Re: [PATCH] platform/x86: dell-wmi-sysman: Clean up security buffer helpers

["Ilpo Järvinen" <ilpo.jarvinen@linux.intel.com>]

[-- Attachment #1: Type: text/plain, Size: 2009 bytes --]

On Tue, 31 Mar 2026, Thorsten Blum wrote:

> On Tue, Mar 31, 2026 at 03:46:11PM +0300, Ilpo Järvinen wrote:
> > On Mon, 9 Mar 2026, Thorsten Blum wrote:
> > > In calculate_security_buffer(), call strlen() once and use ALIGN() to
> > > round up to an even size.
> > > 
> > > In populate_security_buffer(), also avoid recomputing strlen(), rename
> > > the u32 pointer from 'seclen' to 'seclenp' to avoid confusion with the
> > > new length variable, and drop the memcpy() guard since calling it with
> > > size 0 is a no-op and therefore safe.
> > > 
> > > Use 'const char *' for the read-only source string in both helpers.
> > > 
> > > Signed-off-by: Thorsten Blum <thorsten.blum@linux.dev>
> > > ---
> > >  .../dell/dell-wmi-sysman/dell-wmi-sysman.h    |  4 ++--
> > >  .../x86/dell/dell-wmi-sysman/sysman.c         | 20 ++++++++-----------
> > >  2 files changed, 10 insertions(+), 14 deletions(-)
> > > 
> > > [...]
> > 
> > Logicwise this change seem fine to me but I wonder why is buffer char *
> > (again, this is a pre-existing problem). It seems u8 * would be more 
> > appropriate given how it's handled here as something else than characters.
> 
> Yes, I agree. However, using u8 * would also require changing the call
> sites in biosattr-interface.c and passwordattr-interface.c. Happy to
> change it, but it would probably be better as a follow-up patch.

A follow-up patch is fine.

> > Also, you don't need resend if your patch is still listed in the 
> > patchwork, it's not lost as long as it remains in patchwork and I'll 
> > eventually get to it.
> 
> I wasn't aware of the patchwork instance - is it this one?
> 
> https://patchwork.kernel.org/project/platform-driver-x86/list/

Correct.

You can find that URL from MAINTAINERS.

> I see 67 patches in total, but this one is not part of the list, afaict.

It was there until today but I marked it changes required after noting it 
lacks the header so I'm expecting a new version.

-- 
 i.