[RISCV] Enable bidirectional scheduling and tracking register pressure (#115445)


This is based on other targets like PPC/AArch64 and some experiments.

This PR will only enable bidirectional schedu

[RISCV] Enable bidirectional scheduling and tracking register pressure (#115445)


This is based on other targets like PPC/AArch64 and some experiments.

This PR will only enable bidirectional scheduling and tracking register
pressure.

Disclaimer: I haven't tested it on many cores, maybe we should make
some options being features. I believe downstreams must have tried
this before, so feedbacks are welcome.

show more ...

[RISCV] Add 32 bit GPR sub-register for Zfinx. (#108336)

This patches adds a 32 bit register class for use with Zfinx instructions. This makes them more similar to F instructions and allows us to on

[RISCV] Add 32 bit GPR sub-register for Zfinx. (#108336)

This patches adds a 32 bit register class for use with Zfinx instructions. This makes them more similar to F instructions and allows us to only spill 32 bits.

I've added CodeGenOnly instructions for load/store using GPRF32 as that gave better results than insert_subreg/extract_subreg.

Function arguments use this new GPRF32 register class for f32 arguments with Zfinx. Eliminating the need to use RISCVISD::FMV* nodes.

This is similar to #107446 which adds a 16 bit register class.

show more ...

[RISCV] Pass f32/f64 directly without a bitcast for Zfinx/Zdinx. (#107464)

With Zfinx/Zdinx, f32/f64 are legal types for a GPR, we don't need a
bitcast.

This avoids turning fneg/fabs into bitwis

[RISCV] Pass f32/f64 directly without a bitcast for Zfinx/Zdinx. (#107464)

With Zfinx/Zdinx, f32/f64 are legal types for a GPR, we don't need a
bitcast.

This avoids turning fneg/fabs into bitwise operations purely because of
these bitcasts. If the bitwise operations are faster for some reason on
a Zfinx CPU, then that seems like it should be done for all fneg/fabs,
not just ones near function arguments/returns.

I don't have much interest in Zfinx, this just makes the code more
similar to what I proposed for Zhinx in #107446.

show more ...

[RISCV][ISel] Fold FSGNJX idioms (#100718)

This patch folds `fmul X, (fcopysign 1.0, Y)` into `fsgnjx X, Y`. This
pattern exists in some graphics applications/math libraries.
Alive2: https://alive

[RISCV][ISel] Fold FSGNJX idioms (#100718)

This patch folds `fmul X, (fcopysign 1.0, Y)` into `fsgnjx X, Y`. This
pattern exists in some graphics applications/math libraries.
Alive2: https://alive2.llvm.org/ce/z/epyL33

Since fpimm +1.0 is lowered to a load from constant pool after
OpLegalization, I have to introduce a new RISCVISD node FSGNJX and fold
this pattern in DAGCombine.

Closes https://github.com/dtcxzyw/llvm-opt-benchmark/issues/1072.

show more ...

[RISCV] Omit "@plt" in assembly output "call foo@plt" (#72467)

R_RISCV_CALL/R_RISCV_CALL_PLT distinction is not necessary and
R_RISCV_CALL has been deprecated. Since https://reviews.llvm.org/D132530

[RISCV] Omit "@plt" in assembly output "call foo@plt" (#72467)

R_RISCV_CALL/R_RISCV_CALL_PLT distinction is not necessary and
R_RISCV_CALL has been deprecated. Since https://reviews.llvm.org/D132530
`call foo` assembles to R_RISCV_CALL_PLT. The `@plt` suffix is not
useful and can be removed now (matching AArch64 and PowerPC).

GNU assembler assembles `call foo` to RISCV_CALL_PLT since 2022-09
(70f35d72ef04cd23771875c1661c9975044a749c).

Without this patch, unconditionally changing MO_CALL to MO_PLT could
create `jump .L1@plt, a0`, which is invalid in LLVM integrated assembler
and GNU assembler.

show more ...

[RISCV][CodeGen] Support Zfinx codegen

This patch was split from D122918 . Co-Author: @liaolucy @realqhc

Reviewed By: craig.topper

Differential Revision: https://reviews.llvm.org/D148874

[RISCV] Move compressible registers to the beginning of the FP allocation order.

We don't have very many compressible FP instructions, just load and store.
These instruction require the FP register

[RISCV] Move compressible registers to the beginning of the FP allocation order.

We don't have very many compressible FP instructions, just load and store.
These instruction require the FP register to be f8-f15.

This patch changes the FP allocation order to prioritize f10-f15 first.
These are also the FP argument registers. So I allocated them in reverse
order starting at f15 to avoid taking the first argument registers.
This appears to match gcc allocation order.

Reviewed By: asb

Differential Revision: https://reviews.llvm.org/D146488

show more ...

[RISCV] Add ADDW/AND/OR/XOR/SUB/SUBW to getRegAllocHints.

These instructions requires both register operands to be compressible
so I've only applied the hint if we already have a GPRC physical regis

[RISCV] Add ADDW/AND/OR/XOR/SUB/SUBW to getRegAllocHints.

These instructions requires both register operands to be compressible
so I've only applied the hint if we already have a GPRC physical register
assigned for the other register operand.

Reviewed By: reames

Differential Revision: https://reviews.llvm.org/D139079

show more ...

[RISCV] Teach getRegAllocationHints about compressible SRAI/SRLI.

Similar to previous patches for ADDI/ADDIW/SLLI/ADD, but restricted
to only cases where the register is x8-x15(GPRC reg class).

I'v

[RISCV] Teach getRegAllocationHints about compressible SRAI/SRLI.

Similar to previous patches for ADDI/ADDIW/SLLI/ADD, but restricted
to only cases where the register is x8-x15(GPRC reg class).

I've restricted it so that we can be precise about whether the
resulting instruction would be compressible. Changing the register
allocation may make some other instruction not compressible so we
should try to be accurate.

Reviewed By: asb

Differential Revision: https://reviews.llvm.org/D138740

show more ...

[RISCV] Use register allocation hints to improve use of compressed instructions.

Compressed instructions usually require one of the source registers
to also be the source register. The register allo

[RISCV] Use register allocation hints to improve use of compressed instructions.

Compressed instructions usually require one of the source registers
to also be the source register. The register allocator doesn't have
that bias on its own.

This patch adds register allocation hints to introduce this bias.
I've started with ADDI, ADDIW, and SLLI. These all have a 5-bit
field for the register. If the source and dest register are the
same they are guaranteed to compress as long as the immediate is
also 6 bits.

This code was inspired by similar code from the SystemZ target.

Reviewed By: reames

Differential Revision: https://reviews.llvm.org/D138242

show more ...

[RISCV][test][NFC] Regenerate RISC-V tests with update_llc_test_checks.py -u

If a change alters more than a couple of tests it's really handy to be
able to regenerate any that were created by update

[RISCV][test][NFC] Regenerate RISC-V tests with update_llc_test_checks.py -u

If a change alters more than a couple of tests it's really handy to be
able to regenerate any that were created by update_llc_test_checks.py
with something like `update_llc_test_checks.py -u
llvm/test/CodeGen/RISCV`. I noticed this causes some extraneous changes
(perhaps due to hand editing). This commit addresses that by updating
any fails that are modified by update_llc_test_checks.py -u.

show more ...

[RISCV] Use check-prefixes to reduce check lines

Reviewed By: frasercrmck

Differential Revision: https://reviews.llvm.org/D125083

[RISCV] Add more patterns for FNMADD

D54205 handles fnmadd: -rs1 * rs2 - rs3
This patch add fnmadd: -(rs1 * rs2 + rs3) (the nsz flag on the FMA)

Reviewed By: craig.topper

Differential Revision: ht

[RISCV] Add more patterns for FNMADD

D54205 handles fnmadd: -rs1 * rs2 - rs3
This patch add fnmadd: -(rs1 * rs2 + rs3) (the nsz flag on the FMA)

Reviewed By: craig.topper

Differential Revision: https://reviews.llvm.org/D126852

show more ...

[RISCV] Set CostPerUse to 1 iff RVC is enabled

After D86836, we can define multiple cost values for
different cost models. So here we set CostPerUse to
1 iff RVC is enabled to avoid potential impact

[RISCV] Set CostPerUse to 1 iff RVC is enabled

After D86836, we can define multiple cost values for
different cost models. So here we set CostPerUse to
1 iff RVC is enabled to avoid potential impact on RA.

Reviewed By: craig.topper

Differential Revision: https://reviews.llvm.org/D117741

show more ...

[RISCV] Use shift for zero extension when Zbb and Zbp are not enabled

Now AND is used for zero extension when both Zbb and Zbp are not enabled.
It may be better to use shift operation if the trailin

[RISCV] Use shift for zero extension when Zbb and Zbp are not enabled

Now AND is used for zero extension when both Zbb and Zbp are not enabled.
It may be better to use shift operation if the trailing ones mask exceeds simm12.

This patch optimzes LUI+ADDI+AND to SLLI+SRLI.

Reviewed By: craig.topper

Differential Revision: https://reviews.llvm.org/D116720

show more ...

[RISCV] Use FP ABI on some of the FP tests to reduce the number of CHECK lines. NFC

These tests are interested in the FP instructions being used, not
the conversions needed to pass the arguments/ret

[RISCV] Use FP ABI on some of the FP tests to reduce the number of CHECK lines. NFC

These tests are interested in the FP instructions being used, not
the conversions needed to pass the arguments/returns in GPRs.

Reviewed By: asb

Differential Revision: https://reviews.llvm.org/D116869

show more ...

[RISCV] Reduce duplicate FP test cases.

-Remove feq, fle, flt tests from *-arith.ll in favor of *-fcmp.ll which tests all predicates.

Reviewed By: asb

Differential Revision: https://reviews.llvm.o

[RISCV] Reduce duplicate FP test cases.

-Remove feq, fle, flt tests from *-arith.ll in favor of *-fcmp.ll which tests all predicates.

Reviewed By: asb

Differential Revision: https://reviews.llvm.org/D113703

show more ...

[RISCV] Reverse the order of loading/storing callee-saved registers.

Currently, we restore the return address register as the last restoring
instruction in the epilog. The next instruction is `ret`

[RISCV] Reverse the order of loading/storing callee-saved registers.

Currently, we restore the return address register as the last restoring
instruction in the epilog. The next instruction is `ret` usually. It is
a use of return address register. In some microarchitectures, there is
load-to-use data hazard. To avoid the load-to-use data hazard, we could
separate the load instruction from its use as far as possible. In this
patch, we reverse the order of restoring callee-saved registers to
increase the distance of `load ra` and `ret` in the epilog.

Differential Revision: https://reviews.llvm.org/D113967

show more ...

[RISCV] Generate pseudo instruction li

Add an alias of `addi [x], zero, imm` to generate pseudo
instruction li, which makes assembly mush more readable.
For existed tests, users can update them by r

[RISCV] Generate pseudo instruction li

Add an alias of `addi [x], zero, imm` to generate pseudo
instruction li, which makes assembly mush more readable.
For existed tests, users can update them by running script
`llvm/utils/update_llc_test_checks.py`.

Reviewed By: asb

Differential Revision: https://reviews.llvm.org/D112692

show more ...

[RISCV] Add rv32i/rv64i command lines to some floating point tests. NFC

This improves our coverage of soft float libcalls lowering.

Remove most of the test cases from rv64i-single-softfloat.ll. The

[RISCV] Add rv32i/rv64i command lines to some floating point tests. NFC

This improves our coverage of soft float libcalls lowering.

Remove most of the test cases from rv64i-single-softfloat.ll. They
were duplicated in the test files that now test softflow. Only
a couple test cases for constrained FP remain. Those should be
removed when we start supporting constrained FP.

This is follow up from D113528.

show more ...

[RISCV] Add an implementation of isFMAFasterThanFMulAndFAdd

Start with an assumption that FMA is faster than Fmul+FAdd. If thats not true
on some particular implementation we can add a tuning parame

[RISCV] Add an implementation of isFMAFasterThanFMulAndFAdd

Start with an assumption that FMA is faster than Fmul+FAdd. If thats not true
on some particular implementation we can add a tuning parameter in the future.

I've update the fmuladd test cases and added new test cases for fast math flag
based contraction.

Differential Revision: https://reviews.llvm.org/D91987

show more ...

[SelectionDAG][ARM][AArch64][Hexagon][RISCV][X86] Add SDNPCommutative to fma and fmad nodes in tablegen. Remove explicit commuted patterns from targets.

X86 was already specially marking fma as comm

[SelectionDAG][ARM][AArch64][Hexagon][RISCV][X86] Add SDNPCommutative to fma and fmad nodes in tablegen. Remove explicit commuted patterns from targets.

X86 was already specially marking fma as commutable which allowed
tablegen to autogenerate commuted patterns. This moves it to the target
independent definition and fix up the targets to remove now
unneeded patterns.

Unfortunately, the tests change because the commuted version of
the patterns are generating operands in a different than the
explicit patterns.

Differential Revision: https://reviews.llvm.org/D91842

show more ...

[RISCV] Add isel patterns for fnmadd/fnmsub with an fneg on the second operand instead of the first.

The multiply part of FMA is commutable, but TargetSelectionDAG.td
doesn't have it marked as commu

[RISCV] Add isel patterns for fnmadd/fnmsub with an fneg on the second operand instead of the first.

The multiply part of FMA is commutable, but TargetSelectionDAG.td
doesn't have it marked as commutable so tablegen won't automatically
create the additional patterns.

So manually add commuted patterns.

show more ...

[RISCV] Add test cases to show missed opportunities to use fnmadd/fnmsub if the second operand to the fma is negated rather than the first. NFC

We need to add more isel patterns to handle this.

[RISCV] Select +0.0 immediate using fmv.{w,d}.x / fcvt.d.w

Floating point positive zero can be selected using fmv.w.x / fmv.d.x /
fcvt.d.w and the zero source register.

Differential Revision: https

[RISCV] Select +0.0 immediate using fmv.{w,d}.x / fcvt.d.w

Floating point positive zero can be selected using fmv.w.x / fmv.d.x /
fcvt.d.w and the zero source register.

Differential Revision: https://reviews.llvm.org/D75729

show more ...