https://reviews.llvm.org/D57178
Now add a hook in TargetPassConfig to query if CSE needs to be
enabled. By default this hook returns false only for O0 opt level but
this can be overridden by the target.
As a consequence of the default of enabled for non O0, a few tests
needed to be updated to not use CSE (by passing in -O0) to the run
line.
reviewed by: arsenm
llvm-svn: 352126
It should be emitted when any floating-point operations (including
calls) are present in the object, not just when calls to printf/scanf
with floating point args are made.
The difference caused by this is very subtle: in static (/MT) builds,
on x86-32, in a program that uses floating point but doesn't print it,
the default x87 rounding mode may not be set properly upon
initialization.
This commit also removes the walk of the types pointed to by pointer
arguments in calls. (To assist in opaque pointer types migration --
eventually the pointee type won't be available.)
That latter implies that it will no longer consider a call like
`scanf("%f", &floatvar)` as sufficient to emit _fltused on its
own. And without _fltused, `scanf("%f")` will abort with error R6002. This
new behavior is unlikely to bite anyone in practice (you'd have to
read a float, and do nothing with it!), and also, is consistent with
MSVC.
Differential Revision: https://reviews.llvm.org/D56548
llvm-svn: 352076
This patch adds a new ReadAdvance definition named ReadInt2Fpu.
ReadInt2Fpu allows x86 scheduling models to accurately describe delays caused by
data transfers from the integer unit to the floating point unit.
ReadInt2Fpu currently defaults to a delay of zero cycles (i.e. no delay) for all
x86 models excluding BtVer2. That means, this patch is only a functional change
for the Jaguar cpu model only.
Tablegen definitions for instructions (V)PINSR* have been updated to account for
the new ReadInt2Fpu. That read is mapped to the the GPR input operand.
On Jaguar, int-to-fpu transfers are modeled as a +6cy delay. Before this patch,
that extra delay was added to the opcode latency. In practice, the insert opcode
only executes for 1cy. Most of the actual latency is actually contributed by the
so-called operand-latency. According to the AMD SOG for family 16h, (V)PINSR*
latency is defined by expression f+1, where f is defined as a forwarding delay
from the integer unit to the fpu.
When printing instruction latency from MCA (see InstructionInfoView.cpp) and LLC
(only when flag -print-schedule is speified), we now need to account for any
extra forwarding delays. We do this by checking if scheduling classes declare
any negative ReadAdvance entries. Quoting a code comment in TargetSchedule.td:
"A negative advance effectively increases latency, which may be used for
cross-domain stalls". When computing the instruction latency for the purpose of
our scheduling tests, we now add any extra delay to the formula. This avoids
regressing existing codegen and mca schedule tests. It comes with the cost of an
extra (but very simple) hook in MCSchedModel.
Differential Revision: https://reviews.llvm.org/D57056
llvm-svn: 351965
For AMDGPU the shift amount is never 64-bit, and
this needs to use a 32-bit shift.
X86 uses i8, but seemed to be hacking around this before.
llvm-svn: 351882
The existing test unintentionally shows that we have prematurely
optimized the shuffle into a vector concat and lost the undef info,
so it is not affected by a basic improvement to
SimplifyDemandedVectorElts.
llvm-svn: 351834
vecbo (insertsubv undef, X, Z), (insertsubv undef, Y, Z) --> insertsubv VecC, (vecbo X, Y), Z
This is another step in generic vector narrowing. It's also a step towards more horizontal op
formation specifically for x86 (although we still failed to match those in the affected tests).
The scalarization cases are also not optimal (we should be scalarizing those), but it's still
an improvement to use a narrower vector op when we know part of the result must be constant
because both inputs are undef in some vector lanes.
I think a similar match but checking for a constant operand might help some of the cases in
D51553.
Differential Revision: https://reviews.llvm.org/D56875
llvm-svn: 351825
For constant bit select patterns, replace one AND with a ANDNP, allowing us to reuse the constant mask. Only do this if the mask has multiple uses (to avoid losing load folding) or if we have XOP as its VPCMOV can handle most folding commutations.
This also requires computeKnownBitsForTargetNode support for X86ISD::ANDNP and X86ISD::FOR to prevent regressions in fabs/fcopysign patterns.
Differential Revision: https://reviews.llvm.org/D55935
llvm-svn: 351819
Similar to horizontal ops on D56777, the sse2 (but not mmx) bit shift ops has local forwarding disabled, adding +1cy to the use latency for the result.
Differential Revision: https://reviews.llvm.org/D57026
llvm-svn: 351817
Similar to horizontal ops on D56777, the vpermilpd/vpermilps variable mask ops has local forwarding disabled, adding +1cy to the use latency for the result.
Differential Revision: https://reviews.llvm.org/D57022
llvm-svn: 351815
When we are inserting 1 "inline" element, and zeroing 2 of the other elements then we can safely commute the insertps source inputs to improve memory folding.
Differential Revision: https://reviews.llvm.org/D56843
llvm-svn: 351807
D56777 added +1cy local forwarding penalty for horizontal operations, but this penalty only affects sse2/xmm variants, the mmx variants don't suffer the penalty.
Confirmed with @andreadb
llvm-svn: 351755
r327630 introduced new write definitions for float/vector loads.
Before that revision, WriteLoad was used by both integer/float (scalar/vector)
load. So, WriteLoad had to conservatively declare a latency to 5cy. That is
because the load-to-use latency for float/vector load is 5cy.
Now that we have dedicated writes for float/vector loads, there is no reason why
we should keep the latency of WriteLoad to 5cy. At the moment, WriteLoad is only
used by scalar integer loads only; we can assume an optimstic 3cy latency for
them.
This patch changes that latency from 5cy to 3cy, and regenerates the affected
scheduling/mca tests.
Differential Revision: https://reviews.llvm.org/D56922
llvm-svn: 351742
This causes a couple of changes in the upgrade tests as signed/unsigned eq/ne are equivalent and we constant fold true/false codes, these changes are the same as what we already do for avx512 cmp/ucmp.
Noticed while cleaning up vector integer comparison costs for PR40376.
llvm-svn: 351697
We were upgrading these to the new style VPCOM/VPCOMU intrinsics (which includes the condition code immediate), but we'll be getting rid of those shortly, so convert these to generics first.
This causes a couple of changes in the upgrade tests as signed/unsigned eq/ne are equivalent and we constant fold true/false codes, these changes are the same as what we already do for avx512 cmp/ucmp.
Noticed while cleaning up vector integer comparison costs for PR40376.
llvm-svn: 351690
These intrinsics can always be replaced with generic integer comparisons without any regression in codegen, even for -O0/-fast-isel cases.
Noticed while cleaning up vector integer comparison costs for PR40376.
A future commit will remove/autoupgrade the existing VPCOM/VPCOMU llvm intrinsics.
llvm-svn: 351688
The existing tests already show a sub-optimal transform,
but this should make it clear that we can't just match
an 'and' op when creating movmsk instructions.
llvm-svn: 351590
This sends these intrinsics through isel in a much more normal way. This should allow addressing mode matching in isel to make better use of the displacement field.
Differential Revision: https://reviews.llvm.org/D56827
llvm-svn: 351570
Previously we used ISD::SHL and ISD::SRL to represent these in SelectionDAG. ISD::SHL/SRL interpret an out of range shift amount as undefined behavior and will constant fold to undef. While the intrinsics are defined to return 0 for out of range shift amounts. A previous patch added a special node for VPSRAV to produce all sign bits.
This was previously believed safe because undefs frequently get turned into 0 either from the constant pool or a desire to not have a false register dependency. But undef is treated specially in some optimizations. For example, its ignored in detection of vector splats. So if the ISD::SHL/SRL can be constant folded and all of the elements with in bounds shift amounts are the same, we might fold it to single element broadcast from the constant pool. This would not put 0s in the elements with out of bounds shift amounts.
We do have an existing InstCombine optimization to use shl/lshr when the shift amounts are all constant and in bounds. That should prevent some loss of constant folding from this change.
Patch by zhutianyang and Craig Topper
Differential Revision: https://reviews.llvm.org/D56695
llvm-svn: 351381
If we're going to generate a new inverted setcc, we should make sure we will be able to remove the old setcc.
Differential Revision: https://reviews.llvm.org/D56765
llvm-svn: 351378
On Jaguar, horizontal adds/subs have local forwarding disable.
That means, we pay a compulsory extra cycle of write-back stage, and the value
is not available until the end of that stage.
This patch changes the latency of horizontal operations by adding an extra
cycle. With this patch, latency numbers now match what is reported by perf.
I plan to send another patch to also 'fix' the latency of shuffle operations (on
Jaguar, local forwarding is disabled for vector shuffles too).
Differential Revision: https://reviews.llvm.org/D56777
llvm-svn: 351366
I was trying to prevent shuffle regressions while matching more horizontal ops
and ended up here:
shuf (extract X, 0), (extract X, 4), Mask --> extract (shuf X, undef, Mask'), 0
The affected tests were added for:
https://bugs.llvm.org/show_bug.cgi?id=34380
This patch won't change the examples in the bug report itself, but we should be
able to extend this to catch more types.
Differential Revision: https://reviews.llvm.org/D56756
llvm-svn: 351346
Summary:
Make recoverfp intrinsic target-independent so that it can be implemented for AArch64, etc.
Refer D53541 for the context. Clang counterpart D56748.
Reviewers: rnk, efriedma
Reviewed By: rnk, efriedma
Subscribers: javed.absar, kristof.beyls, llvm-commits
Differential Revision: https://reviews.llvm.org/D56747
llvm-svn: 351281
We're trying to have the vXi1 types in IR as much as possible. This prevents the need for bitcasts when the producer of the mask was already a vXi1 value like an icmp. The bitcasts can be subject to code motion and interfere with basic block at a time isel in bad ways.
llvm-svn: 351275
In keeping with our general direction of having the vXi1 type present in IR, this patch converts the mask argument for avx512 gather to vXi1. This can avoid k-register to GPR to k-register transitions late in codegen.
I left the existing intrinsics behind because they have many out of tree users such as ISPC. They generate their own code and don't go through the autoupgrade path which only works for bitcode and ll parsing. Ideally we will get them to migrate to target independent intrinsics, but it might be easier for them to migrate to these new intrinsics.
I'll work on scatter and gatherpf/scatterpf next.
Differential Revision: https://reviews.llvm.org/D56527
llvm-svn: 351234
If we're shuffling with a zero vector, then we are better off not doing VECTOR_SHUFFLE(UNPCK()) as we lose track of those zero elements.
We were already doing this for SSSE3 targets as we have PSHUFB, but its worth doing for all targets.
llvm-svn: 351203
compiler identification lines in test-cases.
(Doing so only because it's then easier to search for references which
are actually important and need fixing.)
llvm-svn: 351200
The motivating case for this is shown in the first regression test. We are
transferring to scalar and back rather than just zero-extending with 'vpmovzxdq'.
That's a special-case for a more general pattern as shown here. In all tests,
we're avoiding the vector-scalar-vector moves in favor of vector ops.
We aren't producing optimal shuffle code in some cases though, so the patch is
limited to reduce regressions.
Differential Revision: https://reviews.llvm.org/D56281
llvm-svn: 351198
Summary:
In r345197 ESP and RSP were added to GR32_TC/GR64_TC, allowing them to
be used for tail calls, but this also caused `findDeadCallerSavedReg` to
think they were acceptable targets for clobbering. Filter them out.
Fixes PR40289.
Patch by Geoffry Song!
Reviewed By: rnk
Differential Revision: https://reviews.llvm.org/D56617
llvm-svn: 351146
If we have PSHUFB and we're shuffling with a zero vector, then we are better off not doing VECTOR_SHUFFLE(UNPCK()) as we lose track of those zero elements.
llvm-svn: 351103
add (extractelt (X, 0), extractelt (X, 1)) --> extractelt (hadd X, X), 0
This is the integer sibling to D56011.
There's an additional restriction to only to do this transform in the
case where we don't have extra extracts from the source vector. Without
that, we can fail to match larger horizontal patterns that are more
beneficial than this minimal case. An improvement to the more general
h-op lowering may allow us to remove the restriction here in a follow-up.
llvm-svn: 351093
Following PR39807, the way in which SimplifyCFG hoists common code on
branch paths was fixed in r347782. However this left extra code hanging
around HoistThenElseCodeToIf that wasn't necessary and needlessly
complicated matters -- we no longer need to look up through the 'if'
basic block to find a location for hoisted 'select' insts, we can instead
use the location chosen by applyMergedLocation.
This patch deletes that extra logic, and updates a regression test to
reflect the new logic (selects get the merged location, not a previous
insts location).
Differential Revision: https://reviews.llvm.org/D55272
llvm-svn: 351058
Part of the effort to refactoring frame pointer code generation. We used
to use two function attributes "no-frame-pointer-elim" and
"no-frame-pointer-elim-non-leaf" to represent three kinds of frame
pointer usage: (all) frames use frame pointer, (non-leaf) frames use
frame pointer, (none) frame use frame pointer. This CL makes the idea
explicit by using only one enum function attribute "frame-pointer"
Option "-frame-pointer=" replaces "-disable-fp-elim" for tools such as
llc.
"no-frame-pointer-elim" and "no-frame-pointer-elim-non-leaf" are still
supported for easy migration to "frame-pointer".
tests are mostly updated with
// replace command line args ‘-disable-fp-elim=false’ with ‘-frame-pointer=none’
grep -iIrnl '\-disable-fp-elim=false' * | xargs sed -i '' -e "s/-disable-fp-elim=false/-frame-pointer=none/g"
// replace command line args ‘-disable-fp-elim’ with ‘-frame-pointer=all’
grep -iIrnl '\-disable-fp-elim' * | xargs sed -i '' -e "s/-disable-fp-elim/-frame-pointer=all/g"
Patch by Yuanfang Chen (tabloid.adroit)!
Differential Revision: https://reviews.llvm.org/D56351
llvm-svn: 351049
This pattern:
t33: v8i32 = insert_subvector undef:v8i32, t35, Constant:i64<0>
t21: v16i32 = insert_subvector undef:v16i32, t33, Constant:i64<0>
...shows up in PR33758:
https://bugs.llvm.org/show_bug.cgi?id=33758
...although this patch doesn't make any difference to the final result on that yet.
In the affected tests here, it looks like it just makes RA wiggle. But we might
as well squash this to prevent it interfering with other pattern-matching.
Differential Revision:
https://reviews.llvm.org/D56604
llvm-svn: 351008
This patch takes some of the code from D49837 to allow us to enable ISD::ABS support for all SSE vector types.
Differential Revision: https://reviews.llvm.org/D56544
llvm-svn: 350998
The 128-bit input produces 64-bits of output and fills the upper 64-bits with 0. The mask only applies to the lower elements. But we can't represent this with a vselect like we normally do.
This also avoids the need to have a special X86ISD::SELECT when avx512bw isn't enabled since vselect v8i16 isn't legal there.
Fixes another instruction for PR34877.
llvm-svn: 350994
We still use i8 for the load/store type. So we need to convert to/from i16 to around the mask type.
By doing this we get an i8->i16 extload which we can then pattern match to a KMOVW if the access is aligned.
llvm-svn: 350989
We can't properly represent this with a vselect since the upper elements of the result are supposed to be zeroed regardless of the mask.
This also reuses the new nodes even when the result type fits in 128 bits if the input is q/d and the result is w/b since vselect w/b using k-register condition isn't legal without avx512bw. Currently we're doing this even when avx512bw is enabled, but I might change that.
This fixes some of PR34877
llvm-svn: 350985
Summary:
When legalizing the result of a SELECT_CC node by promoting the
floating-point type, use the promoted-to type rather than the original
type.
Fix PR40273.
Reviewers: efriedma, majnemer
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D56566
llvm-svn: 350951
Previously, we limited this transform to cases where the
extraction into the build vector happens from vectors of
the same type as the build vector, but that's not required.
There's a slight potential regression seen in the AVX512
result for phadd -- we're using the 256-bit flavor of the
instruction now even though the 128-bit subset is sufficient.
The same problem could already be seen in the AVX2 result.
Follow-up patches will attempt to narrow that back down.
llvm-svn: 350928
We were lowering the last step extract_vector_elt to a bitcast+truncate. Change it to use an extract_vector_elt of index 0 instead. Add isel patterns to do the equivalent of what the bitcast would have done. Plus an isel pattern for an any_extend+extract to prevent some regressions.
Finally add a DAG combine to turn v1i1 scalar_to_vector+extract_vector_elt of 0 into an extract_subvector.
This fixes some of the regressions from D350800.
llvm-svn: 350918
This extends to combineVSelectToShrunkBlend to be able to resimplify SHRUNKBLENDS that have already been created.
This should help some of the regressions from D56387
Differential Revision: https://reviews.llvm.org/D56421
llvm-svn: 350875
When we use the partial-matching function on a 128-bit chunk, we must
account for the possibility that we've matched undef halves of the
original source vectors, so the outputs may need to be reset.
This should allow closing PR40243:
https://bugs.llvm.org/show_bug.cgi?id=40243
llvm-svn: 350830
This is a partial fix for:
https://bugs.llvm.org/show_bug.cgi?id=40243
...as seen in the integer test, we still need to correct the result when using the
existing (old) horizontal op matching function because it does not model the way
x86 256-bit horizontal ops return results (each 128-bit half is its own horizontal-op).
A potential follow-up change for that is discussed in the bug report - see also D56490.
This generally duplicates a lot of the existing matching code, but we can't just remove
that without introducing regressions, so the existing code is renamed and used less often.
Follow-ups may try to reduce that overlap.
Differential Revision: https://reviews.llvm.org/D56450
llvm-svn: 350826
Summary:
This pass replaces GR8/GR16/GR32/GR64 with their equivalent sized mask register classes. But VK32/VK64 aren't legal without AVX512BW. Apparently this mostly appears to work if the register coalescer is able to remove the VK32/VK64 register class reference. Or if we don't ever spill it. But there's no guarantee of that.
Another Intel employee managed to trigger a crash due to this with ISPC. Unfortunately, I've lost the test case he sent me at the time. I'm trying to get him to reproduce it for me. I'd like to get this in before 8.0 branches since its a little scary.
The regressions here are unfortunate, but I think we can make some improvements to DAG combine, load folding, etc. to fix them. Just not sure if we can get that done for 8.0.
Fixes PR39741
Reviewers: RKSimon, spatel
Reviewed By: RKSimon
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D56460
llvm-svn: 350800
merging a src register in ToBeUpdated set.
This is to fix PR40061 related with https://reviews.llvm.org/rL339035.
In https://reviews.llvm.org/rL339035, live interval of source pseudo register
in rematerialized copy may be saved in ToBeUpdated set and its update may be
postponed.
In PR40061, %t2 = %t1 is rematerialized and %t1 is added into toBeUpdated set
to postpone its live interval update. After the rematerialization, the live
interval of %t1 is larger than necessary. Then %t1 is merged into %t3 and %t1
gets removed. After the merge, %t3 contains live interval larger than necessary.
Because %t3 is not in toBeUpdated set, its live interval is not updated after
register coalescing and it will break some assumption in regalloc.
The patch requires the live interval of destination register in a merge to be
updated if the source register is in ToBeUpdated.
Differential revision: https://reviews.llvm.org/D55867
llvm-svn: 350586
Aditya Nandakumar