linux-kernel - Re: [PATCH 04/13] objtool: Update documentation

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <Z9fhmCS68Fc8GGw_@google.com>
Date: Mon, 17 Mar 2025 08:47:20 +0000
From: Brendan Jackman <jackmanb@...gle.com>
To: Josh Poimboeuf <jpoimboe@...nel.org>
Cc: x86@...nel.org, linux-kernel@...r.kernel.org,
	Peter Zijlstra <peterz@...radead.org>,
	Nathan Chancellor <nathan@...nel.org>
Subject: Re: [PATCH 04/13] objtool: Update documentation

On Fri, Mar 14, 2025 at 12:29:02PM -0700, Josh Poimboeuf wrote:
> Fix some outdated information in the objtool doc.

Regarding the switch from ENTRY/ENDPROC to SYM_FUNC_{START,END}, would
you mind adding a bit more context to the commit message for the
benefit of ignorami like me?

BTW in tip/objtool/core include/linux/objtool.h still says:

>  Such normal callable functions
>  are annotated with the ENTRY/ENDPROC macros.

And o we also wanna update include/linux/linkage.h to mention that
these macros are deprecated?

> Signed-off-by: Josh Poimboeuf <jpoimboe@...nel.org>
> ---
>  tools/objtool/Documentation/objtool.txt | 95 ++++++++++++++-----------
>  1 file changed, 53 insertions(+), 42 deletions(-)
> 
> diff --git a/tools/objtool/Documentation/objtool.txt b/tools/objtool/Documentation/objtool.txt
> index 87950a7aaa17..28ac57b9e102 100644
> --- a/tools/objtool/Documentation/objtool.txt
> +++ b/tools/objtool/Documentation/objtool.txt
> @@ -28,6 +28,15 @@ Objtool has the following features:
>    sites, enabling the kernel to patch them inline, to prevent "thunk
>    funneling" for both security and performance reasons
>  
> +- Return thunk validation -- validates return thunks are used for
> +  certain CPU mitigations including Retbleed and SRSO
> +
> +- Return thunk annotation -- annotates all return thunk sites so kernel
> +  can patch them inline, depending on enabled mitigations
> +
> +- Return thunk training valiation -- validate that all entry paths

s/valiation/validation

Also did you mean "untraining"?

> +  untrain a "safe return" before the first return (or call)

Nit - assuming this is about the SRSO mitigation I think it's better
to just refer concretely to "Safe RET" here (with the capitalisation).