https://sourceware.org/bugzilla/show_bug.cgi?id=34224
Bug ID: 34224
Summary: [RISC-V] GNU ld.bfd false truncated JAL at positive
limit after relax
Product: binutils
Version: 2.45
Status: UNCONFIRMED
Severity: normal
Priority: P2
Component: ld
Assignee: unassigned at sourceware dot org
Reporter: l784896635 at gmail dot com
Target Milestone: ---
I ran into this while reducing a RISC-V linker testcase. The reproducer is
already quite small, and I have been seeing the same result consistently across
three reruns.
Summary
=======
riscv64-unknown-elf-ld 2.45 --relax rejects the legal link with relocation
truncated to fit: R_RISCV_JAL against baz, while ld.lld 21.1.3 links the same
object and keeps the final foo to baz call as auipc ra,0x100 plus jalr
ra,0x0(ra).
Expected behavior
=================
A fixed-layout image performs many short direct calls to a nearby helper bar,
then makes one final far call from foo to baz. The last call sits exactly at
the positive jal overflow boundary, so a correct linker must keep it as auipc
plus jalr instead of forcing a jal relaxation.
Environment
===========
- product: binutils
- component: ld
- toolchain route: gas+ld.bfd with ld.lld control
- march: rv64i
- mabi: lp64
- first failing stage: link
Reduced testcase
================
--- case.s ---
.section .text.a, "ax"
.align 2
.globl _start
_start:
.rept 0x100
call bar
.endr
.globl foo
foo:
call baz
.section .text.b, "ax"
.align 2
.globl bar
bar:
nop
.section .text.c, "ax"
.align 2
.globl baz
baz:
nop
--- layout.ld ---
SECTIONS {
.text.a 0x0 : { *(.text.a) }
.text.b : { *(.text.b) }
.text.c 0x100400 : { *(.text.c) }
}
Reproduction notes
==================
- This packaged root does not have a single canonical `run.ps1` wrapper.
- Use the reduced inputs under `case/` and follow the commands documented in
`case/README.md`.
- Stable witness outputs, when present, are preserved under
`verify/run1..run3/`.
What I checked
==============
- Reduced inputs are preserved under case/.
- Stable witness outputs are preserved under verify/run1..run3/.
- The strict recheck says stable normalized run signatures across three runs:
True.
- Tracker guidance link:
https://sourceware.org/binutils/docs/ld/Bug-Reporting.html
- evidence summary: 3 clean reproductions under
hunt/verify/bfd_false_truncated_jal_at_positive_limit/run1..run3 show stable
gas assembly success, stable ld.bfd link-time failure at .text.a+0x400 with
relocation truncated to fit: R_RISCV_JAL against baz, and stable ld.lld success
with final disassembly keeping foo at 0x400 and baz at 0x100400 connected by
auipc plus jalr rather than jal.
Notes
=====
local reduced fixed-address case derived from
bfd-relax-truncated-20260526T083500Z and upstream relax-truncated patch family
Root key: bfd.riscv.false_truncated_jal_at_positive_limit_after_relax
Case id: 20260526-bfd-false-truncated-jal-positive-limit
--
You are receiving this mail because:
You are on the CC list for the bug.
