Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next

Pull networking updates from David Miller: "Highlights (1721 non-merge commits, this has to be a record of some sort): 1) Add 'random' mode to team driver, from Jiri Pirko and Eric Dumazet. 2) Make it so that any driver that supports configuration of multiple MAC addresses can provide the forwarding database add and del calls by providing a default implementation and hooking that up if the driver doesn't have an explicit set of handlers. From Vlad Yasevich. 3) Support GSO segmentation over tunnels and other encapsulating devices such as VXLAN, from Pravin B Shelar. 4) Support L2 GRE tunnels in the flow dissector, from Michael Dalton. 5) Implement Tail Loss Probe (TLP) detection in TCP, from Nandita Dukkipati. 6) In the PHY layer, allow supporting wake-on-lan in situations where the PHY registers have to be written for it to be configured. Use it to support wake-on-lan in mv643xx_eth. From Michael Stapelberg. 7) Significantly improve firewire IPV6 support, from YOSHIFUJI Hideaki. 8) Allow multiple packets to be sent in a single transmission using network coding in batman-adv, from Martin Hundebøll. 9) Add support for T5 cxgb4 chips, from Santosh Rastapur. 10) Generalize the VXLAN forwarding tables so that there is more flexibility in configurating various aspects of the endpoints. From David Stevens. 11) Support RSS and TSO in hardware over GRE tunnels in bxn2x driver, from Dmitry Kravkov. 12) Zero copy support in nfnelink_queue, from Eric Dumazet and Pablo Neira Ayuso. 13) Start adding networking selftests. 14) In situations of overload on the same AF_PACKET fanout socket, or per-cpu packet receive queue, minimize drop by distributing the load to other cpus/fanouts. From Willem de Bruijn and Eric Dumazet. 15) Add support for new payload offset BPF instruction, from Daniel Borkmann. 16) Convert several drivers over to mdoule_platform_driver(), from Sachin Kamat. 17) Provide a minimal BPF JIT image disassembler userspace tool, from Daniel Borkmann. 18) Rewrite F-RTO implementation in TCP to match the final specification of it in RFC4138 and RFC5682. From Yuchung Cheng. 19) Provide netlink socket diag of netlink sockets ("Yo dawg, I hear you like netlink, so I implemented netlink dumping of netlink sockets.") From Andrey Vagin. 20) Remove ugly passing of rtnetlink attributes into rtnl_doit functions, from Thomas Graf. 21) Allow userspace to be able to see if a configuration change occurs in the middle of an address or device list dump, from Nicolas Dichtel. 22) Support RFC3168 ECN protection for ipv6 fragments, from Hannes Frederic Sowa. 23) Increase accuracy of packet length used by packet scheduler, from Jason Wang. 24) Beginning set of changes to make ipv4/ipv6 fragment handling more scalable and less susceptible to overload and locking contention, from Jesper Dangaard Brouer. 25) Get rid of using non-type-safe NLMSG_* macros and use nlmsg_*() instead. From Hong Zhiguo. 26) Optimize route usage in IPVS by avoiding reference counting where possible, from Julian Anastasov. 27) Convert IPVS schedulers to RCU, also from Julian Anastasov. 28) Support cpu fanouts in xt_NFQUEUE netfilter target, from Holger Eitzenberger. 29) Network namespace support for nf_log, ebt_log, xt_LOG, ipt_ULOG, nfnetlink_log, and nfnetlink_queue. From Gao feng. 30) Implement RFC3168 ECN protection, from Hannes Frederic Sowa. 31) Support several new r8169 chips, from Hayes Wang. 32) Support tokenized interface identifiers in ipv6, from Daniel Borkmann. 33) Use usbnet_link_change() helper in USB net driver, from Ming Lei. 34) Add 802.1ad vlan offload support, from Patrick McHardy. 35) Support mmap() based netlink communication, also from Patrick McHardy. 36) Support HW timestamping in mlx4 driver, from Amir Vadai. 37) Rationalize AF_PACKET packet timestamping when transmitting, from Willem de Bruijn and Daniel Borkmann. 38) Bring parity to what's provided by /proc/net/packet socket dumping and the info provided by netlink socket dumping of AF_PACKET sockets. From Nicolas Dichtel. 39) Fix peeking beyond zero sized SKBs in AF_UNIX, from Benjamin Poirier" * git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next: (1722 commits) filter: fix va_list build error af_unix: fix a fatal race with bit fields bnx2x: Prevent memory leak when cnic is absent bnx2x: correct reading of speed capabilities net: sctp: attribute printl with __printf for gcc fmt checks netlink: kconfig: move mmap i/o into netlink kconfig netpoll: convert mutex into a semaphore netlink: Fix skb ref counting. net_sched: act_ipt forward compat with xtables mlx4_en: fix a build error on 32bit arches Revert "bnx2x: allow nvram test to run when device is down" bridge: avoid OOPS if root port not found drivers: net: cpsw: fix kernel warn on cpsw irq enable sh_eth: use random MAC address if no valid one supplied 3c509.c: call SET_NETDEV_DEV for all device types (ISA/ISAPnP/EISA) tg3: fix to append hardware time stamping flags unix/stream: fix peeking with an offset larger than data in queue unix/dgram: fix peeking with an offset larger than data in queue unix/dgram: peek beyond 0-sized skbs openvswitch: Remove unneeded ovs_netdev_get_ifindex() ...
author: Linus Torvalds <torvalds@linux-foundation.org> 2013-05-01 23:08:52 +0200
committer: Linus Torvalds <torvalds@linux-foundation.org> 2013-05-01 23:08:52 +0200
commit: 73287a43cc79ca06629a88d1a199cd283f42456a (patch)
tree: acf4456e260115bea77ee31a29f10ce17f0db45c /drivers/scsi/csiostor
parent: Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/dto... (diff)
parent: filter: fix va_list build error (diff)
download: linux-73287a43cc79ca06629a88d1a199cd283f42456a.tar.xz
linux-73287a43cc79ca06629a88d1a199cd283f42456a.zip
12 files changed, 1205 insertions, 530 deletions
diff --git a/drivers/scsi/csiostor/Makefile b/drivers/scsi/csiostor/Makefile
index b581966c88f9..913b9a92fb06 100644
--- a/drivers/scsi/csiostor/Makefile
+++ b/drivers/scsi/csiostor/Makefile
@@ -8,4 +8,5 @@ ccflags-y += -I$(srctree)/drivers/net/ethernet/chelsio/cxgb4
 obj-$(CONFIG_SCSI_CHELSIO_FCOE) += csiostor.o
 
 csiostor-objs := csio_attr.o csio_init.o csio_lnode.o csio_scsi.o \
-		csio_hw.o csio_isr.o csio_mb.o csio_rnode.o csio_wr.o
+		csio_hw.o csio_hw_t4.o csio_hw_t5.o csio_isr.o \
+		csio_mb.o csio_rnode.o csio_wr.o
diff --git a/drivers/scsi/csiostor/csio_hw.c b/drivers/scsi/csiostor/csio_hw.c
index 7dbaf58fab9f..193605519361 100644
--- a/drivers/scsi/csiostor/csio_hw.c
+++ b/drivers/scsi/csiostor/csio_hw.c
@@ -61,7 +61,7 @@ int csio_msi = 2;
 static int dev_num;
 
 /* FCoE Adapter types & its description */
-static const struct csio_adap_desc csio_fcoe_adapters[] = {
+static const struct csio_adap_desc csio_t4_fcoe_adapters[] = {
 	{"T440-Dbg 10G", "Chelsio T440-Dbg 10G [FCoE]"},
 	{"T420-CR 10G", "Chelsio T420-CR 10G [FCoE]"},
 	{"T422-CR 10G/1G", "Chelsio T422-CR 10G/1G [FCoE]"},
@@ -77,7 +77,38 @@ static const struct csio_adap_desc csio_fcoe_adapters[] = {
 	{"B404-BT 1G", "Chelsio B404-BT 1G [FCoE]"},
 	{"T480-CR 10G", "Chelsio T480-CR 10G [FCoE]"},
 	{"T440-LP-CR 10G", "Chelsio T440-LP-CR 10G [FCoE]"},
-	{"T4 FPGA", "Chelsio T4 FPGA [FCoE]"}
+	{"AMSTERDAM 10G", "Chelsio AMSTERDAM 10G [FCoE]"},
+	{"HUAWEI T480 10G", "Chelsio HUAWEI T480 10G [FCoE]"},
+	{"HUAWEI T440 10G", "Chelsio HUAWEI T440 10G [FCoE]"},
+	{"HUAWEI STG 10G", "Chelsio HUAWEI STG 10G [FCoE]"},
+	{"ACROMAG XAUI 10G", "Chelsio ACROMAG XAUI 10G [FCoE]"},
+	{"ACROMAG SFP+ 10G", "Chelsio ACROMAG SFP+ 10G [FCoE]"},
+	{"QUANTA SFP+ 10G", "Chelsio QUANTA SFP+ 10G [FCoE]"},
+	{"HUAWEI 10Gbase-T", "Chelsio HUAWEI 10Gbase-T [FCoE]"},
+	{"HUAWEI T4TOE 10G", "Chelsio HUAWEI T4TOE 10G [FCoE]"}
+};
+
+static const struct csio_adap_desc csio_t5_fcoe_adapters[] = {
+	{"T580-Dbg 10G", "Chelsio T580-Dbg 10G [FCoE]"},
+	{"T520-CR 10G", "Chelsio T520-CR 10G [FCoE]"},
+	{"T522-CR 10G/1G", "Chelsio T452-CR 10G/1G [FCoE]"},
+	{"T540-CR 10G", "Chelsio T540-CR 10G [FCoE]"},
+	{"T520-BCH 10G", "Chelsio T520-BCH 10G [FCoE]"},
+	{"T540-BCH 10G", "Chelsio T540-BCH 10G [FCoE]"},
+	{"T540-CH 10G", "Chelsio T540-CH 10G [FCoE]"},
+	{"T520-SO 10G", "Chelsio T520-SO 10G [FCoE]"},
+	{"T520-CX4 10G", "Chelsio T520-CX4 10G [FCoE]"},
+	{"T520-BT 10G", "Chelsio T520-BT 10G [FCoE]"},
+	{"T504-BT 1G", "Chelsio T504-BT 1G [FCoE]"},
+	{"B520-SR 10G", "Chelsio B520-SR 10G [FCoE]"},
+	{"B504-BT 1G", "Chelsio B504-BT 1G [FCoE]"},
+	{"T580-CR 10G", "Chelsio T580-CR 10G [FCoE]"},
+	{"T540-LP-CR 10G", "Chelsio T540-LP-CR 10G [FCoE]"},
+	{"AMSTERDAM 10G", "Chelsio AMSTERDAM 10G [FCoE]"},
+	{"T580-LP-CR 40G", "Chelsio T580-LP-CR 40G [FCoE]"},
+	{"T520-LL-CR 10G", "Chelsio T520-LL-CR 10G [FCoE]"},
+	{"T560-CR 40G", "Chelsio T560-CR 40G [FCoE]"},
+	{"T580-CR 40G", "Chelsio T580-CR 40G [FCoE]"}
 };
 
 static void csio_mgmtm_cleanup(struct csio_mgmtm *);
@@ -124,7 +155,7 @@ int csio_is_hw_removing(struct csio_hw *hw)
  *	at the time it indicated completion is stored there.  Returns 0 if the
  *	operation completes and	-EAGAIN	otherwise.
  */
-static int
+int
 csio_hw_wait_op_done_val(struct csio_hw *hw, int reg, uint32_t mask,
 			 int polarity, int attempts, int delay, uint32_t *valp)
 {
@@ -145,6 +176,24 @@ csio_hw_wait_op_done_val(struct csio_hw *hw, int reg, uint32_t mask,
 	}
 }
 
+/*
+ *	csio_hw_tp_wr_bits_indirect - set/clear bits in an indirect TP register
+ *	@hw: the adapter
+ *	@addr: the indirect TP register address
+ *	@mask: specifies the field within the register to modify
+ *	@val: new value for the field
+ *
+ *	Sets a field of an indirect TP register to the given value.
+ */
+void
+csio_hw_tp_wr_bits_indirect(struct csio_hw *hw, unsigned int addr,
+			unsigned int mask, unsigned int val)
+{
+	csio_wr_reg32(hw, addr, TP_PIO_ADDR);
+	val |= csio_rd_reg32(hw, TP_PIO_DATA) & ~mask;
+	csio_wr_reg32(hw, val, TP_PIO_DATA);
+}
+
 void
 csio_set_reg_field(struct csio_hw *hw, uint32_t reg, uint32_t mask,
 		   uint32_t value)
@@ -157,242 +206,22 @@ csio_set_reg_field(struct csio_hw *hw, uint32_t reg, uint32_t mask,
 
 }
 
-/*
- *	csio_hw_mc_read - read from MC through backdoor accesses
- *	@hw: the hw module
- *	@addr: address of first byte requested
- *	@data: 64 bytes of data containing the requested address
- *	@ecc: where to store the corresponding 64-bit ECC word
- *
- *	Read 64 bytes of data from MC starting at a 64-byte-aligned address
- *	that covers the requested address @addr.  If @parity is not %NULL it
- *	is assigned the 64-bit ECC word for the read data.
- */
-int
-csio_hw_mc_read(struct csio_hw *hw, uint32_t addr, __be32 *data,
-		uint64_t *ecc)
-{
-	int i;
-
-	if (csio_rd_reg32(hw, MC_BIST_CMD) & START_BIST)
-		return -EBUSY;
-	csio_wr_reg32(hw, addr & ~0x3fU, MC_BIST_CMD_ADDR);
-	csio_wr_reg32(hw, 64, MC_BIST_CMD_LEN);
-	csio_wr_reg32(hw, 0xc, MC_BIST_DATA_PATTERN);
-	csio_wr_reg32(hw, BIST_OPCODE(1) | START_BIST |  BIST_CMD_GAP(1),
-		      MC_BIST_CMD);
-	i = csio_hw_wait_op_done_val(hw, MC_BIST_CMD, START_BIST,
-		 0, 10, 1, NULL);
-	if (i)
-		return i;
-
-#define MC_DATA(i) MC_BIST_STATUS_REG(MC_BIST_STATUS_RDATA, i)
-
-	for (i = 15; i >= 0; i--)
-		*data++ = htonl(csio_rd_reg32(hw, MC_DATA(i)));
-	if (ecc)
-		*ecc = csio_rd_reg64(hw, MC_DATA(16));
-#undef MC_DATA
-	return 0;
-}
-
-/*
- *	csio_hw_edc_read - read from EDC through backdoor accesses
- *	@hw: the hw module
- *	@idx: which EDC to access
- *	@addr: address of first byte requested
- *	@data: 64 bytes of data containing the requested address
- *	@ecc: where to store the corresponding 64-bit ECC word
- *
- *	Read 64 bytes of data from EDC starting at a 64-byte-aligned address
- *	that covers the requested address @addr.  If @parity is not %NULL it
- *	is assigned the 64-bit ECC word for the read data.
- */
-int
-csio_hw_edc_read(struct csio_hw *hw, int idx, uint32_t addr, __be32 *data,
-		uint64_t *ecc)
-{
-	int i;
-
-	idx *= EDC_STRIDE;
-	if (csio_rd_reg32(hw, EDC_BIST_CMD + idx) & START_BIST)
-		return -EBUSY;
-	csio_wr_reg32(hw, addr & ~0x3fU, EDC_BIST_CMD_ADDR + idx);
-	csio_wr_reg32(hw, 64, EDC_BIST_CMD_LEN + idx);
-	csio_wr_reg32(hw, 0xc, EDC_BIST_DATA_PATTERN + idx);
-	csio_wr_reg32(hw, BIST_OPCODE(1) | BIST_CMD_GAP(1) | START_BIST,
-		     EDC_BIST_CMD + idx);
-	i = csio_hw_wait_op_done_val(hw, EDC_BIST_CMD + idx, START_BIST,
-		 0, 10, 1, NULL);
-	if (i)
-		return i;
-
-#define EDC_DATA(i) (EDC_BIST_STATUS_REG(EDC_BIST_STATUS_RDATA, i) + idx)
-
-	for (i = 15; i >= 0; i--)
-		*data++ = htonl(csio_rd_reg32(hw, EDC_DATA(i)));
-	if (ecc)
-		*ecc = csio_rd_reg64(hw, EDC_DATA(16));
-#undef EDC_DATA
-	return 0;
-}
-
-/*
- *      csio_mem_win_rw - read/write memory through PCIE memory window
- *      @hw: the adapter
- *      @addr: address of first byte requested
- *      @data: MEMWIN0_APERTURE bytes of data containing the requested address
- *      @dir: direction of transfer 1 => read, 0 => write
- *
- *      Read/write MEMWIN0_APERTURE bytes of data from MC starting at a
- *      MEMWIN0_APERTURE-byte-aligned address that covers the requested
- *      address @addr.
- */
-static int
-csio_mem_win_rw(struct csio_hw *hw, u32 addr, u32 *data, int dir)
-{
-	int i;
-
-	/*
-	 * Setup offset into PCIE memory window.  Address must be a
-	 * MEMWIN0_APERTURE-byte-aligned address.  (Read back MA register to
-	 * ensure that changes propagate before we attempt to use the new
-	 * values.)
-	 */
-	csio_wr_reg32(hw, addr & ~(MEMWIN0_APERTURE - 1),
-			PCIE_MEM_ACCESS_OFFSET);
-	csio_rd_reg32(hw, PCIE_MEM_ACCESS_OFFSET);
-
-	/* Collecting data 4 bytes at a time upto MEMWIN0_APERTURE */
-	for (i = 0; i < MEMWIN0_APERTURE; i = i + sizeof(__be32)) {
-		if (dir)
-			*data++ = csio_rd_reg32(hw, (MEMWIN0_BASE + i));
-		else
-			csio_wr_reg32(hw, *data++, (MEMWIN0_BASE + i));
-	}
-
-	return 0;
-}
-
-/*
- *      csio_memory_rw - read/write EDC 0, EDC 1 or MC via PCIE memory window
- *      @hw: the csio_hw
- *      @mtype: memory type: MEM_EDC0, MEM_EDC1 or MEM_MC
- *      @addr: address within indicated memory type
- *      @len: amount of memory to transfer
- *      @buf: host memory buffer
- *      @dir: direction of transfer 1 => read, 0 => write
- *
- *      Reads/writes an [almost] arbitrary memory region in the firmware: the
- *      firmware memory address, length and host buffer must be aligned on
- *      32-bit boudaries.  The memory is transferred as a raw byte sequence
- *      from/to the firmware's memory.  If this memory contains data
- *      structures which contain multi-byte integers, it's the callers
- *      responsibility to perform appropriate byte order conversions.
- */
-static int
-csio_memory_rw(struct csio_hw *hw, int mtype, u32 addr, u32 len,
-		uint32_t *buf, int dir)
-{
-	uint32_t pos, start, end, offset, memoffset;
-	int ret;
-	uint32_t *data;
-
-	/*
-	 * Argument sanity checks ...
-	 */
-	if ((addr & 0x3) || (len & 0x3))
-		return -EINVAL;
-
-	data = kzalloc(MEMWIN0_APERTURE, GFP_KERNEL);
-	if (!data)
-		return -ENOMEM;
-
-	/* Offset into the region of memory which is being accessed
-	 * MEM_EDC0 = 0
-	 * MEM_EDC1 = 1
-	 * MEM_MC   = 2
-	 */
-	memoffset = (mtype * (5 * 1024 * 1024));
-
-	/* Determine the PCIE_MEM_ACCESS_OFFSET */
-	addr = addr + memoffset;
-
-	/*
-	 * The underlaying EDC/MC read routines read MEMWIN0_APERTURE bytes
-	 * at a time so we need to round down the start and round up the end.
-	 * We'll start copying out of the first line at (addr - start) a word
-	 * at a time.
-	 */
-	start = addr & ~(MEMWIN0_APERTURE-1);
-	end = (addr + len + MEMWIN0_APERTURE-1) & ~(MEMWIN0_APERTURE-1);
-	offset = (addr - start)/sizeof(__be32);
-
-	for (pos = start; pos < end; pos += MEMWIN0_APERTURE, offset = 0) {
-		/*
-		 * If we're writing, copy the data from the caller's memory
-		 * buffer
-		 */
-		if (!dir) {
-			/*
-			 * If we're doing a partial write, then we need to do
-			 * a read-modify-write ...
-			 */
-			if (offset || len < MEMWIN0_APERTURE) {
-				ret = csio_mem_win_rw(hw, pos, data, 1);
-				if (ret) {
-					kfree(data);
-					return ret;
-				}
-			}
-			while (offset < (MEMWIN0_APERTURE/sizeof(__be32)) &&
-								len > 0) {
-				data[offset++] = *buf++;
-				len -= sizeof(__be32);
-			}
-		}
-
-		/*
-		 * Transfer a block of memory and bail if there's an error.
-		 */
-		ret = csio_mem_win_rw(hw, pos, data, dir);
-		if (ret) {
-			kfree(data);
-			return ret;
-		}
-
-		/*
-		 * If we're reading, copy the data into the caller's memory
-		 * buffer.
-		 */
-		if (dir)
-			while (offset < (MEMWIN0_APERTURE/sizeof(__be32)) &&
-								len > 0) {
-				*buf++ = data[offset++];
-				len -= sizeof(__be32);
-			}
-	}
-
-	kfree(data);
-
-	return 0;
-}
-
 static int
 csio_memory_write(struct csio_hw *hw, int mtype, u32 addr, u32 len, u32 *buf)
 {
-	return csio_memory_rw(hw, mtype, addr, len, buf, 0);
+	return hw->chip_ops->chip_memory_rw(hw, MEMWIN_CSIOSTOR, mtype,
+					    addr, len, buf, 0);
 }
 
 /*
  * EEPROM reads take a few tens of us while writes can take a bit over 5 ms.
  */
-#define EEPROM_MAX_RD_POLL 40
-#define EEPROM_MAX_WR_POLL 6
-#define EEPROM_STAT_ADDR   0x7bfc
-#define VPD_BASE           0x400
-#define VPD_BASE_OLD	   0
-#define VPD_LEN            512
+#define EEPROM_MAX_RD_POLL	40
+#define EEPROM_MAX_WR_POLL	6
+#define EEPROM_STAT_ADDR	0x7bfc
+#define VPD_BASE		0x400
+#define VPD_BASE_OLD		0
+#define VPD_LEN			1024
 #define VPD_INFO_FLD_HDR_SIZE	3
 
 /*
@@ -817,23 +646,6 @@ out:
 	return 0;
 }
 
-/*
- *	csio_hw_flash_cfg_addr - return the address of the flash
- *				configuration file
- *	@hw: the HW module
- *
- *	Return the address within the flash where the Firmware Configuration
- *	File is stored.
- */
-static unsigned int
-csio_hw_flash_cfg_addr(struct csio_hw *hw)
-{
-	if (hw->params.sf_size == 0x100000)
-		return FPGA_FLASH_CFG_OFFSET;
-	else
-		return FLASH_CFG_OFFSET;
-}
-
 static void
 csio_hw_print_fw_version(struct csio_hw *hw, char *str)
 {
@@ -898,13 +710,13 @@ csio_hw_check_fw_version(struct csio_hw *hw)
 	minor = FW_HDR_FW_VER_MINOR_GET(hw->fwrev);
 	micro = FW_HDR_FW_VER_MICRO_GET(hw->fwrev);
 
-	if (major != FW_VERSION_MAJOR) {            /* major mismatch - fail */
+	if (major != FW_VERSION_MAJOR(hw)) {	/* major mismatch - fail */
 		csio_err(hw, "card FW has major version %u, driver wants %u\n",
-			 major, FW_VERSION_MAJOR);
+			 major, FW_VERSION_MAJOR(hw));
 		return -EINVAL;
 	}
 
-	if (minor == FW_VERSION_MINOR && micro == FW_VERSION_MICRO)
+	if (minor == FW_VERSION_MINOR(hw) && micro == FW_VERSION_MICRO(hw))
 		return 0;        /* perfect match */
 
 	/* Minor/micro version mismatch */
@@ -1044,7 +856,7 @@ static void
 csio_set_pcie_completion_timeout(struct csio_hw *hw, u8 range)
 {
 	uint16_t val;
-	uint32_t pcie_cap;
+	int pcie_cap;
 
 	if (!csio_pci_capability(hw->pdev, PCI_CAP_ID_EXP, &pcie_cap)) {
 		pci_read_config_word(hw->pdev,
@@ -1056,84 +868,6 @@ csio_set_pcie_completion_timeout(struct csio_hw *hw, u8 range)
 	}
 }
 
-
-/*
- * Return the specified PCI-E Configuration Space register from our Physical
- * Function.  We try first via a Firmware LDST Command since we prefer to let
- * the firmware own all of these registers, but if that fails we go for it
- * directly ourselves.
- */
-static uint32_t
-csio_read_pcie_cfg4(struct csio_hw *hw, int reg)
-{
-	u32 val = 0;
-	struct csio_mb *mbp;
-	int rv;
-	struct fw_ldst_cmd *ldst_cmd;
-
-	mbp = mempool_alloc(hw->mb_mempool, GFP_ATOMIC);
-	if (!mbp) {
-		CSIO_INC_STATS(hw, n_err_nomem);
-		pci_read_config_dword(hw->pdev, reg, &val);
-		return val;
-	}
-
-	csio_mb_ldst(hw, mbp, CSIO_MB_DEFAULT_TMO, reg);
-
-	rv = csio_mb_issue(hw, mbp);
-
-	/*
-	 * If the LDST Command suucceeded, exctract the returned register
-	 * value.  Otherwise read it directly ourself.
-	 */
-	if (rv == 0) {
-		ldst_cmd = (struct fw_ldst_cmd *)(mbp->mb);
-		val = ntohl(ldst_cmd->u.pcie.data[0]);
-	} else
-		pci_read_config_dword(hw->pdev, reg, &val);
-
-	mempool_free(mbp, hw->mb_mempool);
-
-	return val;
-} /* csio_read_pcie_cfg4 */
-
-static int
-csio_hw_set_mem_win(struct csio_hw *hw)
-{
-	u32 bar0;
-
-	/*
-	 * Truncation intentional: we only read the bottom 32-bits of the
-	 * 64-bit BAR0/BAR1 ...  We use the hardware backdoor mechanism to
-	 * read BAR0 instead of using pci_resource_start() because we could be
-	 * operating from within a Virtual Machine which is trapping our
-	 * accesses to our Configuration Space and we need to set up the PCI-E
-	 * Memory Window decoders with the actual addresses which will be
-	 * coming across the PCI-E link.
-	 */
-	bar0 = csio_read_pcie_cfg4(hw, PCI_BASE_ADDRESS_0);
-	bar0 &= PCI_BASE_ADDRESS_MEM_MASK;
-
-	/*
-	 * Set up memory window for accessing adapter memory ranges.  (Read
-	 * back MA register to ensure that changes propagate before we attempt
-	 * to use the new values.)
-	 */
-	csio_wr_reg32(hw, (bar0 + MEMWIN0_BASE) | BIR(0) |
-		WINDOW(ilog2(MEMWIN0_APERTURE) - 10),
-		PCIE_MEM_ACCESS_REG(PCIE_MEM_ACCESS_BASE_WIN, 0));
-	csio_wr_reg32(hw, (bar0 + MEMWIN1_BASE) | BIR(0) |
-		WINDOW(ilog2(MEMWIN1_APERTURE) - 10),
-		PCIE_MEM_ACCESS_REG(PCIE_MEM_ACCESS_BASE_WIN, 1));
-	csio_wr_reg32(hw, (bar0 + MEMWIN2_BASE) | BIR(0) |
-		WINDOW(ilog2(MEMWIN2_APERTURE) - 10),
-		PCIE_MEM_ACCESS_REG(PCIE_MEM_ACCESS_BASE_WIN, 2));
-	csio_rd_reg32(hw, PCIE_MEM_ACCESS_REG(PCIE_MEM_ACCESS_BASE_WIN, 2));
-	return 0;
-} /* csio_hw_set_mem_win */
-
-
-
 /*****************************************************************************/
 /* HW State machine assists                                                  */
 /*****************************************************************************/
@@ -1234,7 +968,9 @@ retry:
 		for (;;) {
 			uint32_t pcie_fw;
 
+			spin_unlock_irq(&hw->lock);
 			msleep(50);
+			spin_lock_irq(&hw->lock);
 			waiting -= 50;
 
 			/*
@@ -2121,9 +1857,9 @@ csio_hw_flash_config(struct csio_hw *hw, u32 *fw_cfg_param, char *path)
 	uint32_t *cfg_data;
 	int value_to_add = 0;
 
-	if (request_firmware(&cf, CSIO_CF_FNAME, dev) < 0) {
-		csio_err(hw, "could not find config file " CSIO_CF_FNAME
-			 ",err: %d\n", ret);
+	if (request_firmware(&cf, CSIO_CF_FNAME(hw), dev) < 0) {
+		csio_err(hw, "could not find config file %s, err: %d\n",
+			 CSIO_CF_FNAME(hw), ret);
 		return -ENOENT;
 	}
 
@@ -2147,9 +1883,24 @@ csio_hw_flash_config(struct csio_hw *hw, u32 *fw_cfg_param, char *path)
 
 	ret = csio_memory_write(hw, mtype, maddr,
 				cf->size + value_to_add, cfg_data);
+
+	if ((ret == 0) && (value_to_add != 0)) {
+		union {
+			u32 word;
+			char buf[4];
+		} last;
+		size_t size = cf->size & ~0x3;
+		int i;
+
+		last.word = cfg_data[size >> 2];
+		for (i = value_to_add; i < 4; i++)
+			last.buf[i] = 0;
+		ret = csio_memory_write(hw, mtype, maddr + size, 4, &last.word);
+	}
 	if (ret == 0) {
-		csio_info(hw, "config file upgraded to " CSIO_CF_FNAME "\n");
-		strncpy(path, "/lib/firmware/" CSIO_CF_FNAME, 64);
+		csio_info(hw, "config file upgraded to %s\n",
+			  CSIO_CF_FNAME(hw));
+		snprintf(path, 64, "%s%s", "/lib/firmware/", CSIO_CF_FNAME(hw));
 	}
 
 leave:
@@ -2179,7 +1930,7 @@ csio_hw_use_fwconfig(struct csio_hw *hw, int reset, u32 *fw_cfg_param)
 {
 	unsigned int mtype, maddr;
 	int rv;
-	uint32_t finiver, finicsum, cfcsum;
+	uint32_t finiver = 0, finicsum = 0, cfcsum = 0;
 	int using_flash;
 	char path[64];
 
@@ -2207,7 +1958,7 @@ csio_hw_use_fwconfig(struct csio_hw *hw, int reset, u32 *fw_cfg_param)
 			 * config file from flash.
 			 */
 			mtype = FW_MEMTYPE_CF_FLASH;
-			maddr = csio_hw_flash_cfg_addr(hw);
+			maddr = hw->chip_ops->chip_flash_cfg_addr(hw);
 			using_flash = 1;
 		} else {
 			/*
@@ -2346,30 +2097,32 @@ csio_hw_flash_fw(struct csio_hw *hw)
 	struct pci_dev *pci_dev = hw->pdev;
 	struct device *dev = &pci_dev->dev ;
 
-	if (request_firmware(&fw, CSIO_FW_FNAME, dev) < 0) {
-		csio_err(hw, "could not find firmware image " CSIO_FW_FNAME
-		",err: %d\n", ret);
+	if (request_firmware(&fw, CSIO_FW_FNAME(hw), dev) < 0) {
+		csio_err(hw, "could not find firmware image %s, err: %d\n",
+			 CSIO_FW_FNAME(hw), ret);
 		return -EINVAL;
 	}
 
 	hdr = (const struct fw_hdr *)fw->data;
 	fw_ver = ntohl(hdr->fw_ver);
-	if (FW_HDR_FW_VER_MAJOR_GET(fw_ver) != FW_VERSION_MAJOR)
+	if (FW_HDR_FW_VER_MAJOR_GET(fw_ver) != FW_VERSION_MAJOR(hw))
 		return -EINVAL;      /* wrong major version, won't do */
 
 	/*
 	 * If the flash FW is unusable or we found something newer, load it.
 	 */
-	if (FW_HDR_FW_VER_MAJOR_GET(hw->fwrev) != FW_VERSION_MAJOR ||
+	if (FW_HDR_FW_VER_MAJOR_GET(hw->fwrev) != FW_VERSION_MAJOR(hw) ||
 	    fw_ver > hw->fwrev) {
 		ret = csio_hw_fw_upgrade(hw, hw->pfn, fw->data, fw->size,
 				    /*force=*/false);
 		if (!ret)
-			csio_info(hw, "firmware upgraded to version %pI4 from "
-				  CSIO_FW_FNAME "\n", &hdr->fw_ver);
+			csio_info(hw,
+				  "firmware upgraded to version %pI4 from %s\n",
+				  &hdr->fw_ver, CSIO_FW_FNAME(hw));
 		else
 			csio_err(hw, "firmware upgrade failed! err=%d\n", ret);
-	}
+	} else
+		ret = -EINVAL;
 
 	release_firmware(fw);
 
@@ -2410,7 +2163,7 @@ csio_hw_configure(struct csio_hw *hw)
 	/* Set pci completion timeout value to 4 seconds. */
 	csio_set_pcie_completion_timeout(hw, 0xd);
 
-	csio_hw_set_mem_win(hw);
+	hw->chip_ops->chip_set_mem_win(hw, MEMWIN_CSIOSTOR);
 
 	rv = csio_hw_get_fw_version(hw, &hw->fwrev);
 	if (rv != 0)
@@ -2478,6 +2231,8 @@ csio_hw_configure(struct csio_hw *hw)
 	} else {
 		if (hw->fw_state == CSIO_DEV_STATE_INIT) {
 
+			hw->flags |= CSIO_HWF_USING_SOFT_PARAMS;
+
 			/* device parameters */
 			rv = csio_get_device_params(hw);
 			if (rv != 0)
@@ -2651,7 +2406,7 @@ csio_hw_intr_disable(struct csio_hw *hw)
 
 }
 
-static void
+void
 csio_hw_fatal_err(struct csio_hw *hw)
 {
 	csio_set_reg_field(hw, SGE_CONTROL, GLOBALENABLE, 0);
@@ -2990,14 +2745,6 @@ csio_hws_pcierr(struct csio_hw *hw, enum csio_hw_ev evt)
 /* END: HW SM                                                                */
 /*****************************************************************************/
 
-/* Slow path handlers */
-struct intr_info {
-	unsigned int mask;       /* bits to check in interrupt status */
-	const char *msg;         /* message to print or NULL */
-	short stat_idx;          /* stat counter to increment or -1 */
-	unsigned short fatal;    /* whether the condition reported is fatal */
-};
-
 /*
  *	csio_handle_intr_status - table driven interrupt handler
  *	@hw: HW instance
@@ -3011,7 +2758,7 @@ struct intr_info {
  *	by an entry specifying mask 0.  Returns the number of fatal interrupt
  *	conditions.
  */
-static int
+int
 csio_handle_intr_status(struct csio_hw *hw, unsigned int reg,
 				 const struct intr_info *acts)
 {
@@ -3038,80 +2785,6 @@ csio_handle_intr_status(struct csio_hw *hw, unsigned int reg,
 }
 
 /*
- * Interrupt handler for the PCIE module.
- */
-static void
-csio_pcie_intr_handler(struct csio_hw *hw)
-{
-	static struct intr_info sysbus_intr_info[] = {
-		{ RNPP, "RXNP array parity error", -1, 1 },
-		{ RPCP, "RXPC array parity error", -1, 1 },
-		{ RCIP, "RXCIF array parity error", -1, 1 },
-		{ RCCP, "Rx completions control array parity error", -1, 1 },
-		{ RFTP, "RXFT array parity error", -1, 1 },
-		{ 0, NULL, 0, 0 }
-	};
-	static struct intr_info pcie_port_intr_info[] = {
-		{ TPCP, "TXPC array parity error", -1, 1 },
-		{ TNPP, "TXNP array parity error", -1, 1 },
-		{ TFTP, "TXFT array parity error", -1, 1 },
-		{ TCAP, "TXCA array parity error", -1, 1 },
-		{ TCIP, "TXCIF array parity error", -1, 1 },
-		{ RCAP, "RXCA array parity error", -1, 1 },
-		{ OTDD, "outbound request TLP discarded", -1, 1 },
-		{ RDPE, "Rx data parity error", -1, 1 },
-		{ TDUE, "Tx uncorrectable data error", -1, 1 },
-		{ 0, NULL, 0, 0 }
-	};
-	static struct intr_info pcie_intr_info[] = {
-		{ MSIADDRLPERR, "MSI AddrL parity error", -1, 1 },
-		{ MSIADDRHPERR, "MSI AddrH parity error", -1, 1 },
-		{ MSIDATAPERR, "MSI data parity error", -1, 1 },
-		{ MSIXADDRLPERR, "MSI-X AddrL parity error", -1, 1 },
-		{ MSIXADDRHPERR, "MSI-X AddrH parity error", -1, 1 },
-		{ MSIXDATAPERR, "MSI-X data parity error", -1, 1 },
-		{ MSIXDIPERR, "MSI-X DI parity error", -1, 1 },
-		{ PIOCPLPERR, "PCI PIO completion FIFO parity error", -1, 1 },
-		{ PIOREQPERR, "PCI PIO request FIFO parity error", -1, 1 },
-		{ TARTAGPERR, "PCI PCI target tag FIFO parity error", -1, 1 },
-		{ CCNTPERR, "PCI CMD channel count parity error", -1, 1 },
-		{ CREQPERR, "PCI CMD channel request parity error", -1, 1 },
-		{ CRSPPERR, "PCI CMD channel response parity error", -1, 1 },
-		{ DCNTPERR, "PCI DMA channel count parity error", -1, 1 },
-		{ DREQPERR, "PCI DMA channel request parity error", -1, 1 },
-		{ DRSPPERR, "PCI DMA channel response parity error", -1, 1 },
-		{ HCNTPERR, "PCI HMA channel count parity error", -1, 1 },
-		{ HREQPERR, "PCI HMA channel request parity error", -1, 1 },
-		{ HRSPPERR, "PCI HMA channel response parity error", -1, 1 },
-		{ CFGSNPPERR, "PCI config snoop FIFO parity error", -1, 1 },
-		{ FIDPERR, "PCI FID parity error", -1, 1 },
-		{ INTXCLRPERR, "PCI INTx clear parity error", -1, 1 },
-		{ MATAGPERR, "PCI MA tag parity error", -1, 1 },
-		{ PIOTAGPERR, "PCI PIO tag parity error", -1, 1 },
-		{ RXCPLPERR, "PCI Rx completion parity error", -1, 1 },
-		{ RXWRPERR, "PCI Rx write parity error", -1, 1 },
-		{ RPLPERR, "PCI replay buffer parity error", -1, 1 },
-		{ PCIESINT, "PCI core secondary fault", -1, 1 },
-		{ PCIEPINT, "PCI core primary fault", -1, 1 },
-		{ UNXSPLCPLERR, "PCI unexpected split completion error", -1,
-		  0 },
-		{ 0, NULL, 0, 0 }
-	};
-
-	int fat;
-
-	fat = csio_handle_intr_status(hw,
-				    PCIE_CORE_UTL_SYSTEM_BUS_AGENT_STATUS,
-				    sysbus_intr_info) +
-	      csio_handle_intr_status(hw,
-				    PCIE_CORE_UTL_PCI_EXPRESS_PORT_STATUS,
-				    pcie_port_intr_info) +
-	      csio_handle_intr_status(hw, PCIE_INT_CAUSE, pcie_intr_info);
-	if (fat)
-		csio_hw_fatal_err(hw);
-}
-
-/*
  * TP interrupt handler.
  */
 static void csio_tp_intr_handler(struct csio_hw *hw)
@@ -3517,7 +3190,7 @@ static void csio_ncsi_intr_handler(struct csio_hw *hw)
  */
 static void csio_xgmac_intr_handler(struct csio_hw *hw, int port)
 {
-	uint32_t v = csio_rd_reg32(hw, PORT_REG(port, XGMAC_PORT_INT_CAUSE));
+	uint32_t v = csio_rd_reg32(hw, CSIO_MAC_INT_CAUSE_REG(hw, port));
 
 	v &= TXFIFO_PRTY_ERR | RXFIFO_PRTY_ERR;
 	if (!v)
@@ -3527,7 +3200,7 @@ static void csio_xgmac_intr_handler(struct csio_hw *hw, int port)
 		csio_fatal(hw, "XGMAC %d Tx FIFO parity error\n", port);
 	if (v & RXFIFO_PRTY_ERR)
 		csio_fatal(hw, "XGMAC %d Rx FIFO parity error\n", port);
-	csio_wr_reg32(hw, v, PORT_REG(port, XGMAC_PORT_INT_CAUSE));
+	csio_wr_reg32(hw, v, CSIO_MAC_INT_CAUSE_REG(hw, port));
 	csio_hw_fatal_err(hw);
 }
 
@@ -3596,7 +3269,7 @@ csio_hw_slow_intr_handler(struct csio_hw *hw)
 		csio_xgmac_intr_handler(hw, 3);
 
 	if (cause & PCIE)
-		csio_pcie_intr_handler(hw);
+		hw->chip_ops->chip_pcie_intr_handler(hw);
 
 	if (cause & MC)
 		csio_mem_intr_handler(hw, MEM_MC);
@@ -4257,6 +3930,7 @@ csio_hw_get_device_id(struct csio_hw *hw)
 			     &hw->params.pci.device_id);
 
 	csio_dev_id_cached(hw);
+	hw->chip_id = (hw->params.pci.device_id & CSIO_HW_CHIP_MASK);
 
 } /* csio_hw_get_device_id */
 
@@ -4275,19 +3949,21 @@ csio_hw_set_description(struct csio_hw *hw, uint16_t ven_id, uint16_t dev_id)
 		prot_type = (dev_id & CSIO_ASIC_DEVID_PROTO_MASK);
 		adap_type = (dev_id & CSIO_ASIC_DEVID_TYPE_MASK);
 
-		if (prot_type == CSIO_FPGA) {
+		if (prot_type == CSIO_T4_FCOE_ASIC) {
+			memcpy(hw->hw_ver,
+			       csio_t4_fcoe_adapters[adap_type].model_no, 16);
 			memcpy(hw->model_desc,
-				csio_fcoe_adapters[13].description, 32);
-		} else if (prot_type == CSIO_T4_FCOE_ASIC) {
+			       csio_t4_fcoe_adapters[adap_type].description,
+			       32);
+		} else if (prot_type == CSIO_T5_FCOE_ASIC) {
 			memcpy(hw->hw_ver,
-			       csio_fcoe_adapters[adap_type].model_no, 16);
+			       csio_t5_fcoe_adapters[adap_type].model_no, 16);
 			memcpy(hw->model_desc,
-				csio_fcoe_adapters[adap_type].description, 32);
+			       csio_t5_fcoe_adapters[adap_type].description,
+			       32);
 		} else {
 			char tempName[32] = "Chelsio FCoE Controller";
 			memcpy(hw->model_desc, tempName, 32);
-
-			CSIO_DB_ASSERT(0);
 		}
 	}
 } /* csio_hw_set_description */
@@ -4316,6 +3992,9 @@ csio_hw_init(struct csio_hw *hw)
 
 	strcpy(hw->name, CSIO_HW_NAME);
 
+	/* Initialize the HW chip ops with T4/T5 specific ops */
+	hw->chip_ops = csio_is_t4(hw->chip_id) ? &t4_ops : &t5_ops;
+
 	/* Set the model & its description */
 
 	ven_id = hw->params.pci.vendor_id;
diff --git a/drivers/scsi/csiostor/csio_hw.h b/drivers/scsi/csiostor/csio_hw.h
index 9edcca4c71af..489fc095cb03 100644
--- a/drivers/scsi/csiostor/csio_hw.h
+++ b/drivers/scsi/csiostor/csio_hw.h
@@ -48,6 +48,7 @@
 #include <scsi/scsi_device.h>
 #include <scsi/scsi_transport_fc.h>
 
+#include "csio_hw_chip.h"
 #include "csio_wr.h"
 #include "csio_mb.h"
 #include "csio_scsi.h"
@@ -60,13 +61,6 @@
  */
 #define	FW_HOSTERROR			255
 
-#define CSIO_FW_FNAME		"cxgb4/t4fw.bin"
-#define CSIO_CF_FNAME		"cxgb4/t4-config.txt"
-
-#define FW_VERSION_MAJOR	1
-#define FW_VERSION_MINOR	2
-#define FW_VERSION_MICRO	8
-
 #define CSIO_HW_NAME		"Chelsio FCoE Adapter"
 #define CSIO_MAX_PFN		8
 #define CSIO_MAX_PPORTS		4
@@ -123,8 +117,6 @@ extern int csio_msi;
 #define CSIO_VENDOR_ID				0x1425
 #define CSIO_ASIC_DEVID_PROTO_MASK		0xFF00
 #define CSIO_ASIC_DEVID_TYPE_MASK		0x00FF
-#define CSIO_FPGA				0xA000
-#define CSIO_T4_FCOE_ASIC			0x4600
 
 #define CSIO_GLBL_INTR_MASK		(CIM | MPS | PL | PCIE | MC | EDC0 | \
 					 EDC1 | LE | TP | MA | PM_TX | PM_RX | \
@@ -207,17 +199,6 @@ enum {
 	SF_SIZE = SF_SEC_SIZE * 16,   /* serial flash size */
 };
 
-enum { MEM_EDC0, MEM_EDC1, MEM_MC };
-
-enum {
-	MEMWIN0_APERTURE = 2048,
-	MEMWIN0_BASE     = 0x1b800,
-	MEMWIN1_APERTURE = 32768,
-	MEMWIN1_BASE     = 0x28000,
-	MEMWIN2_APERTURE = 65536,
-	MEMWIN2_BASE     = 0x30000,
-};
-
 /* serial flash and firmware constants */
 enum {
 	SF_ATTEMPTS = 10,             /* max retries for SF operations */
@@ -239,9 +220,6 @@ enum {
 	FLASH_CFG_MAX_SIZE    = 0x10000 , /* max size of the flash config file*/
 	FLASH_CFG_OFFSET      = 0x1f0000,
 	FLASH_CFG_START_SEC   = FLASH_CFG_OFFSET / SF_SEC_SIZE,
-	FPGA_FLASH_CFG_OFFSET = 0xf0000 , /* if FPGA mode, then cfg file is
-					   * at 1MB - 64KB */
-	FPGA_FLASH_CFG_START_SEC  = FPGA_FLASH_CFG_OFFSET / SF_SEC_SIZE,
 };
 
 /*
@@ -259,6 +237,8 @@ enum {
 	FLASH_FW_START = FLASH_START(FLASH_FW_START_SEC),
 	FLASH_FW_MAX_SIZE = FLASH_MAX_SIZE(FLASH_FW_NSECS),
 
+	/* Location of Firmware Configuration File in FLASH. */
+	FLASH_CFG_START = FLASH_START(FLASH_CFG_START_SEC),
 };
 
 #undef FLASH_START
@@ -310,7 +290,7 @@ struct csio_adap_desc {
 struct pci_params {
 	uint16_t   vendor_id;
 	uint16_t   device_id;
-	uint32_t   vpd_cap_addr;
+	int        vpd_cap_addr;
 	uint16_t   speed;
 	uint8_t    width;
 };
@@ -513,6 +493,7 @@ struct csio_hw {
 	uint32_t		fwrev;
 	uint32_t		tp_vers;
 	char			chip_ver;
+	uint16_t		chip_id;		/* Tells T4/T5 chip */
 	uint32_t		cfg_finiver;
 	uint32_t		cfg_finicsum;
 	uint32_t		cfg_cfcsum;
@@ -556,6 +537,9 @@ struct csio_hw {
 							 */
 
 	struct csio_fcoe_res_info  fres_info;		/* Fcoe resource info */
+	struct csio_hw_chip_ops	*chip_ops;		/* T4/T5 Chip specific
+							 * Operations
+							 */
 
 	/* MSIX vectors */
 	struct csio_msix_entries msix_entries[CSIO_MAX_MSIX_VECS];
@@ -636,9 +620,16 @@ csio_us_to_core_ticks(struct csio_hw *hw, uint32_t us)
 #define csio_dbg(__hw, __fmt, ...)
 #endif
 
+int csio_hw_wait_op_done_val(struct csio_hw *, int, uint32_t, int,
+			     int, int, uint32_t *);
+void csio_hw_tp_wr_bits_indirect(struct csio_hw *, unsigned int,
+				 unsigned int, unsigned int);
 int csio_mgmt_req_lookup(struct csio_mgmtm *, struct csio_ioreq *);
 void csio_hw_intr_disable(struct csio_hw *);
-int csio_hw_slow_intr_handler(struct csio_hw *hw);
+int csio_hw_slow_intr_handler(struct csio_hw *);
+int csio_handle_intr_status(struct csio_hw *, unsigned int,
+			    const struct intr_info *);
+
 int csio_hw_start(struct csio_hw *);
 int csio_hw_stop(struct csio_hw *);
 int csio_hw_reset(struct csio_hw *);
@@ -647,19 +638,17 @@ int csio_is_hw_removing(struct csio_hw *);
 
 int csio_fwevtq_handler(struct csio_hw *);
 void csio_evtq_worker(struct work_struct *);
-int csio_enqueue_evt(struct csio_hw *hw, enum csio_evt type,
-				void *evt_msg, uint16_t len);
+int csio_enqueue_evt(struct csio_hw *, enum csio_evt, void *, uint16_t);
 void csio_evtq_flush(struct csio_hw *hw);
 
 int csio_request_irqs(struct csio_hw *);
 void csio_intr_enable(struct csio_hw *);
 void csio_intr_disable(struct csio_hw *, bool);
+void csio_hw_fatal_err(struct csio_hw *);
 
 struct csio_lnode *csio_lnode_alloc(struct csio_hw *);
 int csio_config_queues(struct csio_hw *);
 
-int csio_hw_mc_read(struct csio_hw *, uint32_t, __be32 *, uint64_t *);
-int csio_hw_edc_read(struct csio_hw *, int, uint32_t, __be32 *, uint64_t *);
 int csio_hw_init(struct csio_hw *);
 void csio_hw_exit(struct csio_hw *);
 #endif /* ifndef __CSIO_HW_H__ */
diff --git a/drivers/scsi/csiostor/csio_hw_chip.h b/drivers/scsi/csiostor/csio_hw_chip.h
new file mode 100644
index 000000000000..bca0de61ae80
--- /dev/null
+++ b/drivers/scsi/csiostor/csio_hw_chip.h
@@ -0,0 +1,175 @@
+/*
+ * This file is part of the Chelsio FCoE driver for Linux.
+ *
+ * Copyright (c) 2008-2013 Chelsio Communications, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#ifndef __CSIO_HW_CHIP_H__
+#define __CSIO_HW_CHIP_H__
+
+#include "csio_defs.h"
+
+/* FCoE device IDs for T4 */
+#define CSIO_DEVID_T440DBG_FCOE			0x4600
+#define CSIO_DEVID_T420CR_FCOE			0x4601
+#define CSIO_DEVID_T422CR_FCOE			0x4602
+#define CSIO_DEVID_T440CR_FCOE			0x4603
+#define CSIO_DEVID_T420BCH_FCOE			0x4604
+#define CSIO_DEVID_T440BCH_FCOE			0x4605
+#define CSIO_DEVID_T440CH_FCOE			0x4606
+#define CSIO_DEVID_T420SO_FCOE			0x4607
+#define CSIO_DEVID_T420CX_FCOE			0x4608
+#define CSIO_DEVID_T420BT_FCOE			0x4609
+#define CSIO_DEVID_T404BT_FCOE			0x460A
+#define CSIO_DEVID_B420_FCOE			0x460B
+#define CSIO_DEVID_B404_FCOE			0x460C
+#define CSIO_DEVID_T480CR_FCOE			0x460D
+#define CSIO_DEVID_T440LPCR_FCOE		0x460E
+#define CSIO_DEVID_AMSTERDAM_T4_FCOE		0x460F
+#define CSIO_DEVID_HUAWEI_T480_FCOE		0x4680
+#define CSIO_DEVID_HUAWEI_T440_FCOE		0x4681
+#define CSIO_DEVID_HUAWEI_STG310_FCOE		0x4682
+#define CSIO_DEVID_ACROMAG_XMC_XAUI		0x4683
+#define CSIO_DEVID_ACROMAG_XMC_SFP_FCOE		0x4684
+#define CSIO_DEVID_QUANTA_MEZZ_SFP_FCOE		0x4685
+#define CSIO_DEVID_HUAWEI_10GT_FCOE		0x4686
+#define CSIO_DEVID_HUAWEI_T440_TOE_FCOE		0x4687
+
+/* FCoE device IDs for T5 */
+#define CSIO_DEVID_T580DBG_FCOE			0x5600
+#define CSIO_DEVID_T520CR_FCOE			0x5601
+#define CSIO_DEVID_T522CR_FCOE			0x5602
+#define CSIO_DEVID_T540CR_FCOE			0x5603
+#define CSIO_DEVID_T520BCH_FCOE			0x5604
+#define CSIO_DEVID_T540BCH_FCOE			0x5605
+#define CSIO_DEVID_T540CH_FCOE			0x5606
+#define CSIO_DEVID_T520SO_FCOE			0x5607
+#define CSIO_DEVID_T520CX_FCOE			0x5608
+#define CSIO_DEVID_T520BT_FCOE			0x5609
+#define CSIO_DEVID_T504BT_FCOE			0x560A
+#define CSIO_DEVID_B520_FCOE			0x560B
+#define CSIO_DEVID_B504_FCOE			0x560C
+#define CSIO_DEVID_T580CR2_FCOE			0x560D
+#define CSIO_DEVID_T540LPCR_FCOE		0x560E
+#define CSIO_DEVID_AMSTERDAM_T5_FCOE		0x560F
+#define CSIO_DEVID_T580LPCR_FCOE		0x5610
+#define CSIO_DEVID_T520LLCR_FCOE		0x5611
+#define CSIO_DEVID_T560CR_FCOE			0x5612
+#define CSIO_DEVID_T580CR_FCOE			0x5613
+
+/* Define MACRO values */
+#define CSIO_HW_T4				0x4000
+#define CSIO_T4_FCOE_ASIC			0x4600
+#define CSIO_HW_T5				0x5000
+#define CSIO_T5_FCOE_ASIC			0x5600
+#define CSIO_HW_CHIP_MASK			0xF000
+#define T4_REGMAP_SIZE				(160 * 1024)
+#define T5_REGMAP_SIZE				(332 * 1024)
+#define FW_FNAME_T4				"cxgb4/t4fw.bin"
+#define FW_FNAME_T5				"cxgb4/t5fw.bin"
+#define FW_CFG_NAME_T4				"cxgb4/t4-config.txt"
+#define FW_CFG_NAME_T5				"cxgb4/t5-config.txt"
+
+/* Define static functions */
+static inline int csio_is_t4(uint16_t chip)
+{
+	return (chip == CSIO_HW_T4);
+}
+
+static inline int csio_is_t5(uint16_t chip)
+{
+	return (chip == CSIO_HW_T5);
+}
+
+/* Define MACRO DEFINITIONS */
+#define CSIO_DEVICE(devid, idx)						\
+	{ PCI_VENDOR_ID_CHELSIO, (devid), PCI_ANY_ID, PCI_ANY_ID, 0, 0, (idx) }
+
+#define CSIO_HW_PIDX(hw, index)						\
+	(csio_is_t4(hw->chip_id) ? (PIDX(index)) :			\
+					(PIDX_T5(index) | DBTYPE(1U)))
+
+#define CSIO_HW_LP_INT_THRESH(hw, val)					\
+	(csio_is_t4(hw->chip_id) ? (LP_INT_THRESH(val)) :		\
+					(V_LP_INT_THRESH_T5(val)))
+
+#define CSIO_HW_M_LP_INT_THRESH(hw)					\
+	(csio_is_t4(hw->chip_id) ? (LP_INT_THRESH_MASK) : (M_LP_INT_THRESH_T5))
+
+#define CSIO_MAC_INT_CAUSE_REG(hw, port)				\
+	(csio_is_t4(hw->chip_id) ? (PORT_REG(port, XGMAC_PORT_INT_CAUSE)) : \
+				(T5_PORT_REG(port, MAC_PORT_INT_CAUSE)))
+
+#define FW_VERSION_MAJOR(hw) (csio_is_t4(hw->chip_id) ? 1 : 0)
+#define FW_VERSION_MINOR(hw) (csio_is_t4(hw->chip_id) ? 2 : 0)
+#define FW_VERSION_MICRO(hw) (csio_is_t4(hw->chip_id) ? 8 : 0)
+
+#define CSIO_FW_FNAME(hw)						\
+	(csio_is_t4(hw->chip_id) ? FW_FNAME_T4 : FW_FNAME_T5)
+
+#define CSIO_CF_FNAME(hw)						\
+	(csio_is_t4(hw->chip_id) ? FW_CFG_NAME_T4 : FW_CFG_NAME_T5)
+
+/* Declare ENUMS */
+enum { MEM_EDC0, MEM_EDC1, MEM_MC, MEM_MC0 = MEM_MC, MEM_MC1 };
+
+enum {
+	MEMWIN_APERTURE = 2048,
+	MEMWIN_BASE     = 0x1b800,
+	MEMWIN_CSIOSTOR = 6,		/* PCI-e Memory Window access */
+};
+
+/* Slow path handlers */
+struct intr_info {
+	unsigned int mask;       /* bits to check in interrupt status */
+	const char *msg;         /* message to print or NULL */
+	short stat_idx;          /* stat counter to increment or -1 */
+	unsigned short fatal;    /* whether the condition reported is fatal */
+};
+
+/* T4/T5 Chip specific ops */
+struct csio_hw;
+struct csio_hw_chip_ops {
+	int (*chip_set_mem_win)(struct csio_hw *, uint32_t);
+	void (*chip_pcie_intr_handler)(struct csio_hw *);
+	uint32_t (*chip_flash_cfg_addr)(struct csio_hw *);
+	int (*chip_mc_read)(struct csio_hw *, int, uint32_t,
+					__be32 *, uint64_t *);
+	int (*chip_edc_read)(struct csio_hw *, int, uint32_t,
+					__be32 *, uint64_t *);
+	int (*chip_memory_rw)(struct csio_hw *, u32, int, u32,
+					u32, uint32_t *, int);
+	void (*chip_dfs_create_ext_mem)(struct csio_hw *);
+};
+
+extern struct csio_hw_chip_ops t4_ops;
+extern struct csio_hw_chip_ops t5_ops;
+
+#endif /* #ifndef __CSIO_HW_CHIP_H__ */
diff --git a/drivers/scsi/csiostor/csio_hw_t4.c b/drivers/scsi/csiostor/csio_hw_t4.c
new file mode 100644
index 000000000000..89ecbac5478f
--- /dev/null
+++ b/drivers/scsi/csiostor/csio_hw_t4.c
@@ -0,0 +1,403 @@
+/*
+ * This file is part of the Chelsio FCoE driver for Linux.
+ *
+ * Copyright (c) 2008-2013 Chelsio Communications, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#include "csio_hw.h"
+#include "csio_init.h"
+
+/*
+ * Return the specified PCI-E Configuration Space register from our Physical
+ * Function.  We try first via a Firmware LDST Command since we prefer to let
+ * the firmware own all of these registers, but if that fails we go for it
+ * directly ourselves.
+ */
+static uint32_t
+csio_t4_read_pcie_cfg4(struct csio_hw *hw, int reg)
+{
+	u32 val = 0;
+	struct csio_mb *mbp;
+	int rv;
+	struct fw_ldst_cmd *ldst_cmd;
+
+	mbp = mempool_alloc(hw->mb_mempool, GFP_ATOMIC);
+	if (!mbp) {
+		CSIO_INC_STATS(hw, n_err_nomem);
+		pci_read_config_dword(hw->pdev, reg, &val);
+		return val;
+	}
+
+	csio_mb_ldst(hw, mbp, CSIO_MB_DEFAULT_TMO, reg);
+	rv = csio_mb_issue(hw, mbp);
+
+	/*
+	 * If the LDST Command suucceeded, exctract the returned register
+	 * value.  Otherwise read it directly ourself.
+	 */
+	if (rv == 0) {
+		ldst_cmd = (struct fw_ldst_cmd *)(mbp->mb);
+		val = ntohl(ldst_cmd->u.pcie.data[0]);
+	} else
+		pci_read_config_dword(hw->pdev, reg, &val);
+
+	mempool_free(mbp, hw->mb_mempool);
+
+	return val;
+}
+
+static int
+csio_t4_set_mem_win(struct csio_hw *hw, uint32_t win)
+{
+	u32 bar0;
+	u32 mem_win_base;
+
+	/*
+	 * Truncation intentional: we only read the bottom 32-bits of the
+	 * 64-bit BAR0/BAR1 ...  We use the hardware backdoor mechanism to
+	 * read BAR0 instead of using pci_resource_start() because we could be
+	 * operating from within a Virtual Machine which is trapping our
+	 * accesses to our Configuration Space and we need to set up the PCI-E
+	 * Memory Window decoders with the actual addresses which will be
+	 * coming across the PCI-E link.
+	 */
+	bar0 = csio_t4_read_pcie_cfg4(hw, PCI_BASE_ADDRESS_0);
+	bar0 &= PCI_BASE_ADDRESS_MEM_MASK;
+
+	mem_win_base = bar0 + MEMWIN_BASE;
+
+	/*
+	 * Set up memory window for accessing adapter memory ranges.  (Read
+	 * back MA register to ensure that changes propagate before we attempt
+	 * to use the new values.)
+	 */
+	csio_wr_reg32(hw, mem_win_base | BIR(0) |
+			  WINDOW(ilog2(MEMWIN_APERTURE) - 10),
+			  PCIE_MEM_ACCESS_REG(PCIE_MEM_ACCESS_BASE_WIN, win));
+	csio_rd_reg32(hw,
+		      PCIE_MEM_ACCESS_REG(PCIE_MEM_ACCESS_BASE_WIN, win));
+	return 0;
+}
+
+/*
+ * Interrupt handler for the PCIE module.
+ */
+static void
+csio_t4_pcie_intr_handler(struct csio_hw *hw)
+{
+	static struct intr_info sysbus_intr_info[] = {
+		{ RNPP, "RXNP array parity error", -1, 1 },
+		{ RPCP, "RXPC array parity error", -1, 1 },
+		{ RCIP, "RXCIF array parity error", -1, 1 },
+		{ RCCP, "Rx completions control array parity error", -1, 1 },
+		{ RFTP, "RXFT array parity error", -1, 1 },
+		{ 0, NULL, 0, 0 }
+	};
+	static struct intr_info pcie_port_intr_info[] = {
+		{ TPCP, "TXPC array parity error", -1, 1 },
+		{ TNPP, "TXNP array parity error", -1, 1 },
+		{ TFTP, "TXFT array parity error", -1, 1 },
+		{ TCAP, "TXCA array parity error", -1, 1 },
+		{ TCIP, "TXCIF array parity error", -1, 1 },
+		{ RCAP, "RXCA array parity error", -1, 1 },
+		{ OTDD, "outbound request TLP discarded", -1, 1 },
+		{ RDPE, "Rx data parity error", -1, 1 },
+		{ TDUE, "Tx uncorrectable data error", -1, 1 },
+		{ 0, NULL, 0, 0 }
+	};
+
+	static struct intr_info pcie_intr_info[] = {
+		{ MSIADDRLPERR, "MSI AddrL parity error", -1, 1 },
+		{ MSIADDRHPERR, "MSI AddrH parity error", -1, 1 },
+		{ MSIDATAPERR, "MSI data parity error", -1, 1 },
+		{ MSIXADDRLPERR, "MSI-X AddrL parity error", -1, 1 },
+		{ MSIXADDRHPERR, "MSI-X AddrH parity error", -1, 1 },
+		{ MSIXDATAPERR, "MSI-X data parity error", -1, 1 },
+		{ MSIXDIPERR, "MSI-X DI parity error", -1, 1 },
+		{ PIOCPLPERR, "PCI PIO completion FIFO parity error", -1, 1 },
+		{ PIOREQPERR, "PCI PIO request FIFO parity error", -1, 1 },
+		{ TARTAGPERR, "PCI PCI target tag FIFO parity error", -1, 1 },
+		{ CCNTPERR, "PCI CMD channel count parity error", -1, 1 },
+		{ CREQPERR, "PCI CMD channel request parity error", -1, 1 },
+		{ CRSPPERR, "PCI CMD channel response parity error", -1, 1 },
+		{ DCNTPERR, "PCI DMA channel count parity error", -1, 1 },
+		{ DREQPERR, "PCI DMA channel request parity error", -1, 1 },
+		{ DRSPPERR, "PCI DMA channel response parity error", -1, 1 },
+		{ HCNTPERR, "PCI HMA channel count parity error", -1, 1 },
+		{ HREQPERR, "PCI HMA channel request parity error", -1, 1 },
+		{ HRSPPERR, "PCI HMA channel response parity error", -1, 1 },
+		{ CFGSNPPERR, "PCI config snoop FIFO parity error", -1, 1 },
+		{ FIDPERR, "PCI FID parity error", -1, 1 },
+		{ INTXCLRPERR, "PCI INTx clear parity error", -1, 1 },
+		{ MATAGPERR, "PCI MA tag parity error", -1, 1 },
+		{ PIOTAGPERR, "PCI PIO tag parity error", -1, 1 },
+		{ RXCPLPERR, "PCI Rx completion parity error", -1, 1 },
+		{ RXWRPERR, "PCI Rx write parity error", -1, 1 },
+		{ RPLPERR, "PCI replay buffer parity error", -1, 1 },
+		{ PCIESINT, "PCI core secondary fault", -1, 1 },
+		{ PCIEPINT, "PCI core primary fault", -1, 1 },
+		{ UNXSPLCPLERR, "PCI unexpected split completion error", -1,
+		  0 },
+		{ 0, NULL, 0, 0 }
+	};
+
+	int fat;
+	fat = csio_handle_intr_status(hw,
+				      PCIE_CORE_UTL_SYSTEM_BUS_AGENT_STATUS,
+				      sysbus_intr_info) +
+	      csio_handle_intr_status(hw,
+				      PCIE_CORE_UTL_PCI_EXPRESS_PORT_STATUS,
+				      pcie_port_intr_info) +
+	      csio_handle_intr_status(hw, PCIE_INT_CAUSE, pcie_intr_info);
+	if (fat)
+		csio_hw_fatal_err(hw);
+}
+
+/*
+ * csio_t4_flash_cfg_addr - return the address of the flash configuration file
+ * @hw: the HW module
+ *
+ * Return the address within the flash where the Firmware Configuration
+ * File is stored.
+ */
+static unsigned int
+csio_t4_flash_cfg_addr(struct csio_hw *hw)
+{
+	return FLASH_CFG_OFFSET;
+}
+
+/*
+ *      csio_t4_mc_read - read from MC through backdoor accesses
+ *      @hw: the hw module
+ *      @idx: not used for T4 adapter
+ *      @addr: address of first byte requested
+ *      @data: 64 bytes of data containing the requested address
+ *      @ecc: where to store the corresponding 64-bit ECC word
+ *
+ *      Read 64 bytes of data from MC starting at a 64-byte-aligned address
+ *      that covers the requested address @addr.  If @parity is not %NULL it
+ *      is assigned the 64-bit ECC word for the read data.
+ */
+static int
+csio_t4_mc_read(struct csio_hw *hw, int idx, uint32_t addr, __be32 *data,
+		uint64_t *ecc)
+{
+	int i;
+
+	if (csio_rd_reg32(hw, MC_BIST_CMD) & START_BIST)
+		return -EBUSY;
+	csio_wr_reg32(hw, addr & ~0x3fU, MC_BIST_CMD_ADDR);
+	csio_wr_reg32(hw, 64, MC_BIST_CMD_LEN);
+	csio_wr_reg32(hw, 0xc, MC_BIST_DATA_PATTERN);
+	csio_wr_reg32(hw, BIST_OPCODE(1) | START_BIST | BIST_CMD_GAP(1),
+		      MC_BIST_CMD);
+	i = csio_hw_wait_op_done_val(hw, MC_BIST_CMD, START_BIST,
+				     0, 10, 1, NULL);
+	if (i)
+		return i;
+
+#define MC_DATA(i) MC_BIST_STATUS_REG(MC_BIST_STATUS_RDATA, i)
+
+	for (i = 15; i >= 0; i--)
+		*data++ = htonl(csio_rd_reg32(hw, MC_DATA(i)));
+	if (ecc)
+		*ecc = csio_rd_reg64(hw, MC_DATA(16));
+#undef MC_DATA
+	return 0;
+}
+
+/*
+ *      csio_t4_edc_read - read from EDC through backdoor accesses
+ *      @hw: the hw module
+ *      @idx: which EDC to access
+ *      @addr: address of first byte requested
+ *      @data: 64 bytes of data containing the requested address
+ *      @ecc: where to store the corresponding 64-bit ECC word
+ *
+ *      Read 64 bytes of data from EDC starting at a 64-byte-aligned address
+ *      that covers the requested address @addr.  If @parity is not %NULL it
+ *      is assigned the 64-bit ECC word for the read data.
+ */
+static int
+csio_t4_edc_read(struct csio_hw *hw, int idx, uint32_t addr, __be32 *data,
+		uint64_t *ecc)
+{
+	int i;
+
+	idx *= EDC_STRIDE;
+	if (csio_rd_reg32(hw, EDC_BIST_CMD + idx) & START_BIST)
+		return -EBUSY;
+	csio_wr_reg32(hw, addr & ~0x3fU, EDC_BIST_CMD_ADDR + idx);
+	csio_wr_reg32(hw, 64, EDC_BIST_CMD_LEN + idx);
+	csio_wr_reg32(hw, 0xc, EDC_BIST_DATA_PATTERN + idx);
+	csio_wr_reg32(hw, BIST_OPCODE(1) | BIST_CMD_GAP(1) | START_BIST,
+		      EDC_BIST_CMD + idx);
+	i = csio_hw_wait_op_done_val(hw, EDC_BIST_CMD + idx, START_BIST,
+				     0, 10, 1, NULL);
+	if (i)
+		return i;
+
+#define EDC_DATA(i) (EDC_BIST_STATUS_REG(EDC_BIST_STATUS_RDATA, i) + idx)
+
+	for (i = 15; i >= 0; i--)
+		*data++ = htonl(csio_rd_reg32(hw, EDC_DATA(i)));
+	if (ecc)
+		*ecc = csio_rd_reg64(hw, EDC_DATA(16));
+#undef EDC_DATA
+	return 0;
+}
+
+/*
+ * csio_t4_memory_rw - read/write EDC 0, EDC 1 or MC via PCIE memory window
+ * @hw: the csio_hw
+ * @win: PCI-E memory Window to use
+ * @mtype: memory type: MEM_EDC0, MEM_EDC1, MEM_MC0 (or MEM_MC) or MEM_MC1
+ * @addr: address within indicated memory type
+ * @len: amount of memory to transfer
+ * @buf: host memory buffer
+ * @dir: direction of transfer 1 => read, 0 => write
+ *
+ * Reads/writes an [almost] arbitrary memory region in the firmware: the
+ * firmware memory address, length and host buffer must be aligned on
+ * 32-bit boudaries.  The memory is transferred as a raw byte sequence
+ * from/to the firmware's memory.  If this memory contains data
+ * structures which contain multi-byte integers, it's the callers
+ * responsibility to perform appropriate byte order conversions.
+ */
+static int
+csio_t4_memory_rw(struct csio_hw *hw, u32 win, int mtype, u32 addr,
+		u32 len, uint32_t *buf, int dir)
+{
+	u32 pos, start, offset, memoffset, bar0;
+	u32 edc_size, mc_size, mem_reg, mem_aperture, mem_base;
+
+	/*
+	 * Argument sanity checks ...
+	 */
+	if ((addr & 0x3) || (len & 0x3))
+		return -EINVAL;
+
+	/* Offset into the region of memory which is being accessed
+	 * MEM_EDC0 = 0
+	 * MEM_EDC1 = 1
+	 * MEM_MC   = 2 -- T4
+	 */
+	edc_size  = EDRAM_SIZE_GET(csio_rd_reg32(hw, MA_EDRAM0_BAR));
+	if (mtype != MEM_MC1)
+		memoffset = (mtype * (edc_size * 1024 * 1024));
+	else {
+		mc_size = EXT_MEM_SIZE_GET(csio_rd_reg32(hw,
+							 MA_EXT_MEMORY_BAR));
+		memoffset = (MEM_MC0 * edc_size + mc_size) * 1024 * 1024;
+	}
+
+	/* Determine the PCIE_MEM_ACCESS_OFFSET */
+	addr = addr + memoffset;
+
+	/*
+	 * Each PCI-E Memory Window is programmed with a window size -- or
+	 * "aperture" -- which controls the granularity of its mapping onto
+	 * adapter memory.  We need to grab that aperture in order to know
+	 * how to use the specified window.  The window is also programmed
+	 * with the base address of the Memory Window in BAR0's address
+	 * space.  For T4 this is an absolute PCI-E Bus Address.  For T5
+	 * the address is relative to BAR0.
+	 */
+	mem_reg = csio_rd_reg32(hw,
+			PCIE_MEM_ACCESS_REG(PCIE_MEM_ACCESS_BASE_WIN, win));
+	mem_aperture = 1 << (WINDOW(mem_reg) + 10);
+	mem_base = GET_PCIEOFST(mem_reg) << 10;
+
+	bar0 = csio_t4_read_pcie_cfg4(hw, PCI_BASE_ADDRESS_0);
+	bar0 &= PCI_BASE_ADDRESS_MEM_MASK;
+	mem_base -= bar0;
+
+	start = addr & ~(mem_aperture-1);
+	offset = addr - start;
+
+	csio_dbg(hw, "csio_t4_memory_rw: mem_reg: 0x%x, mem_aperture: 0x%x\n",
+		 mem_reg, mem_aperture);
+	csio_dbg(hw, "csio_t4_memory_rw: mem_base: 0x%x, mem_offset: 0x%x\n",
+		 mem_base, memoffset);
+	csio_dbg(hw, "csio_t4_memory_rw: bar0: 0x%x, start:0x%x, offset:0x%x\n",
+		 bar0, start, offset);
+	csio_dbg(hw, "csio_t4_memory_rw: mtype: %d, addr: 0x%x, len: %d\n",
+		 mtype, addr, len);
+
+	for (pos = start; len > 0; pos += mem_aperture, offset = 0) {
+		/*
+		 * Move PCI-E Memory Window to our current transfer
+		 * position.  Read it back to ensure that changes propagate
+		 * before we attempt to use the new value.
+		 */
+		csio_wr_reg32(hw, pos,
+			PCIE_MEM_ACCESS_REG(PCIE_MEM_ACCESS_OFFSET, win));
+		csio_rd_reg32(hw,
+			PCIE_MEM_ACCESS_REG(PCIE_MEM_ACCESS_OFFSET, win));
+
+		while (offset < mem_aperture && len > 0) {
+			if (dir)
+				*buf++ = csio_rd_reg32(hw, mem_base + offset);
+			else
+				csio_wr_reg32(hw, *buf++, mem_base + offset);
+
+			offset += sizeof(__be32);
+			len -= sizeof(__be32);
+		}
+	}
+	return 0;
+}
+
+/*
+ * csio_t4_dfs_create_ext_mem - setup debugfs for MC to read the values
+ * @hw: the csio_hw
+ *
+ * This function creates files in the debugfs with external memory region MC.
+ */
+static void
+csio_t4_dfs_create_ext_mem(struct csio_hw *hw)
+{
+	u32 size;
+	int i = csio_rd_reg32(hw, MA_TARGET_MEM_ENABLE);
+	if (i & EXT_MEM_ENABLE) {
+		size = csio_rd_reg32(hw, MA_EXT_MEMORY_BAR);
+		csio_add_debugfs_mem(hw, "mc", MEM_MC,
+				     EXT_MEM_SIZE_GET(size));
+	}
+}
+
+/* T4 adapter specific function */
+struct csio_hw_chip_ops t4_ops = {
+	.chip_set_mem_win		= csio_t4_set_mem_win,
+	.chip_pcie_intr_handler		= csio_t4_pcie_intr_handler,
+	.chip_flash_cfg_addr		= csio_t4_flash_cfg_addr,
+	.chip_mc_read			= csio_t4_mc_read,
+	.chip_edc_read			= csio_t4_edc_read,
+	.chip_memory_rw			= csio_t4_memory_rw,
+	.chip_dfs_create_ext_mem	= csio_t4_dfs_create_ext_mem,
+};
diff --git a/drivers/scsi/csiostor/csio_hw_t5.c b/drivers/scsi/csiostor/csio_hw_t5.c
new file mode 100644
index 000000000000..27745c170c24
--- /dev/null
+++ b/drivers/scsi/csiostor/csio_hw_t5.c
@@ -0,0 +1,397 @@
+/*
+ * This file is part of the Chelsio FCoE driver for Linux.
+ *
+ * Copyright (c) 2008-2013 Chelsio Communications, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#include "csio_hw.h"
+#include "csio_init.h"
+
+static int
+csio_t5_set_mem_win(struct csio_hw *hw, uint32_t win)
+{
+	u32 mem_win_base;
+	/*
+	 * Truncation intentional: we only read the bottom 32-bits of the
+	 * 64-bit BAR0/BAR1 ...  We use the hardware backdoor mechanism to
+	 * read BAR0 instead of using pci_resource_start() because we could be
+	 * operating from within a Virtual Machine which is trapping our
+	 * accesses to our Configuration Space and we need to set up the PCI-E
+	 * Memory Window decoders with the actual addresses which will be
+	 * coming across the PCI-E link.
+	 */
+
+	/* For T5, only relative offset inside the PCIe BAR is passed */
+	mem_win_base = MEMWIN_BASE;
+
+	/*
+	 * Set up memory window for accessing adapter memory ranges.  (Read
+	 * back MA register to ensure that changes propagate before we attempt
+	 * to use the new values.)
+	 */
+	csio_wr_reg32(hw, mem_win_base | BIR(0) |
+			  WINDOW(ilog2(MEMWIN_APERTURE) - 10),
+			  PCIE_MEM_ACCESS_REG(PCIE_MEM_ACCESS_BASE_WIN, win));
+	csio_rd_reg32(hw,
+		      PCIE_MEM_ACCESS_REG(PCIE_MEM_ACCESS_BASE_WIN, win));
+
+	return 0;
+}
+
+/*
+ * Interrupt handler for the PCIE module.
+ */
+static void
+csio_t5_pcie_intr_handler(struct csio_hw *hw)
+{
+	static struct intr_info sysbus_intr_info[] = {
+		{ RNPP, "RXNP array parity error", -1, 1 },
+		{ RPCP, "RXPC array parity error", -1, 1 },
+		{ RCIP, "RXCIF array parity error", -1, 1 },
+		{ RCCP, "Rx completions control array parity error", -1, 1 },
+		{ RFTP, "RXFT array parity error", -1, 1 },
+		{ 0, NULL, 0, 0 }
+	};
+	static struct intr_info pcie_port_intr_info[] = {
+		{ TPCP, "TXPC array parity error", -1, 1 },
+		{ TNPP, "TXNP array parity error", -1, 1 },
+		{ TFTP, "TXFT array parity error", -1, 1 },
+		{ TCAP, "TXCA array parity error", -1, 1 },
+		{ TCIP, "TXCIF array parity error", -1, 1 },
+		{ RCAP, "RXCA array parity error", -1, 1 },
+		{ OTDD, "outbound request TLP discarded", -1, 1 },
+		{ RDPE, "Rx data parity error", -1, 1 },
+		{ TDUE, "Tx uncorrectable data error", -1, 1 },
+		{ 0, NULL, 0, 0 }
+	};
+
+	static struct intr_info pcie_intr_info[] = {
+		{ MSTGRPPERR, "Master Response Read Queue parity error",
+		-1, 1 },
+		{ MSTTIMEOUTPERR, "Master Timeout FIFO parity error", -1, 1 },
+		{ MSIXSTIPERR, "MSI-X STI SRAM parity error", -1, 1 },
+		{ MSIXADDRLPERR, "MSI-X AddrL parity error", -1, 1 },
+		{ MSIXADDRHPERR, "MSI-X AddrH parity error", -1, 1 },
+		{ MSIXDATAPERR, "MSI-X data parity error", -1, 1 },
+		{ MSIXDIPERR, "MSI-X DI parity error", -1, 1 },
+		{ PIOCPLGRPPERR, "PCI PIO completion Group FIFO parity error",
+		-1, 1 },
+		{ PIOREQGRPPERR, "PCI PIO request Group FIFO parity error",
+		-1, 1 },
+		{ TARTAGPERR, "PCI PCI target tag FIFO parity error", -1, 1 },
+		{ MSTTAGQPERR, "PCI master tag queue parity error", -1, 1 },
+		{ CREQPERR, "PCI CMD channel request parity error", -1, 1 },
+		{ CRSPPERR, "PCI CMD channel response parity error", -1, 1 },
+		{ DREQWRPERR, "PCI DMA channel write request parity error",
+		-1, 1 },
+		{ DREQPERR, "PCI DMA channel request parity error", -1, 1 },
+		{ DRSPPERR, "PCI DMA channel response parity error", -1, 1 },
+		{ HREQWRPERR, "PCI HMA channel count parity error", -1, 1 },
+		{ HREQPERR, "PCI HMA channel request parity error", -1, 1 },
+		{ HRSPPERR, "PCI HMA channel response parity error", -1, 1 },
+		{ CFGSNPPERR, "PCI config snoop FIFO parity error", -1, 1 },
+		{ FIDPERR, "PCI FID parity error", -1, 1 },
+		{ VFIDPERR, "PCI INTx clear parity error", -1, 1 },
+		{ MAGRPPERR, "PCI MA group FIFO parity error", -1, 1 },
+		{ PIOTAGPERR, "PCI PIO tag parity error", -1, 1 },
+		{ IPRXHDRGRPPERR, "PCI IP Rx header group parity error",
+		-1, 1 },
+		{ IPRXDATAGRPPERR, "PCI IP Rx data group parity error",
+		-1, 1 },
+		{ RPLPERR, "PCI IP replay buffer parity error", -1, 1 },
+		{ IPSOTPERR, "PCI IP SOT buffer parity error", -1, 1 },
+		{ TRGT1GRPPERR, "PCI TRGT1 group FIFOs parity error", -1, 1 },
+		{ READRSPERR, "Outbound read error", -1, 0 },
+		{ 0, NULL, 0, 0 }
+	};
+
+	int fat;
+	fat = csio_handle_intr_status(hw,
+				      PCIE_CORE_UTL_SYSTEM_BUS_AGENT_STATUS,
+				      sysbus_intr_info) +
+	      csio_handle_intr_status(hw,
+				      PCIE_CORE_UTL_PCI_EXPRESS_PORT_STATUS,
+				      pcie_port_intr_info) +
+	      csio_handle_intr_status(hw, PCIE_INT_CAUSE, pcie_intr_info);
+	if (fat)
+		csio_hw_fatal_err(hw);
+}
+
+/*
+ * csio_t5_flash_cfg_addr - return the address of the flash configuration file
+ * @hw: the HW module
+ *
+ * Return the address within the flash where the Firmware Configuration
+ * File is stored.
+ */
+static unsigned int
+csio_t5_flash_cfg_addr(struct csio_hw *hw)
+{
+	return FLASH_CFG_START;
+}
+
+/*
+ *      csio_t5_mc_read - read from MC through backdoor accesses
+ *      @hw: the hw module
+ *      @idx: index to the register
+ *      @addr: address of first byte requested
+ *      @data: 64 bytes of data containing the requested address
+ *      @ecc: where to store the corresponding 64-bit ECC word
+ *
+ *      Read 64 bytes of data from MC starting at a 64-byte-aligned address
+ *      that covers the requested address @addr.  If @parity is not %NULL it
+ *      is assigned the 64-bit ECC word for the read data.
+ */
+static int
+csio_t5_mc_read(struct csio_hw *hw, int idx, uint32_t addr, __be32 *data,
+		uint64_t *ecc)
+{
+	int i;
+	uint32_t mc_bist_cmd_reg, mc_bist_cmd_addr_reg, mc_bist_cmd_len_reg;
+	uint32_t mc_bist_status_rdata_reg, mc_bist_data_pattern_reg;
+
+	mc_bist_cmd_reg = MC_REG(MC_P_BIST_CMD, idx);
+	mc_bist_cmd_addr_reg = MC_REG(MC_P_BIST_CMD_ADDR, idx);
+	mc_bist_cmd_len_reg = MC_REG(MC_P_BIST_CMD_LEN, idx);
+	mc_bist_status_rdata_reg = MC_REG(MC_P_BIST_STATUS_RDATA, idx);
+	mc_bist_data_pattern_reg = MC_REG(MC_P_BIST_DATA_PATTERN, idx);
+
+	if (csio_rd_reg32(hw, mc_bist_cmd_reg) & START_BIST)
+		return -EBUSY;
+	csio_wr_reg32(hw, addr & ~0x3fU, mc_bist_cmd_addr_reg);
+	csio_wr_reg32(hw, 64, mc_bist_cmd_len_reg);
+	csio_wr_reg32(hw, 0xc, mc_bist_data_pattern_reg);
+	csio_wr_reg32(hw, BIST_OPCODE(1) | START_BIST |  BIST_CMD_GAP(1),
+		      mc_bist_cmd_reg);
+	i = csio_hw_wait_op_done_val(hw, mc_bist_cmd_reg, START_BIST,
+				     0, 10, 1, NULL);
+	if (i)
+		return i;
+
+#define MC_DATA(i) MC_BIST_STATUS_REG(MC_BIST_STATUS_RDATA, i)
+
+	for (i = 15; i >= 0; i--)
+		*data++ = htonl(csio_rd_reg32(hw, MC_DATA(i)));
+	if (ecc)
+		*ecc = csio_rd_reg64(hw, MC_DATA(16));
+#undef MC_DATA
+	return 0;
+}
+
+/*
+ *      csio_t5_edc_read - read from EDC through backdoor accesses
+ *      @hw: the hw module
+ *      @idx: which EDC to access
+ *      @addr: address of first byte requested
+ *      @data: 64 bytes of data containing the requested address
+ *      @ecc: where to store the corresponding 64-bit ECC word
+ *
+ *      Read 64 bytes of data from EDC starting at a 64-byte-aligned address
+ *      that covers the requested address @addr.  If @parity is not %NULL it
+ *      is assigned the 64-bit ECC word for the read data.
+ */
+static int
+csio_t5_edc_read(struct csio_hw *hw, int idx, uint32_t addr, __be32 *data,
+		uint64_t *ecc)
+{
+	int i;
+	uint32_t edc_bist_cmd_reg, edc_bist_cmd_addr_reg, edc_bist_cmd_len_reg;
+	uint32_t edc_bist_cmd_data_pattern, edc_bist_status_rdata_reg;
+
+/*
+ * These macro are missing in t4_regs.h file.
+ */
+#define EDC_STRIDE_T5 (EDC_T51_BASE_ADDR - EDC_T50_BASE_ADDR)
+#define EDC_REG_T5(reg, idx) (reg + EDC_STRIDE_T5 * idx)
+
+	edc_bist_cmd_reg = EDC_REG_T5(EDC_H_BIST_CMD, idx);
+	edc_bist_cmd_addr_reg = EDC_REG_T5(EDC_H_BIST_CMD_ADDR, idx);
+	edc_bist_cmd_len_reg = EDC_REG_T5(EDC_H_BIST_CMD_LEN, idx);
+	edc_bist_cmd_data_pattern = EDC_REG_T5(EDC_H_BIST_DATA_PATTERN, idx);
+	edc_bist_status_rdata_reg = EDC_REG_T5(EDC_H_BIST_STATUS_RDATA, idx);
+#undef EDC_REG_T5
+#undef EDC_STRIDE_T5
+
+	if (csio_rd_reg32(hw, edc_bist_cmd_reg) & START_BIST)
+		return -EBUSY;
+	csio_wr_reg32(hw, addr & ~0x3fU, edc_bist_cmd_addr_reg);
+	csio_wr_reg32(hw, 64, edc_bist_cmd_len_reg);
+	csio_wr_reg32(hw, 0xc, edc_bist_cmd_data_pattern);
+	csio_wr_reg32(hw, BIST_OPCODE(1) | START_BIST |  BIST_CMD_GAP(1),
+		      edc_bist_cmd_reg);
+	i = csio_hw_wait_op_done_val(hw, edc_bist_cmd_reg, START_BIST,
+				     0, 10, 1, NULL);
+	if (i)
+		return i;
+
+#define EDC_DATA(i) (EDC_BIST_STATUS_REG(EDC_BIST_STATUS_RDATA, i) + idx)
+
+	for (i = 15; i >= 0; i--)
+		*data++ = htonl(csio_rd_reg32(hw, EDC_DATA(i)));
+	if (ecc)
+		*ecc = csio_rd_reg64(hw, EDC_DATA(16));
+#undef EDC_DATA
+	return 0;
+}
+
+/*
+ * csio_t5_memory_rw - read/write EDC 0, EDC 1 or MC via PCIE memory window
+ * @hw: the csio_hw
+ * @win: PCI-E memory Window to use
+ * @mtype: memory type: MEM_EDC0, MEM_EDC1, MEM_MC0 (or MEM_MC) or MEM_MC1
+ * @addr: address within indicated memory type
+ * @len: amount of memory to transfer
+ * @buf: host memory buffer
+ * @dir: direction of transfer 1 => read, 0 => write
+ *
+ * Reads/writes an [almost] arbitrary memory region in the firmware: the
+ * firmware memory address, length and host buffer must be aligned on
+ * 32-bit boudaries.  The memory is transferred as a raw byte sequence
+ * from/to the firmware's memory.  If this memory contains data
+ * structures which contain multi-byte integers, it's the callers
+ * responsibility to perform appropriate byte order conversions.
+ */
+static int
+csio_t5_memory_rw(struct csio_hw *hw, u32 win, int mtype, u32 addr,
+		u32 len, uint32_t *buf, int dir)
+{
+	u32 pos, start, offset, memoffset;
+	u32 edc_size, mc_size, win_pf, mem_reg, mem_aperture, mem_base;
+
+	/*
+	 * Argument sanity checks ...
+	 */
+	if ((addr & 0x3) || (len & 0x3))
+		return -EINVAL;
+
+	/* Offset into the region of memory which is being accessed
+	 * MEM_EDC0 = 0
+	 * MEM_EDC1 = 1
+	 * MEM_MC   = 2 -- T4
+	 * MEM_MC0  = 2 -- For T5
+	 * MEM_MC1  = 3 -- For T5
+	 */
+	edc_size  = EDRAM_SIZE_GET(csio_rd_reg32(hw, MA_EDRAM0_BAR));
+	if (mtype != MEM_MC1)
+		memoffset = (mtype * (edc_size * 1024 * 1024));
+	else {
+		mc_size = EXT_MEM_SIZE_GET(csio_rd_reg32(hw,
+							 MA_EXT_MEMORY_BAR));
+		memoffset = (MEM_MC0 * edc_size + mc_size) * 1024 * 1024;
+	}
+
+	/* Determine the PCIE_MEM_ACCESS_OFFSET */
+	addr = addr + memoffset;
+
+	/*
+	 * Each PCI-E Memory Window is programmed with a window size -- or
+	 * "aperture" -- which controls the granularity of its mapping onto
+	 * adapter memory.  We need to grab that aperture in order to know
+	 * how to use the specified window.  The window is also programmed
+	 * with the base address of the Memory Window in BAR0's address
+	 * space.  For T4 this is an absolute PCI-E Bus Address.  For T5
+	 * the address is relative to BAR0.
+	 */
+	mem_reg = csio_rd_reg32(hw,
+			PCIE_MEM_ACCESS_REG(PCIE_MEM_ACCESS_BASE_WIN, win));
+	mem_aperture = 1 << (WINDOW(mem_reg) + 10);
+	mem_base = GET_PCIEOFST(mem_reg) << 10;
+
+	start = addr & ~(mem_aperture-1);
+	offset = addr - start;
+	win_pf = V_PFNUM(hw->pfn);
+
+	csio_dbg(hw, "csio_t5_memory_rw: mem_reg: 0x%x, mem_aperture: 0x%x\n",
+		 mem_reg, mem_aperture);
+	csio_dbg(hw, "csio_t5_memory_rw: mem_base: 0x%x, mem_offset: 0x%x\n",
+		 mem_base, memoffset);
+	csio_dbg(hw, "csio_t5_memory_rw: start:0x%x, offset:0x%x, win_pf:%d\n",
+		 start, offset, win_pf);
+	csio_dbg(hw, "csio_t5_memory_rw: mtype: %d, addr: 0x%x, len: %d\n",
+		 mtype, addr, len);
+
+	for (pos = start; len > 0; pos += mem_aperture, offset = 0) {
+		/*
+		 * Move PCI-E Memory Window to our current transfer
+		 * position.  Read it back to ensure that changes propagate
+		 * before we attempt to use the new value.
+		 */
+		csio_wr_reg32(hw, pos | win_pf,
+			PCIE_MEM_ACCESS_REG(PCIE_MEM_ACCESS_OFFSET, win));
+		csio_rd_reg32(hw,
+			PCIE_MEM_ACCESS_REG(PCIE_MEM_ACCESS_OFFSET, win));
+
+		while (offset < mem_aperture && len > 0) {
+			if (dir)
+				*buf++ = csio_rd_reg32(hw, mem_base + offset);
+			else
+				csio_wr_reg32(hw, *buf++, mem_base + offset);
+
+			offset += sizeof(__be32);
+			len -= sizeof(__be32);
+		}
+	}
+	return 0;
+}
+
+/*
+ * csio_t5_dfs_create_ext_mem - setup debugfs for MC0 or MC1 to read the values
+ * @hw: the csio_hw
+ *
+ * This function creates files in the debugfs with external memory region
+ * MC0 & MC1.
+ */
+static void
+csio_t5_dfs_create_ext_mem(struct csio_hw *hw)
+{
+	u32 size;
+	int i = csio_rd_reg32(hw, MA_TARGET_MEM_ENABLE);
+	if (i & EXT_MEM_ENABLE) {
+		size = csio_rd_reg32(hw, MA_EXT_MEMORY_BAR);
+		csio_add_debugfs_mem(hw, "mc0", MEM_MC0,
+				     EXT_MEM_SIZE_GET(size));
+	}
+	if (i & EXT_MEM1_ENABLE) {
+		size = csio_rd_reg32(hw, MA_EXT_MEMORY1_BAR);
+		csio_add_debugfs_mem(hw, "mc1", MEM_MC1,
+				     EXT_MEM_SIZE_GET(size));
+	}
+}
+
+/* T5 adapter specific function */
+struct csio_hw_chip_ops t5_ops = {
+	.chip_set_mem_win		= csio_t5_set_mem_win,
+	.chip_pcie_intr_handler		= csio_t5_pcie_intr_handler,
+	.chip_flash_cfg_addr		= csio_t5_flash_cfg_addr,
+	.chip_mc_read			= csio_t5_mc_read,
+	.chip_edc_read			= csio_t5_edc_read,
+	.chip_memory_rw			= csio_t5_memory_rw,
+	.chip_dfs_create_ext_mem	= csio_t5_dfs_create_ext_mem,
+};
diff --git a/drivers/scsi/csiostor/csio_init.c b/drivers/scsi/csiostor/csio_init.c
index 0604b5ff3638..00346fe939d5 100644
--- a/drivers/scsi/csiostor/csio_init.c
+++ b/drivers/scsi/csiostor/csio_init.c
@@ -81,9 +81,11 @@ csio_mem_read(struct file *file, char __user *buf, size_t count, loff_t *ppos)
 		__be32 data[16];
 
 		if (mem == MEM_MC)
-			ret = csio_hw_mc_read(hw, pos, data, NULL);
+			ret = hw->chip_ops->chip_mc_read(hw, 0, pos,
+							 data, NULL);
 		else
-			ret = csio_hw_edc_read(hw, mem, pos, data, NULL);
+			ret = hw->chip_ops->chip_edc_read(hw, mem, pos,
+							  data, NULL);
 		if (ret)
 			return ret;
 
@@ -108,7 +110,7 @@ static const struct file_operations csio_mem_debugfs_fops = {
 	.llseek  = default_llseek,
 };
 
-static void csio_add_debugfs_mem(struct csio_hw *hw, const char *name,
+void csio_add_debugfs_mem(struct csio_hw *hw, const char *name,
 				 unsigned int idx, unsigned int size_mb)
 {
 	struct dentry *de;
@@ -131,9 +133,8 @@ static int csio_setup_debugfs(struct csio_hw *hw)
 		csio_add_debugfs_mem(hw, "edc0", MEM_EDC0, 5);
 	if (i & EDRAM1_ENABLE)
 		csio_add_debugfs_mem(hw, "edc1", MEM_EDC1, 5);
-	if (i & EXT_MEM_ENABLE)
-		csio_add_debugfs_mem(hw, "mc", MEM_MC,
-		      EXT_MEM_SIZE_GET(csio_rd_reg32(hw, MA_EXT_MEMORY_BAR)));
+
+	hw->chip_ops->chip_dfs_create_ext_mem(hw);
 	return 0;
 }
 
@@ -1169,7 +1170,7 @@ static struct pci_error_handlers csio_err_handler = {
 };
 
 static DEFINE_PCI_DEVICE_TABLE(csio_pci_tbl) = {
-	CSIO_DEVICE(CSIO_DEVID_T440DBG_FCOE, 0),	/* T440DBG FCOE */
+	CSIO_DEVICE(CSIO_DEVID_T440DBG_FCOE, 0),        /* T4 DEBUG FCOE */
 	CSIO_DEVICE(CSIO_DEVID_T420CR_FCOE, 0),		/* T420CR FCOE */
 	CSIO_DEVICE(CSIO_DEVID_T422CR_FCOE, 0),		/* T422CR FCOE */
 	CSIO_DEVICE(CSIO_DEVID_T440CR_FCOE, 0),		/* T440CR FCOE */
@@ -1184,8 +1185,34 @@ static DEFINE_PCI_DEVICE_TABLE(csio_pci_tbl) = {
 	CSIO_DEVICE(CSIO_DEVID_B404_FCOE, 0),		/* B404 FCOE */
 	CSIO_DEVICE(CSIO_DEVID_T480CR_FCOE, 0),		/* T480 CR FCOE */
 	CSIO_DEVICE(CSIO_DEVID_T440LPCR_FCOE, 0),	/* T440 LP-CR FCOE */
-	CSIO_DEVICE(CSIO_DEVID_PE10K, 0),		/* PE10K FCOE */
-	CSIO_DEVICE(CSIO_DEVID_PE10K_PF1, 0),	/* PE10K FCOE on PF1 */
+	CSIO_DEVICE(CSIO_DEVID_AMSTERDAM_T4_FCOE, 0),   /* AMSTERDAM T4 FCOE */
+	CSIO_DEVICE(CSIO_DEVID_HUAWEI_T480_FCOE, 0),    /* HUAWEI T480 FCOE */
+	CSIO_DEVICE(CSIO_DEVID_HUAWEI_T440_FCOE, 0),    /* HUAWEI T440 FCOE */
+	CSIO_DEVICE(CSIO_DEVID_HUAWEI_STG310_FCOE, 0),  /* HUAWEI STG FCOE */
+	CSIO_DEVICE(CSIO_DEVID_ACROMAG_XMC_XAUI, 0),    /* ACROMAG XAUI FCOE */
+	CSIO_DEVICE(CSIO_DEVID_QUANTA_MEZZ_SFP_FCOE, 0),/* QUANTA MEZZ FCOE */
+	CSIO_DEVICE(CSIO_DEVID_HUAWEI_10GT_FCOE, 0),    /* HUAWEI 10GT FCOE */
+	CSIO_DEVICE(CSIO_DEVID_HUAWEI_T440_TOE_FCOE, 0),/* HUAWEI T4 TOE FCOE */
+	CSIO_DEVICE(CSIO_DEVID_T580DBG_FCOE, 0),        /* T5 DEBUG FCOE */
+	CSIO_DEVICE(CSIO_DEVID_T520CR_FCOE, 0),         /* T520CR FCOE */
+	CSIO_DEVICE(CSIO_DEVID_T522CR_FCOE, 0),         /* T522CR FCOE */
+	CSIO_DEVICE(CSIO_DEVID_T540CR_FCOE, 0),         /* T540CR FCOE */
+	CSIO_DEVICE(CSIO_DEVID_T520BCH_FCOE, 0),        /* T520BCH FCOE */
+	CSIO_DEVICE(CSIO_DEVID_T540BCH_FCOE, 0),        /* T540BCH FCOE */
+	CSIO_DEVICE(CSIO_DEVID_T540CH_FCOE, 0),         /* T540CH FCOE */
+	CSIO_DEVICE(CSIO_DEVID_T520SO_FCOE, 0),         /* T520SO FCOE */
+	CSIO_DEVICE(CSIO_DEVID_T520CX_FCOE, 0),         /* T520CX FCOE */
+	CSIO_DEVICE(CSIO_DEVID_T520BT_FCOE, 0),         /* T520BT FCOE */
+	CSIO_DEVICE(CSIO_DEVID_T504BT_FCOE, 0),         /* T504BT FCOE */
+	CSIO_DEVICE(CSIO_DEVID_B520_FCOE, 0),           /* B520 FCOE */
+	CSIO_DEVICE(CSIO_DEVID_B504_FCOE, 0),           /* B504 FCOE */
+	CSIO_DEVICE(CSIO_DEVID_T580CR2_FCOE, 0),	/* T580 CR FCOE */
+	CSIO_DEVICE(CSIO_DEVID_T540LPCR_FCOE, 0),       /* T540 LP-CR FCOE */
+	CSIO_DEVICE(CSIO_DEVID_AMSTERDAM_T5_FCOE, 0),   /* AMSTERDAM T5 FCOE */
+	CSIO_DEVICE(CSIO_DEVID_T580LPCR_FCOE, 0),       /* T580 LP-CR FCOE */
+	CSIO_DEVICE(CSIO_DEVID_T520LLCR_FCOE, 0),       /* T520 LL-CR FCOE */
+	CSIO_DEVICE(CSIO_DEVID_T560CR_FCOE, 0),         /* T560 CR FCOE */
+	CSIO_DEVICE(CSIO_DEVID_T580CR_FCOE, 0),         /* T580 CR FCOE */
 	{ 0, 0, 0, 0, 0, 0, 0 }
 };
 
@@ -1259,4 +1286,5 @@ MODULE_DESCRIPTION(CSIO_DRV_DESC);
 MODULE_LICENSE(CSIO_DRV_LICENSE);
 MODULE_DEVICE_TABLE(pci, csio_pci_tbl);
 MODULE_VERSION(CSIO_DRV_VERSION);
-MODULE_FIRMWARE(CSIO_FW_FNAME);
+MODULE_FIRMWARE(FW_FNAME_T4);
+MODULE_FIRMWARE(FW_FNAME_T5);
diff --git a/drivers/scsi/csiostor/csio_init.h b/drivers/scsi/csiostor/csio_init.h
index 0838fd7ec9c7..5cc5d317a442 100644
--- a/drivers/scsi/csiostor/csio_init.h
+++ b/drivers/scsi/csiostor/csio_init.h
@@ -52,31 +52,6 @@
 #define CSIO_DRV_DESC			"Chelsio FCoE driver"
 #define CSIO_DRV_VERSION		"1.0.0"
 
-#define CSIO_DEVICE(devid, idx)					\
-{ PCI_VENDOR_ID_CHELSIO, (devid), PCI_ANY_ID, PCI_ANY_ID, 0, 0, (idx) }
-
-#define CSIO_IS_T4_FPGA(_dev)		(((_dev) == CSIO_DEVID_PE10K) ||\
-					 ((_dev) == CSIO_DEVID_PE10K_PF1))
-
-/* FCoE device IDs */
-#define CSIO_DEVID_PE10K		0xA000
-#define CSIO_DEVID_PE10K_PF1		0xA001
-#define CSIO_DEVID_T440DBG_FCOE		0x4600
-#define CSIO_DEVID_T420CR_FCOE		0x4601
-#define CSIO_DEVID_T422CR_FCOE		0x4602
-#define CSIO_DEVID_T440CR_FCOE		0x4603
-#define CSIO_DEVID_T420BCH_FCOE		0x4604
-#define CSIO_DEVID_T440BCH_FCOE		0x4605
-#define CSIO_DEVID_T440CH_FCOE		0x4606
-#define CSIO_DEVID_T420SO_FCOE		0x4607
-#define CSIO_DEVID_T420CX_FCOE		0x4608
-#define CSIO_DEVID_T420BT_FCOE		0x4609
-#define CSIO_DEVID_T404BT_FCOE		0x460A
-#define CSIO_DEVID_B420_FCOE		0x460B
-#define CSIO_DEVID_B404_FCOE		0x460C
-#define CSIO_DEVID_T480CR_FCOE		0x460D
-#define CSIO_DEVID_T440LPCR_FCOE	0x460E
-
 extern struct fc_function_template csio_fc_transport_funcs;
 extern struct fc_function_template csio_fc_transport_vport_funcs;
 
@@ -100,6 +75,10 @@ struct csio_lnode *csio_shost_init(struct csio_hw *, struct device *, bool,
 void csio_shost_exit(struct csio_lnode *);
 void csio_lnodes_exit(struct csio_hw *, bool);
 
+/* DebugFS helper routines */
+void csio_add_debugfs_mem(struct csio_hw *, const char *,
+		unsigned int, unsigned int);
+
 static inline struct Scsi_Host *
 csio_ln_to_shost(struct csio_lnode *ln)
 {
diff --git a/drivers/scsi/csiostor/csio_lnode.h b/drivers/scsi/csiostor/csio_lnode.h
index 8d84988ab06d..0f9c04175b11 100644
--- a/drivers/scsi/csiostor/csio_lnode.h
+++ b/drivers/scsi/csiostor/csio_lnode.h
@@ -114,7 +114,7 @@ struct csio_lnode_stats {
 	uint32_t	n_rnode_match;  /* matched rnode */
 	uint32_t	n_dev_loss_tmo; /* Device loss timeout */
 	uint32_t	n_fdmi_err;	/* fdmi err */
-	uint32_t	n_evt_fw[RSCN_DEV_LOST];	/* fw events */
+	uint32_t	n_evt_fw[PROTO_ERR_IMPL_LOGO];	/* fw events */
 	enum csio_ln_ev	n_evt_sm[CSIO_LNE_MAX_EVENT];	/* State m/c events */
 	uint32_t	n_rnode_alloc;	/* rnode allocated */
 	uint32_t	n_rnode_free;	/* rnode freed */
diff --git a/drivers/scsi/csiostor/csio_rnode.c b/drivers/scsi/csiostor/csio_rnode.c
index 51c6a388de2b..e9c3b045f587 100644
--- a/drivers/scsi/csiostor/csio_rnode.c
+++ b/drivers/scsi/csiostor/csio_rnode.c
@@ -302,7 +302,7 @@ csio_confirm_rnode(struct csio_lnode *ln, uint32_t rdev_flowid,
 {
 	uint8_t rport_type;
 	struct csio_rnode *rn, *match_rn;
-	uint32_t vnp_flowid;
+	uint32_t vnp_flowid = 0;
 	__be32 *port_id;
 
 	port_id = (__be32 *)&rdevp->r_id[0];
@@ -350,6 +350,14 @@ csio_confirm_rnode(struct csio_lnode *ln, uint32_t rdev_flowid,
 			 * Else, go ahead and alloc a new rnode.
 			 */
 			if (!memcmp(csio_rn_wwpn(match_rn), rdevp->wwpn, 8)) {
+				if (rn == match_rn)
+					goto found_rnode;
+				csio_ln_dbg(ln,
+					    "nport_id:x%x and wwpn:%llx"
+					    " match for ssni:x%x\n",
+					    rn->nport_id,
+					    wwn_to_u64(rdevp->wwpn),
+					    rdev_flowid);
 				if (csio_is_rnode_ready(rn)) {
 					csio_ln_warn(ln,
 						     "rnode is already"
diff --git a/drivers/scsi/csiostor/csio_rnode.h b/drivers/scsi/csiostor/csio_rnode.h
index a3b434c801da..65940096a80d 100644
--- a/drivers/scsi/csiostor/csio_rnode.h
+++ b/drivers/scsi/csiostor/csio_rnode.h
@@ -63,7 +63,7 @@ struct csio_rnode_stats {
 	uint32_t	n_err_nomem;	/* error nomem */
 	uint32_t	n_evt_unexp;	/* unexpected event */
 	uint32_t	n_evt_drop;	/* unexpected event */
-	uint32_t	n_evt_fw[RSCN_DEV_LOST];	/* fw events */
+	uint32_t	n_evt_fw[PROTO_ERR_IMPL_LOGO];	/* fw events */
 	enum csio_rn_ev	n_evt_sm[CSIO_RNFE_MAX_EVENT];	/* State m/c events */
 	uint32_t	n_lun_rst;	/* Number of resets of
 					 * of LUNs under this
diff --git a/drivers/scsi/csiostor/csio_wr.c b/drivers/scsi/csiostor/csio_wr.c
index c32df1bdaa97..4255ce264abf 100644
--- a/drivers/scsi/csiostor/csio_wr.c
+++ b/drivers/scsi/csiostor/csio_wr.c
@@ -85,8 +85,8 @@ csio_wr_ring_fldb(struct csio_hw *hw, struct csio_q *flq)
 	 */
 	if (flq->inc_idx >= 8) {
 		csio_wr_reg32(hw, DBPRIO(1) | QID(flq->un.fl.flid) |
-			      PIDX(flq->inc_idx / 8),
-			      MYPF_REG(SGE_PF_KDOORBELL));
+				  CSIO_HW_PIDX(hw, flq->inc_idx / 8),
+				  MYPF_REG(SGE_PF_KDOORBELL));
 		flq->inc_idx &= 7;
 	}
 }
@@ -989,7 +989,8 @@ csio_wr_issue(struct csio_hw *hw, int qidx, bool prio)
 	wmb();
 	/* Ring SGE Doorbell writing q->pidx into it */
 	csio_wr_reg32(hw, DBPRIO(prio) | QID(q->un.eq.physeqid) |
-		      PIDX(q->inc_idx), MYPF_REG(SGE_PF_KDOORBELL));
+			  CSIO_HW_PIDX(hw, q->inc_idx),
+			  MYPF_REG(SGE_PF_KDOORBELL));
 	q->inc_idx = 0;
 
 	return 0;
@@ -1331,20 +1332,30 @@ csio_wr_fixup_host_params(struct csio_hw *hw)
 
 	/* FL BUFFER SIZE#0 is Page size i,e already aligned to cache line */
 	csio_wr_reg32(hw, PAGE_SIZE, SGE_FL_BUFFER_SIZE0);
-	csio_wr_reg32(hw,
-		      (csio_rd_reg32(hw, SGE_FL_BUFFER_SIZE2) +
-		      sge->csio_fl_align - 1) & ~(sge->csio_fl_align - 1),
-		      SGE_FL_BUFFER_SIZE2);
-	csio_wr_reg32(hw,
-		      (csio_rd_reg32(hw, SGE_FL_BUFFER_SIZE3) +
-		      sge->csio_fl_align - 1) & ~(sge->csio_fl_align - 1),
-		      SGE_FL_BUFFER_SIZE3);
+
+	/*
+	 * If using hard params, the following will get set correctly
+	 * in csio_wr_set_sge().
+	 */
+	if (hw->flags & CSIO_HWF_USING_SOFT_PARAMS) {
+		csio_wr_reg32(hw,
+			(csio_rd_reg32(hw, SGE_FL_BUFFER_SIZE2) +
+			sge->csio_fl_align - 1) & ~(sge->csio_fl_align - 1),
+			SGE_FL_BUFFER_SIZE2);
+		csio_wr_reg32(hw,
+			(csio_rd_reg32(hw, SGE_FL_BUFFER_SIZE3) +
+			sge->csio_fl_align - 1) & ~(sge->csio_fl_align - 1),
+			SGE_FL_BUFFER_SIZE3);
+	}
 
 	csio_wr_reg32(hw, HPZ0(PAGE_SHIFT - 12), ULP_RX_TDDP_PSZ);
 
 	/* default value of rx_dma_offset of the NIC driver */
 	csio_set_reg_field(hw, SGE_CONTROL, PKTSHIFT_MASK,
 			   PKTSHIFT(CSIO_SGE_RX_DMA_OFFSET));
+
+	csio_hw_tp_wr_bits_indirect(hw, TP_INGRESS_CONFIG,
+				    CSUM_HAS_PSEUDO_HDR, 0);
 }
 
 static void
@@ -1460,18 +1471,21 @@ csio_wr_set_sge(struct csio_hw *hw)
 	 * and generate an interrupt when this occurs so we can recover.
 	 */
 	csio_set_reg_field(hw, SGE_DBFIFO_STATUS,
-			   HP_INT_THRESH(HP_INT_THRESH_MASK) |
-			   LP_INT_THRESH(LP_INT_THRESH_MASK),
-			   HP_INT_THRESH(CSIO_SGE_DBFIFO_INT_THRESH) |
-			   LP_INT_THRESH(CSIO_SGE_DBFIFO_INT_THRESH));
+		   HP_INT_THRESH(HP_INT_THRESH_MASK) |
+		   CSIO_HW_LP_INT_THRESH(hw, CSIO_HW_M_LP_INT_THRESH(hw)),
+		   HP_INT_THRESH(CSIO_SGE_DBFIFO_INT_THRESH) |
+		   CSIO_HW_LP_INT_THRESH(hw, CSIO_SGE_DBFIFO_INT_THRESH));
+
 	csio_set_reg_field(hw, SGE_DOORBELL_CONTROL, ENABLE_DROP,
 			   ENABLE_DROP);
 
 	/* SGE_FL_BUFFER_SIZE0 is set up by csio_wr_fixup_host_params(). */
 
 	CSIO_SET_FLBUF_SIZE(hw, 1, CSIO_SGE_FLBUF_SIZE1);
-	CSIO_SET_FLBUF_SIZE(hw, 2, CSIO_SGE_FLBUF_SIZE2);
-	CSIO_SET_FLBUF_SIZE(hw, 3, CSIO_SGE_FLBUF_SIZE3);
+	csio_wr_reg32(hw, (CSIO_SGE_FLBUF_SIZE2 + sge->csio_fl_align - 1)
+		      & ~(sge->csio_fl_align - 1), SGE_FL_BUFFER_SIZE2);
+	csio_wr_reg32(hw, (CSIO_SGE_FLBUF_SIZE3 + sge->csio_fl_align - 1)
+		      & ~(sge->csio_fl_align - 1), SGE_FL_BUFFER_SIZE3);
 	CSIO_SET_FLBUF_SIZE(hw, 4, CSIO_SGE_FLBUF_SIZE4);
 	CSIO_SET_FLBUF_SIZE(hw, 5, CSIO_SGE_FLBUF_SIZE5);
 	CSIO_SET_FLBUF_SIZE(hw, 6, CSIO_SGE_FLBUF_SIZE6);
@@ -1522,22 +1536,24 @@ void
 csio_wr_sge_init(struct csio_hw *hw)
 {
 	/*
-	 * If we are master:
+	 * If we are master and chip is not initialized:
 	 *    - If we plan to use the config file, we need to fixup some
 	 *      host specific registers, and read the rest of the SGE
 	 *      configuration.
 	 *    - If we dont plan to use the config file, we need to initialize
 	 *      SGE entirely, including fixing the host specific registers.
+	 * If we are master and chip is initialized, just read and work off of
+	 *	the already initialized SGE values.
 	 * If we arent the master, we are only allowed to read and work off of
 	 *      the already initialized SGE values.
 	 *
 	 * Therefore, before calling this function, we assume that the master-
-	 * ship of the card, and whether to use config file or not, have
-	 * already been decided. In other words, CSIO_HWF_USING_SOFT_PARAMS and
-	 * CSIO_HWF_MASTER should be set/unset.
+	 * ship of the card, state and whether to use config file or not, have
+	 * already been decided.
 	 */
 	if (csio_is_hw_master(hw)) {
-		csio_wr_fixup_host_params(hw);
+		if (hw->fw_state != CSIO_DEV_STATE_INIT)
+			csio_wr_fixup_host_params(hw);
 
 		if (hw->flags & CSIO_HWF_USING_SOFT_PARAMS)
 			csio_wr_get_sge(hw);
author	Linus Torvalds <torvalds@linux-foundation.org>	2013-05-01 23:08:52 +0200
committer	Linus Torvalds <torvalds@linux-foundation.org>	2013-05-01 23:08:52 +0200
commit	73287a43cc79ca06629a88d1a199cd283f42456a (patch)
tree	acf4456e260115bea77ee31a29f10ce17f0db45c /drivers/scsi/csiostor
parent	Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/dto... (diff)
parent	filter: fix va_list build error (diff)
download	linux-73287a43cc79ca06629a88d1a199cd283f42456a.tar.xz linux-73287a43cc79ca06629a88d1a199cd283f42456a.zip