[PATCH net-next 00/13] bnxt_en: Ntuple and RSS updates

[PATCH net-next 00/13] bnxt_en: Ntuple and RSS updates

[Michael Chan]

[-- Attachment #1: Type: text/plain, Size: 1466 bytes --]

This patch series adds more ntuple and RSS features following recent
patches to add support for user configured ntuple filters.  Additional
features include L2 ether filters, partial tuple masks, IP filters
besides TCP/UDP, drop action, saving and re-applying user filters
after driver reset, user configured RSS key, and RSS for IPSEC.

Ajit Khaparde (1):
  bnxt_en: Add RSS support for IPSEC headers

Edwin Peer (1):
  bnxt_en: implement fully specified 5-tuple masks

Michael Chan (3):
  bnxt_en: Use firmware provided maximum filter counts.
  bnxt_en: Add ethtool -N support for ether filters.
  bnxt_en: Support ethtool -n to display ether filters.

Pavan Chebbi (6):
  bnxt_en: Add separate function to delete the filter structure
  bnxt_en: Save user configured filters in a lookup list
  bnxt_en: Retain user configured filters when closing
  bnxt_en: Restore all the user created L2 and ntuple filters
  bnxt_en: Add support for user configured RSS key
  bnxt_en: Invalidate user filters when needed

Vikas Gupta (2):
  bnxt_en: Enhance ethtool ntuple support for ip flows besides TCP/UDP
  bnxt_en: Add drop action support for ntuple

 drivers/net/ethernet/broadcom/bnxt/bnxt.c     | 346 ++++++++++++----
 drivers/net/ethernet/broadcom/bnxt/bnxt.h     |  52 ++-
 .../net/ethernet/broadcom/bnxt/bnxt_ethtool.c | 389 +++++++++++++-----
 .../net/ethernet/broadcom/bnxt/bnxt_ethtool.h |   2 +
 4 files changed, 579 insertions(+), 210 deletions(-)

-- 
2.30.1


[-- Attachment #2: S/MIME Cryptographic Signature --]
[-- Type: application/pkcs7-signature, Size: 4209 bytes --]

[PATCH net-next 01/13] bnxt_en: Use firmware provided maximum filter counts.

[Michael Chan]

[-- Attachment #1: Type: text/plain, Size: 6057 bytes --]

While individual filter structures are allocated as needed, there is an
array to keep track of the software filter IDs that we allocate ahead
of time.  Rather than relying on a fixed maximum filter count to
allocate this array, get the maximum from the firmware when available.

Move these filter related maximum counts queried from the firmware to the
bnxt_hw_resc struct.  If the firmware is not providing these maximum
counts, fall back to the hard-coded constant.

Reviewed-by: Pavan Chebbi <pavan.chebbi@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
---
 drivers/net/ethernet/broadcom/bnxt/bnxt.c     | 26 ++++++++++++-------
 drivers/net/ethernet/broadcom/bnxt/bnxt.h     | 13 +++++-----
 .../net/ethernet/broadcom/bnxt/bnxt_ethtool.c |  4 +--
 3 files changed, 26 insertions(+), 17 deletions(-)

diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.c b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
index fde32b32fa81..91ecf514b924 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt.c
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
@@ -4840,7 +4840,7 @@ static int bnxt_alloc_ntp_fltrs(struct bnxt *bp)
 		INIT_HLIST_HEAD(&bp->ntp_fltr_hash_tbl[i]);
 
 	bp->ntp_fltr_count = 0;
-	bp->ntp_fltr_bmap = bitmap_zalloc(BNXT_MAX_FLTR, GFP_KERNEL);
+	bp->ntp_fltr_bmap = bitmap_zalloc(bp->max_fltr, GFP_KERNEL);
 
 	if (!bp->ntp_fltr_bmap)
 		rc = -ENOMEM;
@@ -5480,7 +5480,7 @@ static int bnxt_init_l2_filter(struct bnxt *bp, struct bnxt_l2_filter *fltr,
 		int bit_id;
 
 		bit_id = bitmap_find_free_region(bp->ntp_fltr_bmap,
-						 BNXT_MAX_FLTR, 0);
+						 bp->max_fltr, 0);
 		if (bit_id < 0)
 			return -ENOMEM;
 		fltr->base.sw_id = (u16)bit_id;
@@ -8709,6 +8709,13 @@ static int __bnxt_hwrm_func_qcaps(struct bnxt *bp)
 	hw_resc->max_vnics = le16_to_cpu(resp->max_vnics);
 	hw_resc->max_stat_ctxs = le16_to_cpu(resp->max_stat_ctx);
 
+	hw_resc->max_encap_records = le32_to_cpu(resp->max_encap_records);
+	hw_resc->max_decap_records = le32_to_cpu(resp->max_decap_records);
+	hw_resc->max_tx_em_flows = le32_to_cpu(resp->max_tx_em_flows);
+	hw_resc->max_tx_wm_flows = le32_to_cpu(resp->max_tx_wm_flows);
+	hw_resc->max_rx_em_flows = le32_to_cpu(resp->max_rx_em_flows);
+	hw_resc->max_rx_wm_flows = le32_to_cpu(resp->max_rx_wm_flows);
+
 	if (BNXT_PF(bp)) {
 		struct bnxt_pf_info *pf = &bp->pf;
 
@@ -8717,12 +8724,6 @@ static int __bnxt_hwrm_func_qcaps(struct bnxt *bp)
 		memcpy(pf->mac_addr, resp->mac_address, ETH_ALEN);
 		pf->first_vf_id = le16_to_cpu(resp->first_vf_id);
 		pf->max_vfs = le16_to_cpu(resp->max_vfs);
-		pf->max_encap_records = le32_to_cpu(resp->max_encap_records);
-		pf->max_decap_records = le32_to_cpu(resp->max_decap_records);
-		pf->max_tx_em_flows = le32_to_cpu(resp->max_tx_em_flows);
-		pf->max_tx_wm_flows = le32_to_cpu(resp->max_tx_wm_flows);
-		pf->max_rx_em_flows = le32_to_cpu(resp->max_rx_em_flows);
-		pf->max_rx_wm_flows = le32_to_cpu(resp->max_rx_wm_flows);
 		bp->flags &= ~BNXT_FLAG_WOL_CAP;
 		if (flags & FUNC_QCAPS_RESP_FLAGS_WOL_MAGICPKT_SUPPORTED)
 			bp->flags |= BNXT_FLAG_WOL_CAP;
@@ -13899,7 +13900,7 @@ int bnxt_insert_ntp_filter(struct bnxt *bp, struct bnxt_ntuple_filter *fltr,
 	int bit_id;
 
 	spin_lock_bh(&bp->ntp_fltr_lock);
-	bit_id = bitmap_find_free_region(bp->ntp_fltr_bmap, BNXT_MAX_FLTR, 0);
+	bit_id = bitmap_find_free_region(bp->ntp_fltr_bmap, bp->max_fltr, 0);
 	if (bit_id < 0) {
 		spin_unlock_bh(&bp->ntp_fltr_lock);
 		return -ENOMEM;
@@ -14669,6 +14670,7 @@ void bnxt_print_device_info(struct bnxt *bp)
 
 static int bnxt_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
 {
+	struct bnxt_hw_resc *hw_resc;
 	struct net_device *dev;
 	struct bnxt *bp;
 	int rc, max_irqs;
@@ -14827,6 +14829,12 @@ static int bnxt_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
 	if (rc)
 		goto init_err_pci_clean;
 
+	hw_resc = &bp->hw_resc;
+	bp->max_fltr = hw_resc->max_rx_em_flows + hw_resc->max_rx_wm_flows +
+		       BNXT_L2_FLTR_MAX_FLTR;
+	/* Older firmware may not report these filters properly */
+	if (bp->max_fltr < BNXT_MAX_FLTR)
+		bp->max_fltr = BNXT_MAX_FLTR;
 	bnxt_init_l2_fltr_tbl(bp);
 	bnxt_set_rx_skb_mode(bp, false);
 	bnxt_set_tpa_flags(bp);
diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.h b/drivers/net/ethernet/broadcom/bnxt/bnxt.h
index b2cb3e77559d..4bd1cf01d99e 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt.h
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.h
@@ -1281,6 +1281,12 @@ struct bnxt_hw_resc {
 	u16	max_nqs;
 	u16	max_irqs;
 	u16	resv_irqs;
+	u32	max_encap_records;
+	u32	max_decap_records;
+	u32	max_tx_em_flows;
+	u32	max_tx_wm_flows;
+	u32	max_rx_em_flows;
+	u32	max_rx_wm_flows;
 };
 
 #if defined(CONFIG_BNXT_SRIOV)
@@ -1315,12 +1321,6 @@ struct bnxt_pf_info {
 	u16	active_vfs;
 	u16	registered_vfs;
 	u16	max_vfs;
-	u32	max_encap_records;
-	u32	max_decap_records;
-	u32	max_tx_em_flows;
-	u32	max_tx_wm_flows;
-	u32	max_rx_em_flows;
-	u32	max_rx_wm_flows;
 	unsigned long	*vf_event_bmap;
 	u16	hwrm_cmd_req_pages;
 	u8	vf_resv_strategy;
@@ -2428,6 +2428,7 @@ struct bnxt {
 
 	unsigned long		*ntp_fltr_bmap;
 	int			ntp_fltr_count;
+	int			max_fltr;
 
 #define BNXT_L2_FLTR_MAX_FLTR	1024
 #define BNXT_L2_FLTR_HASH_SIZE	32
diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c b/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c
index 481b835a7703..1c8610386404 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c
@@ -1077,7 +1077,7 @@ static int bnxt_grxclsrule(struct bnxt *bp, struct ethtool_rxnfc *cmd)
 	struct flow_keys *fkeys;
 	int rc = -EINVAL;
 
-	if (fs->location >= BNXT_NTP_FLTR_MAX_FLTR)
+	if (fs->location >= bp->max_fltr)
 		return rc;
 
 	rcu_read_lock();
@@ -1521,7 +1521,7 @@ static int bnxt_get_rxnfc(struct net_device *dev, struct ethtool_rxnfc *cmd,
 
 	case ETHTOOL_GRXCLSRLCNT:
 		cmd->rule_cnt = bp->ntp_fltr_count;
-		cmd->data = BNXT_NTP_FLTR_MAX_FLTR | RX_CLS_LOC_SPECIAL;
+		cmd->data = bp->max_fltr | RX_CLS_LOC_SPECIAL;
 		break;
 
 	case ETHTOOL_GRXCLSRLALL:
-- 
2.30.1


[-- Attachment #2: S/MIME Cryptographic Signature --]
[-- Type: application/pkcs7-signature, Size: 4209 bytes --]

Re: [PATCH net-next 01/13] bnxt_en: Use firmware provided maximum filter counts.

[Michal Swiatkowski]

On Mon, Feb 05, 2024 at 02:31:50PM -0800, Michael Chan wrote:
> While individual filter structures are allocated as needed, there is an
> array to keep track of the software filter IDs that we allocate ahead
> of time.  Rather than relying on a fixed maximum filter count to
> allocate this array, get the maximum from the firmware when available.
> 
> Move these filter related maximum counts queried from the firmware to the
> bnxt_hw_resc struct.  If the firmware is not providing these maximum
> counts, fall back to the hard-coded constant.
> 
> Reviewed-by: Pavan Chebbi <pavan.chebbi@broadcom.com>
> Signed-off-by: Michael Chan <michael.chan@broadcom.com>
> ---
>  drivers/net/ethernet/broadcom/bnxt/bnxt.c     | 26 ++++++++++++-------
>  drivers/net/ethernet/broadcom/bnxt/bnxt.h     | 13 +++++-----
>  .../net/ethernet/broadcom/bnxt/bnxt_ethtool.c |  4 +--
>  3 files changed, 26 insertions(+), 17 deletions(-)
> 

[...]
Reviewed-by: Michal Swiatkowski <michal.swiatkowski@linux.intel.com>

Thanks,
Michal

[PATCH net-next 02/13] bnxt_en: Add ethtool -N support for ether filters.

[Michael Chan]

[-- Attachment #1: Type: text/plain, Size: 5620 bytes --]

Add ETHTOOL_SRXCLSRLINS and ETHTOOL_SRXCLSRLDEL support for inserting
and deleting L2 ether filter rules.  Destination MAC address and
optional VLAN are supported for each filter entry.  This is currently
only supported on older BCM573XX and BCM574XX chips only.

Reviewed-by: Pavan Chebbi <pavan.chebbi@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
---
 drivers/net/ethernet/broadcom/bnxt/bnxt.c     | 34 +++++++++
 drivers/net/ethernet/broadcom/bnxt/bnxt.h     |  3 +
 .../net/ethernet/broadcom/bnxt/bnxt_ethtool.c | 69 ++++++++++++++++++-
 3 files changed, 103 insertions(+), 3 deletions(-)

diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.c b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
index 91ecf514b924..3cc3504181c7 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt.c
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
@@ -5519,6 +5519,40 @@ static struct bnxt_l2_filter *bnxt_alloc_l2_filter(struct bnxt *bp,
 	return fltr;
 }
 
+struct bnxt_l2_filter *bnxt_alloc_new_l2_filter(struct bnxt *bp,
+						struct bnxt_l2_key *key,
+						u16 flags)
+{
+	struct bnxt_l2_filter *fltr;
+	u32 idx;
+	int rc;
+
+	idx = jhash2(&key->filter_key, BNXT_L2_KEY_SIZE, bp->hash_seed) &
+	      BNXT_L2_FLTR_HASH_MASK;
+	spin_lock_bh(&bp->ntp_fltr_lock);
+	fltr = __bnxt_lookup_l2_filter(bp, key, idx);
+	if (fltr) {
+		fltr = ERR_PTR(-EEXIST);
+		goto l2_filter_exit;
+	}
+	fltr = kzalloc(sizeof(*fltr), GFP_ATOMIC);
+	if (!fltr) {
+		fltr = ERR_PTR(-ENOMEM);
+		goto l2_filter_exit;
+	}
+	fltr->base.flags = flags;
+	rc = bnxt_init_l2_filter(bp, fltr, key, idx);
+	if (rc) {
+		spin_unlock_bh(&bp->ntp_fltr_lock);
+		bnxt_del_l2_filter(bp, fltr);
+		return ERR_PTR(rc);
+	}
+
+l2_filter_exit:
+	spin_unlock_bh(&bp->ntp_fltr_lock);
+	return fltr;
+}
+
 static u16 bnxt_vf_target_id(struct bnxt_pf_info *pf, u16 vf_idx)
 {
 #ifdef CONFIG_BNXT_SRIOV
diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.h b/drivers/net/ethernet/broadcom/bnxt/bnxt.h
index 4bd1cf01d99e..21721b8748bc 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt.h
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.h
@@ -2646,6 +2646,9 @@ int bnxt_hwrm_func_drv_rgtr(struct bnxt *bp, unsigned long *bmap,
 			    int bmap_size, bool async_only);
 int bnxt_hwrm_func_drv_unrgtr(struct bnxt *bp);
 void bnxt_del_l2_filter(struct bnxt *bp, struct bnxt_l2_filter *fltr);
+struct bnxt_l2_filter *bnxt_alloc_new_l2_filter(struct bnxt *bp,
+						struct bnxt_l2_key *key,
+						u16 flags);
 int bnxt_hwrm_l2_filter_free(struct bnxt *bp, struct bnxt_l2_filter *fltr);
 int bnxt_hwrm_l2_filter_alloc(struct bnxt *bp, struct bnxt_l2_filter *fltr);
 int bnxt_hwrm_cfa_ntuple_filter_free(struct bnxt *bp,
diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c b/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c
index 1c8610386404..2d8e847e8fdd 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c
@@ -1152,6 +1152,58 @@ static int bnxt_grxclsrule(struct bnxt *bp, struct ethtool_rxnfc *cmd)
 	return rc;
 }
 
+static int bnxt_add_l2_cls_rule(struct bnxt *bp,
+				struct ethtool_rx_flow_spec *fs)
+{
+	u32 ring = ethtool_get_flow_spec_ring(fs->ring_cookie);
+	u8 vf = ethtool_get_flow_spec_ring_vf(fs->ring_cookie);
+	struct ethhdr *h_ether = &fs->h_u.ether_spec;
+	struct ethhdr *m_ether = &fs->m_u.ether_spec;
+	struct bnxt_l2_filter *fltr;
+	struct bnxt_l2_key key;
+	u16 vnic_id;
+	u8 flags;
+	int rc;
+
+	if (BNXT_CHIP_P5_PLUS(bp))
+		return -EOPNOTSUPP;
+
+	if (!is_broadcast_ether_addr(m_ether->h_dest))
+		return -EINVAL;
+	ether_addr_copy(key.dst_mac_addr, h_ether->h_dest);
+	key.vlan = 0;
+	if (fs->flow_type & FLOW_EXT) {
+		struct ethtool_flow_ext *m_ext = &fs->m_ext;
+		struct ethtool_flow_ext *h_ext = &fs->h_ext;
+
+		if (m_ext->vlan_tci != htons(0xfff) || !h_ext->vlan_tci)
+			return -EINVAL;
+		key.vlan = ntohs(h_ext->vlan_tci);
+	}
+
+	if (vf) {
+		flags = BNXT_ACT_FUNC_DST;
+		vnic_id = 0xffff;
+		vf--;
+	} else {
+		flags = BNXT_ACT_RING_DST;
+		vnic_id = bp->vnic_info[ring + 1].fw_vnic_id;
+	}
+	fltr = bnxt_alloc_new_l2_filter(bp, &key, flags);
+	if (IS_ERR(fltr))
+		return PTR_ERR(fltr);
+
+	fltr->base.fw_vnic_id = vnic_id;
+	fltr->base.rxq = ring;
+	fltr->base.vf_idx = vf;
+	rc = bnxt_hwrm_l2_filter_alloc(bp, fltr);
+	if (rc)
+		bnxt_del_l2_filter(bp, fltr);
+	else
+		fs->location = fltr->base.sw_id;
+	return rc;
+}
+
 #define IPV4_ALL_MASK		((__force __be32)~0)
 #define L4_PORT_ALL_MASK	((__force __be16)~0)
 
@@ -1335,7 +1387,7 @@ static int bnxt_srxclsrlins(struct bnxt *bp, struct ethtool_rxnfc *cmd)
 		return -EINVAL;
 	flow_type &= ~FLOW_EXT;
 	if (flow_type == ETHER_FLOW)
-		rc = -EOPNOTSUPP;
+		rc = bnxt_add_l2_cls_rule(bp, fs);
 	else
 		rc = bnxt_add_ntuple_cls_rule(bp, fs);
 	return rc;
@@ -1346,11 +1398,22 @@ static int bnxt_srxclsrldel(struct bnxt *bp, struct ethtool_rxnfc *cmd)
 	struct ethtool_rx_flow_spec *fs = &cmd->fs;
 	struct bnxt_filter_base *fltr_base;
 	struct bnxt_ntuple_filter *fltr;
+	u32 id = fs->location;
 
 	rcu_read_lock();
+	fltr_base = bnxt_get_one_fltr_rcu(bp, bp->l2_fltr_hash_tbl,
+					  BNXT_L2_FLTR_HASH_SIZE, id);
+	if (fltr_base) {
+		struct bnxt_l2_filter *l2_fltr;
+
+		l2_fltr = container_of(fltr_base, struct bnxt_l2_filter, base);
+		rcu_read_unlock();
+		bnxt_hwrm_l2_filter_free(bp, l2_fltr);
+		bnxt_del_l2_filter(bp, l2_fltr);
+		return 0;
+	}
 	fltr_base = bnxt_get_one_fltr_rcu(bp, bp->ntp_fltr_hash_tbl,
-					  BNXT_NTP_FLTR_HASH_SIZE,
-					  fs->location);
+					  BNXT_NTP_FLTR_HASH_SIZE, id);
 	if (!fltr_base) {
 		rcu_read_unlock();
 		return -ENOENT;
-- 
2.30.1


[-- Attachment #2: S/MIME Cryptographic Signature --]
[-- Type: application/pkcs7-signature, Size: 4209 bytes --]

Re: [PATCH net-next 02/13] bnxt_en: Add ethtool -N support for ether filters.

[Michal Swiatkowski]

On Mon, Feb 05, 2024 at 02:31:51PM -0800, Michael Chan wrote:
> Add ETHTOOL_SRXCLSRLINS and ETHTOOL_SRXCLSRLDEL support for inserting
> and deleting L2 ether filter rules.  Destination MAC address and
> optional VLAN are supported for each filter entry.  This is currently
> only supported on older BCM573XX and BCM574XX chips only.
> 
> Reviewed-by: Pavan Chebbi <pavan.chebbi@broadcom.com>
> Signed-off-by: Michael Chan <michael.chan@broadcom.com>
> ---
>  drivers/net/ethernet/broadcom/bnxt/bnxt.c     | 34 +++++++++
>  drivers/net/ethernet/broadcom/bnxt/bnxt.h     |  3 +
>  .../net/ethernet/broadcom/bnxt/bnxt_ethtool.c | 69 ++++++++++++++++++-
>  3 files changed, 103 insertions(+), 3 deletions(-)
> 
> diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.c b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
> index 91ecf514b924..3cc3504181c7 100644
> --- a/drivers/net/ethernet/broadcom/bnxt/bnxt.c
> +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
> @@ -5519,6 +5519,40 @@ static struct bnxt_l2_filter *bnxt_alloc_l2_filter(struct bnxt *bp,
>  	return fltr;
>  }
>  
> +struct bnxt_l2_filter *bnxt_alloc_new_l2_filter(struct bnxt *bp,
> +						struct bnxt_l2_key *key,
> +						u16 flags)
> +{
> +	struct bnxt_l2_filter *fltr;
> +	u32 idx;
> +	int rc;
> +
> +	idx = jhash2(&key->filter_key, BNXT_L2_KEY_SIZE, bp->hash_seed) &
> +	      BNXT_L2_FLTR_HASH_MASK;
> +	spin_lock_bh(&bp->ntp_fltr_lock);
> +	fltr = __bnxt_lookup_l2_filter(bp, key, idx);
> +	if (fltr) {
> +		fltr = ERR_PTR(-EEXIST);
> +		goto l2_filter_exit;
> +	}
> +	fltr = kzalloc(sizeof(*fltr), GFP_ATOMIC);
> +	if (!fltr) {
> +		fltr = ERR_PTR(-ENOMEM);
> +		goto l2_filter_exit;
> +	}
> +	fltr->base.flags = flags;
> +	rc = bnxt_init_l2_filter(bp, fltr, key, idx);
> +	if (rc) {
> +		spin_unlock_bh(&bp->ntp_fltr_lock);
Why filter needs to be deleted without lock? If you can change the order
it looks more natural:

+if (rc) {
+	fltr = ERR_PTR(rc);
+	goto l2_filter_del;
+}

> +		bnxt_del_l2_filter(bp, fltr);
> +		return ERR_PTR(rc);
> +	}
> +
> +l2_filter_exit:
> +	spin_unlock_bh(&bp->ntp_fltr_lock);
> +	return fltr;
> +}
> +
>  static u16 bnxt_vf_target_id(struct bnxt_pf_info *pf, u16 vf_idx)
>  {
>  #ifdef CONFIG_BNXT_SRIOV
> diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.h b/drivers/net/ethernet/broadcom/bnxt/bnxt.h
> index 4bd1cf01d99e..21721b8748bc 100644
> --- a/drivers/net/ethernet/broadcom/bnxt/bnxt.h
> +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.h
> @@ -2646,6 +2646,9 @@ int bnxt_hwrm_func_drv_rgtr(struct bnxt *bp, unsigned long *bmap,
>  			    int bmap_size, bool async_only);
>  int bnxt_hwrm_func_drv_unrgtr(struct bnxt *bp);
>  void bnxt_del_l2_filter(struct bnxt *bp, struct bnxt_l2_filter *fltr);
> +struct bnxt_l2_filter *bnxt_alloc_new_l2_filter(struct bnxt *bp,
> +						struct bnxt_l2_key *key,
> +						u16 flags);
>  int bnxt_hwrm_l2_filter_free(struct bnxt *bp, struct bnxt_l2_filter *fltr);
>  int bnxt_hwrm_l2_filter_alloc(struct bnxt *bp, struct bnxt_l2_filter *fltr);
>  int bnxt_hwrm_cfa_ntuple_filter_free(struct bnxt *bp,
> diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c b/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c
> index 1c8610386404..2d8e847e8fdd 100644
> --- a/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c
> +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c
> @@ -1152,6 +1152,58 @@ static int bnxt_grxclsrule(struct bnxt *bp, struct ethtool_rxnfc *cmd)
>  	return rc;
>  }
>  
> +static int bnxt_add_l2_cls_rule(struct bnxt *bp,
> +				struct ethtool_rx_flow_spec *fs)
> +{
> +	u32 ring = ethtool_get_flow_spec_ring(fs->ring_cookie);
> +	u8 vf = ethtool_get_flow_spec_ring_vf(fs->ring_cookie);
> +	struct ethhdr *h_ether = &fs->h_u.ether_spec;
> +	struct ethhdr *m_ether = &fs->m_u.ether_spec;
> +	struct bnxt_l2_filter *fltr;
> +	struct bnxt_l2_key key;
> +	u16 vnic_id;
> +	u8 flags;
> +	int rc;
> +
> +	if (BNXT_CHIP_P5_PLUS(bp))
> +		return -EOPNOTSUPP;
> +
> +	if (!is_broadcast_ether_addr(m_ether->h_dest))
> +		return -EINVAL;
> +	ether_addr_copy(key.dst_mac_addr, h_ether->h_dest);
> +	key.vlan = 0;
> +	if (fs->flow_type & FLOW_EXT) {
> +		struct ethtool_flow_ext *m_ext = &fs->m_ext;
> +		struct ethtool_flow_ext *h_ext = &fs->h_ext;
> +
> +		if (m_ext->vlan_tci != htons(0xfff) || !h_ext->vlan_tci)
> +			return -EINVAL;
> +		key.vlan = ntohs(h_ext->vlan_tci);
> +	}
> +
> +	if (vf) {
> +		flags = BNXT_ACT_FUNC_DST;
> +		vnic_id = 0xffff;
> +		vf--;
> +	} else {
> +		flags = BNXT_ACT_RING_DST;
> +		vnic_id = bp->vnic_info[ring + 1].fw_vnic_id;
> +	}
> +	fltr = bnxt_alloc_new_l2_filter(bp, &key, flags);
> +	if (IS_ERR(fltr))
> +		return PTR_ERR(fltr);
> +
> +	fltr->base.fw_vnic_id = vnic_id;
> +	fltr->base.rxq = ring;
> +	fltr->base.vf_idx = vf;
> +	rc = bnxt_hwrm_l2_filter_alloc(bp, fltr);
> +	if (rc)
> +		bnxt_del_l2_filter(bp, fltr);
> +	else
> +		fs->location = fltr->base.sw_id;
> +	return rc;
> +}
> +
>  #define IPV4_ALL_MASK		((__force __be32)~0)
>  #define L4_PORT_ALL_MASK	((__force __be16)~0)
>  
> @@ -1335,7 +1387,7 @@ static int bnxt_srxclsrlins(struct bnxt *bp, struct ethtool_rxnfc *cmd)
>  		return -EINVAL;
>  	flow_type &= ~FLOW_EXT;
>  	if (flow_type == ETHER_FLOW)
> -		rc = -EOPNOTSUPP;
> +		rc = bnxt_add_l2_cls_rule(bp, fs);
>  	else
>  		rc = bnxt_add_ntuple_cls_rule(bp, fs);
>  	return rc;
> @@ -1346,11 +1398,22 @@ static int bnxt_srxclsrldel(struct bnxt *bp, struct ethtool_rxnfc *cmd)
>  	struct ethtool_rx_flow_spec *fs = &cmd->fs;
>  	struct bnxt_filter_base *fltr_base;
>  	struct bnxt_ntuple_filter *fltr;
> +	u32 id = fs->location;
>  
>  	rcu_read_lock();
> +	fltr_base = bnxt_get_one_fltr_rcu(bp, bp->l2_fltr_hash_tbl,
> +					  BNXT_L2_FLTR_HASH_SIZE, id);
> +	if (fltr_base) {
> +		struct bnxt_l2_filter *l2_fltr;
> +
> +		l2_fltr = container_of(fltr_base, struct bnxt_l2_filter, base);
> +		rcu_read_unlock();
> +		bnxt_hwrm_l2_filter_free(bp, l2_fltr);
> +		bnxt_del_l2_filter(bp, l2_fltr);
> +		return 0;
> +	}
>  	fltr_base = bnxt_get_one_fltr_rcu(bp, bp->ntp_fltr_hash_tbl,
> -					  BNXT_NTP_FLTR_HASH_SIZE,
> -					  fs->location);
> +					  BNXT_NTP_FLTR_HASH_SIZE, id);
>  	if (!fltr_base) {
>  		rcu_read_unlock();
>  		return -ENOENT;
> -- 
> 2.30.1
> 

Re: [PATCH net-next 02/13] bnxt_en: Add ethtool -N support for ether filters.

[Michael Chan]

[-- Attachment #1: Type: text/plain, Size: 1103 bytes --]

On Mon, Feb 5, 2024 at 11:04 PM Michal Swiatkowski
<michal.swiatkowski@linux.intel.com> wrote:
>
> On Mon, Feb 05, 2024 at 02:31:51PM -0800, Michael Chan wrote:
> > +     spin_lock_bh(&bp->ntp_fltr_lock);
> > +     fltr = __bnxt_lookup_l2_filter(bp, key, idx);
> > +     if (fltr) {
> > +             fltr = ERR_PTR(-EEXIST);
> > +             goto l2_filter_exit;
> > +     }
> > +     fltr = kzalloc(sizeof(*fltr), GFP_ATOMIC);
> > +     if (!fltr) {
> > +             fltr = ERR_PTR(-ENOMEM);
> > +             goto l2_filter_exit;
> > +     }
> > +     fltr->base.flags = flags;
> > +     rc = bnxt_init_l2_filter(bp, fltr, key, idx);
> > +     if (rc) {
> > +             spin_unlock_bh(&bp->ntp_fltr_lock);
> Why filter needs to be deleted without lock? If you can change the order
> it looks more natural:
>
> +if (rc) {
> +       fltr = ERR_PTR(rc);
> +       goto l2_filter_del;
> +}

Thanks for the review.  bnxt_del_l2_filter() will take the same lock
inside the function if it goes ahead to delete the filter.  That's why
the lock needs to be released first.

[-- Attachment #2: S/MIME Cryptographic Signature --]
[-- Type: application/pkcs7-signature, Size: 4209 bytes --]

Re: [PATCH net-next 02/13] bnxt_en: Add ethtool -N support for ether filters.

[Michal Swiatkowski]

On Tue, Feb 06, 2024 at 12:10:32AM -0800, Michael Chan wrote:
> On Mon, Feb 5, 2024 at 11:04 PM Michal Swiatkowski
> <michal.swiatkowski@linux.intel.com> wrote:
> >
> > On Mon, Feb 05, 2024 at 02:31:51PM -0800, Michael Chan wrote:
> > > +     spin_lock_bh(&bp->ntp_fltr_lock);
> > > +     fltr = __bnxt_lookup_l2_filter(bp, key, idx);
> > > +     if (fltr) {
> > > +             fltr = ERR_PTR(-EEXIST);
> > > +             goto l2_filter_exit;
> > > +     }
> > > +     fltr = kzalloc(sizeof(*fltr), GFP_ATOMIC);
> > > +     if (!fltr) {
> > > +             fltr = ERR_PTR(-ENOMEM);
> > > +             goto l2_filter_exit;
> > > +     }
> > > +     fltr->base.flags = flags;
> > > +     rc = bnxt_init_l2_filter(bp, fltr, key, idx);
> > > +     if (rc) {
> > > +             spin_unlock_bh(&bp->ntp_fltr_lock);
> > Why filter needs to be deleted without lock? If you can change the order
> > it looks more natural:
> >
> > +if (rc) {
> > +       fltr = ERR_PTR(rc);
> > +       goto l2_filter_del;
> > +}
> 
> Thanks for the review.  bnxt_del_l2_filter() will take the same lock
> inside the function if it goes ahead to delete the filter.  That's why
> the lock needs to be released first.

Got it, thanks.

[PATCH net-next 03/13] bnxt_en: Support ethtool -n to display ether filters.

[Michael Chan]

[-- Attachment #1: Type: text/plain, Size: 3120 bytes --]

Implement ETHTOOL_GRXCLSRULE for the user defined ether filters.  Use
the common functions to walk the L2 filter hash table.

Reviewed-by: Pavan Chebbi <pavan.chebbi@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
---
 drivers/net/ethernet/broadcom/bnxt/bnxt.c     |  1 +
 .../net/ethernet/broadcom/bnxt/bnxt_ethtool.c | 38 ++++++++++++++++++-
 2 files changed, 38 insertions(+), 1 deletion(-)

diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.c b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
index 3cc3504181c7..da298f4512b5 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt.c
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
@@ -5484,6 +5484,7 @@ static int bnxt_init_l2_filter(struct bnxt *bp, struct bnxt_l2_filter *fltr,
 		if (bit_id < 0)
 			return -ENOMEM;
 		fltr->base.sw_id = (u16)bit_id;
+		bp->ntp_fltr_count++;
 	}
 	head = &bp->l2_fltr_hash_tbl[idx];
 	hlist_add_head_rcu(&fltr->base.hash, head);
diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c b/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c
index 2d8e847e8fdd..4d4dd2b231b8 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c
@@ -1058,11 +1058,17 @@ static struct bnxt_filter_base *bnxt_get_one_fltr_rcu(struct bnxt *bp,
 static int bnxt_grxclsrlall(struct bnxt *bp, struct ethtool_rxnfc *cmd,
 			    u32 *rule_locs)
 {
+	u32 count;
+
 	cmd->data = bp->ntp_fltr_count;
 	rcu_read_lock();
+	count = bnxt_get_all_fltr_ids_rcu(bp, bp->l2_fltr_hash_tbl,
+					  BNXT_L2_FLTR_HASH_SIZE, rule_locs, 0,
+					  cmd->rule_cnt);
 	cmd->rule_cnt = bnxt_get_all_fltr_ids_rcu(bp, bp->ntp_fltr_hash_tbl,
 						  BNXT_NTP_FLTR_HASH_SIZE,
-						  rule_locs, 0, cmd->rule_cnt);
+						  rule_locs, count,
+						  cmd->rule_cnt);
 	rcu_read_unlock();
 
 	return 0;
@@ -1081,6 +1087,36 @@ static int bnxt_grxclsrule(struct bnxt *bp, struct ethtool_rxnfc *cmd)
 		return rc;
 
 	rcu_read_lock();
+	fltr_base = bnxt_get_one_fltr_rcu(bp, bp->l2_fltr_hash_tbl,
+					  BNXT_L2_FLTR_HASH_SIZE,
+					  fs->location);
+	if (fltr_base) {
+		struct ethhdr *h_ether = &fs->h_u.ether_spec;
+		struct ethhdr *m_ether = &fs->m_u.ether_spec;
+		struct bnxt_l2_filter *l2_fltr;
+		struct bnxt_l2_key *l2_key;
+
+		l2_fltr = container_of(fltr_base, struct bnxt_l2_filter, base);
+		l2_key = &l2_fltr->l2_key;
+		fs->flow_type = ETHER_FLOW;
+		ether_addr_copy(h_ether->h_dest, l2_key->dst_mac_addr);
+		eth_broadcast_addr(m_ether->h_dest);
+		if (l2_key->vlan) {
+			struct ethtool_flow_ext *m_ext = &fs->m_ext;
+			struct ethtool_flow_ext *h_ext = &fs->h_ext;
+
+			fs->flow_type |= FLOW_EXT;
+			m_ext->vlan_tci = htons(0xfff);
+			h_ext->vlan_tci = htons(l2_key->vlan);
+		}
+		if (fltr_base->flags & BNXT_ACT_RING_DST)
+			fs->ring_cookie = fltr_base->rxq;
+		if (fltr_base->flags & BNXT_ACT_FUNC_DST)
+			fs->ring_cookie = (u64)(fltr_base->vf_idx + 1) <<
+					  ETHTOOL_RX_FLOW_SPEC_RING_VF_OFF;
+		rcu_read_unlock();
+		return 0;
+	}
 	fltr_base = bnxt_get_one_fltr_rcu(bp, bp->ntp_fltr_hash_tbl,
 					  BNXT_NTP_FLTR_HASH_SIZE,
 					  fs->location);
-- 
2.30.1


[-- Attachment #2: S/MIME Cryptographic Signature --]
[-- Type: application/pkcs7-signature, Size: 4209 bytes --]

Re: [PATCH net-next 03/13] bnxt_en: Support ethtool -n to display ether filters.

[Michal Swiatkowski]

On Mon, Feb 05, 2024 at 02:31:52PM -0800, Michael Chan wrote:
> Implement ETHTOOL_GRXCLSRULE for the user defined ether filters.  Use
> the common functions to walk the L2 filter hash table.
> 
> Reviewed-by: Pavan Chebbi <pavan.chebbi@broadcom.com>
> Signed-off-by: Michael Chan <michael.chan@broadcom.com>
> ---
>  drivers/net/ethernet/broadcom/bnxt/bnxt.c     |  1 +
>  .../net/ethernet/broadcom/bnxt/bnxt_ethtool.c | 38 ++++++++++++++++++-
>  2 files changed, 38 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.c b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
> index 3cc3504181c7..da298f4512b5 100644
> --- a/drivers/net/ethernet/broadcom/bnxt/bnxt.c
> +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
> @@ -5484,6 +5484,7 @@ static int bnxt_init_l2_filter(struct bnxt *bp, struct bnxt_l2_filter *fltr,
>  		if (bit_id < 0)
>  			return -ENOMEM;
>  		fltr->base.sw_id = (u16)bit_id;
> +		bp->ntp_fltr_count++;
>  	}
>  	head = &bp->l2_fltr_hash_tbl[idx];
>  	hlist_add_head_rcu(&fltr->base.hash, head);
> diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c b/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c
> index 2d8e847e8fdd..4d4dd2b231b8 100644
> --- a/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c
> +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c
> @@ -1058,11 +1058,17 @@ static struct bnxt_filter_base *bnxt_get_one_fltr_rcu(struct bnxt *bp,
>  static int bnxt_grxclsrlall(struct bnxt *bp, struct ethtool_rxnfc *cmd,
>  			    u32 *rule_locs)
>  {
> +	u32 count;
> +
>  	cmd->data = bp->ntp_fltr_count;
>  	rcu_read_lock();
> +	count = bnxt_get_all_fltr_ids_rcu(bp, bp->l2_fltr_hash_tbl,
> +					  BNXT_L2_FLTR_HASH_SIZE, rule_locs, 0,
> +					  cmd->rule_cnt);
>  	cmd->rule_cnt = bnxt_get_all_fltr_ids_rcu(bp, bp->ntp_fltr_hash_tbl,
>  						  BNXT_NTP_FLTR_HASH_SIZE,
> -						  rule_locs, 0, cmd->rule_cnt);
> +						  rule_locs, count,
> +						  cmd->rule_cnt);
>  	rcu_read_unlock();
>  
>  	return 0;
> @@ -1081,6 +1087,36 @@ static int bnxt_grxclsrule(struct bnxt *bp, struct ethtool_rxnfc *cmd)
>  		return rc;
>  
>  	rcu_read_lock();
> +	fltr_base = bnxt_get_one_fltr_rcu(bp, bp->l2_fltr_hash_tbl,
> +					  BNXT_L2_FLTR_HASH_SIZE,
> +					  fs->location);
> +	if (fltr_base) {
> +		struct ethhdr *h_ether = &fs->h_u.ether_spec;
> +		struct ethhdr *m_ether = &fs->m_u.ether_spec;
> +		struct bnxt_l2_filter *l2_fltr;
> +		struct bnxt_l2_key *l2_key;
> +
> +		l2_fltr = container_of(fltr_base, struct bnxt_l2_filter, base);
> +		l2_key = &l2_fltr->l2_key;
> +		fs->flow_type = ETHER_FLOW;
> +		ether_addr_copy(h_ether->h_dest, l2_key->dst_mac_addr);
> +		eth_broadcast_addr(m_ether->h_dest);
> +		if (l2_key->vlan) {
> +			struct ethtool_flow_ext *m_ext = &fs->m_ext;
> +			struct ethtool_flow_ext *h_ext = &fs->h_ext;
> +
> +			fs->flow_type |= FLOW_EXT;
> +			m_ext->vlan_tci = htons(0xfff);
> +			h_ext->vlan_tci = htons(l2_key->vlan);
> +		}
> +		if (fltr_base->flags & BNXT_ACT_RING_DST)
> +			fs->ring_cookie = fltr_base->rxq;
> +		if (fltr_base->flags & BNXT_ACT_FUNC_DST)
> +			fs->ring_cookie = (u64)(fltr_base->vf_idx + 1) <<
> +					  ETHTOOL_RX_FLOW_SPEC_RING_VF_OFF;
Wonder if saving ring_cookie in fltr_base won't be cleaner, but I am not
sure if it is worth to introduce new field for only that.

> +		rcu_read_unlock();
> +		return 0;
> +	}
>  	fltr_base = bnxt_get_one_fltr_rcu(bp, bp->ntp_fltr_hash_tbl,
>  					  BNXT_NTP_FLTR_HASH_SIZE,
>  					  fs->location);

Reviewed-by: Michal Swiatkowski <michal.swiatkowski@linux.intel.com>

Thanks,
Michal
> -- 
> 2.30.1
> 

[PATCH net-next 04/13] bnxt_en: implement fully specified 5-tuple masks

[Michael Chan]

[-- Attachment #1: Type: text/plain, Size: 16988 bytes --]

From: Edwin Peer <edwin.peer@broadcom.com>

Support subfield masking for IP addresses and ports. Previously, only
entire fields could be included or excluded in NTUPLE filters.

Reviewed-by: Pavan Chebbi <pavan.chebbi@broadcom.com>
Signed-off-by: Edwin Peer <edwin.peer@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
---
 drivers/net/ethernet/broadcom/bnxt/bnxt.c     | 141 ++++++++++-------
 drivers/net/ethernet/broadcom/bnxt/bnxt.h     |  19 +--
 .../net/ethernet/broadcom/bnxt/bnxt_ethtool.c | 142 +++++-------------
 3 files changed, 134 insertions(+), 168 deletions(-)

diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.c b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
index da298f4512b5..dc54cc0ab075 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt.c
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
@@ -246,6 +246,49 @@ static const u16 bnxt_async_events_arr[] = {
 
 static struct workqueue_struct *bnxt_pf_wq;
 
+#define BNXT_IPV6_MASK_ALL {{{ 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, \
+			       0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff }}}
+#define BNXT_IPV6_MASK_NONE {{{ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 }}}
+
+const struct bnxt_flow_masks BNXT_FLOW_MASK_NONE = {
+	.ports = {
+		.src = 0,
+		.dst = 0,
+	},
+	.addrs = {
+		.v6addrs = {
+			.src = BNXT_IPV6_MASK_NONE,
+			.dst = BNXT_IPV6_MASK_NONE,
+		},
+	},
+};
+
+const struct bnxt_flow_masks BNXT_FLOW_IPV6_MASK_ALL = {
+	.ports = {
+		.src = cpu_to_be16(0xffff),
+		.dst = cpu_to_be16(0xffff),
+	},
+	.addrs = {
+		.v6addrs = {
+			.src = BNXT_IPV6_MASK_ALL,
+			.dst = BNXT_IPV6_MASK_ALL,
+		},
+	},
+};
+
+const struct bnxt_flow_masks BNXT_FLOW_IPV4_MASK_ALL = {
+	.ports = {
+		.src = cpu_to_be16(0xffff),
+		.dst = cpu_to_be16(0xffff),
+	},
+	.addrs = {
+		.v4addrs = {
+			.src = cpu_to_be32(0xffffffff),
+			.dst = cpu_to_be32(0xffffffff),
+		},
+	},
+};
+
 static bool bnxt_vf_pciid(enum board_idx idx)
 {
 	return (idx == NETXTREME_C_VF || idx == NETXTREME_E_VF ||
@@ -5690,6 +5733,7 @@ int bnxt_hwrm_cfa_ntuple_filter_alloc(struct bnxt *bp,
 {
 	struct hwrm_cfa_ntuple_filter_alloc_output *resp;
 	struct hwrm_cfa_ntuple_filter_alloc_input *req;
+	struct bnxt_flow_masks *masks = &fltr->fmasks;
 	struct flow_keys *keys = &fltr->fkeys;
 	struct bnxt_l2_filter *l2_fltr;
 	struct bnxt_vnic_info *vnic;
@@ -5722,25 +5766,15 @@ int bnxt_hwrm_cfa_ntuple_filter_alloc(struct bnxt *bp,
 		req->ethertype = htons(ETH_P_IPV6);
 		req->ip_addr_type =
 			CFA_NTUPLE_FILTER_ALLOC_REQ_IP_ADDR_TYPE_IPV6;
-		if (fltr->ntuple_flags & BNXT_NTUPLE_MATCH_SRC_IP) {
-			*(struct in6_addr *)&req->src_ipaddr[0] =
-				keys->addrs.v6addrs.src;
-			bnxt_fill_ipv6_mask(req->src_ipaddr_mask);
-		}
-		if (fltr->ntuple_flags & BNXT_NTUPLE_MATCH_DST_IP) {
-			*(struct in6_addr *)&req->dst_ipaddr[0] =
-				keys->addrs.v6addrs.dst;
-			bnxt_fill_ipv6_mask(req->dst_ipaddr_mask);
-		}
+		*(struct in6_addr *)&req->src_ipaddr[0] = keys->addrs.v6addrs.src;
+		*(struct in6_addr *)&req->src_ipaddr_mask[0] = masks->addrs.v6addrs.src;
+		*(struct in6_addr *)&req->dst_ipaddr[0] = keys->addrs.v6addrs.dst;
+		*(struct in6_addr *)&req->dst_ipaddr_mask[0] = masks->addrs.v6addrs.dst;
 	} else {
-		if (fltr->ntuple_flags & BNXT_NTUPLE_MATCH_SRC_IP) {
-			req->src_ipaddr[0] = keys->addrs.v4addrs.src;
-			req->src_ipaddr_mask[0] = cpu_to_be32(0xffffffff);
-		}
-		if (fltr->ntuple_flags & BNXT_NTUPLE_MATCH_DST_IP) {
-			req->dst_ipaddr[0] = keys->addrs.v4addrs.dst;
-			req->dst_ipaddr_mask[0] = cpu_to_be32(0xffffffff);
-		}
+		req->src_ipaddr[0] = keys->addrs.v4addrs.src;
+		req->src_ipaddr_mask[0] = masks->addrs.v4addrs.src;
+		req->dst_ipaddr[0] = keys->addrs.v4addrs.dst;
+		req->dst_ipaddr_mask[0] = masks->addrs.v4addrs.dst;
 	}
 	if (keys->control.flags & FLOW_DIS_ENCAPSULATION) {
 		req->enables |= cpu_to_le32(BNXT_NTP_TUNNEL_FLTR_FLAG);
@@ -5748,14 +5782,10 @@ int bnxt_hwrm_cfa_ntuple_filter_alloc(struct bnxt *bp,
 			CFA_NTUPLE_FILTER_ALLOC_REQ_TUNNEL_TYPE_ANYTUNNEL;
 	}
 
-	if (fltr->ntuple_flags & BNXT_NTUPLE_MATCH_SRC_PORT) {
-		req->src_port = keys->ports.src;
-		req->src_port_mask = cpu_to_be16(0xffff);
-	}
-	if (fltr->ntuple_flags & BNXT_NTUPLE_MATCH_DST_PORT) {
-		req->dst_port = keys->ports.dst;
-		req->dst_port_mask = cpu_to_be16(0xffff);
-	}
+	req->src_port = keys->ports.src;
+	req->src_port_mask = masks->ports.src;
+	req->dst_port = keys->ports.dst;
+	req->dst_port_mask = masks->ports.dst;
 
 	resp = hwrm_req_hold(bp, req);
 	rc = hwrm_req_send(bp, req);
@@ -13955,45 +13985,39 @@ int bnxt_insert_ntp_filter(struct bnxt *bp, struct bnxt_ntuple_filter *fltr,
 static bool bnxt_fltr_match(struct bnxt_ntuple_filter *f1,
 			    struct bnxt_ntuple_filter *f2)
 {
+	struct bnxt_flow_masks *masks1 = &f1->fmasks;
+	struct bnxt_flow_masks *masks2 = &f2->fmasks;
 	struct flow_keys *keys1 = &f1->fkeys;
 	struct flow_keys *keys2 = &f2->fkeys;
 
-	if (f1->ntuple_flags != f2->ntuple_flags)
-		return false;
-
 	if (keys1->basic.n_proto != keys2->basic.n_proto ||
 	    keys1->basic.ip_proto != keys2->basic.ip_proto)
 		return false;
 
 	if (keys1->basic.n_proto == htons(ETH_P_IP)) {
-		if (((f1->ntuple_flags & BNXT_NTUPLE_MATCH_SRC_IP) &&
-		     keys1->addrs.v4addrs.src != keys2->addrs.v4addrs.src) ||
-		    ((f1->ntuple_flags & BNXT_NTUPLE_MATCH_DST_IP) &&
-		     keys1->addrs.v4addrs.dst != keys2->addrs.v4addrs.dst))
+		if (keys1->addrs.v4addrs.src != keys2->addrs.v4addrs.src ||
+		    masks1->addrs.v4addrs.src != masks2->addrs.v4addrs.src ||
+		    keys1->addrs.v4addrs.dst != keys2->addrs.v4addrs.dst ||
+		    masks1->addrs.v4addrs.dst != masks2->addrs.v4addrs.dst)
 			return false;
 	} else {
-		if (((f1->ntuple_flags & BNXT_NTUPLE_MATCH_SRC_IP) &&
-		     memcmp(&keys1->addrs.v6addrs.src,
-			    &keys2->addrs.v6addrs.src,
-			    sizeof(keys1->addrs.v6addrs.src))) ||
-		    ((f1->ntuple_flags & BNXT_NTUPLE_MATCH_DST_IP) &&
-		     memcmp(&keys1->addrs.v6addrs.dst,
-			    &keys2->addrs.v6addrs.dst,
-			    sizeof(keys1->addrs.v6addrs.dst))))
+		if (!ipv6_addr_equal(&keys1->addrs.v6addrs.src,
+				     &keys2->addrs.v6addrs.src) ||
+		    !ipv6_addr_equal(&masks1->addrs.v6addrs.src,
+				     &masks2->addrs.v6addrs.src) ||
+		    !ipv6_addr_equal(&keys1->addrs.v6addrs.dst,
+				     &keys2->addrs.v6addrs.dst) ||
+		    !ipv6_addr_equal(&masks1->addrs.v6addrs.dst,
+				     &masks2->addrs.v6addrs.dst))
 			return false;
 	}
 
-	if (((f1->ntuple_flags & BNXT_NTUPLE_MATCH_SRC_PORT) &&
-	     keys1->ports.src != keys2->ports.src) ||
-	    ((f1->ntuple_flags & BNXT_NTUPLE_MATCH_DST_PORT) &&
-	     keys1->ports.dst != keys2->ports.dst))
-		return false;
-
-	if (keys1->control.flags == keys2->control.flags &&
-	    f1->l2_fltr == f2->l2_fltr)
-		return true;
-
-	return false;
+	return keys1->ports.src == keys2->ports.src &&
+	       masks1->ports.src == masks2->ports.src &&
+	       keys1->ports.dst == keys2->ports.dst &&
+	       masks1->ports.dst == masks2->ports.dst &&
+	       keys1->control.flags == keys2->control.flags &&
+	       f1->l2_fltr == f2->l2_fltr;
 }
 
 struct bnxt_ntuple_filter *
@@ -14058,10 +14082,13 @@ static int bnxt_rx_flow_steer(struct net_device *dev, const struct sk_buff *skb,
 		rc = -EPROTONOSUPPORT;
 		goto err_free;
 	}
-	if (fkeys->basic.n_proto == htons(ETH_P_IPV6) &&
-	    bp->hwrm_spec_code < 0x10601) {
-		rc = -EPROTONOSUPPORT;
-		goto err_free;
+	new_fltr->fmasks = BNXT_FLOW_IPV4_MASK_ALL;
+	if (fkeys->basic.n_proto == htons(ETH_P_IPV6)) {
+		if (bp->hwrm_spec_code < 0x10601) {
+			rc = -EPROTONOSUPPORT;
+			goto err_free;
+		}
+		new_fltr->fmasks = BNXT_FLOW_IPV6_MASK_ALL;
 	}
 	flags = fkeys->control.flags;
 	if (((flags & FLOW_DIS_ENCAPSULATION) &&
@@ -14069,9 +14096,7 @@ static int bnxt_rx_flow_steer(struct net_device *dev, const struct sk_buff *skb,
 		rc = -EPROTONOSUPPORT;
 		goto err_free;
 	}
-
 	new_fltr->l2_fltr = l2_fltr;
-	new_fltr->ntuple_flags = BNXT_NTUPLE_MATCH_ALL;
 
 	idx = bnxt_get_ntp_filter_idx(bp, fkeys, skb);
 	rcu_read_lock();
diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.h b/drivers/net/ethernet/broadcom/bnxt/bnxt.h
index 21721b8748bc..aae180fa63b7 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt.h
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.h
@@ -1355,19 +1355,20 @@ struct bnxt_filter_base {
 	struct rcu_head         rcu;
 };
 
+struct bnxt_flow_masks {
+	struct flow_dissector_key_ports ports;
+	struct flow_dissector_key_addrs addrs;
+};
+
+extern const struct bnxt_flow_masks BNXT_FLOW_MASK_NONE;
+extern const struct bnxt_flow_masks BNXT_FLOW_IPV6_MASK_ALL;
+extern const struct bnxt_flow_masks BNXT_FLOW_IPV4_MASK_ALL;
+
 struct bnxt_ntuple_filter {
 	struct bnxt_filter_base	base;
 	struct flow_keys	fkeys;
+	struct bnxt_flow_masks	fmasks;
 	struct bnxt_l2_filter	*l2_fltr;
-	u32			ntuple_flags;
-#define BNXT_NTUPLE_MATCH_SRC_IP	1
-#define BNXT_NTUPLE_MATCH_DST_IP	2
-#define BNXT_NTUPLE_MATCH_SRC_PORT	4
-#define BNXT_NTUPLE_MATCH_DST_PORT	8
-#define BNXT_NTUPLE_MATCH_ALL		(BNXT_NTUPLE_MATCH_SRC_IP |	\
-					 BNXT_NTUPLE_MATCH_DST_IP |	\
-					 BNXT_NTUPLE_MATCH_SRC_PORT |	\
-					 BNXT_NTUPLE_MATCH_DST_PORT)
 	u32			flow_id;
 };
 
diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c b/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c
index 4d4dd2b231b8..b3fd32ff963a 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c
@@ -1080,6 +1080,7 @@ static int bnxt_grxclsrule(struct bnxt *bp, struct ethtool_rxnfc *cmd)
 		(struct ethtool_rx_flow_spec *)&cmd->fs;
 	struct bnxt_filter_base *fltr_base;
 	struct bnxt_ntuple_filter *fltr;
+	struct bnxt_flow_masks *fmasks;
 	struct flow_keys *fkeys;
 	int rc = -EINVAL;
 
@@ -1127,6 +1128,7 @@ static int bnxt_grxclsrule(struct bnxt *bp, struct ethtool_rxnfc *cmd)
 	fltr = container_of(fltr_base, struct bnxt_ntuple_filter, base);
 
 	fkeys = &fltr->fkeys;
+	fmasks = &fltr->fmasks;
 	if (fkeys->basic.n_proto == htons(ETH_P_IP)) {
 		if (fkeys->basic.ip_proto == IPPROTO_TCP)
 			fs->flow_type = TCP_V4_FLOW;
@@ -1135,22 +1137,14 @@ static int bnxt_grxclsrule(struct bnxt *bp, struct ethtool_rxnfc *cmd)
 		else
 			goto fltr_err;
 
-		if (fltr->ntuple_flags & BNXT_NTUPLE_MATCH_SRC_IP) {
-			fs->h_u.tcp_ip4_spec.ip4src = fkeys->addrs.v4addrs.src;
-			fs->m_u.tcp_ip4_spec.ip4src = cpu_to_be32(~0);
-		}
-		if (fltr->ntuple_flags & BNXT_NTUPLE_MATCH_DST_IP) {
-			fs->h_u.tcp_ip4_spec.ip4dst = fkeys->addrs.v4addrs.dst;
-			fs->m_u.tcp_ip4_spec.ip4dst = cpu_to_be32(~0);
-		}
-		if (fltr->ntuple_flags & BNXT_NTUPLE_MATCH_SRC_PORT) {
-			fs->h_u.tcp_ip4_spec.psrc = fkeys->ports.src;
-			fs->m_u.tcp_ip4_spec.psrc = cpu_to_be16(~0);
-		}
-		if (fltr->ntuple_flags & BNXT_NTUPLE_MATCH_DST_PORT) {
-			fs->h_u.tcp_ip4_spec.pdst = fkeys->ports.dst;
-			fs->m_u.tcp_ip4_spec.pdst = cpu_to_be16(~0);
-		}
+		fs->h_u.tcp_ip4_spec.ip4src = fkeys->addrs.v4addrs.src;
+		fs->m_u.tcp_ip4_spec.ip4src = fmasks->addrs.v4addrs.src;
+		fs->h_u.tcp_ip4_spec.ip4dst = fkeys->addrs.v4addrs.dst;
+		fs->m_u.tcp_ip4_spec.ip4dst = fmasks->addrs.v4addrs.dst;
+		fs->h_u.tcp_ip4_spec.psrc = fkeys->ports.src;
+		fs->m_u.tcp_ip4_spec.psrc = fmasks->ports.src;
+		fs->h_u.tcp_ip4_spec.pdst = fkeys->ports.dst;
+		fs->m_u.tcp_ip4_spec.pdst = fmasks->ports.dst;
 	} else {
 		if (fkeys->basic.ip_proto == IPPROTO_TCP)
 			fs->flow_type = TCP_V6_FLOW;
@@ -1159,24 +1153,18 @@ static int bnxt_grxclsrule(struct bnxt *bp, struct ethtool_rxnfc *cmd)
 		else
 			goto fltr_err;
 
-		if (fltr->ntuple_flags & BNXT_NTUPLE_MATCH_SRC_IP) {
-			*(struct in6_addr *)&fs->h_u.tcp_ip6_spec.ip6src[0] =
-				fkeys->addrs.v6addrs.src;
-			bnxt_fill_ipv6_mask(fs->m_u.tcp_ip6_spec.ip6src);
-		}
-		if (fltr->ntuple_flags & BNXT_NTUPLE_MATCH_DST_IP) {
-			*(struct in6_addr *)&fs->h_u.tcp_ip6_spec.ip6dst[0] =
-				fkeys->addrs.v6addrs.dst;
-			bnxt_fill_ipv6_mask(fs->m_u.tcp_ip6_spec.ip6dst);
-		}
-		if (fltr->ntuple_flags & BNXT_NTUPLE_MATCH_SRC_PORT) {
-			fs->h_u.tcp_ip6_spec.psrc = fkeys->ports.src;
-			fs->m_u.tcp_ip6_spec.psrc = cpu_to_be16(~0);
-		}
-		if (fltr->ntuple_flags & BNXT_NTUPLE_MATCH_DST_PORT) {
-			fs->h_u.tcp_ip6_spec.pdst = fkeys->ports.dst;
-			fs->m_u.tcp_ip6_spec.pdst = cpu_to_be16(~0);
-		}
+		*(struct in6_addr *)&fs->h_u.tcp_ip6_spec.ip6src[0] =
+			fkeys->addrs.v6addrs.src;
+		*(struct in6_addr *)&fs->m_u.tcp_ip6_spec.ip6src[0] =
+			fmasks->addrs.v6addrs.src;
+		*(struct in6_addr *)&fs->h_u.tcp_ip6_spec.ip6dst[0] =
+			fkeys->addrs.v6addrs.dst;
+		*(struct in6_addr *)&fs->m_u.tcp_ip6_spec.ip6dst[0] =
+			fmasks->addrs.v6addrs.dst;
+		fs->h_u.tcp_ip6_spec.psrc = fkeys->ports.src;
+		fs->m_u.tcp_ip6_spec.psrc = fmasks->ports.src;
+		fs->h_u.tcp_ip6_spec.pdst = fkeys->ports.dst;
+		fs->m_u.tcp_ip6_spec.pdst = fmasks->ports.dst;
 	}
 
 	fs->ring_cookie = fltr->base.rxq;
@@ -1240,19 +1228,6 @@ static int bnxt_add_l2_cls_rule(struct bnxt *bp,
 	return rc;
 }
 
-#define IPV4_ALL_MASK		((__force __be32)~0)
-#define L4_PORT_ALL_MASK	((__force __be16)~0)
-
-static bool ipv6_mask_is_full(__be32 mask[4])
-{
-	return (mask[0] & mask[1] & mask[2] & mask[3]) == IPV4_ALL_MASK;
-}
-
-static bool ipv6_mask_is_zero(__be32 mask[4])
-{
-	return !(mask[0] | mask[1] | mask[2] | mask[3]);
-}
-
 static int bnxt_add_ntuple_cls_rule(struct bnxt *bp,
 				    struct ethtool_rx_flow_spec *fs)
 {
@@ -1260,6 +1235,7 @@ static int bnxt_add_ntuple_cls_rule(struct bnxt *bp,
 	u32 ring = ethtool_get_flow_spec_ring(fs->ring_cookie);
 	struct bnxt_ntuple_filter *new_fltr, *fltr;
 	struct bnxt_l2_filter *l2_fltr;
+	struct bnxt_flow_masks *fmasks;
 	u32 flow_type = fs->flow_type;
 	struct flow_keys *fkeys;
 	u32 idx;
@@ -1278,6 +1254,7 @@ static int bnxt_add_ntuple_cls_rule(struct bnxt *bp,
 	l2_fltr = bp->vnic_info[0].l2_filters[0];
 	atomic_inc(&l2_fltr->refcnt);
 	new_fltr->l2_fltr = l2_fltr;
+	fmasks = &new_fltr->fmasks;
 	fkeys = &new_fltr->fkeys;
 
 	rc = -EOPNOTSUPP;
@@ -1291,32 +1268,14 @@ static int bnxt_add_ntuple_cls_rule(struct bnxt *bp,
 		if (flow_type == UDP_V4_FLOW)
 			fkeys->basic.ip_proto = IPPROTO_UDP;
 		fkeys->basic.n_proto = htons(ETH_P_IP);
-
-		if (ip_mask->ip4src == IPV4_ALL_MASK) {
-			fkeys->addrs.v4addrs.src = ip_spec->ip4src;
-			new_fltr->ntuple_flags |= BNXT_NTUPLE_MATCH_SRC_IP;
-		} else if (ip_mask->ip4src) {
-			goto ntuple_err;
-		}
-		if (ip_mask->ip4dst == IPV4_ALL_MASK) {
-			fkeys->addrs.v4addrs.dst = ip_spec->ip4dst;
-			new_fltr->ntuple_flags |= BNXT_NTUPLE_MATCH_DST_IP;
-		} else if (ip_mask->ip4dst) {
-			goto ntuple_err;
-		}
-
-		if (ip_mask->psrc == L4_PORT_ALL_MASK) {
-			fkeys->ports.src = ip_spec->psrc;
-			new_fltr->ntuple_flags |= BNXT_NTUPLE_MATCH_SRC_PORT;
-		} else if (ip_mask->psrc) {
-			goto ntuple_err;
-		}
-		if (ip_mask->pdst == L4_PORT_ALL_MASK) {
-			fkeys->ports.dst = ip_spec->pdst;
-			new_fltr->ntuple_flags |= BNXT_NTUPLE_MATCH_DST_PORT;
-		} else if (ip_mask->pdst) {
-			goto ntuple_err;
-		}
+		fkeys->addrs.v4addrs.src = ip_spec->ip4src;
+		fmasks->addrs.v4addrs.src = ip_mask->ip4src;
+		fkeys->addrs.v4addrs.dst = ip_spec->ip4dst;
+		fmasks->addrs.v4addrs.dst = ip_mask->ip4dst;
+		fkeys->ports.src = ip_spec->psrc;
+		fmasks->ports.src = ip_mask->psrc;
+		fkeys->ports.dst = ip_spec->pdst;
+		fmasks->ports.dst = ip_mask->pdst;
 		break;
 	}
 	case TCP_V6_FLOW:
@@ -1329,40 +1288,21 @@ static int bnxt_add_ntuple_cls_rule(struct bnxt *bp,
 			fkeys->basic.ip_proto = IPPROTO_UDP;
 		fkeys->basic.n_proto = htons(ETH_P_IPV6);
 
-		if (ipv6_mask_is_full(ip_mask->ip6src)) {
-			fkeys->addrs.v6addrs.src =
-				*(struct in6_addr *)&ip_spec->ip6src;
-			new_fltr->ntuple_flags |= BNXT_NTUPLE_MATCH_SRC_IP;
-		} else if (!ipv6_mask_is_zero(ip_mask->ip6src)) {
-			goto ntuple_err;
-		}
-		if (ipv6_mask_is_full(ip_mask->ip6dst)) {
-			fkeys->addrs.v6addrs.dst =
-				*(struct in6_addr *)&ip_spec->ip6dst;
-			new_fltr->ntuple_flags |= BNXT_NTUPLE_MATCH_DST_IP;
-		} else if (!ipv6_mask_is_zero(ip_mask->ip6dst)) {
-			goto ntuple_err;
-		}
-
-		if (ip_mask->psrc == L4_PORT_ALL_MASK) {
-			fkeys->ports.src = ip_spec->psrc;
-			new_fltr->ntuple_flags |= BNXT_NTUPLE_MATCH_SRC_PORT;
-		} else if (ip_mask->psrc) {
-			goto ntuple_err;
-		}
-		if (ip_mask->pdst == L4_PORT_ALL_MASK) {
-			fkeys->ports.dst = ip_spec->pdst;
-			new_fltr->ntuple_flags |= BNXT_NTUPLE_MATCH_DST_PORT;
-		} else if (ip_mask->pdst) {
-			goto ntuple_err;
-		}
+		fkeys->addrs.v6addrs.src = *(struct in6_addr *)&ip_spec->ip6src;
+		fmasks->addrs.v6addrs.src = *(struct in6_addr *)&ip_mask->ip6src;
+		fkeys->addrs.v6addrs.dst = *(struct in6_addr *)&ip_spec->ip6dst;
+		fmasks->addrs.v6addrs.dst = *(struct in6_addr *)&ip_mask->ip6dst;
+		fkeys->ports.src = ip_spec->psrc;
+		fmasks->ports.src = ip_mask->psrc;
+		fkeys->ports.dst = ip_spec->pdst;
+		fmasks->ports.dst = ip_mask->pdst;
 		break;
 	}
 	default:
 		rc = -EOPNOTSUPP;
 		goto ntuple_err;
 	}
-	if (!new_fltr->ntuple_flags)
+	if (!memcmp(&BNXT_FLOW_MASK_NONE, fmasks, sizeof(*fmasks)))
 		goto ntuple_err;
 
 	idx = bnxt_get_ntp_filter_idx(bp, fkeys, NULL);
-- 
2.30.1


[-- Attachment #2: S/MIME Cryptographic Signature --]
[-- Type: application/pkcs7-signature, Size: 4209 bytes --]

Re: [PATCH net-next 04/13] bnxt_en: implement fully specified 5-tuple masks

[Michal Swiatkowski]

On Mon, Feb 05, 2024 at 02:31:53PM -0800, Michael Chan wrote:
> From: Edwin Peer <edwin.peer@broadcom.com>
> 
> Support subfield masking for IP addresses and ports. Previously, only
> entire fields could be included or excluded in NTUPLE filters.
> 
> Reviewed-by: Pavan Chebbi <pavan.chebbi@broadcom.com>
> Signed-off-by: Edwin Peer <edwin.peer@broadcom.com>
> Signed-off-by: Michael Chan <michael.chan@broadcom.com>
> ---
>  drivers/net/ethernet/broadcom/bnxt/bnxt.c     | 141 ++++++++++-------
>  drivers/net/ethernet/broadcom/bnxt/bnxt.h     |  19 +--
>  .../net/ethernet/broadcom/bnxt/bnxt_ethtool.c | 142 +++++-------------
>  3 files changed, 134 insertions(+), 168 deletions(-)
> 
[...]

Looks fine
Reviewed-by: Michal Swiatkowski <michal.swiatkowski@linux.intel.com>

Thanks,
Michal

[PATCH net-next 05/13] bnxt_en: Enhance ethtool ntuple support for ip flows besides TCP/UDP

[Michael Chan]

[-- Attachment #1: Type: text/plain, Size: 9670 bytes --]

From: Vikas Gupta <vikas.gupta@broadcom.com>

Enable flow type ipv4/ipv6
1) for protocols ICMPV4 and ICMPV6.
2) for wildcard match. Wildcard matches to TCP/UDP/ICMP.
   Note that, IPPROTO_RAW(255) i.e. a reserved protocol
   considered for a wildcard.

Reviewed-by: Andy Gospodarek <andrew.gospodarek@broadcom.com>
Reviewed-by: Pavan Chebbi <pavan.chebbi@broadcom.com>
Signed-off-by: Vikas Gupta <vikas.gupta@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
---
 drivers/net/ethernet/broadcom/bnxt/bnxt.c     |   4 +
 drivers/net/ethernet/broadcom/bnxt/bnxt.h     |   1 +
 .../net/ethernet/broadcom/bnxt/bnxt_ethtool.c | 124 +++++++++++++++---
 .../net/ethernet/broadcom/bnxt/bnxt_ethtool.h |   2 +
 4 files changed, 113 insertions(+), 18 deletions(-)

diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.c b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
index dc54cc0ab075..ea9c33e328cf 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt.c
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
@@ -8891,6 +8891,10 @@ static int bnxt_hwrm_cfa_adv_flow_mgnt_qcaps(struct bnxt *bp)
 	    CFA_ADV_FLOW_MGNT_QCAPS_RESP_FLAGS_RFS_RING_TBL_IDX_V2_SUPPORTED)
 		bp->fw_cap |= BNXT_FW_CAP_CFA_RFS_RING_TBL_IDX_V2;
 
+	if (flags &
+	    CFA_ADV_FLOW_MGNT_QCAPS_RESP_FLAGS_NTUPLE_FLOW_RX_EXT_IP_PROTO_SUPPORTED)
+		bp->fw_cap |= BNXT_FW_CAP_CFA_NTUPLE_RX_EXT_IP_PROTO;
+
 hwrm_cfa_adv_qcaps_exit:
 	hwrm_req_drop(bp, req);
 	return rc;
diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.h b/drivers/net/ethernet/broadcom/bnxt/bnxt.h
index aae180fa63b7..16f18c70c7bb 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt.h
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.h
@@ -2302,6 +2302,7 @@ struct bnxt {
 	#define BNXT_FW_CAP_PRE_RESV_VNICS		BIT_ULL(35)
 	#define BNXT_FW_CAP_BACKING_STORE_V2		BIT_ULL(36)
 	#define BNXT_FW_CAP_VNIC_TUNNEL_TPA		BIT_ULL(37)
+	#define BNXT_FW_CAP_CFA_NTUPLE_RX_EXT_IP_PROTO	BIT_ULL(38)
 
 	u32			fw_dbg_cap;
 
diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c b/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c
index b3fd32ff963a..b6bc14b5d1bb 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c
@@ -1130,28 +1130,50 @@ static int bnxt_grxclsrule(struct bnxt *bp, struct ethtool_rxnfc *cmd)
 	fkeys = &fltr->fkeys;
 	fmasks = &fltr->fmasks;
 	if (fkeys->basic.n_proto == htons(ETH_P_IP)) {
-		if (fkeys->basic.ip_proto == IPPROTO_TCP)
+		if (fkeys->basic.ip_proto == IPPROTO_ICMP ||
+		    fkeys->basic.ip_proto == IPPROTO_RAW) {
+			fs->flow_type = IP_USER_FLOW;
+			fs->h_u.usr_ip4_spec.ip_ver = ETH_RX_NFC_IP4;
+			if (fkeys->basic.ip_proto == IPPROTO_ICMP)
+				fs->h_u.usr_ip4_spec.proto = IPPROTO_ICMP;
+			else
+				fs->h_u.usr_ip4_spec.proto = IPPROTO_RAW;
+			fs->m_u.usr_ip4_spec.proto = BNXT_IP_PROTO_FULL_MASK;
+		} else if (fkeys->basic.ip_proto == IPPROTO_TCP) {
 			fs->flow_type = TCP_V4_FLOW;
-		else if (fkeys->basic.ip_proto == IPPROTO_UDP)
+		} else if (fkeys->basic.ip_proto == IPPROTO_UDP) {
 			fs->flow_type = UDP_V4_FLOW;
-		else
+		} else {
 			goto fltr_err;
+		}
 
 		fs->h_u.tcp_ip4_spec.ip4src = fkeys->addrs.v4addrs.src;
 		fs->m_u.tcp_ip4_spec.ip4src = fmasks->addrs.v4addrs.src;
 		fs->h_u.tcp_ip4_spec.ip4dst = fkeys->addrs.v4addrs.dst;
 		fs->m_u.tcp_ip4_spec.ip4dst = fmasks->addrs.v4addrs.dst;
-		fs->h_u.tcp_ip4_spec.psrc = fkeys->ports.src;
-		fs->m_u.tcp_ip4_spec.psrc = fmasks->ports.src;
-		fs->h_u.tcp_ip4_spec.pdst = fkeys->ports.dst;
-		fs->m_u.tcp_ip4_spec.pdst = fmasks->ports.dst;
+		if (fs->flow_type == TCP_V4_FLOW ||
+		    fs->flow_type == UDP_V4_FLOW) {
+			fs->h_u.tcp_ip4_spec.psrc = fkeys->ports.src;
+			fs->m_u.tcp_ip4_spec.psrc = fmasks->ports.src;
+			fs->h_u.tcp_ip4_spec.pdst = fkeys->ports.dst;
+			fs->m_u.tcp_ip4_spec.pdst = fmasks->ports.dst;
+		}
 	} else {
-		if (fkeys->basic.ip_proto == IPPROTO_TCP)
+		if (fkeys->basic.ip_proto == IPPROTO_ICMPV6 ||
+		    fkeys->basic.ip_proto == IPPROTO_RAW) {
+			fs->flow_type = IPV6_USER_FLOW;
+			if (fkeys->basic.ip_proto == IPPROTO_ICMPV6)
+				fs->h_u.usr_ip6_spec.l4_proto = IPPROTO_ICMPV6;
+			else
+				fs->h_u.usr_ip6_spec.l4_proto = IPPROTO_RAW;
+			fs->m_u.usr_ip6_spec.l4_proto = BNXT_IP_PROTO_FULL_MASK;
+		} else if (fkeys->basic.ip_proto == IPPROTO_TCP) {
 			fs->flow_type = TCP_V6_FLOW;
-		else if (fkeys->basic.ip_proto == IPPROTO_UDP)
+		} else if (fkeys->basic.ip_proto == IPPROTO_UDP) {
 			fs->flow_type = UDP_V6_FLOW;
-		else
+		} else {
 			goto fltr_err;
+		}
 
 		*(struct in6_addr *)&fs->h_u.tcp_ip6_spec.ip6src[0] =
 			fkeys->addrs.v6addrs.src;
@@ -1161,10 +1183,13 @@ static int bnxt_grxclsrule(struct bnxt *bp, struct ethtool_rxnfc *cmd)
 			fkeys->addrs.v6addrs.dst;
 		*(struct in6_addr *)&fs->m_u.tcp_ip6_spec.ip6dst[0] =
 			fmasks->addrs.v6addrs.dst;
-		fs->h_u.tcp_ip6_spec.psrc = fkeys->ports.src;
-		fs->m_u.tcp_ip6_spec.psrc = fmasks->ports.src;
-		fs->h_u.tcp_ip6_spec.pdst = fkeys->ports.dst;
-		fs->m_u.tcp_ip6_spec.pdst = fmasks->ports.dst;
+		if (fs->flow_type == TCP_V6_FLOW ||
+		    fs->flow_type == UDP_V6_FLOW) {
+			fs->h_u.tcp_ip6_spec.psrc = fkeys->ports.src;
+			fs->m_u.tcp_ip6_spec.psrc = fmasks->ports.src;
+			fs->h_u.tcp_ip6_spec.pdst = fkeys->ports.dst;
+			fs->m_u.tcp_ip6_spec.pdst = fmasks->ports.dst;
+		}
 	}
 
 	fs->ring_cookie = fltr->base.rxq;
@@ -1228,6 +1253,28 @@ static int bnxt_add_l2_cls_rule(struct bnxt *bp,
 	return rc;
 }
 
+static bool bnxt_verify_ntuple_ip4_flow(struct ethtool_usrip4_spec *ip_spec,
+					struct ethtool_usrip4_spec *ip_mask)
+{
+	if (ip_mask->l4_4_bytes || ip_mask->tos ||
+	    ip_spec->ip_ver != ETH_RX_NFC_IP4 ||
+	    ip_mask->proto != BNXT_IP_PROTO_FULL_MASK ||
+	    (ip_spec->proto != IPPROTO_RAW && ip_spec->proto != IPPROTO_ICMP))
+		return false;
+	return true;
+}
+
+static bool bnxt_verify_ntuple_ip6_flow(struct ethtool_usrip6_spec *ip_spec,
+					struct ethtool_usrip6_spec *ip_mask)
+{
+	if (ip_mask->l4_4_bytes || ip_mask->tclass ||
+	    ip_mask->l4_proto != BNXT_IP_PROTO_FULL_MASK ||
+	    (ip_spec->l4_proto != IPPROTO_RAW &&
+	     ip_spec->l4_proto != IPPROTO_ICMPV6))
+		return false;
+	return true;
+}
+
 static int bnxt_add_ntuple_cls_rule(struct bnxt *bp,
 				    struct ethtool_rx_flow_spec *fs)
 {
@@ -1247,6 +1294,18 @@ static int bnxt_add_ntuple_cls_rule(struct bnxt *bp,
 	if ((flow_type & (FLOW_MAC_EXT | FLOW_EXT)) || vf)
 		return -EOPNOTSUPP;
 
+	if (flow_type == IP_USER_FLOW) {
+		if (!bnxt_verify_ntuple_ip4_flow(&fs->h_u.usr_ip4_spec,
+						 &fs->m_u.usr_ip4_spec))
+			return -EOPNOTSUPP;
+	}
+
+	if (flow_type == IPV6_USER_FLOW) {
+		if (!bnxt_verify_ntuple_ip6_flow(&fs->h_u.usr_ip6_spec,
+						 &fs->m_u.usr_ip6_spec))
+			return -EOPNOTSUPP;
+	}
+
 	new_fltr = kzalloc(sizeof(*new_fltr), GFP_KERNEL);
 	if (!new_fltr)
 		return -ENOMEM;
@@ -1259,6 +1318,18 @@ static int bnxt_add_ntuple_cls_rule(struct bnxt *bp,
 
 	rc = -EOPNOTSUPP;
 	switch (flow_type) {
+	case IP_USER_FLOW: {
+		struct ethtool_usrip4_spec *ip_spec = &fs->h_u.usr_ip4_spec;
+		struct ethtool_usrip4_spec *ip_mask = &fs->m_u.usr_ip4_spec;
+
+		fkeys->basic.ip_proto = ip_spec->proto;
+		fkeys->basic.n_proto = htons(ETH_P_IP);
+		fkeys->addrs.v4addrs.src = ip_spec->ip4src;
+		fmasks->addrs.v4addrs.src = ip_mask->ip4src;
+		fkeys->addrs.v4addrs.dst = ip_spec->ip4dst;
+		fmasks->addrs.v4addrs.dst = ip_mask->ip4dst;
+		break;
+	}
 	case TCP_V4_FLOW:
 	case UDP_V4_FLOW: {
 		struct ethtool_tcpip4_spec *ip_spec = &fs->h_u.tcp_ip4_spec;
@@ -1278,6 +1349,18 @@ static int bnxt_add_ntuple_cls_rule(struct bnxt *bp,
 		fmasks->ports.dst = ip_mask->pdst;
 		break;
 	}
+	case IPV6_USER_FLOW: {
+		struct ethtool_usrip6_spec *ip_spec = &fs->h_u.usr_ip6_spec;
+		struct ethtool_usrip6_spec *ip_mask = &fs->m_u.usr_ip6_spec;
+
+		fkeys->basic.ip_proto = ip_spec->l4_proto;
+		fkeys->basic.n_proto = htons(ETH_P_IPV6);
+		fkeys->addrs.v6addrs.src = *(struct in6_addr *)&ip_spec->ip6src;
+		fmasks->addrs.v6addrs.src = *(struct in6_addr *)&ip_mask->ip6src;
+		fkeys->addrs.v6addrs.dst = *(struct in6_addr *)&ip_spec->ip6dst;
+		fmasks->addrs.v6addrs.dst = *(struct in6_addr *)&ip_mask->ip6dst;
+		break;
+	}
 	case TCP_V6_FLOW:
 	case UDP_V6_FLOW: {
 		struct ethtool_tcpip6_spec *ip_spec = &fs->h_u.tcp_ip6_spec;
@@ -1349,6 +1432,15 @@ static int bnxt_srxclsrlins(struct bnxt *bp, struct ethtool_rxnfc *cmd)
 	if (fs->location != RX_CLS_LOC_ANY)
 		return -EINVAL;
 
+	flow_type = fs->flow_type;
+	if ((flow_type == IP_USER_FLOW ||
+	     flow_type == IPV6_USER_FLOW) &&
+	    !(bp->fw_cap & BNXT_FW_CAP_CFA_NTUPLE_RX_EXT_IP_PROTO))
+		return -EOPNOTSUPP;
+	if (flow_type & (FLOW_MAC_EXT | FLOW_RSS))
+		return -EINVAL;
+	flow_type &= ~FLOW_EXT;
+
 	ring = ethtool_get_flow_spec_ring(fs->ring_cookie);
 	vf = ethtool_get_flow_spec_ring_vf(fs->ring_cookie);
 	if (BNXT_VF(bp) && vf)
@@ -1358,10 +1450,6 @@ static int bnxt_srxclsrlins(struct bnxt *bp, struct ethtool_rxnfc *cmd)
 	if (!vf && ring >= bp->rx_nr_rings)
 		return -EINVAL;
 
-	flow_type = fs->flow_type;
-	if (flow_type & (FLOW_MAC_EXT | FLOW_RSS))
-		return -EINVAL;
-	flow_type &= ~FLOW_EXT;
 	if (flow_type == ETHER_FLOW)
 		rc = bnxt_add_l2_cls_rule(bp, fs);
 	else
diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.h b/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.h
index a8ecef8ab82c..f9e36ac80c40 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.h
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.h
@@ -43,6 +43,8 @@ struct bnxt_led_cfg {
 
 #define BNXT_PXP_REG_LEN	0x3110
 
+#define BNXT_IP_PROTO_FULL_MASK	0xFF
+
 extern const struct ethtool_ops bnxt_ethtool_ops;
 
 u32 bnxt_get_rxfh_indir_size(struct net_device *dev);
-- 
2.30.1


[-- Attachment #2: S/MIME Cryptographic Signature --]
[-- Type: application/pkcs7-signature, Size: 4209 bytes --]

[PATCH net-next 06/13] bnxt_en: Add drop action support for ntuple

[Michael Chan]

[-- Attachment #1: Type: text/plain, Size: 2709 bytes --]

From: Vikas Gupta <vikas.gupta@broadcom.com>

Add drop action for protocols TCP/UDP/ICMP
1) Drop action for TCP/UDP is supported via flow type
   tcp4/udp4/tcp6/udp6.
2) Drop action for ICMPV4/ICMPV6/wildcard is supported
   via flow type ipv4/ipv6.

Reviewed-by: Andy Gospodarek <andrew.gospodarek@broadcom.com>
Signed-off-by: Vikas Gupta <vikas.gupta@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
---
 drivers/net/ethernet/broadcom/bnxt/bnxt.c         |  5 +++--
 drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c | 13 +++++++++++--
 2 files changed, 14 insertions(+), 4 deletions(-)

diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.c b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
index ea9c33e328cf..382b4a559c9c 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt.c
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
@@ -5747,8 +5747,9 @@ int bnxt_hwrm_cfa_ntuple_filter_alloc(struct bnxt *bp,
 	l2_fltr = fltr->l2_fltr;
 	req->l2_filter_id = l2_fltr->base.filter_id;
 
-
-	if (bp->fw_cap & BNXT_FW_CAP_CFA_RFS_RING_TBL_IDX_V2) {
+	if (fltr->base.flags & BNXT_ACT_DROP) {
+		flags = CFA_NTUPLE_FILTER_ALLOC_REQ_FLAGS_DROP;
+	} else if (bp->fw_cap & BNXT_FW_CAP_CFA_RFS_RING_TBL_IDX_V2) {
 		flags = CFA_NTUPLE_FILTER_ALLOC_REQ_FLAGS_DEST_RFS_RING_IDX;
 		req->dst_id = cpu_to_le16(fltr->base.rxq);
 	} else {
diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c b/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c
index b6bc14b5d1bb..6721c4a88453 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c
@@ -1192,7 +1192,10 @@ static int bnxt_grxclsrule(struct bnxt *bp, struct ethtool_rxnfc *cmd)
 		}
 	}
 
-	fs->ring_cookie = fltr->base.rxq;
+	if (fltr->base.flags & BNXT_ACT_DROP)
+		fs->ring_cookie = RX_CLS_FLOW_DISC;
+	else
+		fs->ring_cookie = fltr->base.rxq;
 	rc = 0;
 
 fltr_err:
@@ -1398,8 +1401,11 @@ static int bnxt_add_ntuple_cls_rule(struct bnxt *bp,
 	}
 	rcu_read_unlock();
 
-	new_fltr->base.rxq = ring;
 	new_fltr->base.flags = BNXT_ACT_NO_AGING;
+	if (fs->ring_cookie == RX_CLS_FLOW_DISC)
+		new_fltr->base.flags |= BNXT_ACT_DROP;
+	else
+		new_fltr->base.rxq = ring;
 	__set_bit(BNXT_FLTR_VALID, &new_fltr->base.state);
 	rc = bnxt_insert_ntp_filter(bp, new_fltr, idx);
 	if (!rc) {
@@ -1441,6 +1447,9 @@ static int bnxt_srxclsrlins(struct bnxt *bp, struct ethtool_rxnfc *cmd)
 		return -EINVAL;
 	flow_type &= ~FLOW_EXT;
 
+	if (fs->ring_cookie == RX_CLS_FLOW_DISC && flow_type != ETHER_FLOW)
+		return bnxt_add_ntuple_cls_rule(bp, fs);
+
 	ring = ethtool_get_flow_spec_ring(fs->ring_cookie);
 	vf = ethtool_get_flow_spec_ring_vf(fs->ring_cookie);
 	if (BNXT_VF(bp) && vf)
-- 
2.30.1


[-- Attachment #2: S/MIME Cryptographic Signature --]
[-- Type: application/pkcs7-signature, Size: 4209 bytes --]

[PATCH net-next 07/13] bnxt_en: Add separate function to delete the filter structure

[Michael Chan]

[-- Attachment #1: Type: text/plain, Size: 2988 bytes --]

From: Pavan Chebbi <pavan.chebbi@broadcom.com>

Since we are going to do filter deletion at multiple places in the
upcoming patches, add a function that does the deletion.  Future patches
add more code into this function.

Since we are passing the address of the filter base to free the
entire filter structure, add a comment to make sure that the base
is always at the beginning of the structure.

Reviewed-by: Andy Gospodarek <andrew.gospodarek@broadcom.com>
Signed-off-by: Pavan Chebbi <pavan.chebbi@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
---
 drivers/net/ethernet/broadcom/bnxt/bnxt.c | 22 ++++++++++++----------
 drivers/net/ethernet/broadcom/bnxt/bnxt.h |  2 ++
 2 files changed, 14 insertions(+), 10 deletions(-)

diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.c b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
index 382b4a559c9c..d39de72bffea 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt.c
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
@@ -4841,6 +4841,16 @@ static void bnxt_clear_ring_indices(struct bnxt *bp)
 	}
 }
 
+static void bnxt_del_fltr(struct bnxt *bp, struct bnxt_filter_base *fltr)
+{
+	hlist_del(&fltr->hash);
+	if (fltr->flags) {
+		clear_bit(fltr->sw_id, bp->ntp_fltr_bmap);
+		bp->ntp_fltr_count--;
+	}
+	kfree(fltr);
+}
+
 static void bnxt_free_ntp_fltrs(struct bnxt *bp, bool all)
 {
 	int i;
@@ -4858,10 +4868,7 @@ static void bnxt_free_ntp_fltrs(struct bnxt *bp, bool all)
 			bnxt_del_l2_filter(bp, fltr->l2_fltr);
 			if (!all && (fltr->base.flags & BNXT_ACT_FUNC_DST))
 				continue;
-			hlist_del(&fltr->base.hash);
-			clear_bit(fltr->base.sw_id, bp->ntp_fltr_bmap);
-			bp->ntp_fltr_count--;
-			kfree(fltr);
+			bnxt_del_fltr(bp, &fltr->base);
 		}
 	}
 	if (!all)
@@ -4904,12 +4911,7 @@ static void bnxt_free_l2_filters(struct bnxt *bp, bool all)
 		hlist_for_each_entry_safe(fltr, tmp, head, base.hash) {
 			if (!all && (fltr->base.flags & BNXT_ACT_FUNC_DST))
 				continue;
-			hlist_del(&fltr->base.hash);
-			if (fltr->base.flags) {
-				clear_bit(fltr->base.sw_id, bp->ntp_fltr_bmap);
-				bp->ntp_fltr_count--;
-			}
-			kfree(fltr);
+			bnxt_del_fltr(bp, &fltr->base);
 		}
 	}
 }
diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.h b/drivers/net/ethernet/broadcom/bnxt/bnxt.h
index 16f18c70c7bb..d070e3ac9739 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt.h
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.h
@@ -1365,6 +1365,7 @@ extern const struct bnxt_flow_masks BNXT_FLOW_IPV6_MASK_ALL;
 extern const struct bnxt_flow_masks BNXT_FLOW_IPV4_MASK_ALL;
 
 struct bnxt_ntuple_filter {
+	/* base filter must be the first member */
 	struct bnxt_filter_base	base;
 	struct flow_keys	fkeys;
 	struct bnxt_flow_masks	fmasks;
@@ -1395,6 +1396,7 @@ struct bnxt_ipv6_tuple {
 #define BNXT_L2_KEY_SIZE	(sizeof(struct bnxt_l2_key) / 4)
 
 struct bnxt_l2_filter {
+	/* base filter must be the first member */
 	struct bnxt_filter_base	base;
 	struct bnxt_l2_key	l2_key;
 	atomic_t		refcnt;
-- 
2.30.1


[-- Attachment #2: S/MIME Cryptographic Signature --]
[-- Type: application/pkcs7-signature, Size: 4209 bytes --]

[PATCH net-next 08/13] bnxt_en: Save user configured filters in a lookup list

[Michael Chan]

[-- Attachment #1: Type: text/plain, Size: 4352 bytes --]

From: Pavan Chebbi <pavan.chebbi@broadcom.com>

Driver needs to maintain a lookup list of all the user configured
filters. This is required in order to reconfigure these filters upon
interface toggle. We can look up this list to follow the order with
which they should be re-applied.

Reviewed-by: Andy Gospodarek <andrew.gospodarek@broadcom.com>
Signed-off-by: Pavan Chebbi <pavan.chebbi@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
---
 drivers/net/ethernet/broadcom/bnxt/bnxt.c | 23 +++++++++++++++++++++++
 drivers/net/ethernet/broadcom/bnxt/bnxt.h |  5 +++++
 2 files changed, 28 insertions(+)

diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.c b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
index d39de72bffea..b5bd88f33028 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt.c
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
@@ -4841,9 +4841,26 @@ static void bnxt_clear_ring_indices(struct bnxt *bp)
 	}
 }
 
+void bnxt_insert_usr_fltr(struct bnxt *bp, struct bnxt_filter_base *fltr)
+{
+	u8 type = fltr->type, flags = fltr->flags;
+
+	INIT_LIST_HEAD(&fltr->list);
+	if ((type == BNXT_FLTR_TYPE_L2 && flags & BNXT_ACT_RING_DST) ||
+	    (type == BNXT_FLTR_TYPE_NTUPLE && flags & BNXT_ACT_NO_AGING))
+		list_add_tail(&fltr->list, &bp->usr_fltr_list);
+}
+
+void bnxt_del_one_usr_fltr(struct bnxt *bp, struct bnxt_filter_base *fltr)
+{
+	if (!list_empty(&fltr->list))
+		list_del_init(&fltr->list);
+}
+
 static void bnxt_del_fltr(struct bnxt *bp, struct bnxt_filter_base *fltr)
 {
 	hlist_del(&fltr->hash);
+	bnxt_del_one_usr_fltr(bp, fltr);
 	if (fltr->flags) {
 		clear_bit(fltr->sw_id, bp->ntp_fltr_bmap);
 		bp->ntp_fltr_count--;
@@ -5387,6 +5404,7 @@ void bnxt_del_l2_filter(struct bnxt *bp, struct bnxt_l2_filter *fltr)
 		return;
 	}
 	hlist_del_rcu(&fltr->base.hash);
+	bnxt_del_one_usr_fltr(bp, &fltr->base);
 	if (fltr->base.flags) {
 		clear_bit(fltr->base.sw_id, bp->ntp_fltr_bmap);
 		bp->ntp_fltr_count--;
@@ -5533,6 +5551,7 @@ static int bnxt_init_l2_filter(struct bnxt *bp, struct bnxt_l2_filter *fltr,
 	}
 	head = &bp->l2_fltr_hash_tbl[idx];
 	hlist_add_head_rcu(&fltr->base.hash, head);
+	bnxt_insert_usr_fltr(bp, &fltr->base);
 	set_bit(BNXT_FLTR_INSERTED, &fltr->base.state);
 	atomic_set(&fltr->refcnt, 1);
 	return 0;
@@ -13984,6 +14003,7 @@ int bnxt_insert_ntp_filter(struct bnxt *bp, struct bnxt_ntuple_filter *fltr,
 	head = &bp->ntp_fltr_hash_tbl[idx];
 	hlist_add_head_rcu(&fltr->base.hash, head);
 	set_bit(BNXT_FLTR_INSERTED, &fltr->base.state);
+	bnxt_insert_usr_fltr(bp, &fltr->base);
 	bp->ntp_fltr_count++;
 	spin_unlock_bh(&bp->ntp_fltr_lock);
 	return 0;
@@ -14138,6 +14158,7 @@ void bnxt_del_ntp_filter(struct bnxt *bp, struct bnxt_ntuple_filter *fltr)
 		return;
 	}
 	hlist_del_rcu(&fltr->base.hash);
+	bnxt_del_one_usr_fltr(bp, &fltr->base);
 	bp->ntp_fltr_count--;
 	spin_unlock_bh(&bp->ntp_fltr_lock);
 	bnxt_del_l2_filter(bp, fltr->l2_fltr);
@@ -14954,6 +14975,8 @@ static int bnxt_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
 	if (rc)
 		goto init_err_dl;
 
+	INIT_LIST_HEAD(&bp->usr_fltr_list);
+
 	rc = register_netdev(dev);
 	if (rc)
 		goto init_err_cleanup;
diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.h b/drivers/net/ethernet/broadcom/bnxt/bnxt.h
index d070e3ac9739..a6b6db1546cc 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt.h
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.h
@@ -1334,6 +1334,7 @@ struct bnxt_pf_info {
 
 struct bnxt_filter_base {
 	struct hlist_node	hash;
+	struct list_head	list;
 	__le64			filter_id;
 	u8			type;
 #define BNXT_FLTR_TYPE_NTUPLE	1
@@ -2442,6 +2443,8 @@ struct bnxt {
 	u32			hash_seed;
 	u64			toeplitz_prefix;
 
+	struct list_head	usr_fltr_list;
+
 	/* To protect link related settings during link changes and
 	 * ethtool settings changes.
 	 */
@@ -2646,6 +2649,8 @@ u32 bnxt_fw_health_readl(struct bnxt *bp, int reg_idx);
 void bnxt_set_tpa_flags(struct bnxt *bp);
 void bnxt_set_ring_params(struct bnxt *);
 int bnxt_set_rx_skb_mode(struct bnxt *bp, bool page_mode);
+void bnxt_insert_usr_fltr(struct bnxt *bp, struct bnxt_filter_base *fltr);
+void bnxt_del_one_usr_fltr(struct bnxt *bp, struct bnxt_filter_base *fltr);
 int bnxt_hwrm_func_drv_rgtr(struct bnxt *bp, unsigned long *bmap,
 			    int bmap_size, bool async_only);
 int bnxt_hwrm_func_drv_unrgtr(struct bnxt *bp);
-- 
2.30.1


[-- Attachment #2: S/MIME Cryptographic Signature --]
[-- Type: application/pkcs7-signature, Size: 4209 bytes --]

[PATCH net-next 09/13] bnxt_en: Retain user configured filters when closing

[Michael Chan]

[-- Attachment #1: Type: text/plain, Size: 1983 bytes --]

From: Pavan Chebbi <pavan.chebbi@broadcom.com>

Driver should not free user created filters from its memory
when closing since we are going to reconfigure them when
we open again.  If the "all" parameter is false, do not free
user configured filters in bnxt_free_ntp_fltrs() and
bnxt_free_l2_filters().

Reviewed-by: Andy Gospodarek <andrew.gospodarek@broadcom.com>
Signed-off-by: Pavan Chebbi <pavan.chebbi@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
---
 drivers/net/ethernet/broadcom/bnxt/bnxt.c | 9 ++++++---
 1 file changed, 6 insertions(+), 3 deletions(-)

diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.c b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
index b5bd88f33028..41e1c6d2c127 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt.c
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
@@ -4883,7 +4883,8 @@ static void bnxt_free_ntp_fltrs(struct bnxt *bp, bool all)
 		head = &bp->ntp_fltr_hash_tbl[i];
 		hlist_for_each_entry_safe(fltr, tmp, head, base.hash) {
 			bnxt_del_l2_filter(bp, fltr->l2_fltr);
-			if (!all && (fltr->base.flags & BNXT_ACT_FUNC_DST))
+			if (!all && ((fltr->base.flags & BNXT_ACT_FUNC_DST) ||
+				     !list_empty(&fltr->base.list)))
 				continue;
 			bnxt_del_fltr(bp, &fltr->base);
 		}
@@ -4926,7 +4927,8 @@ static void bnxt_free_l2_filters(struct bnxt *bp, bool all)
 
 		head = &bp->l2_fltr_hash_tbl[i];
 		hlist_for_each_entry_safe(fltr, tmp, head, base.hash) {
-			if (!all && (fltr->base.flags & BNXT_ACT_FUNC_DST))
+			if (!all && ((fltr->base.flags & BNXT_ACT_FUNC_DST) ||
+				     !list_empty(&fltr->base.list)))
 				continue;
 			bnxt_del_fltr(bp, &fltr->base);
 		}
@@ -5851,7 +5853,8 @@ static void bnxt_hwrm_clear_vnic_filter(struct bnxt *bp)
 			struct bnxt_l2_filter *fltr = vnic->l2_filters[j];
 
 			bnxt_hwrm_l2_filter_free(bp, fltr);
-			bnxt_del_l2_filter(bp, fltr);
+			if (list_empty(&fltr->base.list))
+				bnxt_del_l2_filter(bp, fltr);
 		}
 		vnic->uc_filter_count = 0;
 	}
-- 
2.30.1


[-- Attachment #2: S/MIME Cryptographic Signature --]
[-- Type: application/pkcs7-signature, Size: 4209 bytes --]

[PATCH net-next 10/13] bnxt_en: Restore all the user created L2 and ntuple filters

[Michael Chan]

[-- Attachment #1: Type: text/plain, Size: 2671 bytes --]

From: Pavan Chebbi <pavan.chebbi@broadcom.com>

Walk the usr_fltr_list and call firmware to add these filters when
we open the NIC.  This will restore all user created filters after
reset.

Reviewed-by: Andy Gospodarek <andrew.gospodarek@broadcom.com>
Signed-off-by: Pavan Chebbi <pavan.chebbi@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
---
 drivers/net/ethernet/broadcom/bnxt/bnxt.c | 40 +++++++++++++++++++++--
 1 file changed, 38 insertions(+), 2 deletions(-)

diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.c b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
index 41e1c6d2c127..a6e2b65c45ce 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt.c
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
@@ -5853,8 +5853,7 @@ static void bnxt_hwrm_clear_vnic_filter(struct bnxt *bp)
 			struct bnxt_l2_filter *fltr = vnic->l2_filters[j];
 
 			bnxt_hwrm_l2_filter_free(bp, fltr);
-			if (list_empty(&fltr->base.list))
-				bnxt_del_l2_filter(bp, fltr);
+			bnxt_del_l2_filter(bp, fltr);
 		}
 		vnic->uc_filter_count = 0;
 	}
@@ -11537,6 +11536,42 @@ static int bnxt_reinit_after_abort(struct bnxt *bp)
 	return rc;
 }
 
+static void bnxt_cfg_one_usr_fltr(struct bnxt *bp, struct bnxt_filter_base *fltr)
+{
+	struct bnxt_ntuple_filter *ntp_fltr;
+	struct bnxt_l2_filter *l2_fltr;
+
+	if (list_empty(&fltr->list))
+		return;
+
+	if (fltr->type == BNXT_FLTR_TYPE_NTUPLE) {
+		ntp_fltr = container_of(fltr, struct bnxt_ntuple_filter, base);
+		l2_fltr = bp->vnic_info[0].l2_filters[0];
+		atomic_inc(&l2_fltr->refcnt);
+		ntp_fltr->l2_fltr = l2_fltr;
+		if (bnxt_hwrm_cfa_ntuple_filter_alloc(bp, ntp_fltr)) {
+			bnxt_del_ntp_filter(bp, ntp_fltr);
+			netdev_err(bp->dev, "restoring previously configured ntuple filter id %d failed\n",
+				   fltr->sw_id);
+		}
+	} else if (fltr->type == BNXT_FLTR_TYPE_L2) {
+		l2_fltr = container_of(fltr, struct bnxt_l2_filter, base);
+		if (bnxt_hwrm_l2_filter_alloc(bp, l2_fltr)) {
+			bnxt_del_l2_filter(bp, l2_fltr);
+			netdev_err(bp->dev, "restoring previously configured l2 filter id %d failed\n",
+				   fltr->sw_id);
+		}
+	}
+}
+
+static void bnxt_cfg_usr_fltrs(struct bnxt *bp)
+{
+	struct bnxt_filter_base *usr_fltr, *tmp;
+
+	list_for_each_entry_safe(usr_fltr, tmp, &bp->usr_fltr_list, list)
+		bnxt_cfg_one_usr_fltr(bp, usr_fltr);
+}
+
 static int __bnxt_open_nic(struct bnxt *bp, bool irq_re_init, bool link_re_init)
 {
 	int rc = 0;
@@ -11623,6 +11658,7 @@ static int __bnxt_open_nic(struct bnxt *bp, bool irq_re_init, bool link_re_init)
 		bnxt_vf_reps_open(bp);
 	bnxt_ptp_init_rtc(bp, true);
 	bnxt_ptp_cfg_tstamp_filters(bp);
+	bnxt_cfg_usr_fltrs(bp);
 	return 0;
 
 open_err_irq:
-- 
2.30.1


[-- Attachment #2: S/MIME Cryptographic Signature --]
[-- Type: application/pkcs7-signature, Size: 4209 bytes --]

[PATCH net-next 11/13] bnxt_en: Add support for user configured RSS key

[Michael Chan]

[-- Attachment #1: Type: text/plain, Size: 2712 bytes --]

From: Pavan Chebbi <pavan.chebbi@broadcom.com>

Store the user configured or generated Toeplitz key in
bp->rss_hash_key.  The key stays constant across ifdown/ifup
unless updated by the user.

Signed-off-by: Pavan Chebbi <pavan.chebbi@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
---
 drivers/net/ethernet/broadcom/bnxt/bnxt.c      | 18 ++++++++++++++++--
 drivers/net/ethernet/broadcom/bnxt/bnxt.h      |  4 ++++
 .../net/ethernet/broadcom/bnxt/bnxt_ethtool.c  |  6 ++++--
 3 files changed, 24 insertions(+), 4 deletions(-)

diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.c b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
index a6e2b65c45ce..27941ade894a 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt.c
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
@@ -4246,9 +4246,23 @@ static void bnxt_init_vnics(struct bnxt *bp)
 				u8 *key = (void *)vnic->rss_hash_key;
 				int k;
 
+				if (!bp->rss_hash_key_valid &&
+				    !bp->rss_hash_key_updated) {
+					get_random_bytes(bp->rss_hash_key,
+							 HW_HASH_KEY_SIZE);
+					bp->rss_hash_key_updated = true;
+				}
+
+				memcpy(vnic->rss_hash_key, bp->rss_hash_key,
+				       HW_HASH_KEY_SIZE);
+
+				if (!bp->rss_hash_key_updated)
+					continue;
+
+				bp->rss_hash_key_updated = false;
+				bp->rss_hash_key_valid = true;
+
 				bp->toeplitz_prefix = 0;
-				get_random_bytes(vnic->rss_hash_key,
-					      HW_HASH_KEY_SIZE);
 				for (k = 0; k < 8; k++) {
 					bp->toeplitz_prefix <<= 8;
 					bp->toeplitz_prefix |= key[k];
diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.h b/drivers/net/ethernet/broadcom/bnxt/bnxt.h
index a6b6db1546cc..28091889615b 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt.h
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.h
@@ -2222,6 +2222,10 @@ struct bnxt {
 #define BNXT_RSS_CAP_NEW_RSS_CAP		BIT(2)
 #define BNXT_RSS_CAP_RSS_TCAM			BIT(3)
 
+	u8			rss_hash_key[HW_HASH_KEY_SIZE];
+	u8			rss_hash_key_valid:1;
+	u8			rss_hash_key_updated:1;
+
 	u16			max_mtu;
 	u8			max_tc;
 	u8			max_lltc;	/* lossless TCs */
diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c b/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c
index 6721c4a88453..9b5382412db3 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c
@@ -1755,8 +1755,10 @@ static int bnxt_set_rxfh(struct net_device *dev,
 	if (rxfh->hfunc && rxfh->hfunc != ETH_RSS_HASH_TOP)
 		return -EOPNOTSUPP;
 
-	if (rxfh->key)
-		return -EOPNOTSUPP;
+	if (rxfh->key) {
+		memcpy(bp->rss_hash_key, rxfh->key, HW_HASH_KEY_SIZE);
+		bp->rss_hash_key_updated = true;
+	}
 
 	if (rxfh->indir) {
 		u32 i, pad, tbl_size = bnxt_get_rxfh_indir_size(dev);
-- 
2.30.1


[-- Attachment #2: S/MIME Cryptographic Signature --]
[-- Type: application/pkcs7-signature, Size: 4209 bytes --]

[PATCH net-next 12/13] bnxt_en: Invalidate user filters when needed

[Michael Chan]

[-- Attachment #1: Type: text/plain, Size: 3467 bytes --]

From: Pavan Chebbi <pavan.chebbi@broadcom.com>

The cached user filters slated to be reapplied need to
be cleared if configured MAC changes, RSS key changes,
number of rings changes, or ntuple is disabled.

Reviewed-by: Andy Gospodarek <andrew.gospodarek@broadcom.com>
Signed-off-by: Pavan Chebbi <pavan.chebbi@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
---
 drivers/net/ethernet/broadcom/bnxt/bnxt.c         | 14 ++++++++++++++
 drivers/net/ethernet/broadcom/bnxt/bnxt.h         |  1 +
 drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c |  3 ++-
 3 files changed, 17 insertions(+), 1 deletion(-)

diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.c b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
index 27941ade894a..856fda0c4c1a 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt.c
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
@@ -4871,6 +4871,17 @@ void bnxt_del_one_usr_fltr(struct bnxt *bp, struct bnxt_filter_base *fltr)
 		list_del_init(&fltr->list);
 }
 
+void bnxt_clear_usr_fltrs(struct bnxt *bp, bool all)
+{
+	struct bnxt_filter_base *usr_fltr, *tmp;
+
+	list_for_each_entry_safe(usr_fltr, tmp, &bp->usr_fltr_list, list) {
+		if (!all && usr_fltr->type == BNXT_FLTR_TYPE_L2)
+			continue;
+		bnxt_del_one_usr_fltr(bp, usr_fltr);
+	}
+}
+
 static void bnxt_del_fltr(struct bnxt *bp, struct bnxt_filter_base *fltr)
 {
 	hlist_del(&fltr->hash);
@@ -12414,6 +12425,8 @@ static int bnxt_set_features(struct net_device *dev, netdev_features_t features)
 
 	if (features & NETIF_F_NTUPLE)
 		flags |= BNXT_FLAG_RFS;
+	else
+		bnxt_clear_usr_fltrs(bp, true);
 
 	changes = flags ^ bp->flags;
 	if (changes & BNXT_FLAG_TPA) {
@@ -13911,6 +13924,7 @@ static int bnxt_change_mac_addr(struct net_device *dev, void *p)
 		return rc;
 
 	eth_hw_addr_set(dev, addr->sa_data);
+	bnxt_clear_usr_fltrs(bp, true);
 	if (netif_running(dev)) {
 		bnxt_close_nic(bp, false, false);
 		rc = bnxt_open_nic(bp, false, false);
diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.h b/drivers/net/ethernet/broadcom/bnxt/bnxt.h
index 28091889615b..2b2f051ee085 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt.h
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.h
@@ -2655,6 +2655,7 @@ void bnxt_set_ring_params(struct bnxt *);
 int bnxt_set_rx_skb_mode(struct bnxt *bp, bool page_mode);
 void bnxt_insert_usr_fltr(struct bnxt *bp, struct bnxt_filter_base *fltr);
 void bnxt_del_one_usr_fltr(struct bnxt *bp, struct bnxt_filter_base *fltr);
+void bnxt_clear_usr_fltrs(struct bnxt *bp, bool all);
 int bnxt_hwrm_func_drv_rgtr(struct bnxt *bp, unsigned long *bmap,
 			    int bmap_size, bool async_only);
 int bnxt_hwrm_func_drv_unrgtr(struct bnxt *bp);
diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c b/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c
index 9b5382412db3..a0962026a85f 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c
@@ -968,6 +968,7 @@ static int bnxt_set_channels(struct net_device *dev,
 		return -EINVAL;
 	}
 
+	bnxt_clear_usr_fltrs(bp, true);
 	if (netif_running(dev)) {
 		if (BNXT_PF(bp)) {
 			/* TODO CHIMP_FW: Send message to all VF's
@@ -1769,7 +1770,7 @@ static int bnxt_set_rxfh(struct net_device *dev,
 		if (pad)
 			memset(&bp->rss_indir_tbl[i], 0, pad * sizeof(u16));
 	}
-
+	bnxt_clear_usr_fltrs(bp, false);
 	if (netif_running(bp->dev)) {
 		bnxt_close_nic(bp, false, false);
 		rc = bnxt_open_nic(bp, false, false);
-- 
2.30.1


[-- Attachment #2: S/MIME Cryptographic Signature --]
[-- Type: application/pkcs7-signature, Size: 4209 bytes --]

[PATCH net-next 13/13] bnxt_en: Add RSS support for IPSEC headers

[Michael Chan]

[-- Attachment #1: Type: text/plain, Size: 5282 bytes --]

From: Ajit Khaparde <ajit.khaparde@broadcom.com>

IPSec uses two distinct protocols, Authentication Header (AH) and
Encapsulating Security Payload (ESP).
Add support to configure RSS based on AH and ESP headers.
This functionality will be enabled based on the capabilities
indicated by the firmware in HWRM_VNIC_QCAPS.

Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
---
Cc: Ajit Khaparde <ajit.khaparde@broadcom.com>
---
 drivers/net/ethernet/broadcom/bnxt/bnxt.c     | 15 ++++++--
 drivers/net/ethernet/broadcom/bnxt/bnxt.h     |  4 +++
 .../net/ethernet/broadcom/bnxt/bnxt_ethtool.c | 34 +++++++++++++++++--
 3 files changed, 49 insertions(+), 4 deletions(-)

diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.c b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
index 856fda0c4c1a..0c471ddad747 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt.c
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
@@ -6099,10 +6099,13 @@ static void
 __bnxt_hwrm_vnic_set_rss(struct bnxt *bp, struct hwrm_vnic_rss_cfg_input *req,
 			 struct bnxt_vnic_info *vnic)
 {
-	if (bp->flags & BNXT_FLAG_CHIP_P5_PLUS)
+	if (bp->flags & BNXT_FLAG_CHIP_P5_PLUS) {
 		bnxt_fill_hw_rss_tbl_p5(bp, vnic);
-	else
+		if (bp->flags & BNXT_FLAG_CHIP_P7)
+			req->flags |= VNIC_RSS_CFG_REQ_FLAGS_IPSEC_HASH_TYPE_CFG_SUPPORT;
+	} else {
 		bnxt_fill_hw_rss_tbl(bp, vnic);
+	}
 
 	if (bp->rss_hash_delta) {
 		req->hash_type = cpu_to_le32(bp->rss_hash_delta);
@@ -6465,6 +6468,14 @@ static int bnxt_hwrm_vnic_qcaps(struct bnxt *bp)
 		}
 		if (flags & VNIC_QCAPS_RESP_FLAGS_HW_TUNNEL_TPA_CAP)
 			bp->fw_cap |= BNXT_FW_CAP_VNIC_TUNNEL_TPA;
+		if (flags & VNIC_QCAPS_RESP_FLAGS_RSS_IPSEC_AH_SPI_IPV4_CAP)
+			bp->rss_cap |= BNXT_RSS_CAP_AH_V4_RSS_CAP;
+		if (flags & VNIC_QCAPS_RESP_FLAGS_RSS_IPSEC_AH_SPI_IPV6_CAP)
+			bp->rss_cap |= BNXT_RSS_CAP_AH_V6_RSS_CAP;
+		if (flags & VNIC_QCAPS_RESP_FLAGS_RSS_IPSEC_ESP_SPI_IPV4_CAP)
+			bp->rss_cap |= BNXT_RSS_CAP_ESP_V4_RSS_CAP;
+		if (flags & VNIC_QCAPS_RESP_FLAGS_RSS_IPSEC_ESP_SPI_IPV6_CAP)
+			bp->rss_cap |= BNXT_RSS_CAP_ESP_V6_RSS_CAP;
 	}
 	hwrm_req_drop(bp, req);
 	return rc;
diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.h b/drivers/net/ethernet/broadcom/bnxt/bnxt.h
index 2b2f051ee085..60bdd0673ec8 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt.h
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.h
@@ -2221,6 +2221,10 @@ struct bnxt {
 #define BNXT_RSS_CAP_UDP_RSS_CAP		BIT(1)
 #define BNXT_RSS_CAP_NEW_RSS_CAP		BIT(2)
 #define BNXT_RSS_CAP_RSS_TCAM			BIT(3)
+#define BNXT_RSS_CAP_AH_V4_RSS_CAP		BIT(4)
+#define BNXT_RSS_CAP_AH_V6_RSS_CAP		BIT(5)
+#define BNXT_RSS_CAP_ESP_V4_RSS_CAP		BIT(6)
+#define BNXT_RSS_CAP_ESP_V6_RSS_CAP		BIT(7)
 
 	u8			rss_hash_key[HW_HASH_KEY_SIZE];
 	u8			rss_hash_key_valid:1;
diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c b/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c
index a0962026a85f..f3ec5b96a5d0 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c
@@ -1533,8 +1533,14 @@ static int bnxt_grxfh(struct bnxt *bp, struct ethtool_rxnfc *cmd)
 			cmd->data |= RXH_IP_SRC | RXH_IP_DST |
 				     RXH_L4_B_0_1 | RXH_L4_B_2_3;
 		fallthrough;
-	case SCTP_V4_FLOW:
 	case AH_ESP_V4_FLOW:
+		if (bp->rss_hash_cfg &
+		    (VNIC_RSS_CFG_REQ_HASH_TYPE_AH_SPI_IPV4 |
+		     VNIC_RSS_CFG_REQ_HASH_TYPE_ESP_SPI_IPV4))
+			cmd->data |= RXH_IP_SRC | RXH_IP_DST |
+				     RXH_L4_B_0_1 | RXH_L4_B_2_3;
+		fallthrough;
+	case SCTP_V4_FLOW:
 	case AH_V4_FLOW:
 	case ESP_V4_FLOW:
 	case IPV4_FLOW:
@@ -1552,8 +1558,14 @@ static int bnxt_grxfh(struct bnxt *bp, struct ethtool_rxnfc *cmd)
 			cmd->data |= RXH_IP_SRC | RXH_IP_DST |
 				     RXH_L4_B_0_1 | RXH_L4_B_2_3;
 		fallthrough;
-	case SCTP_V6_FLOW:
 	case AH_ESP_V6_FLOW:
+		if (bp->rss_hash_cfg &
+		    (VNIC_RSS_CFG_REQ_HASH_TYPE_AH_SPI_IPV6 |
+		     VNIC_RSS_CFG_REQ_HASH_TYPE_ESP_SPI_IPV6))
+			cmd->data |= RXH_IP_SRC | RXH_IP_DST |
+				     RXH_L4_B_0_1 | RXH_L4_B_2_3;
+		fallthrough;
+	case SCTP_V6_FLOW:
 	case AH_V6_FLOW:
 	case ESP_V6_FLOW:
 	case IPV6_FLOW:
@@ -1600,6 +1612,24 @@ static int bnxt_srxfh(struct bnxt *bp, struct ethtool_rxnfc *cmd)
 		rss_hash_cfg &= ~VNIC_RSS_CFG_REQ_HASH_TYPE_UDP_IPV6;
 		if (tuple == 4)
 			rss_hash_cfg |= VNIC_RSS_CFG_REQ_HASH_TYPE_UDP_IPV6;
+	} else if (cmd->flow_type == AH_ESP_V4_FLOW) {
+		if (tuple == 4 && (!(bp->rss_cap & BNXT_RSS_CAP_AH_V4_RSS_CAP) ||
+				   !(bp->rss_cap & BNXT_RSS_CAP_ESP_V4_RSS_CAP)))
+			return -EINVAL;
+		rss_hash_cfg &= ~(VNIC_RSS_CFG_REQ_HASH_TYPE_AH_SPI_IPV4 |
+				  VNIC_RSS_CFG_REQ_HASH_TYPE_ESP_SPI_IPV4);
+		if (tuple == 4)
+			rss_hash_cfg |= VNIC_RSS_CFG_REQ_HASH_TYPE_AH_SPI_IPV4 |
+					VNIC_RSS_CFG_REQ_HASH_TYPE_ESP_SPI_IPV4;
+	} else if (cmd->flow_type == AH_ESP_V6_FLOW) {
+		if (tuple == 4 && (!(bp->rss_cap & BNXT_RSS_CAP_AH_V6_RSS_CAP) ||
+				   !(bp->rss_cap & BNXT_RSS_CAP_ESP_V6_RSS_CAP)))
+			return -EINVAL;
+		rss_hash_cfg &= ~(VNIC_RSS_CFG_REQ_HASH_TYPE_AH_SPI_IPV6 |
+				  VNIC_RSS_CFG_REQ_HASH_TYPE_ESP_SPI_IPV6);
+		if (tuple == 4)
+			rss_hash_cfg |= VNIC_RSS_CFG_REQ_HASH_TYPE_AH_SPI_IPV6 |
+					VNIC_RSS_CFG_REQ_HASH_TYPE_ESP_SPI_IPV6;
 	} else if (tuple == 4) {
 		return -EINVAL;
 	}
-- 
2.30.1


[-- Attachment #2: S/MIME Cryptographic Signature --]
[-- Type: application/pkcs7-signature, Size: 4209 bytes --]

Re: [PATCH net-next 00/13] bnxt_en: Ntuple and RSS updates

[patchwork-bot+netdevbpf]

Hello:

This series was applied to netdev/net-next.git (main)
by Jakub Kicinski <kuba@kernel.org>:

On Mon,  5 Feb 2024 14:31:49 -0800 you wrote:
> This patch series adds more ntuple and RSS features following recent
> patches to add support for user configured ntuple filters.  Additional
> features include L2 ether filters, partial tuple masks, IP filters
> besides TCP/UDP, drop action, saving and re-applying user filters
> after driver reset, user configured RSS key, and RSS for IPSEC.
> 
> Ajit Khaparde (1):
>   bnxt_en: Add RSS support for IPSEC headers
> 
> [...]

Here is the summary with links:
  - [net-next,01/13] bnxt_en: Use firmware provided maximum filter counts.
    https://git.kernel.org/netdev/net-next/c/f42822f22b1c
  - [net-next,02/13] bnxt_en: Add ethtool -N support for ether filters.
    https://git.kernel.org/netdev/net-next/c/e462998abc62
  - [net-next,03/13] bnxt_en: Support ethtool -n to display ether filters.
    https://git.kernel.org/netdev/net-next/c/7c8036fb71ce
  - [net-next,04/13] bnxt_en: implement fully specified 5-tuple masks
    https://git.kernel.org/netdev/net-next/c/c8d129c437f6
  - [net-next,05/13] bnxt_en: Enhance ethtool ntuple support for ip flows besides TCP/UDP
    https://git.kernel.org/netdev/net-next/c/9ba0e56199e3
  - [net-next,06/13] bnxt_en: Add drop action support for ntuple
    https://git.kernel.org/netdev/net-next/c/7efd79c0e689
  - [net-next,07/13] bnxt_en: Add separate function to delete the filter structure
    https://git.kernel.org/netdev/net-next/c/be40b4e9cac8
  - [net-next,08/13] bnxt_en: Save user configured filters in a lookup list
    https://git.kernel.org/netdev/net-next/c/8336a974f37d
  - [net-next,09/13] bnxt_en: Retain user configured filters when closing
    https://git.kernel.org/netdev/net-next/c/25041467d093
  - [net-next,10/13] bnxt_en: Restore all the user created L2 and ntuple filters
    https://git.kernel.org/netdev/net-next/c/44af4b622a3d
  - [net-next,11/13] bnxt_en: Add support for user configured RSS key
    https://git.kernel.org/netdev/net-next/c/5de1fce33695
  - [net-next,12/13] bnxt_en: Invalidate user filters when needed
    https://git.kernel.org/netdev/net-next/c/1018319f949c
  - [net-next,13/13] bnxt_en: Add RSS support for IPSEC headers
    https://git.kernel.org/netdev/net-next/c/0c36211bac9b

You are awesome, thank you!
-- 
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html