This reverts commit r293004 because it broke the buildbots with "unknown CPU"
errors. I tried to fix it in r293026, but that broke on Green Dragon with this
kind of error:
error: expected string not found in input
// CHECK: l{{ +}}df{{ +}}*ABS*{{ +}}{{0+}}{{.+}}preprocessed-input.c{{$}}
^
<stdin>:2:1: note: scanning from here
/Users/buildslave/jenkins/sharedspace/incremental@2/clang-build/tools/clang/test/Frontend/Output/preprocessed-input.c.tmp.o: file format Mach-O 64-bit x86-64
^
<stdin>:2:67: note: possible intended match here
/Users/buildslave/jenkins/sharedspace/incremental@2/clang-build/tools/clang/test/Frontend/Output/preprocessed-input.c.tmp.o: file format Mach-O 64-bit x86-64
I suppose this means that llvm-objdump doesn't support Mach-O, so the test
should indeed check for linux (but not for x86). I'll leave it to someone that
knows better.
llvm-svn: 293032
Summary:
A patch to enable the llvm-xray graph subcommand to color edges and
vertices based on statistics and to annotate vertices with statistics.
Depends on D27243
Reviewers: dblaikie, dberris
Reviewed By: dberris
Subscribers: mgorny, llvm-commits
Differential Revision: https://reviews.llvm.org/D28225
llvm-svn: 293031
Enable the next form (intel style):
"mov <reg64>, <largeImm>"
which is should be available,
where <largeImm> stands for immediates which exceed the range of a singed 32bit integer
Differential Revision: https://reviews.llvm.org/D28988
llvm-svn: 293030
This test broke on a lot of non-x86 buildbots with "unknowm CPU" errors. I don't
see anything platform-specific about this test, and it seems to work fine on ARM
if we just remove the -triple i686 flags from the run line.
llvm-svn: 293026
Conservatively disable sinking and merging inline-asm instructions as doing so
can potentially create arguments that cannot satisfy the inline-asm constraints.
For example, SimplifyCFG used to do the following transformation:
(before)
if.then:
%0 = call i32 asm "rorl $2, $0", "=&r,0,n"(i32 %r6, i32 8)
br label %if.end
if.else:
%1 = call i32 asm "rorl $2, $0", "=&r,0,n"(i32 %r6, i32 6)
br label %if.end
(after)
%.sink = select i1 %tobool, i32 6, i32 8
%0 = call i32 asm "rorl $2, $0", "=&r,0,n"(i32 %r6, i32 %.sink)
This would result in a crash in the backend since only immediate integer operands
are permitted for constraint "n".
rdar://problem/30110806
Differential Revision: https://reviews.llvm.org/D29111
llvm-svn: 293025
clang already emits this with -cl-no-signed-zeros, but codegen
doesn't do anything with it. Treat it like the other fast math
attributes, and change one place to use it.
llvm-svn: 293024
I found root class should be instantiated for variadic tempate to instantiate static member explicitly.
This will fix failures in mingw DLL build.
llvm-svn: 293017
Leave early ifcvt disabled for now since there are some
shader-db regressions.
This causes some immediate improvements, but could be better.
The cost checking that the pass does is based on critical path
length for out of order CPUs which we do not want so it skips out
on many cases we want.
llvm-svn: 293016
Summary:
In this change we introduce the notion of a "flight data recorder" mode
for XRay logging, where XRay logs in-memory first, and write out data
on-demand as required (as opposed to the naive implementation that keeps
logging while tracing is "on"). This depends on D26232 where we
implement the core data structure for holding the buffers that threads
will be using to write out records of operation.
This implementation only currently works on x86_64 and depends heavily
on the TSC math to write out smaller records to the inmemory buffers.
Also, this implementation defines two different kinds of records with
different sizes (compared to the current naive implementation): a
MetadataRecord (16 bytes) and a FunctionRecord (8 bytes). MetadataRecord
entries are meant to write out information like the thread ID for which
the metadata record is defined for, whether the execution of a thread
moved to a different CPU, etc. while a FunctionRecord represents the
different kinds of function call entry/exit records we might encounter
in the course of a thread's execution along with a delta from the last
time the logging handler was called.
While this implementation is not exactly what is described in the
original XRay whitepaper, this one gives us an initial implementation
that we can iterate and build upon.
Reviewers: echristo, rSerge, majnemer
Subscribers: mehdi_amini, llvm-commits, mgorny
Differential Revision: https://reviews.llvm.org/D27038
llvm-svn: 293015
Sometime clang would be supplied -fobjc-arc -f(no)objc-arc-exceptions
and then later disable ARC with -fno-objc-arc, which only negate first
option, but not the latter, resulting usused argument warning. Silence
this warning only when -fno-objc-arc option is present.
Patch by Onha Choe!
llvm-svn: 293014
Looks like our cmake goop for handling .inc->td dependencies doesn't
track the .td files.
This manifests as cmake complaining about missing files since r293009.
Force a rerun to avoid that.
llvm-svn: 293012
loops.
We do this by reconstructing the newly added loops after the unroll
completes to avoid threading pass manager details through all the mess
of the unrolling infrastructure.
I've enabled some extra assertions in the LPM to try and catch issues
here and enabled a bunch of unroller tests to try and make sure this is
sane.
Currently, I'm manually running loop-simplify when needed. That should
go away once it is folded into the LPM infrastructure.
Differential Revision: https://reviews.llvm.org/D28848
llvm-svn: 293011
This surprisingly isn't NFC because there are patterns to select GPR
sub to SUBSWrr (rather than SUBWrr/rs); SUBS is later optimized to
SUB if NZCV is dead. From ISel's perspective, both are fine.
llvm-svn: 293010
Summary:
When we decide that the result of the invoke instruction need to be spilled, we need to insert the spill into a block that is on the normal edge coming out of the invoke instruction. (Prior to this change the code would insert the spill immediately after the invoke instruction, which breaks the IR, since invoke is a terminator instruction).
In the following example, we will split the edge going into %cont and insert the spill there.
```
%r = invoke double @print(double 0.0) to label %cont unwind label %pad
cont:
%0 = call i8 @llvm.coro.suspend(token none, i1 false)
switch i8 %0, label %suspend [i8 0, label %resume
i8 1, label %cleanup]
resume:
call double @print(double %r)
```
Reviewers: majnemer
Reviewed By: majnemer
Subscribers: mehdi_amini, llvm-commits, EricWF
Differential Revision: https://reviews.llvm.org/D29102
llvm-svn: 293006
This patch adds support for codegen of 'target teams' on the host.
This combined directive has two captured statements, one for the
'teams' region, and the other for the 'parallel'.
This target teams region is offloaded using the __tgt_target_teams()
call. The patch sets the number of teams as an argument to
this call.
Reviewers: ABataev
Differential Revision: https://reviews.llvm.org/D29084
llvm-svn: 293005
Summary:
Clang appears to always use name as specified on the command
line, whereas gcc uses the name as specified in the linemarker at the
first line when compiling a preprocessed source. This results mismatch
between two compilers in FILE symbol table entry. This patch makes clang
to resemble gcc's behavior in finding the original source file name and
use it as an input file name.
Even with this patch, values of FILE symbol table entry may still be
different because clang uses dirname+basename for the entry whlie gcc
uses basename only. I'll write a patch for that once this patch is
committed.
Reviewers: dblaikie, inglorion
Reviewed By: inglorion
Subscribers: inglorion, aprantl, bruno
Differential Revision: https://reviews.llvm.org/D28796
llvm-svn: 293004
This patch adds support for codegen of 'target teams' on the host.
This combined directive has two captured statements, one for the
'teams' region, and the other for the 'parallel'.
This target teams region is offloaded using the __tgt_target_teams()
call. The patch sets the number of teams as an argument to
this call.
Reviewers: ABataev
Differential Revision: https://reviews.llvm.org/D29084
llvm-svn: 293001
Summary:
This lets you select which sort of spilling you want, either s[0:1] or 64-bit loads from s[0:1].
Patch By: Dave Airlie
Reviewers: nhaehnle, arsenm, tstellarAMD
Reviewed By: arsenm
Subscribers: mareko, llvm-commits, kzhuravl, wdng, yaxunl, tony-tye
Differential Revision: https://reviews.llvm.org/D25428
llvm-svn: 293000
This patch adds support for the Spmd construct 'target parallel' on the
NVPTX device. This involves ignoring the num_threads clause on the device
since the number of threads in this combined construct is already set on
the host through the call to __tgt_target_teams().
Reviewers: ABataev
Differential Revision: https://reviews.llvm.org/D29083
llvm-svn: 292999
The num_threads-clause on the combined directive applies to the
'parallel' region of this construct. We modify the NumThreadsClause
class to capture the clause expression within the 'target' region.
The offload runtime call for 'target parallel' is changed to
__tgt_target_teams() with 1 team and the number of threads set by
this clause or a default if none.
Reviewers: ABataev
Differential Revision: https://reviews.llvm.org/D29082
llvm-svn: 292997
There was a bug here where we were using p0 instead of s32 for the
selector type in the landingpad. Instead of hardcoding these types we
should get the types from the landingpad instruction directly.
Note that we replicate an assert from SDAG here to only support
two-valued landingpads.
llvm-svn: 292995
for CPU_SUBTYPE_ARM_V7S and CPU_SUBTYPE_ARM_V7K.
For these two cpusubtypes they should default to a cortex-a7 CPU
to give proper disassembly without a -mcpu= flag.
rdar://27431703
llvm-svn: 292993
Summary: This commit adds an option to set PC to the entry point of the file loaded using "target module load" command. In D28804, Greg asked me to separate this part under a different option.
Reviewers: clayborg
Reviewed By: clayborg
Subscribers: lldb-commits
Differential Revision: https://reviews.llvm.org/D28944
llvm-svn: 292989
Summary: LLDB was using packet size advertised by the target as the max memory size to write in one go. It is wrong because packets have other overhead apart from memory payload. Also memory transferred through 'm' and 'M' packets needs 2 bytes in packet to transfer 1 of memory.
Reviewers: clayborg
Reviewed By: clayborg
Subscribers: lldb-commits
Differential Revision: https://reviews.llvm.org/D28808
llvm-svn: 292987
The sequence like this:
v_cmpx_le_f32_e32 vcc, 0, v0
s_branch BB0_30
s_cbranch_execnz BB0_30
; BB#29:
exp null off, off, off, off done vm
s_endpgm
BB0_30:
; %endif110
is likely wrong. The s_branch instruction will unconditionally jump
to BB0_30 and the skip block (exp done + endpgm) inserted for
performing the kill instruction will never be executed. This results
in a GPU hang with Star Ruler 2.
The s_branch instruction is added during the "Control Flow Optimizer"
pass which seems to re-organize the basic blocks, and we assume
that SI_KILL_TERMINATOR is always the last instruction inside a
basic block. Thus, after inserting a skip block we just go to the
next BB without looking at the subsequent instructions after the
kill, and the s_branch op is never removed.
Instead, we should remove the unconditional out branches and let
skip the two instructions if the exec mask is non-zero.
This patch fixes the GPU hang and doesn't introduce any regressions
with "make check".
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=99019
Patch by Samuel Pitoiset <samuel.pitoiset@gmail.com>
llvm-svn: 292985
This switches to the workaround that HSA defaults to
for the mesa path.
This should be applied to the 4.0 branch.
Patch by Vedran Miletić <vedran@miletic.net>
llvm-svn: 292982
Diana Picus