[X86][MC] Add R_X86_64_CODE_4_GOTTPOFF #116633
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
@llvm/pr-subscribers-backend-x86 @llvm/pr-subscribers-llvm-binary-utilities Author: Feng Zou (fzou1) ChangesFor mov name@GOTTPOFF(%rip), %reg add
if the instruction starts at 4 bytes before the relocation offset. It's similar to R_X86_64_GOTTPOFF. Linker can treat mov $name@tpoff, %reg if the first byte of the instruction at the relocation Full diff: https://github.com/llvm/llvm-project/pull/116633.diff 3 Files Affected:
diff --git a/llvm/include/llvm/BinaryFormat/ELFRelocs/x86_64.def b/llvm/include/llvm/BinaryFormat/ELFRelocs/x86_64.def
index 161b1969abfeb4..ff4516d1edf507 100644
--- a/llvm/include/llvm/BinaryFormat/ELFRelocs/x86_64.def
+++ b/llvm/include/llvm/BinaryFormat/ELFRelocs/x86_64.def
@@ -44,3 +44,4 @@ ELF_RELOC(R_X86_64_IRELATIVE, 37)
ELF_RELOC(R_X86_64_GOTPCRELX, 41)
ELF_RELOC(R_X86_64_REX_GOTPCRELX, 42)
ELF_RELOC(R_X86_64_REX2_GOTPCRELX, 43)
+ELF_RELOC(R_X86_64_CODE_4_GOTTPOFF, 44)
\ No newline at end of file
diff --git a/llvm/lib/Target/X86/MCTargetDesc/X86ELFObjectWriter.cpp b/llvm/lib/Target/X86/MCTargetDesc/X86ELFObjectWriter.cpp
index 90222278d1ad6f..ab363e4eb8dce3 100644
--- a/llvm/lib/Target/X86/MCTargetDesc/X86ELFObjectWriter.cpp
+++ b/llvm/lib/Target/X86/MCTargetDesc/X86ELFObjectWriter.cpp
@@ -197,6 +197,9 @@ static unsigned getRelocType64(MCContext &Ctx, SMLoc Loc,
return ELF::R_X86_64_TLSGD;
case MCSymbolRefExpr::VK_GOTTPOFF:
checkIs32(Ctx, Loc, Type);
+ if ((unsigned)Kind == X86::reloc_riprel_4byte_movq_load_rex2 ||
+ (unsigned)Kind == X86::reloc_riprel_4byte_relax_rex2)
+ return ELF::R_X86_64_CODE_4_GOTTPOFF;
return ELF::R_X86_64_GOTTPOFF;
case MCSymbolRefExpr::VK_TLSLD:
checkIs32(Ctx, Loc, Type);
diff --git a/llvm/test/MC/X86/tls.s b/llvm/test/MC/X86/tls.s
new file mode 100644
index 00000000000000..418ab692cac576
--- /dev/null
+++ b/llvm/test/MC/X86/tls.s
@@ -0,0 +1,40 @@
+// RUN: llvm-mc -filetype=obj -triple=x86_64 %s -o %t
+// RUN: llvm-objdump -dr --no-print-imm-hex %t | FileCheck %s
+
+// CHECK: <_start>:
+// CHECK-NEXT: movq (%rip), %r16
+// CHECK-NEXT: R_X86_64_CODE_4_GOTTPOFF tls0-0x4
+// CHECK-NEXT: movq (%rip), %r20
+// CHECK-NEXT: R_X86_64_CODE_4_GOTTPOFF tls0-0x4
+// CHECK-NEXT: movq (%rip), %r16
+// CHECK-NEXT: R_X86_64_CODE_4_GOTTPOFF tls1-0x4
+// CHECK-NEXT: addq (%rip), %r16
+// CHECK-NEXT: R_X86_64_CODE_4_GOTTPOFF tls0-0x4
+// CHECK-NEXT: addq (%rip), %r28
+// CHECK-NEXT: R_X86_64_CODE_4_GOTTPOFF tls0-0x4
+// CHECK-NEXT: addq (%rip), %r16
+// CHECK-NEXT: R_X86_64_CODE_4_GOTTPOFF tls1-0x4
+
+.type tls0,@object
+.section .tbss,"awT",@nobits
+.globl tls0
+.align 4
+tls0:
+.long 0
+.size tls0, 4
+.type tls1,@object
+.globl tls1
+.align 4
+tls1:
+.long 0
+.size tls1, 4
+.section .text
+.globl _start
+_start:
+ # EGPR
+ movq tls0@GOTTPOFF(%rip), %r16
+ movq tls0@GOTTPOFF(%rip), %r20
+ movq tls1@GOTTPOFF(%rip), %r16
+ addq tls0@GOTTPOFF(%rip), %r16
+ addq tls0@GOTTPOFF(%rip), %r28
+ addq tls1@GOTTPOFF(%rip), %r16
\ No newline at end of file
|
fzou1
added a commit
to fzou1/llvm-project
that referenced
this pull request
Nov 18, 2024
For mov name@GOTTPOFF(%rip), %reg add name@GOTTPOFF(%rip), %reg add `R_X86_64_CODE_4_GOTTPOFF` = 44 in llvm#116633. Linker can treat `R_X86_64_CODE_4_GOTTPOFF` as `R_X86_64_GOTTPOFF` or convert the instructions above to mov $name, %reg add $name, %reg if the first byte of the instruction at the relocation `offset - 4` is `0xd5` (namely, encoded w/REX2 prefix) when possible.
MaskRay
reviewed
Nov 19, 2024
| @@ -44,3 +44,4 @@ ELF_RELOC(R_X86_64_IRELATIVE, 37) | |||
| ELF_RELOC(R_X86_64_GOTPCRELX, 41) | |||
| ELF_RELOC(R_X86_64_REX_GOTPCRELX, 42) | |||
| ELF_RELOC(R_X86_64_REX2_GOTPCRELX, 43) | |||
| ELF_RELOC(R_X86_64_CODE_4_GOTTPOFF, 44) | |||
MaskRay
reviewed
Nov 19, 2024
llvm/test/MC/X86/tls.s
Outdated
|
|
||
| .type tls0,@object | ||
| .section .tbss,"awT",@nobits | ||
| .globl tls0 |
There was a problem hiding this comment.
drop .align and .size, unneeded by this test
MaskRay
reviewed
Nov 19, 2024
llvm/test/MC/X86/tls.s
Outdated
| movq tls1@GOTTPOFF(%rip), %r16 | ||
| addq tls0@GOTTPOFF(%rip), %r16 | ||
| addq tls0@GOTTPOFF(%rip), %r28 | ||
| addq tls1@GOTTPOFF(%rip), %r16 |
KanRobert
reviewed
Nov 20, 2024
llvm/test/MC/X86/tls.s
Outdated
| // RUN: llvm-mc -filetype=obj -triple=x86_64 %s -o %t | ||
| // RUN: llvm-objdump -dr --no-print-imm-hex %t | FileCheck %s | ||
|
|
||
| // CHECK: <_start>: |
There was a problem hiding this comment.
Why not use the existing test file llvm/test/MC/ELF/relocation.s?
There was a problem hiding this comment.
There were no checks for both instructions and relocation types emitted by llvm-mc, so I added new test file for this R_X86_64_CODE_4_GOTTPOFF type and other relocation types to be added.
There was a problem hiding this comment.
Why do we need to check the instructions? There should be existing encoding/decoding tests there.
IMHO, checking the similar relocation types in a single file can help us not miss some checks?
There was a problem hiding this comment.
It's not to check encoding/decoding. It showed the map of instruction and relocation information (type, symbol and offset), and that's the output what users got from llvm-mc/llvm. Some tests in llvm/test/MC folder did same thing.
I'm okay with adding the check of relocation information in relocation.s and not adding this test file, if you insist.
There was a problem hiding this comment.
I've removed the test file newly added and add the check in relocation.s.
For mov name@GOTTPOFF(%rip), %reg add name@GOTTPOFF(%rip), %reg add `R_X86_64_CODE_4_GOTTPOFF` = 44 if the instruction starts at 4 bytes before the relocation offset. It's similar to R_X86_64_GOTTPOFF. Linker can treat `R_X86_64_CODE_4_GOTTPOFF` as `R_X86_64_GOTTPOFF` or convert the instructions above to mov $name@tpoff, %reg add $name@tpoff, %reg if the first byte of the instruction at the relocation `offset - 4` is `0xd5` (namely, encoded w/REX2 prefix) when possible.
fzou1
force-pushed
the
mc-code-4-gottpoff
branch
from
1dd700d to
2031939
Compare
|
@MaskRay, any comments? If no objection, @KanRobert /@MaskRay, please help merge this PR. I don't have commit access. Thanks. |
MaskRay
approved these changes
Nov 21, 2024
|
LLVM Buildbot has detected a new failure on builder Full details are available at: https://lab.llvm.org/buildbot/#/builders/30/builds/10712 Here is the relevant piece of the build log for the reference |
fzou1
added a commit
to fzou1/llvm-project
that referenced
this pull request
Nov 21, 2024
For mov name@GOTTPOFF(%rip), %reg add name@GOTTPOFF(%rip), %reg add `R_X86_64_CODE_4_GOTTPOFF` = 44 in llvm#116633. Linker can treat `R_X86_64_CODE_4_GOTTPOFF` as `R_X86_64_GOTTPOFF` or convert the instructions above to mov $name, %reg add $name, %reg if the first byte of the instruction at the relocation `offset - 4` is `0xd5` (namely, encoded w/REX2 prefix) when possible.
fzou1
added a commit
that referenced
this pull request
Nov 27, 2024
For mov name@GOTTPOFF(%rip), %reg add name@GOTTPOFF(%rip), %reg add `R_X86_64_CODE_4_GOTTPOFF` = 44 in #116633. Linker can treat `R_X86_64_CODE_4_GOTTPOFF` as `R_X86_64_GOTTPOFF` or convert the instructions above to mov $name, %reg add $name, %reg if the first byte of the instruction at the relocation `offset - 4` is `0xd5` (namely, encoded w/REX2 prefix) when possible. Binutils patch: bminor/binutils-gdb@a533c8d Binutils mailthread: https://sourceware.org/pipermail/binutils/2023-December/131463.html ABI discussion: https://groups.google.com/g/x86-64-abi/c/ACwD-UQXVDs/m/vrgTenKyFwAJ Blog: https://kanrobert.github.io/rfc/All-about-APX-relocation
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
For
mov name@GOTTPOFF(%rip), %reg
add name@GOTTPOFF(%rip), %reg
add
R_X86_64_CODE_4_GOTTPOFF= 44if the instruction starts at 4 bytes before the relocation offset. It's similar to R_X86_64_GOTTPOFF.
Linker can treat
R_X86_64_CODE_4_GOTTPOFFasR_X86_64_GOTTPOFFor convert the instructions above tomov $name@tpoff, %reg
add $name@tpoff, %reg
if the first byte of the instruction at the relocation
offset - 4is0xd5(namely, encoded w/REX2 prefix) when possible.Binutils patch: bminor/binutils-gdb@a533c8d
Binutils mailthread: https://sourceware.org/pipermail/binutils/2023-December/131463.html
ABI discussion: https://groups.google.com/g/x86-64-abi/c/ACwD-UQXVDs/m/vrgTenKyFwAJ
Blog: https://kanrobert.github.io/rfc/All-about-APX-relocation