linux-kernel - Re: [PATCH] livepatch.h: Add comment to klp transition state

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] [day] [month] [year] [list]

Message-ID: <ZjiWxNLI3yS9nFI1@pathway.suse.cz>
Date: Mon, 6 May 2024 10:37:24 +0200
From: Petr Mladek <pmladek@...e.com>
To: zhang warden <zhangwarden@...il.com>
Cc: Josh Poimboeuf <jpoimboe@...nel.org>, mbenes@...e.cz, jikos@...nel.org,
	joe.lawrence@...hat.com, live-patching@...r.kernel.org,
	linux-kernel@...r.kernel.org
Subject: Re: [PATCH] livepatch.h: Add comment to klp transition state

On Mon 2024-05-06 10:04:26, zhang warden wrote:
> 
> 
> > On May 6, 2024, at 05:00, Josh Poimboeuf <jpoimboe@...nel.org> wrote:
> > 
> > On Mon, Apr 29, 2024 at 03:26:28PM +0800, zhangwarden@...il.com wrote:
> >> From: Wardenjohn <zhangwarden@...il.com>
> >> 
> >> livepatch.h use KLP_UNDEFINED\KLP_UNPATCHED\KLP_PATCHED for klp transition state.
> >> When livepatch is ready but idle, using KLP_UNDEFINED seems very confusing.
> >> In order not to introduce potential risks to kernel, just update comment
> >> to these state.
> >> ---
> >> include/linux/livepatch.h | 6 +++---
> >> 1 file changed, 3 insertions(+), 3 deletions(-)
> >> 
> >> diff --git a/include/linux/livepatch.h b/include/linux/livepatch.h
> >> index 9b9b38e89563..b6a214f2f8e3 100644
> >> --- a/include/linux/livepatch.h
> >> +++ b/include/linux/livepatch.h
> >> @@ -18,9 +18,9 @@
> >> #if IS_ENABLED(CONFIG_LIVEPATCH)
> >> 
> >> /* task patch states */
> >> -#define KLP_UNDEFINED -1
> >> -#define KLP_UNPATCHED  0
> >> -#define KLP_PATCHED  1
> >> +#define KLP_UNDEFINED -1 /* idle, no transition in progress */
> >> +#define KLP_UNPATCHED  0 /* transitioning to unpatched state */
> >> +#define KLP_PATCHED  1 /* transitioning to patched state */
> > 
> > Instead of the comments, how about we just rename them to
> > 
> >  KLP_TRANSITION_IDLE
> >  KLP_TRANSITION_UNPATCHED
> >  KLP_TRANSITION_PATCHED
> > 
> > which shouldn't break userspace AFAIK.

Great idea! It is better then nothing.

> Renaming them may be a better way as my previous patch. I would like to know why renaming KLP_*** into 
> KLP_TRANSITION_*** will not break userspace while 
> Renaming KLP_UNDEWFINED to KLP_IDLE would break the userspace.

As I already wrote in [1], both "task->patch_state == KLP_UNDEFINED"
and "KLP_IDLE" are misleading. They are not talking
about the state of the patch but about the state of the transition.

We could not rename the variables because it would break userspace.
But we could rename the state names at least.

[1] https://lore.kernel.org/r/Zg7EpZol5jB_gHH9@alley

Best Regards,
Petr