Matt Arsenault [Fri, 24 Jun 2016 06:30:11 +0000 (06:30 +0000)]
AMDGPU: Cleanup subtarget handling.
Split AMDGPUSubtarget into amdgcn/r600 specific subclasses.
This removes most of the static_casting of the basic codegen
classes everywhere, and tries to restrict the features
visible on the wrong target.
Some buildbots are complaining about a .s file under test/ that was
inadvertently created by a test earlier today, and is still hanging
around. I'll undo this change in ~1hr (or whenever the bots are done
getting rid of said file).
BitcodeWriter: Remove redundant (and incorrect) check for whether to emit module summary.
The function name Module::empty() is slightly misleading in that it
only tests for the presence of functions in the module. However we
still want to emit the module summary if the module contains only
global variables or aliases. The presence of such entities can be
determined simply by checking the summary directly, as we are doing
below.
[CFLAA] Propagate StratifiedAttrs in interproc. analysis.
This patch also has a refactor that kills StratifiedAttr, and leaves us
with StratifiedAttrs, because having both was mildly redundant.
This patch makes us correctly handle stratified attributes when doing
interprocedural analysis. It also adds another attribute, AttrCaller,
which acts like AttrUnknown. We can filter out AttrCaller values when
during interprocedural analysis, since the caller should have
information about what arguments it's passing to its callee.
Vedant Kumar [Fri, 24 Jun 2016 00:34:51 +0000 (00:34 +0000)]
[llvm-cov] Add SourceNames to SourceCoverageViews, NFC
A SourceName can be a file or a function. It makes sense to attach this
information to a SourceCoverageView, seeing as views (1) already point
to the text corresponding to the relevant source code and (2) are
already used to render that text along with the SourceNames.
This is a nice cleanup which is independent of the upcoming html patch.
While we're at it, document the fields in SourceCoverageView.
Vedant Kumar [Fri, 24 Jun 2016 00:34:48 +0000 (00:34 +0000)]
[llvm-cov] Rename SourceCoverageView::LineCoverageInfo to LineCoverageStats, NFC
Pull LineCoverageInfo out of SourceCoverageView and rename it so that it
doesn't conflict with another class of the same name in
CoverageSummaryInfo.h.
This cuts down on the amount of code we have to move into a `protected`
section of SourceCoverageView for the upcoming html patch. It also makes
the code a bit clearer: having two LineCoverageInfo's is strange.
Chandler Carruth [Thu, 23 Jun 2016 22:51:14 +0000 (22:51 +0000)]
[LCG] Make the name of an SCC include more of the functions in it.
This makes it much easier to debug issues when the logging contains the
name of the SCC. It requires to create a temporary string, but for
logging and debugging uses that seems fine. I've added logic to try to
output all the function names with an elipsis if there are too many.
This was helpful fro me in debugging issues with the new pass manager.
Chris Bieneman [Thu, 23 Jun 2016 22:36:31 +0000 (22:36 +0000)]
[yaml2macho] Removing asserts in favor of explicit yaml parse error
32-bit Mach headers don't have reserved fields. When generating the
mapping for 32-bit headers leaving off the reserved field will result in
parse errors if the field is present in the yaml.
Added a CHECK-NOT line to ensure that mach_header.yaml isn't adding a
reserved field, and a test to ensure that the parser error gets hit with
32-bit headers.
Chris Bieneman [Thu, 23 Jun 2016 22:07:21 +0000 (22:07 +0000)]
[CMake] Add LLVM runtimes directory
Summary:
There are a few LLVM projects that produce runtime libraries. Ideally
runtime libraries should be built differently than other projects,
specifically they should be built using the just-built toolchain.
There is support for building compiler-rt in this way from the clang
build. Moving this logic into the LLVM build is interesting because it
provides a simpler way to extend the just-built toolchain to include
LLD and the LLVM object file tools.
Once this functionality is better fleshed out and tested we’ll want to
encapsulate it in a module that can be used for clang standalone
builds, and we’ll want to make it the default way to build compiler-rt.
With this patch applied there is no immediate change in the build.
Moving compiler-rt out from llvm/projects into llvm/runtimes enables
the functionality.
This code has a few improvements over the method provided by
LLVM_BUILD_EXTERNAL_COMPILER_RT. Specifically the sub-ninja command is
always invoked, so changes to compiler-rt source files will get built
properly, so this patch can be used for iterative development with
just-built tools.
This first patch only works with compiler-rt. Support for other
runtime projects will be coming in follow-up patches.
Kyle Butt [Thu, 23 Jun 2016 21:38:49 +0000 (21:38 +0000)]
Codegen: LICM Remove check for exactly 1 register def.
When considering whether to split an instruction with a memory operand
into an explicit load and a register-based instruction, we currently
check that the resulting instruction has exactly 1 def. This prevents 2
important LICM optimizations: compares with memory operands, and double
indirect calls. All the tests and the test-suite pass without the check.
My guess as to original intent is to limit the additional register pressure
created by the new instruction, but given that we only split out a single
register, it is already limited.
The licm-dominance test now checks actual memory loads for hoisting instead of
undef, and it tests compares.
hoist-invariant-load.ll now checks for 2 hoists, the intended hoist, and a bonus
from calling a got-relative function in a loop.
Some of the bots running GCC 4.7 seem to be having trouble with lambdas
that explicitly capture `this`. Relevant-looking bug:
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=53137
Anna Thomas [Thu, 23 Jun 2016 20:22:22 +0000 (20:22 +0000)]
InstCombine rule to fold trunc when value available
Summary:
This instcombine rule folds away trunc operations that have value available from a prior load or store.
This kind of code can be generated as a result of GVN widening the load or from source code as well.
Aaron Ballman [Thu, 23 Jun 2016 19:02:09 +0000 (19:02 +0000)]
Default to using the Unicode version of Win32 APIs instead of the ANSI version. This helps to catch instances where a developer accidentally forgets to explicitly specify which version of the API to use and accidentally winds up failing to support non-ASCII characters properly.
[CFLAA] Use better interprocedural function summaries.
Previously, we just unified any arguments that seemed to be related to
each other. With this patch, we now respect dereference levels, etc.
which should make us substantially more accurate. Proper handling of
StratifiedAttrs will be done in a later patch.
[X86] Extract HiPE prologue constants into metadata
X86FrameLowering::adjustForHiPEPrologue() contains a hard-coded offset
into an Erlang Runtime System-internal data structure (the PCB). As the
layout of this data structure is prone to change, this poses problems
for maintaining compatibility.
To address this problem, the compiler can produce this information as
module-level named metadata. For example (where P_NSP_LIMIT is the
offending offset):
Sanjay Patel [Thu, 23 Jun 2016 17:41:59 +0000 (17:41 +0000)]
[ValueTracking] simplify logic in ComputeNumSignBits (NFCI)
This was noted in http://reviews.llvm.org/D21610 . The previous code
predated the use of APInt ( http://reviews.llvm.org/rL47654 ), so it
had to account for the fixed width of uint64_t.
Now that we're using the variable width APInt, we can remove some
complexity.
Pablo Barrio [Thu, 23 Jun 2016 16:53:49 +0000 (16:53 +0000)]
[ARM] Lower (select_cc k k (select_cc ~k ~k x)) into (SSAT l_k x)
Summary:
SSAT saturates an integer, making sure that its value lies within
an interval [-k, k]. Since the constant is given to SSAT as the
number of bytes set to one, k + 1 must be a power of 2, otherwise
the optimization is not possible. Also, the select_cc must use <
and > respectively so that they define an interval.
Artur Pilipenko [Thu, 23 Jun 2016 15:25:09 +0000 (15:25 +0000)]
Remangle intrinsics names when types are renamed
This is a fix for the problem mentioned in "LTO and intrinsics mangling" llvm-dev mail thread:
http://lists.llvm.org/pipermail/llvm-dev/2016-April/098387.html
Aaron Ballman [Thu, 23 Jun 2016 14:45:54 +0000 (14:45 +0000)]
Explicitly specify the ANSI version of these Win32 APIs. While these are seemingly unrelated changes, they are all NFC because we currently default to the ANSI versions of the APIs when building for Windows. This simply makes the ANSI usage explicit.
[LoopUnrollAnalyzer] Fix a bug in UnrolledInstAnalyzer::visitLoad.
When simplifying a load we need to make sure that the type of the
simplified value matches the type of the instruction we're processing.
In theory, we can handle casts here as we deal with constant data, but
since it's not implemented at the moment, we at least need to bail out.
Hal Finkel [Thu, 23 Jun 2016 13:46:39 +0000 (13:46 +0000)]
Allow DeadStoreElimination to track combinations of partial later wrties
DeadStoreElimination can currently remove a small store rendered unnecessary by
a later larger one, but could not remove a larger store rendered unnecessary by
a series of later smaller ones. This adds that capability.
It works by keeping a map, which is used as an effective interval map, for each
store later overwritten only partially, and filling in that interval map as
more such stores are discovered. No additional walking or aliasing queries are
used. In the map forms an interval covering the the entire earlier store, then
it is dead and can be removed. The map is used as an interval map by storing a
mapping between the ending offset and the beginning offset of each interval.
I discovered this problem when investigating a performance issue with code like
this on PowerPC:
#include <complex>
using namespace std;
complex<float> bar(complex<float> C);
complex<float> foo(complex<float> C) {
return bar(C)*C;
}
which produces this:
define void @_Z4testSt7complexIfE(%"struct.std::complex"* noalias nocapture sret %agg.result, i64 %c.coerce) {
entry:
%ref.tmp = alloca i64, align 8
%tmpcast = bitcast i64* %ref.tmp to %"struct.std::complex"*
%c.sroa.0.0.extract.shift = lshr i64 %c.coerce, 32
%c.sroa.0.0.extract.trunc = trunc i64 %c.sroa.0.0.extract.shift to i32
%0 = bitcast i32 %c.sroa.0.0.extract.trunc to float
%c.sroa.2.0.extract.trunc = trunc i64 %c.coerce to i32
%1 = bitcast i32 %c.sroa.2.0.extract.trunc to float
call void @_Z3barSt7complexIfE(%"struct.std::complex"* nonnull sret %tmpcast, i64 %c.coerce)
%2 = bitcast %"struct.std::complex"* %agg.result to i64*
%3 = load i64, i64* %ref.tmp, align 8
store i64 %3, i64* %2, align 4 ; <--- ***** THIS SHOULD NOT BE HERE ****
%_M_value.realp.i.i = getelementptr inbounds %"struct.std::complex", %"struct.std::complex"* %agg.result, i64 0, i32 0, i32 0
%4 = lshr i64 %3, 32
%5 = trunc i64 %4 to i32
%6 = bitcast i32 %5 to float
%_M_value.imagp.i.i = getelementptr inbounds %"struct.std::complex", %"struct.std::complex"* %agg.result, i64 0, i32 0, i32 1
%7 = trunc i64 %3 to i32
%8 = bitcast i32 %7 to float
%mul_ad.i.i = fmul fast float %6, %1
%mul_bc.i.i = fmul fast float %8, %0
%mul_i.i.i = fadd fast float %mul_ad.i.i, %mul_bc.i.i
%mul_ac.i.i = fmul fast float %6, %0
%mul_bd.i.i = fmul fast float %8, %1
%mul_r.i.i = fsub fast float %mul_ac.i.i, %mul_bd.i.i
store float %mul_r.i.i, float* %_M_value.realp.i.i, align 4
store float %mul_i.i.i, float* %_M_value.imagp.i.i, align 4
ret void
}
the problem here is not just that the i64 store is unnecessary, but also that
it blocks further backend optimizations of the other uses of that i64 value in
the backend.
In the future, we might want to add a special case for handling smaller
accesses (e.g. using a bit vector) if the map mechanism turns out to be
noticeably inefficient. A sorted vector is also a possible replacement for the
map for small numbers of tracked intervals.
Daniel Sanders [Thu, 23 Jun 2016 12:42:53 +0000 (12:42 +0000)]
[mips] Don't derive the default ABI from the CPU in the backend.
Summary:
The backend has no reason to behave like a driver and should generally do
as it's told (and error out if it can't) instead of trying to figure out
what the API user meant. The default ABI is still derived from the arch
component as a concession to backwards compatibility.
API-users that previously passed an explicit CPU and a triple that was
inconsistent with the CPU (e.g. mips-linux-gnu and mips64r2) may get a
different ABI to what they got before. However, it's expected that there
are no such users on the basis that CodeGen has been asserting that the
triple is consistent with the selected ABI for several releases. API-users
that were consistent or passed '' or 'generic' as the CPU will see no
difference.
Daniel Sanders [Thu, 23 Jun 2016 10:54:09 +0000 (10:54 +0000)]
[mips][ias] Integers are not registers.
Summary:
When parseAnyRegister() encounters a symbol alias, it parses integers and adds
a corresponding expression to the operand list. This is clearly wrong since the
only operands that parseAnyRegister() should be accepting are registers.
It's not clear why this code was added and there are no test cases that cover
it. I think it might be leftover from when searchSymbolAlias() was more widely
used.
Diana Picus [Thu, 23 Jun 2016 09:49:56 +0000 (09:49 +0000)]
[llc] Remove exit-on-error flag (PR27759)
This flag was introduced in r269655 with the new diagnostic handler for llc. Its
purpose was to keep the old behavior for some of the tests that didn't recover
well after an error. Those tests have been fixed, so now it's safe to remove the
flag entirely.
Simon Dardis [Thu, 23 Jun 2016 09:22:11 +0000 (09:22 +0000)]
[misched] Extend scheduler to handle unsupported features
Currently isComplete = 1 requires that every instruction must
be described, declared unsupported or marked as having no
scheduling information for a processor.
For some backends such as MIPS, this requirement entails
long regex lists of instructions that are unsupported.
This patch teaches Tablegen to skip over instructions that
are associated with unsupported feature when checking if the
scheduling model is complete.
Diana Picus [Thu, 23 Jun 2016 09:19:16 +0000 (09:19 +0000)]
[AMDGPU] Remove exit-on-error in test (PR27761)
The exit-on-error flag was necessary in order to avoid an assertion when
handling DYNAMIC_STACKALLOC nodes in SelectionDAGLegalize.
We can avoid the assertion by creating some dummy nodes. This enables us to
remove the exit-on-error flag on the first 2 run lines (SI), but on the third
run line (R600) we would run into another assertion when trying to reserve
indirect registers. This patch also replaces that assertion with an early exit
from the function.
Jonas Paulsson [Thu, 23 Jun 2016 08:13:20 +0000 (08:13 +0000)]
[IfConversion] Bugfix: Don't use undef flag while adding use operands.
IfConversion used to always add the undef flag when adding a use operand
on a newly predicated instruction. This would be an operand for the register
being conditionally redefined. Due to the undef flag, the liveness of this
register prior to the predicated instruction would get lost.
This patch changes this so that such use operands are added only when the
register is live, without the undef flag.
Reviewed by Quentin Colombet.
http://reviews.llvm.org/D209077
Diana Picus [Thu, 23 Jun 2016 07:47:35 +0000 (07:47 +0000)]
[ARM] Do not test for CPUs, use SubtargetFeatures (Part 1). NFCI
This is a cleanup commit similar to r271555, but for ARM.
The end goal is to get rid of the isSwift / isCortexXY / isWhatever methods.
Since the ARM backend seems to have quite a lot of calls to these methods, I
intend to submit 5-6 subtarget features at a time, instead of one big lump.
Sagar Thakur [Thu, 23 Jun 2016 06:39:35 +0000 (06:39 +0000)]
[LLVM][MIPS] Introduce 64 bit atomic check in CheckAtomic.cmake
Patch by Nitesh Jain.
Summary: On some target like MIPS32 we need to explicitly link atomic library for 64 bit atomic operations. This module then can be used in LLDB (http://reviews.llvm.org/D20464) or Libcxx (http://reviews.llvm.org/D16613) for explicitly link to atomic library.
Reid Kleckner [Wed, 22 Jun 2016 23:23:08 +0000 (23:23 +0000)]
Prune some includes from headers and sink some inline functions
MCSymbol.h shouldn't pull in MCAssembler.h, just MCFragment.h.
MCLinkerOptimizationHint.h shouldn't need MCMachObjectWriter.h. The
rest is fixing the fallout.
Chris Bieneman [Wed, 22 Jun 2016 22:19:08 +0000 (22:19 +0000)]
[MachO] Finish moving fat header swap functions to MachO.h
This is a follow-up to r273479. At the time I wrote r273479 I didn't connect the dots that the functions I was adding had to exist somewhere. Turns out, they do. This finishes moving the functions to MachO.h.
Existing MachO fat header tests like test/tools/llvm-readobj/Inputs/macho-universal-archive.x86_64.i386 execute this code.
Sanjoy Das [Wed, 22 Jun 2016 22:16:51 +0000 (22:16 +0000)]
[ImplicitNullChecks] Hoist trivial depdendencies if possible
When trying to convert a loading instruction into a FAULTING_LOAD, we
sometimes face code like this:
if %R10 is not null:
%R9<def> = MOV32ri Immediate
%R9<def, tied> = AND32rm %R9, 0x20(%R10)
else:
goto TRAP
In these cases we would like to use the AND32rm instruction as the
faulting operation by hoisting the "depedency" def-ing %R9 also above
the control flow, transforming the program into:
projects / llvm / log