From: Tom de Vries <tdevries@suse.de>
To: gdb-patches@sourceware.org
Subject: Re: [PATCH 02/11] [gdb/macros] Handle 64-bit dwarf in gdb.base/fission-macro.exp
Date: Tue, 21 Jan 2025 11:26:52 +0100 [thread overview]
Message-ID: <d96a2425-246b-46b5-9583-6cca406f7765@suse.de> (raw)
In-Reply-To: <20240521154415.9543-2-tdevries@suse.de>
On 5/21/24 17:44, Tom de Vries wrote:
> When enabling 64-bit dwarf in the test-case gdb.base/fission-macro.exp, we run
> into:
> ...
> (gdb) break -qualified main^M
> DW_MACRO_define_strx pointing outside of .debug_str.dwo section \
> [in module fission-macro-5-640.dwo]^M
> (gdb) FAIL: gdb.base/fission-macro.exp: dwarf_version=5: dwarf_bits=64: \
> gdb_breakpoint: set breakpoint at main
> ...
>
> A dwarf-5 .debug_str_offsets section starts with a header consisting of:
> - an initial length (4 bytes for 32-bit and 12 bytes for 64-bit),
> - a 2 byte version string, and
> - 2 bytes padding
> so in total 8 bytes for 32-bit and 16 bytes for 64-bit.
>
> This offset is calculated here in dwarf_decode_macros:
> ...
> str_offsets_base = cu->header.addr_size;
> ...
> and in both cases this evaluates to 8.
>
> So in the 64-bit case, we interpret the last 8 bytes of the header as an offset.
>
> Fix this by computing str_offsets_base correctly for dwarf-5, for both the
> 32-bit and 64-bit case.
>
I've combined this patch and "Handle v4 dwarf in
gdb.base/fission-macro.exp", in other words 02/11 and 06/11 and
resubmitted as
https://sourceware.org/pipermail/gdb-patches/2025-January/214901.html .
I've also changed the code to use "cu->header.version" instead of
"cu->per_cu->version ()". ( See also
https://sourceware.org/pipermail/gdb-patches/2025-January/214886.html ).
Thanks,
- Tom
> Tested on x86_64-linux.
>
> Tested test-case using a current gcc trunk build, and gcc 14.
> ---
> gdb/dwarf2/read.c | 11 ++++++++++-
> gdb/testsuite/gdb.base/fission-macro.exp | 2 +-
> 2 files changed, 11 insertions(+), 2 deletions(-)
>
> diff --git a/gdb/dwarf2/read.c b/gdb/dwarf2/read.c
> index e8416905163..4884553000a 100644
> --- a/gdb/dwarf2/read.c
> +++ b/gdb/dwarf2/read.c
> @@ -21232,7 +21232,16 @@ dwarf_decode_macros (struct dwarf2_cu *cu, unsigned int offset,
> str_offsets_section = &cu->dwo_unit->dwo_file
> ->sections.str_offsets;
> str_section = &cu->dwo_unit->dwo_file->sections.str;
> - str_offsets_base = cu->header.addr_size;
> + if (cu->per_cu->version () <= 4)
> + str_offsets_base = cu->header.addr_size;
> + else
> + {
> + bfd *abfd = str_offsets_section->get_bfd_owner ();
> + unsigned int bytes_read = 0;
> + read_initial_length (abfd, str_offsets_section->buffer, &bytes_read, false);
> + const bool is_dwarf64 = bytes_read != 4;
> + str_offsets_base = is_dwarf64 ? 16 : 8;
> + }
> }
> else
> {
> diff --git a/gdb/testsuite/gdb.base/fission-macro.exp b/gdb/testsuite/gdb.base/fission-macro.exp
> index 705e3dd0c4b..4eb9c0efe21 100644
> --- a/gdb/testsuite/gdb.base/fission-macro.exp
> +++ b/gdb/testsuite/gdb.base/fission-macro.exp
> @@ -67,7 +67,7 @@ proc do_tests { dwarf_version dwarf_bits strict_dwarf } {
> }
>
> foreach_with_prefix dwarf_version {5} {
> - foreach_with_prefix dwarf_bits {32} {
> + foreach_with_prefix dwarf_bits {32 64} {
> foreach_with_prefix strict_dwarf {0 1} {
> do_tests $dwarf_version $dwarf_bits $strict_dwarf
> }
next prev parent reply other threads:[~2025-01-21 10:27 UTC|newest]
Thread overview: 19+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-05-21 15:44 [PATCH 01/11] [gdb/testsuite] Add gdb.base/fission-macro.exp Tom de Vries
2024-05-21 15:44 ` [PATCH 02/11] [gdb/macros] Handle 64-bit dwarf in gdb.base/fission-macro.exp Tom de Vries
2025-01-21 10:26 ` Tom de Vries [this message]
2024-05-21 15:44 ` [PATCH 03/11] [gdb/testsuite] Add test-case gdb.base/fission-macro-i.exp Tom de Vries
2024-06-13 16:27 ` Tom de Vries
2024-05-21 15:44 ` [PATCH 04/11] [gdb/macros] Work around gcc PR debug/99319 Tom de Vries
2024-05-21 15:44 ` [PATCH 05/11] [gdb/macros] Work around a gcc PR fixed in gcc 9 Tom de Vries
2024-05-21 15:44 ` [PATCH 06/11] [gdb/macros] Handle v4 dwarf in gdb.base/fission-macro.exp Tom de Vries
2024-05-21 15:44 ` [PATCH 07/11] [gdb/macros] Workaround gcc PR debug/115066 Tom de Vries
2024-05-21 15:44 ` [PATCH 08/11] [gdb/testsuite] Extend gdb.base/fission-macro.exp Tom de Vries
2024-05-21 15:44 ` [PATCH 09/11] [gdb/macros] Fix gdb.base/fission-macro.exp with clang Tom de Vries
2025-01-21 14:28 ` [PING][PATCH " Tom de Vries
2024-05-21 15:44 ` [PATCH 10/11] [gdb/testsuite] Use -g3 in gdb.base/lineinc.exp Tom de Vries
2025-01-21 11:42 ` Tom de Vries
2024-05-21 15:44 ` [PATCH 11/11] [gdb/testsuite] Don't use readelf in gdb/contrib/cc-with-tweaks.sh Tom de Vries
2024-06-13 15:40 ` [PATCH 01/11] [gdb/testsuite] Add gdb.base/fission-macro.exp Tom de Vries
2024-06-17 8:16 ` Christophe Lyon
2024-06-17 8:23 ` Tom de Vries
2024-06-17 8:43 ` Christophe Lyon
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=d96a2425-246b-46b5-9583-6cca406f7765@suse.de \
--to=tdevries@suse.de \
--cc=gdb-patches@sourceware.org \
/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