Re: [PATCH 1/1] dmioem: Update Format in HPE OEM 216

[Jean Delvare]

Hi Jerry,

On Mon, 2023-06-05 at 15:45 -0600, Jerry Hoemann wrote:
> Correct mask in HPE OEM 216 type 1.  Should be byte, not nibble.
> 
> Only display internal version number when bit is set to indicate it is
> valid.
> 
> Signed-off-by: Jerry Hoemann <jerry.hoemann@hpe.com>
> ---
>  dmioem.c | 6 ++++--
>  1 file changed, 4 insertions(+), 2 deletions(-)
> 
> diff --git a/dmioem.c b/dmioem.c
> index 186b250..ef28cce 100644
> --- a/dmioem.c
> +++ b/dmioem.c
> @@ -379,6 +379,7 @@ static void dmi_hp_216_version(u8 format, u8 *data)
>  {
>         const char * const name = "Version Data";
>         const char * const reserved = "Reserved";
> +       char buf[80] = "";
>         int gen;
>  
>         gen = dmi_hpegen(dmi_product);
> @@ -388,8 +389,9 @@ static void dmi_hp_216_version(u8 format, u8 *data)
>                 pr_attr(name, "No Version Data");
>                 break;
>         case 1:
> -               pr_attr(name, "%c.%d.%d", data[0] & (1 << 7) ? 'B' : 'R',
> -                                         data[0] & 0x7, data[1] & 0x7);
> +               if (data[0] >> 7)
> +                       sprintf(buf, " B.0x%02X", data[0] & 0x7F);
> +               pr_attr(name, "0x%02X%s", data[1] & 0x7F, buf);
>                 break;
>         case 2:
>                 pr_attr(name, "%d.%d", data[0] >> 4, data[0] & 0x0f);

As usual, I'm not a big fan of temporary buffers. What about simply
having 2 separate pr_attr() calls, one when the internal version must
be displayed, and one when not?

-- 
Jean Delvare
SUSE L3 Support