Occupancy (i.e., the number of waves per EU) depends, in addition to
register usage, on per-workgroup LDS usage as well as on the range of
possible workgroup sizes. Mirroring the latter, occupancy should
therefore be expressed as a range since different group sizes generally
yield different achievable occupancies.
`getOccupancyWithLocalMemSize` currently returns a scalar occupancy
based on the maximum workgroup size and LDS usage. With respect to the
workgroup size range, this scalar can be the minimum, the maximum, or
neither of the two of the range of achievable occupancies. This commit
fixes the function by making it compute and return the range of
achievable occupancies w.r.t. workgroup size and LDS usage; it also
renames it to `getOccupancyWithWorkGroupSizes` since it is the range of
workgroup sizes that produces the range of achievable occupancies.
Computing the achievable occupancy range is surprisingly involved.
Minimum/maximum workgroup sizes do not necessarily yield maximum/minimum
occupancies i.e., sometimes workgroup sizes inside the range yield the
occupancy bounds. The implementation finds these sizes in constant time;
heavy documentation explains the rationale behind the sometimes
relatively obscure calculations.
As a justifying example, consider a target with 10 waves / EU, 4 EUs/CU,
64-wide waves. Also consider a function with no LDS usage and a flat
workgroup size range of [513,1024].
- A group of 513 items requires 9 waves per group. Only 4 groups made up
of 9 waves each can fit fully on a CU at any given time, for a total of
36 waves on the CU, or 9 per EU. However, filling as much as possible
the remaining 40-36=4 wave slots without decreasing the number of groups
reveals that a larger group of 640 items yields 40 waves on the CU, or
10 per EU.
- Similarly, a group of 1024 items requires 16 waves per group. Only 2
groups made up of 16 waves each can fit fully on a CU ay any given time,
for a total of 32 waves on the CU, or 8 per EU. However, removing as
many waves as possible from the groups without being able to fit another
equal-sized group on the CU reveals that a smaller group of 896 items
yields 28 waves on the CU, or 7 per EU.
Therefore the achievable occupancy range for this function is not [8,9]
as the group size bounds directly yield, but [7,10].
Naturally this change causes a lot of test churn as instruction
scheduling is driven by achievable occupancy estimates. In most unit
tests the flat workgroup size range is the default [1,1024] which,
ignoring potential LDS limitations, would previously produce a scalar
occupancy of 8 (derived from 1024) on a lot of targets, whereas we now
consider the maximum occupancy to be 10 in such cases. Most tests are
updated automatically and checked manually for sanity. I also manually
changed some non-automatically generated assertions when necessary.
Fixes#118220.
This reverts commit adaff46d087799072438dd744b038e6fd50a2d78.
Drop the -O3 checks from default-attributes.hip. I don't know why they
are different on some bots but reverting this is far too disruptive.
Removing it from the codegen pipeline induces a lot of test churn
because llc is no longer optimizing out implicit arguments to kernels.
Mostly mechanical, but there are some creative test updates. I preferred
to take the changes as-is in tests where the ABI isn't relevant. In
cases where it's more relevant, or the optimize out logic was too
ingrained in the test, I pre-run the optimization. Some cases manually
add attributes to disable inputs.
Inspired by some of the cases from D145468
Let SimplifyDemandedBits handle the narrowing of lshr to half-width if we don't require the upper bits, the narrowed shift is profitable and the zext/trunc are free.
A future patch will propose the equivalent shl narrowing combine.
Differential Revision: https://reviews.llvm.org/D146121
Occupancy is expressed as waves per SIMD. This means that we need to
take into account the number of SIMDs per "CU" or, to be more precise,
the number of SIMDs over which a workgroup may be distributed.
getOccupancyWithLocalMemSize was wrong because it didn't take SIMDs
into account at all.
At the same time, we need to take into account that WGP mode offers
access to a larger total amount of LDS, since this can affect how
non-power-of-two LDS allocations are rounded. To make this work
consistently, we distinguish between (available) local memory size and
addressable local memory size (which is always limited by 64kB on
gfx10+, even with WGP mode).
This change results in a massive amount of test churn. A lot of it is
caused by the fact that the default work group size is 1024, which means
that (due to rounding effects) the default occupancy on older hardware
is 8 instead of 10, which affects scheduling via register pressure
estimates. I've adjusted most tests by just running the UTC tools, but
in some cases I manually changed the work group size to 32 or 64 to make
sure that work group size chunkiness has no effect.
Differential Revision: https://reviews.llvm.org/D139468
This patch contains changes necessary to carry physical condition register (SCC) dependencies through the SDNode scheduler. It adds the edge in the SDNodeScheduler dependency graph instead of inserting the SCC copy between each definition and use. This approach lets the scheduler place instructions in an optimal way placing the copy only when the dependency cannot be resolved.
Reviewed By: rampitec
Differential Revision: https://reviews.llvm.org/D133593
This patch allows SimplifyDemandedBits to call SimplifyMultipleUseDemandedBits in cases where the ISD::SRL source operand has other uses, enabling us to peek through the shifted value if we don't demand all the bits/elts.
This is another step towards removing SelectionDAG::GetDemandedBits and just using TargetLowering::SimplifyMultipleUseDemandedBits.
There a few cases where we end up with extra register moves which I think we can accept in exchange for the increased ILP.
Differential Revision: https://reviews.llvm.org/D77804
Previously SIFoldOperands::foldInstOperand would only fold a
non-inlinable immediate into a single user, so as not to increase code
size by adding the same 32-bit literal operand to many instructions.
This patch removes that restriction, so that a non-inlinable immediate
will be folded into any number of users. The rationale is:
- It reduces the number of registers used for holding constant values,
which might increase occupancy. (On the other hand, many of these
registers are SGPRs which no longer affect occupancy on GFX10+.)
- It reduces ALU stalls between the instruction that loads a constant
into a register, and the instruction that uses it.
- The above benefits are expected to outweigh any increase in code size.
Differential Revision: https://reviews.llvm.org/D114643
We have a pattern that undo sub x, c -> add x, -c canonicalization since c is more likely
an inline immediate than -c. This patch enables it to select scalar or vector subtracion by the input node divergence.
Reviewed By: foad
Differential Revision: https://reviews.llvm.org/D121360
Use shufflevector to do the subvector extracts. This allows a lot more
load merging on AMDGPU and also on NVPTX when <2 x half> is involved.
Differential Revision: https://reviews.llvm.org/D117219
Using a BufferSize of one for memory ProcResources will result in better
ILP since it more accurately models the dependencies between memory ops
and their consumers on an in-order processor. After this change, the
scheduler will treat the data edges from loads as blocking so that
stalls are guaranteed when waiting for data to be retreaved from memory.
Since we don't actually track waitcnt here, this should do a better job
at modeling their behavior.
Practically, this means that the scheduler will trigger the 'STALL'
heuristic more often.
This type of change needs to be evaluated experimentally. Preliminary
results are positive.
Fixes: SWDEV-282962
Reviewed By: rampitec
Differential Revision: https://reviews.llvm.org/D114777
Use GCNHazardRecognizer in postra sched.
Updated tests for the new schedules.
Reviewed By: arsenm
Differential Revision: https://reviews.llvm.org/D109536
Change-Id: Ia86ba2ae168f12fb34b4d8efdab491f84d936cde
Description: This change enables the compare operations to be selected to SALU/VALU form
dependent of the SDNode divergence flag.
Reviewed By: rampitec
Differential Revision: https://reviews.llvm.org/D106079
This is to allow 64 bit constant rematerialization. If a constant
is split into two separate moves initializing sub0 and sub1 like
now RA cannot rematerizalize a 64 bit register.
This gives 10-20% uplift in a set of huge apps heavily using double
precession math.
Fixes: SWDEV-292645
Differential Revision: https://reviews.llvm.org/D104874
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: rampitec
Subscribers: llvm-commits, kerbowa, hiraditya, t-tye, Anastasia, tpr, dstuttard, yaxunl, nhaehnle, wdng, jvesely, kzhuravl, thakis
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D82393
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
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
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
Summary:
Mem op clustering adds a weak edge in the DAG between two loads or
stores that should be clustered, but the direction of this edge is
pretty arbitrary (it depends on the sort order of MemOpInfo, which
represents the operands of a load or store). This often means that two
loads or stores will get reordered even if they would naturally have
been scheduled together anyway, which leads to test case churn and goes
against the scheduler's "do no harm" philosophy.
The fix makes sure that the direction of the edge always matches the
original code order of the instructions.
Reviewers: atrick, MatzeB, arsenm, rampitec, t.p.northover
Subscribers: jvesely, wdng, nhaehnle, kristof.beyls, hiraditya, javed.absar, arphaman, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D72706
This is the more natural lowering, and presents more opportunities to
reduce 64-bit ops to 32-bit.
This should also help avoid issues graphics shaders have had with
64-bit values, and simplify argument lowering in globalisel.
llvm-svn: 366578
Make the FP register callee saved.
This is tricky because now the FP needs to be spilled in the prolog
relative to the incoming SP register, rather than the frame register
used throughout the rest of the function. I don't like how this
bypassess the standard mechanism for CSR spills just to get the
correct insert point. I may look for a better solution, since all CSR
VGPRs may also need to have all lanes activated. Another option might
be to make getFrameIndexReference change the base register if the
frame index is a CSR, and then try to figure out the right insertion
point in emitProlog.
If there is a free VGPR lane available for SGPR spilling, try to use
it for the FP. If that would require intrtoducing a new VGPR spill,
try to use a free call clobbered SGPR. Only fallback to introducing a
new VGPR spill as a last resort.
This also doesn't attempt to handle SGPR spilling with scalar stores.
llvm-svn: 365372
Summary: This change is the first part of the AMDGPU target description
change. The aim of it is the effective splitting the vector and scalar
flows at the selection stage. Selection uses predicate functions based
on the framework implemented earlier - https://reviews.llvm.org/D35267
Differential revision: https://reviews.llvm.org/D52019
Reviewers: rampitec
llvm-svn: 342719
