The V_MQSAD_PK_U16_U8, V_QSAD_PK_U16_U8, and V_MQSAD_U32_U8 take more than 1 pass in hardware. For these three instructions, the destination registers must be different than all sources, so that the first pass does not overwrite sources for the following passes.
Differential Revision: https://reviews.llvm.org/D33783
llvm-svn: 304998
In SDAG, we don't expand libcalls with a nobuiltin attribute.
It's not clear if that's correct from the existing code comment:
"Don't do the check if marked as nobuiltin for some reason."
...adding a test here either way to show that there is currently
a different behavior implemented in the CGP-based expansion.
llvm-svn: 304991
The test diff for PowerPC shows we can better optimize if this case is one block.
For x86, there's would be a substantial difference if CGP expansion was enabled because branches are assumed
cheap and SDAG can't optimize across blocks.
Instead of this:
_cmp_eq8:
movq (%rdi), %rax
cmpq (%rsi), %rax
je LBB23_1
## BB#2: ## %res_block
movl $1, %ecx
jmp LBB23_3
LBB23_1:
xorl %ecx, %ecx
LBB23_3: ## %endblock
xorl %eax, %eax
testl %ecx, %ecx
sete %al
retq
We get this:
cmp_eq8:
movq (%rdi), %rcx
xorl %eax, %eax
cmpq (%rsi), %rcx
sete %al
retq
And that matches the optimal codegen that we get from the current expansion in SelectionDAGBuilder::visitMemCmpCall().
If this looks right, then I just need to confirm that vector-sized expansion will work from here, and we can enable
CGP memcmp() expansion for x86. Ie, we'll bypass the power-of-2 special cases currently optimized in SDAG because we
can lower the IR produced here optimally.
Differential Revision: https://reviews.llvm.org/D34005
llvm-svn: 304987
Add a couple of tests to increase coverage for the TableGen'erated code,
in particular for rules where 2 generic instructions may be combined
into a single machine instruction.
llvm-svn: 304971
The zero heuristic assumes that integers are more likely positive than negative,
but this also has the effect of assuming that strcmp return values are more
likely positive than negative. Given that for nonzero strcmp return values it's
the ordering of arguments that determines the sign of the result there's no
reason to assume that's true.
Fix this by inspecting the LHS of the compare and using TargetLibraryInfo to
decide if it's strcmp-like, and if so only assume that nonzero is more likely
than zero i.e. strings are more often different than the same. This causes a
slight code generation change in the spec2006 benchmark 403.gcc, but with no
noticeable performance impact. The intent of this patch is to allow better
optimisation of dhrystone on Cortex-M cpus, but currently it won't as there are
also some changes that need to be made to if-conversion.
Differential Revision: https://reviews.llvm.org/D33934
llvm-svn: 304970
This was discussed in D33338. We have larger pattern-matching ending in a truncate that
we can reduce or remove by handling these smaller patterns first. Further motivation is
that narrower shift ops are easier for value tracking and zext is better than sext.
http://rise4fun.com/Alive/rhh
Name: boolshift
%sext = sext i1 %x to i8
%r = lshr i8 %sext, 7
=>
%r = zext i1 %x to i8
Name: noboolshift
%sext = sext i3 %x to i8
%r = lshr i8 %sext, 7
=>
%sh = lshr i3 %x, 2
%r = zext i3 %sh to i8
Differential Revision: https://reviews.llvm.org/D33879
llvm-svn: 304939
This could be viewed as another shortcoming of the DAGCombiner:
when both operands of a compare are zexted from the same source
type, we should be able to compare the original types.
The effect on PowerPC perf is likely unnoticeable, but there's a
visible regression for x86 if we feed the suboptimal IR for memcmp
expansion to the DAG:
_cmp_eq4_zexted_to_i64:
movl (%rdi), %ecx
movl (%rsi), %edx
xorl %eax, %eax
cmpq %rdx, %rcx
sete %al
_cmp_eq4_better:
movl (%rdi), %ecx
xorl %eax, %eax
cmpl (%rsi), %ecx
sete %al
llvm-svn: 304923
This makes it so that the code quality for CFI checks when compiling
with -O2 and linking with --lto-O0 is similar to that of the rest of
the code.
Reduces the size of a chrome binary built with -O2/--lto-O0 by
about 750KB.
Differential Revision: https://reviews.llvm.org/D33925
llvm-svn: 304921
Changed immediate type for repl.ph from uimm10 to simm10 as per the specs.
Repl.qb still accepts uimm8. Both instructions now mimic the behaviour of
GNU as.
Patch by Stefan Maksimovic.
Differential Revision: https://reviews.llvm.org/D33594
llvm-svn: 304918
If we know that both operands of an unsigned integer vector comparison are non-negative,
then it's safe to directly use a signed-compare-greater-than instruction (the only non-equality
integer vector compare predicate provided by SSE/AVX).
We're intentionally not changing the condition code to signed in order to preserve the
existing transforms that use min/max/psubus below here.
This should solve PR33276:
https://bugs.llvm.org/show_bug.cgi?id=33276
Differential Revision: https://reviews.llvm.org/D33862
llvm-svn: 304909
The clang compiler by default uses FastISel when invoked with -O0, which
is also the default. In that case, passing of -mxgot does not get honored,
i.e. the code path that is to deal with large got is not taken.
Clang produces same output regardless of -mxgot being present or not.
This change checks whether -mxgot is passed as an option, and turns off
FastISel if it is.
Patch by Stefan Maksimovic.
Differential Revision: https://reviews.llvm.org/D33593
llvm-svn: 304906
According to the commit message from r296921, G_MERGE_VALUES and
G_INSERT are to be preferred over G_SEQUENCE. Therefore, stop generating
G_SEQUENCE in the ARM backend and remove the code dealing with it.
This boils down to the code breaking up double values for the soft float
calling convention. Use G_MERGE_VALUES + G_UNMERGE_VALUES instead of
G_SEQUENCE + G_EXTRACT for it. This maps very nicely to VMOVDRR +
VMOVRRD and simplifies the code in the instruction selector.
There's one occurence of G_SEQUENCE left in arm-irtranslator.ll, but
that is part of the target-independent code for translating constant
structs. Therefore, it is beyond the scope of this commit.
llvm-svn: 304902
If there's enough data in fron of it the skipped region would just
become arbitrarily large, and we scan for the CHECK-NOT everywhere.
llvm-svn: 304900
This is identical to the support for the other binary operators:
- widen to s32
- map into GPR
- select ANDrr (via TableGen'erated code)
llvm-svn: 304885
Summary:
This patch updates Triple::isCompatibleWith to make armxx and thumbxx
triples compatible, as long as the subarch, vendor, os, envorionment and
object format match. Thumb/ARM code generation should be controlled
using the thumb-mode per-function target feature rather than by the
triple to allow mixing Thumb and ARM functions.
D33448 updates Clang's codegen to add thumb-mode for all functions with
armxx or thumbxx triples.
Reviewers: echristo, t.p.northover, rafael, kristof.beyls, rengolin, tejohnson
Reviewed By: tejohnson
Subscribers: rinon, eugenis, pcc, srhines, aemerson, mehdi_amini, javed.absar, llvm-commits
Differential Revision: https://reviews.llvm.org/D33287
llvm-svn: 304884
Summary:
Relocations are required for unconditional branches to function symbols with
different execution mode. Without this patch, incorrect branches are
generated for tail calls between functions with different execution
mode.
Reviewers: peter.smith, rafael, echristo, kristof.beyls
Reviewed By: peter.smith
Subscribers: aemerson, javed.absar, llvm-commits
Differential Revision: https://reviews.llvm.org/D33898
llvm-svn: 304882
A few tests in llvm/test/Transforms/Util/PredicateInfo/ are using -reverse-iterate.
The option -reverse-iterate is enabled with +Asserts in usual cases, but it can be turned on/off regardless of LLVM_ENABLE_ASSERTIONS.
I wonder if this were incompatible to https://reviews.llvm.org/D33908 (r304757).
Differential Revision: https://reviews.llvm.org/D33854
llvm-svn: 304851
I'd like to enable CGP memcmp expansion for x86, but the output from CGP would regress the
special cases (memcmp(x,y,N) != 0 for N=1,2,4,8,16,32 bytes) that we already handle.
I'm not sure if we'll actually be able to produce the optimal code given the block-at-a-time
limitation in the DAG. We might have to just avoid those special-cases here in CGP. But
regardless of that, I think this is a win for the more general cases.
http://rise4fun.com/Alive/cbQ
Differential Revision: https://reviews.llvm.org/D33963
llvm-svn: 304849
This patch introduces a new command line option, called brief, to
llvm-dwarfdump. When -brief is used, the attribute forms for the
.debug_info section will not be emitted to output.
Patch by Spyridoula Gravani!
rdar://problem/21474365
Differential Revision: https://reviews.llvm.org/D33867
llvm-svn: 304844
3 of the tests were testing exactly the same thing: memcmp(x, y, 16) != 0.
I changed that to test 4, 7, and 16 bytes, so we can see how those differ.
llvm-svn: 304838
- Add -x <language> option to switch between IR and MIR inputs.
- Change MIR parser to read from stdin when filename is '-'.
- Add a simple mir roundtrip test.
llvm-svn: 304825
Summary:
The patch makes instruction count the highest priority for
LSR solution for X86 (previously registers had highest priority).
Reviewers: qcolombet
Differential Revision: http://reviews.llvm.org/D30562
From: Evgeny Stupachenko <evstupac@gmail.com>
llvm-svn: 304824
Summary:
LVIPrinter pass was previously relying on the LVICache. We now directly call the
the LVI functions which solves the value if the LVI information is not already
available in the cache. This has 2 benefits over the printing of LVI cache:
1. higher coverage (i.e. catches errors) in LVI code when cache value is
invalidated.
2. relies on the core functions, and not dependent on the LVI cache (which may
be scrapped at some point).
It would still catch any cache invalidation errors, since we first go through
the cache.
Reviewers: reames, dberlin, sanjoy
Reviewed by: reames
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D32135
llvm-svn: 304819
The change cleans up and unifies the handling of relocation
entries in WasmObjectWriter. Type index relocation no longer
need to be handled separately.
The only externally visible change should be that type
index relocations are no longer grouped at the end.
Differential Revision: https://reviews.llvm.org/D33918
llvm-svn: 304816
CodeGen uses MO_ExternalSymbol to represent the inline assembly strings.
Empty strings for symbol names appear to be invalid. For now just
special case the output code to avoid hitting an `assert()` in
`printLLVMNameWithoutPrefix()`.
This fixes https://llvm.org/PR33317
llvm-svn: 304815
Patch https://reviews.llvm.org/rL304806 was causing failures in Aarch64
and multiple other targets since the test should be run on X86 only.
Specifying the target triple is not enough. Moving the testcase to the
X86 target directory in LoopIdiom.
llvm-svn: 304809
1. When there is no perfect iteration order, we can't let phi nodes
put themselves in terms of things that come later in the iteration
order, or we will endlessly cycle (the normal RPO algorithm clears the
hashtable to avoid this issue).
2. We are sometimes erasing the wrong expression (causing pessimism)
because our equality says loads and stores are the same.
We introduce an exact equality function and use it when erasing to
make sure we erase only identical expressions, not equivalent ones.
llvm-svn: 304807
