[v2] net/ice: optimize max queue number calculation
Checks
Commit Message
Remove the limitation that max queue pair number must be 2^n.
With this patch, even on a 8 ports device, the max queue pair
number increased from 128 to 254.
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
---
v2:
- fix check patch warning
drivers/net/ice/ice_ethdev.c | 24 ++++++++++++++++++++----
1 file changed, 20 insertions(+), 4 deletions(-)
Comments
> -----Original Message-----
> From: Zhang, Qi Z <qi.z.zhang@intel.com>
> Sent: Friday, April 8, 2022 7:24 PM
> To: Yang, Qiming <qiming.yang@intel.com>; Wu, Wenjun1
> <wenjun1.wu@intel.com>
> Cc: dev@dpdk.org; Zhang, Qi Z <qi.z.zhang@intel.com>
> Subject: [PATCH v2] net/ice: optimize max queue number calculation
>
> Remove the limitation that max queue pair number must be 2^n.
> With this patch, even on a 8 ports device, the max queue pair number
> increased from 128 to 254.
>
> Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
> ---
>
> v2:
> - fix check patch warning
>
> drivers/net/ice/ice_ethdev.c | 24 ++++++++++++++++++++----
> 1 file changed, 20 insertions(+), 4 deletions(-)
>
> diff --git a/drivers/net/ice/ice_ethdev.c b/drivers/net/ice/ice_ethdev.c index
> 73e550f5fb..ff2b3e45d9 100644
> --- a/drivers/net/ice/ice_ethdev.c
> +++ b/drivers/net/ice/ice_ethdev.c
> @@ -819,10 +819,26 @@ ice_vsi_config_tc_queue_mapping(struct ice_vsi
> *vsi,
> return -ENOTSUP;
> }
>
> - vsi->nb_qps = RTE_MIN(vsi->nb_qps, ICE_MAX_Q_PER_TC);
> - fls = (vsi->nb_qps == 0) ? 0 : rte_fls_u32(vsi->nb_qps) - 1;
> - /* Adjust the queue number to actual queues that can be applied */
> - vsi->nb_qps = (vsi->nb_qps == 0) ? 0 : 0x1 << fls;
> + /* vector 0 is reserved and 1 vector for ctrl vsi */
> + if (vsi->adapter->hw.func_caps.common_cap.num_msix_vectors < 2)
> + vsi->nb_qps = 0;
> + else
> + vsi->nb_qps = RTE_MIN
> + ((uint16_t)vsi->adapter-
> >hw.func_caps.common_cap.num_msix_vectors - 2,
> + RTE_MIN(vsi->nb_qps, ICE_MAX_Q_PER_TC));
> +
> + /* nb_qps(hex) -> fls */
> + /* 0000 -> 0 */
> + /* 0001 -> 0 */
> + /* 0002 -> 1 */
> + /* 0003 ~ 0004 -> 2 */
> + /* 0005 ~ 0008 -> 3 */
> + /* 0009 ~ 0010 -> 4 */
> + /* 0011 ~ 0020 -> 5 */
> + /* 0021 ~ 0040 -> 6 */
> + /* 0041 ~ 0080 -> 7 */
> + /* 0081 ~ 0100 -> 8 */
> + fls = (vsi->nb_qps == 0) ? 0 : rte_fls_u32(vsi->nb_qps - 1);
>
> qp_idx = 0;
> /* Set tc and queue mapping with VSI */
> --
> 2.26.2
Acked-by: Wenjun Wu < wenjun1.wu@intel.com>
Thanks
Wenjun
> -----Original Message-----
> From: Wu, Wenjun1 <wenjun1.wu@intel.com>
> Sent: Tuesday, April 26, 2022 9:14 PM
> To: Zhang, Qi Z <qi.z.zhang@intel.com>; Yang, Qiming <qiming.yang@intel.com>
> Cc: dev@dpdk.org
> Subject: RE: [PATCH v2] net/ice: optimize max queue number calculation
>
>
>
> > -----Original Message-----
> > From: Zhang, Qi Z <qi.z.zhang@intel.com>
> > Sent: Friday, April 8, 2022 7:24 PM
> > To: Yang, Qiming <qiming.yang@intel.com>; Wu, Wenjun1
> > <wenjun1.wu@intel.com>
> > Cc: dev@dpdk.org; Zhang, Qi Z <qi.z.zhang@intel.com>
> > Subject: [PATCH v2] net/ice: optimize max queue number calculation
> >
> > Remove the limitation that max queue pair number must be 2^n.
> > With this patch, even on a 8 ports device, the max queue pair number
> > increased from 128 to 254.
> >
> > Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
> > ---
> >
> > v2:
> > - fix check patch warning
> >
> > drivers/net/ice/ice_ethdev.c | 24 ++++++++++++++++++++----
> > 1 file changed, 20 insertions(+), 4 deletions(-)
> >
> > diff --git a/drivers/net/ice/ice_ethdev.c
> > b/drivers/net/ice/ice_ethdev.c index
> > 73e550f5fb..ff2b3e45d9 100644
> > --- a/drivers/net/ice/ice_ethdev.c
> > +++ b/drivers/net/ice/ice_ethdev.c
> > @@ -819,10 +819,26 @@ ice_vsi_config_tc_queue_mapping(struct ice_vsi
> > *vsi,
> > return -ENOTSUP;
> > }
> >
> > - vsi->nb_qps = RTE_MIN(vsi->nb_qps, ICE_MAX_Q_PER_TC);
> > - fls = (vsi->nb_qps == 0) ? 0 : rte_fls_u32(vsi->nb_qps) - 1;
> > - /* Adjust the queue number to actual queues that can be applied */
> > - vsi->nb_qps = (vsi->nb_qps == 0) ? 0 : 0x1 << fls;
> > + /* vector 0 is reserved and 1 vector for ctrl vsi */
> > + if (vsi->adapter->hw.func_caps.common_cap.num_msix_vectors < 2)
> > + vsi->nb_qps = 0;
> > + else
> > + vsi->nb_qps = RTE_MIN
> > + ((uint16_t)vsi->adapter-
> > >hw.func_caps.common_cap.num_msix_vectors - 2,
> > + RTE_MIN(vsi->nb_qps, ICE_MAX_Q_PER_TC));
> > +
> > + /* nb_qps(hex) -> fls */
> > + /* 0000 -> 0 */
> > + /* 0001 -> 0 */
> > + /* 0002 -> 1 */
> > + /* 0003 ~ 0004 -> 2 */
> > + /* 0005 ~ 0008 -> 3 */
> > + /* 0009 ~ 0010 -> 4 */
> > + /* 0011 ~ 0020 -> 5 */
> > + /* 0021 ~ 0040 -> 6 */
> > + /* 0041 ~ 0080 -> 7 */
> > + /* 0081 ~ 0100 -> 8 */
> > + fls = (vsi->nb_qps == 0) ? 0 : rte_fls_u32(vsi->nb_qps - 1);
> >
> > qp_idx = 0;
> > /* Set tc and queue mapping with VSI */
> > --
> > 2.26.2
>
> Acked-by: Wenjun Wu < wenjun1.wu@intel.com>
>
> Thanks
> Wenjun
>
Applied to dpdk-next-net-intel.
Thanks
Qi
@@ -819,10 +819,26 @@ ice_vsi_config_tc_queue_mapping(struct ice_vsi *vsi,
return -ENOTSUP;
}
- vsi->nb_qps = RTE_MIN(vsi->nb_qps, ICE_MAX_Q_PER_TC);
- fls = (vsi->nb_qps == 0) ? 0 : rte_fls_u32(vsi->nb_qps) - 1;
- /* Adjust the queue number to actual queues that can be applied */
- vsi->nb_qps = (vsi->nb_qps == 0) ? 0 : 0x1 << fls;
+ /* vector 0 is reserved and 1 vector for ctrl vsi */
+ if (vsi->adapter->hw.func_caps.common_cap.num_msix_vectors < 2)
+ vsi->nb_qps = 0;
+ else
+ vsi->nb_qps = RTE_MIN
+ ((uint16_t)vsi->adapter->hw.func_caps.common_cap.num_msix_vectors - 2,
+ RTE_MIN(vsi->nb_qps, ICE_MAX_Q_PER_TC));
+
+ /* nb_qps(hex) -> fls */
+ /* 0000 -> 0 */
+ /* 0001 -> 0 */
+ /* 0002 -> 1 */
+ /* 0003 ~ 0004 -> 2 */
+ /* 0005 ~ 0008 -> 3 */
+ /* 0009 ~ 0010 -> 4 */
+ /* 0011 ~ 0020 -> 5 */
+ /* 0021 ~ 0040 -> 6 */
+ /* 0041 ~ 0080 -> 7 */
+ /* 0081 ~ 0100 -> 8 */
+ fls = (vsi->nb_qps == 0) ? 0 : rte_fls_u32(vsi->nb_qps - 1);
qp_idx = 0;
/* Set tc and queue mapping with VSI */