diff mbox series

[net-next,v2,4/6] sfc: add functions to insert encap matches into the MAE

Message ID b9798c4b1f176257cb9b690d350f3a3c66c1b401.1679603051.git.ecree.xilinx@gmail.com (mailing list archive)
State Superseded
Delegated to: Netdev Maintainers
Headers show
Series sfc: support TC decap rules | expand

Checks

Context Check Description
netdev/series_format success Posting correctly formatted
netdev/tree_selection success Clearly marked for net-next
netdev/fixes_present success Fixes tag not required for -next series
netdev/header_inline success No static functions without inline keyword in header files
netdev/build_32bit success Errors and warnings before: 19 this patch: 19
netdev/cc_maintainers success CCed 7 of 7 maintainers
netdev/build_clang success Errors and warnings before: 18 this patch: 18
netdev/verify_signedoff success Signed-off-by tag matches author and committer
netdev/deprecated_api success None detected
netdev/check_selftest success No net selftest shell script
netdev/verify_fixes success No Fixes tag
netdev/build_allmodconfig_warn success Errors and warnings before: 19 this patch: 19
netdev/checkpatch warning WARNING: line length of 84 exceeds 80 columns WARNING: line length of 85 exceeds 80 columns WARNING: line length of 87 exceeds 80 columns WARNING: line length of 88 exceeds 80 columns WARNING: line length of 90 exceeds 80 columns WARNING: line length of 92 exceeds 80 columns WARNING: line length of 93 exceeds 80 columns WARNING: line length of 94 exceeds 80 columns WARNING: line length of 97 exceeds 80 columns
netdev/kdoc success Errors and warnings before: 0 this patch: 0
netdev/source_inline success Was 0 now: 0

Commit Message

edward.cree@amd.com March 23, 2023, 8:45 p.m. UTC
From: Edward Cree <ecree.xilinx@gmail.com>

An encap match corresponds to an entry in the exact-match Outer Rule
 table; the lookup response includes the encap type (protocol) allowing
 the hardware to continue parsing into the inner headers.

Signed-off-by: Edward Cree <ecree.xilinx@gmail.com>
---
 drivers/net/ethernet/sfc/mae.c | 105 +++++++++++++++++++++++++++++++++
 drivers/net/ethernet/sfc/mae.h |   5 ++
 drivers/net/ethernet/sfc/tc.h  |   1 +
 3 files changed, 111 insertions(+)

Comments

Simon Horman March 25, 2023, 12:04 p.m. UTC | #1
On Thu, Mar 23, 2023 at 08:45:12PM +0000, edward.cree@amd.com wrote:
> From: Edward Cree <ecree.xilinx@gmail.com>
> 
> An encap match corresponds to an entry in the exact-match Outer Rule
>  table; the lookup response includes the encap type (protocol) allowing
>  the hardware to continue parsing into the inner headers.
> 
> Signed-off-by: Edward Cree <ecree.xilinx@gmail.com>

Hi Edward,

this also looks good to me.

Minor nit below not withstanding.

Reviewed-by: Simon Horman <simon.horman@corigine.com>

...

> diff --git a/drivers/net/ethernet/sfc/mae.c b/drivers/net/ethernet/sfc/mae.c
> index 2290a63908c5..92f1383ee4b9 100644
> --- a/drivers/net/ethernet/sfc/mae.c
> +++ b/drivers/net/ethernet/sfc/mae.c
> @@ -558,6 +558,20 @@ int efx_mae_free_counter(struct efx_nic *efx, struct efx_tc_counter *cnt)
>  	return 0;
>  }
>  
> +static int efx_mae_encap_type_to_mae_type(enum efx_encap_type type)
> +{
> +	switch (type & EFX_ENCAP_TYPES_MASK) {
> +	case EFX_ENCAP_TYPE_NONE:
> +		return MAE_MCDI_ENCAP_TYPE_NONE;
> +	case EFX_ENCAP_TYPE_VXLAN:
> +		return MAE_MCDI_ENCAP_TYPE_VXLAN;
> +	case EFX_ENCAP_TYPE_GENEVE:
> +		return MAE_MCDI_ENCAP_TYPE_GENEVE;
> +	default:
> +		return -EOPNOTSUPP;
> +	}
> +}
> +
>  int efx_mae_lookup_mport(struct efx_nic *efx, u32 vf_idx, u32 *id)
>  {
>  	struct ef100_nic_data *nic_data = efx->nic_data;
> @@ -915,6 +929,97 @@ int efx_mae_free_action_set_list(struct efx_nic *efx,
>  	return 0;
>  }
>  
> +int efx_mae_register_encap_match(struct efx_nic *efx,
> +				 struct efx_tc_encap_match *encap)
> +{
> +	MCDI_DECLARE_BUF(inbuf, MC_CMD_MAE_OUTER_RULE_INSERT_IN_LEN(MAE_ENC_FIELD_PAIRS_LEN));
> +	MCDI_DECLARE_BUF(outbuf, MC_CMD_MAE_OUTER_RULE_INSERT_OUT_LEN);
> +	MCDI_DECLARE_STRUCT_PTR(match_crit);
> +	size_t outlen;
> +	int rc;
> +
> +	rc = efx_mae_encap_type_to_mae_type(encap->tun_type);
> +	if (rc < 0)
> +		return rc;

...

> diff --git a/drivers/net/ethernet/sfc/tc.h b/drivers/net/ethernet/sfc/tc.h
> index c1485679507c..19782c9a4354 100644
> --- a/drivers/net/ethernet/sfc/tc.h
> +++ b/drivers/net/ethernet/sfc/tc.h
> @@ -70,6 +70,7 @@ struct efx_tc_encap_match {
>  	__be32 src_ip, dst_ip;
>  	struct in6_addr src_ip6, dst_ip6;
>  	__be16 udp_dport;
> +	u16 tun_type; /* enum efx_encap_type */

nit: maybe the type of tyn_type can be enum efx_encap_type.

>  	u32 fw_id; /* index of this entry in firmware encap match table */
>  };
>  
>
Edward Cree March 27, 2023, 8:28 a.m. UTC | #2
On 25/03/2023 12:04, Simon Horman wrote:
> On Thu, Mar 23, 2023 at 08:45:12PM +0000, edward.cree@amd.com wrote:
>> diff --git a/drivers/net/ethernet/sfc/tc.h b/drivers/net/ethernet/sfc/tc.h
>> index c1485679507c..19782c9a4354 100644
>> --- a/drivers/net/ethernet/sfc/tc.h
>> +++ b/drivers/net/ethernet/sfc/tc.h
>> @@ -70,6 +70,7 @@ struct efx_tc_encap_match {
>>  	__be32 src_ip, dst_ip;
>>  	struct in6_addr src_ip6, dst_ip6;
>>  	__be16 udp_dport;
>> +	u16 tun_type; /* enum efx_encap_type */
> 
> nit: maybe the type of tyn_type can be enum efx_encap_type.

Yeah, it probably should.  Looking at my git history I think the
 initial reason was for struct-packing reasons (enums are int-sized
 at minimum, which is excessive for a field whose largest value is
 3), but with the rhash_head that later appears between these two
 fields, using the narrower type doesn't actually avoid a 16-bit
 hole like it appears to here.
Will change in v3.
Simon Horman March 27, 2023, 8:48 a.m. UTC | #3
On Mon, Mar 27, 2023 at 09:28:57AM +0100, Edward Cree wrote:
> On 25/03/2023 12:04, Simon Horman wrote:
> > On Thu, Mar 23, 2023 at 08:45:12PM +0000, edward.cree@amd.com wrote:
> >> diff --git a/drivers/net/ethernet/sfc/tc.h b/drivers/net/ethernet/sfc/tc.h
> >> index c1485679507c..19782c9a4354 100644
> >> --- a/drivers/net/ethernet/sfc/tc.h
> >> +++ b/drivers/net/ethernet/sfc/tc.h
> >> @@ -70,6 +70,7 @@ struct efx_tc_encap_match {
> >>  	__be32 src_ip, dst_ip;
> >>  	struct in6_addr src_ip6, dst_ip6;
> >>  	__be16 udp_dport;
> >> +	u16 tun_type; /* enum efx_encap_type */
> > 
> > nit: maybe the type of tyn_type can be enum efx_encap_type.
> 
> Yeah, it probably should.  Looking at my git history I think the
>  initial reason was for struct-packing reasons (enums are int-sized
>  at minimum, which is excessive for a field whose largest value is
>  3), but with the rhash_head that later appears between these two
>  fields, using the narrower type doesn't actually avoid a 16-bit
>  hole like it appears to here.
> Will change in v3.

Thanks.
diff mbox series

Patch

diff --git a/drivers/net/ethernet/sfc/mae.c b/drivers/net/ethernet/sfc/mae.c
index 2290a63908c5..92f1383ee4b9 100644
--- a/drivers/net/ethernet/sfc/mae.c
+++ b/drivers/net/ethernet/sfc/mae.c
@@ -558,6 +558,20 @@  int efx_mae_free_counter(struct efx_nic *efx, struct efx_tc_counter *cnt)
 	return 0;
 }
 
+static int efx_mae_encap_type_to_mae_type(enum efx_encap_type type)
+{
+	switch (type & EFX_ENCAP_TYPES_MASK) {
+	case EFX_ENCAP_TYPE_NONE:
+		return MAE_MCDI_ENCAP_TYPE_NONE;
+	case EFX_ENCAP_TYPE_VXLAN:
+		return MAE_MCDI_ENCAP_TYPE_VXLAN;
+	case EFX_ENCAP_TYPE_GENEVE:
+		return MAE_MCDI_ENCAP_TYPE_GENEVE;
+	default:
+		return -EOPNOTSUPP;
+	}
+}
+
 int efx_mae_lookup_mport(struct efx_nic *efx, u32 vf_idx, u32 *id)
 {
 	struct ef100_nic_data *nic_data = efx->nic_data;
@@ -915,6 +929,97 @@  int efx_mae_free_action_set_list(struct efx_nic *efx,
 	return 0;
 }
 
+int efx_mae_register_encap_match(struct efx_nic *efx,
+				 struct efx_tc_encap_match *encap)
+{
+	MCDI_DECLARE_BUF(inbuf, MC_CMD_MAE_OUTER_RULE_INSERT_IN_LEN(MAE_ENC_FIELD_PAIRS_LEN));
+	MCDI_DECLARE_BUF(outbuf, MC_CMD_MAE_OUTER_RULE_INSERT_OUT_LEN);
+	MCDI_DECLARE_STRUCT_PTR(match_crit);
+	size_t outlen;
+	int rc;
+
+	rc = efx_mae_encap_type_to_mae_type(encap->tun_type);
+	if (rc < 0)
+		return rc;
+	match_crit = _MCDI_DWORD(inbuf, MAE_OUTER_RULE_INSERT_IN_FIELD_MATCH_CRITERIA);
+	/* The struct contains IP src and dst, and udp dport.
+	 * So we actually need to filter on IP src and dst, L4 dport, and
+	 * ipproto == udp.
+	 */
+	MCDI_SET_DWORD(inbuf, MAE_OUTER_RULE_INSERT_IN_ENCAP_TYPE, rc);
+#ifdef CONFIG_IPV6
+	if (encap->src_ip | encap->dst_ip) {
+#endif
+		MCDI_STRUCT_SET_DWORD_BE(match_crit, MAE_ENC_FIELD_PAIRS_ENC_SRC_IP4_BE,
+					 encap->src_ip);
+		MCDI_STRUCT_SET_DWORD_BE(match_crit, MAE_ENC_FIELD_PAIRS_ENC_SRC_IP4_BE_MASK,
+					 ~(__be32)0);
+		MCDI_STRUCT_SET_DWORD_BE(match_crit, MAE_ENC_FIELD_PAIRS_ENC_DST_IP4_BE,
+					 encap->dst_ip);
+		MCDI_STRUCT_SET_DWORD_BE(match_crit, MAE_ENC_FIELD_PAIRS_ENC_DST_IP4_BE_MASK,
+					 ~(__be32)0);
+		MCDI_STRUCT_SET_WORD_BE(match_crit, MAE_ENC_FIELD_PAIRS_ENC_ETHER_TYPE_BE,
+					htons(ETH_P_IP));
+#ifdef CONFIG_IPV6
+	} else {
+		memcpy(MCDI_STRUCT_PTR(match_crit, MAE_ENC_FIELD_PAIRS_ENC_SRC_IP6_BE),
+		       &encap->src_ip6, sizeof(encap->src_ip6));
+		memset(MCDI_STRUCT_PTR(match_crit, MAE_ENC_FIELD_PAIRS_ENC_SRC_IP6_BE_MASK),
+		       0xff, sizeof(encap->src_ip6));
+		memcpy(MCDI_STRUCT_PTR(match_crit, MAE_ENC_FIELD_PAIRS_ENC_DST_IP6_BE),
+		       &encap->dst_ip6, sizeof(encap->dst_ip6));
+		memset(MCDI_STRUCT_PTR(match_crit, MAE_ENC_FIELD_PAIRS_ENC_DST_IP6_BE_MASK),
+		       0xff, sizeof(encap->dst_ip6));
+		MCDI_STRUCT_SET_WORD_BE(match_crit, MAE_ENC_FIELD_PAIRS_ENC_ETHER_TYPE_BE,
+					htons(ETH_P_IPV6));
+	}
+#endif
+	MCDI_STRUCT_SET_WORD_BE(match_crit, MAE_ENC_FIELD_PAIRS_ENC_ETHER_TYPE_BE_MASK,
+				~(__be16)0);
+	MCDI_STRUCT_SET_WORD_BE(match_crit, MAE_ENC_FIELD_PAIRS_ENC_L4_DPORT_BE,
+				encap->udp_dport);
+	MCDI_STRUCT_SET_WORD_BE(match_crit, MAE_ENC_FIELD_PAIRS_ENC_L4_DPORT_BE_MASK,
+				~(__be16)0);
+	MCDI_STRUCT_SET_BYTE(match_crit, MAE_ENC_FIELD_PAIRS_ENC_IP_PROTO, IPPROTO_UDP);
+	MCDI_STRUCT_SET_BYTE(match_crit, MAE_ENC_FIELD_PAIRS_ENC_IP_PROTO_MASK, ~0);
+	rc = efx_mcdi_rpc(efx, MC_CMD_MAE_OUTER_RULE_INSERT, inbuf,
+			  sizeof(inbuf), outbuf, sizeof(outbuf), &outlen);
+	if (rc)
+		return rc;
+	if (outlen < sizeof(outbuf))
+		return -EIO;
+	encap->fw_id = MCDI_DWORD(outbuf, MAE_OUTER_RULE_INSERT_OUT_OR_ID);
+	return 0;
+}
+
+int efx_mae_unregister_encap_match(struct efx_nic *efx,
+				   struct efx_tc_encap_match *encap)
+{
+	MCDI_DECLARE_BUF(outbuf, MC_CMD_MAE_OUTER_RULE_REMOVE_OUT_LEN(1));
+	MCDI_DECLARE_BUF(inbuf, MC_CMD_MAE_OUTER_RULE_REMOVE_IN_LEN(1));
+	size_t outlen;
+	int rc;
+
+	MCDI_SET_DWORD(inbuf, MAE_OUTER_RULE_REMOVE_IN_OR_ID, encap->fw_id);
+	rc = efx_mcdi_rpc(efx, MC_CMD_MAE_OUTER_RULE_REMOVE, inbuf,
+			  sizeof(inbuf), outbuf, sizeof(outbuf), &outlen);
+	if (rc)
+		return rc;
+	if (outlen < sizeof(outbuf))
+		return -EIO;
+	/* FW freed a different ID than we asked for, should also never happen.
+	 * Warn because it means we've now got a different idea to the FW of
+	 * what encap_mds exist, which could cause mayhem later.
+	 */
+	if (WARN_ON(MCDI_DWORD(outbuf, MAE_OUTER_RULE_REMOVE_OUT_REMOVED_OR_ID) != encap->fw_id))
+		return -EIO;
+	/* We're probably about to free @encap, but let's just make sure its
+	 * fw_id is blatted so that it won't look valid if it leaks out.
+	 */
+	encap->fw_id = MC_CMD_MAE_OUTER_RULE_INSERT_OUT_OUTER_RULE_ID_NULL;
+	return 0;
+}
+
 static int efx_mae_populate_match_criteria(MCDI_DECLARE_STRUCT_PTR(match_crit),
 					   const struct efx_tc_match *match)
 {
diff --git a/drivers/net/ethernet/sfc/mae.h b/drivers/net/ethernet/sfc/mae.h
index 2ccbc62d79b9..79d315a129ab 100644
--- a/drivers/net/ethernet/sfc/mae.h
+++ b/drivers/net/ethernet/sfc/mae.h
@@ -94,6 +94,11 @@  int efx_mae_alloc_action_set_list(struct efx_nic *efx,
 int efx_mae_free_action_set_list(struct efx_nic *efx,
 				 struct efx_tc_action_set_list *acts);
 
+int efx_mae_register_encap_match(struct efx_nic *efx,
+				 struct efx_tc_encap_match *encap);
+int efx_mae_unregister_encap_match(struct efx_nic *efx,
+				   struct efx_tc_encap_match *encap);
+
 int efx_mae_insert_rule(struct efx_nic *efx, const struct efx_tc_match *match,
 			u32 prio, u32 acts_id, u32 *id);
 int efx_mae_delete_rule(struct efx_nic *efx, u32 id);
diff --git a/drivers/net/ethernet/sfc/tc.h b/drivers/net/ethernet/sfc/tc.h
index c1485679507c..19782c9a4354 100644
--- a/drivers/net/ethernet/sfc/tc.h
+++ b/drivers/net/ethernet/sfc/tc.h
@@ -70,6 +70,7 @@  struct efx_tc_encap_match {
 	__be32 src_ip, dst_ip;
 	struct in6_addr src_ip6, dst_ip6;
 	__be16 udp_dport;
+	u16 tun_type; /* enum efx_encap_type */
 	u32 fw_id; /* index of this entry in firmware encap match table */
 };