public inbox for binutils@sourceware.org
 help / color / mirror / Atom feed
From: "Willgerodt, Felix" <felix.willgerodt@intel.com>
To: "binutils@sourceware.org" <binutils@sourceware.org>
Cc: "gdb-patches@sourceware.org" <gdb-patches@sourceware.org>
Subject: [PING] [PATCH 1/1] bfd, gdb: fix missing "Core was generated by" when loading a x32 corefile.
Date: Mon, 20 Feb 2023 12:50:36 +0000	[thread overview]
Message-ID: <MN2PR11MB456669E8FCBFA087E013CA658EA49@MN2PR11MB4566.namprd11.prod.outlook.com> (raw)
In-Reply-To: <20230124134202.2452845-1-felix.willgerodt@intel.com>

*Ping*

Thanks,
Felix

> -----Original Message-----
> From: Willgerodt, Felix <felix.willgerodt@intel.com>
> Sent: Dienstag, 24. Januar 2023 14:42
> To: binutils@sourceware.org
> Cc: gdb-patches@sourceware.org; Willgerodt, Felix
> <felix.willgerodt@intel.com>; Christoph Rotte <christoph.rotte@intel.com>
> Subject: [PATCH 1/1] bfd, gdb: fix missing "Core was generated by" when
> loading a x32 corefile.
> 
> When loading a corefile that was generated using 'gcore', gdb usually
> displays "Core was generated by" followed by the path of the
> executable.  If the binary, however, was compiled with '-mx32', the
> 'Elf_Internal_Note' (bfd/elf64-x86-64.c:elf_x86_64_grok_psinfo)
> contains a 4 byte offset compared to the default 'struct elf_prpsinfo'
> on Linux/x32:
> 
>   note of 124 byte (initial default case on Linux/x32):
> 
> 	 | pid | program | command
>   offset |  12 |      28 |      44
> 
>   note of 128 byte (Linux/x32):
> 
> 	 | pid | program | command
>   offset |  12 |      32 |      48
> 
> Initially, the 'Elf_Internal_Note' had a size of 124 bytes for x32
> and for 32 bit.
> Commit a2f63b2 ("ELF/BFD,GDB: Handle both variants of the 32-bit Linux
> core PRPSINFO note") changed that by differentiating between 2 and 4
> bytes for 'uid' and 'gid' and not handling the resulting offset in the
> function 'elf_x86_64_grok_psinfo'.
> 
> The reason for this are the different sizes for '__kernel_uid_t'
> and '__kernel_gid_t' (16 vs 32).
> 'bfd/elf.c:elfcore_write_linux_prpsinfo32' differentiates between
> 'elf_external_linux_prpsinfo32_ugid16' and
> 'elf_external_linux_prpsinfo32_ugid32'.
> 
> Consequently, the function 'elf_x86_64_grok_psinfo' cannot find the
> respective information and does not assign the path of the executable to
> the current 'bfd' object.  The function directly returns with the
> respective path string still being null.
> 
> The proposed fix circumvents this by adding a separate case for
> 128 byte notes with the respective offsets.
> 
> 'gdb.base/gcore-relro.exp' already tests the expected behaviour;
> the fail before this patch can be reproduced using
> '--target_board=unix/-mx32'.
> 
> Co-authored-by: Christoph Rotte  <christoph.rotte@intel.com>
> 
> bfd/ChangeLog:
> 2023-01-24  Felix Willgerodt  <felix.willgerodt@intel.com>
> 
> 	* elf64-x86-64.c (elf_x86_64_grok_psinfo): Check for
> 	elf_external_linux_prpsinfo32_ugid32.
> ---
>  bfd/elf64-x86-64.c | 13 ++++++++++++-
>  1 file changed, 12 insertions(+), 1 deletion(-)
> 
> diff --git a/bfd/elf64-x86-64.c b/bfd/elf64-x86-64.c
> index 914f82d0151..30789a9da8c 100644
> --- a/bfd/elf64-x86-64.c
> +++ b/bfd/elf64-x86-64.c
> @@ -389,7 +389,8 @@ elf_x86_64_grok_psinfo (bfd *abfd, Elf_Internal_Note
> *note)
>        default:
>  	return false;
> 
> -      case 124:		/* sizeof(struct elf_prpsinfo) on Linux/x32 */
> +      case 124:
> +	/* sizeof (struct elf_external_linux_prpsinfo32_ugid16).  */
>  	elf_tdata (abfd)->core->pid
>  	  = bfd_get_32 (abfd, note->descdata + 12);
>  	elf_tdata (abfd)->core->program
> @@ -398,6 +399,16 @@ elf_x86_64_grok_psinfo (bfd *abfd, Elf_Internal_Note
> *note)
>  	  = _bfd_elfcore_strndup (abfd, note->descdata + 44, 80);
>  	break;
> 
> +      case 128:
> +	/* sizeof (struct elf_external_linux_prpsinfo32_ugid32).  */
> +	elf_tdata (abfd)->core->pid
> +	  = bfd_get_32 (abfd, note->descdata + 12);
> +	elf_tdata (abfd)->core->program
> +	  = _bfd_elfcore_strndup (abfd, note->descdata + 32, 16);
> +	elf_tdata (abfd)->core->command
> +	  = _bfd_elfcore_strndup (abfd, note->descdata + 48, 80);
> +	break;
> +
>        case 136:		/* sizeof(struct elf_prpsinfo) on Linux/x86_64 */
>  	elf_tdata (abfd)->core->pid
>  	  = bfd_get_32 (abfd, note->descdata + 24);
> --
> 2.34.3

Intel Deutschland GmbH
Registered Address: Am Campeon 10, 85579 Neubiberg, Germany
Tel: +49 89 99 8853-0, www.intel.de <http://www.intel.de>
Managing Directors: Christin Eisenschmid, Sharon Heck, Tiffany Doon Silva  
Chairperson of the Supervisory Board: Nicole Lau
Registered Office: Munich
Commercial Register: Amtsgericht Muenchen HRB 186928

  reply	other threads:[~2023-02-20 12:50 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-01-24 13:42 Felix Willgerodt
2023-02-20 12:50 ` Willgerodt, Felix [this message]
2023-02-27 13:05   ` [PING] " Nick Clifton

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=MN2PR11MB456669E8FCBFA087E013CA658EA49@MN2PR11MB4566.namprd11.prod.outlook.com \
    --to=felix.willgerodt@intel.com \
    --cc=binutils@sourceware.org \
    --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;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).