Mirror of the gdb-patches mailing list
 help / color / mirror / Atom feed
From: "Marek Pikuła" <m.pikula@partner.samsung.com>
To: gdb-patches@sourceware.org
Cc: s.rutka@samsung.com, palmer@dabbelt.com, andrew@sifive.com,
	jim.wilson.gcc@gmail.com, nelson@rivosinc.com,
	f.wasil@samsung.com,
	"Marek Pikuła" <m.pikula@partner.samsung.com>
Subject: [PATCH v2] RISC-V: Fix ISA string detection for disassembly
Date: Wed, 28 May 2025 13:54:47 +0200	[thread overview]
Message-ID: <20250528115447.2750177-1-m.pikula@partner.samsung.com> (raw)
In-Reply-To: <CGME20250528115450eucas1p1e57249bb51f4b7f41741772f60186468@eucas1p1.samsung.com>

Commit 3f61a38 introduced a regression where the ISA string was no
longer detected based on the ELF header. The mechanism was changed from
directly referencing `abfd` to using `disassembler_info->section`, which
was not properly initialized for RISC-V.

The previous implementation ignored the object in scope, leading to
issues such as failing to decode RVV instructions when a library was
compiled as `rv64gcv` and the main application as `rv64gc`.

This patch resolves both problems by initializing
`disassembler_info->section` with the object currently in scope,
ensuring correct ISA string detection during disassembly.

Signed-off-by: Marek Pikuła <m.pikula@partner.samsung.com>
---
 gdb/riscv-tdep.c | 16 ++++++++++++++++
 1 file changed, 16 insertions(+)

diff --git a/gdb/riscv-tdep.c b/gdb/riscv-tdep.c
index 8998a297315..01e5dd00153 100644
--- a/gdb/riscv-tdep.c
+++ b/gdb/riscv-tdep.c
@@ -4163,6 +4163,19 @@ riscv_gnu_triplet_regexp (struct gdbarch *gdbarch)
   return "riscv(32|64)?";
 }

+/* Implement the "print_insn" gdbarch method.  */
+static int
+riscv_print_insn (bfd_vma addr, struct disassemble_info *info)
+{
+  /* Initialize the BFD section to enable ISA string detection depending on the
+     object in scope.  */
+  struct obj_section *s = find_pc_section (addr);
+  if (s != nullptr)
+    info->section = s->the_bfd_section;
+
+  return default_print_insn (addr, info);
+}
+
 /* Implementation of `gdbarch_stap_is_single_operand', as defined in
    gdbarch.h.  */

@@ -4429,6 +4442,9 @@ riscv_gdbarch_init (struct gdbarch_info info,
 					  disassembler_options_riscv ());
   set_gdbarch_disassembler_options (gdbarch, &riscv_disassembler_options);

+  /* Disassembler print_insn.  */
+  set_gdbarch_print_insn (gdbarch, riscv_print_insn);
+
   /* SystemTap Support.  */
   set_gdbarch_stap_is_single_operand (gdbarch, riscv_stap_is_single_operand);
   set_gdbarch_stap_register_indirection_prefixes
--
2.49.0

       reply	other threads:[~2025-05-28 11:55 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
     [not found] <CGME20250528115450eucas1p1e57249bb51f4b7f41741772f60186468@eucas1p1.samsung.com>
2025-05-28 11:54 ` Marek Pikuła [this message]
2025-05-28 13:14   ` Marek Pikula
2025-07-09 15:53     ` Marek Pikula
2025-07-17 17:07       ` Andrew Burgess
2025-07-17 17:14         ` Andrew Burgess

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20250528115447.2750177-1-m.pikula@partner.samsung.com \
    --to=m.pikula@partner.samsung.com \
    --cc=andrew@sifive.com \
    --cc=f.wasil@samsung.com \
    --cc=gdb-patches@sourceware.org \
    --cc=jim.wilson.gcc@gmail.com \
    --cc=nelson@rivosinc.com \
    --cc=palmer@dabbelt.com \
    --cc=s.rutka@samsung.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox