Many updates to cxgbe(4)

- Device configuration via plain text config file.  Also able to operate
  when not attached to the chip as the master driver.

- Generic "work request" queue that serves as the base for both ctrl and
  ofld tx queues.

- Generic interrupt handler routine that can process any event on any
  kind of ingress queue (via a dispatch table).

- A couple of new driver ioctls.  cxgbetool can now install a firmware
  to the card ("loadfw" command) and can read the card's memory
  ("memdump" and "tcb" commands).

- Lots of assorted information within dev.t4nex.X.misc.*  This is
  primarily for debugging and won't show up in sysctl -a.

- Code to manage the L2 tables on the chip.

- Updates to cxgbe(4) man page to go with the tunables that have changed.

- Updates to the shared code in common/

- Updates to the driver-firmware interface (now at fw 1.4.16.0)

MFC after:	1 month

1 files changed, 59 insertions, 12 deletions

diff --git a/sys/dev/cxgbe/offload.h b/sys/dev/cxgbe/offload.h
index f31b840..f6ada9d 100644
--- a/sys/dev/cxgbe/offload.h
+++ b/sys/dev/cxgbe/offload.h
@@ -31,15 +31,18 @@
 #ifndef __T4_OFFLOAD_H__
 #define __T4_OFFLOAD_H__
 
-/* CPL message priority levels */
-enum {
-	CPL_PRIORITY_DATA     = 0,  /* data messages */
-	CPL_PRIORITY_SETUP    = 1,  /* connection setup messages */
-	CPL_PRIORITY_TEARDOWN = 0,  /* connection teardown messages */
-	CPL_PRIORITY_LISTEN   = 1,  /* listen start/stop messages */
-	CPL_PRIORITY_ACK      = 1,  /* RX ACK messages */
-	CPL_PRIORITY_CONTROL  = 1   /* control messages */
-};
+/* XXX: flagrant misuse of mbuf fields (during tx by TOM) */
+#define MBUF_EQ(m)		(*((void **)(&(m)->m_pkthdr.rcvif)))
+/* These have to work for !M_PKTHDR so we use a field from m_hdr. */
+#define MBUF_TX_CREDITS(m)	((m)->m_hdr.pad[0])
+#define MBUF_DMA_MAPPED(m)	((m)->m_hdr.pad[1])
+
+#define INIT_ULPTX_WR(w, wrlen, atomic, tid) do { \
+	(w)->wr.wr_hi = htonl(V_FW_WR_OP(FW_ULPTX_WR) | V_FW_WR_ATOMIC(atomic)); \
+	(w)->wr.wr_mid = htonl(V_FW_WR_LEN16(DIV_ROUND_UP(wrlen, 16)) | \
+			       V_FW_WR_FLOWID(tid)); \
+	(w)->wr.wr_lo = cpu_to_be64(0); \
+} while (0)
 
 #define INIT_TP_WR(w, tid) do { \
 	(w)->wr.wr_hi = htonl(V_FW_WR_OP(FW_TP_WR) | \
@@ -49,13 +52,19 @@ enum {
 	(w)->wr.wr_lo = cpu_to_be64(0); \
 } while (0)
 
+#define INIT_TP_WR_MIT_CPL(w, cpl, tid) do { \
+	INIT_TP_WR(w, tid); \
+	OPCODE_TID(w) = htonl(MK_OPCODE_TID(cpl, tid)); \
+} while (0)
+
 /*
  * Max # of ATIDs.  The absolute HW max is 16K but we keep it lower.
  */
 #define MAX_ATIDS 8192U
 
-struct serv_entry {
+union serv_entry {
 	void *data;
+	union serv_entry *next;
 };
 
 union aopen_entry {
@@ -71,8 +80,7 @@ struct tid_info {
 	void **tid_tab;
 	unsigned int ntids;
 
-	struct serv_entry *stid_tab;
-	unsigned long *stid_bmap;
+	union serv_entry *stid_tab;
 	unsigned int nstids;
 	unsigned int stid_base;
 
@@ -84,10 +92,15 @@ struct tid_info {
 	unsigned int ftid_base;
 	unsigned int ftids_in_use;
 
+	struct mtx atid_lock;
 	union aopen_entry *afree;
 	unsigned int atids_in_use;
 
+	struct mtx stid_lock;
+	union serv_entry *sfree;
 	unsigned int stids_in_use;
+
+	unsigned int tids_in_use;
 };
 
 struct t4_range {
@@ -101,6 +114,40 @@ struct t4_virt_res {                      /* virtualized HW resources */
 	struct t4_range stag;
 	struct t4_range rq;
 	struct t4_range pbl;
+	struct t4_range qp;
+	struct t4_range cq;
+	struct t4_range ocq;
 };
 
+#ifndef TCP_OFFLOAD_DISABLE
+enum {
+	ULD_TOM = 1,
+};
+
+struct adapter;
+struct port_info;
+struct uld_info {
+	SLIST_ENTRY(uld_info) link;
+	int refcount;
+	int uld_id;
+	int (*attach)(struct adapter *, void **);
+	int (*detach)(void *);
+};
+
+struct uld_softc {
+	struct uld_info *uld;
+	void *softc;
+};
+
+struct tom_tunables {
+	int sndbuf;
+	int ddp;
+	int indsz;
+	int ddp_thres;
+};
+
+int t4_register_uld(struct uld_info *);
+int t4_unregister_uld(struct uld_info *);
+#endif
+
 #endif