This was promoting booleans to i32 to perform a comparison against
them to feed to a select condition. Just use the booleans
directly. This produces the same final code, since the combiner is
unable to undo the mess this creates. I untangled this logic when I
ported this code to GlobalISel, so port the cleanups back.
Extract the existing code from getInstructionThroughput into
TTImpl::getUserCost. The duplicated code in the AMDGPU backend has
also been removed.
Differential Revision: https://reviews.llvm.org/D81448
Add the remaining arithmetic opcodes into the generic implementation
of getUserCost and then call this from getInstructionThroughput. Most
of the backends have been modified to return the base implementation
for cost kinds other RecipThroughput. The outlier here is AMDGPU
which already uses getArithmeticInstrCost for all the cost kinds.
This change means that most of the opcodes can be removed from that
backends implementation of getUserCost.
Differential Revision: https://reviews.llvm.org/D80992
It was annoying enough that every custom lowering needed to set the
insert point, but this was made worse since now these all needed to be
updated to setInstrAndDebugLoc. Consolidate these so every
legalization action has the right insert position by default.
This should fix dropping debug info in every custom AMDGPU
legalization.
Summary:
Make use of both the - (1) clustered bytes and (2) cluster length, to decide on
the max number of mem ops that can be clustered. On an average, when loads
are dword or smaller, consider `5` as max threshold, otherwise `4`. This heuristic
is purely based on different experimentation conducted, and there is no analytical
logic here.
Reviewers: foad, rampitec, arsenm, vpykhtin
Reviewed By: foad, rampitec
Subscribers: kzhuravl, jvesely, wdng, nhaehnle, yaxunl, dstuttard, tpr, Anastasia, t-tye, hiraditya, kerbowa, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D81085
The annoying behavior where the output is different due to the
legality check struck again, plus the subtarget predicate wasn't
really correctly set for DS FP atomics.
Some of the FP min/max instructions seem to be in the gfx6/gfx7
manuals, but IIRC this might have been one of the cases where the
manual got ahead of the actual hardware support, but I've left these
as-is for now since the assembler tests seem to expect them.
The current set is an incomprehensible mess riddled with ordering
hacks for various limitations in the legalizer at the time of writing,
many of which have been fixed. This takes a very small step in
correcting this.
The core first change is to start checking for fully legal cases
first, rather than trying to figure out all of the actions that could
need to be performed. It's recommended to check the legal cases first
for faster legality checks in the common case. This still has a table
listing some common cases, but it needs measuring whether this really
helps or not.
More significantly, stop trying to allow any arbitrary type with a
legal bitwidth as a legal memory type, and start using the bitcast
legalize action for them. Allowing loads of these weird vector types
produced new burdens we don't need for handling all of the
legalization artifacts. Unlike the SelectionDAG handling, this is
still not casting 64 or 16-bit element vectors to 32-bit
vectors. These cases should still be handled by increasing/decreasing
the number of 16-bit elements. This is primarily to fix 8-bit element
vectors.
Another change is to stop trying to handle the load-widening based on
a higher alignment. We should still do this, but the way it was
handled wasn't really correct. We really need to modify the MMO's size
at the same time, and not just increase the result type. The
LegalizerHelper does not do this, and I think this would really
require a separate WidenMemory action (or to add a memory action
payload to the LegalizeMutation). These will now fail to legalize.
The structure of the legalizer rules makes writing concise rules here
difficult. It would be easier if the same function could answer the
query the query, and report the action to perform at the same
time. Instead these two are split into distinct predicate and action
functions. This is mostly tolerable for other cases, but the
load/store rules get pretty complicated so it's difficult to keep two
versions of these functions in sync.
Global TableGen let override blocks are pretty dangerous and override
any local special cases. In this case, the broader HasFlatGlobalInsts
was overriding the more specific predicate for
FeatureAtomicFaddInsts. Make sure HasFlatGlobalInsts is implied by
FeatureAtomicFaddInsts, and make sure the right predicate is used.
One issue with independently setting the subtarget features on
incompatible targets is all of the encoding families do not define all
opcodes. This will hit an assert on gfx10 for example, since we set
the encoding independently based on the generation and not based on a
feature.
This may be missing a few overrides to set it off still in some
special cases. Since the flags set during selection should now be
reliably preserved, this should not change codegen for non-strictfp
functions.
This wasn't getting much value from the DAG or depth arguments, since
it's only called on the frame index root nodes. FrameIndexes can also
only return a scalar value, so it also didn't need DemandedElts.
Casts and intrinsics are now handled by the default implementation
of getUserCost, so remove them from the backends switch statement.
https://reviews.llvm.org/D80994
Explicitly set the exec mask for SGPR spills and reloads.
This fixes a bug where SGPR spills to memory could be incorrect
if the exec mask was 0 (or differed between spill and reload).
Additionally pack scalar subregisters (upto 16/32 per VGPR),
so that the majority of scalar types can be spilt or reloaded
with a simple memory access. This should amortize some of the
additional overhead of manipulating the exec mask.
Reviewed By: arsenm
Differential Revision: https://reviews.llvm.org/D80282
These are scalar instructions that change vector instructions, so they
should not be executed without any active lanes.
The implementation of -amdgpu-skip-threshold also seem to be backwards
from expected, since decreasing it prevents removal.
The pass which infers when it's legal to load a global address space
as SMRD was only considering amdgpu_kernel, and ignoring the shader
entry type calling conventions.
In this awkward case, we have to emit custom pseudo-constrained FP
wrappers. InstrEmitter concludes that since a mayRaiseFPException
instruction had a chain, it can't add nofpexcept.
Test deferred until mayRaiseFPException is really set on everything.
Summary:
While clustering mem ops, AMDGPU target needs to consider number of clustered bytes
to decide on max number of mem ops that can be clustered. This patch adds support to pass
number of clustered bytes to target mem ops clustering logic.
Reviewers: foad, rampitec, arsenm, vpykhtin, javedabsar
Reviewed By: foad
Subscribers: MatzeB, kzhuravl, jvesely, wdng, nhaehnle, yaxunl, dstuttard, tpr, t-tye, hiraditya, javed.absar, kerbowa, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D80545
This is a custom inserter because it was less work than teaching
tablegen a way to indicate that it is sometimes OK to have a no side
effect instruction in the output of a side effecting pattern.
The asm is needed to look like a read of the mode register to prevent
it from being deleted. However, there seems to be a bug where the mode
register def instructions are moved across the asm sideeffect by the
post-RA scheduler.
Another oddity is the immediate is formatted differently between
s_denorm_mode and s_round_mode.
Tweak a few constant expressions involving numbers::pi etc to avoid
rounding errors. NFCI though it's possible some of these will now be
more accurate in the last bit.
I get confused by a lot of the predicate names here, since I would
assume they apply to vectors as well. Rename to reflect they only
apply to scalars.
Also add a few predicates AMDGPU uses that should be generally useful.
Also add any() to complement all. I've wanted to use this a few times
but then worked around it not being there.
If this mask only clears bits in the low 32-bit half of a flat
pointer, these bits are always preserved in the result address
space. If the high bits are modified, they may need to be preserved
for some kind of user pointer tagging.
Summary:
PHIs result register class is set to VGPR or SGPR depending on the cross block value divergence.
In some cases uniform PHI need to be converted to return VGPR to prevent the oddnumber of moves values from VGPR to SGPR and back.
PHI should certainly return VGPR if it has at least one VGPR input. This change adds the exception.
We don't want to convert uniform PHI to VGPRs in case the only VGPR input is a VGPR to SGPR COPY and definition od the
source VGPR in this COPY is move immediate.
bb.0:
%0:vgpr_32 = V_MOV_B32_e32 0, implicit $exec
%2:sreg_32 = .....
bb.1:
%3:sreg_32 = PHI %1, %bb.3, %2, %bb.1
S_BRANCH %bb.3
bb.3:
%1:sreg_32 = COPY %0
S_BRANCH %bb.2
Reviewers: rampitec
Reviewed By: rampitec
Subscribers: arsenm, kzhuravl, jvesely, wdng, nhaehnle, yaxunl, dstuttard, tpr, t-tye, hiraditya, kerbowa, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D80434
This one is slightly odd since it counts as an address expression,
which previously could never fail. Allow the existing TTI hook to
return the value to use, and re-use it for handling how to handle
ptrmask.
Handles the no-op addrspacecasts for AMDGPU. We could probably do
something better based on analysis of the mask value based on the
address space, but leave that for now.
OpenMP emits these for some reason, so handle them. Assume these use
4096 bytes by default, with a flag to override this. Also change the
related stack assumption for calls to have a flag.
We do not have register classes for all possible vector
sizes, so round it up for extract vector element.
Also fixes selection of G_MERGE_VALUES when vectors are
not a power of two.
This has required to refactor getRegSplitParts() in way
that it can handle not just power of two vectors.
Ideally we would like RegSplitParts to be generated by
tablegen.
Differential Revision: https://reviews.llvm.org/D80457
This is the groundwork required to implement strictfp. For now, this
should be NFC for regular instructoins (many instructions just gain an
extra use of a reserved register). Regalloc won't rematerialize
instructions with reads of physical registers, but we were suffering
from that anyway with the exec reads.
Should add it for all the related FP uses (possibly with some
extras). I did not add it to either the gpr index mode instructions
(or every single VALU instruction) since it's a ridiculous feature
already modeled as an arbitrary side effect.
Also work towards marking instructions with FP exceptions. This
doesn't actually set the bit yet since this would start to change
codegen. It seems nofpexcept is currently not implied from the regular
IR FP operations. Add it to some MIR tests where I think it might
matter.
This will enable selecting non-entry block allocas. Skip the SP write
check in the base isSchedulingBoundary implementation to preserve the
previous scheduling behavior and avoid test churn. It's apparently for
compile time reasons, but if we were to use this more work would be
needed since in some of the failing tests, we seem to incorrectly get
hazard nops inserted.
All 3 passes that change instruction encodings were dropping MI
flags. This avoids scheduling regressions caused by setting
mayRaiseFPExceptions on FP instructions for non-strictfp functions.
The vector equivalent has backwards operands, but the scalar version
does not. The passes that use these hooks aren't enabled by default,
so this doesn't really change anything.
I'm guessing this was a holdover from when 0 was an invalid stack
pointer, but surprised nobody has discovered this before.
Also don't allow offset folding for -1 pointers, since it looks weird
to partially fold this.
I consider this to be a hack, since we probably should not mark any
16-bit extract as legal, and require all extracts to be done on
multiples of 32. There are quite a few more battles to fight in the
legalizer for sub-dword vectors, so just select this for now so we can
pass OpenCL conformance without crashing.
Also fix the same assert for G_INSERTs. Unlike G_EXTRACT there's not a
trivial way to select this so just fail on it.
Confusingly, these were unrelated and had different semantics. The
G_PTR_MASK instruction predates the llvm.ptrmask intrinsic, but has a
different format. G_PTR_MASK only allows clearing the low bits of a
pointer, and only a constant number of bits. The ptrmask intrinsic
allows an arbitrary mask. Replace G_PTR_MASK to match the intrinsic.
Only selects the cases that look like the old instruction. More work
is needed to select the general case. Also new legalization code is
still needed to deal with the case where the incoming mask size does
not match the pointer size, which has a specified behavior in the
langref.
This is currently missing most of the hard parts to lower correctly,
so disable it for now. This fixes at least one OpenCL conformance test
and allows it to pass with fallback. Hide this behind an option for
now.
Recommitting most of the remaining changes from
259eb619ff, but excluding the call to
getUserCost from getInstructionThroughput. Though there's still no
test changes, I doubt that this is an NFC...
With the two getIntrinsicInstrCosts folded into one, now fold in the
scalar/code-size orientated getIntrinsicCost. The remaining scalar
intrinsics were memcpy, cttz and ctlz which now have special handling
in the BasicTTI implementation.
This had required a change in the AMDGPU backend for fabs as it
should always be 'free'. I've also changed the X86 backend to return
the BaseT implementation when the CostKind isn't RecipThroughput.
Differential Revision: https://reviews.llvm.org/D80012
Summary: 'A' constraint requires an immediate int or fp constant that can be inlined in an instruction encoding.
Reviewers: arsenm, rampitec
Differential Revision: https://reviews.llvm.org/D78494
Replace TargetMachine.h include with forward declaration and CodeGen.h include in AMDGPU.h.
Exposes a couple of implicit dependencies that require additional forward declarations/includes.
Unlike SelectionDAGBuilder, IRTranslator omits the unconditional
branch in fallthrough cases. Confusingly, the control flow pseudos
function in the opposite way the intrinsics are used, and the branch
targets always need to be swapped. We're inverting the target blocks,
so we need to figure out the old fallthrough block and insert a branch
to the original unconditional branch target.
This only affects assembly and -filetype=asm codegen of PAL metadata.
Differential Revision: https://reviews.llvm.org/D78860
Change-Id: I7b822e1917bf7b403486820d31afc483be207652
Promote alloca to vector before SROA and loop unroll. If we manage
to eliminate allocas before unroll we may choose to unroll less.
Differential Revision: https://reviews.llvm.org/D80386
It was set in total vector size while the idea was to limit
a number of instructions. Now it started to work with doubles
and thresholds needs to be updated.
Differential Revision: https://reviews.llvm.org/D80322
With the two getIntrinsicInstrCosts folded into one, now fold in the
scalar/code-size orientated getIntrinsicCost. This involved sinking
cost of the TTIImpl into the base implementation, as it performs no
target checks. The opcodes remaining were memcpy, cttz and ctlz which
now have special handling in the BasicTTI implementation.
getInstructionThroughput can now directly return the result of
getUserCost.
This had required a change in the AMDGPU backend for fabs and its
always 'free'. I've also changed the X86 backend to return '1' for
any intrinsic when the CostKind isn't RecipThroughput.
Though this intended to be a non-functional change, there are many
paths being combined here so I would be very surprised if this didn't
have an effect.
Differential Revision: https://reviews.llvm.org/D80012
Even though series of cmd/cndmask can produce quite a lot of
code that is still better than a loop. In case of doubles we
would even produce two loops.
Differential Revision: https://reviews.llvm.org/D80032
Combine the two API calls into one by introducing a structure to hold
the relevant data. This has the added benefit of moving the boiler
plate code for arguments and flags, into the constructors. This is
intended to be a non-functional change, but the complicated web of
logic involved here makes it very hard to guarantee.
Differential Revision: https://reviews.llvm.org/D79941
We have the getNegatibleCost/getNegatedExpression to evaluate the cost and negate the expression.
However, during negating the expression, the cost might change as we are changing the DAG,
and then, hit the assertion if we negated the wrong expression as the cost is not trustful anymore.
This patch is target to remove the getNegatibleCost to avoid the out of sync with getNegatedExpression,
and check the cost during negating the expression. It also reduce the duplicated code between
getNegatibleCost and getNegatedExpression. And fix the crash for the test in D76638
Reviewed By: RKSimon, spatel
Differential Revision: https://reviews.llvm.org/D77319
Relying on any MachineFunction state in the MachineFunctionInfo
constructor is hazardous, because the construction time is unclear and
determined by the first use. The function may be only partially
constructed, which is part of why we have many of these hacky string
attributes to track what we need for ABI lowering.
For SelectionDAG, all stack objects are created up-front before
calling convention lowering so stack objects are visible at
construction time. For GlobalISel, none of the IR function has been
visited yet and the allocas haven't been added to the MachineFrameInfo
yet. This should fix failing to set flat_scratch_init in GlobalISel
when needed.
This pass really needs to be turned into some kind of analysis, but I
haven't found a nice way use one here.
This should be directly implied from the register class, and there's
no need to special case live ins here. This was getting the wrong
answer for the queue ptr argument in callable functions, since it's
not an explicit IR argument and is always uniform.
Fixes not using scalar loads for the aperture in addrspacecast
lowering, and any other places that use implicit SGPR arguments.
This was assumed to be a simple move, and interpreting the immediate
modifier operand as a materialized immediate. Apparently the SDWA pass
never produces these, but GlobalISel does emit these for some vector
shuffles.
When the callee requires a dynamic stack realignment,
it is not possible to correcty access the incoming
stack arguments using the stack pointer. We reserve a
base pointer in such cases to access the function arguments
inside the callee. The base pointer will hold the incoming
stack pointer value before any kind of delta added to it.
Reviewed By: arsenm, scott.linder
Differential Revision: https://reviews.llvm.org/D78811
Along the lines of D77454 and D79968. Unlike loads and stores, the
default alignment is getPrefTypeAlign, to match the existing handling in
various places, including SelectionDAG and InstCombine.
Differential Revision: https://reviews.llvm.org/D80044
Summary:
When spilling in the entry function we should be able to borrow
StackPtrOffsetReg as a last resort. This restores behaviour
removed in D75138, and fixes failures when shaders use all
SGPRs, VGPRs and spill in the entry function.
Reviewers: scott.linder, arsenm, tpr
Reviewed By: scott.linder, arsenm
Subscribers: qcolombet, foad, kzhuravl, jvesely, wdng, nhaehnle, yaxunl, dstuttard, t-tye, hiraditya, kerbowa, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D79776
This reverts commit 454de99a6f.
The problem was that one of the ctor arguments of CallAnalyzer was left
to be const std::function<>&. A function_ref was passed for it, and then
the ctor stored the value in a function_ref field. So a std::function<>
would be created as a temporary, and not survive past the ctor
invocation, while the field would.
Tested locally by following https://github.com/google/sanitizers/wiki/SanitizerBotReproduceBuild
Original Differential Revision: https://reviews.llvm.org/D79917
This is D77454, except for stores. All the infrastructure work was done
for loads, so the remaining changes necessary are relatively small.
Differential Revision: https://reviews.llvm.org/D79968
Summary:
Replacing uses of std::function pointers or refs, or Optional, to
function_ref, since the usage pattern allows that. If the function is
optional, using a default parameter value (nullptr). This led to a few
parameter reshufles, to push all optionals to the end of the parameter
list.
Reviewers: davidxl, dblaikie
Subscribers: arsenm, jvesely, nhaehnle, eraman, hiraditya, haicheng, kerbowa, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D79917
We need to use it to handle <16 x double> indirect indexes
in the AMDGPU BE.
The only visible change from adding it is in ARM cost model.
To me it looks reasonable. With doubling a vector size it
quadruples the cost up to the size 8 and then it did only
double it. Now it also quadruples, which seems a logical
progression to me.
Actual AMDGPU code is to follow, this is a common part, plus
load/store legalization in the AMDGPU BE not to break what
works now.
Differential Revision: https://reviews.llvm.org/D79952
Enable clausing of memory loads on gfx10 by adding a new pass to insert
the s_clause instructions that mark the start of each hard clause.
Differential Revision: https://reviews.llvm.org/D79792
SelectMOVRELOffset prevents peeling of a constant from an index
if final base could be negative. isBaseWithConstantOffset() succeeds
if a value is an "add" or "or" operator. In case of "or" it shall
be an add-like "or" which never changes a sign of the sum given a
non-negative offset. I.e. we can safely allow peeling if operator is
an "or".
Differential Revision: https://reviews.llvm.org/D79898
We can produce such vectors in the Promote Alloca pass,
but we are unable to use movrel to operate it and lower
via scratch. Making it legal makes SI_INDIRECT patterns
work.
There is more work to do in subsequent changes:
1. We initialize m0 twice to access each dword. It shall
be possible to only do it once and increment base register
number instead.
2. We also need v16i64/v16f64 but these first need to be
added to tablegen.
Differential Revision: https://reviews.llvm.org/D79808
When splitting loads in RegBankSelect G_EXTRACT_VECTOR_ELT were being added
which could not be selected. Since invoking the legalizer will generate
instructions that split and combine wide loads, we can remove the redundant
repair instructions which are added by RegBankSelect.
Differential Revision: https://reviews.llvm.org/D75547
If there are no available lanes in a reserved VGPR, no free SGPR, and no unused CSR
VGPR when trying to save the FP it needs to be spilled to memory as a last
resort. This can be done in the prolog/epilog if we manually add the spill
and manage exec.
Differential Revision: https://reviews.llvm.org/D79610
Just do not touch loads and stores which are already vector.
Previously pass was just unable to see these loads and stores
because these were hidden bitcasts.
Differential Revision: https://reviews.llvm.org/D79738
We have the getNegatibleCost/getNegatedExpression to evaluate the cost and negate the expression.
However, during negating the expression, the cost might change as we are changing the DAG,
and then, hit the assertion if we negated the wrong expression as the cost is not trustful anymore.
This patch is target to remove the getNegatibleCost to avoid the out of sync with getNegatedExpression,
and check the cost during negating the expression. It also reduce the duplicated code between
getNegatibleCost and getNegatedExpression. And fix the crash for the test in D76638
Reviewed By: RKSimon, spatel
Differential Revision: https://reviews.llvm.org/D77319
Currently this code exists in widenScalar for G_MERGE_VALUE
sources. I'm not sure if the existing expansion in widenScalar should
be removed or not. The widenScalar variant tries to extend to the
requested size, but this just uses the original bitwidth.
Check the address space first before searching for the object
definition to save compile time. As an added bonus, this will now
treat casts to constant addrspace as constant.
We also seemed to be missing targeted tests for this, so add a few
missing other cases too.
This is mostly useful if alloca element type is not integer
and then casted to an integer for load or store. We now can
vectorize an [i32] alloca but cannot do so for [float].
There also a separate patch needed to properly lower 64 bit
types after they vectorized. At the moment these are lowered
via scratch anyway.
Differential Revision: https://reviews.llvm.org/D79641
This is preparation for D79294, which removes an expensive
InstSimplify optimization, on the assumption that it will be
picked up by InstCombine instead. Of course, this does not hold
up if a backend performs non-trivial IR expansions without running
a canonicalization pipeline afterwards, which turned up as an
issue in the context of AMDGPU div/rem expansion.
This patch mitigates the issue by explicitly performing a known
bits calculation where it matters. No test changes, as those would
only be visible after the other patch lands.
Differential Revision: https://reviews.llvm.org/D79596
We do not want to break asm syntax. These suffixes are
quite useful for debugging, so add an option to print
them. Right now it is NFC.
Differential Revision: https://reviews.llvm.org/D79435
When called from the post-RA scheduler, hazards have already been
handled by getHazardType returning NoopHazard, so PreEmitNoops always
returns zero. Remove it. NFC.
Historical note: PreEmitNoops was added to the hazard recognizer
interface as an optional feature to support dispatch group formation on
the POWER target:
http://lists.llvm.org/pipermail/llvm-commits/Week-of-Mon-20131202/197470.html
So it seems right that we shouldn't need to implement it.
We do still implement the other overload PreEmitNoops(MachineInstr *)
because that is used by the PostRAHazardRecognizer pass.
Differential Revision: https://reviews.llvm.org/D79476
Since SRSRC has alignment requirements, first find non GIT pointer clobbered
registers for SRSRC and then if those registers clobber preloaded Scratch Wave
Offset register, copy the Scratch Wave Offset register to a free SGPR.
The AMDGPU target has a convention that defined all VGPRs
(execept the initial 32 argument registers) as callee-saved.
This convention is not efficient always, esp. when the callee
requiring more registers, ended up emitting a large number of
spills, even though its caller requires only a few.
This patch revises the ABI by introducing more scratch registers
that a callee can freely use.
The 256 vgpr registers now become:
32 argument registers
112 scratch registers and
112 callee saved registers.
The scratch registers and the CSRs are intermixed at regular
intervals (a split boundary of 8) to obtain a better occupancy.
Reviewers: arsenm, t-tye, rampitec, b-sumner, mjbedy, tpr
Reviewed By: arsenm, t-tye
Differential Revision: https://reviews.llvm.org/D76356
VMEM soft clauses only contain VMEM and FLAT instructions. Teaching
GCNHazardRecognizer::checkSoftClauseHazards that other kinds of
instructions will naturally break the clause means there are far fewer
cases where it has to insert an s_nop instruction to forcibly break the
clause.
Differential Revision: https://reviews.llvm.org/D79353
Marking a section as ALLOC tells the ELF loader to load the section into memory.
As we do not want to load the notes into VRAM, the flag should not be there.
On AMDHSA, .note is still marked as ALLOC, apparently this is currently
needed for OpenCL (see https://reviews.llvm.org/D74995).
Differential Revision: https://reviews.llvm.org/D76278
Make the kind of cost explicit throughout the cost model which,
apart from making the cost clear, will allow the generic parts to
calculate better costs. It will also allow some backends to
approximate and correlate the different costs if they wish. Another
benefit is that it will also help simplify the cost model around
immediate and intrinsic costs, where we currently have multiple APIs.
RFC thread:
http://lists.llvm.org/pipermail/llvm-dev/2020-April/141263.html
Differential Revision: https://reviews.llvm.org/D79002
This a hack to fix illegal 32 to 16 bit copies.
The problem is when we make 16 bit subregs legal it creates
a huge amount of failures which can only be resolved at once
without a temporary hack like this.
The next step is to change operands, instruction definitions
and patterns until this hack is not needed.
Differential Revision: https://reviews.llvm.org/D79119
Summary: This change enables all kind of carry out ISD opcodes to be selected according to the node divergence.
Reviewers: rampitec, arsenm, vpykhtin
Reviewed By: rampitec
Subscribers: kzhuravl, jvesely, wdng, nhaehnle, yaxunl, dstuttard, tpr, t-tye, hiraditya, kerbowa, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D78091
VMEM loads of the same type (sampler vs no sampler) are guaranteed to
write their result registers in order, so there is no need for an
s_waitcnt even if they write to overlapping vgprs.
Differential Revision: https://reviews.llvm.org/D79176
It allows it not to crash and analyze 16 bit subregs if those
appear in the instructions. At the same time it does not attempt
to reassign these. It still can correctly identify register
banks to let larger registers to be reassigned.
More work will be needed here when real instructions will use
these registers and more tests as well.
Differential Revision: https://reviews.llvm.org/D78772
These are used in SReg_32 and when we start to use SGPR_LO16
there will be compaints that not all registers in RC support
all subreg indexes. For now it is NFC.
Unused regunits are reserved so that verifier does not complain
about missing phys reg live-ins.
Differential Revision: https://reviews.llvm.org/D78591
These are used in SReg_32 and when we start to use SGPR_LO16
there will be compaints that not all registers in RC support
all subreg indexes. For now it is NFC.
Unused regunits are reserved so that verifier does not complain
about missing phys reg live-ins.
Differential Revision: https://reviews.llvm.org/D78591
There are several different types of cost that TTI tries to provide
explicit information for: throughput, latency, code size along with
a vague 'intersection of code-size cost and execution cost'.
The vectorizer is a keen user of RecipThroughput and there's at least
'getInstructionThroughput' and 'getArithmeticInstrCost' designed to
help with this cost. The latency cost has a single use and a single
implementation. The intersection cost appears to cover most of the
rest of the API.
getUserCost is explicitly called from within TTI when the user has
been explicit in wanting the code size (also only one use) as well
as a few passes which are concerned with a mixture of size and/or
a relative cost. In many cases these costs are closely related, such
as when multiple instructions are required, but one evident diverging
cost in this function is for div/rem.
This patch adds an argument so that the cost required is explicit,
so that we can make the important distinction when necessary.
Differential Revision: https://reviews.llvm.org/D78635
This method has been commented as deprecated for a while. Remove
it and replace all uses with the equivalent getCalledOperand().
I also made a few cleanups in here. For example, to removes use
of getElementType on a pointer when we could just use getFunctionType
from the call.
Differential Revision: https://reviews.llvm.org/D78882
This is to fix performance regressions introduced by
86c944d790.
The old search would collect all potentially mergeable instructions in
the entire block. In this case, the same address is written in
multiple places in the block on the other side of a fence. When sorted
by offset, the two unmergeable, identical addresses would be next to
each other and the merge would give up.
Break the search space when we encounter an instruction we won't be
able to merge across. This will keep the identical addresses in
different merge attempts.
This may also improve compile time by reducing the merge list size.
Summary:
Frontend guarantees that coherent accesses have
corresponding cache policy bits set (glc, dlc).
Therefore there is no need for extra instructions
that invalidate cache.
Subscribers: arsenm, kzhuravl, jvesely, wdng, nhaehnle, yaxunl, dstuttard, tpr, t-tye, hiraditya, kerbowa, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D78800
This was backwards from intended and missing a test. We perhaps should
just ignored the FP mode here, since it shouldn't be legal to mix code
with different default modes in the absence of strictfp.
If f32 denormals were enabled pre-gfx9, we would still try to
implement this with v_max_f32. Pre-gfx9, these instructions ignored
the denormal mode and did not flush. Switch to the multiply form for
f32 as a workaround which should always work in any case.
This fixes conformance failures when the library implementation of
fmin/fmax were accidentally not inlined, forcing the assumption of no
flushing on targets where denormals are not enabled by default. This
is a workaround, since really we should not be mixing code with
different FP mode expectations, but prefer the lowering that will work
in any mode.
Now this will always use max to implement canonicalize on gfx9+. This
is only really beneficial for f64. For f32/f16 it's a neutral choice
(and worse in terms of code size in 1 case), but possibly worse for
the compiler since it does add an extra register use operand. Leave
this change for later.
This avoids more long lists of register classes that have to be updated
every time we add a new one. NFC.
Differential Revision: https://reviews.llvm.org/D78570
12994a70cf did this for 128-bit classes:
SGPR_128 only includes the real allocatable SGPRs, and SReg_128 adds
the additional non-allocatable TTMP registers. There's no point in
allocating SReg_128 vregs. This shrinks the size of the classes
regalloc needs to consider, which is usually good.
This patch extends it to all classes > 64 bits, for consistency.
Differential Revision: https://reviews.llvm.org/D78622
Add 96-bit, 160-bit and 256-bit AReg classes to match VReg and SReg.
NFC as far as I know, but it may avoid weird legalization problems.
Differential Revision: https://reviews.llvm.org/D78348
This was yet another function that had to be updated whenever you added
a new register class. Remove it by refactoring its only caller to use
standard helper functions from SIRegisterInfo.
Differential Revision: https://reviews.llvm.org/D78557
Summary:
Before this patch, `relaxInstruction` takes three arguments, the first
argument refers to the instruction before relaxation and the third
argument is the output instruction after relaxation. There are two quite
strange things:
1) The first argument's type is `const MCInst &`, the third
argument's type is `MCInst &`, but they may be aliased to the same
variable
2) The backends of ARM, AMDGPU, RISC-V, Hexagon assume that the third
argument is a fresh uninitialized `MCInst` even if `relaxInstruction`
may be called like `relaxInstruction(Relaxed, STI, Relaxed)` in a
loop.
In this patch, we drop the thrid argument, and let `relaxInstruction`
directly modify the given instruction. Also, this patch fixes the bug https://bugs.llvm.org/show_bug.cgi?id=45580, which is introduced by D77851, and
breaks the assumption of ARM, AMDGPU, RISC-V, Hexagon.
Reviewers: Razer6, MaskRay, jyknight, asb, luismarques, enderby, rtaylor, colinl, bcain
Reviewed By: Razer6, MaskRay, bcain
Subscribers: bcain, nickdesaulniers, nathanchance, wuzish, annita.zhang, arsenm, dschuff, jyknight, dylanmckay, sdardis, nemanjai, jvesely, nhaehnle, tpr, sbc100, jgravelle-google, kristof.beyls, hiraditya, aheejin, kbarton, fedor.sergeev, asb, rbar, johnrusso, simoncook, sabuasal, niosHD, jrtc27, MaskRay, zzheng, edward-jones, atanasyan, rogfer01, MartinMosbeck, brucehoult, the_o, PkmX, jocewei, Jim, lenary, s.egerton, pzheng, sameer.abuasal, apazos, luismarques, kerbowa, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D78364
The API for shuffles and reductions uses generic Type parameters,
instead of VectorType, and so assertions and casts are used a lot.
This patch makes those types explicit, which means that the clients
can't be lazy, but results in less ambiguity, and that can only be a
good thing.
Bugzilla: https://bugs.llvm.org/show_bug.cgi?id=45562
Differential Revision: https://reviews.llvm.org/D78357
Summary:
Introduce new helper functions getVGPRClassForBitWidth,
getAGPRClassForBitWidth, getSGPRClassForBitWidth and use them to
refactor various other functions that all contained their own lists of
valid register class widths. NFC.
Reviewers: arsenm, rampitec
Subscribers: kzhuravl, jvesely, wdng, nhaehnle, yaxunl, dstuttard, tpr, t-tye, hiraditya, kerbowa, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D78311
Summary:
Use more logic and fewer tables. This reduces the line count and
reduces the effort required to introduce more register classes of
different sizes in future.
Reviewers: arsenm, rampitec, nhaehnle
Subscribers: kzhuravl, jvesely, wdng, yaxunl, dstuttard, tpr, t-tye, hiraditya, kerbowa, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D78351
This does for G_EXTRACT_VECTOR_ELT what 588bd7be36 did for G_TRUNC.
Ideally types without a corresponding register class wouldn't reach
here, but we're currently missing some (in particular a 192-bit class
is missing).
This allows targets to know exactly which operands are contributing to
the dependency, which is required for targets with per-operand
scheduling models.
Differential Revision: https://reviews.llvm.org/D77135
These are needed as a counterpart for VGPR subregs even though
there are no scalar instructions which can operate 16 bit values.
When we are materializing a constant that is done into an SGPR
and that SGPR may/will be copied into a 16 bit VGPR subreg. Such
copy is illegal. There are also similar problems if a source
operand of a 16 bit VALU instruction is an SGPR. In addition
we need to get a register with a lo16 subregister of an SGPR
RC during selection and this fails as well.
All of that makes me believe we need these subregisters as a
syntactic glue.
Differential Revision: https://reviews.llvm.org/D78250
An irreducible SCC is one which has multiple "header" blocks, i.e., blocks
with control-flow edges incident from outside the SCC. This pass converts an
irreducible SCC into a natural loop by introducing a single new header
block and redirecting all the edges on the original headers to this
new block.
This is a useful workaround for a limitation in the structurizer
which, which produces incorrect control flow in the presence of
irreducible regions. The AMDGPU backend provides an option to
enable this pass before the structurizer, which may eventually be
enabled by default.
Reviewed By: nhaehnle
Differential Revision: https://reviews.llvm.org/D77198
This restores commit 2ada8e2525.
Originally reverted with commit 44e09b59b8.
This reverts commit 2ada8e2525.
Buildbots produced compilation errors which I was not able to quickly
reproduce locally. Need more time to investigate.
An irreducible SCC is one which has multiple "header" blocks, i.e., blocks
with control-flow edges incident from outside the SCC. This pass converts an
irreducible SCC into a natural loop by introducing a single new header
block and redirecting all the edges on the original headers to this
new block.
This is a useful workaround for a limitation in the structurizer
which, which produces incorrect control flow in the presence of
irreducible regions. The AMDGPU backend provides an option to
enable this pass before the structurizer, which may eventually be
enabled by default.
Reviewed By: nhaehnle
Differential Revision: https://reviews.llvm.org/D77198
Ports the existing DAG combines, minus the simplify demanded bits
which seems to have no equivalent now. Without these, this isn't
particularly helpful in most of the IR sample cases.
Since 1725f28841, this should check
isFMADLegalForFAddFSub rather than the the plain isOperationLegal.
This would assert in a subset of cases due to an oddity in how FMAD is
selected. We will allow FMA formation pre-legalize, but not FMAD even
in cases where it would be valid.
The current hook requires passing in the root fadd/fsub. However, in
this distributed case, this would be far more complicated to pass in
the relevant operand. AMDGPU doesn't get any value from the node, and
only needs the type and is the only implementor, so I'm not sure why
we have this complexity. Just rename and expand the assert to avoid
the more complicated checks spread through the distribution logic.
Summary:
Updated CallPromotionUtils and impacted sites. Parameters that are
expected to be non-null, and return values that are guranteed non-null,
were replaced with CallBase references rather than pointers.
Left FIXME in places where more changes are facilitated by CallBase, but
aren't CallSites: Instruction* parameters or return values, for example,
where the contract that they are actually CallBase values.
Reviewers: davidxl, dblaikie, wmi
Reviewed By: dblaikie
Subscribers: arsenm, jvesely, nhaehnle, eraman, hiraditya, kerbowa, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D77930
Selection would fail after the post legalize combiner put an illegal
zextload back together.
The base combiner has parameter to only allow legal operations, but
they appear to not be used. I also don't see a nice way to remove a
single entry from all_combines, so just hack around this.
These will be widened in the DAG. In the meanwhile early
widening prevents otherwise possible vectorization of
such loads.
Differential Revision: https://reviews.llvm.org/D77835
Summary:
Remove usages of asserting vector getters in Type in preparation for the
VectorType refactor. The existence of these functions complicates the
refactor while adding little value.
Reviewers: arsenm, efriedma, sdesmalen
Reviewed By: arsenm
Subscribers: wdng, arsenm, jvesely, nhaehnle, hiraditya, kerbowa, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D77268
This patch extends existing constant folding in logical operations to
handle S_XNOR, S_NAND, S_NOR, S_ANDN2, S_ORN2, V_LSHL_ADD_U32 and
V_AND_OR_B32. Also added a couple of tests for existing folds.
We can only collapse adjacent SI_END_CF if outer statement
belongs to a simple SI_IF, otherwise correct mask is not in the
register we expect, but is an argument of an S_XOR instruction.
Even if SI_IF is simple it might be lowered using S_XOR because
lowering is dependent on a basic block layout. It is not
considered simple if instruction consuming its output is
not an SI_END_CF. Since that SI_END_CF might have already been
lowered to an S_OR isSimpleIf() check may return false.
This situation is an opportunity for a further optimization
of SI_IF lowering, but that is a separate optimization. In the
meanwhile move SI_END_CF post the lowering when we already know
how the rest of the CFG was lowered since a non-simple SI_IF
case still needs to be handled.
Differential Revision: https://reviews.llvm.org/D77610
Now that we have scalable vectors, there's a distinction that isn't
getting captured in the original SequentialType: some vectors don't have
a known element count, so counting the number of elements doesn't make
sense.
In some cases, there's a better way to express the commonality using
other methods. If we're dealing with GEPs, there's GEP methods; if we're
dealing with a ConstantDataSequential, we can query its element type
directly.
In the relatively few remaining cases, I just decided to write out
the type checks. We're talking about relatively few places, and I think
the abstraction doesn't really carry its weight. (See thread "[RFC]
Refactor class hierarchy of VectorType in the IR" on llvmdev.)
Differential Revision: https://reviews.llvm.org/D75661
The extracts from control flow intrinsics are already properly handled
by divergence analysis. The inline asm case isn't dead, but has also
never really worked correctly so leave it as-is for now.
We can fix register class of PHI based on its all AGPR uses.
That leaves behind all PHIs which were already processed
earlier. Propagate RC back to PHI operands of a PHI.
Differential Revision: https://reviews.llvm.org/D77344
Summary:
This fixes a few issues related to SMRD offsets. On gfx9 and gfx10 we have a
signed byte offset immediate, however we can overflow into a negative since we
treat it as unsigned.
Also, the SMRD SOFFSET sgpr is an unsigned offset on all subtargets. We
sometimes tried to use negative values here.
Third, S_BUFFER instructions should never use a signed offset immediate.
Differential Revision: https://reviews.llvm.org/D77082
This will likely introduce catastrophic performance regressions on
older subtargets, but should be correct. A follow up change will
remove the old fp32-denormals subtarget features, and switch to using
the new denormal-fp-math/denormal-fp-math-f32 attributes. Frontends
should be making sure to add the denormal-fp-math-f32 attribute when
appropriate to avoid performance regressions.
This is a workaround for clang adding noinline to all functions at
-O0. Previously, we would just add alwaysinline, and the verifier
would complain about having both noinline and alwaysinline. We
currently can't truly codegen this case as a freestanding function, so
override the user forcing noinline.
SILoadStoreOptimizer::checkAndPrepareMerge() expects base and
paired instruction to come in order and scans MBB from base to
the paired instruction. An original order can be changed if
there were a dependent instruction in between and base instruction
was moved.
Fixed by bailing the optimization. In theory it might be possible
still to perform a merge by swapping instructions, but on practice
it bails anyway because it finds dependency on that same instruction
which has resulted in the base move.
Differential Revision: https://reviews.llvm.org/D77245
Instead, represent the mask as out-of-line data in the instruction. This
should be more efficient in the places that currently use
getShuffleVector(), and paves the way for further changes to add new
shuffles for scalable vectors.
This doesn't change the syntax in textual IR. And I don't currently plan
to change the bitcode encoding in this patch, although we'll probably
need to do something once we extend shufflevector for scalable types.
I expect that once this is finished, we can then replace the raw "mask"
with something more appropriate for scalable vectors. Not sure exactly
what this looks like at the moment, but there are a few different ways
we could handle it. Maybe we could try to describe specific shuffles.
Or maybe we could define it in terms of a function to convert a fixed-length
array into an appropriate scalable vector, using a "step", or something
like that.
Differential Revision: https://reviews.llvm.org/D72467
We have loads preserving low and high 16 bits of their
destinations. However, we always use a whole 32 bit register
for these. The same happens with 16 bit stores, we have to
use full 32 bit register so if high bits are clobbered the
register needs to be copied. One example of such code is
added to the load-hi16.ll.
The proper solution to the problem is to define 16 bit subregs
and use them in the operations which do not read another half
of a VGPR or preserve it if the VGPR is written.
This patch simply defines subregisters and register classes.
At the moment there should be no difference in code generation.
A lot more work is needed to actually use these new register
classes. Therefore, there are no new tests at this time.
Register weight calculation has changed with new subregs so
appropriate changes were made to keep all calculations just
as they are now, especially calculations of register pressure.
Differential Revision: https://reviews.llvm.org/D74873
Add a new llvm.amdgcn.ballot intrinsic modeled on the ballot function
in GLSL and other shader languages. It returns a bitfield containing the
result of its boolean argument in all active lanes, and zero in all
inactive lanes.
This is intended to replace the existing llvm.amdgcn.icmp and
llvm.amdgcn.fcmp intrinsics after a suitable transition period.
Use the new intrinsic in the atomic optimizer pass.
Differential Revision: https://reviews.llvm.org/D65088
We currently don't have a way to map to the equivalent intrinsic
opcode, so track immediate 0s in place of the address for the
selection to know to change the final opcode.
Summary:
This change adds amdgcn.reloc.constant intrinsic to the amdgpu backend, which will compile into a relocation entry in the resulting elf.
The intrinsics takes a MetadataNode (String) as its only argument, which specifies the symbol name of the relocation entry.
`SelectionDAGBuilder::getValueImpl` is changed to allow metadata operands passed through to ISel.
Author: csyonghe <yonghe@google.com>
Reviewers: tpr, nhaehnle
Reviewed By: nhaehnle
Subscribers: arsenm, kzhuravl, jvesely, wdng, yaxunl, dstuttard, t-tye, hiraditya, kerbowa, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D76440
For each natural loop with multiple exit blocks, this pass creates a
new block N such that all exiting blocks now branch to N, and then
control flow is redistributed to all the original exit blocks.
The bulk of the tranformation is a new function introduced in
BasicBlockUtils that an redirect control flow from a set of incoming
blocks to a set of outgoing blocks via a common "hub".
This is a useful workaround for a limitation in the structurizer which
incorrectly orders blocks when processing a nest of loops. This pass
bypasses that issue by ensuring that each natural loop is recognized
as a separate region. Since the structurizer is a region pass, it no
longer sees a nest of loops in a single region, and instead processes
each "level" in the nesting as a separate region.
The AMDGPU backend provides a new option to enable this pass before
the structurizer, which may eventually be enabled by default.
Reviewers: madhur13490, arsenm, nhaehnle
Reviewed By: nhaehnle
Differential Revision: https://reviews.llvm.org/D75865
I still think the call lowering type legalization logic split between
the generic code and target is too confusing, but largely induced by
the reliance on the DAG infrastructure.
This fixes
warning: loop variable 'Def' of type 'const llvm::Register' creates a copy from type 'const llvm::Register' [-Wrange-loop-analysis]
llvm::Register just contains a single unsigned and should be copied.
Reviewers: rampitec
Reviewed By: rampitec
Differential Revision: https://reviews.llvm.org/D77011
Summary:
"Per CU" is a bit simplistic for gfx10, but I couldn't think of a better
name.
Reviewers: arsenm, rampitec, nhaehnle, dstuttard, tpr
Subscribers: kzhuravl, jvesely, wdng, yaxunl, t-tye, hiraditya, kerbowa, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D76861
Having arbitrary passes looking at the TargetOptions is pretty
messy. This was also disregarding if a function already had an
explicit attribute setting on it. opt/llc now add the attributes to
functions that don't specify the attribute. clang and lld do not call
the function to do this, which they maybe should.
This was also treating unsafe-fp-math as implying the others, and
setting the other attributes based on it. This is not done anywhere
else, and I'm not sure is correct based on the current description of
the option bit.
Effectively reverts 1d8cf2be89
Summary:
This is patch is part of a series to introduce an Alignment type.
See this thread for context: http://lists.llvm.org/pipermail/llvm-dev/2019-July/133851.html
See this patch for the introduction of the type: https://reviews.llvm.org/D64790
Reviewers: courbet
Subscribers: arsenm, dschuff, sdardis, nemanjai, jvesely, nhaehnle, sbc100, jgravelle-google, hiraditya, aheejin, kbarton, jrtc27, atanasyan, jfb, kerbowa, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D76925
Summary: I think Max in the name was misleading. NFC.
Reviewers: arsenm, rampitec
Subscribers: kzhuravl, jvesely, wdng, nhaehnle, yaxunl, dstuttard, tpr, t-tye, hiraditya, kerbowa, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D76860
Summary:
These methods were identical. I chose to remove getMaxWavesPerCU because
I think Max in the name was misleading. NFC.
Reviewers: arsenm, rampitec
Subscribers: kzhuravl, jvesely, wdng, nhaehnle, yaxunl, dstuttard, tpr, t-tye, hiraditya, kerbowa, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D76859
Summary:
The PC_32 DWARF register is for a 32-bit process address space which we
don't implement in AMDGCN; another way of putting this is that the size
of the PC register is not a function of the wavefront size. If we ever
implement a 32-bit process address space we will need to add two more
DwarfFlavours i.e. we will need to represent the product of (wave32,
wave64) x (64-bit address space, 32-bit address space).
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D76732
Follow-up of D72172 and D72180
This patch passes `uint64_t Address` to print methods of PC-relative
operands so that subsequent target specific patches can change
`*InstPrinter::print{Operand,PCRelImm,...}` to customize the output.
Add MCInstPrinter::PrintBranchImmAsAddress which is set to true by
llvm-objdump.
```
// Current llvm-objdump -d output
aarch64: 20000: bl #0
ppc: 20000: bl .+4
x86: 20000: callq 0
// Ideal output
aarch64: 20000: bl 0x20000
ppc: 20000: bl 0x20004
x86: 20000: callq 0x20005
// GNU objdump -d. The lack of 0x is not ideal because the result cannot be re-assembled
aarch64: 20000: bl 20000
ppc: 20000: bl 0x20004
x86: 20000: callq 20005
```
In `lib/Target/X86/X86GenAsmWriter1.inc` (generated by `llvm-tblgen -gen-asm-writer`):
```
case 12:
// CALL64pcrel32, CALLpcrel16, CALLpcrel32, EH_SjLj_Setup, JCXZ, JECXZ, J...
- printPCRelImm(MI, 0, O);
+ printPCRelImm(MI, Address, 0, O);
return;
```
Some targets have 2 `printOperand` overloads, one without `Address` and
one with `Address`. They should annotate derived `Operand` properly with
`let OperandType = "OPERAND_PCREL"`.
Reviewed By: jhenderson
Differential Revision: https://reviews.llvm.org/D76574
AMDGPUPropagateAttributes pass was skipping some of the functions
when cloning. Functions were added to root set and then skipped
on the next interation because they are already in the root set,
while were meant to be processed with different features.
Differential Revision: https://reviews.llvm.org/D76815
AMDGPUPropagateAttributes can swap names while cloning a function.
Only do it if original symbol was not externally visible.
Differential Revision: https://reviews.llvm.org/D76789
This pass can handle all the optimization
opportunities found just before code emission.
Presently it includes the handling of vcc branch
optimization that was handled earlier in SIInsertSkips.
Reviewed By: arsenm
Differential Revision: https://reviews.llvm.org/D76712
There seems to be a small benefit to the legalized sequence for v2f16
round with packed instructions, so allow vectorizing it by reducing
the cost.
An unintended side effect is vectorization of f32 round also
happens. The current FMA logic seems off to me, and isn't checking for
packed instructions.
Implement the DWARF register mapping described in llvm/docs/AMDGPUUsage.rst.
This enables generating appropriate DWARF register numbers for wave64 and
wave32 modes.
Summary:
Currently we custom select add/sub with carry out to scalar form relying on later replacing them to vector form if necessary.
This change enables custom selection code to take the divergence of adde/addc SDNodes into account and select the appropriate form in one step.
Reviewers: arsenm, vpykhtin, rampitec
Reviewed By: arsenm, vpykhtin
Subscribers: kzhuravl, jvesely, wdng, nhaehnle, yaxunl, dstuttard, tpr, t-tye, hiraditya, kerbowa
Differential Revision: https://reviews.llvm.org/D76371
If there were no free VGPRs we would need two emergency spill slots for register
scavenging during PEI/frame index elimination. Reuse 'ResultReg' for scale
calculation so that only one spill is needed.
Differential Revision: https://reviews.llvm.org/D76387
Apart from the argument registers, set the CostPerUse
value as per the ratio reg_index/allocation_granularity.
It is a pre-commit for introducing the scratch registers
in the ABI. This change should help in a balanced
register allocation.
Differential Revision: https://reviews.llvm.org/D76417
Remove the gap left between the stack pointer (s32) and frame pointer
(s34) now that the scratch wave offset is no longer a part of the
calling convention ABI.
Update llvm/docs/AMDGPUUsage.rst to reflect the change.
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D75657
Add the scratch wave offset to the scratch buffer descriptor (SRSrc) in
the entry function prologue. This allows us to removes the scratch wave
offset register from the calling convention ABI.
As part of this change, allow the use of an inline constant zero for the
SOffset of MUBUF instructions accessing the stack in entry functions
when a frame pointer is not requested/required. Entry functions with
calls still need to set up the calling convention ABI stack pointer
register, and reference it in order to address arguments of called
functions. The ABI stack pointer register remains unswizzled, but is now
wave-relative instead of queue-relative.
Non-entry functions also use an inline constant zero SOffset for
wave-relative scratch access, but continue to use the stack and frame
pointers as before. When the stack or frame pointer is converted to a
swizzled offset it is now scaled directly, as the scratch wave offset no
longer needs to be subtracted first.
Update llvm/docs/AMDGPUUsage.rst to reflect these changes to the calling
convention.
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D75138
Remove dead code and factor repeated conditions out into a single check.
Rename and move code to make it more obvious what is running only for
entry functions. Simplify function arguments to make it clearer what the
relevant inputs are. Make flat scratch init accept an MBB iterator and
move it to where it was logically being emitted within the prologue.
These changes will make a future update to the calling convention
simpler.
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D75092
This isn't really usable, and requires using the
-amdgpu-fixed-function-abi flag to work.
Assumes a uniform call target, and will hit a verifier error if the
call target ends up in a VGPR. Also doesn't attempt to do anything
sensible for the reported register/stack usage.
This reverts commit 9bca8fc4cf.
Rearrange handling to avoid changing the instruction in the case where
it's going to be erased and replaced with undef.
Summary:
For the case where "done" bits on existing exports are removed
by unifyReturnBlockSet(), unify all return blocks - even the
uniformly reached ones. We do not want to end up with a non-unified,
uniformly reached block containing a normal export with the "done"
bit cleared.
That case is believed to be rare - possible with infinite loops
in pixel shaders.
This is a fix for D71192.
Subscribers: arsenm, kzhuravl, jvesely, wdng, nhaehnle, yaxunl, dstuttard, tpr, t-tye, hiraditya, kerbowa, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D76364
Summary:
This is patch is part of a series to introduce an Alignment type.
See this thread for context: http://lists.llvm.org/pipermail/llvm-dev/2019-July/133851.html
See this patch for the introduction of the type: https://reviews.llvm.org/D64790
Reviewers: courbet
Subscribers: jholewinski, arsenm, dschuff, jyknight, sdardis, nemanjai, jvesely, nhaehnle, sbc100, jgravelle-google, hiraditya, aheejin, kbarton, fedor.sergeev, asb, rbar, johnrusso, simoncook, sabuasal, niosHD, jrtc27, MaskRay, zzheng, edward-jones, atanasyan, rogfer01, MartinMosbeck, brucehoult, the_o, PkmX, jocewei, Jim, lenary, s.egerton, pzheng, sameer.abuasal, apazos, luismarques, kerbowa, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D76348
Summary:
Explanation is in a comment in the diff, but essentially printing a
physical register name here is ambiguous. Until we can implement
printing a DWARF register name here just use the encoding directly.
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D76253
For normal loads, fully eliminate the load. For the TFE case, adjust
the dmask value in the instruction so the selector doesn't need to
handle it. For the TFE special case, I guess it would be possible to
replace the loaded data register with undef, but as-is this will start
treating it as a well defined value.
Trim elements that won't be written. The equivalent still needs to be
done for writes. Also start widening 3 elements to 4
elements. Selection will get the count from the dmask.
We were letting G_ANYEXT with a vcc register bank through, which was
incorrect and would select to an invalid copy. Fix this up like G_ZEXT
and G_SEXT. Also drop old code to fixup the non-boolean case in
RegBankSelect. We now have to perform that expansion during selection,
so there's no benefit to doing it during RegBankSelect.
Fixes integers that don't evenly divide to i32 pieces. We should
probably extract some of the code in the legalizer to start handling
argument breakdowns. I'm dissatisfied with the argument lowering's
handling of vectors for example, and we should not be producing the
weird G_EXTRACTs we do now.
This patch removes compiler runtime assertions that ensure the implicit
conversion are only guaranteed to work for fixed-width vectors.
With the assert it would be impossible to get _anything_ to build until
the
entire codebase has been upgraded, even when the indiscriminate uses of
the size as uint64_t would work fine for both scalable and fixed-width
types.
This issue will need to be addressed differently, with build-time errors
rather than assertion failures, but that effort falls beyond the scope
of this patch.
Returning the scalable size and avoiding the assert in getFixedSize()
is a temporary stop-gap in order to use LLVM for compiling and using
the SVE ACLE intrinsics.
Reviewers: efriedma, huntergr, rovka, ctetreau, rengolin
Reviewed By: efriedma
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D75297
Only collapse inner endcf if the outer one belongs to SI_IF.
If it does belong to SI_ELSE then mask being restored in fact
a partial inverse of what we need.
Differential Revision: https://reviews.llvm.org/D76154
There are some functional regressions and I suspect our
scopes are not as perfectly enclosed as I expected.
Disable it for now.
Differential Revision: https://reviews.llvm.org/D76148
For context, the proposed RISC-V bit manipulation extension has a subset
of instructions which require one of two SubtargetFeatures to be
enabled, 'zbb' or 'zbp', and there is no defined feature which both of
these can imply to use as a constraint either (see comments in D65649).
AssemblerPredicates allow multiple SubtargetFeatures to be declared in
the "AssemblerCondString" field, separated by commas, and this means
that the two features must both be enabled. There is no equivalent to
say that _either_ feature X or feature Y must be enabled, short of
creating a dummy SubtargetFeature for this purpose and having features X
and Y imply the new feature.
To solve the case where X or Y is needed without adding a new feature,
and to better match a typical TableGen style, this replaces the existing
"AssemblerCondString" with a dag "AssemblerCondDag" which represents the
same information. Two operators are defined for use with
AssemblerCondDag, "all_of", which matches the current behaviour, and
"any_of", which adds the new proposed ORing features functionality.
This was originally proposed in the RFC at
http://lists.llvm.org/pipermail/llvm-dev/2020-February/139138.html
Changes to all current backends are mechanical to support the replaced
functionality, and are NFCI.
At this stage, it is illegal to combine features with ands and ors in a
single AssemblerCondDag. I suspect this case is sufficiently rare that
adding more complex changes to support it are unnecessary.
Differential Revision: https://reviews.llvm.org/D74338
Currently we infer whether the flat-scratch-init kernel input should
be enabled based on calls. Move this handling, so we can decide if the
full set of ABI inputs is needed in kernels. Ideally we would have an
analysis of some sort, rather than the function attributes.
The patch removes late endcf handling and only leaves the
related portion with redundant exec mask copy elimination.
Differential Revision: https://reviews.llvm.org/D76095
This patch allows ISD::FSHR(i32) patterns to lower to ALIGNBIT instructions.
This improves test coverage of ISD::FSHR matching - x86 has both FSHL/FSHR instructions and we prefer FSHL by default.
Differential Revision: https://reviews.llvm.org/D76070
This is to replace the optimization from the SIOptimizeExecMaskingPreRA.
We have less opportunities in the control flow lowering because many
VGPR copies are still in place and will be removed later, but we know
for sure an instruction is SI_END_CF and not just an arbitrary S_OR_B64
with EXEC.
The subsequent change needs to convert s_and_saveexec into s_and and
address new TODO lines in tests, then code block guarded by the
-amdgpu-remove-redundant-endcf option in the pre-RA exec mask optimizer
will be removed.
Differential Revision: https://reviews.llvm.org/D76033
The note section type implies a specific format that this section does
not have thus tools like readelf fail here. Progbits has no format and
another pipeline compiler already sets the type to progbits.
Differential Revision: https://reviews.llvm.org/D75913
Instead, emit a trap and a warning. We force inlining of this
situation, so any function where this happens should be dead as
indirect or external calls are not yet supported. This should avoid
erroring on dead code.
The assumption is that conditional regions are perfectly nested
and a mask restored at the exit from the inner block will be
completely covered by a mask restored in the outer.
It turns out with our current structurizer this is not always
the case.
Disable the optimization for now, but I want to keep it around
for a while to either try after further structurizer changes or
to move it into control flow lowering where we have more info
and reuse the test.
Differential Revision: https://reviews.llvm.org/D75958
Summary:
There's a lot of test case churn but the overall effect is to increase
the number of back-to-back v_sub,v_subbrev pairs, which can execute with
no delay even on gfx10.
Reviewers: arsenm, rampitec, nhaehnle
Subscribers: kzhuravl, jvesely, wdng, yaxunl, dstuttard, tpr, t-tye, hiraditya, kerbowa, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D75999
This was failng on any pre-assigned copy to the VCC bank.
This is something of a workaround for the default implementation in
getInstrMappingImpl, and how it treats copy-like operations in
general.
Copy-like operations are considered to only have one result register
bank, rather than separate banks for each source like a normal
instruction. To avoid potentially mishandling reg_sequence with
impossible operand combinations, the generic implementation errors on
impossible costs. If the bank was already assigned, is treated it
as-if it were an unsatisfiable REG_SEQUENCE mapping. We really don't
get any value from any of what getInstrMappingImpl tries to do for
copies, so just directly emit the simple mapping we really want.
Refines the gather/scatter cost model, but also changes the TTI
function getIntrinsicInstrCost to accept an additional parameter
which is needed for the gather/scatter cost evaluation.
This did require trivial changes in some non-ARM backends to
adopt the new parameter.
Extending gathers and truncating scatters are now priced cheaper.
Differential Revision: https://reviews.llvm.org/D75525
This avoids regressions in a future patch. I'm confused by the use of
the gfx9 usage legacy_mad. Was this a pointless instruction rename, or
uses fmul_legacy handling? Why is regular mac avilable in that case?
Summary:
As far as I can tell on gfx10 conversions to/from f32 (that are not
converting f32 to/from f64) are full rate instructions, but they were
marked as quarter rate instructions.
I have fixed this for gfx10 only. I assume the scheduling model was
correct for older architectures, though I don't have any documentation
handy to confirm that.
Reviewers: rampitec, arsenm
Subscribers: kzhuravl, jvesely, wdng, nhaehnle, yaxunl, dstuttard, tpr, t-tye, hiraditya, kerbowa, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D75392
In case the source value ends up in a VGPR, insert a readfirstlane to
avoid producing an illegal copy later. If it turns out to be
unnecessary, it can be folded out.
Summary:
When SI_INDIRECT_DST_V* pseudos has indexes in VGPR, they get expanded into the self-looped basic block that modifies EXEC in a loop.
To keep EXEC consistent it is stored before and then re-stored after the pseudo expansion result.
%95:vreg_512 = SI_INDIRECT_DST_V16 %93:vreg_512(tied-def 0), %94:sreg_32, 0, killed %1500:vgpr_32
results to
s_mov_b64 s[6:7], exec
BB0_16:
v_readfirstlane_b32 s8, v28
v_cmp_eq_u32_e32 vcc, s8, v28
s_and_saveexec_b64 vcc, vcc
s_set_gpr_idx_on s8, gpr_idx(DST)
v_mov_b32_e32 v6, v25
s_set_gpr_idx_off
s_xor_b64 exec, exec, vcc
s_cbranch_execnz BB0_16
; %bb.17:
s_mov_b64 exec, s[6:7]
The bug appeared in case this expansion occurs in the ELSE block of the CF.
Originally
%110:vreg_512 = SI_INDIRECT_DST_V16 %103:vreg_512(tied-def 0), %85:vgpr_32, 0, %107:vgpr_32,
%112:sreg_64 = SI_ELSE %108:sreg_64, %bb.19, 0, implicit-def dead $exec, implicit-def dead $scc, implicit $exec
expanded to
****************** <== here exec has "THEN" context
s_mov_b64 s[6:7], exec
BB0_16:
v_readfirstlane_b32 s8, v28
v_cmp_eq_u32_e32 vcc, s8, v28
s_and_saveexec_b64 vcc, vcc
s_set_gpr_idx_on s8, gpr_idx(DST)
v_mov_b32_e32 v6, v25
s_set_gpr_idx_off
s_xor_b64 exec, exec, vcc
s_cbranch_execnz BB0_16
; %bb.17:
s_or_saveexec_b64 s[4:5], s[4:5] <-- exec mask is restored for "ELSE" but immediately overwritten.
s_mov_b64 exec, s[6:7]
The rest of the "ELSE" block is executed not by the workitems which constitute the "else mask" but by those which constitute "then mask"
SILowerControlFlow::emitElse always considers the basic block begin() as an insertion point for s_or_saveexec.
Proposed fix: The SI_INDIRECT_DST_V* procedure should split the reminder block to create landing pad for the EXEC restoration.
Reviewers: rampitec, vpykhtin, nhaehnle
Reviewed By: vpykhtin
Subscribers: arsenm, kzhuravl, jvesely, wdng, yaxunl, dstuttard, tpr, t-tye, hiraditya, kerbowa, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D75472
When expanding scalar packed operations, we should not introduce
illegal vector casts LegalizerHelper introduces. We're not in a
legalizer context, and there's no RegBankSelect apply or legalize
worklist.
We weren't considering the packed case correctly, and this was passing
through to the selector. The selector only checked the size, so this
would incorrectly compile to a single 32-bit scalar add.
As usual, the LegalizerHelper is somewhat awkward to use from
applyMappingImpl. I think this is the first place we've needed
multi-step legalization here though.
Summary:
Instruction variants like S_MOV_B32_term should have the same SchedRW
class as the base instruction, S_MOV_B32. This probably doesn't make any
difference in practice because as terminators, they'll always be
scheduled at the end of a basic block, but it's simply more correct than
giving them all the default SchedRW class of Write32Bit, which implies a
VALU operation.
Reviewers: rampitec, arsenm, nhaehnle
Subscribers: kzhuravl, jvesely, wdng, yaxunl, dstuttard, tpr, t-tye, hiraditya, kerbowa, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D75860
Most importantly, this fixes ret i8. Also make sure to handle
signext/zeroext for odd types > i32. Some of the corresponding
argument passing fixes also need to be handled.
Interpret these as extending to the next multiple of 32-bits. This had
no effect with i48 for example, which is really split into {i32, i16},
which should extend the high part.
Use a hack to only enable this for GlobalISel.
Technically this also works with SelectionDAG, but the divergence
selection isn't reliable enough and a few cases fail, but I have no
desire to spend time writing the manual expansion code for it. The DAG
actually does a better job since it catches using v_add_lshl_u32 in
the mixed SGPR/VGPR cases.
Summary:
This seems like an obvious error - cut and paste issue?
The change does make a change to one of the lit tests - it stops s_buffer_load
re-ordering past an MUBUF instruction (which is not surprising).
Change-Id: I80be99de5b62af4f42e91af2591b76a52ac9efa6
Subscribers: arsenm, kzhuravl, jvesely, wdng, nhaehnle, yaxunl, tpr, t-tye, hiraditya, kerbowa, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D75686
Create a wider source vector, and unmerge with dead defs like the
legalizer. The legalization handling for G_EXTRACT is incomplete, and
it's preferrable to keep everything in 32-bit pieces.
We should probably start moving these functions into utils, since we
have a growing number of places that do almost the same thing.
If SimplifyDemandedBits succeeds in simplifying the byte src, add the CVT_F32_UBYTE node back to the worklist as we might be able to simplify further.
Yet another step towards removing SelectionDAG::GetDemandedBits.
The computation here didn't really make sense to me, and reported
wildy different results depending on the flat work group size
attribute.
I think this should really report a range derived from the possible
work group size bounds, and only allow an occupancy that is a multiple
of the group size.
Summary:
pickNodeBidirectional tried to compare the best top candidate and the
best bottom candidate by examining TopCand.Reason and BotCand.Reason.
This is unsound because, after calling pickNodeFromQueue, Cand.Reason
does not reflect the most important reason why Cand was chosen. Rather
it reflects the most recent reason why it beat some other potential
candidate, which could have been for some low priority tie breaker
reason.
I have seen this cause problems where TopCand is a good candidate, but
because TopCand.Reason is ORDER (which is very low priority) it is
repeatedly ignored in favour of a mediocre BotCand. This is not how
bidirectional scheduling is supposed to work.
To fix this I changed the code to always compare TopCand and BotCand
directly, like the generic implementation of pickNodeBidirectional does.
This removes some uncommented AMDGPU-specific logic; if this logic turns
out to be important then perhaps it could be moved into an override of
tryCandidate instead.
Graphics shader benchmarking on gfx10 shows a lot more positive than
negative effects from this change.
Reviewers: arsenm, tstellar, rampitec, kzhuravl, vpykhtin, dstuttard, tpr, atrick, MatzeB
Subscribers: jvesely, wdng, nhaehnle, yaxunl, t-tye, hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D68338
Summary:
Final patch in series to fix inlining between functions with different
nobuiltin attributes/options, which was specifically an issue in LTO.
See discussion on D61634 for background.
The prior patch in this series (D67923) enabled per-Function TLI
construction that identified the nobuiltin attributes.
Here I have allowed inlining to proceed if the callee's nobuiltins are a
subset of the caller's nobuiltins, but not in the reverse case, which
should be conservatively correct. This is controlled by a new option,
-inline-caller-superset-nobuiltin, which is enabled by default.
Reviewers: hfinkel, gchatelet, chandlerc, davidxl
Subscribers: arsenm, jvesely, nhaehnle, mehdi_amini, eraman, hiraditya, haicheng, dexonsmith, kerbowa, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D74162
Matt Arsenault