aboutsummaryrefslogtreecommitdiff
path: root/drivers
diff options
context:
space:
mode:
authorDavid S. Miller2022-03-04 09:45:51 +0000
committerDavid S. Miller2022-03-04 09:45:51 +0000
commit4ee508ff78c83c4bf855148f026315fa58c7baf4 (patch)
tree04ae015988dbef78cebc6807495ad4339e84ecff /drivers
parentf2ecfa06afc692f85e749a219e116acdc6501080 (diff)
parent3d5985a185e6abfc0b38ed187819016a79eca864 (diff)
Merge branch '100GbE' of git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/next-queue
Tony Nguyen says: ==================== 100GbE Intel Wired LAN Driver Updates 2022-03-03 Jacob Keller says: This series refactors the ice networking driver VF storage from a simple static array to a hash table. It also introduces krefs and proper locking and protection to prevent common use-after-free and concurrency issues. There are two motivations for this work. First is to make the ice driver more resilient by preventing a whole class of use-after-free bugs that can occur around concurrent access to VF structures while removing VFs. The second is to prepare the ice driver for future virtualization work to support Scalable IOV, an alternative VF implementation compared to Single Root IOV. The new VF implementation will allow for more dynamic VF creation and removal, necessitating a more robust implementation for VF storage that can't rely on the existing mechanisms to prevent concurrent access violations. The first few patches are cleanup and preparatory work needed to make the conversion to the hash table safe. Following this preparatory work is a patch to migrate the VF structures and variables to a new sub-structure for code clarity. Next introduce new interface functions to abstract the VF storage. Finally, the driver is actually converted to the hash table and kref implementation. ==================== Signed-off-by: David S. Miller <davem@davemloft.net>
Diffstat (limited to 'drivers')
-rw-r--r--drivers/net/ethernet/intel/ice/ice.h13
-rw-r--r--drivers/net/ethernet/intel/ice/ice_base.c4
-rw-r--r--drivers/net/ethernet/intel/ice/ice_eswitch.c161
-rw-r--r--drivers/net/ethernet/intel/ice/ice_ethtool.c20
-rw-r--r--drivers/net/ethernet/intel/ice/ice_lib.c203
-rw-r--r--drivers/net/ethernet/intel/ice/ice_lib.h3
-rw-r--r--drivers/net/ethernet/intel/ice/ice_main.c64
-rw-r--r--drivers/net/ethernet/intel/ice/ice_repr.c54
-rw-r--r--drivers/net/ethernet/intel/ice/ice_txrx.c2
-rw-r--r--drivers/net/ethernet/intel/ice/ice_vf_vsi_vlan_ops.c19
-rw-r--r--drivers/net/ethernet/intel/ice/ice_virtchnl_fdir.c13
-rw-r--r--drivers/net/ethernet/intel/ice/ice_virtchnl_pf.c787
-rw-r--r--drivers/net/ethernet/intel/ice/ice_virtchnl_pf.h83
13 files changed, 872 insertions, 554 deletions
diff --git a/drivers/net/ethernet/intel/ice/ice.h b/drivers/net/ethernet/intel/ice/ice.h
index 1ca1273a7cd3..dc42ff92dbad 100644
--- a/drivers/net/ethernet/intel/ice/ice.h
+++ b/drivers/net/ethernet/intel/ice/ice.h
@@ -109,7 +109,6 @@
/* All VF control VSIs share the same IRQ, so assign a unique ID for them */
#define ICE_RES_VF_CTRL_VEC_ID (ICE_RES_RDMA_VEC_ID - 1)
#define ICE_INVAL_Q_INDEX 0xffff
-#define ICE_INVAL_VFID 256
#define ICE_MAX_RXQS_PER_TC 256 /* Used when setting VSI context per TC Rx queues */
@@ -333,7 +332,7 @@ struct ice_vsi {
u16 vsi_num; /* HW (absolute) index of this VSI */
u16 idx; /* software index in pf->vsi[] */
- s16 vf_id; /* VF ID for SR-IOV VSIs */
+ struct ice_vf *vf; /* VF associated with this VSI */
u16 ethtype; /* Ethernet protocol for pause frame */
u16 num_gfltr;
@@ -529,15 +528,7 @@ struct ice_pf {
struct ice_vsi **vsi; /* VSIs created by the driver */
struct ice_sw *first_sw; /* first switch created by firmware */
u16 eswitch_mode; /* current mode of eswitch */
- /* Virtchnl/SR-IOV config info */
- struct ice_vf *vf;
- u16 num_alloc_vfs; /* actual number of VFs allocated */
- u16 num_vfs_supported; /* num VFs supported for this PF */
- u16 num_qps_per_vf;
- u16 num_msix_per_vf;
- /* used to ratelimit the MDD event logging */
- unsigned long last_printed_mdd_jiffies;
- DECLARE_BITMAP(malvfs, ICE_MAX_VF_COUNT);
+ struct ice_vfs vfs;
DECLARE_BITMAP(features, ICE_F_MAX);
DECLARE_BITMAP(state, ICE_STATE_NBITS);
DECLARE_BITMAP(flags, ICE_PF_FLAGS_NBITS);
diff --git a/drivers/net/ethernet/intel/ice/ice_base.c b/drivers/net/ethernet/intel/ice/ice_base.c
index 2360e6abdb1e..a3094470d31d 100644
--- a/drivers/net/ethernet/intel/ice/ice_base.c
+++ b/drivers/net/ethernet/intel/ice/ice_base.c
@@ -323,7 +323,7 @@ ice_setup_tx_ctx(struct ice_tx_ring *ring, struct ice_tlan_ctx *tlan_ctx, u16 pf
break;
case ICE_VSI_VF:
/* Firmware expects vmvf_num to be absolute VF ID */
- tlan_ctx->vmvf_num = hw->func_caps.vf_base_id + vsi->vf_id;
+ tlan_ctx->vmvf_num = hw->func_caps.vf_base_id + vsi->vf->vf_id;
tlan_ctx->vmvf_type = ICE_TLAN_CTX_VMVF_TYPE_VF;
break;
case ICE_VSI_SWITCHDEV_CTRL:
@@ -429,7 +429,7 @@ static int ice_setup_rx_ctx(struct ice_rx_ring *ring)
*/
if (ice_is_dvm_ena(hw))
if (vsi->type == ICE_VSI_VF &&
- ice_vf_is_port_vlan_ena(&vsi->back->vf[vsi->vf_id]))
+ ice_vf_is_port_vlan_ena(vsi->vf))
rlan_ctx.l2tsel = 1;
else
rlan_ctx.l2tsel = 0;
diff --git a/drivers/net/ethernet/intel/ice/ice_eswitch.c b/drivers/net/ethernet/intel/ice/ice_eswitch.c
index 95c81fc9ec9f..9a84d746a6c4 100644
--- a/drivers/net/ethernet/intel/ice/ice_eswitch.c
+++ b/drivers/net/ethernet/intel/ice/ice_eswitch.c
@@ -176,10 +176,20 @@ static void ice_eswitch_remap_rings_to_vectors(struct ice_pf *pf)
int q_id;
ice_for_each_txq(vsi, q_id) {
- struct ice_repr *repr = pf->vf[q_id].repr;
- struct ice_q_vector *q_vector = repr->q_vector;
- struct ice_tx_ring *tx_ring = vsi->tx_rings[q_id];
- struct ice_rx_ring *rx_ring = vsi->rx_rings[q_id];
+ struct ice_q_vector *q_vector;
+ struct ice_tx_ring *tx_ring;
+ struct ice_rx_ring *rx_ring;
+ struct ice_repr *repr;
+ struct ice_vf *vf;
+
+ vf = ice_get_vf_by_id(pf, q_id);
+ if (WARN_ON(!vf))
+ continue;
+
+ repr = vf->repr;
+ q_vector = repr->q_vector;
+ tx_ring = vsi->tx_rings[q_id];
+ rx_ring = vsi->rx_rings[q_id];
q_vector->vsi = vsi;
q_vector->reg_idx = vsi->q_vectors[0]->reg_idx;
@@ -199,6 +209,38 @@ static void ice_eswitch_remap_rings_to_vectors(struct ice_pf *pf)
rx_ring->q_vector = q_vector;
rx_ring->next = NULL;
rx_ring->netdev = repr->netdev;
+
+ ice_put_vf(vf);
+ }
+}
+
+/**
+ * ice_eswitch_release_reprs - clear PR VSIs configuration
+ * @pf: poiner to PF struct
+ * @ctrl_vsi: pointer to switchdev control VSI
+ */
+static void
+ice_eswitch_release_reprs(struct ice_pf *pf, struct ice_vsi *ctrl_vsi)
+{
+ struct ice_vf *vf;
+ unsigned int bkt;
+
+ lockdep_assert_held(&pf->vfs.table_lock);
+
+ ice_for_each_vf(pf, bkt, vf) {
+ struct ice_vsi *vsi = vf->repr->src_vsi;
+
+ /* Skip VFs that aren't configured */
+ if (!vf->repr->dst)
+ continue;
+
+ ice_vsi_update_security(vsi, ice_vsi_ctx_set_antispoof);
+ metadata_dst_free(vf->repr->dst);
+ vf->repr->dst = NULL;
+ ice_fltr_add_mac_and_broadcast(vsi, vf->hw_lan_addr.addr,
+ ICE_FWD_TO_VSI);
+
+ netif_napi_del(&vf->repr->q_vector->napi);
}
}
@@ -210,11 +252,13 @@ static int ice_eswitch_setup_reprs(struct ice_pf *pf)
{
struct ice_vsi *ctrl_vsi = pf->switchdev.control_vsi;
int max_vsi_num = 0;
- int i;
+ struct ice_vf *vf;
+ unsigned int bkt;
+
+ lockdep_assert_held(&pf->vfs.table_lock);
- ice_for_each_vf(pf, i) {
- struct ice_vsi *vsi = pf->vf[i].repr->src_vsi;
- struct ice_vf *vf = &pf->vf[i];
+ ice_for_each_vf(pf, bkt, vf) {
+ struct ice_vsi *vsi = vf->repr->src_vsi;
ice_remove_vsi_fltr(&pf->hw, vsi->idx);
vf->repr->dst = metadata_dst_alloc(0, METADATA_HW_PORT_MUX,
@@ -231,6 +275,7 @@ static int ice_eswitch_setup_reprs(struct ice_pf *pf)
vf->hw_lan_addr.addr,
ICE_FWD_TO_VSI);
metadata_dst_free(vf->repr->dst);
+ vf->repr->dst = NULL;
goto err;
}
@@ -239,6 +284,7 @@ static int ice_eswitch_setup_reprs(struct ice_pf *pf)
vf->hw_lan_addr.addr,
ICE_FWD_TO_VSI);
metadata_dst_free(vf->repr->dst);
+ vf->repr->dst = NULL;
ice_vsi_update_security(vsi, ice_vsi_ctx_set_antispoof);
goto err;
}
@@ -252,8 +298,8 @@ static int ice_eswitch_setup_reprs(struct ice_pf *pf)
netif_keep_dst(vf->repr->netdev);
}
- ice_for_each_vf(pf, i) {
- struct ice_repr *repr = pf->vf[i].repr;
+ ice_for_each_vf(pf, bkt, vf) {
+ struct ice_repr *repr = vf->repr;
struct ice_vsi *vsi = repr->src_vsi;
struct metadata_dst *dst;
@@ -266,43 +312,12 @@ static int ice_eswitch_setup_reprs(struct ice_pf *pf)
return 0;
err:
- for (i = i - 1; i >= 0; i--) {
- struct ice_vsi *vsi = pf->vf[i].repr->src_vsi;
- struct ice_vf *vf = &pf->vf[i];
-
- ice_vsi_update_security(vsi, ice_vsi_ctx_set_antispoof);
- metadata_dst_free(vf->repr->dst);
- ice_fltr_add_mac_and_broadcast(vsi, vf->hw_lan_addr.addr,
- ICE_FWD_TO_VSI);
- }
+ ice_eswitch_release_reprs(pf, ctrl_vsi);
return -ENODEV;
}
/**
- * ice_eswitch_release_reprs - clear PR VSIs configuration
- * @pf: poiner to PF struct
- * @ctrl_vsi: pointer to switchdev control VSI
- */
-static void
-ice_eswitch_release_reprs(struct ice_pf *pf, struct ice_vsi *ctrl_vsi)
-{
- int i;
-
- ice_for_each_vf(pf, i) {
- struct ice_vsi *vsi = pf->vf[i].repr->src_vsi;
- struct ice_vf *vf = &pf->vf[i];
-
- ice_vsi_update_security(vsi, ice_vsi_ctx_set_antispoof);
- metadata_dst_free(vf->repr->dst);
- ice_fltr_add_mac_and_broadcast(vsi, vf->hw_lan_addr.addr,
- ICE_FWD_TO_VSI);
-
- netif_napi_del(&vf->repr->q_vector->napi);
- }
-}
-
-/**
* ice_eswitch_update_repr - reconfigure VF port representor
* @vsi: VF VSI for which port representor is configured
*/
@@ -316,7 +331,7 @@ void ice_eswitch_update_repr(struct ice_vsi *vsi)
if (!ice_is_switchdev_running(pf))
return;
- vf = &pf->vf[vsi->vf_id];
+ vf = vsi->vf;
repr = vf->repr;
repr->src_vsi = vsi;
repr->dst->u.port_info.port_id = vsi->vsi_num;
@@ -324,7 +339,8 @@ void ice_eswitch_update_repr(struct ice_vsi *vsi)
ret = ice_vsi_update_security(vsi, ice_vsi_ctx_clear_antispoof);
if (ret) {
ice_fltr_add_mac_and_broadcast(vsi, vf->hw_lan_addr.addr, ICE_FWD_TO_VSI);
- dev_err(ice_pf_to_dev(pf), "Failed to update VF %d port representor", vsi->vf_id);
+ dev_err(ice_pf_to_dev(pf), "Failed to update VF %d port representor",
+ vsi->vf->vf_id);
}
}
@@ -408,7 +424,7 @@ static void ice_eswitch_release_env(struct ice_pf *pf)
static struct ice_vsi *
ice_eswitch_vsi_setup(struct ice_pf *pf, struct ice_port_info *pi)
{
- return ice_vsi_setup(pf, pi, ICE_VSI_SWITCHDEV_CTRL, ICE_INVAL_VFID, NULL);
+ return ice_vsi_setup(pf, pi, ICE_VSI_SWITCHDEV_CTRL, NULL, NULL);
}
/**
@@ -417,10 +433,13 @@ ice_eswitch_vsi_setup(struct ice_pf *pf, struct ice_port_info *pi)
*/
static void ice_eswitch_napi_del(struct ice_pf *pf)
{
- int i;
+ struct ice_vf *vf;
+ unsigned int bkt;
- ice_for_each_vf(pf, i)
- netif_napi_del(&pf->vf[i].repr->q_vector->napi);
+ lockdep_assert_held(&pf->vfs.table_lock);
+
+ ice_for_each_vf(pf, bkt, vf)
+ netif_napi_del(&vf->repr->q_vector->napi);
}
/**
@@ -429,10 +448,13 @@ static void ice_eswitch_napi_del(struct ice_pf *pf)
*/
static void ice_eswitch_napi_enable(struct ice_pf *pf)
{
- int i;
+ struct ice_vf *vf;
+ unsigned int bkt;
+
+ lockdep_assert_held(&pf->vfs.table_lock);
- ice_for_each_vf(pf, i)
- napi_enable(&pf->vf[i].repr->q_vector->napi);
+ ice_for_each_vf(pf, bkt, vf)
+ napi_enable(&vf->repr->q_vector->napi);
}
/**
@@ -441,10 +463,13 @@ static void ice_eswitch_napi_enable(struct ice_pf *pf)
*/
static void ice_eswitch_napi_disable(struct ice_pf *pf)
{
- int i;
+ struct ice_vf *vf;
+ unsigned int bkt;
+
+ lockdep_assert_held(&pf->vfs.table_lock);
- ice_for_each_vf(pf, i)
- napi_disable(&pf->vf[i].repr->q_vector->napi);
+ ice_for_each_vf(pf, bkt, vf)
+ napi_disable(&vf->repr->q_vector->napi);
}
/**
@@ -522,7 +547,7 @@ ice_eswitch_mode_set(struct devlink *devlink, u16 mode,
if (pf->eswitch_mode == mode)
return 0;
- if (pf->num_alloc_vfs) {
+ if (ice_has_vfs(pf)) {
dev_info(ice_pf_to_dev(pf), "Changing eswitch mode is allowed only if there is no VFs created");
NL_SET_ERR_MSG_MOD(extack, "Changing eswitch mode is allowed only if there is no VFs created");
return -EOPNOTSUPP;
@@ -613,16 +638,17 @@ int ice_eswitch_configure(struct ice_pf *pf)
*/
static void ice_eswitch_start_all_tx_queues(struct ice_pf *pf)
{
- struct ice_repr *repr;
- int i;
+ struct ice_vf *vf;
+ unsigned int bkt;
+
+ lockdep_assert_held(&pf->vfs.table_lock);
if (test_bit(ICE_DOWN, pf->state))
return;
- ice_for_each_vf(pf, i) {
- repr = pf->vf[i].repr;
- if (repr)
- ice_repr_start_tx_queues(repr);
+ ice_for_each_vf(pf, bkt, vf) {
+ if (vf->repr)
+ ice_repr_start_tx_queues(vf->repr);
}
}
@@ -632,16 +658,17 @@ static void ice_eswitch_start_all_tx_queues(struct ice_pf *pf)
*/
void ice_eswitch_stop_all_tx_queues(struct ice_pf *pf)
{
- struct ice_repr *repr;
- int i;
+ struct ice_vf *vf;
+ unsigned int bkt;
+
+ lockdep_assert_held(&pf->vfs.table_lock);
if (test_bit(ICE_DOWN, pf->state))
return;
- ice_for_each_vf(pf, i) {
- repr = pf->vf[i].repr;
- if (repr)
- ice_repr_stop_tx_queues(repr);
+ ice_for_each_vf(pf, bkt, vf) {
+ if (vf->repr)
+ ice_repr_stop_tx_queues(vf->repr);
}
}
diff --git a/drivers/net/ethernet/intel/ice/ice_ethtool.c b/drivers/net/ethernet/intel/ice/ice_ethtool.c
index a3492754d0d3..399625892f9e 100644
--- a/drivers/net/ethernet/intel/ice/ice_ethtool.c
+++ b/drivers/net/ethernet/intel/ice/ice_ethtool.c
@@ -316,16 +316,20 @@ out:
*/
static bool ice_active_vfs(struct ice_pf *pf)
{
- unsigned int i;
-
- ice_for_each_vf(pf, i) {
- struct ice_vf *vf = &pf->vf[i];
+ bool active = false;
+ struct ice_vf *vf;
+ unsigned int bkt;
- if (test_bit(ICE_VF_STATE_ACTIVE, vf->vf_states))
- return true;
+ rcu_read_lock();
+ ice_for_each_vf_rcu(pf, bkt, vf) {
+ if (test_bit(ICE_VF_STATE_ACTIVE, vf->vf_states)) {
+ active = true;
+ break;
+ }
}
+ rcu_read_unlock();
- return false;
+ return active;
}
/**
@@ -1298,7 +1302,7 @@ static int ice_set_priv_flags(struct net_device *netdev, u32 flags)
}
if (test_bit(ICE_FLAG_VF_VLAN_PRUNING, change_flags) &&
- pf->num_alloc_vfs) {
+ ice_has_vfs(pf)) {
dev_err(dev, "vf-vlan-pruning: VLAN pruning cannot be changed while VFs are active.\n");
/* toggle bit back to previous state */
change_bit(ICE_FLAG_VF_VLAN_PRUNING, pf->flags);
diff --git a/drivers/net/ethernet/intel/ice/ice_lib.c b/drivers/net/ethernet/intel/ice/ice_lib.c
index 662493d1002b..113a2c56c14c 100644
--- a/drivers/net/ethernet/intel/ice/ice_lib.c
+++ b/drivers/net/ethernet/intel/ice/ice_lib.c
@@ -166,21 +166,19 @@ static void ice_vsi_set_num_desc(struct ice_vsi *vsi)
/**
* ice_vsi_set_num_qs - Set number of queues, descriptors and vectors for a VSI
* @vsi: the VSI being configured
- * @vf_id: ID of the VF being configured
+ * @vf: the VF associated with this VSI, if any
*
* Return 0 on success and a negative value on error
*/
-static void ice_vsi_set_num_qs(struct ice_vsi *vsi, u16 vf_id)
+static void ice_vsi_set_num_qs(struct ice_vsi *vsi, struct ice_vf *vf)
{
+ enum ice_vsi_type vsi_type = vsi->type;
struct ice_pf *pf = vsi->back;
- struct ice_vf *vf = NULL;
- if (vsi->type == ICE_VSI_VF)
- vsi->vf_id = vf_id;
- else
- vsi->vf_id = ICE_INVAL_VFID;
+ if (WARN_ON(vsi_type == ICE_VSI_VF && !vf))
+ return;
- switch (vsi->type) {
+ switch (vsi_type) {
case ICE_VSI_PF:
if (vsi->req_txq) {
vsi->alloc_txq = vsi->req_txq;
@@ -217,22 +215,21 @@ static void ice_vsi_set_num_qs(struct ice_vsi *vsi, u16 vf_id)
/* The number of queues for ctrl VSI is equal to number of VFs.
* Each ring is associated to the corresponding VF_PR netdev.
*/
- vsi->alloc_txq = pf->num_alloc_vfs;
- vsi->alloc_rxq = pf->num_alloc_vfs;
+ vsi->alloc_txq = ice_get_num_vfs(pf);
+ vsi->alloc_rxq = vsi->alloc_txq;
vsi->num_q_vectors = 1;
break;
case ICE_VSI_VF:
- vf = &pf->vf[vsi->vf_id];
if (vf->num_req_qs)
vf->num_vf_qs = vf->num_req_qs;
vsi->alloc_txq = vf->num_vf_qs;
vsi->alloc_rxq = vf->num_vf_qs;
- /* pf->num_msix_per_vf includes (VF miscellaneous vector +
+ /* pf->vfs.num_msix_per includes (VF miscellaneous vector +
* data queue interrupts). Since vsi->num_q_vectors is number
* of queues vectors, subtract 1 (ICE_NONQ_VECS_VF) from the
* original vector count
*/
- vsi->num_q_vectors = pf->num_msix_per_vf - ICE_NONQ_VECS_VF;
+ vsi->num_q_vectors = pf->vfs.num_msix_per - ICE_NONQ_VECS_VF;
break;
case ICE_VSI_CTRL:
vsi->alloc_txq = 1;
@@ -248,7 +245,7 @@ static void ice_vsi_set_num_qs(struct ice_vsi *vsi, u16 vf_id)
vsi->alloc_rxq = 1;
break;
default:
- dev_warn(ice_pf_to_dev(pf), "Unknown VSI type %d\n", vsi->type);
+ dev_warn(ice_pf_to_dev(pf), "Unknown VSI type %d\n", vsi_type);
break;
}
@@ -299,7 +296,7 @@ void ice_vsi_delete(struct ice_vsi *vsi)
return;
if (vsi->type == ICE_VSI_VF)
- ctxt->vf_num = vsi->vf_id;
+ ctxt->vf_num = vsi->vf->vf_id;
ctxt->vsi_num = vsi->vsi_num;
memcpy(&ctxt->info, &vsi->info, sizeof(ctxt->info));
@@ -384,8 +381,7 @@ int ice_vsi_clear(struct ice_vsi *vsi)
pf->vsi[vsi->idx] = NULL;
if (vsi->idx < pf->next_vsi && vsi->type != ICE_VSI_CTRL)
pf->next_vsi = vsi->idx;
- if (vsi->idx < pf->next_vsi && vsi->type == ICE_VSI_CTRL &&
- vsi->vf_id != ICE_INVAL_VFID)
+ if (vsi->idx < pf->next_vsi && vsi->type == ICE_VSI_CTRL && vsi->vf)
pf->next_vsi = vsi->idx;
ice_vsi_free_arrays(vsi);
@@ -437,13 +433,16 @@ static irqreturn_t ice_eswitch_msix_clean_rings(int __always_unused irq, void *d
{
struct ice_q_vector *q_vector = (struct ice_q_vector *)data;
struct ice_pf *pf = q_vector->vsi->back;
- int i;
+ struct ice_vf *vf;
+ unsigned int bkt;
if (!q_vector->tx.tx_ring && !q_vector->rx.rx_ring)
return IRQ_HANDLED;
- ice_for_each_vf(pf, i)
- napi_schedule(&pf->vf[i].repr->q_vector->napi);
+ rcu_read_lock();
+ ice_for_each_vf_rcu(pf, bkt, vf)
+ napi_schedule(&vf->repr->q_vector->napi);
+ rcu_read_unlock();
return IRQ_HANDLED;
}
@@ -453,17 +452,24 @@ static irqreturn_t ice_eswitch_msix_clean_rings(int __always_unused irq, void *d
* @pf: board private structure
* @vsi_type: type of VSI
* @ch: ptr to channel
- * @vf_id: ID of the VF being configured
+ * @vf: VF for ICE_VSI_VF and ICE_VSI_CTRL
+ *
+ * The VF pointer is used for ICE_VSI_VF and ICE_VSI_CTRL. For ICE_VSI_CTRL,
+ * it may be NULL in the case there is no association with a VF. For
+ * ICE_VSI_VF the VF pointer *must not* be NULL.
*
* returns a pointer to a VSI on success, NULL on failure.
*/
static struct ice_vsi *
ice_vsi_alloc(struct ice_pf *pf, enum ice_vsi_type vsi_type,
- struct ice_channel *ch, u16 vf_id)
+ struct ice_channel *ch, struct ice_vf *vf)
{
struct device *dev = ice_pf_to_dev(pf);
struct ice_vsi *vsi = NULL;
+ if (WARN_ON(vsi_type == ICE_VSI_VF && !vf))
+ return NULL;
+
/* Need to protect the allocation of the VSIs at the PF level */
mutex_lock(&pf->sw_mutex);
@@ -485,9 +491,9 @@ ice_vsi_alloc(struct ice_pf *pf, enum ice_vsi_type vsi_type,
set_bit(ICE_VSI_DOWN, vsi->state);
if (vsi_type == ICE_VSI_VF)
- ice_vsi_set_num_qs(vsi, vf_id);
+ ice_vsi_set_num_qs(vsi, vf);
else if (vsi_type != ICE_VSI_CHNL)
- ice_vsi_set_num_qs(vsi, ICE_INVAL_VFID);
+ ice_vsi_set_num_qs(vsi, NULL);
switch (vsi->type) {
case ICE_VSI_SWITCHDEV_CTRL:
@@ -510,10 +516,16 @@ ice_vsi_alloc(struct ice_pf *pf, enum ice_vsi_type vsi_type,
/* Setup ctrl VSI MSIX irq handler */
vsi->irq_handler = ice_msix_clean_ctrl_vsi;
+
+ /* For the PF control VSI this is NULL, for the VF control VSI
+ * this will be the first VF to allocate it.
+ */
+ vsi->vf = vf;
break;
case ICE_VSI_VF:
if (ice_vsi_alloc_arrays(vsi))
goto err_rings;
+ vsi->vf = vf;
break;
case ICE_VSI_CHNL:
if (!ch)
@@ -531,7 +543,7 @@ ice_vsi_alloc(struct ice_pf *pf, enum ice_vsi_type vsi_type,
goto unlock_pf;
}
- if (vsi->type == ICE_VSI_CTRL && vf_id == ICE_INVAL_VFID) {
+ if (vsi->type == ICE_VSI_CTRL && !vf) {
/* Use the last VSI slot as the index for PF control VSI */
vsi->idx = pf->num_alloc_vsi - 1;
pf->ctrl_vsi_idx = vsi->idx;
@@ -546,8 +558,8 @@ ice_vsi_alloc(struct ice_pf *pf, enum ice_vsi_type vsi_type,
pf->next_vsi);
}
- if (vsi->type == ICE_VSI_CTRL && vf_id != ICE_INVAL_VFID)
- pf->vf[vf_id].ctrl_vsi_idx = vsi->idx;
+ if (vsi->type == ICE_VSI_CTRL && vf)
+ vf->ctrl_vsi_idx = vsi->idx;
goto unlock_pf;
err_rings:
@@ -1130,7 +1142,7 @@ static int ice_vsi_init(struct ice_vsi *vsi, bool init_vsi)
case ICE_VSI_VF:
ctxt->flags = ICE_AQ_VSI_TYPE_VF;
/* VF number here is the absolute VF number (0-255) */
- ctxt->vf_num = vsi->vf_id + hw->func_caps.vf_base_id;
+ ctxt->vf_num = vsi->vf->vf_id + hw->func_caps.vf_base_id;
break;
default:
ret = -ENODEV;
@@ -1322,6 +1334,36 @@ ice_get_res(struct ice_pf *pf, struct ice_res_tracker *res, u16 needed, u16 id)
}
/**
+ * ice_get_vf_ctrl_res - Get VF control VSI resource
+ * @pf: pointer to the PF structure
+ * @vsi: the VSI to allocate a resource for
+ *
+ * Look up whether another VF has already allocated the control VSI resource.
+ * If so, re-use this resource so that we share it among all VFs.
+ *
+ * Otherwise, allocate the resource and return it.
+ */
+static int ice_get_vf_ctrl_res(struct ice_pf *pf, struct ice_vsi *vsi)
+{
+ struct ice_vf *vf;
+ unsigned int bkt;
+ int base;
+
+ rcu_read_lock();
+ ice_for_each_vf_rcu(pf, bkt, vf) {
+ if (vf != vsi->vf && vf->ctrl_vsi_idx != ICE_NO_VSI) {
+ base = pf->vsi[vf->ctrl_vsi_idx]->base_vector;
+ rcu_read_unlock();
+ return base;
+ }
+ }
+ rcu_read_unlock();
+
+ return ice_get_res(pf, pf->irq_tracker, vsi->num_q_vectors,
+ ICE_RES_VF_CTRL_VEC_ID);
+}
+
+/**
* ice_vsi_setup_vector_base - Set up the base vector for the given VSI
* @vsi: ptr to the VSI
*
@@ -1353,20 +1395,8 @@ static int ice_vsi_setup_vector_base(struct ice_vsi *vsi)
num_q_vectors = vsi->num_q_vectors;
/* reserve slots from OS requested IRQs */
- if (vsi->type == ICE_VSI_CTRL && vsi->vf_id != ICE_INVAL_VFID) {
- int i;
-
- ice_for_each_vf(pf, i) {
- struct ice_vf *vf = &pf->vf[i];
-
- if (i != vsi->vf_id && vf->ctrl_vsi_idx != ICE_NO_VSI) {
- base = pf->vsi[vf->ctrl_vsi_idx]->base_vector;
- break;
- }
- }
- if (i == pf->num_alloc_vfs)
- base = ice_get_res(pf, pf->irq_tracker, num_q_vectors,
- ICE_RES_VF_CTRL_VEC_ID);
+ if (vsi->type == ICE_VSI_CTRL && vsi->vf) {
+ base = ice_get_vf_ctrl_res(pf, vsi);
} else {
base = ice_get_res(pf, pf->irq_tracker, num_q_vectors,
vsi->idx);
@@ -2218,7 +2248,7 @@ ice_vsi_set_q_vectors_reg_idx(struct ice_vsi *vsi)
}
if (vsi->type == ICE_VSI_VF) {
- struct ice_vf *vf = &vsi->back->vf[vsi->vf_id];
+ struct ice_vf *vf = vsi->vf;
q_vector->reg_idx = ice_calc_vf_reg_idx(vf, q_vector);
} else {
@@ -2403,9 +2433,8 @@ static void ice_set_agg_vsi(struct ice_vsi *vsi)
* @pf: board private structure
* @pi: pointer to the port_info instance
* @vsi_type: VSI type
- * @vf_id: defines VF ID to which this VSI connects. This field is meant to be
- * used only for ICE_VSI_VF VSI type. For other VSI types, should
- * fill-in ICE_INVAL_VFID as input.
+ * @vf: pointer to VF to which this VSI connects. This field is used primarily
+ * for the ICE_VSI_VF type. Other VSI types should pass NULL.
* @ch: ptr to channel
*
* This allocates the sw VSI structure and its queue resources.
@@ -2415,7 +2444,8 @@ static void ice_set_agg_vsi(struct ice_vsi *vsi)
*/
struct ice_vsi *
ice_vsi_setup(struct ice_pf *pf, struct ice_port_info *pi,
- enum ice_vsi_type vsi_type, u16 vf_id, struct ice_channel *ch)
+ enum ice_vsi_type vsi_type, struct ice_vf *vf,
+ struct ice_channel *ch)
{
u16 max_txqs[ICE_MAX_TRAFFIC_CLASS] = { 0 };
struct device *dev = ice_pf_to_dev(pf);
@@ -2423,11 +2453,11 @@ ice_vsi_setup(struct ice_pf *pf, struct ice_port_info *pi,
int ret, i;
if (vsi_type == ICE_VSI_CHNL)
- vsi = ice_vsi_alloc(pf, vsi_type, ch, ICE_INVAL_VFID);
+ vsi = ice_vsi_alloc(pf, vsi_type, ch, NULL);
else if (vsi_type == ICE_VSI_VF || vsi_type == ICE_VSI_CTRL)
- vsi = ice_vsi_alloc(pf, vsi_type, NULL, vf_id);
+ vsi = ice_vsi_alloc(pf, vsi_type, NULL, vf);
else
- vsi = ice_vsi_alloc(pf, vsi_type, NULL, ICE_INVAL_VFID);
+ vsi = ice_vsi_alloc(pf, vsi_type, NULL, NULL);
if (!vsi) {
dev_err(dev, "could not allocate VSI\n");
@@ -2439,9 +2469,6 @@ ice_vsi_setup(struct ice_pf *pf, struct ice_port_info *pi,
if (vsi->type == ICE_VSI_PF)
vsi->ethtype = ETH_P_PAUSE;
- if (vsi->type == ICE_VSI_VF || vsi->type == ICE_VSI_CTRL)
- vsi->vf_id = vf_id;
-
ice_alloc_fd_res(vsi);
if (vsi_type != ICE_VSI_CHNL) {
@@ -2862,6 +2889,37 @@ void ice_napi_del(struct ice_vsi *vsi)
}
/**
+ * ice_free_vf_ctrl_res - Free the VF control VSI resource
+ * @pf: pointer to PF structure
+ * @vsi: the VSI to free resources for
+ *
+ * Check if the VF control VSI resource is still in use. If no VF is using it
+ * any more, release the VSI resource. Otherwise, leave it to be cleaned up
+ * once no other VF uses it.
+ */
+static void ice_free_vf_ctrl_res(struct ice_pf *pf, struct ice_vsi *vsi)
+{
+ struct ice_vf *vf;
+ unsigned int bkt;
+
+ rcu_read_lock();
+ ice_for_each_vf_rcu(pf, bkt, vf) {
+ if (vf != vsi->vf && vf->ctrl_vsi_idx != ICE_NO_VSI) {
+ rcu_read_unlock();
+ return;
+ }
+ }
+ rcu_read_unlock();
+
+ /* No other VFs left that have control VSI. It is now safe to reclaim
+ * SW interrupts back to the common pool.
+ */
+ ice_free_res(pf->irq_tracker, vsi->base_vector,
+ ICE_RES_VF_CTRL_VEC_ID);
+ pf->num_avail_sw_msix += vsi->num_q_vectors;
+}
+
+/**
* ice_vsi_release - Delete a VSI and free its resources
* @vsi: the VSI being removed
*
@@ -2904,23 +2962,8 @@ int ice_vsi_release(struct ice_vsi *vsi)
* many interrupts each VF needs. SR-IOV MSIX resources are also
* cleared in the same manner.
*/
- if (vsi->type == ICE_VSI_CTRL && vsi->vf_id != ICE_INVAL_VFID) {
- int i;
-
- ice_for_each_vf(pf, i) {
- struct ice_vf *vf = &pf->vf[i];
-
- if (i != vsi->vf_id && vf->ctrl_vsi_idx != ICE_NO_VSI)
- break;
- }
- if (i == pf->num_alloc_vfs) {
- /* No other VFs left that have control VSI, reclaim SW
- * interrupts back to the common pool
- */
- ice_free_res(pf->irq_tracker, vsi->base_vector,
- ICE_RES_VF_CTRL_VEC_ID);
- pf->num_avail_sw_msix += vsi->num_q_vectors;
- }
+ if (vsi->type == ICE_VSI_CTRL && vsi->vf) {
+ ice_free_vf_ctrl_res(pf, vsi);
} else if (vsi->type != ICE_VSI_VF) {
/* reclaim SW interrupts back to the common pool */
ice_free_res(pf->irq_tracker, vsi->base_vector, vsi->idx);
@@ -3104,7 +3147,6 @@ int ice_vsi_rebuild(struct ice_vsi *vsi, bool init_vsi)
u16 max_txqs[ICE_MAX_TRAFFIC_CLASS] = { 0 };
struct ice_coalesce_stored *coalesce;
int prev_num_q_vectors = 0;
- struct ice_vf *vf = NULL;
enum ice_vsi_type vtype;
struct ice_pf *pf;
int ret, i;
@@ -3114,8 +3156,8 @@ int ice_vsi_rebuild(struct ice_vsi *vsi, bool init_vsi)
pf = vsi->back;
vtype = vsi->type;
- if (vtype == ICE_VSI_VF)
- vf = &pf->vf[vsi->vf_id];
+ if (WARN_ON(vtype == ICE_VSI_VF) && !vsi->vf)
+ return -EINVAL;
ice_vsi_init_vlan_ops(vsi);
@@ -3154,9 +3196,9 @@ int ice_vsi_rebuild(struct ice_vsi *vsi, bool init_vsi)
ice_vsi_clear_rings(vsi);
ice_vsi_free_arrays(vsi);
if (vtype == ICE_VSI_VF)
- ice_vsi_set_num_qs(vsi, vf->vf_id);
+ ice_vsi_set_num_qs(vsi, vsi->vf);
else
- ice_vsi_set_num_qs(vsi, ICE_INVAL_VFID);
+ ice_vsi_set_num_qs(vsi, NULL);
ret = ice_vsi_alloc_arrays(vsi);
if (ret < 0)
@@ -4013,9 +4055,14 @@ static u16 ice_vsi_num_zero_vlans(struct ice_vsi *vsi)
#define ICE_DVM_NUM_ZERO_VLAN_FLTRS 2
#define ICE_SVM_NUM_ZERO_VLAN_FLTRS 1
/* no VLAN 0 filter is created when a port VLAN is active */
- if (vsi->type == ICE_VSI_VF &&
- ice_vf_is_port_vlan_ena(&vsi->back->vf[vsi->vf_id]))
- return 0;
+ if (vsi->type == ICE_VSI_VF) {
+ if (WARN_ON(!vsi->vf))
+ return 0;
+
+ if (ice_vf_is_port_vlan_ena(vsi->vf))
+ return 0;
+ }
+
if (ice_is_dvm_ena(&vsi->back->hw))
return ICE_DVM_NUM_ZERO_VLAN_FLTRS;
else
diff --git a/drivers/net/ethernet/intel/ice/ice_lib.h b/drivers/net/ethernet/intel/ice/ice_lib.h
index 491f13f98797..0095329949d4 100644
--- a/drivers/net/ethernet/intel/ice/ice_lib.h
+++ b/drivers/net/ethernet/intel/ice/ice_lib.h
@@ -52,7 +52,8 @@ void ice_vsi_cfg_netdev_tc(struct ice_vsi *vsi, u8 ena_tc);
struct ice_vsi *
ice_vsi_setup(struct ice_pf *pf, struct ice_port_info *pi,
- enum ice_vsi_type vsi_type, u16 vf_id, struct ice_channel *ch);
+ enum ice_vsi_type vsi_type, struct ice_vf *vf,
+ struct ice_channel *ch);
void ice_napi_del(struct ice_vsi *vsi);
diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c
index 95dfd04e8c3a..289e5c99e313 100644
--- a/drivers/net/ethernet/intel/ice/ice_main.c
+++ b/drivers/net/ethernet/intel/ice/ice_main.c
@@ -505,7 +505,8 @@ ice_prepare_for_reset(struct ice_pf *pf, enum ice_reset_req reset_type)
{
struct ice_hw *hw = &pf->hw;
struct ice_vsi *vsi;
- unsigned int i;
+ struct ice_vf *vf;
+ unsigned int bkt;
dev_dbg(ice_pf_to_dev(pf), "reset_type=%d\n", reset_type);
@@ -520,8 +521,10 @@ ice_prepare_for_reset(struct ice_pf *pf, enum ice_reset_req reset_type)
ice_vc_notify_reset(pf);
/* Disable VFs until reset is completed */
- ice_for_each_vf(pf, i)
- ice_set_vf_state_qs_dis(&pf->vf[i]);
+ mutex_lock(&pf->vfs.table_lock);
+ ice_for_each_vf(pf, bkt, vf)
+ ice_set_vf_state_qs_dis(vf);
+ mutex_unlock(&pf->vfs.table_lock);
if (ice_is_eswitch_mode_switchdev(pf)) {
if (reset_type != ICE_RESET_PFR)
@@ -1666,7 +1669,8 @@ static void ice_handle_mdd_event(struct ice_pf *pf)
{
struct device *dev = ice_pf_to_dev(pf);
struct ice_hw *hw = &pf->hw;
- unsigned int i;
+ struct ice_vf *vf;
+ unsigned int bkt;
u32 reg;
if (!test_and_clear_bit(ICE_MDD_EVENT_PENDING, pf->state)) {
@@ -1754,47 +1758,46 @@ static void ice_handle_mdd_event(struct ice_pf *pf)
/* Check to see if one of the VFs caused an MDD event, and then
* increment counters and set print pending
*/
- ice_for_each_vf(pf, i) {
- struct ice_vf *vf = &pf->vf[i];
-
- reg = rd32(hw, VP_MDET_TX_PQM(i));
+ mutex_lock(&pf->vfs.table_lock);
+ ice_for_each_vf(pf, bkt, vf) {
+ reg = rd32(hw, VP_MDET_TX_PQM(vf->vf_id));
if (reg & VP_MDET_TX_PQM_VALID_M) {
- wr32(hw, VP_MDET_TX_PQM(i), 0xFFFF);
+ wr32(hw, VP_MDET_TX_PQM(vf->vf_id), 0xFFFF);
vf->mdd_tx_events.count++;
set_bit(ICE_MDD_VF_PRINT_PENDING, pf->state);
if (netif_msg_tx_err(pf))
dev_info(dev, "Malicious Driver Detection event TX_PQM detected on VF %d\n",
- i);
+ vf->vf_id);
}
- reg = rd32(hw, VP_MDET_TX_TCLAN(i));
+ reg = rd32(hw, VP_MDET_TX_TCLAN(vf->vf_id));
if (reg & VP_MDET_TX_TCLAN_VALID_M) {
- wr32(hw, VP_MDET_TX_TCLAN(i), 0xFFFF);
+ wr32(hw, VP_MDET_TX_TCLAN(vf->vf_id), 0xFFFF);
vf->mdd_tx_events.count++;
set_bit(ICE_MDD_VF_PRINT_PENDING, pf->state);
if (netif_msg_tx_err(pf))
dev_info(dev, "Malicious Driver Detection event TX_TCLAN detected on VF %d\n",
- i);
+ vf->vf_id);
}
- reg = rd32(hw, VP_MDET_TX_TDPU(i));
+ reg = rd32(hw, VP_MDET_TX_TDPU(vf->vf_id));
if (reg & VP_MDET_TX_TDPU_VALID_M) {
- wr32(hw, VP_MDET_TX_TDPU(i), 0xFFFF);
+ wr32(hw, VP_MDET_TX_TDPU(vf->vf_id), 0xFFFF);
vf->mdd_tx_events.count++;
set_bit(ICE_MDD_VF_PRINT_PENDING, pf->state);
if (netif_msg_tx_err(pf))
dev_info(dev, "Malicious Driver Detection event TX_TDPU detected on VF %d\n",
- i);
+ vf->vf_id);
}
- reg = rd32(hw, VP_MDET_RX(i));
+ reg = rd32(hw, VP_MDET_RX(vf->vf_id));
if (reg & VP_MDET_RX_VALID_M) {
- wr32(hw, VP_MDET_RX(i), 0xFFFF);
+ wr32(hw, VP_MDET_RX(vf->vf_id), 0xFFFF);
vf->mdd_rx_events.count++;
set_bit(ICE_MDD_VF_PRINT_PENDING, pf->state);
if (netif_msg_rx_err(pf))
dev_info(dev, "Malicious Driver Detection event RX detected on VF %d\n",
- i);
+ vf->vf_id);
/* Since the queue is disabled on VF Rx MDD events, the
* PF can be configured to reset the VF through ethtool
@@ -1805,12 +1808,13 @@ static void ice_handle_mdd_event(struct ice_pf *pf)
* reset, so print the event prior to reset.
*/
ice_print_vf_rx_mdd_event(vf);
- mutex_lock(&pf->vf[i].cfg_lock);
- ice_reset_vf(&pf->vf[i], false);
- mutex_unlock(&pf->vf[i].cfg_lock);
+ mutex_lock(&vf->cfg_lock);
+ ice_reset_vf(vf, false);
+ mutex_unlock(&vf->cfg_lock);
}
}
}
+ mutex_unlock(&pf->vfs.table_lock);
ice_print_vfs_mdd_events(pf);
}
@@ -2439,7 +2443,7 @@ static int ice_vsi_req_irq_msix(struct ice_vsi *vsi, char *basename)
/* skip this unused q_vector */
continue;
}
- if (vsi->type == ICE_VSI_CTRL && vsi->vf_id != ICE_INVAL_VFID)
+ if (vsi->type == ICE_VSI_CTRL && vsi->vf)
err = devm_request_irq(dev, irq_num, vsi->irq_handler,
IRQF_SHARED, q_vector->name,
q_vector);
@@ -3386,14 +3390,14 @@ void ice_fill_rss_lut(u8 *lut, u16 rss_table_size, u16 rss_size)
static struct ice_vsi *
ice_pf_vsi_setup(struct ice_pf *pf, struct ice_port_info *pi)
{
- return ice_vsi_setup(pf, pi, ICE_VSI_PF, ICE_INVAL_VFID, NULL);
+ return ice_vsi_setup(pf, pi, ICE_VSI_PF, NULL, NULL);
}
static struct ice_vsi *
ice_chnl_vsi_setup(struct ice_pf *pf, struct ice_port_info *pi,
struct ice_channel *ch)
{
- return ice_vsi_setup(pf, pi, ICE_VSI_CHNL, ICE_INVAL_VFID, ch);
+ return ice_vsi_setup(pf, pi, ICE_VSI_CHNL, NULL, ch);
}
/**
@@ -3407,7 +3411,7 @@ ice_chnl_vsi_setup(struct ice_pf *pf, struct ice_port_info *pi,
static struct ice_vsi *
ice_ctrl_vsi_setup(struct ice_pf *pf, struct ice_port_info *pi)
{
- return ice_vsi_setup(pf, pi, ICE_VSI_CTRL, ICE_INVAL_VFID, NULL);
+ return ice_vsi_setup(pf, pi, ICE_VSI_CTRL, NULL, NULL);
}
/**
@@ -3421,7 +3425,7 @@ ice_ctrl_vsi_setup(struct ice_pf *pf, struct ice_port_info *pi)
struct ice_vsi *
ice_lb_vsi_setup(struct ice_pf *pf, struct ice_port_info *pi)
{
- return ice_vsi_setup(pf, pi, ICE_VSI_LB, ICE_INVAL_VFID, NULL);
+ return ice_vsi_setup(pf, pi, ICE_VSI_LB, NULL, NULL);
}
/**
@@ -3680,6 +3684,7 @@ static void ice_deinit_pf(struct ice_pf *pf)
mutex_destroy(&pf->sw_mutex);
mutex_destroy(&pf->tc_mutex);
mutex_destroy(&pf->avail_q_mutex);
+ mutex_destroy(&pf->vfs.table_lock);
if (pf->avail_txqs) {
bitmap_free(pf->avail_txqs);
@@ -3712,7 +3717,7 @@ static void ice_set_pf_caps(struct ice_pf *pf)
clear_bit(ICE_FLAG_SRIOV_CAPABLE, pf->flags);
if (func_caps->common_cap.sr_iov_1_1) {
set_bit(ICE_FLAG_SRIOV_CAPABLE, pf->flags);
- pf->num_vfs_supported = min_t(int, func_caps->num_allocd_vfs,
+ pf->vfs.num_supported = min_t(int, func_caps->num_allocd_vfs,
ICE_MAX_VF_COUNT);
}
clear_bit(ICE_FLAG_RSS_ENA, pf->flags);
@@ -3779,6 +3784,9 @@ static int ice_init_pf(struct ice_pf *pf)
return -ENOMEM;
}
+ mutex_init(&pf->vfs.table_lock);
+ hash_init(pf->vfs.table);
+
return 0;
}
diff --git a/drivers/net/ethernet/intel/ice/ice_repr.c b/drivers/net/ethernet/intel/ice/ice_repr.c
index dcc310e29300..2adfaf21e056 100644
--- a/drivers/net/ethernet/intel/ice/ice_repr.c
+++ b/drivers/net/ethernet/intel/ice/ice_repr.c
@@ -284,6 +284,8 @@ static int ice_repr_add(struct ice_vf *vf)
devlink_port_type_eth_set(&vf->devlink_port, repr->netdev);
+ ice_vc_change_ops_to_repr(&vf->vc_ops);
+
return 0;
err_netdev:
@@ -311,6 +313,9 @@ err_alloc_rule:
*/
static void ice_repr_rem(struct ice_vf *vf)
{
+ if (!vf->repr)
+ return;
+
ice_devlink_destroy_vf_port(vf);
kfree(vf->repr->q_vector);
vf->repr->q_vector = NULL;
@@ -323,6 +328,23 @@ static void ice_repr_rem(struct ice_vf *vf)
#endif
kfree(vf->repr);
vf->repr = NULL;
+
+ ice_vc_set_dflt_vf_ops(&vf->vc_ops);
+}
+
+/**
+ * ice_repr_rem_from_all_vfs - remove port representor for all VFs
+ * @pf: pointer to PF structure
+ */
+void ice_repr_rem_from_all_vfs(struct ice_pf *pf)
+{
+ struct ice_vf *vf;
+ unsigned int bkt;
+
+ lockdep_assert_held(&pf->vfs.table_lock);
+
+ ice_for_each_vf(pf, bkt, vf)
+ ice_repr_rem(vf);
}
/**
@@ -331,49 +353,27 @@ static void ice_repr_rem(struct ice_vf *vf)
*/
int ice_repr_add_for_all_vfs(struct ice_pf *pf)
{
+ struct ice_vf *vf;
+ unsigned int bkt;
int err;
- int i;
- ice_for_each_vf(pf, i) {
- struct ice_vf *vf = &pf->vf[i];
+ lockdep_assert_held(&pf->vfs.table_lock);
+ ice_for_each_vf(pf, bkt, vf) {
err = ice_repr_add(vf);
if (err)
goto err;
-
- ice_vc_change_ops_to_repr(&vf->vc_ops);
}
return 0;
err:
- for (i = i - 1; i >= 0; i--) {
- struct ice_vf *vf = &pf->vf[i];
-
- ice_repr_rem(vf);
- ice_vc_set_dflt_vf_ops(&vf->vc_ops);
- }
+ ice_repr_rem_from_all_vfs(pf);
return err;
}
/**
- * ice_repr_rem_from_all_vfs - remove port representor for all VFs
- * @pf: pointer to PF structure
- */
-void ice_repr_rem_from_all_vfs(struct ice_pf *pf)
-{
- int i;
-
- ice_for_each_vf(pf, i) {
- struct ice_vf *vf = &pf->vf[i];
-
- ice_repr_rem(vf);
- ice_vc_set_dflt_vf_ops(&vf->vc_ops);
- }
-}
-
-/**
* ice_repr_start_tx_queues - start Tx queues of port representor
* @repr: pointer to repr structure
*/
diff --git a/drivers/net/ethernet/intel/ice/ice_txrx.c b/drivers/net/ethernet/intel/ice/ice_txrx.c
index ff93ec71aed6..853f57a9589a 100644
--- a/drivers/net/ethernet/intel/ice/ice_txrx.c
+++ b/drivers/net/ethernet/intel/ice/ice_txrx.c
@@ -1165,7 +1165,7 @@ int ice_clean_rx_irq(struct ice_rx_ring *rx_ring, int budget)
struct ice_vsi *ctrl_vsi = rx_ring->vsi;
if (rx_desc->wb.rxdid == FDIR_DESC_RXDID &&
- ctrl_vsi->vf_id != ICE_INVAL_VFID)
+ ctrl_vsi->vf)
ice_vc_fdir_irq_handler(ctrl_vsi, rx_desc);
ice_put_rx_buf(rx_ring, NULL, 0);
cleaned_count++;
diff --git a/drivers/net/ethernet/intel/ice/ice_vf_vsi_vlan_ops.c b/drivers/net/ethernet/intel/ice/ice_vf_vsi_vlan_ops.c
index 39f2d36cabba..b16f946185f2 100644
--- a/drivers/net/ethernet/intel/ice/ice_vf_vsi_vlan_ops.c
+++ b/drivers/net/ethernet/intel/ice/ice_vf_vsi_vlan_ops.c
@@ -34,9 +34,10 @@ void ice_vf_vsi_init_vlan_ops(struct ice_vsi *vsi)
{
struct ice_vsi_vlan_ops *vlan_ops;
struct ice_pf *pf = vsi->back;
- struct ice_vf *vf;
+ struct ice_vf *vf = vsi->vf;
- vf = &pf->vf[vsi->vf_id];
+ if (WARN_ON(!vf))
+ return;
if (ice_is_dvm_ena(&pf->hw)) {
vlan_ops = &vsi->outer_vlan_ops;
@@ -126,9 +127,14 @@ void ice_vf_vsi_init_vlan_ops(struct ice_vsi *vsi)
*/
void ice_vf_vsi_cfg_dvm_legacy_vlan_mode(struct ice_vsi *vsi)
{
- struct ice_vf *vf = &vsi->back->vf[vsi->vf_id];
- struct device *dev = ice_pf_to_dev(vf->pf);
struct ice_vsi_vlan_ops *vlan_ops;
+ struct ice_vf *vf = vsi->vf;
+ struct device *dev;
+
+ if (WARN_ON(!vf))
+ return;
+
+ dev = ice_pf_to_dev(vf->pf);
if (!ice_is_dvm_ena(&vsi->back->hw) || ice_vf_is_port_vlan_ena(vf))
return;
@@ -192,7 +198,10 @@ void ice_vf_vsi_cfg_dvm_legacy_vlan_mode(struct ice_vsi *vsi)
*/
void ice_vf_vsi_cfg_svm_legacy_vlan_mode(struct ice_vsi *vsi)
{
- struct ice_vf *vf = &vsi->back->vf[vsi->vf_id];
+ struct ice_vf *vf = vsi->vf;
+
+ if (WARN_ON(!vf))
+ return;
if (ice_is_dvm_ena(&vsi->back->hw) || ice_vf_is_port_vlan_ena(vf))
return;
diff --git a/drivers/net/ethernet/intel/ice/ice_virtchnl_fdir.c b/drivers/net/ethernet/intel/ice/ice_virtchnl_fdir.c
index d64df81d4893..07989f1d08ef 100644
--- a/drivers/net/ethernet/intel/ice/ice_virtchnl_fdir.c
+++ b/drivers/net/ethernet/intel/ice/ice_virtchnl_fdir.c
@@ -1288,15 +1288,16 @@ ice_vc_fdir_irq_handler(struct ice_vsi *ctrl_vsi,
union ice_32b_rx_flex_desc *rx_desc)
{
struct ice_pf *pf = ctrl_vsi->back;
+ struct ice_vf *vf = ctrl_vsi->vf;
struct ice_vf_fdir_ctx *ctx_done;
struct ice_vf_fdir_ctx *ctx_irq;
struct ice_vf_fdir *fdir;
unsigned long flags;
struct device *dev;
- struct ice_vf *vf;
int ret;
- vf = &pf->vf[ctrl_vsi->vf_id];
+ if (WARN_ON(!vf))
+ return;
fdir = &vf->fdir;
ctx_done = &fdir->ctx_done;
@@ -1571,15 +1572,16 @@ err_exit:
*/
void ice_flush_fdir_ctx(struct ice_pf *pf)
{
- int i;
+ struct ice_vf *vf;
+ unsigned int bkt;
if (!test_and_clear_bit(ICE_FD_VF_FLUSH_CTX, pf->state))
return;
- ice_for_each_vf(pf, i) {
+ mutex_lock(&pf->vfs.table_lock);
+ ice_for_each_vf(pf, bkt, vf) {
struct device *dev = ice_pf_to_dev(pf);
enum virtchnl_fdir_prgm_status status;
- struct ice_vf *vf = &pf->vf[i];
struct ice_vf_fdir_ctx *ctx;
unsigned long flags;
int ret;
@@ -1633,6 +1635,7 @@ err_exit:
ctx->flags &= ~ICE_VF_FDIR_CTX_VALID;
spin_unlock_irqrestore(&vf->fdir.ctx_lock, flags);
}
+ mutex_unlock(&pf->vfs.table_lock);
}
/**
diff --git a/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.c b/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.c
index 455ce5dd0031..4840570c494d 100644
--- a/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.c
+++ b/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.c
@@ -175,18 +175,107 @@ struct ice_vsi *ice_get_vf_vsi(struct ice_vf *vf)
}
/**
- * ice_validate_vf_id - helper to check if VF ID is valid
- * @pf: pointer to the PF structure
- * @vf_id: the ID of the VF to check
+ * ice_get_vf_by_id - Get pointer to VF by ID
+ * @pf: the PF private structure
+ * @vf_id: the VF ID to locate
+ *
+ * Locate and return a pointer to the VF structure associated with a given ID.
+ * Returns NULL if the ID does not have a valid VF structure associated with
+ * it.
+ *
+ * This function takes a reference to the VF, which must be released by
+ * calling ice_put_vf() once the caller is finished accessing the VF structure
+ * returned.
*/
-static int ice_validate_vf_id(struct ice_pf *pf, u16 vf_id)
+struct ice_vf *ice_get_vf_by_id(struct ice_pf *pf, u16 vf_id)
{
- /* vf_id range is only valid for 0-255, and should always be unsigned */
- if (vf_id >= pf->num_alloc_vfs) {
- dev_err(ice_pf_to_dev(pf), "Invalid VF ID: %u\n", vf_id);
- return -EINVAL;
+ struct ice_vf *vf;
+
+ rcu_read_lock();
+ hash_for_each_possible_rcu(pf->vfs.table, vf, entry, vf_id) {
+ if (vf->vf_id == vf_id) {
+ struct ice_vf *found;
+
+ if (kref_get_unless_zero(&vf->refcnt))
+ found = vf;
+ else
+ found = NULL;
+
+ rcu_read_unlock();
+ return found;
+ }
}
- return 0;
+ rcu_read_unlock();
+
+ return NULL;
+}
+
+/**
+ * ice_release_vf - Release VF associated with a refcount
+ * @ref: the kref decremented to zero
+ *
+ * Callback function for kref_put to release a VF once its reference count has
+ * hit zero.
+ */
+static void ice_release_vf(struct kref *ref)
+{
+ struct ice_vf *vf = container_of(ref, struct ice_vf, refcnt);
+
+ mutex_destroy(&vf->cfg_lock);
+
+ kfree_rcu(vf, rcu);
+}
+
+/**
+ * ice_put_vf - Release a reference to a VF
+ * @vf: the VF structure to decrease reference count on
+ *
+ * This must be called after ice_get_vf_by_id() once the reference to the VF
+ * structure is no longer used. Otherwise, the VF structure will never be
+ * freed.
+ */
+void ice_put_vf(struct ice_vf *vf)
+{
+ kref_put(&vf->refcnt, ice_release_vf);
+}
+
+/**
+ * ice_has_vfs - Return true if the PF has any associated VFs
+ * @pf: the PF private structure
+ *
+ * Return whether or not the PF has any allocated VFs.
+ *
+ * Note that this function only guarantees that there are no VFs at the point
+ * of calling it. It does not guarantee that no more VFs will be added.
+ */
+bool ice_has_vfs(struct ice_pf *pf)
+{
+ /* A simple check that the hash table is not empty does not require
+ * the mutex or rcu_read_lock.
+ */
+ return !hash_empty(pf->vfs.table);
+}
+
+/**
+ * ice_get_num_vfs - Get number of allocated VFs
+ * @pf: the PF private structure
+ *
+ * Return the total number of allocated VFs. NOTE: VF IDs are not guaranteed
+ * to be contiguous. Do not assume that a VF ID is guaranteed to be less than
+ * the output of this function.
+ */
+u16 ice_get_num_vfs(struct ice_pf *pf)
+{
+ struct ice_vf *vf;
+ unsigned int bkt;
+ u16 num_vfs = 0;
+
+ rcu_read_lock();
+ ice_for_each_vf_rcu(pf, bkt, vf)
+ num_vfs++;
+ rcu_read_unlock();
+
+ return num_vfs;
}
/**
@@ -205,6 +294,32 @@ static int ice_check_vf_init(struct ice_pf *pf, struct ice_vf *vf)
}
/**
+ * ice_free_vf_entries - Free all VF entries from the hash table
+ * @pf: pointer to the PF structure
+ *
+ * Iterate over the VF hash table, removing and releasing all VF entries.
+ * Called during VF teardown or as cleanup during failed VF initialization.
+ */
+static void ice_free_vf_entries(struct ice_pf *pf)
+{
+ struct ice_vfs *vfs = &pf->vfs;
+ struct hlist_node *tmp;
+ struct ice_vf *vf;
+ unsigned int bkt;
+
+ /* Remove all VFs from the hash table and release their main
+ * reference. Once all references to the VF are dropped, ice_put_vf()
+ * will call ice_release_vf which will remove the VF memory.
+ */
+ lockdep_assert_held(&vfs->table_lock);
+
+ hash_for_each_safe(vfs->table, bkt, tmp, vf, entry) {
+ hash_del_rcu(&vf->entry);
+ ice_put_vf(vf);
+ }
+}
+
+/**
* ice_vc_vf_broadcast - Broadcast a message to all VFs on PF
* @pf: pointer to the PF structure
* @v_opcode: operation code
@@ -217,11 +332,11 @@ ice_vc_vf_broadcast(struct ice_pf *pf, enum virtchnl_ops v_opcode,
enum virtchnl_status_code v_retval, u8 *msg, u16 msglen)
{
struct ice_hw *hw = &pf->hw;
- unsigned int i;
-
- ice_for_each_vf(pf, i) {
- struct ice_vf *vf = &pf->vf[i];
+ struct ice_vf *vf;
+ unsigned int bkt;
+ mutex_lock(&pf->vfs.table_lock);
+ ice_for_each_vf(pf, bkt, vf) {
/* Not all vfs are enabled so skip the ones that are not */
if (!test_bit(ICE_VF_STATE_INIT, vf->vf_states) &&
!test_bit(ICE_VF_STATE_ACTIVE, vf->vf_states))
@@ -233,6 +348,7 @@ ice_vc_vf_broadcast(struct ice_pf *pf, enum virtchnl_ops v_opcode,
ice_aq_send_msg_to_vf(hw, vf->vf_id, v_opcode, v_retval, msg,
msglen, NULL);
}
+ mutex_unlock(&pf->vfs.table_lock);
}
/**
@@ -381,7 +497,7 @@ static void ice_free_vf_res(struct ice_vf *vf)
vf->num_mac = 0;
}
- last_vector_idx = vf->first_vector_idx + pf->num_msix_per_vf - 1;
+ last_vector_idx = vf->first_vector_idx + pf->vfs.num_msix_per - 1;
/* clear VF MDD event information */
memset(&vf->mdd_tx_events, 0, sizeof(vf->mdd_tx_events));
@@ -417,7 +533,7 @@ static void ice_dis_vf_mappings(struct ice_vf *vf)
wr32(hw, VPINT_ALLOC_PCI(vf->vf_id), 0);
first = vf->first_vector_idx;
- last = first + pf->num_msix_per_vf - 1;
+ last = first + pf->vfs.num_msix_per - 1;
for (v = first; v <= last; v++) {
u32 reg;
@@ -499,14 +615,14 @@ static void ice_dis_vf_qs(struct ice_vf *vf)
void ice_free_vfs(struct ice_pf *pf)
{
struct device *dev = ice_pf_to_dev(pf);
+ struct ice_vfs *vfs = &pf->vfs;
struct ice_hw *hw = &pf->hw;
- unsigned int tmp, i;
+ struct ice_vf *vf;
+ unsigned int bkt;
- if (!pf->vf)
+ if (!ice_has_vfs(pf))
return;
- ice_eswitch_release(pf);
-
while (test_and_set_bit(ICE_VF_DIS, pf->state))
usleep_range(1000, 2000);
@@ -519,12 +635,11 @@ void ice_free_vfs(struct ice_pf *pf)
else
dev_warn(dev, "VFs are assigned - not disabling SR-IOV\n");
- tmp = pf->num_alloc_vfs;
- pf->num_qps_per_vf = 0;
- pf->num_alloc_vfs = 0;
- for (i = 0; i < tmp; i++) {
- struct ice_vf *vf = &pf->vf[i];
+ mutex_lock(&vfs->table_lock);
+
+ ice_eswitch_release(pf);
+ ice_for_each_vf(pf, bkt, vf) {
mutex_lock(&vf->cfg_lock);
ice_dis_vf_qs(vf);
@@ -536,42 +651,30 @@ void ice_free_vfs(struct ice_pf *pf)
ice_free_vf_res(vf);
}
- mutex_unlock(&vf->cfg_lock);
+ if (!pci_vfs_assigned(pf->pdev)) {
+ u32 reg_idx, bit_idx;
+
+ reg_idx = (hw->func_caps.vf_base_id + vf->vf_id) / 32;
+ bit_idx = (hw->func_caps.vf_base_id + vf->vf_id) % 32;
+ wr32(hw, GLGEN_VFLRSTAT(reg_idx), BIT(bit_idx));
+ }
- mutex_destroy(&vf->cfg_lock);
+ /* clear malicious info since the VF is getting released */
+ if (ice_mbx_clear_malvf(&hw->mbx_snapshot, pf->vfs.malvfs,
+ ICE_MAX_VF_COUNT, vf->vf_id))
+ dev_dbg(dev, "failed to clear malicious VF state for VF %u\n",
+ vf->vf_id);
+
+ mutex_unlock(&vf->cfg_lock);
}
if (ice_sriov_free_msix_res(pf))
dev_err(dev, "Failed to free MSIX resources used by SR-IOV\n");
- devm_kfree(dev, pf->vf);
- pf->vf = NULL;
-
- /* This check is for when the driver is unloaded while VFs are
- * assigned. Setting the number of VFs to 0 through sysfs is caught
- * before this function ever gets called.
- */
- if (!pci_vfs_assigned(pf->pdev)) {
- unsigned int vf_id;
-
- /* Acknowledge VFLR for all VFs. Without this, VFs will fail to
- * work correctly when SR-IOV gets re-enabled.
- */
- for (vf_id = 0; vf_id < tmp; vf_id++) {
- u32 reg_idx, bit_idx;
-
- reg_idx = (hw->func_caps.vf_base_id + vf_id) / 32;
- bit_idx = (hw->func_caps.vf_base_id + vf_id) % 32;
- wr32(hw, GLGEN_VFLRSTAT(reg_idx), BIT(bit_idx));
- }
- }
+ vfs->num_qps_per = 0;
+ ice_free_vf_entries(pf);
- /* clear malicious info if the VFs are getting released */
- for (i = 0; i < tmp; i++)
- if (ice_mbx_clear_malvf(&hw->mbx_snapshot, pf->malvfs,
- ICE_MAX_VF_COUNT, i))
- dev_dbg(dev, "failed to clear malicious VF state for VF %u\n",
- i);
+ mutex_unlock(&vfs->table_lock);
clear_bit(ICE_VF_DIS, pf->state);
clear_bit(ICE_FLAG_SRIOV_ENA, pf->flags);
@@ -667,7 +770,7 @@ static struct ice_vsi *ice_vf_vsi_setup(struct ice_vf *vf)
struct ice_pf *pf = vf->pf;
struct ice_vsi *vsi;
- vsi = ice_vsi_setup(pf, pi, ICE_VSI_VF, vf->vf_id, NULL);
+ vsi = ice_vsi_setup(pf, pi, ICE_VSI_VF, vf, NULL);
if (!vsi) {
dev_err(ice_pf_to_dev(pf), "Failed to create VF VSI\n");
@@ -694,7 +797,7 @@ struct ice_vsi *ice_vf_ctrl_vsi_setup(struct ice_vf *vf)
struct ice_pf *pf = vf->pf;
struct ice_vsi *vsi;
- vsi = ice_vsi_setup(pf, pi, ICE_VSI_CTRL, vf->vf_id, NULL);
+ vsi = ice_vsi_setup(pf, pi, ICE_VSI_CTRL, vf, NULL);
if (!vsi) {
dev_err(ice_pf_to_dev(pf), "Failed to create VF control VSI\n");
ice_vf_ctrl_invalidate_vsi(vf);
@@ -717,7 +820,7 @@ struct ice_vsi *ice_vf_ctrl_vsi_setup(struct ice_vf *vf)
*/
static int ice_calc_vf_first_vector_idx(struct ice_pf *pf, struct ice_vf *vf)
{
- return pf->sriov_base_vector + vf->vf_id * pf->num_msix_per_vf;
+ return pf->sriov_base_vector + vf->vf_id * pf->vfs.num_msix_per;
}
/**
@@ -974,12 +1077,12 @@ static void ice_ena_vf_msix_mappings(struct ice_vf *vf)
hw = &pf->hw;
pf_based_first_msix = vf->first_vector_idx;
- pf_based_last_msix = (pf_based_first_msix + pf->num_msix_per_vf) - 1;
+ pf_based_last_msix = (pf_based_first_msix + pf->vfs.num_msix_per) - 1;
device_based_first_msix = pf_based_first_msix +
pf->hw.func_caps.common_cap.msix_vector_first_id;
device_based_last_msix =
- (device_based_first_msix + pf->num_msix_per_vf) - 1;
+ (device_based_first_msix + pf->vfs.num_msix_per) - 1;
device_based_vf_id = vf->vf_id + hw->func_caps.vf_base_id;
reg = (((device_based_first_msix << VPINT_ALLOC_FIRST_S) &
@@ -1070,45 +1173,6 @@ static void ice_ena_vf_mappings(struct ice_vf *vf)
}
/**
- * ice_determine_res
- * @pf: pointer to the PF structure
- * @avail_res: available resources in the PF structure
- * @max_res: maximum resources that can be given per VF
- * @min_res: minimum resources that can be given per VF
- *
- * Returns non-zero value if resources (queues/vectors) are available or
- * returns zero if PF cannot accommodate for all num_alloc_vfs.
- */
-static int
-ice_determine_res(struct ice_pf *pf, u16 avail_res, u16 max_res, u16 min_res)
-{
- bool checked_min_res = false;
- int res;
-
- /* start by checking if PF can assign max number of resources for
- * all num_alloc_vfs.
- * if yes, return number per VF
- * If no, divide by 2 and roundup, check again
- * repeat the loop till we reach a point where even minimum resources
- * are not available, in that case return 0
- */
- res = max_res;
- while ((res >= min_res) && !checked_min_res) {
- int num_all_res;
-
- num_all_res = pf->num_alloc_vfs * res;
- if (num_all_res <= avail_res)
- return res;
-
- if (res == min_res)
- checked_min_res = true;
-
- res = DIV_ROUND_UP(res, 2);
- }
- return 0;
-}
-
-/**
* ice_calc_vf_reg_idx - Calculate the VF's register index in the PF space
* @vf: VF to calculate the register index for
* @q_vector: a q_vector associated to the VF
@@ -1123,7 +1187,7 @@ int ice_calc_vf_reg_idx(struct ice_vf *vf, struct ice_q_vector *q_vector)
pf = vf->pf;
/* always add one to account for the OICR being the first MSIX */
- return pf->sriov_base_vector + pf->num_msix_per_vf * vf->vf_id +
+ return pf->sriov_base_vector + pf->vfs.num_msix_per * vf->vf_id +
q_vector->v_idx + 1;
}
@@ -1187,6 +1251,7 @@ static int ice_sriov_set_msix_res(struct ice_pf *pf, u16 num_msix_needed)
/**
* ice_set_per_vf_res - check if vectors and queues are available
* @pf: pointer to the PF structure
+ * @num_vfs: the number of SR-IOV VFs being configured
*
* First, determine HW interrupts from common pool. If we allocate fewer VFs, we
* get more vectors and can enable more queues per VF. Note that this does not
@@ -1205,20 +1270,22 @@ static int ice_sriov_set_msix_res(struct ice_pf *pf, u16 num_msix_needed)
* Lastly, set queue and MSI-X VF variables tracked by the PF so it can be used
* by each VF during VF initialization and reset.
*/
-static int ice_set_per_vf_res(struct ice_pf *pf)
+static int ice_set_per_vf_res(struct ice_pf *pf, u16 num_vfs)
{
int max_valid_res_idx = ice_get_max_valid_res_idx(pf->irq_tracker);
+ u16 num_msix_per_vf, num_txq, num_rxq, avail_qs;
int msix_avail_per_vf, msix_avail_for_sriov;
struct device *dev = ice_pf_to_dev(pf);
- u16 num_msix_per_vf, num_txq, num_rxq;
- if (!pf->num_alloc_vfs || max_valid_res_idx < 0)
+ lockdep_assert_held(&pf->vfs.table_lock);
+
+ if (!num_vfs || max_valid_res_idx < 0)
return -EINVAL;
/* determine MSI-X resources per VF */
msix_avail_for_sriov = pf->hw.func_caps.common_cap.num_msix_vectors -
pf->irq_tracker->num_entries;
- msix_avail_per_vf = msix_avail_for_sriov / pf->num_alloc_vfs;
+ msix_avail_per_vf = msix_avail_for_sriov / num_vfs;
if (msix_avail_per_vf >= ICE_NUM_VF_MSIX_MED) {
num_msix_per_vf = ICE_NUM_VF_MSIX_MED;
} else if (msix_avail_per_vf >= ICE_NUM_VF_MSIX_SMALL) {
@@ -1230,40 +1297,43 @@ static int ice_set_per_vf_res(struct ice_pf *pf)
} else {
dev_err(dev, "Only %d MSI-X interrupts available for SR-IOV. Not enough to support minimum of %d MSI-X interrupts per VF for %d VFs\n",
msix_avail_for_sriov, ICE_MIN_INTR_PER_VF,
- pf->num_alloc_vfs);
+ num_vfs);
return -EIO;
}
- /* determine queue resources per VF */
- num_txq = ice_determine_res(pf, ice_get_avail_txq_count(pf),
- min_t(u16,
- num_msix_per_vf - ICE_NONQ_VECS_VF,
- ICE_MAX_RSS_QS_PER_VF),
- ICE_MIN_QS_PER_VF);
+ num_txq = min_t(u16, num_msix_per_vf - ICE_NONQ_VECS_VF,
+ ICE_MAX_RSS_QS_PER_VF);
+ avail_qs = ice_get_avail_txq_count(pf) / num_vfs;
+ if (!avail_qs)
+ num_txq = 0;
+ else if (num_txq > avail_qs)
+ num_txq = rounddown_pow_of_two(avail_qs);
- num_rxq = ice_determine_res(pf, ice_get_avail_rxq_count(pf),
- min_t(u16,
- num_msix_per_vf - ICE_NONQ_VECS_VF,
- ICE_MAX_RSS_QS_PER_VF),
- ICE_MIN_QS_PER_VF);
+ num_rxq = min_t(u16, num_msix_per_vf - ICE_NONQ_VECS_VF,
+ ICE_MAX_RSS_QS_PER_VF);
+ avail_qs = ice_get_avail_rxq_count(pf) / num_vfs;
+ if (!avail_qs)
+ num_rxq = 0;
+ else if (num_rxq > avail_qs)
+ num_rxq = rounddown_pow_of_two(avail_qs);
- if (!num_txq || !num_rxq) {
+ if (num_txq < ICE_MIN_QS_PER_VF || num_rxq < ICE_MIN_QS_PER_VF) {
dev_err(dev, "Not enough queues to support minimum of %d queue pairs per VF for %d VFs\n",
- ICE_MIN_QS_PER_VF, pf->num_alloc_vfs);
+ ICE_MIN_QS_PER_VF, num_vfs);
return -EIO;
}
- if (ice_sriov_set_msix_res(pf, num_msix_per_vf * pf->num_alloc_vfs)) {
+ if (ice_sriov_set_msix_res(pf, num_msix_per_vf * num_vfs)) {
dev_err(dev, "Unable to set MSI-X resources for %d VFs\n",
- pf->num_alloc_vfs);
+ num_vfs);
return -EINVAL;
}
/* only allow equal Tx/Rx queue count (i.e. queue pairs) */
- pf->num_qps_per_vf = min_t(int, num_txq, num_rxq);
- pf->num_msix_per_vf = num_msix_per_vf;
+ pf->vfs.num_qps_per = min_t(int, num_txq, num_rxq);
+ pf->vfs.num_msix_per = num_msix_per_vf;
dev_info(dev, "Enabling %d VFs with %d vectors and %d queues per VF\n",
- pf->num_alloc_vfs, pf->num_msix_per_vf, pf->num_qps_per_vf);
+ num_vfs, pf->vfs.num_msix_per, pf->vfs.num_qps_per);
return 0;
}
@@ -1510,24 +1580,30 @@ bool ice_reset_all_vfs(struct ice_pf *pf, bool is_vflr)
struct device *dev = ice_pf_to_dev(pf);
struct ice_hw *hw = &pf->hw;
struct ice_vf *vf;
- int v, i;
+ unsigned int bkt;
/* If we don't have any VFs, then there is nothing to reset */
- if (!pf->num_alloc_vfs)
+ if (!ice_has_vfs(pf))
return false;
+ mutex_lock(&pf->vfs.table_lock);
+
/* clear all malicious info if the VFs are getting reset */
- ice_for_each_vf(pf, i)
- if (ice_mbx_clear_malvf(&hw->mbx_snapshot, pf->malvfs, ICE_MAX_VF_COUNT, i))
- dev_dbg(dev, "failed to clear malicious VF state for VF %u\n", i);
+ ice_for_each_vf(pf, bkt, vf)
+ if (ice_mbx_clear_malvf(&hw->mbx_snapshot, pf->vfs.malvfs,
+ ICE_MAX_VF_COUNT, vf->vf_id))
+ dev_dbg(dev, "failed to clear malicious VF state for VF %u\n",
+ vf->vf_id);
/* If VFs have been disabled, there is no need to reset */
- if (test_and_set_bit(ICE_VF_DIS, pf->state))
+ if (test_and_set_bit(ICE_VF_DIS, pf->state)) {
+ mutex_unlock(&pf->vfs.table_lock);
return false;
+ }
/* Begin reset on all VFs at once */
- ice_for_each_vf(pf, v)
- ice_trigger_vf_reset(&pf->vf[v], is_vflr, true);
+ ice_for_each_vf(pf, bkt, vf)
+ ice_trigger_vf_reset(vf, is_vflr, true);
/* HW requires some time to make sure it can flush the FIFO for a VF
* when it resets it. Poll the VPGEN_VFRSTAT register for each VF in
@@ -1535,36 +1611,34 @@ bool ice_reset_all_vfs(struct ice_pf *pf, bool is_vflr)
* the VFs using a simple iterator that increments once that VF has
* finished resetting.
*/
- for (i = 0, v = 0; i < 10 && v < pf->num_alloc_vfs; i++) {
- /* Check each VF in sequence */
- while (v < pf->num_alloc_vfs) {
- u32 reg;
-
- vf = &pf->vf[v];
- reg = rd32(hw, VPGEN_VFRSTAT(vf->vf_id));
- if (!(reg & VPGEN_VFRSTAT_VFRD_M)) {
- /* only delay if the check failed */
- usleep_range(10, 20);
+ ice_for_each_vf(pf, bkt, vf) {
+ bool done = false;
+ unsigned int i;
+ u32 reg;
+
+ for (i = 0; i < 10; i++) {
+ reg = rd32(&pf->hw, VPGEN_VFRSTAT(vf->vf_id));
+ if (reg & VPGEN_VFRSTAT_VFRD_M) {
+ done = true;
break;
}
- /* If the current VF has finished resetting, move on
- * to the next VF in sequence.
+ /* only delay if check failed */
+ usleep_range(10, 20);
+ }
+
+ if (!done) {
+ /* Display a warning if at least one VF didn't manage
+ * to reset in time, but continue on with the
+ * operation.
*/
- v++;
+ dev_warn(dev, "VF %u reset check timeout\n", vf->vf_id);
+ break;
}
}
- /* Display a warning if at least one VF didn't manage to reset in
- * time, but continue on with the operation.
- */
- if (v < pf->num_alloc_vfs)
- dev_warn(dev, "VF reset check timeout\n");
-
/* free VF resources to begin resetting the VSI state */
- ice_for_each_vf(pf, v) {
- vf = &pf->vf[v];
-
+ ice_for_each_vf(pf, bkt, vf) {
mutex_lock(&vf->cfg_lock);
vf->driver_caps = 0;
@@ -1592,6 +1666,8 @@ bool ice_reset_all_vfs(struct ice_pf *pf, bool is_vflr)
ice_flush(hw);
clear_bit(ICE_VF_DIS, pf->state);
+ mutex_unlock(&pf->vfs.table_lock);
+
return true;
}
@@ -1728,7 +1804,8 @@ bool ice_reset_vf(struct ice_vf *vf, bool is_vflr)
ice_eswitch_replay_vf_mac_rule(vf);
/* if the VF has been reset allow it to come up again */
- if (ice_mbx_clear_malvf(&hw->mbx_snapshot, pf->malvfs, ICE_MAX_VF_COUNT, vf->vf_id))
+ if (ice_mbx_clear_malvf(&hw->mbx_snapshot, pf->vfs.malvfs,
+ ICE_MAX_VF_COUNT, vf->vf_id))
dev_dbg(dev, "failed to clear malicious VF state for VF %u\n", i);
return true;
@@ -1740,10 +1817,13 @@ bool ice_reset_vf(struct ice_vf *vf, bool is_vflr)
*/
void ice_vc_notify_link_state(struct ice_pf *pf)
{
- int i;
+ struct ice_vf *vf;
+ unsigned int bkt;
- ice_for_each_vf(pf, i)
- ice_vc_notify_vf_link_state(&pf->vf[i]);
+ mutex_lock(&pf->vfs.table_lock);
+ ice_for_each_vf(pf, bkt, vf)
+ ice_vc_notify_vf_link_state(vf);
+ mutex_unlock(&pf->vfs.table_lock);
}
/**
@@ -1756,7 +1836,7 @@ void ice_vc_notify_reset(struct ice_pf *pf)
{
struct virtchnl_pf_event pfe;
- if (!pf->num_alloc_vfs)
+ if (!ice_has_vfs(pf))
return;
pfe.event = VIRTCHNL_EVENT_RESET_IMPENDING;
@@ -1772,14 +1852,7 @@ void ice_vc_notify_reset(struct ice_pf *pf)
static void ice_vc_notify_vf_reset(struct ice_vf *vf)
{
struct virtchnl_pf_event pfe;
- struct ice_pf *pf;
-
- if (!vf)
- return;
-
- pf = vf->pf;
- if (ice_validate_vf_id(pf, vf->vf_id))
- return;
+ struct ice_pf *pf = vf->pf;
/* Bail out if VF is in disabled state, neither initialized, nor active
* state - otherwise proceed with notifications
@@ -1865,11 +1938,14 @@ release_vsi:
static int ice_start_vfs(struct ice_pf *pf)
{
struct ice_hw *hw = &pf->hw;
- int retval, i;
+ unsigned int bkt, it_cnt;
+ struct ice_vf *vf;
+ int retval;
- ice_for_each_vf(pf, i) {
- struct ice_vf *vf = &pf->vf[i];
+ lockdep_assert_held(&pf->vfs.table_lock);
+ it_cnt = 0;
+ ice_for_each_vf(pf, bkt, vf) {
ice_clear_vf_reset_trigger(vf);
retval = ice_init_vf_vsi_res(vf);
@@ -1882,40 +1958,63 @@ static int ice_start_vfs(struct ice_pf *pf)
set_bit(ICE_VF_STATE_INIT, vf->vf_states);
ice_ena_vf_mappings(vf);
wr32(hw, VFGEN_RSTAT(vf->vf_id), VIRTCHNL_VFR_VFACTIVE);
+ it_cnt++;
}
ice_flush(hw);
return 0;
teardown:
- for (i = i - 1; i >= 0; i--) {
- struct ice_vf *vf = &pf->vf[i];
+ ice_for_each_vf(pf, bkt, vf) {
+ if (it_cnt == 0)
+ break;
ice_dis_vf_mappings(vf);
ice_vf_vsi_release(vf);
+ it_cnt--;
}
return retval;
}
/**
- * ice_set_dflt_settings_vfs - set VF defaults during initialization/creation
- * @pf: PF holding reference to all VFs for default configuration
+ * ice_create_vf_entries - Allocate and insert VF entries
+ * @pf: pointer to the PF structure
+ * @num_vfs: the number of VFs to allocate
+ *
+ * Allocate new VF entries and insert them into the hash table. Set some
+ * basic default fields for initializing the new VFs.
+ *
+ * After this function exits, the hash table will have num_vfs entries
+ * inserted.
+ *
+ * Returns 0 on success or an integer error code on failure.
*/
-static void ice_set_dflt_settings_vfs(struct ice_pf *pf)
+static int ice_create_vf_entries(struct ice_pf *pf, u16 num_vfs)
{
- int i;
+ struct ice_vfs *vfs = &pf->vfs;
+ struct ice_vf *vf;
+ u16 vf_id;
+ int err;
+
+ lockdep_assert_held(&vfs->table_lock);
- ice_for_each_vf(pf, i) {
- struct ice_vf *vf = &pf->vf[i];
+ for (vf_id = 0; vf_id < num_vfs; vf_id++) {
+ vf = kzalloc(sizeof(*vf), GFP_KERNEL);
+ if (!vf) {
+ err = -ENOMEM;
+ goto err_free_entries;
+ }
+ kref_init(&vf->refcnt);
vf->pf = pf;
- vf->vf_id = i;
+ vf->vf_id = vf_id;
+
vf->vf_sw_id = pf->first_sw;
/* assign default capabilities */
set_bit(ICE_VIRTCHNL_VF_CAP_L2, &vf->vf_caps);
vf->spoofchk = true;
- vf->num_vf_qs = pf->num_qps_per_vf;
+ vf->num_vf_qs = pf->vfs.num_qps_per;
ice_vc_set_default_allowlist(vf);
/* ctrl_vsi_idx will be set to a valid value only when VF
@@ -1927,27 +2026,15 @@ static void ice_set_dflt_settings_vfs(struct ice_pf *pf)
ice_vc_set_dflt_vf_ops(&vf->vc_ops);
mutex_init(&vf->cfg_lock);
- }
-}
-
-/**
- * ice_alloc_vfs - allocate num_vfs in the PF structure
- * @pf: PF to store the allocated VFs in
- * @num_vfs: number of VFs to allocate
- */
-static int ice_alloc_vfs(struct ice_pf *pf, int num_vfs)
-{
- struct ice_vf *vfs;
-
- vfs = devm_kcalloc(ice_pf_to_dev(pf), num_vfs, sizeof(*vfs),
- GFP_KERNEL);
- if (!vfs)
- return -ENOMEM;
- pf->vf = vfs;
- pf->num_alloc_vfs = num_vfs;
+ hash_add_rcu(vfs->table, &vf->entry, vf_id);
+ }
return 0;
+
+err_free_entries:
+ ice_free_vf_entries(pf);
+ return err;
}
/**
@@ -1968,28 +2055,29 @@ static int ice_ena_vfs(struct ice_pf *pf, u16 num_vfs)
ice_flush(hw);
ret = pci_enable_sriov(pf->pdev, num_vfs);
- if (ret) {
- pf->num_alloc_vfs = 0;
+ if (ret)
goto err_unroll_intr;
- }
- ret = ice_alloc_vfs(pf, num_vfs);
- if (ret)
- goto err_pci_disable_sriov;
+ mutex_lock(&pf->vfs.table_lock);
- if (ice_set_per_vf_res(pf)) {
+ if (ice_set_per_vf_res(pf, num_vfs)) {
dev_err(dev, "Not enough resources for %d VFs, try with fewer number of VFs\n",
num_vfs);
ret = -ENOSPC;
goto err_unroll_sriov;
}
- ice_set_dflt_settings_vfs(pf);
+ ret = ice_create_vf_entries(pf, num_vfs);
+ if (ret) {
+ dev_err(dev, "Failed to allocate VF entries for %d VFs\n",
+ num_vfs);
+ goto err_unroll_sriov;
+ }
if (ice_start_vfs(pf)) {
dev_err(dev, "Failed to start VF(s)\n");
ret = -EAGAIN;
- goto err_unroll_sriov;
+ goto err_unroll_vf_entries;
}
clear_bit(ICE_VF_DIS, pf->state);
@@ -2002,13 +2090,14 @@ static int ice_ena_vfs(struct ice_pf *pf, u16 num_vfs)
if (test_and_clear_bit(ICE_OICR_INTR_DIS, pf->state))
ice_irq_dynamic_ena(hw, NULL, NULL);
+ mutex_unlock(&pf->vfs.table_lock);
+
return 0;
+err_unroll_vf_entries:
+ ice_free_vf_entries(pf);
err_unroll_sriov:
- devm_kfree(dev, pf->vf);
- pf->vf = NULL;
- pf->num_alloc_vfs = 0;
-err_pci_disable_sriov:
+ mutex_unlock(&pf->vfs.table_lock);
pci_disable_sriov(pf->pdev);
err_unroll_intr:
/* rearm interrupts here */
@@ -2035,9 +2124,9 @@ static int ice_pci_sriov_ena(struct ice_pf *pf, int num_vfs)
else if (pre_existing_vfs && pre_existing_vfs == num_vfs)
return 0;
- if (num_vfs > pf->num_vfs_supported) {
+ if (num_vfs > pf->vfs.num_supported) {
dev_err(dev, "Can't enable %d VFs, max VFs supported is %d\n",
- num_vfs, pf->num_vfs_supported);
+ num_vfs, pf->vfs.num_supported);
return -EOPNOTSUPP;
}
@@ -2135,19 +2224,20 @@ int ice_sriov_configure(struct pci_dev *pdev, int num_vfs)
void ice_process_vflr_event(struct ice_pf *pf)
{
struct ice_hw *hw = &pf->hw;
- unsigned int vf_id;
+ struct ice_vf *vf;
+ unsigned int bkt;
u32 reg;
if (!test_and_clear_bit(ICE_VFLR_EVENT_PENDING, pf->state) ||
- !pf->num_alloc_vfs)
+ !ice_has_vfs(pf))
return;
- ice_for_each_vf(pf, vf_id) {
- struct ice_vf *vf = &pf->vf[vf_id];
+ mutex_lock(&pf->vfs.table_lock);
+ ice_for_each_vf(pf, bkt, vf) {
u32 reg_idx, bit_idx;
- reg_idx = (hw->func_caps.vf_base_id + vf_id) / 32;
- bit_idx = (hw->func_caps.vf_base_id + vf_id) % 32;
+ reg_idx = (hw->func_caps.vf_base_id + vf->vf_id) / 32;
+ bit_idx = (hw->func_caps.vf_base_id + vf->vf_id) % 32;
/* read GLGEN_VFLRSTAT register to find out the flr VFs */
reg = rd32(hw, GLGEN_VFLRSTAT(reg_idx));
if (reg & BIT(bit_idx)) {
@@ -2157,6 +2247,7 @@ void ice_process_vflr_event(struct ice_pf *pf)
mutex_unlock(&vf->cfg_lock);
}
}
+ mutex_unlock(&pf->vfs.table_lock);
}
/**
@@ -2176,22 +2267,36 @@ static void ice_vc_reset_vf(struct ice_vf *vf)
*
* If no VF is found who owns the pfq then return NULL, otherwise return a
* pointer to the VF who owns the pfq
+ *
+ * If this function returns non-NULL, it acquires a reference count of the VF
+ * structure. The caller is responsible for calling ice_put_vf() to drop this
+ * reference.
*/
static struct ice_vf *ice_get_vf_from_pfq(struct ice_pf *pf, u16 pfq)
{
- unsigned int vf_id;
+ struct ice_vf *vf;
+ unsigned int bkt;
- ice_for_each_vf(pf, vf_id) {
- struct ice_vf *vf = &pf->vf[vf_id];
+ rcu_read_lock();
+ ice_for_each_vf_rcu(pf, bkt, vf) {
struct ice_vsi *vsi;
u16 rxq_idx;
vsi = ice_get_vf_vsi(vf);
ice_for_each_rxq(vsi, rxq_idx)
- if (vsi->rxq_map[rxq_idx] == pfq)
- return vf;
+ if (vsi->rxq_map[rxq_idx] == pfq) {
+ struct ice_vf *found;
+
+ if (kref_get_unless_zero(&vf->refcnt))
+ found = vf;
+ else
+ found = NULL;
+ rcu_read_unlock();
+ return found;
+ }
}
+ rcu_read_unlock();
return NULL;
}
@@ -2235,6 +2340,8 @@ ice_vf_lan_overflow_event(struct ice_pf *pf, struct ice_rq_event_info *event)
mutex_lock(&vf->cfg_lock);
ice_vc_reset_vf(vf);
mutex_unlock(&vf->cfg_lock);
+
+ ice_put_vf(vf);
}
/**
@@ -2255,13 +2362,7 @@ ice_vc_send_msg_to_vf(struct ice_vf *vf, u32 v_opcode,
struct ice_pf *pf;
int aq_ret;
- if (!vf)
- return -EINVAL;
-
pf = vf->pf;
- if (ice_validate_vf_id(pf, vf->vf_id))
- return -EINVAL;
-
dev = ice_pf_to_dev(pf);
/* single place to detect unsuccessful return values */
@@ -2452,7 +2553,7 @@ static int ice_vc_get_vf_res_msg(struct ice_vf *vf, u8 *msg)
vfres->num_vsis = 1;
/* Tx and Rx queue are equal for VF */
vfres->num_queue_pairs = vsi->num_txq;
- vfres->max_vectors = pf->num_msix_per_vf;
+ vfres->max_vectors = pf->vfs.num_msix_per;
vfres->rss_key_size = ICE_VSIQF_HKEY_ARRAY_SIZE;
vfres->rss_lut_size = ICE_VSIQF_HLUT_ARRAY_SIZE;
vfres->max_mtu = ice_vc_get_max_frame_size(vf);
@@ -2526,7 +2627,7 @@ bool ice_vc_isvalid_vsi_id(struct ice_vf *vf, u16 vsi_id)
vsi = ice_find_vsi_from_id(pf, vsi_id);
- return (vsi && (vsi->vf_id == vf->vf_id));
+ return (vsi && (vsi->vf == vf));
}
/**
@@ -3017,30 +3118,34 @@ int ice_set_vf_spoofchk(struct net_device *netdev, int vf_id, bool ena)
int ret;
dev = ice_pf_to_dev(pf);
- if (ice_validate_vf_id(pf, vf_id))
+
+ vf = ice_get_vf_by_id(pf, vf_id);
+ if (!vf)
return -EINVAL;
- vf = &pf->vf[vf_id];
ret = ice_check_vf_ready_for_cfg(vf);
if (ret)
- return ret;
+ goto out_put_vf;
vf_vsi = ice_get_vf_vsi(vf);
if (!vf_vsi) {
netdev_err(netdev, "VSI %d for VF %d is null\n",
vf->lan_vsi_idx, vf->vf_id);
- return -EINVAL;
+ ret = -EINVAL;
+ goto out_put_vf;
}
if (vf_vsi->type != ICE_VSI_VF) {
netdev_err(netdev, "Type %d of VSI %d for VF %d is no ICE_VSI_VF\n",
vf_vsi->type, vf_vsi->vsi_num, vf->vf_id);
- return -ENODEV;
+ ret = -ENODEV;
+ goto out_put_vf;
}
if (ena == vf->spoofchk) {
dev_dbg(dev, "VF spoofchk already %s\n", ena ? "ON" : "OFF");
- return 0;
+ ret = 0;
+ goto out_put_vf;
}
if (ena)
@@ -3053,6 +3158,8 @@ int ice_set_vf_spoofchk(struct net_device *netdev, int vf_id, bool ena)
else
vf->spoofchk = ena;
+out_put_vf:
+ ice_put_vf(vf);
return ret;
}
@@ -3065,18 +3172,22 @@ int ice_set_vf_spoofchk(struct net_device *netdev, int vf_id, bool ena)
*/
bool ice_is_any_vf_in_promisc(struct ice_pf *pf)
{
- int vf_idx;
-
- ice_for_each_vf(pf, vf_idx) {
- struct ice_vf *vf = &pf->vf[vf_idx];
+ bool is_vf_promisc = false;
+ struct ice_vf *vf;
+ unsigned int bkt;
+ rcu_read_lock();
+ ice_for_each_vf_rcu(pf, bkt, vf) {
/* found a VF that has promiscuous mode configured */
if (test_bit(ICE_VF_STATE_UC_PROMISC, vf->vf_states) ||
- test_bit(ICE_VF_STATE_MC_PROMISC, vf->vf_states))
- return true;
+ test_bit(ICE_VF_STATE_MC_PROMISC, vf->vf_states)) {
+ is_vf_promisc = true;
+ break;
+ }
}
+ rcu_read_unlock();
- return false;
+ return is_vf_promisc;
}
/**
@@ -3596,7 +3707,7 @@ static int ice_vc_cfg_irq_map_msg(struct ice_vf *vf, u8 *msg)
* there is actually at least a single VF queue vector mapped
*/
if (!test_bit(ICE_VF_STATE_ACTIVE, vf->vf_states) ||
- pf->num_msix_per_vf < num_q_vectors_mapped ||
+ pf->vfs.num_msix_per < num_q_vectors_mapped ||
!num_q_vectors_mapped) {
v_ret = VIRTCHNL_STATUS_ERR_PARAM;
goto error_param;
@@ -3618,7 +3729,7 @@ static int ice_vc_cfg_irq_map_msg(struct ice_vf *vf, u8 *msg)
/* vector_id is always 0-based for each VF, and can never be
* larger than or equal to the max allowed interrupts per VF
*/
- if (!(vector_id < pf->num_msix_per_vf) ||
+ if (!(vector_id < pf->vfs.num_msix_per) ||
!ice_vc_isvalid_vsi_id(vf, vsi_id) ||
(!vector_id && (map->rxq_map || map->txq_map))) {
v_ret = VIRTCHNL_STATUS_ERR_PARAM;
@@ -4209,8 +4320,6 @@ ice_set_vf_port_vlan(struct net_device *netdev, int vf_id, u16 vlan_id, u8 qos,
int ret;
dev = ice_pf_to_dev(pf);
- if (ice_validate_vf_id(pf, vf_id))
- return -EINVAL;
if (vlan_id >= VLAN_N_VID || qos > 7) {
dev_err(dev, "Invalid Port VLAN parameters for VF %d, ID %d, QoS %d\n",
@@ -4224,10 +4333,13 @@ ice_set_vf_port_vlan(struct net_device *netdev, int vf_id, u16 vlan_id, u8 qos,
return -EPROTONOSUPPORT;
}
- vf = &pf->vf[vf_id];
+ vf = ice_get_vf_by_id(pf, vf_id);
+ if (!vf)
+ return -EINVAL;
+
ret = ice_check_vf_ready_for_cfg(vf);
if (ret)
- return ret;
+ goto out_put_vf;
if (ice_vf_get_port_vlan_prio(vf) == qos &&
ice_vf_get_port_vlan_tpid(vf) == local_vlan_proto &&
@@ -4235,7 +4347,8 @@ ice_set_vf_port_vlan(struct net_device *netdev, int vf_id, u16 vlan_id, u8 qos,
/* duplicate request, so just return success */
dev_dbg(dev, "Duplicate port VLAN %u, QoS %u, TPID 0x%04x request\n",
vlan_id, qos, local_vlan_proto);
- return 0;
+ ret = 0;
+ goto out_put_vf;
}
mutex_lock(&vf->cfg_lock);
@@ -4250,7 +4363,9 @@ ice_set_vf_port_vlan(struct net_device *netdev, int vf_id, u16 vlan_id, u8 qos,
ice_vc_reset_vf(vf);
mutex_unlock(&vf->cfg_lock);
- return 0;
+out_put_vf:
+ ice_put_vf(vf);
+ return ret;
}
/**
@@ -5772,12 +5887,13 @@ void ice_vc_process_vf_msg(struct ice_pf *pf, struct ice_rq_event_info *event)
int err = 0;
dev = ice_pf_to_dev(pf);
- if (ice_validate_vf_id(pf, vf_id)) {
- err = -EINVAL;
- goto error_handler;
- }
- vf = &pf->vf[vf_id];
+ vf = ice_get_vf_by_id(pf, vf_id);
+ if (!vf) {
+ dev_err(dev, "Unable to locate VF for message from VF ID %d, opcode %d, len %d\n",
+ vf_id, v_opcode, msglen);
+ return;
+ }
/* Check if VF is disabled. */
if (test_bit(ICE_VF_STATE_DIS, vf->vf_states)) {
@@ -5800,6 +5916,7 @@ void ice_vc_process_vf_msg(struct ice_pf *pf, struct ice_rq_event_info *event)
ice_vc_send_msg_to_vf(vf, v_opcode,
VIRTCHNL_STATUS_ERR_NOT_SUPPORTED, NULL,
0);
+ ice_put_vf(vf);
return;
}
@@ -5809,6 +5926,7 @@ error_handler:
NULL, 0);
dev_err(dev, "Invalid message from VF %d, opcode %d, len %d, error %d\n",
vf_id, v_opcode, msglen, err);
+ ice_put_vf(vf);
return;
}
@@ -5818,6 +5936,7 @@ error_handler:
if (!mutex_trylock(&vf->cfg_lock)) {
dev_info(dev, "VF %u is being configured in another context that will trigger a VFR, so there is no need to handle this message\n",
vf->vf_id);
+ ice_put_vf(vf);
return;
}
@@ -5932,6 +6051,7 @@ error_handler:
}
mutex_unlock(&vf->cfg_lock);
+ ice_put_vf(vf);
}
/**
@@ -5947,14 +6067,15 @@ ice_get_vf_cfg(struct net_device *netdev, int vf_id, struct ifla_vf_info *ivi)
{
struct ice_pf *pf = ice_netdev_to_pf(netdev);
struct ice_vf *vf;
+ int ret;
- if (ice_validate_vf_id(pf, vf_id))
+ vf = ice_get_vf_by_id(pf, vf_id);
+ if (!vf)
return -EINVAL;
- vf = &pf->vf[vf_id];
-
- if (ice_check_vf_init(pf, vf))
- return -EBUSY;
+ ret = ice_check_vf_ready_for_cfg(vf);
+ if (ret)
+ goto out_put_vf;
ivi->vf = vf_id;
ether_addr_copy(ivi->mac, vf->hw_lan_addr.addr);
@@ -5975,7 +6096,10 @@ ice_get_vf_cfg(struct net_device *netdev, int vf_id, struct ifla_vf_info *ivi)
ivi->linkstate = IFLA_VF_LINK_STATE_DISABLE;
ivi->max_tx_rate = vf->max_tx_rate;
ivi->min_tx_rate = vf->min_tx_rate;
- return 0;
+
+out_put_vf:
+ ice_put_vf(vf);
+ return ret;
}
/**
@@ -6025,28 +6149,31 @@ int ice_set_vf_mac(struct net_device *netdev, int vf_id, u8 *mac)
struct ice_vf *vf;
int ret;
- if (ice_validate_vf_id(pf, vf_id))
- return -EINVAL;
-
if (is_multicast_ether_addr(mac)) {
netdev_err(netdev, "%pM not a valid unicast address\n", mac);
return -EINVAL;
}
- vf = &pf->vf[vf_id];
+ vf = ice_get_vf_by_id(pf, vf_id);
+ if (!vf)
+ return -EINVAL;
+
/* nothing left to do, unicast MAC already set */
if (ether_addr_equal(vf->dev_lan_addr.addr, mac) &&
- ether_addr_equal(vf->hw_lan_addr.addr, mac))
- return 0;
+ ether_addr_equal(vf->hw_lan_addr.addr, mac)) {
+ ret = 0;
+ goto out_put_vf;
+ }
ret = ice_check_vf_ready_for_cfg(vf);
if (ret)
- return ret;
+ goto out_put_vf;
if (ice_unicast_mac_exists(pf, mac)) {
netdev_err(netdev, "Unicast MAC %pM already exists on this PF. Preventing setting VF %u unicast MAC address to %pM\n",
mac, vf_id, mac);
- return -EINVAL;
+ ret = -EINVAL;
+ goto out_put_vf;
}
mutex_lock(&vf->cfg_lock);
@@ -6070,7 +6197,10 @@ int ice_set_vf_mac(struct net_device *netdev, int vf_id, u8 *mac)
ice_vc_reset_vf(vf);
mutex_unlock(&vf->cfg_lock);
- return 0;
+
+out_put_vf:
+ ice_put_vf(vf);
+ return ret;
}
/**
@@ -6092,17 +6222,19 @@ int ice_set_vf_trust(struct net_device *netdev, int vf_id, bool trusted)
return -EOPNOTSUPP;
}
- if (ice_validate_vf_id(pf, vf_id))
+ vf = ice_get_vf_by_id(pf, vf_id);
+ if (!vf)
return -EINVAL;
- vf = &pf->vf[vf_id];
ret = ice_check_vf_ready_for_cfg(vf);
if (ret)
- return ret;
+ goto out_put_vf;
/* Check if already trusted */
- if (trusted == vf->trusted)
- return 0;
+ if (trusted == vf->trusted) {
+ ret = 0;
+ goto out_put_vf;
+ }
mutex_lock(&vf->cfg_lock);
@@ -6113,7 +6245,9 @@ int ice_set_vf_trust(struct net_device *netdev, int vf_id, bool trusted)
mutex_unlock(&vf->cfg_lock);
- return 0;
+out_put_vf:
+ ice_put_vf(vf);
+ return ret;
}
/**
@@ -6130,13 +6264,13 @@ int ice_set_vf_link_state(struct net_device *netdev, int vf_id, int link_state)
struct ice_vf *vf;
int ret;
- if (ice_validate_vf_id(pf, vf_id))
+ vf = ice_get_vf_by_id(pf, vf_id);
+ if (!vf)
return -EINVAL;
- vf = &pf->vf[vf_id];
ret = ice_check_vf_ready_for_cfg(vf);
if (ret)
- return ret;
+ goto out_put_vf;
switch (link_state) {
case IFLA_VF_LINK_STATE_AUTO:
@@ -6151,12 +6285,15 @@ int ice_set_vf_link_state(struct net_device *netdev, int vf_id, int link_state)
vf->link_up = false;
break;
default:
- return -EINVAL;
+ ret = -EINVAL;
+ goto out_put_vf;
}
ice_vc_notify_vf_link_state(vf);
- return 0;
+out_put_vf:
+ ice_put_vf(vf);
+ return ret;
}
/**
@@ -6165,10 +6302,14 @@ int ice_set_vf_link_state(struct net_device *netdev, int vf_id, int link_state)
*/
static int ice_calc_all_vfs_min_tx_rate(struct ice_pf *pf)
{
- int rate = 0, i;
+ struct ice_vf *vf;
+ unsigned int bkt;
+ int rate = 0;
- ice_for_each_vf(pf, i)
- rate += pf->vf[i].min_tx_rate;
+ rcu_read_lock();
+ ice_for_each_vf_rcu(pf, bkt, vf)
+ rate += vf->min_tx_rate;
+ rcu_read_unlock();
return rate;
}
@@ -6223,13 +6364,14 @@ ice_set_vf_bw(struct net_device *netdev, int vf_id, int min_tx_rate,
int ret;
dev = ice_pf_to_dev(pf);
- if (ice_validate_vf_id(pf, vf_id))
+
+ vf = ice_get_vf_by_id(pf, vf_id);
+ if (!vf)
return -EINVAL;
- vf = &pf->vf[vf_id];
ret = ice_check_vf_ready_for_cfg(vf);
if (ret)
- return ret;
+ goto out_put_vf;
vsi = ice_get_vf_vsi(vf);
@@ -6239,23 +6381,27 @@ ice_set_vf_bw(struct net_device *netdev, int vf_id, int min_tx_rate,
if (max_tx_rate && min_tx_rate > max_tx_rate) {
dev_err(dev, "Cannot set min Tx rate %d Mbps greater than max Tx rate %d Mbps\n",
min_tx_rate, max_tx_rate);
- return -EINVAL;
+ ret = -EINVAL;
+ goto out_put_vf;
}
if (min_tx_rate && ice_is_dcb_active(pf)) {
dev_err(dev, "DCB on PF is currently enabled. VF min Tx rate limiting not allowed on this PF.\n");
- return -EOPNOTSUPP;
+ ret = -EOPNOTSUPP;
+ goto out_put_vf;
}
- if (ice_min_tx_rate_oversubscribed(vf, min_tx_rate))
- return -EINVAL;
+ if (ice_min_tx_rate_oversubscribed(vf, min_tx_rate)) {
+ ret = -EINVAL;
+ goto out_put_vf;
+ }
if (vf->min_tx_rate != (unsigned int)min_tx_rate) {
ret = ice_set_min_bw_limit(vsi, (u64)min_tx_rate * 1000);
if (ret) {
dev_err(dev, "Unable to set min-tx-rate for VF %d\n",
vf->vf_id);
- return ret;
+ goto out_put_vf;
}
vf->min_tx_rate = min_tx_rate;
@@ -6266,13 +6412,15 @@ ice_set_vf_bw(struct net_device *netdev, int vf_id, int min_tx_rate,
if (ret) {
dev_err(dev, "Unable to set max-tx-rate for VF %d\n",
vf->vf_id);
- return ret;
+ goto out_put_vf;
}
vf->max_tx_rate = max_tx_rate;
}
- return 0;
+out_put_vf:
+ ice_put_vf(vf);
+ return ret;
}
/**
@@ -6290,17 +6438,19 @@ int ice_get_vf_stats(struct net_device *netdev, int vf_id,
struct ice_vf *vf;
int ret;
- if (ice_validate_vf_id(pf, vf_id))
+ vf = ice_get_vf_by_id(pf, vf_id);
+ if (!vf)
return -EINVAL;
- vf = &pf->vf[vf_id];
ret = ice_check_vf_ready_for_cfg(vf);
if (ret)
- return ret;
+ goto out_put_vf;
vsi = ice_get_vf_vsi(vf);
- if (!vsi)
- return -EINVAL;
+ if (!vsi) {
+ ret = -EINVAL;
+ goto out_put_vf;
+ }
ice_update_eth_stats(vsi);
stats = &vsi->eth_stats;
@@ -6318,7 +6468,9 @@ int ice_get_vf_stats(struct net_device *netdev, int vf_id,
vf_stats->rx_dropped = stats->rx_discards;
vf_stats->tx_dropped = stats->tx_discards;
- return 0;
+out_put_vf:
+ ice_put_vf(vf);
+ return ret;
}
/**
@@ -6349,21 +6501,21 @@ void ice_print_vfs_mdd_events(struct ice_pf *pf)
{
struct device *dev = ice_pf_to_dev(pf);
struct ice_hw *hw = &pf->hw;
- int i;
+ struct ice_vf *vf;
+ unsigned int bkt;
/* check that there are pending MDD events to print */
if (!test_and_clear_bit(ICE_MDD_VF_PRINT_PENDING, pf->state))
return;
/* VF MDD event logs are rate limited to one second intervals */
- if (time_is_after_jiffies(pf->last_printed_mdd_jiffies + HZ * 1))
+ if (time_is_after_jiffies(pf->vfs.last_printed_mdd_jiffies + HZ * 1))
return;
- pf->last_printed_mdd_jiffies = jiffies;
-
- ice_for_each_vf(pf, i) {
- struct ice_vf *vf = &pf->vf[i];
+ pf->vfs.last_printed_mdd_jiffies = jiffies;
+ mutex_lock(&pf->vfs.table_lock);
+ ice_for_each_vf(pf, bkt, vf) {
/* only print Rx MDD event message if there are new events */
if (vf->mdd_rx_events.count != vf->mdd_rx_events.last_printed) {
vf->mdd_rx_events.last_printed =
@@ -6377,10 +6529,11 @@ void ice_print_vfs_mdd_events(struct ice_pf *pf)
vf->mdd_tx_events.count;
dev_info(dev, "%d Tx Malicious Driver Detection events detected on PF %d VF %d MAC %pM.\n",
- vf->mdd_tx_events.count, hw->pf_id, i,
+ vf->mdd_tx_events.count, hw->pf_id, vf->vf_id,
vf->dev_lan_addr.addr);
}
}
+ mutex_unlock(&pf->vfs.table_lock);
}
/**
@@ -6432,13 +6585,12 @@ ice_is_malicious_vf(struct ice_pf *pf, struct ice_rq_event_info *event,
struct ice_vf *vf;
int status;
- if (ice_validate_vf_id(pf, vf_id))
+ vf = ice_get_vf_by_id(pf, vf_id);
+ if (!vf)
return false;
- vf = &pf->vf[vf_id];
- /* Check if VF is disabled. */
if (test_bit(ICE_VF_STATE_DIS, vf->vf_states))
- return false;
+ goto out_put_vf;
mbxdata.num_msg_proc = num_msg_proc;
mbxdata.num_pending_arq = num_msg_pending;
@@ -6449,7 +6601,7 @@ ice_is_malicious_vf(struct ice_pf *pf, struct ice_rq_event_info *event,
/* check to see if we have a malicious VF */
status = ice_mbx_vf_state_handler(&pf->hw, &mbxdata, vf_id, &malvf);
if (status)
- return false;
+ goto out_put_vf;
if (malvf) {
bool report_vf = false;
@@ -6457,7 +6609,7 @@ ice_is_malicious_vf(struct ice_pf *pf, struct ice_rq_event_info *event,
/* if the VF is malicious and we haven't let the user
* know about it, then let them know now
*/
- status = ice_mbx_report_malvf(&pf->hw, pf->malvfs,
+ status = ice_mbx_report_malvf(&pf->hw, pf->vfs.malvfs,
ICE_MAX_VF_COUNT, vf_id,
&report_vf);
if (status)
@@ -6471,12 +6623,9 @@ ice_is_malicious_vf(struct ice_pf *pf, struct ice_rq_event_info *event,
&vf->dev_lan_addr.addr[0],
pf_vsi->netdev->dev_addr);
}
-
- return true;
}
- /* if there was an error in detection or the VF is not malicious then
- * return false
- */
- return false;
+out_put_vf:
+ ice_put_vf(vf);
+ return malvf;
}
diff --git a/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.h b/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.h
index 4f4961043638..02e3d306f6dd 100644
--- a/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.h
+++ b/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.h
@@ -39,8 +39,50 @@
#define ICE_MAX_VF_RESET_TRIES 40
#define ICE_MAX_VF_RESET_SLEEP_MS 20
-#define ice_for_each_vf(pf, i) \
- for ((i) = 0; (i) < (pf)->num_alloc_vfs; (i)++)
+/* VF Hash Table access functions
+ *
+ * These functions provide abstraction for interacting with the VF hash table.
+ * In general, direct access to the hash table should be avoided outside of
+ * these functions where possible.
+ *
+ * The VF entries in the hash table are protected by reference counting to
+ * track lifetime of accesses from the table. The ice_get_vf_by_id() function
+ * obtains a reference to the VF structure which must be dropped by using
+ * ice_put_vf().
+ */
+
+/**
+ * ice_for_each_vf - Iterate over each VF entry
+ * @pf: pointer to the PF private structure
+ * @bkt: bucket index used for iteration
+ * @vf: pointer to the VF entry currently being processed in the loop.
+ *
+ * The bkt variable is an unsigned integer iterator used to traverse the VF
+ * entries. It is *not* guaranteed to be the VF's vf_id. Do not assume it is.
+ * Use vf->vf_id to get the id number if needed.
+ *
+ * The caller is expected to be under the table_lock mutex for the entire
+ * loop. Use this iterator if your loop is long or if it might sleep.
+ */
+#define ice_for_each_vf(pf, bkt, vf) \
+ hash_for_each((pf)->vfs.table, (bkt), (vf), entry)
+
+/**
+ * ice_for_each_vf_rcu - Iterate over each VF entry protected by RCU
+ * @pf: pointer to the PF private structure
+ * @bkt: bucket index used for iteration
+ * @vf: pointer to the VF entry currently being processed in the loop.
+ *
+ * The bkt variable is an unsigned integer iterator used to traverse the VF
+ * entries. It is *not* guaranteed to be the VF's vf_id. Do not assume it is.
+ * Use vf->vf_id to get the id number if needed.
+ *
+ * The caller is expected to be under rcu_read_lock() for the entire loop.
+ * Only use this iterator if your loop is short and you can guarantee it does
+ * not sleep.
+ */
+#define ice_for_each_vf_rcu(pf, bkt, vf) \
+ hash_for_each_rcu((pf)->vfs.table, (bkt), (vf), entry)
/* Specific VF states */
enum ice_vf_states {
@@ -104,8 +146,22 @@ struct ice_vc_vf_ops {
int (*dis_vlan_insertion_v2_msg)(struct ice_vf *vf, u8 *msg);
};
+/* Virtchnl/SR-IOV config info */
+struct ice_vfs {
+ DECLARE_HASHTABLE(table, 8); /* table of VF entries */
+ struct mutex table_lock; /* Lock for protecting the hash table */
+ u16 num_supported; /* max supported VFs on this PF */
+ u16 num_qps_per; /* number of queue pairs per VF */
+ u16 num_msix_per; /* number of MSI-X vectors per VF */
+ unsigned long last_printed_mdd_jiffies; /* MDD message rate limit */
+ DECLARE_BITMAP(malvfs, ICE_MAX_VF_COUNT); /* malicious VF indicator */
+};
+
/* VF information structure */
struct ice_vf {
+ struct hlist_node entry;
+ struct rcu_head rcu;
+ struct kref refcnt;
struct ice_pf *pf;
/* Used during virtchnl message handling and NDO ops against the VF
@@ -162,6 +218,10 @@ struct ice_vf {
};
#ifdef CONFIG_PCI_IOV
+struct ice_vf *ice_get_vf_by_id(struct ice_pf *pf, u16 vf_id);
+void ice_put_vf(struct ice_vf *vf);
+bool ice_has_vfs(struct ice_pf *pf);
+u16 ice_get_num_vfs(struct ice_pf *pf);
struct ice_vsi *ice_get_vf_vsi(struct ice_vf *vf);
void ice_process_vflr_event(struct ice_pf *pf);
int ice_sriov_configure(struct pci_dev *pdev, int num_vfs);
@@ -221,6 +281,25 @@ ice_vc_send_msg_to_vf(struct ice_vf *vf, u32 v_opcode,
bool ice_vc_isvalid_vsi_id(struct ice_vf *vf, u16 vsi_id);
bool ice_vf_is_port_vlan_ena(struct ice_vf *vf);
#else /* CONFIG_PCI_IOV */
+static inline struct ice_vf *ice_get_vf_by_id(struct ice_pf *pf, u16 vf_id)
+{
+ return NULL;
+}
+
+static inline void ice_put_vf(struct ice_vf *vf)
+{
+}
+
+static inline bool ice_has_vfs(struct ice_pf *pf)
+{
+ return false;
+}
+
+static inline u16 ice_get_num_vfs(struct ice_pf *pf)
+{
+ return 0;
+}
+
static inline void ice_process_vflr_event(struct ice_pf *pf) { }
static inline void ice_free_vfs(struct ice_pf *pf) { }
static inline
generated by cgit v1.2.3 (git 2.39.1) at 2024-12-19 22:30:47 +0000