On Thu, Jan 15, 2026 at 10:38:22AM +0000, Ciara Loftus wrote:
> A value of zero in bit 4 of the generic data means that strict priority
> is enabled. Prior to this commit it was being interpreted as WFQ being
> enabled. Fix this.
>
> Fixes: ab4eaf9a8a31 ("net/ice: dump Tx scheduling tree")
> Cc: [email protected]
>
> Signed-off-by: Ciara Loftus <[email protected]>
> ---
The relevant masks and defines are in some way badly named since the field
is therefore for enabling WFQ rather than Strict Priority. Running an AI
analysis of this block returns that the code is correct because it assumes
that the SP bit being enabled means that strict-priority is enabled (where
in datasheet the SP stands for Single Priority!)
Anyway, after checking datasheet agree fix is correct.
Acked-by: Bruce Richardson <[email protected]>
> drivers/net/intel/ice/ice_diagnose.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/net/intel/ice/ice_diagnose.c
> b/drivers/net/intel/ice/ice_diagnose.c
> index 298d1eda5c..a58279a116 100644
> --- a/drivers/net/intel/ice/ice_diagnose.c
> +++ b/drivers/net/intel/ice/ice_diagnose.c
> @@ -641,7 +641,7 @@ void print_node(const struct rte_eth_dev_data *ethdata,
> fprintf(stream, "\t\t\t\t<tr>\n");
> fprintf(stream, "\t\t\t\t\t<td> priority mode</td>\n");
> fprintf(stream, "\t\t\t\t\t<td>");
> - print_priority_mode(stream, ((data->data.generic >> 4) & 0x1) != 0);
> + print_priority_mode(stream, ((data->data.generic >> 4) & 0x1) == 0);
> fprintf(stream, "</td>\n");
> fprintf(stream, "\t\t\t\t</tr>\n");
>
> --
> 2.43.0
>
