cxgb4: Support CPL_SGE_EGR_UPDATEs encapsulated in a CPL_FW4_MSG

Newer firmware can post CPL_SGE_EGR_UPDATE message encapsulated in a
CPL_FW4_MSG as follows

flit0 rss_header (if DropRSS == 0 in IQ context)
flit1 CPL_FW4_MSG cpl
flit2 rss_header w/opcode CPL_SGE_EGR_UPDATE
flit3 CPL_SGE_EGR_UPDATE cpl

So FW4_MSG CPLs with a newly created type of FW_TYPE_RSSCPL have the
CPL_SGE_EGR_UPDATE CPL message in flit 2 of the FW4_MSG. Firmware can still
post regular CPL_SGE_EGR_UPDATE messages, so the drivers need to handle
both.

This patch also writes a new parameter to firmware requesting encapsulated
EGR_UPDATE. This allows firmware with this support to not break older drivers.

Signed-off-by: Vipul Pandya <vipul@chelsio.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

1 files changed, 35 insertions, 3 deletions

diff --git a/drivers/net/ethernet/chelsio/cxgb4/t4fw_api.h b/drivers/net/ethernet/chelsio/cxgb4/t4fw_api.h
index 93444325b1e8..d1c755f78aaf 100644
--- a/drivers/net/ethernet/chelsio/cxgb4/t4fw_api.h
+++ b/drivers/net/ethernet/chelsio/cxgb4/t4fw_api.h
@@ -973,7 +973,9 @@ enum fw_params_param_pfvf {
         FW_PARAMS_PARAM_PFVF_EQ_START   = 0x2B,
         FW_PARAMS_PARAM_PFVF_EQ_END     = 0x2C,
         FW_PARAMS_PARAM_PFVF_ACTIVE_FILTER_START = 0x2D,
-        FW_PARAMS_PARAM_PFVF_ACTIVE_FILTER_END = 0x2E
+        FW_PARAMS_PARAM_PFVF_ACTIVE_FILTER_END = 0x2E,
+        FW_PARAMS_PARAM_PFVF_ETHOFLD_END = 0x30,
+        FW_PARAMS_PARAM_PFVF_CPLFW4MSG_ENCAP = 0x31
 };
 
 /*
@@ -1758,6 +1760,25 @@ enum fw_port_module_type {
         FW_PORT_MOD_TYPE_NONE = FW_PORT_CMD_MODTYPE_MASK
 };
 
+enum fw_port_mod_sub_type {
+        FW_PORT_MOD_SUB_TYPE_NA,
+        FW_PORT_MOD_SUB_TYPE_MV88E114X = 0x1,
+        FW_PORT_MOD_SUB_TYPE_TN8022 = 0x2,
+        FW_PORT_MOD_SUB_TYPE_AQ1202 = 0x3,
+        FW_PORT_MOD_SUB_TYPE_88x3120 = 0x4,
+        FW_PORT_MOD_SUB_TYPE_BCM84834 = 0x5,
+        FW_PORT_MOD_SUB_TYPE_BT_VSC8634 = 0x8,
+
+        /* The following will never been in the VPD.  They are TWINAX cable
+         * lengths decoded from SFP+ module i2c PROMs.  These should
+         * almost certainly go somewhere else ...
+         */
+        FW_PORT_MOD_SUB_TYPE_TWINAX_1 = 0x9,
+        FW_PORT_MOD_SUB_TYPE_TWINAX_3 = 0xA,
+        FW_PORT_MOD_SUB_TYPE_TWINAX_5 = 0xB,
+        FW_PORT_MOD_SUB_TYPE_TWINAX_7 = 0xC,
+};
+
 /* port stats */
 #define FW_NUM_PORT_STATS 50
 #define FW_NUM_PORT_TX_STATS 23
@@ -2123,11 +2144,11 @@ struct fw_hdr {
         u8 intfver_ri;
         u8 intfver_iscsipdu;
         u8 intfver_iscsi;
+        u8 intfver_fcoepdu;
         u8 intfver_fcoe;
-        u8 reserved2;
+        __u32   reserved2;
         __u32   reserved3;
         __u32   reserved4;
-        __u32   reserved5;
         __be32  flags;
         __be32  reserved6[23];
 };
@@ -2137,6 +2158,17 @@ struct fw_hdr {
 #define FW_HDR_FW_VER_MICRO_GET(x) (((x) >> 8) & 0xff)
 #define FW_HDR_FW_VER_BUILD_GET(x) (((x) >> 0) & 0xff)
 
+enum fw_hdr_intfver {
+        FW_HDR_INTFVER_NIC      = 0x00,
+        FW_HDR_INTFVER_VNIC     = 0x00,
+        FW_HDR_INTFVER_OFLD     = 0x00,
+        FW_HDR_INTFVER_RI       = 0x00,
+        FW_HDR_INTFVER_ISCSIPDU = 0x00,
+        FW_HDR_INTFVER_ISCSI    = 0x00,
+        FW_HDR_INTFVER_FCOEPDU  = 0x00,
+        FW_HDR_INTFVER_FCOE     = 0x00,
+};
+
 enum fw_hdr_flags {
         FW_HDR_FLAGS_RESET_HALT = 0x00000001,
 };