| Message ID | 20240909-e810-live-migration-jk-per-queue-max-frame-rx-buf-len-v1-1-f955205bc78f@intel.com |
|---|---|
| State | Accepted |
| Delegated to: | Anthony Nguyen |
| Headers | show |
| Series | ice: store max frame and rx_buf_len only in ice_rx_ring | expand |
> -----Original Message----- > From: Intel-wired-lan <intel-wired-lan-bounces@osuosl.org> On Behalf Of Jacob > Keller > Sent: Tuesday, September 10, 2024 1:08 AM > To: Intel Wired LAN <intel-wired-lan@lists.osuosl.org>; Nguyen, Anthony L > <anthony.l.nguyen@intel.com> > Cc: Keller, Jacob E <jacob.e.keller@intel.com>; Kitszel, Przemyslaw > <przemyslaw.kitszel@intel.com> > Subject: [Intel-wired-lan] [PATCH iwl-next 1/2] ice: consistently use q_idx in > ice_vc_cfg_qs_msg() > > The ice_vc_cfg_qs_msg() function is used to configure VF queues in response to a > VIRTCHNL_OP_CONFIG_VSI_QUEUES command. > > The virtchnl command contains an array of queue pair data for configuring Tx and > Rx queues. This data includes a queue ID. When configuring the queues, the > driver generally uses this queue ID to determine which Tx and Rx ring to program. > However, a handful of places use the index into the queue pair data from the VF. > While most VF implementations appear to send this data in order, it is not > mandated by the virtchnl and it is not verified that the queue pair data comes in > order. > > Fix the driver to consistently use the q_idx field instead of the 'i' > iterator value when accessing the rings. For the Rx case, introduce a local ring > variable to keep lines short. > > Fixes: 7ad15440acf8 ("ice: Refactor VIRTCHNL_OP_CONFIG_VSI_QUEUES > handling") > Signed-off-by: Jacob Keller <jacob.e.keller@intel.com> > --- > drivers/net/ethernet/intel/ice/ice_virtchnl.c | 21 ++++++++++----------- > 1 file changed, 10 insertions(+), 11 deletions(-) > > diff --git a/drivers/net/ethernet/intel/ice/ice_virtchnl.c > b/drivers/net/ethernet/intel/ice/ice_virtchnl.c > index 59f62306b9cb..b6ec01f6fa73 100644 > --- a/drivers/net/ethernet/intel/ice/ice_virtchnl.c > +++ b/drivers/net/ethernet/intel/ice/ice_virtchnl.c > @@ -1715,8 +1715,8 @@ static int ice_vc_cfg_qs_msg(struct ice_vf *vf, u8 Tested-by: Rafal Romanowski <rafal.romanowski@intel.com>
diff --git a/drivers/net/ethernet/intel/ice/ice_virtchnl.c b/drivers/net/ethernet/intel/ice/ice_virtchnl.c index 59f62306b9cb..b6ec01f6fa73 100644 --- a/drivers/net/ethernet/intel/ice/ice_virtchnl.c +++ b/drivers/net/ethernet/intel/ice/ice_virtchnl.c @@ -1715,8 +1715,8 @@ static int ice_vc_cfg_qs_msg(struct ice_vf *vf, u8 *msg) /* copy Tx queue info from VF into VSI */ if (qpi->txq.ring_len > 0) { - vsi->tx_rings[i]->dma = qpi->txq.dma_ring_addr; - vsi->tx_rings[i]->count = qpi->txq.ring_len; + vsi->tx_rings[q_idx]->dma = qpi->txq.dma_ring_addr; + vsi->tx_rings[q_idx]->count = qpi->txq.ring_len; /* Disable any existing queue first */ if (ice_vf_vsi_dis_single_txq(vf, vsi, q_idx)) @@ -1725,7 +1725,7 @@ static int ice_vc_cfg_qs_msg(struct ice_vf *vf, u8 *msg) /* Configure a queue with the requested settings */ if (ice_vsi_cfg_single_txq(vsi, vsi->tx_rings, q_idx)) { dev_warn(ice_pf_to_dev(pf), "VF-%d failed to configure TX queue %d\n", - vf->vf_id, i); + vf->vf_id, q_idx); goto error_param; } } @@ -1733,24 +1733,23 @@ static int ice_vc_cfg_qs_msg(struct ice_vf *vf, u8 *msg) /* copy Rx queue info from VF into VSI */ if (qpi->rxq.ring_len > 0) { u16 max_frame_size = ice_vc_get_max_frame_size(vf); + struct ice_rx_ring *ring = vsi->rx_rings[q_idx]; u32 rxdid; - vsi->rx_rings[i]->dma = qpi->rxq.dma_ring_addr; - vsi->rx_rings[i]->count = qpi->rxq.ring_len; + ring->dma = qpi->rxq.dma_ring_addr; + ring->count = qpi->rxq.ring_len; if (qpi->rxq.crc_disable) - vsi->rx_rings[q_idx]->flags |= - ICE_RX_FLAGS_CRC_STRIP_DIS; + ring->flags |= ICE_RX_FLAGS_CRC_STRIP_DIS; else - vsi->rx_rings[q_idx]->flags &= - ~ICE_RX_FLAGS_CRC_STRIP_DIS; + ring->flags &= ~ICE_RX_FLAGS_CRC_STRIP_DIS; if (qpi->rxq.databuffer_size != 0 && (qpi->rxq.databuffer_size > ((16 * 1024) - 128) || qpi->rxq.databuffer_size < 1024)) goto error_param; vsi->rx_buf_len = qpi->rxq.databuffer_size; - vsi->rx_rings[i]->rx_buf_len = vsi->rx_buf_len; + ring->rx_buf_len = vsi->rx_buf_len; if (qpi->rxq.max_pkt_size > max_frame_size || qpi->rxq.max_pkt_size < 64) goto error_param; @@ -1765,7 +1764,7 @@ static int ice_vc_cfg_qs_msg(struct ice_vf *vf, u8 *msg) if (ice_vsi_cfg_single_rxq(vsi, q_idx)) { dev_warn(ice_pf_to_dev(pf), "VF-%d failed to configure RX queue %d\n", - vf->vf_id, i); + vf->vf_id, q_idx); goto error_param; }
The ice_vc_cfg_qs_msg() function is used to configure VF queues in response to a VIRTCHNL_OP_CONFIG_VSI_QUEUES command. The virtchnl command contains an array of queue pair data for configuring Tx and Rx queues. This data includes a queue ID. When configuring the queues, the driver generally uses this queue ID to determine which Tx and Rx ring to program. However, a handful of places use the index into the queue pair data from the VF. While most VF implementations appear to send this data in order, it is not mandated by the virtchnl and it is not verified that the queue pair data comes in order. Fix the driver to consistently use the q_idx field instead of the 'i' iterator value when accessing the rings. For the Rx case, introduce a local ring variable to keep lines short. Fixes: 7ad15440acf8 ("ice: Refactor VIRTCHNL_OP_CONFIG_VSI_QUEUES handling") Signed-off-by: Jacob Keller <jacob.e.keller@intel.com> --- drivers/net/ethernet/intel/ice/ice_virtchnl.c | 21 ++++++++++----------- 1 file changed, 10 insertions(+), 11 deletions(-)