Re: [PATCH 3/7] Don't hardcode CONTEXT size for a NOTE_INFO_THREAD win32pstatus note

[Christian Biesinger <cbiesinger@google.com>]

On Wed, Jul 1, 2020 at 4:33 PM Jon Turney <jon.turney@dronecode.org.uk> wrote:
>
> Don't hardcode the size of the Win32 API thread CONTEXT type read from a
> NOTE_INFO_THREAD win32pstatus note (since it's different on different
> architectures).
>
> bfd/ChangeLog:
>
> 2020-07-01  Jon Turney  <jon.turney@dronecode.org.uk>
>
>         * elf.c (elfcore_grok_win32pstatus): Don't hardcode the size of
>         the Win32 API thread CONTEXT type read from a NOTE_INFO_THREAD
>         win32pstatus note.
> ---
>  bfd/ChangeLog | 6 ++++++
>  bfd/elf.c     | 2 +-
>  2 files changed, 7 insertions(+), 1 deletion(-)
>
> diff --git a/bfd/elf.c b/bfd/elf.c
> index a7790a4eec4..a61e2b7dd1d 100644
> --- a/bfd/elf.c
> +++ b/bfd/elf.c
> @@ -10171,7 +10171,7 @@ elfcore_grok_win32pstatus (bfd *abfd, Elf_Internal_Note *note)
>         return FALSE;
>
>        /* sizeof (thread_info.thread_context) */
> -      sect->size = 716;
> +      sect->size = note->descsz - 12;

I guess the "sizeof" comment is now wrong -- where does the "12" come
from though?

>        /* offsetof (thread_info.thread_context) */
>        sect->filepos = note->descpos + 12;
>        sect->alignment_power = 2;
> --
> 2.27.0
>