure how to trigger this with SIFoldOperands
enabled.
Tom Stellard [Wed, 3 Dec 2014 04:28:32 +0000 (04:28 +0000)]
StructurizeCFG: Use LoopInfo analysis for better loop detection
We were assuming that each back-edge in a region represented a unique
loop, which is not always the case. We need to use LoopInfo to
correctly determine which back-edges are loops.
Nick Lewycky [Wed, 3 Dec 2014 02:45:01 +0000 (02:45 +0000)]
Emit the entry block first and the exit block second, then all the blocks in between afterwards. This is what gcc always does, and some out of tree tools depend on that.
This redefines the `prefix` attribute introduced previously and
introduces a `prologue` attribute. There are a two primary usecases
that these attributes aim to serve,
1. Function prologue sigils
2. Function hot-patching: Enable the user to insert `nop` operations
at the beginning of the function which can later be safely replaced
with a call to some instrumentation facility
3. Runtime metadata: Allow a compiler to insert data for use by the
runtime during execution. GHC is one example of a compiler that
needs this functionality for its tables-next-to-code functionality.
Previously `prefix` served cases (1) and (2) quite well by allowing the user
to introduce arbitrary data at the entrypoint but before the function
body. Case (3), however, was poorly handled by this approach as it
required that prefix data was valid executable code.
Here we redefine the notion of prefix data to instead be data which
occurs immediately before the function entrypoint (i.e. the symbol
address). Since prefix data now occurs before the function entrypoint,
there is no need for the data to be valid code.
The previous notion of prefix data now goes under the name "prologue
data" to emphasize its duality with the function epilogue.
The intention here is to handle cases (1) and (2) with prologue data and
case (3) with prefix data.
References
----------
This idea arose out of discussions[1] with Reid Kleckner in response to a
proposal to introduce the notion of symbol offsets to enable handling of
case (3).
The X86AsmParser intel handling was refactored in r216481, making it
try each different memory operand size to see which one matches.
Operand sizes larger than 80 ("[xyz]mmword ptr") were forgotten, which
led to an "invalid operand" error for code such as:
movdqa [rax], xmm0
Tim Northover [Tue, 2 Dec 2014 23:53:43 +0000 (23:53 +0000)]
AArch64: strengthen Darwin ABI alignment assumptions
A global variable without an explicit alignment specified should be assumed to
be ABI-aligned according to its type, like on other platforms. This allows us
to use better memory operations when accessing it.
Tim Northover [Tue, 2 Dec 2014 23:13:39 +0000 (23:13 +0000)]
AArch64: don't be too greedy when folding :lo12: accesses into mem ops.
This frequently leads to cases like:
ldr xD, [xN, :lo12:var]
add xA, xN, :lo12:var
ldr xD, [xA, #8]
where the ADD would have been needed anyway, and the two distinct addressing
modes can prevent the formation of an ldp. Because of how we handle ADRP
(aggressively forming an ADRP/ADD pseudo-inst at ISel time), this pattern also
results in duplicated ADRP instructions (one on its own to cover the ldr, and
one combined with the add).
Such loops shouldn't be vectorized due to the loops form.
After applying loop-rotate (+simplifycfg) the tests again start to check
what they are intended to check.
Simon Pilgrim [Tue, 2 Dec 2014 22:31:23 +0000 (22:31 +0000)]
[X86][SSE] Keep 4i32 vector insertions in integer domain on SSE4.1 targets
4i32 shuffles for single insertions into zero vectors lowers to X86vzmovl which was using (v)blendps - causing domain switch stalls. This patch fixes this by using (v)pblendw instead.
The updated tests on test/CodeGen/X86/sse41.ll still contain a domain stall due to the use of insertps - I'm looking at fixing this in a future patch.
Hal Finkel [Tue, 2 Dec 2014 22:01:00 +0000 (22:01 +0000)]
[PowerPC] Implement readcyclecounter for PPC32
We've long supported readcyclecounter on PPC64, but it is easier there (the
read of the 64-bit time-base register can be accomplished via a single
instruction). This now provides an implementation for PPC32 as well. On PPC32,
the time-base register is still 64 bits, but can only be read 32 bits at a time
via two separate SPRs. The ISA manual explains how to do this properly (it
involves re-reading the upper bits and looping if the counter has wrapped while
being read).
This requires PPC to implement a custom integer splitting legalization for the
READCYCLECOUNTER node, turning it into a target-specific SDAG node, which then
gets turned into a pseudo-instruction, which is then expanded to the necessary
sequence (which has three SPR reads, the comparison and the branch).
Thanks to Paul Hargrove for pointing out to me that this was still unimplemented.
Lang Hames [Tue, 2 Dec 2014 21:36:24 +0000 (21:36 +0000)]
[AArch64][Stackmaps] Optimize stackmap shadows on AArch64.
Reduce the number of nops emitted for stackmap shadows on AArch64 by counting
non-stackmap instructions up to the next branch target towards the requested
shadow.
projects / llvm / log