On Fri, Feb 07, 2025 at 08:55:04PM +0800, Jiayuan Chen wrote:
> The dumpstack.c file has undergone many modifications, and the
> print_context_stack() function was removed or rewritten a long time ago,
> so it's better to remove the incorrect guidance.
"print_context_stack() was removed in 2016 by commit c8fe4609827ae
("x86/dumpstack: Remove dump_trace() and related callbacks"). Remove
the now-inaccurate guide."
>
> I also want to preserve the original contributor info by keeping email
> address and name.
>
"While at it, also link to Ingo's explanatory message."
> The question about the '?' preceding function names in an x86 stacktrace
> -keeps popping up, here's an indepth explanation. It helps if the reader
> -stares at print_context_stack() and the whole machinery in and around
> -arch/x86/kernel/dumpstack.c.
> +keeps popping up. This provides guidance about it. It helps if the reader
> +stares at printk_stack_addressk() and its callers and pays special
"... Here's the explanation, that helps when the reader ..."
> +attention to the 'reliable' parameter ('?' basically means that the
> +address is unreliable).
>
> -Adapted from Ingo's mail, Message-ID: <20150521101614.GA10889@xxxxxxxxx>:
> +The detail about '?' can be found in the comments within dumpstack.c:
> +::
"The answer can be found in the comments within show_trace_log_lvl() body
in arch/x86/kernel/dumpstack.c::"
><snipped>...
> -If the address does not fit into our expected frame pointer chain we
> -still print it, but we print a '?'. It can mean two things:
> +You can also get more info from Ingo's original emal. [1]_
"For more information, see also Ingo's email. [1]_"
><snipped>...
> +.. [1] https://lore.kernel.org/lkml/20150521101614.GA10889@xxxxxxxxx/
Thanks.
--
An old man doll... just what I always wanted! - Clara
Attachment:
signature.asc
Description: PGP signature
