Emit CFI directives in epilogue and enable CFIFixup pass for RISC-V.
Needs ReviewPublic
Actions

Authored by varunkumare99 on Feb 3 2023, 1:11 AM.

Details

Reviewers

asb
shiva0217
luismarques
probinson
MaskRay
chill

Summary

This patch tries to add cfi directives in the epilogue which were disabled in D69723
Also enables CFIFixup pass which was added in D114545, to emit cfi_remember_state and cfi_restore_state when reordering is done.

This is a prototype on adding cfi directives in the epilogue and emitting cfi_remember_state and cfi_restore_state in RISC-V.
Kindly advise on this and future cases to be handled. I am trying the same from my end.

Diff Detail

Unit TestsFailed
View All

	Time	Test
	80 ms	x64 debian > LLVM.CodeGen/RISCV::O0-pipeline.ll Script: -- : 'RUN: at line 1'; /var/lib/buildkite-agent/builds/llvm-project/build/bin/llc -mtriple=riscv32 -O0 -debug-pass=Structure < /var/lib/buildkite-agent/builds/llvm-project/llvm/test/CodeGen/RISCV/O0-pipeline.ll -o /dev/null 2>&1 \| grep -v "Verify generated machine code" \| /var/lib/buildkite-agent/builds/llvm-project/build/bin/FileCheck /var/lib/buildkite-agent/builds/llvm-project/llvm/test/CodeGen/RISCV/O0-pipeline.ll --check-prefixes=CHECK
	70 ms	x64 debian > LLVM.CodeGen/RISCV::O3-pipeline.ll Script: -- : 'RUN: at line 1'; /var/lib/buildkite-agent/builds/llvm-project/build/bin/llc -mtriple=riscv32 -O3 -debug-pass=Structure < /var/lib/buildkite-agent/builds/llvm-project/llvm/test/CodeGen/RISCV/O3-pipeline.ll -o /dev/null 2>&1 \| grep -v "Verify generated machine code" \| /var/lib/buildkite-agent/builds/llvm-project/build/bin/FileCheck /var/lib/buildkite-agent/builds/llvm-project/llvm/test/CodeGen/RISCV/O3-pipeline.ll --check-prefixes=CHECK
	170 ms	x64 debian > LLVM.CodeGen/RISCV::addimm-mulimm.ll Script: -- : 'RUN: at line 5'; /var/lib/buildkite-agent/builds/llvm-project/build/bin/llc -mtriple=riscv32 -mattr=+m,+zba -verify-machineinstrs < /var/lib/buildkite-agent/builds/llvm-project/llvm/test/CodeGen/RISCV/addimm-mulimm.ll \| /var/lib/buildkite-agent/builds/llvm-project/build/bin/FileCheck -check-prefix=RV32IMB /var/lib/buildkite-agent/builds/llvm-project/llvm/test/CodeGen/RISCV/addimm-mulimm.ll
	70 ms	x64 debian > LLVM.CodeGen/RISCV::addrspacecast.ll Script: -- : 'RUN: at line 2'; /var/lib/buildkite-agent/builds/llvm-project/build/bin/llc -mtriple=riscv32 -verify-machineinstrs < /var/lib/buildkite-agent/builds/llvm-project/llvm/test/CodeGen/RISCV/addrspacecast.ll \| /var/lib/buildkite-agent/builds/llvm-project/build/bin/FileCheck /var/lib/buildkite-agent/builds/llvm-project/llvm/test/CodeGen/RISCV/addrspacecast.ll --check-prefix=RV32I
	90 ms	x64 debian > LLVM.CodeGen/RISCV::aext-to-sext.ll Script: -- : 'RUN: at line 2'; /var/lib/buildkite-agent/builds/llvm-project/build/bin/llc -mtriple=riscv64 -verify-machineinstrs < /var/lib/buildkite-agent/builds/llvm-project/llvm/test/CodeGen/RISCV/aext-to-sext.ll \| /var/lib/buildkite-agent/builds/llvm-project/build/bin/FileCheck /var/lib/buildkite-agent/builds/llvm-project/llvm/test/CodeGen/RISCV/aext-to-sext.ll -check-prefix=RV64I
		View Full Test Results (577 Failed)

Event Timeline

varunkumare99 created this revision.Feb 3 2023, 1:11 AM

Herald added a project: Restricted Project. · View Herald TranscriptFeb 3 2023, 1:11 AM

Herald added subscribers: luke, VincentWu, vkmr and 22 others. · View Herald Transcript

varunkumare99 requested review of this revision.Feb 3 2023, 1:11 AM

Herald added subscribers: llvm-commits, pcwang-thead, MaskRay. · View Herald TranscriptFeb 3 2023, 1:11 AM

Harbormaster completed remote builds in B211665: Diff 494533.Feb 3 2023, 2:32 AM

varunkumare99 added a reviewer: lenary.Feb 8 2023, 8:44 PM

lenary resigned from this revision.Feb 10 2023, 3:51 AM

Hi Varun - thanks for filing bug #60698 to help track this issue, and for noticing that there should now be some upstream support for cfi_remember_state and cfi_restore_state.

The first question that comes to my mind is what degree of testing you've done with this?

In D143248#4122933, @asb wrote:

Hi Varun - thanks for filing bug #60698 to help track this issue, and for noticing that there should now be some upstream support for cfi_remember_state and cfi_restore_state.

The first question that comes to my mind is what degree of testing you've done with this?

Hi Alex,
the patch updates over 500 testcases for RISCV(as it adds cfi info the epilogue), which results in a patch of over 30M. Due to which I'm not able to upload.

varunkumare99 edited reviewers, added: probinson, MaskRay, chill; removed: lenary.Feb 15 2023, 4:02 AM

In D143248#4128698, @varunkumare99 wrote:

In D143248#4122933, @asb wrote:

Hi Varun - thanks for filing bug #60698 to help track this issue, and for noticing that there should now be some upstream support for cfi_remember_state and cfi_restore_state.

The first question that comes to my mind is what degree of testing you've done with this?

Hi Alex,
the patch updates over 500 testcases for RISCV(as it adds cfi info the epilogue), which results in a patch of over 30M. Due to which I'm not able to upload.

If that's the case it sounds like we have a lot of test cases that are emitting CFI directives that really should have nounwind.

Hi Varun,

It might be good to have a simple case to illustrate that cfi_remember_state and cfi_restore_state can be generated properly with epilogue CFI.
For examle, could the following case generated with cfi_remember_state and cfi_restore_state as GCC?

extern void foo(void);
void f(int *i) {
  if (*i == 0) {
    *i = 1;
    foo();
    return;
  } else {
    __builtin_printf("Hi");
    *i=0;
  }
}

This patch emits cfi_same_value directives when the unwind table needs to be reset to the initial one.
By default emission unwind tables is disabled for RISC-V.
Need to pass -fasynchronous-unwind-tables to emit unwind tables.

In D143248#4131555, @shiva0217 wrote:
Hi Varun,

It might be good to have a simple case to illustrate that cfi_remember_state and cfi_restore_state can be generated properly with epilogue CFI.
For examle, could the following case generated with cfi_remember_state and cfi_restore_state as GCC?
extern void foo(void);
void f(int *i) {
  if (*i == 0) {
    *i = 1;
    foo();
    return;
  } else {
    __builtin_printf("Hi");
    *i=0;
  }
}

Hi Shiva, the following is the assembly emitted.

f:
.Lfunc_begin0:
	.loc	0 2 0
	.cfi_startproc
	addi	sp, sp, -16
	.cfi_def_cfa_offset 16
	sd	ra, 8(sp)
	sd	s0, 0(sp)
	.cfi_offset ra, -8
	.cfi_offset s0, -16
	.cfi_remember_state
	mv	s0, a0
.Ltmp0:
	.loc	0 3 6 prologue_end
	lw	a0, 0(a0)
.Ltmp1:
	.loc	0 3 6 is_stmt 0
	beqz	a0, .LBB0_2
.Ltmp2:
	.loc	0 8 3 is_stmt 1
	lui	a0, %hi(.L.str)
	addi	a0, a0, %lo(.L.str)
	call	printf
.Ltmp3:
	.loc	0 9 5
	sw	zero, 0(s0)
.Ltmp4:
	.loc	0 11 1
	ld	ra, 8(sp)
	ld	s0, 0(sp)
.Ltmp5:
	.cfi_restore ra
	.cfi_restore s0
	.loc	0 11 1 epilogue_begin is_stmt 0
	addi	sp, sp, 16
	.cfi_def_cfa_offset 0
	ret
.LBB0_2:
	.cfi_restore_state
.Ltmp6:
	.loc	0 0 1
	li	a0, 1
.Ltmp7:
	.loc	0 4 6 is_stmt 1
	sw	a0, 0(s0)
	.loc	0 5 3
	ld	ra, 8(sp)
	ld	s0, 0(sp)
.Ltmp8:
	.cfi_restore ra
	.cfi_restore s0
	.loc	0 5 3 epilogue_begin is_stmt 0
	addi	sp, sp, 16
	.cfi_def_cfa_offset 0
.Ltmp9:
	tail	foo
.Ltmp10:
.Lfunc_end0:
	.size	f, .Lfunc_end0-f
	.cfi_endproc

Harbormaster completed remote builds in B218064: Diff 503307.Wed, Mar 8, 5:40 AM

In D143248#4128698, @varunkumare99 wrote:

In D143248#4122933, @asb wrote:

Hi Varun - thanks for filing bug #60698 to help track this issue, and for noticing that there should now be some upstream support for cfi_remember_state and cfi_restore_state.

The first question that comes to my mind is what degree of testing you've done with this?

Hi Alex,
the patch updates over 500 testcases for RISCV(as it adds cfi info the epilogue), which results in a patch of over 30M. Due to which I'm not able to upload.