[Intel-wired-lan] [PATCH iwl-net 00/11] ice: fix AF_XDP ZC timeout and concurrency issues

[Intel-wired-lan] [PATCH iwl-net 00/11] ice: fix AF_XDP ZC timeout and concurrency issues

[Maciej Fijalkowski]

Hi,

changes included in this patchset address an issue that customer has
been facing when AF_XDP ZC Tx sockets were used in combination with flow
control and regular Tx traffic.

After executing:
ethtool --set-priv-flags $dev link-down-on-close on
ethtool -A $dev rx on tx on

launching multiple ZC Tx sockets on $dev + pinging remote interface (so
that regular Tx traffic is present) and then going through down/up of
$dev, Tx timeout occured and then most of the time ice driver was unable
to recover from that state.

Patches 1-8 combined together solve the described above issue on
customer side. Main focus here is to forbid producing Tx descriptors
when either carrier is not yet initialized or process of bringing
interface down has already started.

On top of that, Larysa goes further with fixing XSK pool setup issues
and race condition when Tx timeout strikes at the same time as
ndo_bpf().

Thanks,
Maciej


Larysa Zaremba (3):
  ice: move locking outside of ice_qp_ena and ice_qp_dis
  ice: lock with PF state instead of VSI state in ice_xsk_pool_setup()
  ice: protect ring configuration with a mutex

Maciej Fijalkowski (7):
  ice: don't busy wait for Rx queue disable in ice_qp_dis()
  ice: replace synchronize_rcu with synchronize_net
  ice: modify error handling when setting XSK pool in ndo_bpf
  ice: toggle netif_carrier when setting up XSK pool
  ice: improve updating ice_{t,r}x_ring::xsk_pool
  ice: add missing WRITE_ONCE when clearing ice_rx_ring::xdp_prog
  ice: xsk: fix txq interrupt mapping

Michal Kubiak (1):
  ice: respect netif readiness in AF_XDP ZC related ndo's

 drivers/net/ethernet/intel/ice/ice.h      |   8 +-
 drivers/net/ethernet/intel/ice/ice_base.c |   4 +-
 drivers/net/ethernet/intel/ice/ice_lib.c  |  23 +++-
 drivers/net/ethernet/intel/ice/ice_main.c |  41 +++++-
 drivers/net/ethernet/intel/ice/ice_txrx.c |   6 +-
 drivers/net/ethernet/intel/ice/ice_xsk.c  | 154 ++++++++++++----------
 drivers/net/ethernet/intel/ice/ice_xsk.h  |   4 +-
 7 files changed, 151 insertions(+), 89 deletions(-)

-- 
2.34.1

[Intel-wired-lan] [PATCH iwl-net 01/11] ice: respect netif readiness in AF_XDP ZC related ndo's

[Maciej Fijalkowski]

From: Michal Kubiak <michal.kubiak@intel.com>

Address a scenario in which XSK ZC Tx produces descriptors to XDP Tx
ring when link is either not yet fully initialized or process of
stopping the netdev has already started. To avoid this, add checks
against carrier readiness in ice_xsk_wakeup() and in ice_xmit_zc().
One could argue that bailing out early in ice_xsk_wakeup() would be
sufficient but given the fact that we produce Tx descriptors on behalf
of NAPI that is triggered for Rx traffic, the latter is also needed.

Bringing link up is an asynchronous event executed within
ice_service_task so even though interface has been brought up there is
still a time frame where link is not yet ok.

Without this patch, when AF_XDP ZC Tx is used simultaneously with stack
Tx, Tx timeouts occur after going through link flap (admin brings
interface down then up again). HW seem to be unable to transmit
descriptor to the wire after HW tail register bump which in turn causes
bit __QUEUE_STATE_STACK_XOFF to be set forever as
netdev_tx_completed_queue() sees no cleaned bytes on the input.

Fixes: 126cdfe1007a ("ice: xsk: Improve AF_XDP ZC Tx and use batching API")
Fixes: 2d4238f55697 ("ice: Add support for AF_XDP")
Signed-off-by: Michal Kubiak <michal.kubiak@intel.com>
Signed-off-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
---
 drivers/net/ethernet/intel/ice/ice_xsk.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)

diff --git a/drivers/net/ethernet/intel/ice/ice_xsk.c b/drivers/net/ethernet/intel/ice/ice_xsk.c
index 2015f66b0cf9..1bd4b054dd80 100644
--- a/drivers/net/ethernet/intel/ice/ice_xsk.c
+++ b/drivers/net/ethernet/intel/ice/ice_xsk.c
@@ -1048,6 +1048,10 @@ bool ice_xmit_zc(struct ice_tx_ring *xdp_ring)
 
 	ice_clean_xdp_irq_zc(xdp_ring);
 
+	if (!netif_carrier_ok(xdp_ring->vsi->netdev) ||
+	    !netif_running(xdp_ring->vsi->netdev))
+		return true;
+
 	budget = ICE_DESC_UNUSED(xdp_ring);
 	budget = min_t(u16, budget, ICE_RING_QUARTER(xdp_ring));
 
@@ -1091,7 +1095,7 @@ ice_xsk_wakeup(struct net_device *netdev, u32 queue_id,
 	struct ice_vsi *vsi = np->vsi;
 	struct ice_tx_ring *ring;
 
-	if (test_bit(ICE_VSI_DOWN, vsi->state))
+	if (test_bit(ICE_VSI_DOWN, vsi->state) || !netif_carrier_ok(netdev))
 		return -ENETDOWN;
 
 	if (!ice_is_xdp_ena_vsi(vsi))
-- 
2.34.1

[Intel-wired-lan] [PATCH iwl-net 02/11] ice: don't busy wait for Rx queue disable in ice_qp_dis()

[Maciej Fijalkowski]

When ice driver is spammed with multiple xdpsock instances and flow
control is enabled, there are cases when Rx queue gets stuck and unable
to reflect the disable state in QRX_CTRL register. Similar issue has
previously been addressed in commit 13a6233b033f ("ice: Add support to
enable/disable all Rx queues before waiting").

To workaround this, let us simply not wait for a disabled state as later
patch will make sure that regardless of the encountered error in the
process of disabling a queue pair, the Rx queue will be enabled.

Fixes: 2d4238f55697 ("ice: Add support for AF_XDP")
Signed-off-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
---
 drivers/net/ethernet/intel/ice/ice_xsk.c | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

diff --git a/drivers/net/ethernet/intel/ice/ice_xsk.c b/drivers/net/ethernet/intel/ice/ice_xsk.c
index 1bd4b054dd80..4f606a1055b0 100644
--- a/drivers/net/ethernet/intel/ice/ice_xsk.c
+++ b/drivers/net/ethernet/intel/ice/ice_xsk.c
@@ -199,10 +199,8 @@ static int ice_qp_dis(struct ice_vsi *vsi, u16 q_idx)
 		if (err)
 			return err;
 	}
-	err = ice_vsi_ctrl_one_rx_ring(vsi, false, q_idx, true);
-	if (err)
-		return err;
 
+	ice_vsi_ctrl_one_rx_ring(vsi, false, q_idx, false);
 	ice_qp_clean_rings(vsi, q_idx);
 	ice_qp_reset_stats(vsi, q_idx);
 
-- 
2.34.1

[Intel-wired-lan] [PATCH iwl-net 03/11] ice: replace synchronize_rcu with synchronize_net

[Maciej Fijalkowski]

Given that ice_qp_dis() is called under rtnl_lock, synchronize_net() can
be called instead of synchronize_rcu() so that XDP rings can finish its
job in a faster way. Also let us do this as earlier in XSK queue disable
flow.

Additionally, turn off regular Tx queue before disabling irqs and NAPI.

Fixes: 2d4238f55697 ("ice: Add support for AF_XDP")
Signed-off-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
---
 drivers/net/ethernet/intel/ice/ice_xsk.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/drivers/net/ethernet/intel/ice/ice_xsk.c b/drivers/net/ethernet/intel/ice/ice_xsk.c
index 4f606a1055b0..e93cb0ca4106 100644
--- a/drivers/net/ethernet/intel/ice/ice_xsk.c
+++ b/drivers/net/ethernet/intel/ice/ice_xsk.c
@@ -53,7 +53,6 @@ static void ice_qp_clean_rings(struct ice_vsi *vsi, u16 q_idx)
 {
 	ice_clean_tx_ring(vsi->tx_rings[q_idx]);
 	if (ice_is_xdp_ena_vsi(vsi)) {
-		synchronize_rcu();
 		ice_clean_tx_ring(vsi->xdp_rings[q_idx]);
 	}
 	ice_clean_rx_ring(vsi->rx_rings[q_idx]);
@@ -180,11 +179,12 @@ static int ice_qp_dis(struct ice_vsi *vsi, u16 q_idx)
 		usleep_range(1000, 2000);
 	}
 
+	synchronize_net();
+	netif_tx_stop_queue(netdev_get_tx_queue(vsi->netdev, q_idx));
+
 	ice_qvec_dis_irq(vsi, rx_ring, q_vector);
 	ice_qvec_toggle_napi(vsi, q_vector, false);
 
-	netif_tx_stop_queue(netdev_get_tx_queue(vsi->netdev, q_idx));
-
 	ice_fill_txq_meta(vsi, tx_ring, &txq_meta);
 	err = ice_vsi_stop_tx_ring(vsi, ICE_NO_RESET, 0, tx_ring, &txq_meta);
 	if (err)
-- 
2.34.1

[Intel-wired-lan] [PATCH iwl-net 04/11] ice: modify error handling when setting XSK pool in ndo_bpf

[Maciej Fijalkowski]

Don't bail out right when spotting an error within ice_qp_{dis,ena}()
but rather track error and go through whole flow of disabling and
enabling queue pair.

Fixes: 2d4238f55697 ("ice: Add support for AF_XDP")
Signed-off-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
---
 drivers/net/ethernet/intel/ice/ice_xsk.c | 30 +++++++++++++-----------
 1 file changed, 16 insertions(+), 14 deletions(-)

diff --git a/drivers/net/ethernet/intel/ice/ice_xsk.c b/drivers/net/ethernet/intel/ice/ice_xsk.c
index e93cb0ca4106..3dcab89be256 100644
--- a/drivers/net/ethernet/intel/ice/ice_xsk.c
+++ b/drivers/net/ethernet/intel/ice/ice_xsk.c
@@ -163,6 +163,7 @@ static int ice_qp_dis(struct ice_vsi *vsi, u16 q_idx)
 	struct ice_tx_ring *tx_ring;
 	struct ice_rx_ring *rx_ring;
 	int timeout = 50;
+	int fail = 0;
 	int err;
 
 	if (q_idx >= vsi->num_rxq || q_idx >= vsi->num_txq)
@@ -187,8 +188,8 @@ static int ice_qp_dis(struct ice_vsi *vsi, u16 q_idx)
 
 	ice_fill_txq_meta(vsi, tx_ring, &txq_meta);
 	err = ice_vsi_stop_tx_ring(vsi, ICE_NO_RESET, 0, tx_ring, &txq_meta);
-	if (err)
-		return err;
+	if (!fail)
+		fail = err;
 	if (ice_is_xdp_ena_vsi(vsi)) {
 		struct ice_tx_ring *xdp_ring = vsi->xdp_rings[q_idx];
 
@@ -196,15 +197,15 @@ static int ice_qp_dis(struct ice_vsi *vsi, u16 q_idx)
 		ice_fill_txq_meta(vsi, xdp_ring, &txq_meta);
 		err = ice_vsi_stop_tx_ring(vsi, ICE_NO_RESET, 0, xdp_ring,
 					   &txq_meta);
-		if (err)
-			return err;
+		if (!fail)
+			fail = err;
 	}
 
 	ice_vsi_ctrl_one_rx_ring(vsi, false, q_idx, false);
 	ice_qp_clean_rings(vsi, q_idx);
 	ice_qp_reset_stats(vsi, q_idx);
 
-	return 0;
+	return fail;
 }
 
 /**
@@ -217,32 +218,33 @@ static int ice_qp_dis(struct ice_vsi *vsi, u16 q_idx)
 static int ice_qp_ena(struct ice_vsi *vsi, u16 q_idx)
 {
 	struct ice_q_vector *q_vector;
+	int fail = 0;
 	int err;
 
 	err = ice_vsi_cfg_single_txq(vsi, vsi->tx_rings, q_idx);
-	if (err)
-		return err;
+	if (!fail)
+		fail = err;
 
 	if (ice_is_xdp_ena_vsi(vsi)) {
 		struct ice_tx_ring *xdp_ring = vsi->xdp_rings[q_idx];
 
 		err = ice_vsi_cfg_single_txq(vsi, vsi->xdp_rings, q_idx);
-		if (err)
-			return err;
+		if (!fail)
+			fail = err;
 		ice_set_ring_xdp(xdp_ring);
 		ice_tx_xsk_pool(vsi, q_idx);
 	}
 
 	err = ice_vsi_cfg_single_rxq(vsi, q_idx);
-	if (err)
-		return err;
+	if (!fail)
+		fail = err;
 
 	q_vector = vsi->rx_rings[q_idx]->q_vector;
 	ice_qvec_cfg_msix(vsi, q_vector);
 
 	err = ice_vsi_ctrl_one_rx_ring(vsi, true, q_idx, true);
-	if (err)
-		return err;
+	if (!fail)
+		fail = err;
 
 	ice_qvec_toggle_napi(vsi, q_vector, true);
 	ice_qvec_ena_irq(vsi, q_vector);
@@ -250,7 +252,7 @@ static int ice_qp_ena(struct ice_vsi *vsi, u16 q_idx)
 	netif_tx_start_queue(netdev_get_tx_queue(vsi->netdev, q_idx));
 	clear_bit(ICE_CFG_BUSY, vsi->state);
 
-	return 0;
+	return fail;
 }
 
 /**
-- 
2.34.1

[Intel-wired-lan] [PATCH iwl-net 05/11] ice: toggle netif_carrier when setting up XSK pool

[Maciej Fijalkowski]

In order to prevent Tx timeout issues, toggle the netif_carrier when
disabling and enabling queue pair during XSK pool setup. One of
conditions checked on running in the background dev_watchdog() is
netif_carrier_ok(), so let us turn it off when we disable the queues
that belong to a q_vector where XSK pool is being configured.

Fixes: 2d4238f55697 ("ice: Add support for AF_XDP")
Signed-off-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
---
 drivers/net/ethernet/intel/ice/ice_xsk.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/drivers/net/ethernet/intel/ice/ice_xsk.c b/drivers/net/ethernet/intel/ice/ice_xsk.c
index 3dcab89be256..8c5006f37310 100644
--- a/drivers/net/ethernet/intel/ice/ice_xsk.c
+++ b/drivers/net/ethernet/intel/ice/ice_xsk.c
@@ -181,6 +181,7 @@ static int ice_qp_dis(struct ice_vsi *vsi, u16 q_idx)
 	}
 
 	synchronize_net();
+	netif_carrier_off(vsi->netdev);
 	netif_tx_stop_queue(netdev_get_tx_queue(vsi->netdev, q_idx));
 
 	ice_qvec_dis_irq(vsi, rx_ring, q_vector);
@@ -250,6 +251,7 @@ static int ice_qp_ena(struct ice_vsi *vsi, u16 q_idx)
 	ice_qvec_ena_irq(vsi, q_vector);
 
 	netif_tx_start_queue(netdev_get_tx_queue(vsi->netdev, q_idx));
+	netif_carrier_on(vsi->netdev);
 	clear_bit(ICE_CFG_BUSY, vsi->state);
 
 	return fail;
-- 
2.34.1

[Intel-wired-lan] [PATCH iwl-net 06/11] ice: improve updating ice_{t, r}x_ring::xsk_pool

[Maciej Fijalkowski]

xsk_buff_pool pointers that ice ring structs hold are updated via
ndo_bpf that is executed in process context while it can be read by
remote CPU at the same time within NAPI poll. Use synchronize_net()
after pointer update and {READ,WRITE}_ONCE() when working with mentioned
pointer.

Fixes: 2d4238f55697 ("ice: Add support for AF_XDP")
Signed-off-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
---
 drivers/net/ethernet/intel/ice/ice.h      |  6 +-
 drivers/net/ethernet/intel/ice/ice_base.c |  4 +-
 drivers/net/ethernet/intel/ice/ice_main.c |  2 +-
 drivers/net/ethernet/intel/ice/ice_txrx.c |  4 +-
 drivers/net/ethernet/intel/ice/ice_xsk.c  | 73 +++++++++++++----------
 drivers/net/ethernet/intel/ice/ice_xsk.h  |  4 +-
 6 files changed, 54 insertions(+), 39 deletions(-)

diff --git a/drivers/net/ethernet/intel/ice/ice.h b/drivers/net/ethernet/intel/ice/ice.h
index da8c8afebc93..701a61d791dd 100644
--- a/drivers/net/ethernet/intel/ice/ice.h
+++ b/drivers/net/ethernet/intel/ice/ice.h
@@ -771,12 +771,12 @@ static inline struct xsk_buff_pool *ice_get_xp_from_qid(struct ice_vsi *vsi,
  * Returns a pointer to xsk_buff_pool structure if there is a buffer pool
  * present, NULL otherwise.
  */
-static inline struct xsk_buff_pool *ice_xsk_pool(struct ice_rx_ring *ring)
+static inline void ice_xsk_pool(struct ice_rx_ring *ring)
 {
 	struct ice_vsi *vsi = ring->vsi;
 	u16 qid = ring->q_index;
 
-	return ice_get_xp_from_qid(vsi, qid);
+	WRITE_ONCE(ring->xsk_pool, ice_get_xp_from_qid(vsi, qid));
 }
 
 /**
@@ -801,7 +801,7 @@ static inline void ice_tx_xsk_pool(struct ice_vsi *vsi, u16 qid)
 	if (!ring)
 		return;
 
-	ring->xsk_pool = ice_get_xp_from_qid(vsi, qid);
+	WRITE_ONCE(ring->xsk_pool, ice_get_xp_from_qid(vsi, qid));
 }
 
 /**
diff --git a/drivers/net/ethernet/intel/ice/ice_base.c b/drivers/net/ethernet/intel/ice/ice_base.c
index 5d396c1a7731..f3dfce136106 100644
--- a/drivers/net/ethernet/intel/ice/ice_base.c
+++ b/drivers/net/ethernet/intel/ice/ice_base.c
@@ -536,7 +536,7 @@ static int ice_vsi_cfg_rxq(struct ice_rx_ring *ring)
 				return err;
 		}
 
-		ring->xsk_pool = ice_xsk_pool(ring);
+		ice_xsk_pool(ring);
 		if (ring->xsk_pool) {
 			xdp_rxq_info_unreg(&ring->xdp_rxq);
 
@@ -597,7 +597,7 @@ static int ice_vsi_cfg_rxq(struct ice_rx_ring *ring)
 			return 0;
 		}
 
-		ok = ice_alloc_rx_bufs_zc(ring, num_bufs);
+		ok = ice_alloc_rx_bufs_zc(ring, ring->xsk_pool, num_bufs);
 		if (!ok) {
 			u16 pf_q = ring->vsi->rxq_map[ring->q_index];
 
diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c
index 1b61ca3a6eb6..15a6805ac2a1 100644
--- a/drivers/net/ethernet/intel/ice/ice_main.c
+++ b/drivers/net/ethernet/intel/ice/ice_main.c
@@ -2946,7 +2946,7 @@ static void ice_vsi_rx_napi_schedule(struct ice_vsi *vsi)
 	ice_for_each_rxq(vsi, i) {
 		struct ice_rx_ring *rx_ring = vsi->rx_rings[i];
 
-		if (rx_ring->xsk_pool)
+		if (READ_ONCE(rx_ring->xsk_pool))
 			napi_schedule(&rx_ring->q_vector->napi);
 	}
 }
diff --git a/drivers/net/ethernet/intel/ice/ice_txrx.c b/drivers/net/ethernet/intel/ice/ice_txrx.c
index 8bb743f78fcb..f4b2b1bca234 100644
--- a/drivers/net/ethernet/intel/ice/ice_txrx.c
+++ b/drivers/net/ethernet/intel/ice/ice_txrx.c
@@ -1523,7 +1523,7 @@ int ice_napi_poll(struct napi_struct *napi, int budget)
 	ice_for_each_tx_ring(tx_ring, q_vector->tx) {
 		bool wd;
 
-		if (tx_ring->xsk_pool)
+		if (READ_ONCE(tx_ring->xsk_pool))
 			wd = ice_xmit_zc(tx_ring);
 		else if (ice_ring_is_xdp(tx_ring))
 			wd = true;
@@ -1556,7 +1556,7 @@ int ice_napi_poll(struct napi_struct *napi, int budget)
 		 * comparison in the irq context instead of many inside the
 		 * ice_clean_rx_irq function and makes the codebase cleaner.
 		 */
-		cleaned = rx_ring->xsk_pool ?
+		cleaned = READ_ONCE(rx_ring->xsk_pool) ?
 			  ice_clean_rx_irq_zc(rx_ring, budget_per_ring) :
 			  ice_clean_rx_irq(rx_ring, budget_per_ring);
 		work_done += cleaned;
diff --git a/drivers/net/ethernet/intel/ice/ice_xsk.c b/drivers/net/ethernet/intel/ice/ice_xsk.c
index 8c5006f37310..e554cf424fb3 100644
--- a/drivers/net/ethernet/intel/ice/ice_xsk.c
+++ b/drivers/net/ethernet/intel/ice/ice_xsk.c
@@ -250,6 +250,8 @@ static int ice_qp_ena(struct ice_vsi *vsi, u16 q_idx)
 	ice_qvec_toggle_napi(vsi, q_vector, true);
 	ice_qvec_ena_irq(vsi, q_vector);
 
+	/* make sure NAPI sees updated ice_{t,x}_ring::xsk_pool */
+	synchronize_net();
 	netif_tx_start_queue(netdev_get_tx_queue(vsi->netdev, q_idx));
 	netif_carrier_on(vsi->netdev);
 	clear_bit(ICE_CFG_BUSY, vsi->state);
@@ -469,7 +471,8 @@ static u16 ice_fill_rx_descs(struct xsk_buff_pool *pool, struct xdp_buff **xdp,
  *
  * Returns true if all allocations were successful, false if any fail.
  */
-static bool __ice_alloc_rx_bufs_zc(struct ice_rx_ring *rx_ring, u16 count)
+static bool __ice_alloc_rx_bufs_zc(struct ice_rx_ring *rx_ring,
+				   struct xsk_buff_pool *xsk_pool,  u16 count)
 {
 	u32 nb_buffs_extra = 0, nb_buffs = 0;
 	union ice_32b_rx_flex_desc *rx_desc;
@@ -481,8 +484,7 @@ static bool __ice_alloc_rx_bufs_zc(struct ice_rx_ring *rx_ring, u16 count)
 	xdp = ice_xdp_buf(rx_ring, ntu);
 
 	if (ntu + count >= rx_ring->count) {
-		nb_buffs_extra = ice_fill_rx_descs(rx_ring->xsk_pool, xdp,
-						   rx_desc,
+		nb_buffs_extra = ice_fill_rx_descs(xsk_pool, xdp, rx_desc,
 						   rx_ring->count - ntu);
 		if (nb_buffs_extra != rx_ring->count - ntu) {
 			ntu += nb_buffs_extra;
@@ -495,7 +497,7 @@ static bool __ice_alloc_rx_bufs_zc(struct ice_rx_ring *rx_ring, u16 count)
 		ice_release_rx_desc(rx_ring, 0);
 	}
 
-	nb_buffs = ice_fill_rx_descs(rx_ring->xsk_pool, xdp, rx_desc, count);
+	nb_buffs = ice_fill_rx_descs(xsk_pool, xdp, rx_desc, count);
 
 	ntu += nb_buffs;
 	if (ntu == rx_ring->count)
@@ -518,7 +520,8 @@ static bool __ice_alloc_rx_bufs_zc(struct ice_rx_ring *rx_ring, u16 count)
  *
  * Returns true if all calls to internal alloc routine succeeded
  */
-bool ice_alloc_rx_bufs_zc(struct ice_rx_ring *rx_ring, u16 count)
+bool ice_alloc_rx_bufs_zc(struct ice_rx_ring *rx_ring,
+			  struct xsk_buff_pool *xsk_pool, u16 count)
 {
 	u16 rx_thresh = ICE_RING_QUARTER(rx_ring);
 	u16 leftover, i, tail_bumps;
@@ -527,9 +530,9 @@ bool ice_alloc_rx_bufs_zc(struct ice_rx_ring *rx_ring, u16 count)
 	leftover = count - (tail_bumps * rx_thresh);
 
 	for (i = 0; i < tail_bumps; i++)
-		if (!__ice_alloc_rx_bufs_zc(rx_ring, rx_thresh))
+		if (!__ice_alloc_rx_bufs_zc(rx_ring, xsk_pool, rx_thresh))
 			return false;
-	return __ice_alloc_rx_bufs_zc(rx_ring, leftover);
+	return __ice_alloc_rx_bufs_zc(rx_ring, xsk_pool, leftover);
 }
 
 /**
@@ -650,7 +653,7 @@ static u32 ice_clean_xdp_irq_zc(struct ice_tx_ring *xdp_ring)
 	if (xdp_ring->next_to_clean >= cnt)
 		xdp_ring->next_to_clean -= cnt;
 	if (xsk_frames)
-		xsk_tx_completed(xdp_ring->xsk_pool, xsk_frames);
+		xsk_tx_completed(READ_ONCE(xdp_ring->xsk_pool), xsk_frames);
 
 	return completed_frames;
 }
@@ -702,7 +705,8 @@ static int ice_xmit_xdp_tx_zc(struct xdp_buff *xdp,
 		dma_addr_t dma;
 
 		dma = xsk_buff_xdp_get_dma(xdp);
-		xsk_buff_raw_dma_sync_for_device(xdp_ring->xsk_pool, dma, size);
+		xsk_buff_raw_dma_sync_for_device(READ_ONCE(xdp_ring->xsk_pool),
+						 dma, size);
 
 		tx_buf->xdp = xdp;
 		tx_buf->type = ICE_TX_BUF_XSK_TX;
@@ -760,7 +764,8 @@ ice_run_xdp_zc(struct ice_rx_ring *rx_ring, struct xdp_buff *xdp,
 		err = xdp_do_redirect(rx_ring->netdev, xdp, xdp_prog);
 		if (!err)
 			return ICE_XDP_REDIR;
-		if (xsk_uses_need_wakeup(rx_ring->xsk_pool) && err == -ENOBUFS)
+		if (xsk_uses_need_wakeup(READ_ONCE(rx_ring->xsk_pool)) &&
+		    err == -ENOBUFS)
 			result = ICE_XDP_EXIT;
 		else
 			result = ICE_XDP_CONSUMED;
@@ -829,8 +834,8 @@ ice_add_xsk_frag(struct ice_rx_ring *rx_ring, struct xdp_buff *first,
  */
 int ice_clean_rx_irq_zc(struct ice_rx_ring *rx_ring, int budget)
 {
+	struct xsk_buff_pool *xsk_pool = READ_ONCE(rx_ring->xsk_pool);
 	unsigned int total_rx_bytes = 0, total_rx_packets = 0;
-	struct xsk_buff_pool *xsk_pool = rx_ring->xsk_pool;
 	u32 ntc = rx_ring->next_to_clean;
 	u32 ntu = rx_ring->next_to_use;
 	struct xdp_buff *first = NULL;
@@ -942,7 +947,8 @@ int ice_clean_rx_irq_zc(struct ice_rx_ring *rx_ring, int budget)
 	rx_ring->next_to_clean = ntc;
 	entries_to_alloc = ICE_RX_DESC_UNUSED(rx_ring);
 	if (entries_to_alloc > ICE_RING_QUARTER(rx_ring))
-		failure |= !ice_alloc_rx_bufs_zc(rx_ring, entries_to_alloc);
+		failure |= !ice_alloc_rx_bufs_zc(rx_ring, xsk_pool,
+						 entries_to_alloc);
 
 	ice_finalize_xdp_rx(xdp_ring, xdp_xmit, 0);
 	ice_update_rx_ring_stats(rx_ring, total_rx_packets, total_rx_bytes);
@@ -968,14 +974,15 @@ int ice_clean_rx_irq_zc(struct ice_rx_ring *rx_ring, int budget)
  * @desc: AF_XDP descriptor to pull the DMA address and length from
  * @total_bytes: bytes accumulator that will be used for stats update
  */
-static void ice_xmit_pkt(struct ice_tx_ring *xdp_ring, struct xdp_desc *desc,
+static void ice_xmit_pkt(struct ice_tx_ring *xdp_ring,
+			 struct xsk_buff_pool *xsk_pool, struct xdp_desc *desc,
 			 unsigned int *total_bytes)
 {
 	struct ice_tx_desc *tx_desc;
 	dma_addr_t dma;
 
-	dma = xsk_buff_raw_get_dma(xdp_ring->xsk_pool, desc->addr);
-	xsk_buff_raw_dma_sync_for_device(xdp_ring->xsk_pool, dma, desc->len);
+	dma = xsk_buff_raw_get_dma(xsk_pool, desc->addr);
+	xsk_buff_raw_dma_sync_for_device(xsk_pool, dma, desc->len);
 
 	tx_desc = ICE_TX_DESC(xdp_ring, xdp_ring->next_to_use++);
 	tx_desc->buf_addr = cpu_to_le64(dma);
@@ -991,7 +998,9 @@ static void ice_xmit_pkt(struct ice_tx_ring *xdp_ring, struct xdp_desc *desc,
  * @descs: AF_XDP descriptors to pull the DMA addresses and lengths from
  * @total_bytes: bytes accumulator that will be used for stats update
  */
-static void ice_xmit_pkt_batch(struct ice_tx_ring *xdp_ring, struct xdp_desc *descs,
+static void ice_xmit_pkt_batch(struct ice_tx_ring *xdp_ring,
+			       struct xsk_buff_pool *xsk_pool,
+			       struct xdp_desc *descs,
 			       unsigned int *total_bytes)
 {
 	u16 ntu = xdp_ring->next_to_use;
@@ -1001,8 +1010,8 @@ static void ice_xmit_pkt_batch(struct ice_tx_ring *xdp_ring, struct xdp_desc *de
 	loop_unrolled_for(i = 0; i < PKTS_PER_BATCH; i++) {
 		dma_addr_t dma;
 
-		dma = xsk_buff_raw_get_dma(xdp_ring->xsk_pool, descs[i].addr);
-		xsk_buff_raw_dma_sync_for_device(xdp_ring->xsk_pool, dma, descs[i].len);
+		dma = xsk_buff_raw_get_dma(xsk_pool, descs[i].addr);
+		xsk_buff_raw_dma_sync_for_device(xsk_pool, dma, descs[i].len);
 
 		tx_desc = ICE_TX_DESC(xdp_ring, ntu++);
 		tx_desc->buf_addr = cpu_to_le64(dma);
@@ -1022,17 +1031,19 @@ static void ice_xmit_pkt_batch(struct ice_tx_ring *xdp_ring, struct xdp_desc *de
  * @nb_pkts: count of packets to be send
  * @total_bytes: bytes accumulator that will be used for stats update
  */
-static void ice_fill_tx_hw_ring(struct ice_tx_ring *xdp_ring, struct xdp_desc *descs,
-				u32 nb_pkts, unsigned int *total_bytes)
+static void ice_fill_tx_hw_ring(struct ice_tx_ring *xdp_ring,
+				struct xsk_buff_pool *xsk_pool,
+				struct xdp_desc *descs, u32 nb_pkts,
+				unsigned int *total_bytes)
 {
 	u32 batched, leftover, i;
 
 	batched = ALIGN_DOWN(nb_pkts, PKTS_PER_BATCH);
 	leftover = nb_pkts & (PKTS_PER_BATCH - 1);
 	for (i = 0; i < batched; i += PKTS_PER_BATCH)
-		ice_xmit_pkt_batch(xdp_ring, &descs[i], total_bytes);
+		ice_xmit_pkt_batch(xdp_ring, xsk_pool, &descs[i], total_bytes);
 	for (; i < batched + leftover; i++)
-		ice_xmit_pkt(xdp_ring, &descs[i], total_bytes);
+		ice_xmit_pkt(xdp_ring, xsk_pool, &descs[i], total_bytes);
 }
 
 /**
@@ -1043,7 +1054,8 @@ static void ice_fill_tx_hw_ring(struct ice_tx_ring *xdp_ring, struct xdp_desc *d
  */
 bool ice_xmit_zc(struct ice_tx_ring *xdp_ring)
 {
-	struct xdp_desc *descs = xdp_ring->xsk_pool->tx_descs;
+	struct xsk_buff_pool *xsk_pool = READ_ONCE(xdp_ring->xsk_pool);
+	struct xdp_desc *descs = xsk_pool->tx_descs;
 	u32 nb_pkts, nb_processed = 0;
 	unsigned int total_bytes = 0;
 	int budget;
@@ -1057,25 +1069,26 @@ bool ice_xmit_zc(struct ice_tx_ring *xdp_ring)
 	budget = ICE_DESC_UNUSED(xdp_ring);
 	budget = min_t(u16, budget, ICE_RING_QUARTER(xdp_ring));
 
-	nb_pkts = xsk_tx_peek_release_desc_batch(xdp_ring->xsk_pool, budget);
+	nb_pkts = xsk_tx_peek_release_desc_batch(xsk_pool, budget);
 	if (!nb_pkts)
 		return true;
 
 	if (xdp_ring->next_to_use + nb_pkts >= xdp_ring->count) {
 		nb_processed = xdp_ring->count - xdp_ring->next_to_use;
-		ice_fill_tx_hw_ring(xdp_ring, descs, nb_processed, &total_bytes);
+		ice_fill_tx_hw_ring(xdp_ring, xsk_pool, descs, nb_processed,
+				    &total_bytes);
 		xdp_ring->next_to_use = 0;
 	}
 
-	ice_fill_tx_hw_ring(xdp_ring, &descs[nb_processed], nb_pkts - nb_processed,
-			    &total_bytes);
+	ice_fill_tx_hw_ring(xdp_ring, xsk_pool, &descs[nb_processed],
+			    nb_pkts - nb_processed, &total_bytes);
 
 	ice_set_rs_bit(xdp_ring);
 	ice_xdp_ring_update_tail(xdp_ring);
 	ice_update_tx_ring_stats(xdp_ring, nb_pkts, total_bytes);
 
-	if (xsk_uses_need_wakeup(xdp_ring->xsk_pool))
-		xsk_set_tx_need_wakeup(xdp_ring->xsk_pool);
+	if (xsk_uses_need_wakeup(xsk_pool))
+		xsk_set_tx_need_wakeup(xsk_pool);
 
 	return nb_pkts < budget;
 }
@@ -1108,7 +1121,7 @@ ice_xsk_wakeup(struct net_device *netdev, u32 queue_id,
 
 	ring = vsi->rx_rings[queue_id]->xdp_ring;
 
-	if (!ring->xsk_pool)
+	if (!READ_ONCE(ring->xsk_pool))
 		return -EINVAL;
 
 	/* The idea here is that if NAPI is running, mark a miss, so
diff --git a/drivers/net/ethernet/intel/ice/ice_xsk.h b/drivers/net/ethernet/intel/ice/ice_xsk.h
index 6fa181f080ef..4cd2d62a0836 100644
--- a/drivers/net/ethernet/intel/ice/ice_xsk.h
+++ b/drivers/net/ethernet/intel/ice/ice_xsk.h
@@ -22,7 +22,8 @@ int ice_xsk_pool_setup(struct ice_vsi *vsi, struct xsk_buff_pool *pool,
 		       u16 qid);
 int ice_clean_rx_irq_zc(struct ice_rx_ring *rx_ring, int budget);
 int ice_xsk_wakeup(struct net_device *netdev, u32 queue_id, u32 flags);
-bool ice_alloc_rx_bufs_zc(struct ice_rx_ring *rx_ring, u16 count);
+bool ice_alloc_rx_bufs_zc(struct ice_rx_ring *rx_ring,
+			  struct xsk_buff_pool *xsk_pool, u16 count);
 bool ice_xsk_any_rx_ring_ena(struct ice_vsi *vsi);
 void ice_xsk_clean_rx_ring(struct ice_rx_ring *rx_ring);
 void ice_xsk_clean_xdp_ring(struct ice_tx_ring *xdp_ring);
@@ -51,6 +52,7 @@ ice_clean_rx_irq_zc(struct ice_rx_ring __always_unused *rx_ring,
 
 static inline bool
 ice_alloc_rx_bufs_zc(struct ice_rx_ring __always_unused *rx_ring,
+		     struct xsk_buff_pool __always_unused *xsk_pool,
 		     u16 __always_unused count)
 {
 	return false;
-- 
2.34.1

Re: [Intel-wired-lan] [PATCH iwl-net 06/11] ice: improve updating ice_{t, r}x_ring::xsk_pool

[kernel test robot]

Hi Maciej,

kernel test robot noticed the following build warnings:

[auto build test WARNING on tnguy-net-queue/dev-queue]

url:    https://github.com/intel-lab-lkp/linux/commits/Maciej-Fijalkowski/ice-respect-netif-readiness-in-AF_XDP-ZC-related-ndo-s/20240528-211914
base:   https://git.kernel.org/pub/scm/linux/kernel/git/tnguy/net-queue.git dev-queue
patch link:    https://lore.kernel.org/r/20240528131429.3012910-7-maciej.fijalkowski%40intel.com
patch subject: [Intel-wired-lan] [PATCH iwl-net 06/11] ice: improve updating ice_{t, r}x_ring::xsk_pool
config: alpha-allyesconfig (https://download.01.org/0day-ci/archive/20240529/202405290101.PV6Uluyq-lkp@intel.com/config)
compiler: alpha-linux-gcc (GCC) 13.2.0
reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20240529/202405290101.PV6Uluyq-lkp@intel.com/reproduce)

If you fix the issue in a separate patch/commit (i.e. not just a new version of
the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <lkp@intel.com>
| Closes: https://lore.kernel.org/oe-kbuild-all/202405290101.PV6Uluyq-lkp@intel.com/

All warnings (new ones prefixed by >>):

>> drivers/net/ethernet/intel/ice/ice_xsk.c:476: warning: Function parameter or struct member 'xsk_pool' not described in '__ice_alloc_rx_bufs_zc'
>> drivers/net/ethernet/intel/ice/ice_xsk.c:525: warning: Function parameter or struct member 'xsk_pool' not described in 'ice_alloc_rx_bufs_zc'
>> drivers/net/ethernet/intel/ice/ice_xsk.c:980: warning: Function parameter or struct member 'xsk_pool' not described in 'ice_xmit_pkt'
>> drivers/net/ethernet/intel/ice/ice_xsk.c:1005: warning: Function parameter or struct member 'xsk_pool' not described in 'ice_xmit_pkt_batch'
>> drivers/net/ethernet/intel/ice/ice_xsk.c:1038: warning: Function parameter or struct member 'xsk_pool' not described in 'ice_fill_tx_hw_ring'


vim +476 drivers/net/ethernet/intel/ice/ice_xsk.c

3876ff525de70ae Maciej Fijalkowski     2022-01-25  462  
3876ff525de70ae Maciej Fijalkowski     2022-01-25  463  /**
3876ff525de70ae Maciej Fijalkowski     2022-01-25  464   * __ice_alloc_rx_bufs_zc - allocate a number of Rx buffers
3876ff525de70ae Maciej Fijalkowski     2022-01-25  465   * @rx_ring: Rx ring
3876ff525de70ae Maciej Fijalkowski     2022-01-25  466   * @count: The number of buffers to allocate
3876ff525de70ae Maciej Fijalkowski     2022-01-25  467   *
3876ff525de70ae Maciej Fijalkowski     2022-01-25  468   * Place the @count of descriptors onto Rx ring. Handle the ring wrap
3876ff525de70ae Maciej Fijalkowski     2022-01-25  469   * for case where space from next_to_use up to the end of ring is less
3876ff525de70ae Maciej Fijalkowski     2022-01-25  470   * than @count. Finally do a tail bump.
3876ff525de70ae Maciej Fijalkowski     2022-01-25  471   *
3876ff525de70ae Maciej Fijalkowski     2022-01-25  472   * Returns true if all allocations were successful, false if any fail.
3876ff525de70ae Maciej Fijalkowski     2022-01-25  473   */
290b7dad8f9f257 Maciej Fijalkowski     2024-05-28  474  static bool __ice_alloc_rx_bufs_zc(struct ice_rx_ring *rx_ring,
290b7dad8f9f257 Maciej Fijalkowski     2024-05-28  475  				   struct xsk_buff_pool *xsk_pool,  u16 count)
3876ff525de70ae Maciej Fijalkowski     2022-01-25 @476  {
d1fc4c6feac18f8 Maciej Fijalkowski     2022-03-17  477  	u32 nb_buffs_extra = 0, nb_buffs = 0;
3876ff525de70ae Maciej Fijalkowski     2022-01-25  478  	union ice_32b_rx_flex_desc *rx_desc;
3876ff525de70ae Maciej Fijalkowski     2022-01-25  479  	u16 ntu = rx_ring->next_to_use;
3876ff525de70ae Maciej Fijalkowski     2022-01-25  480  	u16 total_count = count;
3876ff525de70ae Maciej Fijalkowski     2022-01-25  481  	struct xdp_buff **xdp;
3876ff525de70ae Maciej Fijalkowski     2022-01-25  482  
3876ff525de70ae Maciej Fijalkowski     2022-01-25  483  	rx_desc = ICE_RX_DESC(rx_ring, ntu);
3876ff525de70ae Maciej Fijalkowski     2022-01-25  484  	xdp = ice_xdp_buf(rx_ring, ntu);
3876ff525de70ae Maciej Fijalkowski     2022-01-25  485  
3876ff525de70ae Maciej Fijalkowski     2022-01-25  486  	if (ntu + count >= rx_ring->count) {
290b7dad8f9f257 Maciej Fijalkowski     2024-05-28  487  		nb_buffs_extra = ice_fill_rx_descs(xsk_pool, xdp, rx_desc,
3876ff525de70ae Maciej Fijalkowski     2022-01-25  488  						   rx_ring->count - ntu);
d1fc4c6feac18f8 Maciej Fijalkowski     2022-03-17  489  		if (nb_buffs_extra != rx_ring->count - ntu) {
d1fc4c6feac18f8 Maciej Fijalkowski     2022-03-17  490  			ntu += nb_buffs_extra;
d1fc4c6feac18f8 Maciej Fijalkowski     2022-03-17  491  			goto exit;
d1fc4c6feac18f8 Maciej Fijalkowski     2022-03-17  492  		}
3876ff525de70ae Maciej Fijalkowski     2022-01-25  493  		rx_desc = ICE_RX_DESC(rx_ring, 0);
3876ff525de70ae Maciej Fijalkowski     2022-01-25  494  		xdp = ice_xdp_buf(rx_ring, 0);
3876ff525de70ae Maciej Fijalkowski     2022-01-25  495  		ntu = 0;
3876ff525de70ae Maciej Fijalkowski     2022-01-25  496  		count -= nb_buffs_extra;
3876ff525de70ae Maciej Fijalkowski     2022-01-25  497  		ice_release_rx_desc(rx_ring, 0);
3876ff525de70ae Maciej Fijalkowski     2022-01-25  498  	}
3876ff525de70ae Maciej Fijalkowski     2022-01-25  499  
290b7dad8f9f257 Maciej Fijalkowski     2024-05-28  500  	nb_buffs = ice_fill_rx_descs(xsk_pool, xdp, rx_desc, count);
3876ff525de70ae Maciej Fijalkowski     2022-01-25  501  
db804cfc21e969a Magnus Karlsson        2021-09-22  502  	ntu += nb_buffs;
8b51a13c37c24c0 Maciej Fijalkowski     2021-12-13  503  	if (ntu == rx_ring->count)
2d4238f55697221 Krzysztof Kazimierczak 2019-11-04  504  		ntu = 0;
2d4238f55697221 Krzysztof Kazimierczak 2019-11-04  505  
d1fc4c6feac18f8 Maciej Fijalkowski     2022-03-17  506  exit:
3876ff525de70ae Maciej Fijalkowski     2022-01-25  507  	if (rx_ring->next_to_use != ntu)
2d4238f55697221 Krzysztof Kazimierczak 2019-11-04  508  		ice_release_rx_desc(rx_ring, ntu);
2d4238f55697221 Krzysztof Kazimierczak 2019-11-04  509  
3876ff525de70ae Maciej Fijalkowski     2022-01-25  510  	return total_count == (nb_buffs_extra + nb_buffs);
3876ff525de70ae Maciej Fijalkowski     2022-01-25  511  }
3876ff525de70ae Maciej Fijalkowski     2022-01-25  512  
3876ff525de70ae Maciej Fijalkowski     2022-01-25  513  /**
3876ff525de70ae Maciej Fijalkowski     2022-01-25  514   * ice_alloc_rx_bufs_zc - allocate a number of Rx buffers
3876ff525de70ae Maciej Fijalkowski     2022-01-25  515   * @rx_ring: Rx ring
3876ff525de70ae Maciej Fijalkowski     2022-01-25  516   * @count: The number of buffers to allocate
3876ff525de70ae Maciej Fijalkowski     2022-01-25  517   *
3876ff525de70ae Maciej Fijalkowski     2022-01-25  518   * Wrapper for internal allocation routine; figure out how many tail
3876ff525de70ae Maciej Fijalkowski     2022-01-25  519   * bumps should take place based on the given threshold
3876ff525de70ae Maciej Fijalkowski     2022-01-25  520   *
3876ff525de70ae Maciej Fijalkowski     2022-01-25  521   * Returns true if all calls to internal alloc routine succeeded
3876ff525de70ae Maciej Fijalkowski     2022-01-25  522   */
290b7dad8f9f257 Maciej Fijalkowski     2024-05-28  523  bool ice_alloc_rx_bufs_zc(struct ice_rx_ring *rx_ring,
290b7dad8f9f257 Maciej Fijalkowski     2024-05-28  524  			  struct xsk_buff_pool *xsk_pool, u16 count)
3876ff525de70ae Maciej Fijalkowski     2022-01-25 @525  {
3876ff525de70ae Maciej Fijalkowski     2022-01-25  526  	u16 rx_thresh = ICE_RING_QUARTER(rx_ring);
b3056ae2b57858b Maciej Fijalkowski     2022-09-01  527  	u16 leftover, i, tail_bumps;
3876ff525de70ae Maciej Fijalkowski     2022-01-25  528  
b3056ae2b57858b Maciej Fijalkowski     2022-09-01  529  	tail_bumps = count / rx_thresh;
b3056ae2b57858b Maciej Fijalkowski     2022-09-01  530  	leftover = count - (tail_bumps * rx_thresh);
3876ff525de70ae Maciej Fijalkowski     2022-01-25  531  
3876ff525de70ae Maciej Fijalkowski     2022-01-25  532  	for (i = 0; i < tail_bumps; i++)
290b7dad8f9f257 Maciej Fijalkowski     2024-05-28  533  		if (!__ice_alloc_rx_bufs_zc(rx_ring, xsk_pool, rx_thresh))
3876ff525de70ae Maciej Fijalkowski     2022-01-25  534  			return false;
290b7dad8f9f257 Maciej Fijalkowski     2024-05-28  535  	return __ice_alloc_rx_bufs_zc(rx_ring, xsk_pool, leftover);
2d4238f55697221 Krzysztof Kazimierczak 2019-11-04  536  }
2d4238f55697221 Krzysztof Kazimierczak 2019-11-04  537  

-- 
0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests/wiki

[Intel-wired-lan] [PATCH iwl-net 07/11] ice: add missing WRITE_ONCE when clearing ice_rx_ring::xdp_prog

[Maciej Fijalkowski]

It is read by data path and modified from process context on remote cpu
so it is needed to use WRITE_ONCE to clear the pointer.

Fixes: efc2214b6047 ("ice: Add support for XDP")
Signed-off-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
---
 drivers/net/ethernet/intel/ice/ice_txrx.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/net/ethernet/intel/ice/ice_txrx.c b/drivers/net/ethernet/intel/ice/ice_txrx.c
index f4b2b1bca234..4c115531beba 100644
--- a/drivers/net/ethernet/intel/ice/ice_txrx.c
+++ b/drivers/net/ethernet/intel/ice/ice_txrx.c
@@ -456,7 +456,7 @@ void ice_free_rx_ring(struct ice_rx_ring *rx_ring)
 	if (rx_ring->vsi->type == ICE_VSI_PF)
 		if (xdp_rxq_info_is_reg(&rx_ring->xdp_rxq))
 			xdp_rxq_info_unreg(&rx_ring->xdp_rxq);
-	rx_ring->xdp_prog = NULL;
+	WRITE_ONCE(rx_ring->xdp_prog, NULL);
 	if (rx_ring->xsk_pool) {
 		kfree(rx_ring->xdp_buf);
 		rx_ring->xdp_buf = NULL;
-- 
2.34.1

[Intel-wired-lan] [PATCH iwl-net 08/11] ice: xsk: fix txq interrupt mapping

[Maciej Fijalkowski]

ice_cfg_txq_interrupt() internally handles XDP Tx ring. Do not use
ice_for_each_tx_ring() in ice_qvec_cfg_msix() as this causing us to
treat XDP ring that belongs to queue vector as Tx ring and therefore
misconfiguring the interrupts.

Fixes: 2d4238f55697 ("ice: Add support for AF_XDP")
Signed-off-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
---
 drivers/net/ethernet/intel/ice/ice_xsk.c | 23 +++++++++++++----------
 1 file changed, 13 insertions(+), 10 deletions(-)

diff --git a/drivers/net/ethernet/intel/ice/ice_xsk.c b/drivers/net/ethernet/intel/ice/ice_xsk.c
index e554cf424fb3..3135fc0aaf73 100644
--- a/drivers/net/ethernet/intel/ice/ice_xsk.c
+++ b/drivers/net/ethernet/intel/ice/ice_xsk.c
@@ -113,23 +113,26 @@ ice_qvec_dis_irq(struct ice_vsi *vsi, struct ice_rx_ring *rx_ring,
  * @q_vector: queue vector
  */
 static void
-ice_qvec_cfg_msix(struct ice_vsi *vsi, struct ice_q_vector *q_vector)
+ice_qvec_cfg_msix(struct ice_vsi *vsi, struct ice_q_vector *q_vector, u16 qid)
 {
 	u16 reg_idx = q_vector->reg_idx;
 	struct ice_pf *pf = vsi->back;
 	struct ice_hw *hw = &pf->hw;
-	struct ice_tx_ring *tx_ring;
-	struct ice_rx_ring *rx_ring;
+	int q, _qid = qid;
 
 	ice_cfg_itr(hw, q_vector);
 
-	ice_for_each_tx_ring(tx_ring, q_vector->tx)
-		ice_cfg_txq_interrupt(vsi, tx_ring->reg_idx, reg_idx,
-				      q_vector->tx.itr_idx);
+	for (q = 0; q < q_vector->num_ring_tx; q++) {
+		ice_cfg_txq_interrupt(vsi, _qid, reg_idx, q_vector->tx.itr_idx);
+		_qid++;
+	}
 
-	ice_for_each_rx_ring(rx_ring, q_vector->rx)
-		ice_cfg_rxq_interrupt(vsi, rx_ring->reg_idx, reg_idx,
-				      q_vector->rx.itr_idx);
+	_qid = qid;
+
+	for (q = 0; q < q_vector->num_ring_rx; q++) {
+		ice_cfg_rxq_interrupt(vsi, _qid, reg_idx, q_vector->rx.itr_idx);
+		_qid++;
+	}
 
 	ice_flush(hw);
 }
@@ -241,7 +244,7 @@ static int ice_qp_ena(struct ice_vsi *vsi, u16 q_idx)
 		fail = err;
 
 	q_vector = vsi->rx_rings[q_idx]->q_vector;
-	ice_qvec_cfg_msix(vsi, q_vector);
+	ice_qvec_cfg_msix(vsi, q_vector, q_idx);
 
 	err = ice_vsi_ctrl_one_rx_ring(vsi, true, q_idx, true);
 	if (!fail)
-- 
2.34.1

Re: [Intel-wired-lan] [PATCH iwl-net 08/11] ice: xsk: fix txq interrupt mapping

[kernel test robot]

Hi Maciej,

kernel test robot noticed the following build warnings:

[auto build test WARNING on tnguy-net-queue/dev-queue]

url:    https://github.com/intel-lab-lkp/linux/commits/Maciej-Fijalkowski/ice-respect-netif-readiness-in-AF_XDP-ZC-related-ndo-s/20240528-211914
base:   https://git.kernel.org/pub/scm/linux/kernel/git/tnguy/net-queue.git dev-queue
patch link:    https://lore.kernel.org/r/20240528131429.3012910-9-maciej.fijalkowski%40intel.com
patch subject: [Intel-wired-lan] [PATCH iwl-net 08/11] ice: xsk: fix txq interrupt mapping
config: alpha-allyesconfig (https://download.01.org/0day-ci/archive/20240529/202405290301.TKfQE09f-lkp@intel.com/config)
compiler: alpha-linux-gcc (GCC) 13.2.0
reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20240529/202405290301.TKfQE09f-lkp@intel.com/reproduce)

If you fix the issue in a separate patch/commit (i.e. not just a new version of
the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <lkp@intel.com>
| Closes: https://lore.kernel.org/oe-kbuild-all/202405290301.TKfQE09f-lkp@intel.com/

All warnings (new ones prefixed by >>):

>> drivers/net/ethernet/intel/ice/ice_xsk.c:117: warning: Function parameter or struct member 'qid' not described in 'ice_qvec_cfg_msix'
   drivers/net/ethernet/intel/ice/ice_xsk.c:479: warning: Function parameter or struct member 'xsk_pool' not described in '__ice_alloc_rx_bufs_zc'
   drivers/net/ethernet/intel/ice/ice_xsk.c:528: warning: Function parameter or struct member 'xsk_pool' not described in 'ice_alloc_rx_bufs_zc'
   drivers/net/ethernet/intel/ice/ice_xsk.c:983: warning: Function parameter or struct member 'xsk_pool' not described in 'ice_xmit_pkt'
   drivers/net/ethernet/intel/ice/ice_xsk.c:1008: warning: Function parameter or struct member 'xsk_pool' not described in 'ice_xmit_pkt_batch'
   drivers/net/ethernet/intel/ice/ice_xsk.c:1041: warning: Function parameter or struct member 'xsk_pool' not described in 'ice_fill_tx_hw_ring'


vim +117 drivers/net/ethernet/intel/ice/ice_xsk.c

2d4238f55697221 Krzysztof Kazimierczak 2019-11-04  109  
2d4238f55697221 Krzysztof Kazimierczak 2019-11-04  110  /**
2d4238f55697221 Krzysztof Kazimierczak 2019-11-04  111   * ice_qvec_cfg_msix - Enable IRQ for given queue vector
2d4238f55697221 Krzysztof Kazimierczak 2019-11-04  112   * @vsi: the VSI that contains queue vector
2d4238f55697221 Krzysztof Kazimierczak 2019-11-04  113   * @q_vector: queue vector
2d4238f55697221 Krzysztof Kazimierczak 2019-11-04  114   */
2d4238f55697221 Krzysztof Kazimierczak 2019-11-04  115  static void
d77176d15a1b0ed Maciej Fijalkowski     2024-05-28  116  ice_qvec_cfg_msix(struct ice_vsi *vsi, struct ice_q_vector *q_vector, u16 qid)
2d4238f55697221 Krzysztof Kazimierczak 2019-11-04 @117  {
2d4238f55697221 Krzysztof Kazimierczak 2019-11-04  118  	u16 reg_idx = q_vector->reg_idx;
2d4238f55697221 Krzysztof Kazimierczak 2019-11-04  119  	struct ice_pf *pf = vsi->back;
2d4238f55697221 Krzysztof Kazimierczak 2019-11-04  120  	struct ice_hw *hw = &pf->hw;
d77176d15a1b0ed Maciej Fijalkowski     2024-05-28  121  	int q, _qid = qid;
2d4238f55697221 Krzysztof Kazimierczak 2019-11-04  122  
2d4238f55697221 Krzysztof Kazimierczak 2019-11-04  123  	ice_cfg_itr(hw, q_vector);
2d4238f55697221 Krzysztof Kazimierczak 2019-11-04  124  
d77176d15a1b0ed Maciej Fijalkowski     2024-05-28  125  	for (q = 0; q < q_vector->num_ring_tx; q++) {
d77176d15a1b0ed Maciej Fijalkowski     2024-05-28  126  		ice_cfg_txq_interrupt(vsi, _qid, reg_idx, q_vector->tx.itr_idx);
d77176d15a1b0ed Maciej Fijalkowski     2024-05-28  127  		_qid++;
d77176d15a1b0ed Maciej Fijalkowski     2024-05-28  128  	}
2d4238f55697221 Krzysztof Kazimierczak 2019-11-04  129  
d77176d15a1b0ed Maciej Fijalkowski     2024-05-28  130  	_qid = qid;
d77176d15a1b0ed Maciej Fijalkowski     2024-05-28  131  
d77176d15a1b0ed Maciej Fijalkowski     2024-05-28  132  	for (q = 0; q < q_vector->num_ring_rx; q++) {
d77176d15a1b0ed Maciej Fijalkowski     2024-05-28  133  		ice_cfg_rxq_interrupt(vsi, _qid, reg_idx, q_vector->rx.itr_idx);
d77176d15a1b0ed Maciej Fijalkowski     2024-05-28  134  		_qid++;
d77176d15a1b0ed Maciej Fijalkowski     2024-05-28  135  	}
2d4238f55697221 Krzysztof Kazimierczak 2019-11-04  136  
2d4238f55697221 Krzysztof Kazimierczak 2019-11-04  137  	ice_flush(hw);
2d4238f55697221 Krzysztof Kazimierczak 2019-11-04  138  }
2d4238f55697221 Krzysztof Kazimierczak 2019-11-04  139  

-- 
0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests/wiki

[Intel-wired-lan] [PATCH iwl-net 09/11] ice: move locking outside of ice_qp_ena and ice_qp_dis

[Maciej Fijalkowski]

From: Larysa Zaremba <larysa.zaremba@intel.com>

Currently, ice_qp_ena() is called, even if ICE_CFG_BUSY could not be
acquired by ice_qp_dis(), in such case there is nothing to undo.
Move locking logic out of these functions, so:
* we immediately return, if the lock could not be acquired
* ice_qp_ena() does not operate in an unsafe context
* ice_qp_ena() does not clear ICE_CFG_BUSY, when it is not held

Fixes: 2d4238f55697 ("ice: Add support for AF_XDP")
Signed-off-by: Larysa Zaremba <larysa.zaremba@intel.com>
---
 drivers/net/ethernet/intel/ice/ice_xsk.c | 18 +++++++++---------
 1 file changed, 9 insertions(+), 9 deletions(-)

diff --git a/drivers/net/ethernet/intel/ice/ice_xsk.c b/drivers/net/ethernet/intel/ice/ice_xsk.c
index 3135fc0aaf73..fe4aa4b537dd 100644
--- a/drivers/net/ethernet/intel/ice/ice_xsk.c
+++ b/drivers/net/ethernet/intel/ice/ice_xsk.c
@@ -165,7 +165,6 @@ static int ice_qp_dis(struct ice_vsi *vsi, u16 q_idx)
 	struct ice_q_vector *q_vector;
 	struct ice_tx_ring *tx_ring;
 	struct ice_rx_ring *rx_ring;
-	int timeout = 50;
 	int fail = 0;
 	int err;
 
@@ -176,13 +175,6 @@ static int ice_qp_dis(struct ice_vsi *vsi, u16 q_idx)
 	rx_ring = vsi->rx_rings[q_idx];
 	q_vector = rx_ring->q_vector;
 
-	while (test_and_set_bit(ICE_CFG_BUSY, vsi->state)) {
-		timeout--;
-		if (!timeout)
-			return -EBUSY;
-		usleep_range(1000, 2000);
-	}
-
 	synchronize_net();
 	netif_carrier_off(vsi->netdev);
 	netif_tx_stop_queue(netdev_get_tx_queue(vsi->netdev, q_idx));
@@ -257,7 +249,6 @@ static int ice_qp_ena(struct ice_vsi *vsi, u16 q_idx)
 	synchronize_net();
 	netif_tx_start_queue(netdev_get_tx_queue(vsi->netdev, q_idx));
 	netif_carrier_on(vsi->netdev);
-	clear_bit(ICE_CFG_BUSY, vsi->state);
 
 	return fail;
 }
@@ -390,6 +381,14 @@ int ice_xsk_pool_setup(struct ice_vsi *vsi, struct xsk_buff_pool *pool, u16 qid)
 
 	if (if_running) {
 		struct ice_rx_ring *rx_ring = vsi->rx_rings[qid];
+		int timeout = 50;
+
+		while (test_and_set_bit(ICE_CFG_BUSY, vsi->state)) {
+			timeout--;
+			if (!timeout)
+				return -EBUSY;
+			usleep_range(1000, 2000);
+		}
 
 		ret = ice_qp_dis(vsi, qid);
 		if (ret) {
@@ -412,6 +411,7 @@ int ice_xsk_pool_setup(struct ice_vsi *vsi, struct xsk_buff_pool *pool, u16 qid)
 			napi_schedule(&vsi->rx_rings[qid]->xdp_ring->q_vector->napi);
 		else if (ret)
 			netdev_err(vsi->netdev, "ice_qp_ena error = %d\n", ret);
+		clear_bit(ICE_CFG_BUSY, vsi->state);
 	}
 
 failure:
-- 
2.34.1

[Intel-wired-lan] [PATCH iwl-net 10/11] ice: lock with PF state instead of VSI state in ice_xsk_pool_setup()

[Maciej Fijalkowski]

From: Larysa Zaremba <larysa.zaremba@intel.com>

The main intent of using ICE_CFG_BUSY is to prevent the reset from
starting, when other configuration is being processed. pf->state is checked
before starting reset, but ice_xsk_pool_setup sets the flag in vsi->state,
which is almost useless. Also, ICE_CFG_BUSY belongs to enum ice_pf_state,
not ice_vsi_state.

Change vsi->state to pf->state in ice_xsk_pool_setup() locking code, so
reset would not interfere with AF_XDP configuration.

Fixes: 2d4238f55697 ("ice: Add support for AF_XDP")
Signed-off-by: Larysa Zaremba <larysa.zaremba@intel.com>
---
 drivers/net/ethernet/intel/ice/ice_xsk.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/drivers/net/ethernet/intel/ice/ice_xsk.c b/drivers/net/ethernet/intel/ice/ice_xsk.c
index fe4aa4b537dd..225d027d3d7a 100644
--- a/drivers/net/ethernet/intel/ice/ice_xsk.c
+++ b/drivers/net/ethernet/intel/ice/ice_xsk.c
@@ -370,6 +370,7 @@ int ice_xsk_pool_setup(struct ice_vsi *vsi, struct xsk_buff_pool *pool, u16 qid)
 {
 	bool if_running, pool_present = !!pool;
 	int ret = 0, pool_failure = 0;
+	struct ice_pf *pf = vsi->back;
 
 	if (qid >= vsi->num_rxq || qid >= vsi->num_txq) {
 		netdev_err(vsi->netdev, "Please use queue id in scope of combined queues count\n");
@@ -383,7 +384,7 @@ int ice_xsk_pool_setup(struct ice_vsi *vsi, struct xsk_buff_pool *pool, u16 qid)
 		struct ice_rx_ring *rx_ring = vsi->rx_rings[qid];
 		int timeout = 50;
 
-		while (test_and_set_bit(ICE_CFG_BUSY, vsi->state)) {
+		while (test_and_set_bit(ICE_CFG_BUSY, pf->state)) {
 			timeout--;
 			if (!timeout)
 				return -EBUSY;
@@ -411,7 +412,7 @@ int ice_xsk_pool_setup(struct ice_vsi *vsi, struct xsk_buff_pool *pool, u16 qid)
 			napi_schedule(&vsi->rx_rings[qid]->xdp_ring->q_vector->napi);
 		else if (ret)
 			netdev_err(vsi->netdev, "ice_qp_ena error = %d\n", ret);
-		clear_bit(ICE_CFG_BUSY, vsi->state);
+		clear_bit(ICE_CFG_BUSY, pf->state);
 	}
 
 failure:
-- 
2.34.1

[Intel-wired-lan] [PATCH iwl-net 11/11] ice: protect ring configuration with a mutex

[Maciej Fijalkowski]

From: Larysa Zaremba <larysa.zaremba@intel.com>

Add a ring_lock mutex to protect sections, where software rings are
affected. Particularly, to prevent system crash, when tx_timeout
and .ndo_bpf() happen at the same time.

Fixes: 2d4238f55697 ("ice: Add support for AF_XDP")
Signed-off-by: Larysa Zaremba <larysa.zaremba@intel.com>
---
 drivers/net/ethernet/intel/ice/ice.h      |  2 ++
 drivers/net/ethernet/intel/ice/ice_lib.c  | 23 ++++++++++---
 drivers/net/ethernet/intel/ice/ice_main.c | 39 ++++++++++++++++++++---
 drivers/net/ethernet/intel/ice/ice_xsk.c  | 13 ++------
 4 files changed, 57 insertions(+), 20 deletions(-)

diff --git a/drivers/net/ethernet/intel/ice/ice.h b/drivers/net/ethernet/intel/ice/ice.h
index 701a61d791dd..7c1e24afa34b 100644
--- a/drivers/net/ethernet/intel/ice/ice.h
+++ b/drivers/net/ethernet/intel/ice/ice.h
@@ -307,6 +307,7 @@ enum ice_pf_state {
 	ICE_PHY_INIT_COMPLETE,
 	ICE_FD_VF_FLUSH_CTX,		/* set at FD Rx IRQ or timeout */
 	ICE_AUX_ERR_PENDING,
+	ICE_RTNL_WAITS_FOR_RESET,
 	ICE_STATE_NBITS		/* must be last */
 };
 
@@ -941,6 +942,7 @@ int ice_prepare_xdp_rings(struct ice_vsi *vsi, struct bpf_prog *prog,
 			  enum ice_xdp_cfg cfg_type);
 int ice_destroy_xdp_rings(struct ice_vsi *vsi, enum ice_xdp_cfg cfg_type);
 void ice_map_xdp_rings(struct ice_vsi *vsi);
+bool ice_rebuild_pending(struct ice_vsi *vsi);
 int
 ice_xdp_xmit(struct net_device *dev, int n, struct xdp_frame **frames,
 	     u32 flags);
diff --git a/drivers/net/ethernet/intel/ice/ice_lib.c b/drivers/net/ethernet/intel/ice/ice_lib.c
index 7629b0190578..a5dc6fc6e63d 100644
--- a/drivers/net/ethernet/intel/ice/ice_lib.c
+++ b/drivers/net/ethernet/intel/ice/ice_lib.c
@@ -2426,7 +2426,10 @@ void ice_vsi_decfg(struct ice_vsi *vsi)
 		dev_err(ice_pf_to_dev(pf), "Failed to remove RDMA scheduler config for VSI %u, err %d\n",
 			vsi->vsi_num, err);
 
-	if (ice_is_xdp_ena_vsi(vsi))
+	/* xdp_rings can be absent, if program was attached amid reset,
+	 * VSI rebuild is supposed to create them later
+	 */
+	if (ice_is_xdp_ena_vsi(vsi) && vsi->xdp_rings)
 		/* return value check can be skipped here, it always returns
 		 * 0 if reset is in progress
 		 */
@@ -2737,12 +2740,24 @@ ice_queue_set_napi(struct ice_vsi *vsi, unsigned int queue_index,
 	if (current_work() == &pf->serv_task ||
 	    test_bit(ICE_PREPARED_FOR_RESET, pf->state) ||
 	    test_bit(ICE_DOWN, pf->state) ||
-	    test_bit(ICE_SUSPENDED, pf->state))
+	    test_bit(ICE_SUSPENDED, pf->state)) {
+		bool rtnl_held_here = true;
+
+		while (!rtnl_trylock()) {
+			if (test_bit(ICE_RTNL_WAITS_FOR_RESET, pf->state)) {
+				rtnl_held_here = false;
+				break;
+			}
+			usleep_range(1000, 2000);
+		}
 		__ice_queue_set_napi(vsi->netdev, queue_index, type, napi,
-				     false);
-	else
+				     true);
+		if (rtnl_held_here)
+			rtnl_unlock();
+	} else {
 		__ice_queue_set_napi(vsi->netdev, queue_index, type, napi,
 				     true);
+	}
 }
 
 /**
diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c
index 15a6805ac2a1..7724ed8fc1b1 100644
--- a/drivers/net/ethernet/intel/ice/ice_main.c
+++ b/drivers/net/ethernet/intel/ice/ice_main.c
@@ -2986,6 +2986,20 @@ static int ice_max_xdp_frame_size(struct ice_vsi *vsi)
 		return ICE_RXBUF_3072;
 }
 
+/**
+ * ice_rebuild_pending - ice_vsi_rebuild will be performed, when locks are released
+ * @vsi: VSI to setup XDP for
+ *
+ * ice_vsi_close() in the reset path is called under rtnl_lock(),
+ * so it happened strictly before or after .ndo_bpf().
+ * In case it has happened before, we do not have anything attached to rings
+ */
+bool ice_rebuild_pending(struct ice_vsi *vsi)
+{
+	return ice_is_reset_in_progress(vsi->back->state) &&
+	       !vsi->rx_rings[0]->desc;
+}
+
 /**
  * ice_xdp_setup_prog - Add or remove XDP eBPF program
  * @vsi: VSI to setup XDP for
@@ -3009,7 +3023,7 @@ ice_xdp_setup_prog(struct ice_vsi *vsi, struct bpf_prog *prog,
 	}
 
 	/* hot swap progs and avoid toggling link */
-	if (ice_is_xdp_ena_vsi(vsi) == !!prog) {
+	if (ice_is_xdp_ena_vsi(vsi) == !!prog || ice_rebuild_pending(vsi)) {
 		ice_vsi_assign_bpf_prog(vsi, prog);
 		return 0;
 	}
@@ -3081,21 +3095,33 @@ static int ice_xdp(struct net_device *dev, struct netdev_bpf *xdp)
 {
 	struct ice_netdev_priv *np = netdev_priv(dev);
 	struct ice_vsi *vsi = np->vsi;
+	struct ice_pf *pf = vsi->back;
+	int ret;
 
 	if (vsi->type != ICE_VSI_PF) {
 		NL_SET_ERR_MSG_MOD(xdp->extack, "XDP can be loaded only on PF VSI");
 		return -EINVAL;
 	}
 
+	while (test_and_set_bit(ICE_CFG_BUSY, pf->state)) {
+		set_bit(ICE_RTNL_WAITS_FOR_RESET, pf->state);
+		usleep_range(1000, 2000);
+	}
+	clear_bit(ICE_RTNL_WAITS_FOR_RESET, pf->state);
+
 	switch (xdp->command) {
 	case XDP_SETUP_PROG:
-		return ice_xdp_setup_prog(vsi, xdp->prog, xdp->extack);
+		ret = ice_xdp_setup_prog(vsi, xdp->prog, xdp->extack);
+		break;
 	case XDP_SETUP_XSK_POOL:
-		return ice_xsk_pool_setup(vsi, xdp->xsk.pool,
-					  xdp->xsk.queue_id);
+		ret = ice_xsk_pool_setup(vsi, xdp->xsk.pool, xdp->xsk.queue_id);
+		break;
 	default:
-		return -EINVAL;
+		ret = -EINVAL;
 	}
+
+	clear_bit(ICE_CFG_BUSY, pf->state);
+	return ret;
 }
 
 /**
@@ -7672,7 +7698,10 @@ static void ice_rebuild(struct ice_pf *pf, enum ice_reset_req reset_type)
 		ice_gnss_init(pf);
 
 	/* rebuild PF VSI */
+	while (test_and_set_bit(ICE_CFG_BUSY, pf->state))
+		usleep_range(1000, 2000);
 	err = ice_vsi_rebuild_by_type(pf, ICE_VSI_PF);
+	clear_bit(ICE_CFG_BUSY, pf->state);
 	if (err) {
 		dev_err(dev, "PF VSI rebuild failed: %d\n", err);
 		goto err_vsi_rebuild;
diff --git a/drivers/net/ethernet/intel/ice/ice_xsk.c b/drivers/net/ethernet/intel/ice/ice_xsk.c
index 225d027d3d7a..962af14f9fd5 100644
--- a/drivers/net/ethernet/intel/ice/ice_xsk.c
+++ b/drivers/net/ethernet/intel/ice/ice_xsk.c
@@ -370,7 +370,6 @@ int ice_xsk_pool_setup(struct ice_vsi *vsi, struct xsk_buff_pool *pool, u16 qid)
 {
 	bool if_running, pool_present = !!pool;
 	int ret = 0, pool_failure = 0;
-	struct ice_pf *pf = vsi->back;
 
 	if (qid >= vsi->num_rxq || qid >= vsi->num_txq) {
 		netdev_err(vsi->netdev, "Please use queue id in scope of combined queues count\n");
@@ -378,18 +377,11 @@ int ice_xsk_pool_setup(struct ice_vsi *vsi, struct xsk_buff_pool *pool, u16 qid)
 		goto failure;
 	}
 
-	if_running = netif_running(vsi->netdev) && ice_is_xdp_ena_vsi(vsi);
+	if_running = !ice_rebuild_pending(vsi) &&
+		     (netif_running(vsi->netdev) && ice_is_xdp_ena_vsi(vsi));
 
 	if (if_running) {
 		struct ice_rx_ring *rx_ring = vsi->rx_rings[qid];
-		int timeout = 50;
-
-		while (test_and_set_bit(ICE_CFG_BUSY, pf->state)) {
-			timeout--;
-			if (!timeout)
-				return -EBUSY;
-			usleep_range(1000, 2000);
-		}
 
 		ret = ice_qp_dis(vsi, qid);
 		if (ret) {
@@ -412,7 +404,6 @@ int ice_xsk_pool_setup(struct ice_vsi *vsi, struct xsk_buff_pool *pool, u16 qid)
 			napi_schedule(&vsi->rx_rings[qid]->xdp_ring->q_vector->napi);
 		else if (ret)
 			netdev_err(vsi->netdev, "ice_qp_ena error = %d\n", ret);
-		clear_bit(ICE_CFG_BUSY, pf->state);
 	}
 
 failure:
-- 
2.34.1

Re: [Intel-wired-lan] [PATCH iwl-net 11/11] ice: protect ring configuration with a mutex

[Larysa Zaremba]

On Tue, May 28, 2024 at 03:14:29PM +0200, Maciej Fijalkowski wrote:
> From: Larysa Zaremba <larysa.zaremba@intel.com>
> 
> Add a ring_lock mutex to protect sections, where software rings are
> affected. Particularly, to prevent system crash, when tx_timeout
> and .ndo_bpf() happen at the same time.
> 
> Fixes: 2d4238f55697 ("ice: Add support for AF_XDP")
> Signed-off-by: Larysa Zaremba <larysa.zaremba@intel.com>

This is not the latest version of my patches. Code is the same, but newer 
version has better patch division and more relevant commit messages. Maciej will 
update the patches in v2.

> ---
>  drivers/net/ethernet/intel/ice/ice.h      |  2 ++
>  drivers/net/ethernet/intel/ice/ice_lib.c  | 23 ++++++++++---
>  drivers/net/ethernet/intel/ice/ice_main.c | 39 ++++++++++++++++++++---
>  drivers/net/ethernet/intel/ice/ice_xsk.c  | 13 ++------
>  4 files changed, 57 insertions(+), 20 deletions(-)
> 
> diff --git a/drivers/net/ethernet/intel/ice/ice.h b/drivers/net/ethernet/intel/ice/ice.h
> index 701a61d791dd..7c1e24afa34b 100644
> --- a/drivers/net/ethernet/intel/ice/ice.h
> +++ b/drivers/net/ethernet/intel/ice/ice.h
> @@ -307,6 +307,7 @@ enum ice_pf_state {
>  	ICE_PHY_INIT_COMPLETE,
>  	ICE_FD_VF_FLUSH_CTX,		/* set at FD Rx IRQ or timeout */
>  	ICE_AUX_ERR_PENDING,
> +	ICE_RTNL_WAITS_FOR_RESET,
>  	ICE_STATE_NBITS		/* must be last */
>  };
>  
> @@ -941,6 +942,7 @@ int ice_prepare_xdp_rings(struct ice_vsi *vsi, struct bpf_prog *prog,
>  			  enum ice_xdp_cfg cfg_type);
>  int ice_destroy_xdp_rings(struct ice_vsi *vsi, enum ice_xdp_cfg cfg_type);
>  void ice_map_xdp_rings(struct ice_vsi *vsi);
> +bool ice_rebuild_pending(struct ice_vsi *vsi);
>  int
>  ice_xdp_xmit(struct net_device *dev, int n, struct xdp_frame **frames,
>  	     u32 flags);
> diff --git a/drivers/net/ethernet/intel/ice/ice_lib.c b/drivers/net/ethernet/intel/ice/ice_lib.c
> index 7629b0190578..a5dc6fc6e63d 100644
> --- a/drivers/net/ethernet/intel/ice/ice_lib.c
> +++ b/drivers/net/ethernet/intel/ice/ice_lib.c
> @@ -2426,7 +2426,10 @@ void ice_vsi_decfg(struct ice_vsi *vsi)
>  		dev_err(ice_pf_to_dev(pf), "Failed to remove RDMA scheduler config for VSI %u, err %d\n",
>  			vsi->vsi_num, err);
>  
> -	if (ice_is_xdp_ena_vsi(vsi))
> +	/* xdp_rings can be absent, if program was attached amid reset,
> +	 * VSI rebuild is supposed to create them later
> +	 */
> +	if (ice_is_xdp_ena_vsi(vsi) && vsi->xdp_rings)
>  		/* return value check can be skipped here, it always returns
>  		 * 0 if reset is in progress
>  		 */
> @@ -2737,12 +2740,24 @@ ice_queue_set_napi(struct ice_vsi *vsi, unsigned int queue_index,
>  	if (current_work() == &pf->serv_task ||
>  	    test_bit(ICE_PREPARED_FOR_RESET, pf->state) ||
>  	    test_bit(ICE_DOWN, pf->state) ||
> -	    test_bit(ICE_SUSPENDED, pf->state))
> +	    test_bit(ICE_SUSPENDED, pf->state)) {
> +		bool rtnl_held_here = true;
> +
> +		while (!rtnl_trylock()) {
> +			if (test_bit(ICE_RTNL_WAITS_FOR_RESET, pf->state)) {
> +				rtnl_held_here = false;
> +				break;
> +			}
> +			usleep_range(1000, 2000);
> +		}
>  		__ice_queue_set_napi(vsi->netdev, queue_index, type, napi,
> -				     false);
> -	else
> +				     true);
> +		if (rtnl_held_here)
> +			rtnl_unlock();
> +	} else {
>  		__ice_queue_set_napi(vsi->netdev, queue_index, type, napi,
>  				     true);
> +	}
>  }
>  
>  /**
> diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c
> index 15a6805ac2a1..7724ed8fc1b1 100644
> --- a/drivers/net/ethernet/intel/ice/ice_main.c
> +++ b/drivers/net/ethernet/intel/ice/ice_main.c
> @@ -2986,6 +2986,20 @@ static int ice_max_xdp_frame_size(struct ice_vsi *vsi)
>  		return ICE_RXBUF_3072;
>  }
>  
> +/**
> + * ice_rebuild_pending - ice_vsi_rebuild will be performed, when locks are released
> + * @vsi: VSI to setup XDP for
> + *
> + * ice_vsi_close() in the reset path is called under rtnl_lock(),
> + * so it happened strictly before or after .ndo_bpf().
> + * In case it has happened before, we do not have anything attached to rings
> + */
> +bool ice_rebuild_pending(struct ice_vsi *vsi)
> +{
> +	return ice_is_reset_in_progress(vsi->back->state) &&
> +	       !vsi->rx_rings[0]->desc;
> +}
> +
>  /**
>   * ice_xdp_setup_prog - Add or remove XDP eBPF program
>   * @vsi: VSI to setup XDP for
> @@ -3009,7 +3023,7 @@ ice_xdp_setup_prog(struct ice_vsi *vsi, struct bpf_prog *prog,
>  	}
>  
>  	/* hot swap progs and avoid toggling link */
> -	if (ice_is_xdp_ena_vsi(vsi) == !!prog) {
> +	if (ice_is_xdp_ena_vsi(vsi) == !!prog || ice_rebuild_pending(vsi)) {
>  		ice_vsi_assign_bpf_prog(vsi, prog);
>  		return 0;
>  	}
> @@ -3081,21 +3095,33 @@ static int ice_xdp(struct net_device *dev, struct netdev_bpf *xdp)
>  {
>  	struct ice_netdev_priv *np = netdev_priv(dev);
>  	struct ice_vsi *vsi = np->vsi;
> +	struct ice_pf *pf = vsi->back;
> +	int ret;
>  
>  	if (vsi->type != ICE_VSI_PF) {
>  		NL_SET_ERR_MSG_MOD(xdp->extack, "XDP can be loaded only on PF VSI");
>  		return -EINVAL;
>  	}
>  
> +	while (test_and_set_bit(ICE_CFG_BUSY, pf->state)) {
> +		set_bit(ICE_RTNL_WAITS_FOR_RESET, pf->state);
> +		usleep_range(1000, 2000);
> +	}
> +	clear_bit(ICE_RTNL_WAITS_FOR_RESET, pf->state);
> +
>  	switch (xdp->command) {
>  	case XDP_SETUP_PROG:
> -		return ice_xdp_setup_prog(vsi, xdp->prog, xdp->extack);
> +		ret = ice_xdp_setup_prog(vsi, xdp->prog, xdp->extack);
> +		break;
>  	case XDP_SETUP_XSK_POOL:
> -		return ice_xsk_pool_setup(vsi, xdp->xsk.pool,
> -					  xdp->xsk.queue_id);
> +		ret = ice_xsk_pool_setup(vsi, xdp->xsk.pool, xdp->xsk.queue_id);
> +		break;
>  	default:
> -		return -EINVAL;
> +		ret = -EINVAL;
>  	}
> +
> +	clear_bit(ICE_CFG_BUSY, pf->state);
> +	return ret;
>  }
>  
>  /**
> @@ -7672,7 +7698,10 @@ static void ice_rebuild(struct ice_pf *pf, enum ice_reset_req reset_type)
>  		ice_gnss_init(pf);
>  
>  	/* rebuild PF VSI */
> +	while (test_and_set_bit(ICE_CFG_BUSY, pf->state))
> +		usleep_range(1000, 2000);
>  	err = ice_vsi_rebuild_by_type(pf, ICE_VSI_PF);
> +	clear_bit(ICE_CFG_BUSY, pf->state);
>  	if (err) {
>  		dev_err(dev, "PF VSI rebuild failed: %d\n", err);
>  		goto err_vsi_rebuild;
> diff --git a/drivers/net/ethernet/intel/ice/ice_xsk.c b/drivers/net/ethernet/intel/ice/ice_xsk.c
> index 225d027d3d7a..962af14f9fd5 100644
> --- a/drivers/net/ethernet/intel/ice/ice_xsk.c
> +++ b/drivers/net/ethernet/intel/ice/ice_xsk.c
> @@ -370,7 +370,6 @@ int ice_xsk_pool_setup(struct ice_vsi *vsi, struct xsk_buff_pool *pool, u16 qid)
>  {
>  	bool if_running, pool_present = !!pool;
>  	int ret = 0, pool_failure = 0;
> -	struct ice_pf *pf = vsi->back;
>  
>  	if (qid >= vsi->num_rxq || qid >= vsi->num_txq) {
>  		netdev_err(vsi->netdev, "Please use queue id in scope of combined queues count\n");
> @@ -378,18 +377,11 @@ int ice_xsk_pool_setup(struct ice_vsi *vsi, struct xsk_buff_pool *pool, u16 qid)
>  		goto failure;
>  	}
>  
> -	if_running = netif_running(vsi->netdev) && ice_is_xdp_ena_vsi(vsi);
> +	if_running = !ice_rebuild_pending(vsi) &&
> +		     (netif_running(vsi->netdev) && ice_is_xdp_ena_vsi(vsi));
>  
>  	if (if_running) {
>  		struct ice_rx_ring *rx_ring = vsi->rx_rings[qid];
> -		int timeout = 50;
> -
> -		while (test_and_set_bit(ICE_CFG_BUSY, pf->state)) {
> -			timeout--;
> -			if (!timeout)
> -				return -EBUSY;
> -			usleep_range(1000, 2000);
> -		}
>  
>  		ret = ice_qp_dis(vsi, qid);
>  		if (ret) {
> @@ -412,7 +404,6 @@ int ice_xsk_pool_setup(struct ice_vsi *vsi, struct xsk_buff_pool *pool, u16 qid)
>  			napi_schedule(&vsi->rx_rings[qid]->xdp_ring->q_vector->napi);
>  		else if (ret)
>  			netdev_err(vsi->netdev, "ice_qp_ena error = %d\n", ret);
> -		clear_bit(ICE_CFG_BUSY, pf->state);
>  	}
>  
>  failure:
> -- 
> 2.34.1
>