* [V3 PATCH 1/9] csiostor: Chelsio FCoE offload driver submission (sources part 1).
From: Naresh Kumar Inna @ 2012-09-11 14:38 UTC (permalink / raw)
To: JBottomley, linux-scsi, dm, leedom; +Cc: netdev, naresh, chethan
In-Reply-To: <1347374347-3852-1-git-send-email-naresh@chelsio.com>
This patch contains the hardware interfacing functionality (chip/firmware
initialization/setup) and error handling. It also has slow path event
handling functionality, the Makefile and Kconfig changes.
Signed-off-by: Naresh Kumar Inna <naresh@chelsio.com>
---
V3: Added const qualifier to adapter descriptor table.
drivers/scsi/Kconfig | 1 +
drivers/scsi/Makefile | 1 +
drivers/scsi/csiostor/Kconfig | 20 +
drivers/scsi/csiostor/Makefile | 11 +
drivers/scsi/csiostor/csio_hw.c | 4396 +++++++++++++++++++++++++++++++++++++++
5 files changed, 4429 insertions(+), 0 deletions(-)
create mode 100644 drivers/scsi/csiostor/Kconfig
create mode 100644 drivers/scsi/csiostor/Makefile
create mode 100644 drivers/scsi/csiostor/csio_hw.c
diff --git a/drivers/scsi/Kconfig b/drivers/scsi/Kconfig
index 74bf1aa..af7a3e7 100644
--- a/drivers/scsi/Kconfig
+++ b/drivers/scsi/Kconfig
@@ -1812,6 +1812,7 @@ config SCSI_VIRTIO
This is the virtual HBA driver for virtio. If the kernel will
be used in a virtual machine, say Y or M.
+source "drivers/scsi/csiostor/Kconfig"
endif # SCSI_LOWLEVEL
diff --git a/drivers/scsi/Makefile b/drivers/scsi/Makefile
index 888f73a..8739aa7 100644
--- a/drivers/scsi/Makefile
+++ b/drivers/scsi/Makefile
@@ -90,6 +90,7 @@ obj-$(CONFIG_SCSI_QLA_FC) += qla2xxx/
obj-$(CONFIG_SCSI_QLA_ISCSI) += libiscsi.o qla4xxx/
obj-$(CONFIG_SCSI_LPFC) += lpfc/
obj-$(CONFIG_SCSI_BFA_FC) += bfa/
+obj-$(CONFIG_SCSI_CHELSIO_FCOE) += csiostor/
obj-$(CONFIG_SCSI_PAS16) += pas16.o
obj-$(CONFIG_SCSI_T128) += t128.o
obj-$(CONFIG_SCSI_DMX3191D) += dmx3191d.o
diff --git a/drivers/scsi/csiostor/Kconfig b/drivers/scsi/csiostor/Kconfig
new file mode 100644
index 0000000..c2acf02
--- /dev/null
+++ b/drivers/scsi/csiostor/Kconfig
@@ -0,0 +1,20 @@
+config SCSI_CHELSIO_FCOE
+ tristate "Chelsio Communications FCoE support"
+ depends on PCI && SCSI
+ select SCSI_FC_ATTRS
+ select FW_LOADER
+ help
+ This driver supports FCoE Offload functionality over
+ Chelsio T4-based 10Gb Converged Network Adapters.
+
+ For general information about Chelsio and our products, visit
+ our website at <http://www.chelsio.com>.
+
+ For customer support, please visit our customer support page at
+ <http://www.chelsio.com/support.html>.
+
+ Please send feedback to <linux-bugs@chelsio.com>.
+
+ To compile this driver as a module choose M here; the module
+ will be called csiostor.
+
diff --git a/drivers/scsi/csiostor/Makefile b/drivers/scsi/csiostor/Makefile
new file mode 100644
index 0000000..b581966
--- /dev/null
+++ b/drivers/scsi/csiostor/Makefile
@@ -0,0 +1,11 @@
+#
+## Chelsio FCoE driver
+#
+##
+
+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
diff --git a/drivers/scsi/csiostor/csio_hw.c b/drivers/scsi/csiostor/csio_hw.c
new file mode 100644
index 0000000..990dae9
--- /dev/null
+++ b/drivers/scsi/csiostor/csio_hw.c
@@ -0,0 +1,4396 @@
+/*
+ * This file is part of the Chelsio FCoE driver for Linux.
+ *
+ * Copyright (c) 2008-2012 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
+ * COPYING in the main directory of this source tree, or the
+ * 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 <linux/pci.h>
+#include <linux/pci_regs.h>
+#include <linux/firmware.h>
+#include <linux/stddef.h>
+#include <linux/delay.h>
+#include <linux/string.h>
+#include <linux/compiler.h>
+#include <linux/jiffies.h>
+#include <linux/kernel.h>
+#include <linux/log2.h>
+
+#include "csio_hw.h"
+#include "csio_lnode.h"
+#include "csio_rnode.h"
+
+int csio_force_master;
+int csio_dbg_level = 0xFEFF;
+unsigned int csio_port_mask = 0xf;
+
+/* Default FW event queue entries. */
+static uint32_t csio_evtq_sz = CSIO_EVTQ_SIZE;
+
+/* Default MSI param level */
+int csio_msi = 2;
+
+/* FCoE function instances */
+static int dev_num;
+
+/* FCoE Adapter types & its description */
+static const struct csio_adap_desc csio_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]"},
+ {"T440-CR 10G", "Chelsio T440-CR 10G [FCoE]"},
+ {"T420-BCH 10G", "Chelsio T420-BCH 10G [FCoE]"},
+ {"T440-BCH 10G", "Chelsio T440-BCH 10G [FCoE]"},
+ {"T440-CH 10G", "Chelsio T440-CH 10G [FCoE]"},
+ {"T420-SO 10G", "Chelsio T420-SO 10G [FCoE]"},
+ {"T420-CX4 10G", "Chelsio T420-CX4 10G [FCoE]"},
+ {"T420-BT 10G", "Chelsio T420-BT 10G [FCoE]"},
+ {"T404-BT 1G", "Chelsio T404-BT 1G [FCoE]"},
+ {"B420-SR 10G", "Chelsio B420-SR 10G [FCoE]"},
+ {"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]"}
+};
+
+static void csio_mgmtm_cleanup(struct csio_mgmtm *);
+static void csio_hw_mbm_cleanup(struct csio_hw *);
+
+/* State machine forward declarations */
+static void csio_hws_uninit(struct csio_hw *, enum csio_hw_ev);
+static void csio_hws_configuring(struct csio_hw *, enum csio_hw_ev);
+static void csio_hws_initializing(struct csio_hw *, enum csio_hw_ev);
+static void csio_hws_ready(struct csio_hw *, enum csio_hw_ev);
+static void csio_hws_quiescing(struct csio_hw *, enum csio_hw_ev);
+static void csio_hws_quiesced(struct csio_hw *, enum csio_hw_ev);
+static void csio_hws_resetting(struct csio_hw *, enum csio_hw_ev);
+static void csio_hws_removing(struct csio_hw *, enum csio_hw_ev);
+static void csio_hws_pcierr(struct csio_hw *, enum csio_hw_ev);
+
+static void csio_hw_initialize(struct csio_hw *hw);
+static void csio_evtq_stop(struct csio_hw *hw);
+static void csio_evtq_start(struct csio_hw *hw);
+
+int csio_is_hw_ready(struct csio_hw *hw)
+{
+ return csio_match_state(hw, csio_hws_ready);
+}
+
+int csio_is_hw_removing(struct csio_hw *hw)
+{
+ return csio_match_state(hw, csio_hws_removing);
+}
+
+
+/*
+ * csio_hw_wait_op_done_val - wait until an operation is completed
+ * @hw: the HW module
+ * @reg: the register to check for completion
+ * @mask: a single-bit field within @reg that indicates completion
+ * @polarity: the value of the field when the operation is completed
+ * @attempts: number of check iterations
+ * @delay: delay in usecs between iterations
+ * @valp: where to store the value of the register at completion time
+ *
+ * Wait until an operation is completed by checking a bit in a register
+ * up to @attempts times. If @valp is not NULL the value of the register
+ * at the time it indicated completion is stored there. Returns 0 if the
+ * operation completes and -EAGAIN otherwise.
+ */
+static 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)
+{
+ uint32_t val;
+ while (1) {
+ val = csio_rd_reg32(hw, reg);
+
+ if (!!(val & mask) == polarity) {
+ if (valp)
+ *valp = val;
+ return 0;
+ }
+
+ if (--attempts == 0)
+ return -EAGAIN;
+ if (delay)
+ udelay(delay);
+ }
+}
+
+void
+csio_set_reg_field(struct csio_hw *hw, uint32_t reg, uint32_t mask,
+ uint32_t value)
+{
+ uint32_t val = csio_rd_reg32(hw, reg) & ~mask;
+
+ csio_wr_reg32(hw, val | value, reg);
+ /* Flush */
+ csio_rd_reg32(hw, reg);
+
+}
+
+/*
+ * 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, uint32_t *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, uint32_t *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, __be32 *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;
+ __be32 *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, __be32 *buf)
+{
+ return csio_memory_rw(hw, 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 VPD_INFO_FLD_HDR_SIZE 3
+
+/*
+ * csio_hw_seeprom_read - read a serial EEPROM location
+ * @hw: hw to read
+ * @addr: EEPROM virtual address
+ * @data: where to store the read data
+ *
+ * Read a 32-bit word from a location in serial EEPROM using the card's PCI
+ * VPD capability. Note that this function must be called with a virtual
+ * address.
+ */
+static int
+csio_hw_seeprom_read(struct csio_hw *hw, uint32_t addr, uint32_t *data)
+{
+ uint16_t val = 0;
+ int attempts = EEPROM_MAX_RD_POLL;
+ uint32_t base = hw->params.pci.vpd_cap_addr;
+
+ if (addr >= EEPROMVSIZE || (addr & 3))
+ return -EINVAL;
+
+ pci_write_config_word(hw->pdev, base + PCI_VPD_ADDR, (uint16_t)addr);
+
+ do {
+ udelay(10);
+ pci_read_config_word(hw->pdev, base + PCI_VPD_ADDR, &val);
+ } while (!(val & PCI_VPD_ADDR_F) && --attempts);
+
+ if (!(val & PCI_VPD_ADDR_F)) {
+ csio_err(hw, "reading EEPROM address 0x%x failed\n", addr);
+ return -EINVAL;
+ }
+
+ pci_read_config_dword(hw->pdev, base + PCI_VPD_DATA, data);
+ *data = le32_to_cpu(*data);
+ return 0;
+}
+
+/*
+ * Partial EEPROM Vital Product Data structure. Includes only the ID and
+ * VPD-R sections.
+ */
+struct t4_vpd_hdr {
+ u8 id_tag;
+ u8 id_len[2];
+ u8 id_data[ID_LEN];
+ u8 vpdr_tag;
+ u8 vpdr_len[2];
+};
+
+/*
+ * csio_hw_get_vpd_keyword_val - Locates an information field keyword in
+ * the VPD
+ * @v: Pointer to buffered vpd data structure
+ * @kw: The keyword to search for
+ *
+ * Returns the value of the information field keyword or
+ * -EINVAL otherwise.
+ */
+static int
+csio_hw_get_vpd_keyword_val(const struct t4_vpd_hdr *v, const char *kw)
+{
+ int32_t i;
+ int32_t offset , len;
+ const uint8_t *buf = &v->id_tag;
+ const uint8_t *vpdr_len = &v->vpdr_tag;
+ offset = sizeof(struct t4_vpd_hdr);
+ len = (uint16_t)vpdr_len[1] + ((uint16_t)vpdr_len[2] << 8);
+
+ if (len + sizeof(struct t4_vpd_hdr) > VPD_LEN)
+ return -EINVAL;
+
+ for (i = offset; (i + VPD_INFO_FLD_HDR_SIZE) <= (offset + len);) {
+ if (memcmp(buf + i , kw, 2) == 0) {
+ i += VPD_INFO_FLD_HDR_SIZE;
+ return i;
+ }
+
+ i += VPD_INFO_FLD_HDR_SIZE + buf[i+2];
+ }
+
+ return -EINVAL;
+}
+
+static int
+csio_pci_capability(struct pci_dev *pdev, int cap, int *pos)
+{
+ *pos = pci_find_capability(pdev, cap);
+ if (*pos)
+ return 0;
+
+ return -1;
+}
+
+/*
+ * csio_hw_get_vpd_params - read VPD parameters from VPD EEPROM
+ * @hw: HW module
+ * @p: where to store the parameters
+ *
+ * Reads card parameters stored in VPD EEPROM.
+ */
+static int
+csio_hw_get_vpd_params(struct csio_hw *hw, struct csio_vpd *p)
+{
+ int i, ret, ec, sn, addr;
+ uint8_t *vpd, csum;
+ const struct t4_vpd_hdr *v;
+ /* To get around compilation warning from strstrip */
+ char *s;
+
+ if (csio_is_valid_vpd(hw))
+ return 0;
+
+ ret = csio_pci_capability(hw->pdev, PCI_CAP_ID_VPD,
+ &hw->params.pci.vpd_cap_addr);
+ if (ret)
+ return -EINVAL;
+
+ vpd = kzalloc(VPD_LEN, GFP_ATOMIC);
+ if (vpd == NULL)
+ return -ENOMEM;
+
+ /*
+ * Card information normally starts at VPD_BASE but early cards had
+ * it at 0.
+ */
+ ret = csio_hw_seeprom_read(hw, VPD_BASE, (uint32_t *)(vpd));
+ addr = *vpd == 0x82 ? VPD_BASE : VPD_BASE_OLD;
+
+ for (i = 0; i < VPD_LEN; i += 4) {
+ ret = csio_hw_seeprom_read(hw, addr + i, (uint32_t *)(vpd + i));
+ if (ret) {
+ kfree(vpd);
+ return ret;
+ }
+ }
+
+ /* Reset the VPD flag! */
+ hw->flags &= (~CSIO_HWF_VPD_VALID);
+
+ v = (const struct t4_vpd_hdr *)vpd;
+
+#define FIND_VPD_KW(var, name) do { \
+ var = csio_hw_get_vpd_keyword_val(v, name); \
+ if (var < 0) { \
+ csio_err(hw, "missing VPD keyword " name "\n"); \
+ kfree(vpd); \
+ return -EINVAL; \
+ } \
+} while (0)
+
+ FIND_VPD_KW(i, "RV");
+ for (csum = 0; i >= 0; i--)
+ csum += vpd[i];
+
+ if (csum) {
+ csio_err(hw, "corrupted VPD EEPROM, actual csum %u\n", csum);
+ kfree(vpd);
+ return -EINVAL;
+ }
+ FIND_VPD_KW(ec, "EC");
+ FIND_VPD_KW(sn, "SN");
+#undef FIND_VPD_KW
+
+ memcpy(p->id, v->id_data, ID_LEN);
+ s = strstrip(p->id);
+ memcpy(p->ec, vpd + ec, EC_LEN);
+ s = strstrip(p->ec);
+ i = vpd[sn - VPD_INFO_FLD_HDR_SIZE + 2];
+ memcpy(p->sn, vpd + sn, min(i, SERNUM_LEN));
+ s = strstrip(p->sn);
+
+ csio_valid_vpd_copied(hw);
+
+ kfree(vpd);
+ return 0;
+}
+
+/*
+ * csio_hw_sf1_read - read data from the serial flash
+ * @hw: the HW module
+ * @byte_cnt: number of bytes to read
+ * @cont: whether another operation will be chained
+ * @lock: whether to lock SF for PL access only
+ * @valp: where to store the read data
+ *
+ * Reads up to 4 bytes of data from the serial flash. The location of
+ * the read needs to be specified prior to calling this by issuing the
+ * appropriate commands to the serial flash.
+ */
+static int
+csio_hw_sf1_read(struct csio_hw *hw, uint32_t byte_cnt, int32_t cont,
+ int32_t lock, uint32_t *valp)
+{
+ int ret;
+
+ if (!byte_cnt || byte_cnt > 4)
+ return -EINVAL;
+ if (csio_rd_reg32(hw, SF_OP) & SF_BUSY)
+ return -EBUSY;
+
+ cont = cont ? SF_CONT : 0;
+ lock = lock ? SF_LOCK : 0;
+
+ csio_wr_reg32(hw, lock | cont | BYTECNT(byte_cnt - 1), SF_OP);
+ ret = csio_hw_wait_op_done_val(hw, SF_OP, SF_BUSY, 0, SF_ATTEMPTS,
+ 10, NULL);
+ if (!ret)
+ *valp = csio_rd_reg32(hw, SF_DATA);
+ return ret;
+}
+
+/*
+ * csio_hw_sf1_write - write data to the serial flash
+ * @hw: the HW module
+ * @byte_cnt: number of bytes to write
+ * @cont: whether another operation will be chained
+ * @lock: whether to lock SF for PL access only
+ * @val: value to write
+ *
+ * Writes up to 4 bytes of data to the serial flash. The location of
+ * the write needs to be specified prior to calling this by issuing the
+ * appropriate commands to the serial flash.
+ */
+static int
+csio_hw_sf1_write(struct csio_hw *hw, uint32_t byte_cnt, uint32_t cont,
+ int32_t lock, uint32_t val)
+{
+ if (!byte_cnt || byte_cnt > 4)
+ return -EINVAL;
+ if (csio_rd_reg32(hw, SF_OP) & SF_BUSY)
+ return -EBUSY;
+
+ cont = cont ? SF_CONT : 0;
+ lock = lock ? SF_LOCK : 0;
+
+ csio_wr_reg32(hw, val, SF_DATA);
+ csio_wr_reg32(hw, cont | BYTECNT(byte_cnt - 1) | OP_WR | lock, SF_OP);
+
+ return csio_hw_wait_op_done_val(hw, SF_OP, SF_BUSY, 0, SF_ATTEMPTS,
+ 10, NULL);
+}
+
+/*
+ * csio_hw_flash_wait_op - wait for a flash operation to complete
+ * @hw: the HW module
+ * @attempts: max number of polls of the status register
+ * @delay: delay between polls in ms
+ *
+ * Wait for a flash operation to complete by polling the status register.
+ */
+static int
+csio_hw_flash_wait_op(struct csio_hw *hw, int32_t attempts, int32_t delay)
+{
+ int ret;
+ uint32_t status;
+
+ while (1) {
+ ret = csio_hw_sf1_write(hw, 1, 1, 1, SF_RD_STATUS);
+ if (ret != 0)
+ return ret;
+
+ ret = csio_hw_sf1_read(hw, 1, 0, 1, &status);
+ if (ret != 0)
+ return ret;
+
+ if (!(status & 1))
+ return 0;
+ if (--attempts == 0)
+ return -EAGAIN;
+ if (delay)
+ msleep(delay);
+ }
+}
+
+/*
+ * csio_hw_read_flash - read words from serial flash
+ * @hw: the HW module
+ * @addr: the start address for the read
+ * @nwords: how many 32-bit words to read
+ * @data: where to store the read data
+ * @byte_oriented: whether to store data as bytes or as words
+ *
+ * Read the specified number of 32-bit words from the serial flash.
+ * If @byte_oriented is set the read data is stored as a byte array
+ * (i.e., big-endian), otherwise as 32-bit words in the platform's
+ * natural endianess.
+ */
+static int
+csio_hw_read_flash(struct csio_hw *hw, uint32_t addr, uint32_t nwords,
+ uint32_t *data, int32_t byte_oriented)
+{
+ int ret;
+
+ if (addr + nwords * sizeof(uint32_t) > hw->params.sf_size || (addr & 3))
+ return -EINVAL;
+
+ addr = swab32(addr) | SF_RD_DATA_FAST;
+
+ ret = csio_hw_sf1_write(hw, 4, 1, 0, addr);
+ if (ret != 0)
+ return ret;
+
+ ret = csio_hw_sf1_read(hw, 1, 1, 0, data);
+ if (ret != 0)
+ return ret;
+
+ for ( ; nwords; nwords--, data++) {
+ ret = csio_hw_sf1_read(hw, 4, nwords > 1, nwords == 1, data);
+ if (nwords == 1)
+ csio_wr_reg32(hw, 0, SF_OP); /* unlock SF */
+ if (ret)
+ return ret;
+ if (byte_oriented)
+ *data = htonl(*data);
+ }
+ return 0;
+}
+
+/*
+ * csio_hw_write_flash - write up to a page of data to the serial flash
+ * @hw: the hw
+ * @addr: the start address to write
+ * @n: length of data to write in bytes
+ * @data: the data to write
+ *
+ * Writes up to a page of data (256 bytes) to the serial flash starting
+ * at the given address. All the data must be written to the same page.
+ */
+static int
+csio_hw_write_flash(struct csio_hw *hw, uint32_t addr,
+ uint32_t n, const uint8_t *data)
+{
+ int ret = -EINVAL;
+ uint32_t buf[64];
+ uint32_t i, c, left, val, offset = addr & 0xff;
+
+ if (addr >= hw->params.sf_size || offset + n > SF_PAGE_SIZE)
+ return -EINVAL;
+
+ val = swab32(addr) | SF_PROG_PAGE;
+
+ ret = csio_hw_sf1_write(hw, 1, 0, 1, SF_WR_ENABLE);
+ if (ret != 0)
+ goto unlock;
+
+ ret = csio_hw_sf1_write(hw, 4, 1, 1, val);
+ if (ret != 0)
+ goto unlock;
+
+ for (left = n; left; left -= c) {
+ c = min(left, 4U);
+ for (val = 0, i = 0; i < c; ++i)
+ val = (val << 8) + *data++;
+
+ ret = csio_hw_sf1_write(hw, c, c != left, 1, val);
+ if (ret)
+ goto unlock;
+ }
+ ret = csio_hw_flash_wait_op(hw, 8, 1);
+ if (ret)
+ goto unlock;
+
+ csio_wr_reg32(hw, 0, SF_OP); /* unlock SF */
+
+ /* Read the page to verify the write succeeded */
+ ret = csio_hw_read_flash(hw, addr & ~0xff, ARRAY_SIZE(buf), buf, 1);
+ if (ret)
+ return ret;
+
+ if (memcmp(data - n, (uint8_t *)buf + offset, n)) {
+ csio_err(hw,
+ "failed to correctly write the flash page at %#x\n",
+ addr);
+ return -EINVAL;
+ }
+
+ return 0;
+
+unlock:
+ csio_wr_reg32(hw, 0, SF_OP); /* unlock SF */
+ return ret;
+}
+
+/*
+ * csio_hw_flash_erase_sectors - erase a range of flash sectors
+ * @hw: the HW module
+ * @start: the first sector to erase
+ * @end: the last sector to erase
+ *
+ * Erases the sectors in the given inclusive range.
+ */
+static int
+csio_hw_flash_erase_sectors(struct csio_hw *hw, int32_t start, int32_t end)
+{
+ int ret = 0;
+
+ while (start <= end) {
+
+ ret = csio_hw_sf1_write(hw, 1, 0, 1, SF_WR_ENABLE);
+ if (ret != 0)
+ goto out;
+
+ ret = csio_hw_sf1_write(hw, 4, 0, 1,
+ SF_ERASE_SECTOR | (start << 8));
+ if (ret != 0)
+ goto out;
+
+ ret = csio_hw_flash_wait_op(hw, 14, 500);
+ if (ret != 0)
+ goto out;
+
+ start++;
+ }
+out:
+ if (ret)
+ csio_err(hw, "erase of flash sector %d failed, error %d\n",
+ start, ret);
+ csio_wr_reg32(hw, 0, SF_OP); /* unlock SF */
+ 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)
+{
+ csio_info(hw, "%s: %u.%u.%u.%u\n", str,
+ FW_HDR_FW_VER_MAJOR_GET(hw->fwrev),
+ FW_HDR_FW_VER_MINOR_GET(hw->fwrev),
+ FW_HDR_FW_VER_MICRO_GET(hw->fwrev),
+ FW_HDR_FW_VER_BUILD_GET(hw->fwrev));
+}
+
+/*
+ * csio_hw_get_fw_version - read the firmware version
+ * @hw: HW module
+ * @vers: where to place the version
+ *
+ * Reads the FW version from flash.
+ */
+static int
+csio_hw_get_fw_version(struct csio_hw *hw, uint32_t *vers)
+{
+ return csio_hw_read_flash(hw, FW_IMG_START +
+ offsetof(struct fw_hdr, fw_ver), 1,
+ vers, 0);
+}
+
+/*
+ * csio_hw_get_tp_version - read the TP microcode version
+ * @hw: HW module
+ * @vers: where to place the version
+ *
+ * Reads the TP microcode version from flash.
+ */
+static int
+csio_hw_get_tp_version(struct csio_hw *hw, u32 *vers)
+{
+ return csio_hw_read_flash(hw, FLASH_FW_START +
+ offsetof(struct fw_hdr, tp_microcode_ver), 1,
+ vers, 0);
+}
+
+/*
+ * csio_hw_check_fw_version - check if the FW is compatible with
+ * this driver
+ * @hw: HW module
+ *
+ * Checks if an adapter's FW is compatible with the driver. Returns 0
+ * if there's exact match, a negative error if the version could not be
+ * read or there's a major/minor version mismatch/minor.
+ */
+static int
+csio_hw_check_fw_version(struct csio_hw *hw)
+{
+ int ret, major, minor, micro;
+
+ ret = csio_hw_get_fw_version(hw, &hw->fwrev);
+ if (!ret)
+ ret = csio_hw_get_tp_version(hw, &hw->tp_vers);
+ if (ret)
+ return ret;
+
+ major = FW_HDR_FW_VER_MAJOR_GET(hw->fwrev);
+ 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 */
+ csio_err(hw, "card FW has major version %u, driver wants %u\n",
+ major, FW_VERSION_MAJOR);
+ return -EINVAL;
+ }
+
+ if (minor == FW_VERSION_MINOR && micro == FW_VERSION_MICRO)
+ return 0; /* perfect match */
+
+ /* Minor/micro version mismatch */
+ return -EINVAL;
+}
+
+/*
+ * csio_hw_fw_dload - download firmware.
+ * @hw: HW module
+ * @fw_data: firmware image to write.
+ * @size: image size
+ *
+ * Write the supplied firmware image to the card's serial flash.
+ */
+static int
+csio_hw_fw_dload(struct csio_hw *hw, uint8_t *fw_data, uint32_t size)
+{
+ uint32_t csum;
+ int32_t addr;
+ int ret;
+ uint32_t i;
+ uint8_t first_page[SF_PAGE_SIZE];
+ const uint32_t *p = (const uint32_t *)fw_data;
+ struct fw_hdr *hdr = (struct fw_hdr *)fw_data;
+ uint32_t sf_sec_size;
+
+ if ((!hw->params.sf_size) || (!hw->params.sf_nsec)) {
+ csio_err(hw, "Serial Flash data invalid\n");
+ return -EINVAL;
+ }
+
+ if (!size) {
+ csio_err(hw, "FW image has no data\n");
+ return -EINVAL;
+ }
+
+ if (size & 511) {
+ csio_err(hw, "FW image size not multiple of 512 bytes\n");
+ return -EINVAL;
+ }
+
+ if (ntohs(hdr->len512) * 512 != size) {
+ csio_err(hw, "FW image size differs from size in FW header\n");
+ return -EINVAL;
+ }
+
+ if (size > FW_MAX_SIZE) {
+ csio_err(hw, "FW image too large, max is %u bytes\n",
+ FW_MAX_SIZE);
+ return -EINVAL;
+ }
+
+ for (csum = 0, i = 0; i < size / sizeof(csum); i++)
+ csum += ntohl(p[i]);
+
+ if (csum != 0xffffffff) {
+ csio_err(hw, "corrupted firmware image, checksum %#x\n", csum);
+ return -EINVAL;
+ }
+
+ sf_sec_size = hw->params.sf_size / hw->params.sf_nsec;
+ i = DIV_ROUND_UP(size, sf_sec_size); /* # of sectors spanned */
+
+ csio_dbg(hw, "Erasing sectors... start:%d end:%d\n",
+ FW_START_SEC, FW_START_SEC + i - 1);
+
+ ret = csio_hw_flash_erase_sectors(hw, FW_START_SEC,
+ FW_START_SEC + i - 1);
+ if (ret) {
+ csio_err(hw, "Flash Erase failed\n");
+ goto out;
+ }
+
+ /*
+ * We write the correct version at the end so the driver can see a bad
+ * version if the FW write fails. Start by writing a copy of the
+ * first page with a bad version.
+ */
+ memcpy(first_page, fw_data, SF_PAGE_SIZE);
+ ((struct fw_hdr *)first_page)->fw_ver = htonl(0xffffffff);
+ ret = csio_hw_write_flash(hw, FW_IMG_START, SF_PAGE_SIZE, first_page);
+ if (ret)
+ goto out;
+
+ csio_dbg(hw, "Writing Flash .. start:%d end:%d\n",
+ FW_IMG_START, FW_IMG_START + size);
+
+ addr = FW_IMG_START;
+ for (size -= SF_PAGE_SIZE; size; size -= SF_PAGE_SIZE) {
+ addr += SF_PAGE_SIZE;
+ fw_data += SF_PAGE_SIZE;
+ ret = csio_hw_write_flash(hw, addr, SF_PAGE_SIZE, fw_data);
+ if (ret)
+ goto out;
+ }
+
+ ret = csio_hw_write_flash(hw,
+ FW_IMG_START +
+ offsetof(struct fw_hdr, fw_ver),
+ sizeof(hdr->fw_ver),
+ (const uint8_t *)&hdr->fw_ver);
+
+out:
+ if (ret)
+ csio_err(hw, "firmware download failed, error %d\n", ret);
+ return ret;
+}
+
+static int
+csio_hw_get_flash_params(struct csio_hw *hw)
+{
+ int ret;
+ uint32_t info = 0;
+
+ ret = csio_hw_sf1_write(hw, 1, 1, 0, SF_RD_ID);
+ if (!ret)
+ ret = csio_hw_sf1_read(hw, 3, 0, 1, &info);
+ csio_wr_reg32(hw, 0, SF_OP); /* unlock SF */
+ if (ret != 0)
+ return ret;
+
+ if ((info & 0xff) != 0x20) /* not a Numonix flash */
+ return -EINVAL;
+ info >>= 16; /* log2 of size */
+ if (info >= 0x14 && info < 0x18)
+ hw->params.sf_nsec = 1 << (info - 16);
+ else if (info == 0x18)
+ hw->params.sf_nsec = 64;
+ else
+ return -EINVAL;
+ hw->params.sf_size = 1 << info;
+
+ return 0;
+}
+
+static void
+csio_set_pcie_completion_timeout(struct csio_hw *hw, u8 range)
+{
+ uint16_t val;
+ uint32_t pcie_cap;
+
+ if (!csio_pci_capability(hw->pdev, PCI_CAP_ID_EXP, &pcie_cap)) {
+ pci_read_config_word(hw->pdev,
+ pcie_cap + PCI_EXP_DEVCTL2, &val);
+ val &= 0xfff0;
+ val |= range ;
+ pci_write_config_word(hw->pdev,
+ pcie_cap + PCI_EXP_DEVCTL2, val);
+ }
+}
+
+
+/*
+ * 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 */
+/*****************************************************************************/
+
+static int
+csio_hw_dev_ready(struct csio_hw *hw)
+{
+ uint32_t reg;
+ int cnt = 6;
+
+ while (((reg = csio_rd_reg32(hw, PL_WHOAMI)) == 0xFFFFFFFF) &&
+ (--cnt != 0))
+ mdelay(100);
+
+ if ((cnt == 0) && (((int32_t)(SOURCEPF_GET(reg)) < 0) ||
+ (SOURCEPF_GET(reg) >= CSIO_MAX_PFN))) {
+ csio_err(hw, "PL_WHOAMI returned 0x%x, cnt:%d\n", reg, cnt);
+ return -EIO;
+ }
+
+ hw->pfn = SOURCEPF_GET(reg);
+
+ return 0;
+}
+
+/*
+ * csio_do_hello - Perform the HELLO FW Mailbox command and process response.
+ * @hw: HW module
+ * @state: Device state
+ *
+ * FW_HELLO_CMD has to be polled for completion.
+ */
+static int
+csio_do_hello(struct csio_hw *hw, enum csio_dev_state *state)
+{
+ struct csio_mb *mbp;
+ int rv = 0;
+ enum csio_dev_master master;
+ enum fw_retval retval;
+ uint8_t mpfn;
+ char state_str[16];
+ int retries = FW_CMD_HELLO_RETRIES;
+
+ memset(state_str, 0, sizeof(state_str));
+
+ mbp = mempool_alloc(hw->mb_mempool, GFP_ATOMIC);
+ if (!mbp) {
+ rv = -ENOMEM;
+ CSIO_INC_STATS(hw, n_err_nomem);
+ goto out;
+ }
+
+ master = csio_force_master ? CSIO_MASTER_MUST : CSIO_MASTER_MAY;
+
+retry:
+ csio_mb_hello(hw, mbp, CSIO_MB_DEFAULT_TMO, hw->pfn,
+ hw->pfn, master, NULL);
+
+ rv = csio_mb_issue(hw, mbp);
+ if (rv) {
+ csio_err(hw, "failed to issue HELLO cmd. ret:%d.\n", rv);
+ goto out_free_mb;
+ }
+
+ csio_mb_process_hello_rsp(hw, mbp, &retval, state, &mpfn);
+ if (retval != FW_SUCCESS) {
+ csio_err(hw, "HELLO cmd failed with ret: %d\n", retval);
+ rv = -EINVAL;
+ goto out_free_mb;
+ }
+
+ /* Firmware has designated us to be master */
+ if (hw->pfn == mpfn) {
+ hw->flags |= CSIO_HWF_MASTER;
+ } else if (*state == CSIO_DEV_STATE_UNINIT) {
+ /*
+ * If we're not the Master PF then we need to wait around for
+ * the Master PF Driver to finish setting up the adapter.
+ *
+ * Note that we also do this wait if we're a non-Master-capable
+ * PF and there is no current Master PF; a Master PF may show up
+ * momentarily and we wouldn't want to fail pointlessly. (This
+ * can happen when an OS loads lots of different drivers rapidly
+ * at the same time). In this case, the Master PF returned by
+ * the firmware will be PCIE_FW_MASTER_MASK so the test below
+ * will work ...
+ */
+
+ int waiting = FW_CMD_HELLO_TIMEOUT;
+
+ /*
+ * Wait for the firmware to either indicate an error or
+ * initialized state. If we see either of these we bail out
+ * and report the issue to the caller. If we exhaust the
+ * "hello timeout" and we haven't exhausted our retries, try
+ * again. Otherwise bail with a timeout error.
+ */
+ for (;;) {
+ uint32_t pcie_fw;
+
+ msleep(50);
+ waiting -= 50;
+
+ /*
+ * If neither Error nor Initialialized are indicated
+ * by the firmware keep waiting till we exaust our
+ * timeout ... and then retry if we haven't exhausted
+ * our retries ...
+ */
+ pcie_fw = csio_rd_reg32(hw, PCIE_FW);
+ if (!(pcie_fw & (PCIE_FW_ERR|PCIE_FW_INIT))) {
+ if (waiting <= 0) {
+ if (retries-- > 0)
+ goto retry;
+
+ rv = -ETIMEDOUT;
+ break;
+ }
+ continue;
+ }
+
+ /*
+ * We either have an Error or Initialized condition
+ * report errors preferentially.
+ */
+ if (state) {
+ if (pcie_fw & PCIE_FW_ERR) {
+ *state = CSIO_DEV_STATE_ERR;
+ rv = -ETIMEDOUT;
+ } else if (pcie_fw & PCIE_FW_INIT)
+ *state = CSIO_DEV_STATE_INIT;
+ }
+
+ /*
+ * If we arrived before a Master PF was selected and
+ * there's not a valid Master PF, grab its identity
+ * for our caller.
+ */
+ if (mpfn == PCIE_FW_MASTER_MASK &&
+ (pcie_fw & PCIE_FW_MASTER_VLD))
+ mpfn = PCIE_FW_MASTER_GET(pcie_fw);
+ break;
+ }
+ hw->flags &= ~CSIO_HWF_MASTER;
+ }
+
+ switch (*state) {
+ case CSIO_DEV_STATE_UNINIT:
+ strcpy(state_str, "Initializing");
+ break;
+ case CSIO_DEV_STATE_INIT:
+ strcpy(state_str, "Initialized");
+ break;
+ case CSIO_DEV_STATE_ERR:
+ strcpy(state_str, "Error");
+ break;
+ default:
+ strcpy(state_str, "Unknown");
+ break;
+ }
+
+ if (hw->pfn == mpfn)
+ csio_info(hw, "PF: %d, Coming up as MASTER, HW state: %s\n",
+ hw->pfn, state_str);
+ else
+ csio_info(hw,
+ "PF: %d, Coming up as SLAVE, Master PF: %d, HW state: %s\n",
+ hw->pfn, mpfn, state_str);
+
+out_free_mb:
+ mempool_free(mbp, hw->mb_mempool);
+out:
+ return rv;
+}
+
+/*
+ * csio_do_bye - Perform the BYE FW Mailbox command and process response.
+ * @hw: HW module
+ *
+ */
+static int
+csio_do_bye(struct csio_hw *hw)
+{
+ struct csio_mb *mbp;
+ enum fw_retval retval;
+
+ mbp = mempool_alloc(hw->mb_mempool, GFP_ATOMIC);
+ if (!mbp) {
+ CSIO_INC_STATS(hw, n_err_nomem);
+ return -ENOMEM;
+ }
+
+ csio_mb_bye(hw, mbp, CSIO_MB_DEFAULT_TMO, NULL);
+
+ if (csio_mb_issue(hw, mbp)) {
+ csio_err(hw, "Issue of BYE command failed\n");
+ mempool_free(mbp, hw->mb_mempool);
+ return -EINVAL;
+ }
+
+ retval = csio_mb_fw_retval(mbp);
+ if (retval != FW_SUCCESS) {
+ mempool_free(mbp, hw->mb_mempool);
+ return -EINVAL;
+ }
+
+ mempool_free(mbp, hw->mb_mempool);
+
+ return 0;
+}
+
+/*
+ * csio_do_reset- Perform the device reset.
+ * @hw: HW module
+ * @fw_rst: FW reset
+ *
+ * If fw_rst is set, issues FW reset mbox cmd otherwise
+ * does PIO reset.
+ * Performs reset of the function.
+ */
+static int
+csio_do_reset(struct csio_hw *hw, bool fw_rst)
+{
+ struct csio_mb *mbp;
+ enum fw_retval retval;
+
+ if (!fw_rst) {
+ /* PIO reset */
+ csio_wr_reg32(hw, PIORSTMODE | PIORST, PL_RST);
+ mdelay(2000);
+ return 0;
+ }
+
+ mbp = mempool_alloc(hw->mb_mempool, GFP_ATOMIC);
+ if (!mbp) {
+ CSIO_INC_STATS(hw, n_err_nomem);
+ return -ENOMEM;
+ }
+
+ csio_mb_reset(hw, mbp, CSIO_MB_DEFAULT_TMO,
+ PIORSTMODE | PIORST, 0, NULL);
+
+ if (csio_mb_issue(hw, mbp)) {
+ csio_err(hw, "Issue of RESET command failed.n");
+ mempool_free(mbp, hw->mb_mempool);
+ return -EINVAL;
+ }
+
+ retval = csio_mb_fw_retval(mbp);
+ if (retval != FW_SUCCESS) {
+ csio_err(hw, "RESET cmd failed with ret:0x%x.\n", retval);
+ mempool_free(mbp, hw->mb_mempool);
+ return -EINVAL;
+ }
+
+ mempool_free(mbp, hw->mb_mempool);
+
+ return 0;
+}
+
+static int
+csio_hw_validate_caps(struct csio_hw *hw, struct csio_mb *mbp)
+{
+ struct fw_caps_config_cmd *rsp = (struct fw_caps_config_cmd *)mbp->mb;
+ uint16_t caps;
+
+ caps = ntohs(rsp->fcoecaps);
+
+ if (!(caps & FW_CAPS_CONFIG_FCOE_INITIATOR)) {
+ csio_err(hw, "No FCoE Initiator capability in the firmware.\n");
+ return -EINVAL;
+ }
+
+ if (!(caps & FW_CAPS_CONFIG_FCOE_CTRL_OFLD)) {
+ csio_err(hw, "No FCoE Control Offload capability\n");
+ return -EINVAL;
+ }
+
+ return 0;
+}
+
+/*
+ * csio_hw_fw_halt - issue a reset/halt to FW and put uP into RESET
+ * @hw: the HW module
+ * @mbox: mailbox to use for the FW RESET command (if desired)
+ * @force: force uP into RESET even if FW RESET command fails
+ *
+ * Issues a RESET command to firmware (if desired) with a HALT indication
+ * and then puts the microprocessor into RESET state. The RESET command
+ * will only be issued if a legitimate mailbox is provided (mbox <=
+ * PCIE_FW_MASTER_MASK).
+ *
+ * This is generally used in order for the host to safely manipulate the
+ * adapter without fear of conflicting with whatever the firmware might
+ * be doing. The only way out of this state is to RESTART the firmware
+ * ...
+ */
+static int
+csio_hw_fw_halt(struct csio_hw *hw, uint32_t mbox, int32_t force)
+{
+ enum fw_retval retval = 0;
+
+ /*
+ * If a legitimate mailbox is provided, issue a RESET command
+ * with a HALT indication.
+ */
+ if (mbox <= PCIE_FW_MASTER_MASK) {
+ struct csio_mb *mbp;
+
+ mbp = mempool_alloc(hw->mb_mempool, GFP_ATOMIC);
+ if (!mbp) {
+ CSIO_INC_STATS(hw, n_err_nomem);
+ return -ENOMEM;
+ }
+
+ csio_mb_reset(hw, mbp, CSIO_MB_DEFAULT_TMO,
+ PIORSTMODE | PIORST, FW_RESET_CMD_HALT,
+ NULL);
+
+ if (csio_mb_issue(hw, mbp)) {
+ csio_err(hw, "Issue of RESET command failed!\n");
+ mempool_free(mbp, hw->mb_mempool);
+ return -EINVAL;
+ }
+
+ retval = csio_mb_fw_retval(mbp);
+ mempool_free(mbp, hw->mb_mempool);
+ }
+
+ /*
+ * Normally we won't complete the operation if the firmware RESET
+ * command fails but if our caller insists we'll go ahead and put the
+ * uP into RESET. This can be useful if the firmware is hung or even
+ * missing ... We'll have to take the risk of putting the uP into
+ * RESET without the cooperation of firmware in that case.
+ *
+ * We also force the firmware's HALT flag to be on in case we bypassed
+ * the firmware RESET command above or we're dealing with old firmware
+ * which doesn't have the HALT capability. This will serve as a flag
+ * for the incoming firmware to know that it's coming out of a HALT
+ * rather than a RESET ... if it's new enough to understand that ...
+ */
+ if (retval == 0 || force) {
+ csio_set_reg_field(hw, CIM_BOOT_CFG, UPCRST, UPCRST);
+ csio_set_reg_field(hw, PCIE_FW, PCIE_FW_HALT, PCIE_FW_HALT);
+ }
+
+ /*
+ * And we always return the result of the firmware RESET command
+ * even when we force the uP into RESET ...
+ */
+ return retval ? -EINVAL : 0;
+}
+
+/*
+ * csio_hw_fw_restart - restart the firmware by taking the uP out of RESET
+ * @hw: the HW module
+ * @reset: if we want to do a RESET to restart things
+ *
+ * Restart firmware previously halted by csio_hw_fw_halt(). On successful
+ * return the previous PF Master remains as the new PF Master and there
+ * is no need to issue a new HELLO command, etc.
+ *
+ * We do this in two ways:
+ *
+ * 1. If we're dealing with newer firmware we'll simply want to take
+ * the chip's microprocessor out of RESET. This will cause the
+ * firmware to start up from its start vector. And then we'll loop
+ * until the firmware indicates it's started again (PCIE_FW.HALT
+ * reset to 0) or we timeout.
+ *
+ * 2. If we're dealing with older firmware then we'll need to RESET
+ * the chip since older firmware won't recognize the PCIE_FW.HALT
+ * flag and automatically RESET itself on startup.
+ */
+static int
+csio_hw_fw_restart(struct csio_hw *hw, uint32_t mbox, int32_t reset)
+{
+ if (reset) {
+ /*
+ * Since we're directing the RESET instead of the firmware
+ * doing it automatically, we need to clear the PCIE_FW.HALT
+ * bit.
+ */
+ csio_set_reg_field(hw, PCIE_FW, PCIE_FW_HALT, 0);
+
+ /*
+ * If we've been given a valid mailbox, first try to get the
+ * firmware to do the RESET. If that works, great and we can
+ * return success. Otherwise, if we haven't been given a
+ * valid mailbox or the RESET command failed, fall back to
+ * hitting the chip with a hammer.
+ */
+ if (mbox <= PCIE_FW_MASTER_MASK) {
+ csio_set_reg_field(hw, CIM_BOOT_CFG, UPCRST, 0);
+ msleep(100);
+ if (csio_do_reset(hw, true) == 0)
+ return 0;
+ }
+
+ csio_wr_reg32(hw, PIORSTMODE | PIORST, PL_RST);
+ msleep(2000);
+ } else {
+ int ms;
+
+ csio_set_reg_field(hw, CIM_BOOT_CFG, UPCRST, 0);
+ for (ms = 0; ms < FW_CMD_MAX_TIMEOUT; ) {
+ if (!(csio_rd_reg32(hw, PCIE_FW) & PCIE_FW_HALT))
+ return 0;
+ msleep(100);
+ ms += 100;
+ }
+ return -ETIMEDOUT;
+ }
+ return 0;
+}
+
+/*
+ * csio_hw_fw_upgrade - perform all of the steps necessary to upgrade FW
+ * @hw: the HW module
+ * @mbox: mailbox to use for the FW RESET command (if desired)
+ * @fw_data: the firmware image to write
+ * @size: image size
+ * @force: force upgrade even if firmware doesn't cooperate
+ *
+ * Perform all of the steps necessary for upgrading an adapter's
+ * firmware image. Normally this requires the cooperation of the
+ * existing firmware in order to halt all existing activities
+ * but if an invalid mailbox token is passed in we skip that step
+ * (though we'll still put the adapter microprocessor into RESET in
+ * that case).
+ *
+ * On successful return the new firmware will have been loaded and
+ * the adapter will have been fully RESET losing all previous setup
+ * state. On unsuccessful return the adapter may be completely hosed ...
+ * positive errno indicates that the adapter is ~probably~ intact, a
+ * negative errno indicates that things are looking bad ...
+ */
+static int
+csio_hw_fw_upgrade(struct csio_hw *hw, uint32_t mbox,
+ const u8 *fw_data, uint32_t size, int32_t force)
+{
+ const struct fw_hdr *fw_hdr = (const struct fw_hdr *)fw_data;
+ int reset, ret;
+
+ ret = csio_hw_fw_halt(hw, mbox, force);
+ if (ret != 0 && !force)
+ return ret;
+
+ ret = csio_hw_fw_dload(hw, (uint8_t *) fw_data, size);
+ if (ret != 0)
+ return ret;
+
+ /*
+ * Older versions of the firmware don't understand the new
+ * PCIE_FW.HALT flag and so won't know to perform a RESET when they
+ * restart. So for newly loaded older firmware we'll have to do the
+ * RESET for it so it starts up on a clean slate. We can tell if
+ * the newly loaded firmware will handle this right by checking
+ * its header flags to see if it advertises the capability.
+ */
+ reset = ((ntohl(fw_hdr->flags) & FW_HDR_FLAGS_RESET_HALT) == 0);
+ return csio_hw_fw_restart(hw, mbox, reset);
+}
+
+
+/*
+ * csio_hw_fw_config_file - setup an adapter via a Configuration File
+ * @hw: the HW module
+ * @mbox: mailbox to use for the FW command
+ * @mtype: the memory type where the Configuration File is located
+ * @maddr: the memory address where the Configuration File is located
+ * @finiver: return value for CF [fini] version
+ * @finicsum: return value for CF [fini] checksum
+ * @cfcsum: return value for CF computed checksum
+ *
+ * Issue a command to get the firmware to process the Configuration
+ * File located at the specified mtype/maddress. If the Configuration
+ * File is processed successfully and return value pointers are
+ * provided, the Configuration File "[fini] section version and
+ * checksum values will be returned along with the computed checksum.
+ * It's up to the caller to decide how it wants to respond to the
+ * checksums not matching but it recommended that a prominant warning
+ * be emitted in order to help people rapidly identify changed or
+ * corrupted Configuration Files.
+ *
+ * Also note that it's possible to modify things like "niccaps",
+ * "toecaps",etc. between processing the Configuration File and telling
+ * the firmware to use the new configuration. Callers which want to
+ * do this will need to "hand-roll" their own CAPS_CONFIGS commands for
+ * Configuration Files if they want to do this.
+ */
+static int
+csio_hw_fw_config_file(struct csio_hw *hw,
+ unsigned int mtype, unsigned int maddr,
+ uint32_t *finiver, uint32_t *finicsum, uint32_t *cfcsum)
+{
+ struct csio_mb *mbp;
+ struct fw_caps_config_cmd *caps_cmd;
+ int rv = -EINVAL;
+ enum fw_retval ret;
+
+ mbp = mempool_alloc(hw->mb_mempool, GFP_ATOMIC);
+ if (!mbp) {
+ CSIO_INC_STATS(hw, n_err_nomem);
+ return -ENOMEM;
+ }
+ /*
+ * Tell the firmware to process the indicated Configuration File.
+ * If there are no errors and the caller has provided return value
+ * pointers for the [fini] section version, checksum and computed
+ * checksum, pass those back to the caller.
+ */
+ caps_cmd = (struct fw_caps_config_cmd *)(mbp->mb);
+ CSIO_INIT_MBP(mbp, caps_cmd, CSIO_MB_DEFAULT_TMO, hw, NULL, 1);
+ caps_cmd->op_to_write =
+ htonl(FW_CMD_OP(FW_CAPS_CONFIG_CMD) |
+ FW_CMD_REQUEST |
+ FW_CMD_READ);
+ caps_cmd->cfvalid_to_len16 =
+ htonl(FW_CAPS_CONFIG_CMD_CFVALID |
+ FW_CAPS_CONFIG_CMD_MEMTYPE_CF(mtype) |
+ FW_CAPS_CONFIG_CMD_MEMADDR64K_CF(maddr >> 16) |
+ FW_LEN16(*caps_cmd));
+
+ if (csio_mb_issue(hw, mbp)) {
+ csio_err(hw, "Issue of FW_CAPS_CONFIG_CMD failed!\n");
+ goto out;
+ }
+
+ ret = csio_mb_fw_retval(mbp);
+ if (ret != FW_SUCCESS) {
+ csio_dbg(hw, "FW_CAPS_CONFIG_CMD returned %d!\n", rv);
+ goto out;
+ }
+
+ if (finiver)
+ *finiver = ntohl(caps_cmd->finiver);
+ if (finicsum)
+ *finicsum = ntohl(caps_cmd->finicsum);
+ if (cfcsum)
+ *cfcsum = ntohl(caps_cmd->cfcsum);
+
+ /* Validate device capabilities */
+ if (csio_hw_validate_caps(hw, mbp)) {
+ rv = -ENOENT;
+ goto out;
+ }
+
+ /*
+ * And now tell the firmware to use the configuration we just loaded.
+ */
+ caps_cmd->op_to_write =
+ htonl(FW_CMD_OP(FW_CAPS_CONFIG_CMD) |
+ FW_CMD_REQUEST |
+ FW_CMD_WRITE);
+ caps_cmd->cfvalid_to_len16 = htonl(FW_LEN16(*caps_cmd));
+
+ if (csio_mb_issue(hw, mbp)) {
+ csio_err(hw, "Issue of FW_CAPS_CONFIG_CMD failed!\n");
+ goto out;
+ }
+
+ ret = csio_mb_fw_retval(mbp);
+ if (ret != FW_SUCCESS) {
+ csio_dbg(hw, "FW_CAPS_CONFIG_CMD returned %d!\n", rv);
+ goto out;
+ }
+
+ rv = 0;
+out:
+ mempool_free(mbp, hw->mb_mempool);
+ return rv;
+}
+
+/*
+ * csio_get_device_params - Get device parameters.
+ * @hw: HW module
+ *
+ */
+static int
+csio_get_device_params(struct csio_hw *hw)
+{
+ struct csio_wrm *wrm = csio_hw_to_wrm(hw);
+ struct csio_mb *mbp;
+ enum fw_retval retval;
+ u32 param[6];
+ int i, j = 0;
+
+ /* Initialize portids to -1 */
+ for (i = 0; i < CSIO_MAX_PPORTS; i++)
+ hw->pport[i].portid = -1;
+
+ mbp = mempool_alloc(hw->mb_mempool, GFP_ATOMIC);
+ if (!mbp) {
+ CSIO_INC_STATS(hw, n_err_nomem);
+ return -ENOMEM;
+ }
+
+ /* Get port vec information. */
+ param[0] = FW_PARAM_DEV(PORTVEC);
+
+ /* Get Core clock. */
+ param[1] = FW_PARAM_DEV(CCLK);
+
+ /* Get EQ id start and end. */
+ param[2] = FW_PARAM_PFVF(EQ_START);
+ param[3] = FW_PARAM_PFVF(EQ_END);
+
+ /* Get IQ id start and end. */
+ param[4] = FW_PARAM_PFVF(IQFLINT_START);
+ param[5] = FW_PARAM_PFVF(IQFLINT_END);
+
+ csio_mb_params(hw, mbp, CSIO_MB_DEFAULT_TMO, hw->pfn, 0,
+ ARRAY_SIZE(param), param, NULL, false, NULL);
+ if (csio_mb_issue(hw, mbp)) {
+ csio_err(hw, "Issue of FW_PARAMS_CMD(read) failed!\n");
+ mempool_free(mbp, hw->mb_mempool);
+ return -EINVAL;
+ }
+
+ csio_mb_process_read_params_rsp(hw, mbp, &retval,
+ ARRAY_SIZE(param), param);
+ if (retval != FW_SUCCESS) {
+ csio_err(hw, "FW_PARAMS_CMD(read) failed with ret:0x%x!\n",
+ retval);
+ mempool_free(mbp, hw->mb_mempool);
+ return -EINVAL;
+ }
+
+ /* cache the information. */
+ hw->port_vec = param[0];
+ hw->vpd.cclk = param[1];
+ wrm->fw_eq_start = param[2];
+ wrm->fw_iq_start = param[4];
+
+ /* Using FW configured max iqs & eqs */
+ if ((hw->flags & CSIO_HWF_USING_SOFT_PARAMS) ||
+ !csio_is_hw_master(hw)) {
+ hw->cfg_niq = param[5] - param[4] + 1;
+ hw->cfg_neq = param[3] - param[2] + 1;
+ csio_dbg(hw, "Using fwconfig max niqs %d neqs %d\n",
+ hw->cfg_niq, hw->cfg_neq);
+ }
+
+ hw->port_vec &= csio_port_mask;
+
+ hw->num_pports = hweight32(hw->port_vec);
+
+ csio_dbg(hw, "Port vector: 0x%x, #ports: %d\n",
+ hw->port_vec, hw->num_pports);
+
+ for (i = 0; i < hw->num_pports; i++) {
+ while ((hw->port_vec & (1 << j)) == 0)
+ j++;
+ hw->pport[i].portid = j++;
+ csio_dbg(hw, "Found Port:%d\n", hw->pport[i].portid);
+ }
+ mempool_free(mbp, hw->mb_mempool);
+
+ return 0;
+}
+
+
+/*
+ * csio_config_device_caps - Get and set device capabilities.
+ * @hw: HW module
+ *
+ */
+static int
+csio_config_device_caps(struct csio_hw *hw)
+{
+ struct csio_mb *mbp;
+ enum fw_retval retval;
+ int rv = -EINVAL;
+
+ mbp = mempool_alloc(hw->mb_mempool, GFP_ATOMIC);
+ if (!mbp) {
+ CSIO_INC_STATS(hw, n_err_nomem);
+ return -ENOMEM;
+ }
+
+ /* Get device capabilities */
+ csio_mb_caps_config(hw, mbp, CSIO_MB_DEFAULT_TMO, 0, 0, 0, 0, NULL);
+
+ if (csio_mb_issue(hw, mbp)) {
+ csio_err(hw, "Issue of FW_CAPS_CONFIG_CMD(r) failed!\n");
+ goto out;
+ }
+
+ retval = csio_mb_fw_retval(mbp);
+ if (retval != FW_SUCCESS) {
+ csio_err(hw, "FW_CAPS_CONFIG_CMD(r) returned %d!\n", retval);
+ goto out;
+ }
+
+ /* Validate device capabilities */
+ if (csio_hw_validate_caps(hw, mbp))
+ goto out;
+
+ /* Don't config device capabilities if already configured */
+ if (hw->fw_state == CSIO_DEV_STATE_INIT) {
+ rv = 0;
+ goto out;
+ }
+
+ /* Write back desired device capabilities */
+ csio_mb_caps_config(hw, mbp, CSIO_MB_DEFAULT_TMO, true, true,
+ false, true, NULL);
+
+ if (csio_mb_issue(hw, mbp)) {
+ csio_err(hw, "Issue of FW_CAPS_CONFIG_CMD(w) failed!\n");
+ goto out;
+ }
+
+ retval = csio_mb_fw_retval(mbp);
+ if (retval != FW_SUCCESS) {
+ csio_err(hw, "FW_CAPS_CONFIG_CMD(w) returned %d!\n", retval);
+ goto out;
+ }
+
+ rv = 0;
+out:
+ mempool_free(mbp, hw->mb_mempool);
+ return rv;
+}
+
+static int
+csio_config_global_rss(struct csio_hw *hw)
+{
+ struct csio_mb *mbp;
+ enum fw_retval retval;
+
+ mbp = mempool_alloc(hw->mb_mempool, GFP_ATOMIC);
+ if (!mbp) {
+ CSIO_INC_STATS(hw, n_err_nomem);
+ return -ENOMEM;
+ }
+
+ csio_rss_glb_config(hw, mbp, CSIO_MB_DEFAULT_TMO,
+ FW_RSS_GLB_CONFIG_CMD_MODE_BASICVIRTUAL,
+ FW_RSS_GLB_CONFIG_CMD_TNLMAPEN |
+ FW_RSS_GLB_CONFIG_CMD_HASHTOEPLITZ |
+ FW_RSS_GLB_CONFIG_CMD_TNLALLLKP,
+ NULL);
+
+ if (csio_mb_issue(hw, mbp)) {
+ csio_err(hw, "Issue of FW_RSS_GLB_CONFIG_CMD failed!\n");
+ mempool_free(mbp, hw->mb_mempool);
+ return -EINVAL;
+ }
+
+ retval = csio_mb_fw_retval(mbp);
+ if (retval != FW_SUCCESS) {
+ csio_err(hw, "FW_RSS_GLB_CONFIG_CMD returned 0x%x!\n", retval);
+ mempool_free(mbp, hw->mb_mempool);
+ return -EINVAL;
+ }
+
+ mempool_free(mbp, hw->mb_mempool);
+
+ return 0;
+}
+
+/*
+ * csio_config_pfvf - Configure Physical/Virtual functions settings.
+ * @hw: HW module
+ *
+ */
+static int
+csio_config_pfvf(struct csio_hw *hw)
+{
+ struct csio_mb *mbp;
+ enum fw_retval retval;
+
+ mbp = mempool_alloc(hw->mb_mempool, GFP_ATOMIC);
+ if (!mbp) {
+ CSIO_INC_STATS(hw, n_err_nomem);
+ return -ENOMEM;
+ }
+
+ /*
+ * For now, allow all PFs to access to all ports using a pmask
+ * value of 0xF (M_FW_PFVF_CMD_PMASK). Once we have VFs, we will
+ * need to provide access based on some rule.
+ */
+ csio_mb_pfvf(hw, mbp, CSIO_MB_DEFAULT_TMO, hw->pfn, 0, CSIO_NEQ,
+ CSIO_NETH_CTRL, CSIO_NIQ_FLINT, 0, 0, CSIO_NVI, CSIO_CMASK,
+ CSIO_PMASK, CSIO_NEXACTF, CSIO_R_CAPS, CSIO_WX_CAPS, NULL);
+
+ if (csio_mb_issue(hw, mbp)) {
+ csio_err(hw, "Issue of FW_PFVF_CMD failed!\n");
+ mempool_free(mbp, hw->mb_mempool);
+ return -EINVAL;
+ }
+
+ retval = csio_mb_fw_retval(mbp);
+ if (retval != FW_SUCCESS) {
+ csio_err(hw, "FW_PFVF_CMD returned 0x%x!\n", retval);
+ mempool_free(mbp, hw->mb_mempool);
+ return -EINVAL;
+ }
+
+ mempool_free(mbp, hw->mb_mempool);
+
+ return 0;
+}
+
+/*
+ * csio_enable_ports - Bring up all available ports.
+ * @hw: HW module.
+ *
+ */
+static int
+csio_enable_ports(struct csio_hw *hw)
+{
+ struct csio_mb *mbp;
+ enum fw_retval retval;
+ uint8_t portid;
+ int i;
+
+ mbp = mempool_alloc(hw->mb_mempool, GFP_ATOMIC);
+ if (!mbp) {
+ CSIO_INC_STATS(hw, n_err_nomem);
+ return -ENOMEM;
+ }
+
+ for (i = 0; i < hw->num_pports; i++) {
+ portid = hw->pport[i].portid;
+
+ /* Read PORT information */
+ csio_mb_port(hw, mbp, CSIO_MB_DEFAULT_TMO, portid,
+ false, 0, 0, NULL);
+
+ if (csio_mb_issue(hw, mbp)) {
+ csio_err(hw, "failed to issue FW_PORT_CMD(r) port:%d\n",
+ portid);
+ mempool_free(mbp, hw->mb_mempool);
+ return -EINVAL;
+ }
+
+ csio_mb_process_read_port_rsp(hw, mbp, &retval,
+ &hw->pport[i].pcap);
+ if (retval != FW_SUCCESS) {
+ csio_err(hw, "FW_PORT_CMD(r) port:%d failed: 0x%x\n",
+ portid, retval);
+ mempool_free(mbp, hw->mb_mempool);
+ return -EINVAL;
+ }
+
+ /* Write back PORT information */
+ csio_mb_port(hw, mbp, CSIO_MB_DEFAULT_TMO, portid, true,
+ (PAUSE_RX | PAUSE_TX), hw->pport[i].pcap, NULL);
+
+ if (csio_mb_issue(hw, mbp)) {
+ csio_err(hw, "failed to issue FW_PORT_CMD(w) port:%d\n",
+ portid);
+ mempool_free(mbp, hw->mb_mempool);
+ return -EINVAL;
+ }
+
+ retval = csio_mb_fw_retval(mbp);
+ if (retval != FW_SUCCESS) {
+ csio_err(hw, "FW_PORT_CMD(w) port:%d failed :0x%x\n",
+ portid, retval);
+ mempool_free(mbp, hw->mb_mempool);
+ return -EINVAL;
+ }
+
+ } /* For all ports */
+
+ mempool_free(mbp, hw->mb_mempool);
+
+ return 0;
+}
+
+/*
+ * csio_get_fcoe_resinfo - Read fcoe fw resource info.
+ * @hw: HW module
+ * Issued with lock held.
+ */
+static int
+csio_get_fcoe_resinfo(struct csio_hw *hw)
+{
+ struct csio_fcoe_res_info *res_info = &hw->fres_info;
+ struct fw_fcoe_res_info_cmd *rsp;
+ struct csio_mb *mbp;
+ enum fw_retval retval;
+
+ mbp = mempool_alloc(hw->mb_mempool, GFP_ATOMIC);
+ if (!mbp) {
+ CSIO_INC_STATS(hw, n_err_nomem);
+ return -ENOMEM;
+ }
+
+ /* Get FCoE FW resource information */
+ csio_fcoe_read_res_info_init_mb(hw, mbp, CSIO_MB_DEFAULT_TMO, NULL);
+
+ if (csio_mb_issue(hw, mbp)) {
+ csio_err(hw, "failed to issue FW_FCOE_RES_INFO_CMD\n");
+ mempool_free(mbp, hw->mb_mempool);
+ return -EINVAL;
+ }
+
+ rsp = (struct fw_fcoe_res_info_cmd *)(mbp->mb);
+ retval = FW_CMD_RETVAL_GET(ntohl(rsp->retval_len16));
+ if (retval != FW_SUCCESS) {
+ csio_err(hw, "FW_FCOE_RES_INFO_CMD failed with ret x%x\n",
+ retval);
+ mempool_free(mbp, hw->mb_mempool);
+ return -EINVAL;
+ }
+
+ res_info->e_d_tov = ntohs(rsp->e_d_tov);
+ res_info->r_a_tov_seq = ntohs(rsp->r_a_tov_seq);
+ res_info->r_a_tov_els = ntohs(rsp->r_a_tov_els);
+ res_info->r_r_tov = ntohs(rsp->r_r_tov);
+ res_info->max_xchgs = ntohl(rsp->max_xchgs);
+ res_info->max_ssns = ntohl(rsp->max_ssns);
+ res_info->used_xchgs = ntohl(rsp->used_xchgs);
+ res_info->used_ssns = ntohl(rsp->used_ssns);
+ res_info->max_fcfs = ntohl(rsp->max_fcfs);
+ res_info->max_vnps = ntohl(rsp->max_vnps);
+ res_info->used_fcfs = ntohl(rsp->used_fcfs);
+ res_info->used_vnps = ntohl(rsp->used_vnps);
+
+ csio_dbg(hw, "max ssns:%d max xchgs:%d\n", res_info->max_ssns,
+ res_info->max_xchgs);
+ mempool_free(mbp, hw->mb_mempool);
+
+ return 0;
+}
+
+static int
+csio_hw_check_fwconfig(struct csio_hw *hw, u32 *param)
+{
+ struct csio_mb *mbp;
+ enum fw_retval retval;
+ u32 _param[1];
+
+ mbp = mempool_alloc(hw->mb_mempool, GFP_ATOMIC);
+ if (!mbp) {
+ CSIO_INC_STATS(hw, n_err_nomem);
+ return -ENOMEM;
+ }
+
+ /*
+ * Find out whether we're dealing with a version of
+ * the firmware which has configuration file support.
+ */
+ _param[0] = (FW_PARAMS_MNEM(FW_PARAMS_MNEM_DEV) |
+ FW_PARAMS_PARAM_X(FW_PARAMS_PARAM_DEV_CF));
+
+ csio_mb_params(hw, mbp, CSIO_MB_DEFAULT_TMO, hw->pfn, 0,
+ ARRAY_SIZE(_param), _param, NULL, false, NULL);
+ if (csio_mb_issue(hw, mbp)) {
+ csio_err(hw, "Issue of FW_PARAMS_CMD(read) failed!\n");
+ mempool_free(mbp, hw->mb_mempool);
+ return -EINVAL;
+ }
+
+ csio_mb_process_read_params_rsp(hw, mbp, &retval,
+ ARRAY_SIZE(_param), _param);
+ if (retval != FW_SUCCESS) {
+ csio_err(hw, "FW_PARAMS_CMD(read) failed with ret:0x%x!\n",
+ retval);
+ mempool_free(mbp, hw->mb_mempool);
+ return -EINVAL;
+ }
+
+ mempool_free(mbp, hw->mb_mempool);
+ *param = _param[0];
+
+ return 0;
+}
+
+static int
+csio_hw_flash_config(struct csio_hw *hw, u32 *fw_cfg_param, char *path)
+{
+ int ret = 0;
+ const struct firmware *cf;
+ struct pci_dev *pci_dev = hw->pdev;
+ struct device *dev = &pci_dev->dev;
+
+ unsigned int mtype = 0, maddr = 0;
+ 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);
+ return -ENOENT;
+ }
+
+ if (cf->size%4 != 0)
+ value_to_add = 4 - (cf->size % 4);
+
+ cfg_data = kzalloc(cf->size+value_to_add, GFP_KERNEL);
+ if (cfg_data == NULL)
+ return -ENOMEM;
+
+ memcpy((void *)cfg_data, (const void *)cf->data, cf->size);
+
+ if (csio_hw_check_fwconfig(hw, fw_cfg_param) != 0)
+ return -EINVAL;
+
+ mtype = FW_PARAMS_PARAM_Y_GET(*fw_cfg_param);
+ maddr = FW_PARAMS_PARAM_Z_GET(*fw_cfg_param) << 16;
+
+ ret = csio_memory_write(hw, mtype, maddr,
+ cf->size + value_to_add, cfg_data);
+ if (ret == 0) {
+ csio_info(hw, "config file upgraded to " CSIO_CF_FNAME "\n");
+ strncpy(path, "/lib/firmware/" CSIO_CF_FNAME, 64);
+ }
+
+ kfree(cfg_data);
+ release_firmware(cf);
+
+ return ret;
+}
+
+/*
+ * HW initialization: contact FW, obtain config, perform basic init.
+ *
+ * If the firmware we're dealing with has Configuration File support, then
+ * we use that to perform all configuration -- either using the configuration
+ * file stored in flash on the adapter or using a filesystem-local file
+ * if available.
+ *
+ * If we don't have configuration file support in the firmware, then we'll
+ * have to set things up the old fashioned way with hard-coded register
+ * writes and firmware commands ...
+ */
+
+/*
+ * Attempt to initialize the HW via a Firmware Configuration File.
+ */
+static int
+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;
+ int using_flash;
+ char path[64];
+
+ /*
+ * Reset device if necessary
+ */
+ if (reset) {
+ rv = csio_do_reset(hw, true);
+ if (rv != 0)
+ goto bye;
+ }
+
+ /*
+ * If we have a configuration file in host ,
+ * then use that. Otherwise, use the configuration file stored
+ * in the HW flash ...
+ */
+ spin_unlock_irq(&hw->lock);
+ rv = csio_hw_flash_config(hw, fw_cfg_param, path);
+ spin_lock_irq(&hw->lock);
+ if (rv != 0) {
+ if (rv == -ENOENT) {
+ /*
+ * config file was not found. Use default
+ * config file from flash.
+ */
+ mtype = FW_MEMTYPE_CF_FLASH;
+ maddr = csio_hw_flash_cfg_addr(hw);
+ using_flash = 1;
+ } else {
+ /*
+ * we revert back to the hardwired config if
+ * flashing failed.
+ */
+ goto bye;
+ }
+ } else {
+ mtype = FW_PARAMS_PARAM_Y_GET(*fw_cfg_param);
+ maddr = FW_PARAMS_PARAM_Z_GET(*fw_cfg_param) << 16;
+ using_flash = 0;
+ }
+
+ hw->cfg_store = (uint8_t)mtype;
+
+ /*
+ * Issue a Capability Configuration command to the firmware to get it
+ * to parse the Configuration File.
+ */
+ rv = csio_hw_fw_config_file(hw, mtype, maddr, &finiver,
+ &finicsum, &cfcsum);
+ if (rv != 0)
+ goto bye;
+
+ hw->cfg_finiver = finiver;
+ hw->cfg_finicsum = finicsum;
+ hw->cfg_cfcsum = cfcsum;
+ hw->cfg_csum_status = true;
+
+ if (finicsum != cfcsum) {
+ csio_warn(hw,
+ "Config File checksum mismatch: csum=%#x, computed=%#x\n",
+ finicsum, cfcsum);
+
+ hw->cfg_csum_status = false;
+ }
+
+ /*
+ * Note that we're operating with parameters
+ * not supplied by the driver, rather than from hard-wired
+ * initialization constants buried in the driver.
+ */
+ hw->flags |= CSIO_HWF_USING_SOFT_PARAMS;
+
+ /* device parameters */
+ rv = csio_get_device_params(hw);
+ if (rv != 0)
+ goto bye;
+
+ /* Configure SGE */
+ csio_wr_sge_init(hw);
+
+ /*
+ * And finally tell the firmware to initialize itself using the
+ * parameters from the Configuration File.
+ */
+ /* Post event to notify completion of configuration */
+ csio_post_event(&hw->sm, CSIO_HWE_INIT);
+
+ csio_info(hw,
+ "Firmware Configuration File %s, version %#x, computed checksum %#x\n",
+ (using_flash ? "in device FLASH" : path), finiver, cfcsum);
+
+ return 0;
+
+ /*
+ * Something bad happened. Return the error ...
+ */
+bye:
+ hw->flags &= ~CSIO_HWF_USING_SOFT_PARAMS;
+ csio_dbg(hw, "Configuration file error %d\n", rv);
+ return rv;
+}
+
+/*
+ * Attempt to initialize the adapter via hard-coded, driver supplied
+ * parameters ...
+ */
+static int
+csio_hw_no_fwconfig(struct csio_hw *hw, int reset)
+{
+ int rv;
+ /*
+ * Reset device if necessary
+ */
+ if (reset) {
+ rv = csio_do_reset(hw, true);
+ if (rv != 0)
+ goto out;
+ }
+
+ /* Get and set device capabilities */
+ rv = csio_config_device_caps(hw);
+ if (rv != 0)
+ goto out;
+
+ /* Config Global RSS command */
+ rv = csio_config_global_rss(hw);
+ if (rv != 0)
+ goto out;
+
+ /* Configure PF/VF capabilities of device */
+ rv = csio_config_pfvf(hw);
+ if (rv != 0)
+ goto out;
+
+ /* device parameters */
+ rv = csio_get_device_params(hw);
+ if (rv != 0)
+ goto out;
+
+ /* Configure SGE */
+ csio_wr_sge_init(hw);
+
+ /* Post event to notify completion of configuration */
+ csio_post_event(&hw->sm, CSIO_HWE_INIT);
+
+out:
+ return rv;
+}
+
+/*
+ * Returns -EINVAL if attempts to flash the firmware failed
+ * else returns 0,
+ * if flashing was not attempted because the card had the
+ * latest firmware ECANCELED is returned
+ */
+static int
+csio_hw_flash_fw(struct csio_hw *hw)
+{
+ int ret = -ECANCELED;
+ const struct firmware *fw;
+ const struct fw_hdr *hdr;
+ u32 fw_ver;
+ 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);
+ 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)
+ 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 ||
+ 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);
+ else
+ csio_err(hw, "firmware upgrade failed! err=%d\n", ret);
+ }
+
+ release_firmware(fw);
+
+ return ret;
+}
+
+
+/*
+ * csio_hw_configure - Configure HW
+ * @hw - HW module
+ *
+ */
+static void
+csio_hw_configure(struct csio_hw *hw)
+{
+ int reset = 1;
+ int rv;
+ u32 param[1];
+
+ rv = csio_hw_dev_ready(hw);
+ if (rv != 0) {
+ CSIO_INC_STATS(hw, n_err_fatal);
+ csio_post_event(&hw->sm, CSIO_HWE_FATAL);
+ goto out;
+ }
+
+ /* HW version */
+ hw->chip_ver = (char)csio_rd_reg32(hw, PL_REV);
+
+ /* Needed for FW download */
+ rv = csio_hw_get_flash_params(hw);
+ if (rv != 0) {
+ csio_err(hw, "Failed to get serial flash params rv:%d\n", rv);
+ csio_post_event(&hw->sm, CSIO_HWE_FATAL);
+ goto out;
+ }
+
+ /* Set pci completion timeout value to 4 seconds. */
+ csio_set_pcie_completion_timeout(hw, 0xd);
+
+ csio_hw_set_mem_win(hw);
+
+ rv = csio_hw_get_fw_version(hw, &hw->fwrev);
+ if (rv != 0)
+ goto out;
+
+ csio_hw_print_fw_version(hw, "Firmware revision");
+
+ rv = csio_do_hello(hw, &hw->fw_state);
+ if (rv != 0) {
+ CSIO_INC_STATS(hw, n_err_fatal);
+ csio_post_event(&hw->sm, CSIO_HWE_FATAL);
+ goto out;
+ }
+
+ /* Read vpd */
+ rv = csio_hw_get_vpd_params(hw, &hw->vpd);
+ if (rv != 0)
+ goto out;
+
+ if (csio_is_hw_master(hw) && hw->fw_state != CSIO_DEV_STATE_INIT) {
+ rv = csio_hw_check_fw_version(hw);
+ if (rv == -EINVAL) {
+
+ /* Do firmware update */
+ spin_unlock_irq(&hw->lock);
+ rv = csio_hw_flash_fw(hw);
+ spin_lock_irq(&hw->lock);
+
+ if (rv == 0) {
+ reset = 0;
+ /*
+ * Note that the chip was reset as part of the
+ * firmware upgrade so we don't reset it again
+ * below and grab the new firmware version.
+ */
+ rv = csio_hw_check_fw_version(hw);
+ }
+ }
+ /*
+ * If the firmware doesn't support Configuration
+ * Files, use the old Driver-based, hard-wired
+ * initialization. Otherwise, try using the
+ * Configuration File support and fall back to the
+ * Driver-based initialization if there's no
+ * Configuration File found.
+ */
+ if (csio_hw_check_fwconfig(hw, param) == 0) {
+ rv = csio_hw_use_fwconfig(hw, reset, param);
+ if (rv == -ENOENT)
+ goto out;
+ if (rv != 0) {
+ csio_info(hw,
+ "No Configuration File present "
+ "on adapter. Using hard-wired "
+ "configuration parameters.\n");
+ rv = csio_hw_no_fwconfig(hw, reset);
+ }
+ } else {
+ rv = csio_hw_no_fwconfig(hw, reset);
+ }
+
+ if (rv != 0)
+ goto out;
+
+ } else {
+ if (hw->fw_state == CSIO_DEV_STATE_INIT) {
+
+ /* device parameters */
+ rv = csio_get_device_params(hw);
+ if (rv != 0)
+ goto out;
+
+ /* Get device capabilities */
+ rv = csio_config_device_caps(hw);
+ if (rv != 0)
+ goto out;
+
+ /* Configure SGE */
+ csio_wr_sge_init(hw);
+
+ /* Post event to notify completion of configuration */
+ csio_post_event(&hw->sm, CSIO_HWE_INIT);
+ goto out;
+ }
+ } /* if not master */
+
+out:
+ return;
+}
+
+/*
+ * csio_hw_initialize - Initialize HW
+ * @hw - HW module
+ *
+ */
+static void
+csio_hw_initialize(struct csio_hw *hw)
+{
+ struct csio_mb *mbp;
+ enum fw_retval retval;
+ int rv;
+ int i;
+
+ if (csio_is_hw_master(hw) && hw->fw_state != CSIO_DEV_STATE_INIT) {
+ mbp = mempool_alloc(hw->mb_mempool, GFP_ATOMIC);
+ if (!mbp)
+ goto out;
+
+ csio_mb_initialize(hw, mbp, CSIO_MB_DEFAULT_TMO, NULL);
+
+ if (csio_mb_issue(hw, mbp)) {
+ csio_err(hw, "Issue of FW_INITIALIZE_CMD failed!\n");
+ goto free_and_out;
+ }
+
+ retval = csio_mb_fw_retval(mbp);
+ if (retval != FW_SUCCESS) {
+ csio_err(hw, "FW_INITIALIZE_CMD returned 0x%x!\n",
+ retval);
+ goto free_and_out;
+ }
+
+ mempool_free(mbp, hw->mb_mempool);
+ }
+
+ rv = csio_get_fcoe_resinfo(hw);
+ if (rv != 0) {
+ csio_err(hw, "Failed to read fcoe resource info: %d\n", rv);
+ goto out;
+ }
+
+ spin_unlock_irq(&hw->lock);
+ rv = csio_config_queues(hw);
+ spin_lock_irq(&hw->lock);
+
+ if (rv != 0) {
+ csio_err(hw, "Config of queues failed!: %d\n", rv);
+ goto out;
+ }
+
+ for (i = 0; i < hw->num_pports; i++)
+ hw->pport[i].mod_type = FW_PORT_MOD_TYPE_NA;
+
+ if (csio_is_hw_master(hw) && hw->fw_state != CSIO_DEV_STATE_INIT) {
+ rv = csio_enable_ports(hw);
+ if (rv != 0) {
+ csio_err(hw, "Failed to enable ports: %d\n", rv);
+ goto out;
+ }
+ }
+
+ csio_post_event(&hw->sm, CSIO_HWE_INIT_DONE);
+ return;
+
+free_and_out:
+ mempool_free(mbp, hw->mb_mempool);
+out:
+ return;
+}
+
+#define PF_INTR_MASK (PFSW | PFCIM)
+
+/*
+ * csio_hw_intr_enable - Enable HW interrupts
+ * @hw: Pointer to HW module.
+ *
+ * Enable interrupts in HW registers.
+ */
+static void
+csio_hw_intr_enable(struct csio_hw *hw)
+{
+ uint16_t vec = (uint16_t)csio_get_mb_intr_idx(csio_hw_to_mbm(hw));
+ uint32_t pf = SOURCEPF_GET(csio_rd_reg32(hw, PL_WHOAMI));
+ uint32_t pl = csio_rd_reg32(hw, PL_INT_ENABLE);
+
+ /*
+ * Set aivec for MSI/MSIX. PCIE_PF_CFG.INTXType is set up
+ * by FW, so do nothing for INTX.
+ */
+ if (hw->intr_mode == CSIO_IM_MSIX)
+ csio_set_reg_field(hw, MYPF_REG(PCIE_PF_CFG),
+ AIVEC(AIVEC_MASK), vec);
+ else if (hw->intr_mode == CSIO_IM_MSI)
+ csio_set_reg_field(hw, MYPF_REG(PCIE_PF_CFG),
+ AIVEC(AIVEC_MASK), 0);
+
+ csio_wr_reg32(hw, PF_INTR_MASK, MYPF_REG(PL_PF_INT_ENABLE));
+
+ /* Turn on MB interrupts - this will internally flush PIO as well */
+ csio_mb_intr_enable(hw);
+
+ /* These are common registers - only a master can modify them */
+ if (csio_is_hw_master(hw)) {
+ /*
+ * Disable the Serial FLASH interrupt, if enabled!
+ */
+ pl &= (~SF);
+ csio_wr_reg32(hw, pl, PL_INT_ENABLE);
+
+ csio_wr_reg32(hw, ERR_CPL_EXCEED_IQE_SIZE |
+ EGRESS_SIZE_ERR | ERR_INVALID_CIDX_INC |
+ ERR_CPL_OPCODE_0 | ERR_DROPPED_DB |
+ ERR_DATA_CPL_ON_HIGH_QID1 |
+ ERR_DATA_CPL_ON_HIGH_QID0 | ERR_BAD_DB_PIDX3 |
+ ERR_BAD_DB_PIDX2 | ERR_BAD_DB_PIDX1 |
+ ERR_BAD_DB_PIDX0 | ERR_ING_CTXT_PRIO |
+ ERR_EGR_CTXT_PRIO | INGRESS_SIZE_ERR,
+ SGE_INT_ENABLE3);
+ csio_set_reg_field(hw, PL_INT_MAP0, 0, 1 << pf);
+ }
+
+ hw->flags |= CSIO_HWF_HW_INTR_ENABLED;
+
+}
+
+/*
+ * csio_hw_intr_disable - Disable HW interrupts
+ * @hw: Pointer to HW module.
+ *
+ * Turn off Mailbox and PCI_PF_CFG interrupts.
+ */
+void
+csio_hw_intr_disable(struct csio_hw *hw)
+{
+ uint32_t pf = SOURCEPF_GET(csio_rd_reg32(hw, PL_WHOAMI));
+
+ if (!(hw->flags & CSIO_HWF_HW_INTR_ENABLED))
+ return;
+
+ hw->flags &= ~CSIO_HWF_HW_INTR_ENABLED;
+
+ csio_wr_reg32(hw, 0, MYPF_REG(PL_PF_INT_ENABLE));
+ if (csio_is_hw_master(hw))
+ csio_set_reg_field(hw, PL_INT_MAP0, 1 << pf, 0);
+
+ /* Turn off MB interrupts */
+ csio_mb_intr_disable(hw);
+
+}
+
+static void
+csio_hw_fatal_err(struct csio_hw *hw)
+{
+ csio_set_reg_field(hw, SGE_CONTROL, GLOBALENABLE, 0);
+ csio_hw_intr_disable(hw);
+
+ /* Do not reset HW, we may need FW state for debugging */
+ csio_fatal(hw, "HW Fatal error encountered!\n");
+}
+
+/*****************************************************************************/
+/* START: HW SM */
+/*****************************************************************************/
+/*
+ * csio_hws_uninit - Uninit state
+ * @hw - HW module
+ * @evt - Event
+ *
+ */
+static void
+csio_hws_uninit(struct csio_hw *hw, enum csio_hw_ev evt)
+{
+ hw->prev_evt = hw->cur_evt;
+ hw->cur_evt = evt;
+ CSIO_INC_STATS(hw, n_evt_sm[evt]);
+
+ switch (evt) {
+ case CSIO_HWE_CFG:
+ csio_set_state(&hw->sm, csio_hws_configuring);
+ csio_hw_configure(hw);
+ break;
+
+ default:
+ CSIO_INC_STATS(hw, n_evt_unexp);
+ break;
+ }
+}
+
+/*
+ * csio_hws_configuring - Configuring state
+ * @hw - HW module
+ * @evt - Event
+ *
+ */
+static void
+csio_hws_configuring(struct csio_hw *hw, enum csio_hw_ev evt)
+{
+ hw->prev_evt = hw->cur_evt;
+ hw->cur_evt = evt;
+ CSIO_INC_STATS(hw, n_evt_sm[evt]);
+
+ switch (evt) {
+ case CSIO_HWE_INIT:
+ csio_set_state(&hw->sm, csio_hws_initializing);
+ csio_hw_initialize(hw);
+ break;
+
+ case CSIO_HWE_INIT_DONE:
+ csio_set_state(&hw->sm, csio_hws_ready);
+ /* Fan out event to all lnode SMs */
+ csio_notify_lnodes(hw, CSIO_LN_NOTIFY_HWREADY);
+ break;
+
+ case CSIO_HWE_FATAL:
+ csio_set_state(&hw->sm, csio_hws_uninit);
+ break;
+
+ case CSIO_HWE_PCI_REMOVE:
+ csio_do_bye(hw);
+ break;
+ default:
+ CSIO_INC_STATS(hw, n_evt_unexp);
+ break;
+ }
+}
+
+/*
+ * csio_hws_initializing - Initialiazing state
+ * @hw - HW module
+ * @evt - Event
+ *
+ */
+static void
+csio_hws_initializing(struct csio_hw *hw, enum csio_hw_ev evt)
+{
+ hw->prev_evt = hw->cur_evt;
+ hw->cur_evt = evt;
+ CSIO_INC_STATS(hw, n_evt_sm[evt]);
+
+ switch (evt) {
+ case CSIO_HWE_INIT_DONE:
+ csio_set_state(&hw->sm, csio_hws_ready);
+
+ /* Fan out event to all lnode SMs */
+ csio_notify_lnodes(hw, CSIO_LN_NOTIFY_HWREADY);
+
+ /* Enable interrupts */
+ csio_hw_intr_enable(hw);
+ break;
+
+ case CSIO_HWE_FATAL:
+ csio_set_state(&hw->sm, csio_hws_uninit);
+ break;
+
+ case CSIO_HWE_PCI_REMOVE:
+ csio_do_bye(hw);
+ break;
+
+ default:
+ CSIO_INC_STATS(hw, n_evt_unexp);
+ break;
+ }
+}
+
+/*
+ * csio_hws_ready - Ready state
+ * @hw - HW module
+ * @evt - Event
+ *
+ */
+static void
+csio_hws_ready(struct csio_hw *hw, enum csio_hw_ev evt)
+{
+ /* Remember the event */
+ hw->evtflag = evt;
+
+ hw->prev_evt = hw->cur_evt;
+ hw->cur_evt = evt;
+ CSIO_INC_STATS(hw, n_evt_sm[evt]);
+
+ switch (evt) {
+ case CSIO_HWE_HBA_RESET:
+ case CSIO_HWE_FW_DLOAD:
+ case CSIO_HWE_SUSPEND:
+ case CSIO_HWE_PCI_REMOVE:
+ case CSIO_HWE_PCIERR_DETECTED:
+ csio_set_state(&hw->sm, csio_hws_quiescing);
+ /* cleanup all outstanding cmds */
+ if (evt == CSIO_HWE_HBA_RESET ||
+ evt == CSIO_HWE_PCIERR_DETECTED)
+ csio_scsim_cleanup_io(csio_hw_to_scsim(hw), false);
+ else
+ csio_scsim_cleanup_io(csio_hw_to_scsim(hw), true);
+
+ csio_hw_intr_disable(hw);
+ csio_hw_mbm_cleanup(hw);
+ csio_evtq_stop(hw);
+ csio_notify_lnodes(hw, CSIO_LN_NOTIFY_HWSTOP);
+ csio_evtq_flush(hw);
+ csio_mgmtm_cleanup(csio_hw_to_mgmtm(hw));
+ csio_post_event(&hw->sm, CSIO_HWE_QUIESCED);
+ break;
+
+ case CSIO_HWE_FATAL:
+ csio_set_state(&hw->sm, csio_hws_uninit);
+ break;
+
+ default:
+ CSIO_INC_STATS(hw, n_evt_unexp);
+ break;
+ }
+}
+
+/*
+ * csio_hws_quiescing - Quiescing state
+ * @hw - HW module
+ * @evt - Event
+ *
+ */
+static void
+csio_hws_quiescing(struct csio_hw *hw, enum csio_hw_ev evt)
+{
+ hw->prev_evt = hw->cur_evt;
+ hw->cur_evt = evt;
+ CSIO_INC_STATS(hw, n_evt_sm[evt]);
+
+ switch (evt) {
+ case CSIO_HWE_QUIESCED:
+ switch (hw->evtflag) {
+ case CSIO_HWE_FW_DLOAD:
+ csio_set_state(&hw->sm, csio_hws_resetting);
+ /* Download firmware */
+ /* Fall through */
+
+ case CSIO_HWE_HBA_RESET:
+ csio_set_state(&hw->sm, csio_hws_resetting);
+ /* Start reset of the HBA */
+ csio_notify_lnodes(hw, CSIO_LN_NOTIFY_HWRESET);
+ csio_wr_destroy_queues(hw, false);
+ csio_do_reset(hw, false);
+ csio_post_event(&hw->sm, CSIO_HWE_HBA_RESET_DONE);
+ break;
+
+ case CSIO_HWE_PCI_REMOVE:
+ csio_set_state(&hw->sm, csio_hws_removing);
+ csio_notify_lnodes(hw, CSIO_LN_NOTIFY_HWREMOVE);
+ csio_wr_destroy_queues(hw, true);
+ /* Now send the bye command */
+ csio_do_bye(hw);
+ break;
+
+ case CSIO_HWE_SUSPEND:
+ csio_set_state(&hw->sm, csio_hws_quiesced);
+ break;
+
+ case CSIO_HWE_PCIERR_DETECTED:
+ csio_set_state(&hw->sm, csio_hws_pcierr);
+ csio_wr_destroy_queues(hw, false);
+ break;
+
+ default:
+ CSIO_INC_STATS(hw, n_evt_unexp);
+ break;
+
+ }
+ break;
+
+ default:
+ CSIO_INC_STATS(hw, n_evt_unexp);
+ break;
+ }
+}
+
+/*
+ * csio_hws_quiesced - Quiesced state
+ * @hw - HW module
+ * @evt - Event
+ *
+ */
+static void
+csio_hws_quiesced(struct csio_hw *hw, enum csio_hw_ev evt)
+{
+ hw->prev_evt = hw->cur_evt;
+ hw->cur_evt = evt;
+ CSIO_INC_STATS(hw, n_evt_sm[evt]);
+
+ switch (evt) {
+ case CSIO_HWE_RESUME:
+ csio_set_state(&hw->sm, csio_hws_configuring);
+ csio_hw_configure(hw);
+ break;
+
+ default:
+ CSIO_INC_STATS(hw, n_evt_unexp);
+ break;
+ }
+}
+
+/*
+ * csio_hws_resetting - HW Resetting state
+ * @hw - HW module
+ * @evt - Event
+ *
+ */
+static void
+csio_hws_resetting(struct csio_hw *hw, enum csio_hw_ev evt)
+{
+ hw->prev_evt = hw->cur_evt;
+ hw->cur_evt = evt;
+ CSIO_INC_STATS(hw, n_evt_sm[evt]);
+
+ switch (evt) {
+ case CSIO_HWE_HBA_RESET_DONE:
+ csio_evtq_start(hw);
+ csio_set_state(&hw->sm, csio_hws_configuring);
+ csio_hw_configure(hw);
+ break;
+
+ default:
+ CSIO_INC_STATS(hw, n_evt_unexp);
+ break;
+ }
+}
+
+/*
+ * csio_hws_removing - PCI Hotplug removing state
+ * @hw - HW module
+ * @evt - Event
+ *
+ */
+static void
+csio_hws_removing(struct csio_hw *hw, enum csio_hw_ev evt)
+{
+ hw->prev_evt = hw->cur_evt;
+ hw->cur_evt = evt;
+ CSIO_INC_STATS(hw, n_evt_sm[evt]);
+
+ switch (evt) {
+ case CSIO_HWE_HBA_RESET:
+ if (!csio_is_hw_master(hw))
+ break;
+ /*
+ * The BYE should have alerady been issued, so we cant
+ * use the mailbox interface. Hence we use the PL_RST
+ * register directly.
+ */
+ csio_err(hw, "Resetting HW and waiting 2 seconds...\n");
+ csio_wr_reg32(hw, PIORSTMODE | PIORST, PL_RST);
+ mdelay(2000);
+ break;
+
+ /* Should never receive any new events */
+ default:
+ CSIO_INC_STATS(hw, n_evt_unexp);
+ break;
+
+ }
+}
+
+/*
+ * csio_hws_pcierr - PCI Error state
+ * @hw - HW module
+ * @evt - Event
+ *
+ */
+static void
+csio_hws_pcierr(struct csio_hw *hw, enum csio_hw_ev evt)
+{
+ hw->prev_evt = hw->cur_evt;
+ hw->cur_evt = evt;
+ CSIO_INC_STATS(hw, n_evt_sm[evt]);
+
+ switch (evt) {
+ case CSIO_HWE_PCIERR_SLOT_RESET:
+ csio_evtq_start(hw);
+ csio_set_state(&hw->sm, csio_hws_configuring);
+ csio_hw_configure(hw);
+ break;
+
+ default:
+ CSIO_INC_STATS(hw, n_evt_unexp);
+ break;
+ }
+}
+
+/*****************************************************************************/
+/* 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
+ * @reg: the interrupt status register to process
+ * @acts: table of interrupt actions
+ *
+ * A table driven interrupt handler that applies a set of masks to an
+ * interrupt status word and performs the corresponding actions if the
+ * interrupts described by the mask have occured. The actions include
+ * optionally emitting a warning or alert message. The table is terminated
+ * by an entry specifying mask 0. Returns the number of fatal interrupt
+ * conditions.
+ */
+static int
+csio_handle_intr_status(struct csio_hw *hw, unsigned int reg,
+ const struct intr_info *acts)
+{
+ int fatal = 0;
+ unsigned int mask = 0;
+ unsigned int status = csio_rd_reg32(hw, reg);
+
+ for ( ; acts->mask; ++acts) {
+ if (!(status & acts->mask))
+ continue;
+ if (acts->fatal) {
+ fatal++;
+ csio_fatal(hw, "Fatal %s (0x%x)\n",
+ acts->msg, status & acts->mask);
+ } else if (acts->msg)
+ csio_info(hw, "%s (0x%x)\n",
+ acts->msg, status & acts->mask);
+ mask |= acts->mask;
+ }
+ status &= mask;
+ if (status) /* clear processed interrupts */
+ csio_wr_reg32(hw, status, reg);
+ return fatal;
+}
+
+/*
+ * 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)
+{
+ static struct intr_info tp_intr_info[] = {
+ { 0x3fffffff, "TP parity error", -1, 1 },
+ { FLMTXFLSTEMPTY, "TP out of Tx pages", -1, 1 },
+ { 0, NULL, 0, 0 }
+ };
+
+ if (csio_handle_intr_status(hw, TP_INT_CAUSE, tp_intr_info))
+ csio_hw_fatal_err(hw);
+}
+
+/*
+ * SGE interrupt handler.
+ */
+static void csio_sge_intr_handler(struct csio_hw *hw)
+{
+ uint64_t v;
+
+ static struct intr_info sge_intr_info[] = {
+ { ERR_CPL_EXCEED_IQE_SIZE,
+ "SGE received CPL exceeding IQE size", -1, 1 },
+ { ERR_INVALID_CIDX_INC,
+ "SGE GTS CIDX increment too large", -1, 0 },
+ { ERR_CPL_OPCODE_0, "SGE received 0-length CPL", -1, 0 },
+ { ERR_DROPPED_DB, "SGE doorbell dropped", -1, 0 },
+ { ERR_DATA_CPL_ON_HIGH_QID1 | ERR_DATA_CPL_ON_HIGH_QID0,
+ "SGE IQID > 1023 received CPL for FL", -1, 0 },
+ { ERR_BAD_DB_PIDX3, "SGE DBP 3 pidx increment too large", -1,
+ 0 },
+ { ERR_BAD_DB_PIDX2, "SGE DBP 2 pidx increment too large", -1,
+ 0 },
+ { ERR_BAD_DB_PIDX1, "SGE DBP 1 pidx increment too large", -1,
+ 0 },
+ { ERR_BAD_DB_PIDX0, "SGE DBP 0 pidx increment too large", -1,
+ 0 },
+ { ERR_ING_CTXT_PRIO,
+ "SGE too many priority ingress contexts", -1, 0 },
+ { ERR_EGR_CTXT_PRIO,
+ "SGE too many priority egress contexts", -1, 0 },
+ { INGRESS_SIZE_ERR, "SGE illegal ingress QID", -1, 0 },
+ { EGRESS_SIZE_ERR, "SGE illegal egress QID", -1, 0 },
+ { 0, NULL, 0, 0 }
+ };
+
+ v = (uint64_t)csio_rd_reg32(hw, SGE_INT_CAUSE1) |
+ ((uint64_t)csio_rd_reg32(hw, SGE_INT_CAUSE2) << 32);
+ if (v) {
+ csio_fatal(hw, "SGE parity error (%#llx)\n",
+ (unsigned long long)v);
+ csio_wr_reg32(hw, (uint32_t)(v & 0xFFFFFFFF),
+ SGE_INT_CAUSE1);
+ csio_wr_reg32(hw, (uint32_t)(v >> 32), SGE_INT_CAUSE2);
+ }
+
+ v |= csio_handle_intr_status(hw, SGE_INT_CAUSE3, sge_intr_info);
+
+ if (csio_handle_intr_status(hw, SGE_INT_CAUSE3, sge_intr_info) ||
+ v != 0)
+ csio_hw_fatal_err(hw);
+}
+
+#define CIM_OBQ_INTR (OBQULP0PARERR | OBQULP1PARERR | OBQULP2PARERR |\
+ OBQULP3PARERR | OBQSGEPARERR | OBQNCSIPARERR)
+#define CIM_IBQ_INTR (IBQTP0PARERR | IBQTP1PARERR | IBQULPPARERR |\
+ IBQSGEHIPARERR | IBQSGELOPARERR | IBQNCSIPARERR)
+
+/*
+ * CIM interrupt handler.
+ */
+static void csio_cim_intr_handler(struct csio_hw *hw)
+{
+ static struct intr_info cim_intr_info[] = {
+ { PREFDROPINT, "CIM control register prefetch drop", -1, 1 },
+ { CIM_OBQ_INTR, "CIM OBQ parity error", -1, 1 },
+ { CIM_IBQ_INTR, "CIM IBQ parity error", -1, 1 },
+ { MBUPPARERR, "CIM mailbox uP parity error", -1, 1 },
+ { MBHOSTPARERR, "CIM mailbox host parity error", -1, 1 },
+ { TIEQINPARERRINT, "CIM TIEQ outgoing parity error", -1, 1 },
+ { TIEQOUTPARERRINT, "CIM TIEQ incoming parity error", -1, 1 },
+ { 0, NULL, 0, 0 }
+ };
+ static struct intr_info cim_upintr_info[] = {
+ { RSVDSPACEINT, "CIM reserved space access", -1, 1 },
+ { ILLTRANSINT, "CIM illegal transaction", -1, 1 },
+ { ILLWRINT, "CIM illegal write", -1, 1 },
+ { ILLRDINT, "CIM illegal read", -1, 1 },
+ { ILLRDBEINT, "CIM illegal read BE", -1, 1 },
+ { ILLWRBEINT, "CIM illegal write BE", -1, 1 },
+ { SGLRDBOOTINT, "CIM single read from boot space", -1, 1 },
+ { SGLWRBOOTINT, "CIM single write to boot space", -1, 1 },
+ { BLKWRBOOTINT, "CIM block write to boot space", -1, 1 },
+ { SGLRDFLASHINT, "CIM single read from flash space", -1, 1 },
+ { SGLWRFLASHINT, "CIM single write to flash space", -1, 1 },
+ { BLKWRFLASHINT, "CIM block write to flash space", -1, 1 },
+ { SGLRDEEPROMINT, "CIM single EEPROM read", -1, 1 },
+ { SGLWREEPROMINT, "CIM single EEPROM write", -1, 1 },
+ { BLKRDEEPROMINT, "CIM block EEPROM read", -1, 1 },
+ { BLKWREEPROMINT, "CIM block EEPROM write", -1, 1 },
+ { SGLRDCTLINT , "CIM single read from CTL space", -1, 1 },
+ { SGLWRCTLINT , "CIM single write to CTL space", -1, 1 },
+ { BLKRDCTLINT , "CIM block read from CTL space", -1, 1 },
+ { BLKWRCTLINT , "CIM block write to CTL space", -1, 1 },
+ { SGLRDPLINT , "CIM single read from PL space", -1, 1 },
+ { SGLWRPLINT , "CIM single write to PL space", -1, 1 },
+ { BLKRDPLINT , "CIM block read from PL space", -1, 1 },
+ { BLKWRPLINT , "CIM block write to PL space", -1, 1 },
+ { REQOVRLOOKUPINT , "CIM request FIFO overwrite", -1, 1 },
+ { RSPOVRLOOKUPINT , "CIM response FIFO overwrite", -1, 1 },
+ { TIMEOUTINT , "CIM PIF timeout", -1, 1 },
+ { TIMEOUTMAINT , "CIM PIF MA timeout", -1, 1 },
+ { 0, NULL, 0, 0 }
+ };
+
+ int fat;
+
+ fat = csio_handle_intr_status(hw, CIM_HOST_INT_CAUSE,
+ cim_intr_info) +
+ csio_handle_intr_status(hw, CIM_HOST_UPACC_INT_CAUSE,
+ cim_upintr_info);
+ if (fat)
+ csio_hw_fatal_err(hw);
+}
+
+/*
+ * ULP RX interrupt handler.
+ */
+static void csio_ulprx_intr_handler(struct csio_hw *hw)
+{
+ static struct intr_info ulprx_intr_info[] = {
+ { 0x1800000, "ULPRX context error", -1, 1 },
+ { 0x7fffff, "ULPRX parity error", -1, 1 },
+ { 0, NULL, 0, 0 }
+ };
+
+ if (csio_handle_intr_status(hw, ULP_RX_INT_CAUSE, ulprx_intr_info))
+ csio_hw_fatal_err(hw);
+}
+
+/*
+ * ULP TX interrupt handler.
+ */
+static void csio_ulptx_intr_handler(struct csio_hw *hw)
+{
+ static struct intr_info ulptx_intr_info[] = {
+ { PBL_BOUND_ERR_CH3, "ULPTX channel 3 PBL out of bounds", -1,
+ 0 },
+ { PBL_BOUND_ERR_CH2, "ULPTX channel 2 PBL out of bounds", -1,
+ 0 },
+ { PBL_BOUND_ERR_CH1, "ULPTX channel 1 PBL out of bounds", -1,
+ 0 },
+ { PBL_BOUND_ERR_CH0, "ULPTX channel 0 PBL out of bounds", -1,
+ 0 },
+ { 0xfffffff, "ULPTX parity error", -1, 1 },
+ { 0, NULL, 0, 0 }
+ };
+
+ if (csio_handle_intr_status(hw, ULP_TX_INT_CAUSE, ulptx_intr_info))
+ csio_hw_fatal_err(hw);
+}
+
+/*
+ * PM TX interrupt handler.
+ */
+static void csio_pmtx_intr_handler(struct csio_hw *hw)
+{
+ static struct intr_info pmtx_intr_info[] = {
+ { PCMD_LEN_OVFL0, "PMTX channel 0 pcmd too large", -1, 1 },
+ { PCMD_LEN_OVFL1, "PMTX channel 1 pcmd too large", -1, 1 },
+ { PCMD_LEN_OVFL2, "PMTX channel 2 pcmd too large", -1, 1 },
+ { ZERO_C_CMD_ERROR, "PMTX 0-length pcmd", -1, 1 },
+ { 0xffffff0, "PMTX framing error", -1, 1 },
+ { OESPI_PAR_ERROR, "PMTX oespi parity error", -1, 1 },
+ { DB_OPTIONS_PAR_ERROR, "PMTX db_options parity error", -1,
+ 1 },
+ { ICSPI_PAR_ERROR, "PMTX icspi parity error", -1, 1 },
+ { C_PCMD_PAR_ERROR, "PMTX c_pcmd parity error", -1, 1},
+ { 0, NULL, 0, 0 }
+ };
+
+ if (csio_handle_intr_status(hw, PM_TX_INT_CAUSE, pmtx_intr_info))
+ csio_hw_fatal_err(hw);
+}
+
+/*
+ * PM RX interrupt handler.
+ */
+static void csio_pmrx_intr_handler(struct csio_hw *hw)
+{
+ static struct intr_info pmrx_intr_info[] = {
+ { ZERO_E_CMD_ERROR, "PMRX 0-length pcmd", -1, 1 },
+ { 0x3ffff0, "PMRX framing error", -1, 1 },
+ { OCSPI_PAR_ERROR, "PMRX ocspi parity error", -1, 1 },
+ { DB_OPTIONS_PAR_ERROR, "PMRX db_options parity error", -1,
+ 1 },
+ { IESPI_PAR_ERROR, "PMRX iespi parity error", -1, 1 },
+ { E_PCMD_PAR_ERROR, "PMRX e_pcmd parity error", -1, 1},
+ { 0, NULL, 0, 0 }
+ };
+
+ if (csio_handle_intr_status(hw, PM_RX_INT_CAUSE, pmrx_intr_info))
+ csio_hw_fatal_err(hw);
+}
+
+/*
+ * CPL switch interrupt handler.
+ */
+static void csio_cplsw_intr_handler(struct csio_hw *hw)
+{
+ static struct intr_info cplsw_intr_info[] = {
+ { CIM_OP_MAP_PERR, "CPLSW CIM op_map parity error", -1, 1 },
+ { CIM_OVFL_ERROR, "CPLSW CIM overflow", -1, 1 },
+ { TP_FRAMING_ERROR, "CPLSW TP framing error", -1, 1 },
+ { SGE_FRAMING_ERROR, "CPLSW SGE framing error", -1, 1 },
+ { CIM_FRAMING_ERROR, "CPLSW CIM framing error", -1, 1 },
+ { ZERO_SWITCH_ERROR, "CPLSW no-switch error", -1, 1 },
+ { 0, NULL, 0, 0 }
+ };
+
+ if (csio_handle_intr_status(hw, CPL_INTR_CAUSE, cplsw_intr_info))
+ csio_hw_fatal_err(hw);
+}
+
+/*
+ * LE interrupt handler.
+ */
+static void csio_le_intr_handler(struct csio_hw *hw)
+{
+ static struct intr_info le_intr_info[] = {
+ { LIPMISS, "LE LIP miss", -1, 0 },
+ { LIP0, "LE 0 LIP error", -1, 0 },
+ { PARITYERR, "LE parity error", -1, 1 },
+ { UNKNOWNCMD, "LE unknown command", -1, 1 },
+ { REQQPARERR, "LE request queue parity error", -1, 1 },
+ { 0, NULL, 0, 0 }
+ };
+
+ if (csio_handle_intr_status(hw, LE_DB_INT_CAUSE, le_intr_info))
+ csio_hw_fatal_err(hw);
+}
+
+/*
+ * MPS interrupt handler.
+ */
+static void csio_mps_intr_handler(struct csio_hw *hw)
+{
+ static struct intr_info mps_rx_intr_info[] = {
+ { 0xffffff, "MPS Rx parity error", -1, 1 },
+ { 0, NULL, 0, 0 }
+ };
+ static struct intr_info mps_tx_intr_info[] = {
+ { TPFIFO, "MPS Tx TP FIFO parity error", -1, 1 },
+ { NCSIFIFO, "MPS Tx NC-SI FIFO parity error", -1, 1 },
+ { TXDATAFIFO, "MPS Tx data FIFO parity error", -1, 1 },
+ { TXDESCFIFO, "MPS Tx desc FIFO parity error", -1, 1 },
+ { BUBBLE, "MPS Tx underflow", -1, 1 },
+ { SECNTERR, "MPS Tx SOP/EOP error", -1, 1 },
+ { FRMERR, "MPS Tx framing error", -1, 1 },
+ { 0, NULL, 0, 0 }
+ };
+ static struct intr_info mps_trc_intr_info[] = {
+ { FILTMEM, "MPS TRC filter parity error", -1, 1 },
+ { PKTFIFO, "MPS TRC packet FIFO parity error", -1, 1 },
+ { MISCPERR, "MPS TRC misc parity error", -1, 1 },
+ { 0, NULL, 0, 0 }
+ };
+ static struct intr_info mps_stat_sram_intr_info[] = {
+ { 0x1fffff, "MPS statistics SRAM parity error", -1, 1 },
+ { 0, NULL, 0, 0 }
+ };
+ static struct intr_info mps_stat_tx_intr_info[] = {
+ { 0xfffff, "MPS statistics Tx FIFO parity error", -1, 1 },
+ { 0, NULL, 0, 0 }
+ };
+ static struct intr_info mps_stat_rx_intr_info[] = {
+ { 0xffffff, "MPS statistics Rx FIFO parity error", -1, 1 },
+ { 0, NULL, 0, 0 }
+ };
+ static struct intr_info mps_cls_intr_info[] = {
+ { MATCHSRAM, "MPS match SRAM parity error", -1, 1 },
+ { MATCHTCAM, "MPS match TCAM parity error", -1, 1 },
+ { HASHSRAM, "MPS hash SRAM parity error", -1, 1 },
+ { 0, NULL, 0, 0 }
+ };
+
+ int fat;
+
+ fat = csio_handle_intr_status(hw, MPS_RX_PERR_INT_CAUSE,
+ mps_rx_intr_info) +
+ csio_handle_intr_status(hw, MPS_TX_INT_CAUSE,
+ mps_tx_intr_info) +
+ csio_handle_intr_status(hw, MPS_TRC_INT_CAUSE,
+ mps_trc_intr_info) +
+ csio_handle_intr_status(hw, MPS_STAT_PERR_INT_CAUSE_SRAM,
+ mps_stat_sram_intr_info) +
+ csio_handle_intr_status(hw, MPS_STAT_PERR_INT_CAUSE_TX_FIFO,
+ mps_stat_tx_intr_info) +
+ csio_handle_intr_status(hw, MPS_STAT_PERR_INT_CAUSE_RX_FIFO,
+ mps_stat_rx_intr_info) +
+ csio_handle_intr_status(hw, MPS_CLS_INT_CAUSE,
+ mps_cls_intr_info);
+
+ csio_wr_reg32(hw, 0, MPS_INT_CAUSE);
+ csio_rd_reg32(hw, MPS_INT_CAUSE); /* flush */
+ if (fat)
+ csio_hw_fatal_err(hw);
+}
+
+#define MEM_INT_MASK (PERR_INT_CAUSE | ECC_CE_INT_CAUSE | ECC_UE_INT_CAUSE)
+
+/*
+ * EDC/MC interrupt handler.
+ */
+static void csio_mem_intr_handler(struct csio_hw *hw, int idx)
+{
+ static const char name[3][5] = { "EDC0", "EDC1", "MC" };
+
+ unsigned int addr, cnt_addr, v;
+
+ if (idx <= MEM_EDC1) {
+ addr = EDC_REG(EDC_INT_CAUSE, idx);
+ cnt_addr = EDC_REG(EDC_ECC_STATUS, idx);
+ } else {
+ addr = MC_INT_CAUSE;
+ cnt_addr = MC_ECC_STATUS;
+ }
+
+ v = csio_rd_reg32(hw, addr) & MEM_INT_MASK;
+ if (v & PERR_INT_CAUSE)
+ csio_fatal(hw, "%s FIFO parity error\n", name[idx]);
+ if (v & ECC_CE_INT_CAUSE) {
+ uint32_t cnt = ECC_CECNT_GET(csio_rd_reg32(hw, cnt_addr));
+
+ csio_wr_reg32(hw, ECC_CECNT_MASK, cnt_addr);
+ csio_warn(hw, "%u %s correctable ECC data error%s\n",
+ cnt, name[idx], cnt > 1 ? "s" : "");
+ }
+ if (v & ECC_UE_INT_CAUSE)
+ csio_fatal(hw, "%s uncorrectable ECC data error\n", name[idx]);
+
+ csio_wr_reg32(hw, v, addr);
+ if (v & (PERR_INT_CAUSE | ECC_UE_INT_CAUSE))
+ csio_hw_fatal_err(hw);
+}
+
+/*
+ * MA interrupt handler.
+ */
+static void csio_ma_intr_handler(struct csio_hw *hw)
+{
+ uint32_t v, status = csio_rd_reg32(hw, MA_INT_CAUSE);
+
+ if (status & MEM_PERR_INT_CAUSE)
+ csio_fatal(hw, "MA parity error, parity status %#x\n",
+ csio_rd_reg32(hw, MA_PARITY_ERROR_STATUS));
+ if (status & MEM_WRAP_INT_CAUSE) {
+ v = csio_rd_reg32(hw, MA_INT_WRAP_STATUS);
+ csio_fatal(hw,
+ "MA address wrap-around error by client %u to address %#x\n",
+ MEM_WRAP_CLIENT_NUM_GET(v), MEM_WRAP_ADDRESS_GET(v) << 4);
+ }
+ csio_wr_reg32(hw, status, MA_INT_CAUSE);
+ csio_hw_fatal_err(hw);
+}
+
+/*
+ * SMB interrupt handler.
+ */
+static void csio_smb_intr_handler(struct csio_hw *hw)
+{
+ static struct intr_info smb_intr_info[] = {
+ { MSTTXFIFOPARINT, "SMB master Tx FIFO parity error", -1, 1 },
+ { MSTRXFIFOPARINT, "SMB master Rx FIFO parity error", -1, 1 },
+ { SLVFIFOPARINT, "SMB slave FIFO parity error", -1, 1 },
+ { 0, NULL, 0, 0 }
+ };
+
+ if (csio_handle_intr_status(hw, SMB_INT_CAUSE, smb_intr_info))
+ csio_hw_fatal_err(hw);
+}
+
+/*
+ * NC-SI interrupt handler.
+ */
+static void csio_ncsi_intr_handler(struct csio_hw *hw)
+{
+ static struct intr_info ncsi_intr_info[] = {
+ { CIM_DM_PRTY_ERR, "NC-SI CIM parity error", -1, 1 },
+ { MPS_DM_PRTY_ERR, "NC-SI MPS parity error", -1, 1 },
+ { TXFIFO_PRTY_ERR, "NC-SI Tx FIFO parity error", -1, 1 },
+ { RXFIFO_PRTY_ERR, "NC-SI Rx FIFO parity error", -1, 1 },
+ { 0, NULL, 0, 0 }
+ };
+
+ if (csio_handle_intr_status(hw, NCSI_INT_CAUSE, ncsi_intr_info))
+ csio_hw_fatal_err(hw);
+}
+
+/*
+ * XGMAC interrupt handler.
+ */
+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));
+
+ v &= TXFIFO_PRTY_ERR | RXFIFO_PRTY_ERR;
+ if (!v)
+ return;
+
+ if (v & TXFIFO_PRTY_ERR)
+ 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_hw_fatal_err(hw);
+}
+
+/*
+ * PL interrupt handler.
+ */
+static void csio_pl_intr_handler(struct csio_hw *hw)
+{
+ static struct intr_info pl_intr_info[] = {
+ { FATALPERR, "T4 fatal parity error", -1, 1 },
+ { PERRVFID, "PL VFID_MAP parity error", -1, 1 },
+ { 0, NULL, 0, 0 }
+ };
+
+ if (csio_handle_intr_status(hw, PL_PL_INT_CAUSE, pl_intr_info))
+ csio_hw_fatal_err(hw);
+}
+
+/*
+ * csio_hw_slow_intr_handler - control path interrupt handler
+ * @hw: HW module
+ *
+ * Interrupt handler for non-data global interrupt events, e.g., errors.
+ * The designation 'slow' is because it involves register reads, while
+ * data interrupts typically don't involve any MMIOs.
+ */
+int
+csio_hw_slow_intr_handler(struct csio_hw *hw)
+{
+ uint32_t cause = csio_rd_reg32(hw, PL_INT_CAUSE);
+
+ if (!(cause & CSIO_GLBL_INTR_MASK)) {
+ CSIO_INC_STATS(hw, n_plint_unexp);
+ return 0;
+ }
+
+ csio_dbg(hw, "Slow interrupt! cause: 0x%x\n", cause);
+
+ CSIO_INC_STATS(hw, n_plint_cnt);
+
+ if (cause & CIM)
+ csio_cim_intr_handler(hw);
+
+ if (cause & MPS)
+ csio_mps_intr_handler(hw);
+
+ if (cause & NCSI)
+ csio_ncsi_intr_handler(hw);
+
+ if (cause & PL)
+ csio_pl_intr_handler(hw);
+
+ if (cause & SMB)
+ csio_smb_intr_handler(hw);
+
+ if (cause & XGMAC0)
+ csio_xgmac_intr_handler(hw, 0);
+
+ if (cause & XGMAC1)
+ csio_xgmac_intr_handler(hw, 1);
+
+ if (cause & XGMAC_KR0)
+ csio_xgmac_intr_handler(hw, 2);
+
+ if (cause & XGMAC_KR1)
+ csio_xgmac_intr_handler(hw, 3);
+
+ if (cause & PCIE)
+ csio_pcie_intr_handler(hw);
+
+ if (cause & MC)
+ csio_mem_intr_handler(hw, MEM_MC);
+
+ if (cause & EDC0)
+ csio_mem_intr_handler(hw, MEM_EDC0);
+
+ if (cause & EDC1)
+ csio_mem_intr_handler(hw, MEM_EDC1);
+
+ if (cause & LE)
+ csio_le_intr_handler(hw);
+
+ if (cause & TP)
+ csio_tp_intr_handler(hw);
+
+ if (cause & MA)
+ csio_ma_intr_handler(hw);
+
+ if (cause & PM_TX)
+ csio_pmtx_intr_handler(hw);
+
+ if (cause & PM_RX)
+ csio_pmrx_intr_handler(hw);
+
+ if (cause & ULP_RX)
+ csio_ulprx_intr_handler(hw);
+
+ if (cause & CPL_SWITCH)
+ csio_cplsw_intr_handler(hw);
+
+ if (cause & SGE)
+ csio_sge_intr_handler(hw);
+
+ if (cause & ULP_TX)
+ csio_ulptx_intr_handler(hw);
+
+ /* Clear the interrupts just processed for which we are the master. */
+ csio_wr_reg32(hw, cause & CSIO_GLBL_INTR_MASK, PL_INT_CAUSE);
+ csio_rd_reg32(hw, PL_INT_CAUSE); /* flush */
+
+ return 1;
+}
+
+/*****************************************************************************
+ * HW <--> mailbox interfacing routines.
+ ****************************************************************************/
+/*
+ * csio_mberr_worker - Worker thread (dpc) for mailbox/error completions
+ *
+ * @data: Private data pointer.
+ *
+ * Called from worker thread context.
+ */
+static void
+csio_mberr_worker(void *data)
+{
+ struct csio_hw *hw = (struct csio_hw *)data;
+ struct csio_mbm *mbm = &hw->mbm;
+ LIST_HEAD(cbfn_q);
+ struct csio_mb *mbp_next;
+ int rv;
+
+ del_timer_sync(&mbm->timer);
+
+ spin_lock_irq(&hw->lock);
+ if (list_empty(&mbm->cbfn_q)) {
+ spin_unlock_irq(&hw->lock);
+ return;
+ }
+
+ list_splice_tail_init(&mbm->cbfn_q, &cbfn_q);
+ mbm->stats.n_cbfnq = 0;
+
+ /* Try to start waiting mailboxes */
+ if (!list_empty(&mbm->req_q)) {
+ mbp_next = list_first_entry(&mbm->req_q, struct csio_mb, list);
+ list_del_init(&mbp_next->list);
+
+ rv = csio_mb_issue(hw, mbp_next);
+ if (rv != 0)
+ list_add_tail(&mbp_next->list, &mbm->req_q);
+ else
+ CSIO_DEC_STATS(mbm, n_activeq);
+ }
+ spin_unlock_irq(&hw->lock);
+
+ /* Now callback completions */
+ csio_mb_completions(hw, &cbfn_q);
+}
+
+/*
+ * csio_hw_mb_timer - Top-level Mailbox timeout handler.
+ *
+ * @data: private data pointer
+ *
+ **/
+static void
+csio_hw_mb_timer(uintptr_t data)
+{
+ struct csio_hw *hw = (struct csio_hw *)data;
+ struct csio_mb *mbp = NULL;
+
+ spin_lock_irq(&hw->lock);
+ mbp = csio_mb_tmo_handler(hw);
+ spin_unlock_irq(&hw->lock);
+
+ /* Call back the function for the timed-out Mailbox */
+ if (mbp)
+ mbp->mb_cbfn(hw, mbp);
+
+}
+
+/*
+ * csio_hw_mbm_cleanup - Cleanup Mailbox module.
+ * @hw: HW module
+ *
+ * Called with lock held, should exit with lock held.
+ * Cancels outstanding mailboxes (waiting, in-flight) and gathers them
+ * into a local queue. Drops lock and calls the completions. Holds
+ * lock and returns.
+ */
+static void
+csio_hw_mbm_cleanup(struct csio_hw *hw)
+{
+ LIST_HEAD(cbfn_q);
+
+ csio_mb_cancel_all(hw, &cbfn_q);
+
+ spin_unlock_irq(&hw->lock);
+ csio_mb_completions(hw, &cbfn_q);
+ spin_lock_irq(&hw->lock);
+}
+
+/*****************************************************************************
+ * Event handling
+ ****************************************************************************/
+int
+csio_enqueue_evt(struct csio_hw *hw, enum csio_evt type, void *evt_msg,
+ uint16_t len)
+{
+ struct csio_evt_msg *evt_entry = NULL;
+
+ if (type >= CSIO_EVT_MAX)
+ return -EINVAL;
+
+ if (len > CSIO_EVT_MSG_SIZE)
+ return -EINVAL;
+
+ if (hw->flags & CSIO_HWF_FWEVT_STOP)
+ return -EINVAL;
+
+ if (list_empty(&hw->evt_free_q)) {
+ csio_err(hw, "Failed to alloc evt entry, msg type %d len %d\n",
+ type, len);
+ return -ENOMEM;
+ }
+
+ evt_entry = list_first_entry(&hw->evt_free_q,
+ struct csio_evt_msg, list);
+ list_del_init(&evt_entry->list);
+
+ /* copy event msg and queue the event */
+ evt_entry->type = type;
+ memcpy((void *)evt_entry->data, evt_msg, len);
+ list_add_tail(&evt_entry->list, &hw->evt_active_q);
+
+ CSIO_DEC_STATS(hw, n_evt_freeq);
+ CSIO_INC_STATS(hw, n_evt_activeq);
+
+ return 0;
+}
+
+static int
+csio_enqueue_evt_lock(struct csio_hw *hw, enum csio_evt type, void *evt_msg,
+ uint16_t len, bool msg_sg)
+{
+ struct csio_evt_msg *evt_entry = NULL;
+ struct csio_fl_dma_buf *fl_sg;
+ uint32_t off = 0;
+ unsigned long flags;
+ int n;
+
+ if (type >= CSIO_EVT_MAX)
+ return -EINVAL;
+
+ if (len > CSIO_EVT_MSG_SIZE)
+ return -EINVAL;
+
+ spin_lock_irqsave(&hw->lock, flags);
+ if (hw->flags & CSIO_HWF_FWEVT_STOP) {
+ spin_unlock_irqrestore(&hw->lock, flags);
+ return -EINVAL;
+ }
+
+ if (list_empty(&hw->evt_free_q)) {
+ csio_err(hw, "Failed to alloc evt entry, msg type %d len %d\n",
+ type, len);
+ spin_unlock_irqrestore(&hw->lock, flags);
+ return -ENOMEM;
+ }
+
+ evt_entry = list_first_entry(&hw->evt_free_q,
+ struct csio_evt_msg, list);
+ list_del_init(&evt_entry->list);
+
+ /* copy event msg and queue the event */
+ evt_entry->type = type;
+
+ /* If Payload in SG list*/
+ if (msg_sg) {
+ fl_sg = (struct csio_fl_dma_buf *) evt_msg;
+ for (n = 0; (n < CSIO_MAX_FLBUF_PER_IQWR && off < len); n++) {
+ memcpy((void *)((uintptr_t)evt_entry->data + off),
+ fl_sg->flbufs[n].vaddr,
+ fl_sg->flbufs[n].len);
+ off += fl_sg->flbufs[n].len;
+ }
+ } else
+ memcpy((void *)evt_entry->data, evt_msg, len);
+
+ list_add_tail(&evt_entry->list, &hw->evt_active_q);
+ spin_unlock_irqrestore(&hw->lock, flags);
+
+ CSIO_DEC_STATS(hw, n_evt_freeq);
+ CSIO_INC_STATS(hw, n_evt_activeq);
+
+ return 0;
+}
+
+static void
+csio_free_evt(struct csio_hw *hw, struct csio_evt_msg *evt_entry)
+{
+ if (evt_entry) {
+ spin_lock_irq(&hw->lock);
+ list_del_init(&evt_entry->list);
+ list_add_tail(&evt_entry->list, &hw->evt_free_q);
+ CSIO_DEC_STATS(hw, n_evt_activeq);
+ CSIO_INC_STATS(hw, n_evt_freeq);
+ spin_unlock_irq(&hw->lock);
+ }
+}
+
+void
+csio_evtq_flush(struct csio_hw *hw)
+{
+ uint32_t count;
+ count = 30;
+ while (hw->flags & CSIO_HWF_FWEVT_PENDING && count--) {
+ spin_unlock_irq(&hw->lock);
+ msleep(2000);
+ spin_lock_irq(&hw->lock);
+ }
+
+ CSIO_DB_ASSERT(!(hw->flags & CSIO_HWF_FWEVT_PENDING));
+}
+
+static void
+csio_evtq_stop(struct csio_hw *hw)
+{
+ hw->flags |= CSIO_HWF_FWEVT_STOP;
+}
+
+static void
+csio_evtq_start(struct csio_hw *hw)
+{
+ hw->flags &= ~CSIO_HWF_FWEVT_STOP;
+}
+
+static void
+csio_evtq_cleanup(struct csio_hw *hw)
+{
+ struct list_head *evt_entry, *next_entry;
+
+ /* Release outstanding events from activeq to freeq*/
+ if (!list_empty(&hw->evt_active_q))
+ list_splice_tail_init(&hw->evt_active_q, &hw->evt_free_q);
+
+ hw->stats.n_evt_activeq = 0;
+ hw->flags &= ~CSIO_HWF_FWEVT_PENDING;
+
+ /* Freeup event entry */
+ list_for_each_safe(evt_entry, next_entry, &hw->evt_free_q) {
+ kfree(evt_entry);
+ CSIO_DEC_STATS(hw, n_evt_freeq);
+ }
+
+ hw->stats.n_evt_freeq = 0;
+}
+
+
+static void
+csio_process_fwevtq_entry(struct csio_hw *hw, void *wr, uint32_t len,
+ struct csio_fl_dma_buf *flb, void *priv)
+{
+ __u8 op;
+ __be64 *data;
+ void *msg = NULL;
+ uint32_t msg_len = 0;
+ bool msg_sg = 0;
+
+ op = ((struct rss_header *) wr)->opcode;
+ if (op == CPL_FW6_PLD) {
+ CSIO_INC_STATS(hw, n_cpl_fw6_pld);
+ if (!flb || !flb->totlen) {
+ CSIO_INC_STATS(hw, n_cpl_unexp);
+ return;
+ }
+
+ msg = (void *) flb;
+ msg_len = flb->totlen;
+ msg_sg = 1;
+
+ data = (__be64 *) msg;
+ } else if (op == CPL_FW6_MSG || op == CPL_FW4_MSG) {
+
+ CSIO_INC_STATS(hw, n_cpl_fw6_msg);
+ /* skip RSS header */
+ msg = (void *)((uintptr_t)wr + sizeof(__be64));
+ msg_len = (op == CPL_FW6_MSG) ? sizeof(struct cpl_fw6_msg) :
+ sizeof(struct cpl_fw4_msg);
+
+ data = (__be64 *) msg;
+ } else {
+ csio_warn(hw, "unexpected CPL %#x on FW event queue\n", op);
+ CSIO_INC_STATS(hw, n_cpl_unexp);
+ return;
+ }
+
+ /*
+ * Enqueue event to EventQ. Events processing happens
+ * in Event worker thread context
+ */
+ if (csio_enqueue_evt_lock(hw, CSIO_EVT_FW, msg,
+ (uint16_t)msg_len, msg_sg))
+ CSIO_INC_STATS(hw, n_evt_drop);
+}
+
+void
+csio_evtq_worker(struct work_struct *work)
+{
+ struct csio_hw *hw = container_of(work, struct csio_hw, evtq_work);
+ struct list_head *evt_entry, *next_entry;
+ LIST_HEAD(evt_q);
+ struct csio_evt_msg *evt_msg;
+ struct cpl_fw6_msg *msg;
+ struct csio_rnode *rn;
+ int rv = 0;
+ uint8_t evtq_stop = 0;
+
+ csio_dbg(hw, "event worker thread active evts#%d\n",
+ hw->stats.n_evt_activeq);
+
+ spin_lock_irq(&hw->lock);
+ while (!list_empty(&hw->evt_active_q)) {
+ list_splice_tail_init(&hw->evt_active_q, &evt_q);
+ spin_unlock_irq(&hw->lock);
+
+ list_for_each_safe(evt_entry, next_entry, &evt_q) {
+ evt_msg = (struct csio_evt_msg *) evt_entry;
+
+ /* Drop events if queue is STOPPED */
+ spin_lock_irq(&hw->lock);
+ if (hw->flags & CSIO_HWF_FWEVT_STOP)
+ evtq_stop = 1;
+ spin_unlock_irq(&hw->lock);
+ if (evtq_stop) {
+ CSIO_INC_STATS(hw, n_evt_drop);
+ goto free_evt;
+ }
+
+ switch (evt_msg->type) {
+ case CSIO_EVT_FW:
+ msg = (struct cpl_fw6_msg *)(evt_msg->data);
+
+ if ((msg->opcode == CPL_FW6_MSG ||
+ msg->opcode == CPL_FW4_MSG) &&
+ !msg->type) {
+ rv = csio_mb_fwevt_handler(hw,
+ msg->data);
+ if (!rv)
+ break;
+ /* Handle any remaining fw events */
+ csio_fcoe_fwevt_handler(hw,
+ msg->opcode, msg->data);
+ } else if (msg->opcode == CPL_FW6_PLD) {
+
+ csio_fcoe_fwevt_handler(hw,
+ msg->opcode, msg->data);
+ } else {
+ csio_warn(hw,
+ "Unhandled FW msg op %x type %x\n",
+ msg->opcode, msg->type);
+ CSIO_INC_STATS(hw, n_evt_drop);
+ }
+ break;
+
+ case CSIO_EVT_MBX:
+ csio_mberr_worker(hw);
+ break;
+
+ case CSIO_EVT_DEV_LOSS:
+ memcpy(&rn, evt_msg->data, sizeof(rn));
+ csio_rnode_devloss_handler(rn);
+ break;
+
+ default:
+ csio_warn(hw, "Unhandled event %x on evtq\n",
+ evt_msg->type);
+ CSIO_INC_STATS(hw, n_evt_unexp);
+ break;
+ }
+free_evt:
+ csio_free_evt(hw, evt_msg);
+ }
+
+ spin_lock_irq(&hw->lock);
+ }
+ hw->flags &= ~CSIO_HWF_FWEVT_PENDING;
+ spin_unlock_irq(&hw->lock);
+}
+
+int
+csio_fwevtq_handler(struct csio_hw *hw)
+{
+ int rv;
+
+ if (csio_q_iqid(hw, hw->fwevt_iq_idx) == CSIO_MAX_QID) {
+ CSIO_INC_STATS(hw, n_int_stray);
+ return -EINVAL;
+ }
+
+ rv = csio_wr_process_iq_idx(hw, hw->fwevt_iq_idx,
+ csio_process_fwevtq_entry, NULL);
+ return rv;
+}
+
+/****************************************************************************
+ * Entry points
+ ****************************************************************************/
+
+/* Management module */
+/*
+ * csio_mgmt_req_lookup - Lookup the given IO req exist in Active Q.
+ * mgmt - mgmt module
+ * @io_req - io request
+ *
+ * Return - 0:if given IO Req exists in active Q.
+ * -EINVAL :if lookup fails.
+ */
+int
+csio_mgmt_req_lookup(struct csio_mgmtm *mgmtm, struct csio_ioreq *io_req)
+{
+ struct list_head *tmp;
+
+ /* Lookup ioreq in the ACTIVEQ */
+ list_for_each(tmp, &mgmtm->active_q) {
+ if (io_req == (struct csio_ioreq *)tmp)
+ return 0;
+ }
+ return -EINVAL;
+}
+
+#define ECM_MIN_TMO 1000 /* Minimum timeout value for req */
+
+/*
+ * csio_mgmts_tmo_handler - MGMT IO Timeout handler.
+ * @data - Event data.
+ *
+ * Return - none.
+ */
+static void
+csio_mgmt_tmo_handler(uintptr_t data)
+{
+ struct csio_mgmtm *mgmtm = (struct csio_mgmtm *) data;
+ struct list_head *tmp;
+ struct csio_ioreq *io_req;
+
+ csio_dbg(mgmtm->hw, "Mgmt timer invoked!\n");
+
+ spin_lock_irq(&mgmtm->hw->lock);
+
+ list_for_each(tmp, &mgmtm->active_q) {
+ io_req = (struct csio_ioreq *) tmp;
+ io_req->tmo -= min_t(uint32_t, io_req->tmo, ECM_MIN_TMO);
+
+ if (!io_req->tmo) {
+ /* Dequeue the request from retry Q. */
+ tmp = csio_list_prev(tmp);
+ list_del_init(&io_req->sm.sm_list);
+ if (io_req->io_cbfn) {
+ /* io_req will be freed by completion handler */
+ io_req->wr_status = -ETIMEDOUT;
+ io_req->io_cbfn(mgmtm->hw, io_req);
+ } else {
+ CSIO_DB_ASSERT(0);
+ }
+ }
+ }
+
+ /* If retry queue is not empty, re-arm timer */
+ if (!list_empty(&mgmtm->active_q))
+ mod_timer(&mgmtm->mgmt_timer,
+ jiffies + msecs_to_jiffies(ECM_MIN_TMO));
+ spin_unlock_irq(&mgmtm->hw->lock);
+}
+
+static void
+csio_mgmtm_cleanup(struct csio_mgmtm *mgmtm)
+{
+ struct csio_hw *hw = mgmtm->hw;
+ struct csio_ioreq *io_req;
+ struct list_head *tmp;
+ uint32_t count;
+
+ count = 30;
+ /* Wait for all outstanding req to complete gracefully */
+ while ((!list_empty(&mgmtm->active_q)) && count--) {
+ spin_unlock_irq(&hw->lock);
+ msleep(2000);
+ spin_lock_irq(&hw->lock);
+ }
+
+ /* release outstanding req from ACTIVEQ */
+ list_for_each(tmp, &mgmtm->active_q) {
+ io_req = (struct csio_ioreq *) tmp;
+ tmp = csio_list_prev(tmp);
+ list_del_init(&io_req->sm.sm_list);
+ mgmtm->stats.n_active--;
+ if (io_req->io_cbfn) {
+ /* io_req will be freed by completion handler */
+ io_req->wr_status = -ETIMEDOUT;
+ io_req->io_cbfn(mgmtm->hw, io_req);
+ }
+ }
+}
+
+/*
+ * csio_mgmt_init - Mgmt module init entry point
+ * @mgmtsm - mgmt module
+ * @hw - HW module
+ *
+ * Initialize mgmt timer, resource wait queue, active queue,
+ * completion q. Allocate Egress and Ingress
+ * WR queues and save off the queue index returned by the WR
+ * module for future use. Allocate and save off mgmt reqs in the
+ * mgmt_req_freelist for future use. Make sure their SM is initialized
+ * to uninit state.
+ * Returns: 0 - on success
+ * -ENOMEM - on error.
+ */
+static int
+csio_mgmtm_init(struct csio_mgmtm *mgmtm, struct csio_hw *hw)
+{
+ struct timer_list *timer = &mgmtm->mgmt_timer;
+
+ init_timer(timer);
+ timer->function = csio_mgmt_tmo_handler;
+ timer->data = (unsigned long)mgmtm;
+
+ INIT_LIST_HEAD(&mgmtm->active_q);
+ INIT_LIST_HEAD(&mgmtm->cbfn_q);
+
+ mgmtm->hw = hw;
+ /*mgmtm->iq_idx = hw->fwevt_iq_idx;*/
+
+ return 0;
+}
+
+/*
+ * csio_mgmtm_exit - MGMT module exit entry point
+ * @mgmtsm - mgmt module
+ *
+ * This function called during MGMT module uninit.
+ * Stop timers, free ioreqs allocated.
+ * Returns: None
+ *
+ */
+static void
+csio_mgmtm_exit(struct csio_mgmtm *mgmtm)
+{
+ del_timer_sync(&mgmtm->mgmt_timer);
+}
+
+
+/**
+ * csio_hw_start - Kicks off the HW State machine
+ * @hw: Pointer to HW module.
+ *
+ * It is assumed that the initialization is a synchronous operation.
+ * So when we return afer posting the event, the HW SM should be in
+ * the ready state, if there were no errors during init.
+ */
+int
+csio_hw_start(struct csio_hw *hw)
+{
+ spin_lock_irq(&hw->lock);
+ csio_post_event(&hw->sm, CSIO_HWE_CFG);
+ spin_unlock_irq(&hw->lock);
+
+ if (csio_is_hw_ready(hw))
+ return 0;
+ else
+ return -EINVAL;
+}
+
+int
+csio_hw_stop(struct csio_hw *hw)
+{
+ csio_post_event(&hw->sm, CSIO_HWE_PCI_REMOVE);
+
+ if (csio_is_hw_removing(hw))
+ return 0;
+ else
+ return -EINVAL;
+}
+
+/* Max reset retries */
+#define CSIO_MAX_RESET_RETRIES 3
+
+/**
+ * csio_hw_reset - Reset the hardware
+ * @hw: HW module.
+ *
+ * Caller should hold lock across this function.
+ */
+int
+csio_hw_reset(struct csio_hw *hw)
+{
+ if (!csio_is_hw_master(hw))
+ return -EPERM;
+
+ if (hw->rst_retries >= CSIO_MAX_RESET_RETRIES) {
+ csio_dbg(hw, "Max hw reset attempts reached..");
+ return -EINVAL;
+ }
+
+ hw->rst_retries++;
+ csio_post_event(&hw->sm, CSIO_HWE_HBA_RESET);
+
+ if (csio_is_hw_ready(hw)) {
+ hw->rst_retries = 0;
+ hw->stats.n_reset_start = jiffies_to_msecs(jiffies);
+ return 0;
+ } else
+ return -EINVAL;
+}
+
+/*
+ * csio_hw_get_device_id - Caches the Adapter's vendor & device id.
+ * @hw: HW module.
+ */
+static void
+csio_hw_get_device_id(struct csio_hw *hw)
+{
+ /* Is the adapter device id cached already ?*/
+ if (csio_is_dev_id_cached(hw))
+ return;
+
+ /* Get the PCI vendor & device id */
+ pci_read_config_word(hw->pdev, PCI_VENDOR_ID,
+ &hw->params.pci.vendor_id);
+ pci_read_config_word(hw->pdev, PCI_DEVICE_ID,
+ &hw->params.pci.device_id);
+
+ csio_dev_id_cached(hw);
+
+} /* csio_hw_get_device_id */
+
+/*
+ * csio_hw_set_description - Set the model, description of the hw.
+ * @hw: HW module.
+ * @ven_id: PCI Vendor ID
+ * @dev_id: PCI Device ID
+ */
+static void
+csio_hw_set_description(struct csio_hw *hw, uint16_t ven_id, uint16_t dev_id)
+{
+ uint32_t adap_type, prot_type;
+
+ if (ven_id == CSIO_VENDOR_ID) {
+ prot_type = (dev_id & CSIO_ASIC_DEVID_PROTO_MASK);
+ adap_type = (dev_id & CSIO_ASIC_DEVID_TYPE_MASK);
+
+ if (prot_type == CSIO_FPGA) {
+ memcpy(hw->model_desc,
+ csio_fcoe_adapters[13].description, 32);
+ } else if (prot_type == CSIO_T4_FCOE_ASIC) {
+ memcpy(hw->hw_ver,
+ csio_fcoe_adapters[adap_type].model_no, 16);
+ memcpy(hw->model_desc,
+ csio_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 */
+
+/**
+ * csio_hw_init - Initialize HW module.
+ * @hw: Pointer to HW module.
+ *
+ * Initialize the members of the HW module.
+ */
+int
+csio_hw_init(struct csio_hw *hw)
+{
+ int rv = -EINVAL;
+ uint32_t i;
+ uint16_t ven_id, dev_id;
+ struct csio_evt_msg *evt_entry;
+
+ INIT_LIST_HEAD(&hw->sm.sm_list);
+ csio_init_state(&hw->sm, csio_hws_uninit);
+ spin_lock_init(&hw->lock);
+ INIT_LIST_HEAD(&hw->sln_head);
+
+ /* Get the PCI vendor & device id */
+ csio_hw_get_device_id(hw);
+
+ strcpy(hw->name, CSIO_HW_NAME);
+
+ /* Set the model & its description */
+
+ ven_id = hw->params.pci.vendor_id;
+ dev_id = hw->params.pci.device_id;
+
+ csio_hw_set_description(hw, ven_id, dev_id);
+
+ /* Initialize default log level */
+ hw->params.log_level = (uint32_t) csio_dbg_level;
+
+ csio_set_fwevt_intr_idx(hw, -1);
+ csio_set_nondata_intr_idx(hw, -1);
+
+ /* Init all the modules: Mailbox, WorkRequest and Transport */
+ if (csio_mbm_init(csio_hw_to_mbm(hw), hw, csio_hw_mb_timer))
+ goto err;
+
+ rv = csio_wrm_init(csio_hw_to_wrm(hw), hw);
+ if (rv)
+ goto err_mbm_exit;
+
+ rv = csio_scsim_init(csio_hw_to_scsim(hw), hw);
+ if (rv)
+ goto err_wrm_exit;
+
+ rv = csio_mgmtm_init(csio_hw_to_mgmtm(hw), hw);
+ if (rv)
+ goto err_scsim_exit;
+ /* Pre-allocate evtq and initialize them */
+ INIT_LIST_HEAD(&hw->evt_active_q);
+ INIT_LIST_HEAD(&hw->evt_free_q);
+ for (i = 0; i < csio_evtq_sz; i++) {
+
+ evt_entry = kzalloc(sizeof(struct csio_evt_msg), GFP_KERNEL);
+ if (!evt_entry) {
+ csio_err(hw, "Failed to initialize eventq");
+ goto err_evtq_cleanup;
+ }
+
+ list_add_tail(&evt_entry->list, &hw->evt_free_q);
+ CSIO_INC_STATS(hw, n_evt_freeq);
+ }
+
+ hw->dev_num = dev_num;
+ dev_num++;
+
+ return 0;
+
+err_evtq_cleanup:
+ csio_evtq_cleanup(hw);
+ csio_mgmtm_exit(csio_hw_to_mgmtm(hw));
+err_scsim_exit:
+ csio_scsim_exit(csio_hw_to_scsim(hw));
+err_wrm_exit:
+ csio_wrm_exit(csio_hw_to_wrm(hw), hw);
+err_mbm_exit:
+ csio_mbm_exit(csio_hw_to_mbm(hw));
+err:
+ return rv;
+}
+
+/**
+ * csio_hw_exit - Un-initialize HW module.
+ * @hw: Pointer to HW module.
+ *
+ */
+void
+csio_hw_exit(struct csio_hw *hw)
+{
+ csio_evtq_cleanup(hw);
+ csio_mgmtm_exit(csio_hw_to_mgmtm(hw));
+ csio_scsim_exit(csio_hw_to_scsim(hw));
+ csio_wrm_exit(csio_hw_to_wrm(hw), hw);
+ csio_mbm_exit(csio_hw_to_mbm(hw));
+}
--
1.7.1
^ permalink raw reply related
* [V3 PATCH 0/9] csiostor: Chelsio FCoE offload driver submission
From: Naresh Kumar Inna @ 2012-09-11 14:38 UTC (permalink / raw)
To: JBottomley, linux-scsi, dm, leedom; +Cc: netdev, naresh, chethan
This is the initial submission of the Chelsio FCoE offload driver (csiostor)
to the upstream kernel. This driver currently supports FCoE offload
functionality over Chelsio T4-based 10Gb Converged Network Adapters.
The following patches contain the driver sources for csiostor driver and
updates to firmware/hardware header files shared between csiostor,
cxgb4 (Chelsio T4-based NIC driver) and cxgb4vf (Chelsio T4-based Virtual
Function NIC driver). The csiostor driver is dependent on these
header updates. These patches have been generated against scsi 'misc' branch.
csiostor is a low level SCSI driver that interfaces with PCI, SCSI midlayer and
FC transport subsystems. This driver claims the FCoE PCIe function on
Chelsio Converged Network Adapters. It relies on firmware events for slow path
operations like discovery, thereby offloading session management. The driver
programs firmware via Work Request interfaces for fast path I/O offload
features.
V3 patches address review comments from Ben Hutchings and Stephen Hemminger.
These changes are listed in the individual patch emails.
Here is the brief description of patches:
[V3 PATCH 1/9]: Hardware interface, Makefile and Kconfig changes.
[V3 PATCH 2/9]: Driver initialization and Work Request services.
[V3 PATCH 3/9]: FC transport interfaces and mailbox services.
[V3 PATCH 4/9]: Local and remote port state tracking functionality.
[V3 PATCH 5/9]: Interrupt handling and fast path I/O functionality.
[V3 PATCH 6/9]: Header files part 1.
[V3 PATCH 7/9]: Header files part 2.
[V3 PATCH 8/9]: Updates to header files shared between cxgb4, cxgb4vf and
csiostor.
[V3 PATCH 9/9]: Header file compatibility fixes to cxgb4vf.
Naresh Kumar Inna (9):
csiostor: Chelsio FCoE offload driver submission (sources part 1).
csiostor: Chelsio FCoE offload driver submission (sources part 2).
csiostor: Chelsio FCoE offload driver submission (sources part 3).
csiostor: Chelsio FCoE offload driver submission (sources part 4).
csiostor: Chelsio FCoE offload driver submission (sources part 5).
csiostor: Chelsio FCoE offload driver submission (headers part 1).
csiostor: Chelsio FCoE offload driver submission (headers part 2).
cxgb4: Chelsio FCoE offload driver submission (cxgb4 common header
updates).
cxgb4vf: Chelsio FCoE offload driver submission (header compatibility
fixes).
drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c | 2 +-
drivers/net/ethernet/chelsio/cxgb4/sge.c | 10 +-
drivers/net/ethernet/chelsio/cxgb4/t4_hw.c | 16 +-
drivers/net/ethernet/chelsio/cxgb4/t4_msg.h | 1 +
drivers/net/ethernet/chelsio/cxgb4/t4_regs.h | 69 +-
drivers/net/ethernet/chelsio/cxgb4/t4fw_api.h | 104 +-
drivers/net/ethernet/chelsio/cxgb4vf/sge.c | 11 +-
drivers/scsi/Kconfig | 1 +
drivers/scsi/Makefile | 1 +
drivers/scsi/csiostor/Kconfig | 20 +
drivers/scsi/csiostor/Makefile | 11 +
drivers/scsi/csiostor/csio_attr.c | 809 +++++
drivers/scsi/csiostor/csio_defs.h | 108 +
drivers/scsi/csiostor/csio_hw.c | 4396 +++++++++++++++++++++++
drivers/scsi/csiostor/csio_hw.h | 666 ++++
drivers/scsi/csiostor/csio_init.c | 1272 +++++++
drivers/scsi/csiostor/csio_init.h | 158 +
drivers/scsi/csiostor/csio_isr.c | 624 ++++
drivers/scsi/csiostor/csio_lnode.c | 2148 +++++++++++
drivers/scsi/csiostor/csio_lnode.h | 255 ++
drivers/scsi/csiostor/csio_mb.c | 1769 +++++++++
drivers/scsi/csiostor/csio_mb.h | 278 ++
drivers/scsi/csiostor/csio_rnode.c | 889 +++++
drivers/scsi/csiostor/csio_rnode.h | 141 +
drivers/scsi/csiostor/csio_scsi.c | 2561 +++++++++++++
drivers/scsi/csiostor/csio_scsi.h | 342 ++
drivers/scsi/csiostor/csio_wr.c | 1632 +++++++++
drivers/scsi/csiostor/csio_wr.h | 512 +++
drivers/scsi/csiostor/t4fw_api_stor.h | 578 +++
29 files changed, 19347 insertions(+), 37 deletions(-)
create mode 100644 drivers/scsi/csiostor/Kconfig
create mode 100644 drivers/scsi/csiostor/Makefile
create mode 100644 drivers/scsi/csiostor/csio_attr.c
create mode 100644 drivers/scsi/csiostor/csio_defs.h
create mode 100644 drivers/scsi/csiostor/csio_hw.c
create mode 100644 drivers/scsi/csiostor/csio_hw.h
create mode 100644 drivers/scsi/csiostor/csio_init.c
create mode 100644 drivers/scsi/csiostor/csio_init.h
create mode 100644 drivers/scsi/csiostor/csio_isr.c
create mode 100644 drivers/scsi/csiostor/csio_lnode.c
create mode 100644 drivers/scsi/csiostor/csio_lnode.h
create mode 100644 drivers/scsi/csiostor/csio_mb.c
create mode 100644 drivers/scsi/csiostor/csio_mb.h
create mode 100644 drivers/scsi/csiostor/csio_rnode.c
create mode 100644 drivers/scsi/csiostor/csio_rnode.h
create mode 100644 drivers/scsi/csiostor/csio_scsi.c
create mode 100644 drivers/scsi/csiostor/csio_scsi.h
create mode 100644 drivers/scsi/csiostor/csio_wr.c
create mode 100644 drivers/scsi/csiostor/csio_wr.h
create mode 100644 drivers/scsi/csiostor/t4fw_api_stor.h
^ permalink raw reply
* [PATCH v2 1/2] iproute2: add libgenl files
From: Julian Anastasov @ 2012-09-11 9:04 UTC (permalink / raw)
To: Stephen Hemminger; +Cc: netdev
In-Reply-To: <1347354274-3119-1-git-send-email-ja@ssi.bg>
Create libgenl.h and libgenl.c. They will contain
common code for GENL users such as ipl2tp, tcp_metrics, etc.
Signed-off-by: Julian Anastasov <ja@ssi.bg>
---
include/libgenl.h | 29 +++++++++++++++++++++++
lib/Makefile | 2 +-
lib/libgenl.c | 65 +++++++++++++++++++++++++++++++++++++++++++++++++++++
3 files changed, 95 insertions(+), 1 deletions(-)
create mode 100644 include/libgenl.h
create mode 100644 lib/libgenl.c
diff --git a/include/libgenl.h b/include/libgenl.h
new file mode 100644
index 0000000..7862292
--- /dev/null
+++ b/include/libgenl.h
@@ -0,0 +1,29 @@
+#ifndef __LIBGENL_H__
+#define __LIBGENL_H__
+
+#include "libnetlink.h"
+
+#define GENL_DEFINE_REQUEST(req, hdrsize, bufsiz) \
+struct { \
+ struct nlmsghdr n; \
+ struct genlmsghdr g; \
+ char buf[NLMSG_ALIGN(hdrsize) + (bufsiz)]; \
+} req
+
+#define GENL_INIT_REQUEST(req, family, hdrsize, ver, cmd_, flags) \
+ do { \
+ req.n = (struct nlmsghdr) { \
+ .nlmsg_type = (family), \
+ .nlmsg_flags = (flags), \
+ .nlmsg_len = NLMSG_LENGTH(GENL_HDRLEN + \
+ (hdrsize)), \
+ }; \
+ req.g = (struct genlmsghdr) { \
+ .cmd = (cmd_), \
+ .version = (ver), \
+ }; \
+ } while (0)
+
+extern int genl_resolve_family(struct rtnl_handle *grth, const char *family);
+
+#endif /* __LIBGENL_H__ */
diff --git a/lib/Makefile b/lib/Makefile
index da2f0fc..bfbe672 100644
--- a/lib/Makefile
+++ b/lib/Makefile
@@ -2,7 +2,7 @@ CFLAGS += -fPIC
UTILOBJ=utils.o rt_names.o ll_types.o ll_proto.o ll_addr.o inet_proto.o
-NLOBJ=ll_map.o libnetlink.o
+NLOBJ=libgenl.o ll_map.o libnetlink.o
all: libnetlink.a libutil.a
diff --git a/lib/libgenl.c b/lib/libgenl.c
new file mode 100644
index 0000000..e7ddf95
--- /dev/null
+++ b/lib/libgenl.c
@@ -0,0 +1,65 @@
+/*
+ * libgenl.c GENL library
+ */
+
+#include <stdio.h>
+#include <stdlib.h>
+#include <unistd.h>
+
+#include <linux/genetlink.h>
+#include "libgenl.h"
+
+static int genl_parse_getfamily(struct nlmsghdr *nlh)
+{
+ struct rtattr *tb[CTRL_ATTR_MAX + 1];
+ struct genlmsghdr *ghdr = NLMSG_DATA(nlh);
+ int len = nlh->nlmsg_len;
+ struct rtattr *attrs;
+
+ if (nlh->nlmsg_type != GENL_ID_CTRL) {
+ fprintf(stderr, "Not a controller message, nlmsg_len=%d "
+ "nlmsg_type=0x%x\n", nlh->nlmsg_len, nlh->nlmsg_type);
+ return -1;
+ }
+
+ len -= NLMSG_LENGTH(GENL_HDRLEN);
+
+ if (len < 0) {
+ fprintf(stderr, "wrong controller message len %d\n", len);
+ return -1;
+ }
+
+ if (ghdr->cmd != CTRL_CMD_NEWFAMILY) {
+ fprintf(stderr, "Unknown controller command %d\n", ghdr->cmd);
+ return -1;
+ }
+
+ attrs = (struct rtattr *) ((char *) ghdr + GENL_HDRLEN);
+ parse_rtattr(tb, CTRL_ATTR_MAX, attrs, len);
+
+ if (tb[CTRL_ATTR_FAMILY_ID] == NULL) {
+ fprintf(stderr, "Missing family id TLV\n");
+ return -1;
+ }
+
+ return rta_getattr_u16(tb[CTRL_ATTR_FAMILY_ID]);
+}
+
+int genl_resolve_family(struct rtnl_handle *grth, const char *family)
+{
+ GENL_DEFINE_REQUEST(req, 0, 1024);
+
+ GENL_INIT_REQUEST(req, GENL_ID_CTRL, 0, 0, CTRL_CMD_GETFAMILY,
+ NLM_F_REQUEST);
+
+ addattr_l(&req.n, 1024, CTRL_ATTR_FAMILY_NAME,
+ family, strlen(family) + 1);
+
+ if (rtnl_talk(grth, &req.n, 0, 0, &req.n) < 0) {
+ fprintf(stderr, "Error talking to the kernel\n");
+ return -2;
+ }
+
+ return genl_parse_getfamily(&req.n);
+}
+
--
1.7.3.4
^ permalink raw reply related
* [PATCH v2 2/2] iproute2: use libgenl in ipl2tp
From: Julian Anastasov @ 2012-09-11 9:04 UTC (permalink / raw)
To: Stephen Hemminger; +Cc: netdev
In-Reply-To: <1347354274-3119-1-git-send-email-ja@ssi.bg>
Use the common code from libgenl.c to parse family.
Signed-off-by: Julian Anastasov <ja@ssi.bg>
---
ip/ipl2tp.c | 64 +---------------------------------------------------------
1 files changed, 2 insertions(+), 62 deletions(-)
diff --git a/ip/ipl2tp.c b/ip/ipl2tp.c
index 1cbed8d..c7893c2 100644
--- a/ip/ipl2tp.c
+++ b/ip/ipl2tp.c
@@ -25,6 +25,7 @@
#include <linux/genetlink.h>
#include <linux/l2tp.h>
+#include "libgenl.h"
#include "utils.h"
#include "ip_common.h"
@@ -747,67 +748,6 @@ static int do_show(int argc, char **argv)
return 0;
}
-static int genl_parse_getfamily(struct nlmsghdr *nlh)
-{
- struct rtattr *tb[CTRL_ATTR_MAX + 1];
- struct genlmsghdr *ghdr = NLMSG_DATA(nlh);
- int len = nlh->nlmsg_len;
- struct rtattr *attrs;
-
- if (nlh->nlmsg_type != GENL_ID_CTRL) {
- fprintf(stderr, "Not a controller message, nlmsg_len=%d "
- "nlmsg_type=0x%x\n", nlh->nlmsg_len, nlh->nlmsg_type);
- return -1;
- }
-
- if (ghdr->cmd != CTRL_CMD_NEWFAMILY) {
- fprintf(stderr, "Unknown controller command %d\n", ghdr->cmd);
- return -1;
- }
-
- len -= NLMSG_LENGTH(GENL_HDRLEN);
-
- if (len < 0) {
- fprintf(stderr, "wrong controller message len %d\n", len);
- return -1;
- }
-
- attrs = (struct rtattr *) ((char *) ghdr + GENL_HDRLEN);
- parse_rtattr(tb, CTRL_ATTR_MAX, attrs, len);
-
- if (tb[CTRL_ATTR_FAMILY_ID] == NULL) {
- fprintf(stderr, "Missing family id TLV\n");
- return -1;
- }
-
- return rta_getattr_u16(tb[CTRL_ATTR_FAMILY_ID]);
-}
-
-int genl_ctrl_resolve_family(const char *family)
-{
- struct {
- struct nlmsghdr n;
- struct genlmsghdr g;
- char buf[1024];
- } req;
-
- memset(&req, 0, sizeof(req));
- req.n.nlmsg_len = NLMSG_LENGTH(GENL_HDRLEN);
- req.n.nlmsg_flags = NLM_F_REQUEST;
- req.n.nlmsg_type = GENL_ID_CTRL;
- req.g.cmd = CTRL_CMD_GETFAMILY;
-
- addattr_l(&req.n, 1024, CTRL_ATTR_FAMILY_NAME,
- family, strlen(family) + 1);
-
- if (rtnl_talk(&genl_rth, &req.n, 0, 0, &req.n) < 0) {
- fprintf(stderr, "Error talking to the kernel\n");
- return -2;
- }
-
- return genl_parse_getfamily(&req.n);
-}
-
int do_ipl2tp(int argc, char **argv)
{
if (genl_family < 0) {
@@ -816,7 +756,7 @@ int do_ipl2tp(int argc, char **argv)
exit(1);
}
- genl_family = genl_ctrl_resolve_family(L2TP_GENL_NAME);
+ genl_family = genl_resolve_family(&genl_rth, L2TP_GENL_NAME);
if (genl_family < 0)
exit(1);
}
--
1.7.3.4
^ permalink raw reply related
* [PATCH v2 0/2] iproute2: add GENL helpers
From: Julian Anastasov @ 2012-09-11 9:04 UTC (permalink / raw)
To: Stephen Hemminger; +Cc: netdev
Put common code for GENL users in new files.
v2:
- remove memset for buffer in GENL_INIT_REQUEST and use
C99 style initializers
- rename funcs: libgenl_ to genl_
Julian Anastasov (2):
iproute2: add libgenl files
iproute2: use libgenl in ipl2tp
include/libgenl.h | 29 +++++++++++++++++++++++
ip/ipl2tp.c | 64 +--------------------------------------------------
lib/Makefile | 2 +-
lib/libgenl.c | 65 +++++++++++++++++++++++++++++++++++++++++++++++++++++
4 files changed, 97 insertions(+), 63 deletions(-)
create mode 100644 include/libgenl.h
create mode 100644 lib/libgenl.c
--
1.7.3.4
^ permalink raw reply
* Re: [PATCH 2/3] ipv6, route: remove BACKTRACK() macro
From: Cong Wang @ 2012-09-11 8:39 UTC (permalink / raw)
To: David Miller; +Cc: netdev
In-Reply-To: <20120910.214245.368123682481775366.davem@davemloft.net>
On Mon, 2012-09-10 at 21:42 -0400, David Miller wrote:
>
> Yes, I know this conflicts with cases like how we killed all of
> the netlink macros with embedded gotos.
>
> But this patch made things worse and added code duplication. Fix
> it without the code duplication side effect and it'll be fine.
Ok, will do.
Thanks, David!
^ permalink raw reply
* RE: [PATCH net-next] r8169: use unlimited DMA burst for TX
From: hayeswang @ 2012-09-11 8:09 UTC (permalink / raw)
To: 'Francois Romieu', 'David Miller'
Cc: mschmidt, netdev, ivecera
In-Reply-To: <20120910215521.GA930@electric-eye.fr.zoreil.com>
Hi,
Francois Romieu [mailto:romieu@fr.zoreil.com]
> Sent: Tuesday, September 11, 2012 5:55 AM
> To: David Miller
> Cc: mschmidt@redhat.com; netdev@vger.kernel.org; Hayeswang;
> nic_swsd; ivecera@redhat.com
> Subject: Re: [PATCH net-next] r8169: use unlimited DMA burst for TX
>
> Hayes, should we:
> - mimic Realtek's 8168, 8169 and 810x drivers ?
> - always set TX_DMA_BURST at the max value ?
> - do something different (per chipset) ?
Our hw engineer suggets to set unlimited for both TX_DMA_BURST and RX_DMA_BURST
for all chipsets.
Best Regards,
Hayes
^ permalink raw reply
* [PATCH net-next v3 4/4] ipv6: use DST_* macro to set obselete field
From: Nicolas Dichtel @ 2012-09-11 8:09 UTC (permalink / raw)
To: vyasevich, davem, eric.dumazet, sds, james.l.morris, eparis
Cc: sri, linux-sctp, netdev, Nicolas Dichtel
In-Reply-To: <1347350987-8054-1-git-send-email-nicolas.dichtel@6wind.com>
Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
---
net/ipv6/route.c | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/net/ipv6/route.c b/net/ipv6/route.c
index 561f249..0c6f132 100644
--- a/net/ipv6/route.c
+++ b/net/ipv6/route.c
@@ -226,7 +226,7 @@ static struct rt6_info ip6_null_entry_template = {
.dst = {
.__refcnt = ATOMIC_INIT(1),
.__use = 1,
- .obsolete = -1,
+ .obsolete = DST_OBSOLETE_FORCE_CHK,
.error = -ENETUNREACH,
.input = ip6_pkt_discard,
.output = ip6_pkt_discard_out,
@@ -246,7 +246,7 @@ static struct rt6_info ip6_prohibit_entry_template = {
.dst = {
.__refcnt = ATOMIC_INIT(1),
.__use = 1,
- .obsolete = -1,
+ .obsolete = DST_OBSOLETE_FORCE_CHK,
.error = -EACCES,
.input = ip6_pkt_prohibit,
.output = ip6_pkt_prohibit_out,
@@ -261,7 +261,7 @@ static struct rt6_info ip6_blk_hole_entry_template = {
.dst = {
.__refcnt = ATOMIC_INIT(1),
.__use = 1,
- .obsolete = -1,
+ .obsolete = DST_OBSOLETE_FORCE_CHK,
.error = -EINVAL,
.input = dst_discard,
.output = dst_discard,
--
1.7.12
^ permalink raw reply related
* [PATCH net-next v3 3/4] ipv6: use net->rt_genid to check dst validity
From: Nicolas Dichtel @ 2012-09-11 8:09 UTC (permalink / raw)
To: vyasevich, davem, eric.dumazet, sds, james.l.morris, eparis
Cc: sri, linux-sctp, netdev, Nicolas Dichtel
In-Reply-To: <1347350987-8054-1-git-send-email-nicolas.dichtel@6wind.com>
IPv6 dst should take care of rt_genid too. When a xfrm policy is inserted or
deleted, all dst should be invalidated.
To force the validation, dst entries should be created with ->obsolete set to
DST_OBSOLETE_FORCE_CHK. This was already the case for all functions calling
ip6_dst_alloc(), except for ip6_rt_copy().
As a consequence, we can remove the specific code in inet6_connection_sock.
Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
---
include/net/ip6_fib.h | 5 ++---
net/ipv6/inet6_connection_sock.c | 23 +----------------------
net/ipv6/route.c | 13 +++++++++----
3 files changed, 12 insertions(+), 29 deletions(-)
diff --git a/include/net/ip6_fib.h b/include/net/ip6_fib.h
index cd64cf3..8a2a203 100644
--- a/include/net/ip6_fib.h
+++ b/include/net/ip6_fib.h
@@ -112,9 +112,8 @@ struct rt6_info {
struct inet6_dev *rt6i_idev;
unsigned long _rt6i_peer;
-#ifdef CONFIG_XFRM
- u32 rt6i_flow_cache_genid;
-#endif
+ u32 rt6i_genid;
+
/* more non-fragment space at head required */
unsigned short rt6i_nfheader_len;
diff --git a/net/ipv6/inet6_connection_sock.c b/net/ipv6/inet6_connection_sock.c
index 0251a60..c4f9341 100644
--- a/net/ipv6/inet6_connection_sock.c
+++ b/net/ipv6/inet6_connection_sock.c
@@ -175,33 +175,12 @@ void __inet6_csk_dst_store(struct sock *sk, struct dst_entry *dst,
const struct in6_addr *saddr)
{
__ip6_dst_store(sk, dst, daddr, saddr);
-
-#ifdef CONFIG_XFRM
- {
- struct rt6_info *rt = (struct rt6_info *)dst;
- rt->rt6i_flow_cache_genid = atomic_read(&flow_cache_genid);
- }
-#endif
}
static inline
struct dst_entry *__inet6_csk_dst_check(struct sock *sk, u32 cookie)
{
- struct dst_entry *dst;
-
- dst = __sk_dst_check(sk, cookie);
-
-#ifdef CONFIG_XFRM
- if (dst) {
- struct rt6_info *rt = (struct rt6_info *)dst;
- if (rt->rt6i_flow_cache_genid != atomic_read(&flow_cache_genid)) {
- __sk_dst_reset(sk);
- dst = NULL;
- }
- }
-#endif
-
- return dst;
+ return __sk_dst_check(sk, cookie);
}
static struct dst_entry *inet6_csk_route_socket(struct sock *sk,
diff --git a/net/ipv6/route.c b/net/ipv6/route.c
index 339d921..561f249 100644
--- a/net/ipv6/route.c
+++ b/net/ipv6/route.c
@@ -281,13 +281,14 @@ static inline struct rt6_info *ip6_dst_alloc(struct net *net,
struct fib6_table *table)
{
struct rt6_info *rt = dst_alloc(&net->ipv6.ip6_dst_ops, dev,
- 0, DST_OBSOLETE_NONE, flags);
+ 0, DST_OBSOLETE_FORCE_CHK, flags);
if (rt) {
struct dst_entry *dst = &rt->dst;
memset(dst + 1, 0, sizeof(*rt) - sizeof(*dst));
rt6_init_peer(rt, table ? &table->tb6_peers : net->ipv6.peers);
+ rt->rt6i_genid = rt_genid(net);
}
return rt;
}
@@ -1031,6 +1032,13 @@ static struct dst_entry *ip6_dst_check(struct dst_entry *dst, u32 cookie)
rt = (struct rt6_info *) dst;
+ /* All IPV6 dsts are created with ->obsolete set to the value
+ * DST_OBSOLETE_FORCE_CHK which forces validation calls down
+ * into this function always.
+ */
+ if (rt->rt6i_genid != rt_genid(dev_net(rt->dst.dev)))
+ return NULL;
+
if (rt->rt6i_node && (rt->rt6i_node->fn_sernum == cookie)) {
if (rt->rt6i_peer_genid != rt6_peer_genid()) {
if (!rt6_has_peer(rt))
@@ -1397,8 +1405,6 @@ int ip6_route_add(struct fib6_config *cfg)
goto out;
}
- rt->dst.obsolete = -1;
-
if (cfg->fc_flags & RTF_EXPIRES)
rt6_set_expires(rt, jiffies +
clock_t_to_jiffies(cfg->fc_expires));
@@ -2093,7 +2099,6 @@ struct rt6_info *addrconf_dst_alloc(struct inet6_dev *idev,
rt->dst.input = ip6_input;
rt->dst.output = ip6_output;
rt->rt6i_idev = idev;
- rt->dst.obsolete = -1;
rt->rt6i_flags = RTF_UP | RTF_NONEXTHOP;
if (anycast)
--
1.7.12
^ permalink raw reply related
* [PATCH net-next v3 2/4] xfrm: invalidate dst on policy insertion/deletion
From: Nicolas Dichtel @ 2012-09-11 8:09 UTC (permalink / raw)
To: vyasevich, davem, eric.dumazet, sds, james.l.morris, eparis
Cc: sri, linux-sctp, netdev, Nicolas Dichtel
In-Reply-To: <1347350987-8054-1-git-send-email-nicolas.dichtel@6wind.com>
When a policy is inserted or deleted, all dst should be recalculated.
Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
---
net/xfrm/xfrm_policy.c | 1 +
security/selinux/include/xfrm.h | 1 +
2 files changed, 2 insertions(+)
diff --git a/net/xfrm/xfrm_policy.c b/net/xfrm/xfrm_policy.c
index 741a32a..67f456d 100644
--- a/net/xfrm/xfrm_policy.c
+++ b/net/xfrm/xfrm_policy.c
@@ -602,6 +602,7 @@ int xfrm_policy_insert(int dir, struct xfrm_policy *policy, int excl)
xfrm_pol_hold(policy);
net->xfrm.policy_count[dir]++;
atomic_inc(&flow_cache_genid);
+ rt_genid_bump(net);
if (delpol)
__xfrm_policy_unlink(delpol, dir);
policy->index = delpol ? delpol->index : xfrm_gen_index(net, dir);
diff --git a/security/selinux/include/xfrm.h b/security/selinux/include/xfrm.h
index c220f31..65f67cb 100644
--- a/security/selinux/include/xfrm.h
+++ b/security/selinux/include/xfrm.h
@@ -51,6 +51,7 @@ int selinux_xfrm_decode_session(struct sk_buff *skb, u32 *sid, int ckall);
static inline void selinux_xfrm_notify_policyload(void)
{
atomic_inc(&flow_cache_genid);
+ rt_genid_bump(&init_net);
}
#else
static inline int selinux_xfrm_enabled(void)
--
1.7.12
^ permalink raw reply related
* [PATCH net-next v3 1/4] netns: move net->ipv4.rt_genid to net->rt_genid
From: Nicolas Dichtel @ 2012-09-11 8:09 UTC (permalink / raw)
To: vyasevich, davem, eric.dumazet, sds, james.l.morris, eparis
Cc: sri, linux-sctp, netdev, Nicolas Dichtel
In-Reply-To: <1347350987-8054-1-git-send-email-nicolas.dichtel@6wind.com>
This commit prepares the use of rt_genid by both IPv4 and IPv6.
Initialization is left in IPv4 part.
Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
---
include/net/net_namespace.h | 10 ++++++++++
include/net/netns/ipv4.h | 1 -
net/ipv4/route.c | 9 ++-------
3 files changed, 12 insertions(+), 8 deletions(-)
diff --git a/include/net/net_namespace.h b/include/net/net_namespace.h
index 5ae57f1..037190a 100644
--- a/include/net/net_namespace.h
+++ b/include/net/net_namespace.h
@@ -107,6 +107,7 @@ struct net {
#endif
struct netns_ipvs *ipvs;
struct sock *diag_nlsk;
+ atomic_t rt_genid;
};
/*
@@ -312,5 +313,14 @@ static inline void unregister_net_sysctl_table(struct ctl_table_header *header)
}
#endif
+static inline int rt_genid(struct net *net)
+{
+ return atomic_read(&net->rt_genid);
+}
+
+static inline void rt_genid_bump(struct net *net)
+{
+ atomic_inc(&net->rt_genid);
+}
#endif /* __NET_NET_NAMESPACE_H */
diff --git a/include/net/netns/ipv4.h b/include/net/netns/ipv4.h
index 7d00583..2ae2b83 100644
--- a/include/net/netns/ipv4.h
+++ b/include/net/netns/ipv4.h
@@ -64,7 +64,6 @@ struct netns_ipv4 {
kgid_t sysctl_ping_group_range[2];
long sysctl_tcp_mem[3];
- atomic_t rt_genid;
atomic_t dev_addr_genid;
#ifdef CONFIG_IP_MROUTE
diff --git a/net/ipv4/route.c b/net/ipv4/route.c
index d39edf1..8ff984b 100644
--- a/net/ipv4/route.c
+++ b/net/ipv4/route.c
@@ -202,11 +202,6 @@ EXPORT_SYMBOL(ip_tos2prio);
static DEFINE_PER_CPU(struct rt_cache_stat, rt_cache_stat);
#define RT_CACHE_STAT_INC(field) __this_cpu_inc(rt_cache_stat.field)
-static inline int rt_genid(struct net *net)
-{
- return atomic_read(&net->ipv4.rt_genid);
-}
-
#ifdef CONFIG_PROC_FS
static void *rt_cache_seq_start(struct seq_file *seq, loff_t *pos)
{
@@ -449,7 +444,7 @@ static inline bool rt_is_expired(const struct rtable *rth)
void rt_cache_flush(struct net *net)
{
- atomic_inc(&net->ipv4.rt_genid);
+ rt_genid_bump(net);
}
static struct neighbour *ipv4_neigh_lookup(const struct dst_entry *dst,
@@ -2506,7 +2501,7 @@ static __net_initdata struct pernet_operations sysctl_route_ops = {
static __net_init int rt_genid_init(struct net *net)
{
- atomic_set(&net->ipv4.rt_genid, 0);
+ atomic_set(&net->rt_genid, 0);
get_random_bytes(&net->ipv4.dev_addr_genid,
sizeof(net->ipv4.dev_addr_genid));
return 0;
--
1.7.12
^ permalink raw reply related
* [PATCH net-next v3 0/4] Take care of xfrm policy when checking dst entries
From: Nicolas Dichtel @ 2012-09-11 8:09 UTC (permalink / raw)
To: vyasevich, davem, eric.dumazet, sds, james.l.morris, eparis
Cc: sri, linux-sctp, netdev
In-Reply-To: <504DFF90.3010802@6wind.com>
The goal of these patches is to fix the following problem: a session is
established (TCP, SCTP) and after a new policy is inserted. The current
code does not recalculate the route, thus the traffic is not encrypted.
The patch propose to check flow_cache_genid value when checking a dst
entry, which is incremented each time a policy is inserted or deleted.
v2: use net->ipv4.rt_genid instead of flow_cache_genid (and thus save a test
in fast path). Also move it to net->rt_genid, to be able to use it for IPv6
too. Note that IPv6 will have one more test in fast path.
v3: remove unrelated "#ifdef CONFIG_XFRM" in IPv6 part
bump rt_genid in selinux code (same place than flow_cache_genid)
Patches are tested with TCP and SCTP, IPv4 and IPv6.
Comments are welcome.
Regards,
Nicolas
^ permalink raw reply
* Re: [net-next.git 5/8] stmmac: get/set coalesce parameters via ethtool
From: Giuseppe CAVALLARO @ 2012-09-11 6:58 UTC (permalink / raw)
To: Ben Hutchings; +Cc: netdev, davem
In-Reply-To: <1347295510.2823.4.camel@bwh-desktop.uk.solarflarecom.com>
Hello Ben
On 9/10/2012 6:45 PM, Ben Hutchings wrote:
> You should also check that the settings you don't support
> (rx_max_coalesced_frames, use_adaptive_rx_coalesce,
> use_adaptive_tx_coalesce, *_irq) are equal to 0. (This isn't done
> consistently in all drivers - but I believe best practice is to reject
> rather than quietly ignoring settings you don't support.)
Agreed. I've added this check in ethtool and now, for example, I get the
following on un-supported setting
$ ethtool -C eth0 tx-usecs-irq 50
Cannot set device ring parameters: Operation not supported
$ ethtool -C eth0 tx-usecs-irq 0
tx-usecs-irq unmodified, ignoring
no ring parameters changed, aborting
>
> I don't have time to review the other changes properly, but at a quick
> look they seem to be OK.
Thanks a lot
peppe
>
> Ben.
>
^ permalink raw reply
* [net-next.git 5/8 (V2)] stmmac: get/set coalesce parameters via ethtool
From: Giuseppe CAVALLARO @ 2012-09-11 6:55 UTC (permalink / raw)
To: netdev; +Cc: bhutchings, davem, Giuseppe Cavallaro
In-Reply-To: <1347346514-23411-1-git-send-email-peppe.cavallaro@st.com>
This patch is to get/set the tx/rx coalesce parameters
via ethtool interface.
Tests have been done on several platform with
different GMAC chips w/o w/ RX watchdog feature.
V2: reject coalesce settings that are not supported.
Signed-off-by: Giuseppe Cavallaro <peppe.cavallaro@st.com>
---
drivers/net/ethernet/stmicro/stmmac/common.h | 8 ++-
drivers/net/ethernet/stmicro/stmmac/stmmac.h | 1 +
.../net/ethernet/stmicro/stmmac/stmmac_ethtool.c | 85 ++++++++++++++++++++
drivers/net/ethernet/stmicro/stmmac/stmmac_main.c | 20 +++---
4 files changed, 103 insertions(+), 11 deletions(-)
diff --git a/drivers/net/ethernet/stmicro/stmmac/common.h b/drivers/net/ethernet/stmicro/stmmac/common.h
index 02eb2da..b9033cc 100644
--- a/drivers/net/ethernet/stmicro/stmmac/common.h
+++ b/drivers/net/ethernet/stmicro/stmmac/common.h
@@ -169,7 +169,13 @@ struct stmmac_extra_stats {
#define DMA_HW_FEAT_SAVLANINS 0x08000000 /* Source Addr or VLAN Insertion */
#define DMA_HW_FEAT_ACTPHYIF 0x70000000 /* Active/selected PHY interface */
#define DEFAULT_DMA_PBL 8
-#define DEFAULT_DMA_RIWT 0xff /* Max RI Watchdog Timer count */
+
+/* Coalesce defines */
+#define MAX_DMA_RIWT 0xff /* Max RI Watchdog Timer count */
+#define MIN_DMA_RIWT 0x20
+#define STMMAC_COAL_TX_TIMER 40000
+#define STMMAC_MAX_COAL_TX_TICK 100000
+#define STMMAC_TX_MAX_FRAMES 32
enum rx_frame_status { /* IPC status */
good_frame = 0,
diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac.h b/drivers/net/ethernet/stmicro/stmmac/stmmac.h
index 6901e3c..ad4f6b9 100644
--- a/drivers/net/ethernet/stmicro/stmmac/stmmac.h
+++ b/drivers/net/ethernet/stmicro/stmmac/stmmac.h
@@ -93,6 +93,7 @@ struct stmmac_priv {
u32 tx_count_frames;
u32 tx_coal_frames;
u32 tx_coal_timer;
+ u32 rx_riwt;
};
extern int phyaddr;
diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_ethtool.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_ethtool.c
index 48ad0bc..d27cc18 100644
--- a/drivers/net/ethernet/stmicro/stmmac/stmmac_ethtool.c
+++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_ethtool.c
@@ -522,6 +522,89 @@ static int stmmac_ethtool_op_set_eee(struct net_device *dev,
return phy_ethtool_set_eee(priv->phydev, edata);
}
+static u32 stmmac_usec2riwt(u32 usec, struct stmmac_priv *priv)
+{
+ unsigned long clk = clk_get_rate(priv->stmmac_clk);
+
+ if (!clk)
+ return 0;
+
+ return (usec * (clk / 1000000)) / 256;
+}
+
+static u32 stmmac_riwt2usec(u32 riwt, struct stmmac_priv *priv)
+{
+ unsigned long clk = clk_get_rate(priv->stmmac_clk);
+
+ if (!clk)
+ return 0;
+
+ return (riwt * 256) / (clk / 1000000);
+}
+
+static int stmmac_get_coalesce(struct net_device *dev,
+ struct ethtool_coalesce *ec)
+{
+ struct stmmac_priv *priv = netdev_priv(dev);
+
+ ec->tx_coalesce_usecs = priv->tx_coal_timer;
+ ec->tx_max_coalesced_frames = priv->tx_coal_frames;
+
+ if (priv->use_riwt)
+ ec->rx_coalesce_usecs = stmmac_riwt2usec(priv->rx_riwt, priv);
+
+ return 0;
+}
+
+static int stmmac_set_coalesce(struct net_device *dev,
+ struct ethtool_coalesce *ec)
+{
+ struct stmmac_priv *priv = netdev_priv(dev);
+ unsigned int rx_riwt;
+
+ /* Check not supported parameters */
+ if ((ec->rx_max_coalesced_frames) || (ec->rx_coalesce_usecs_irq) ||
+ (ec->rx_max_coalesced_frames_irq) || (ec->tx_coalesce_usecs_irq) ||
+ (ec->use_adaptive_rx_coalesce) || (ec->use_adaptive_tx_coalesce) ||
+ (ec->pkt_rate_low) || (ec->rx_coalesce_usecs_low) ||
+ (ec->rx_max_coalesced_frames_low) || (ec->tx_coalesce_usecs_high) ||
+ (ec->tx_max_coalesced_frames_low) || (ec->pkt_rate_high) ||
+ (ec->tx_coalesce_usecs_low) || (ec->rx_coalesce_usecs_high) ||
+ (ec->rx_max_coalesced_frames_high) ||
+ (ec->tx_max_coalesced_frames_irq) ||
+ (ec->stats_block_coalesce_usecs) ||
+ (ec->tx_max_coalesced_frames_high) || (ec->rate_sample_interval))
+ return -EOPNOTSUPP;
+
+ /* No rx interrupts will be generated if both are zero */
+ if (ec->rx_coalesce_usecs == 0)
+ return -EINVAL;
+
+ /* No tx interrupts will be generated if both are zero */
+ if ((ec->tx_coalesce_usecs == 0) &&
+ (ec->tx_max_coalesced_frames == 0))
+ return -EINVAL;
+
+ if ((ec->tx_coalesce_usecs > STMMAC_COAL_TX_TIMER) ||
+ (ec->tx_max_coalesced_frames > STMMAC_TX_MAX_FRAMES))
+ return -EINVAL;
+
+ rx_riwt = stmmac_usec2riwt(ec->rx_coalesce_usecs, priv);
+
+ if ((rx_riwt > MAX_DMA_RIWT) || (rx_riwt < MIN_DMA_RIWT))
+ return -EINVAL;
+ else if (!priv->use_riwt)
+ return -EOPNOTSUPP;
+
+ /* Only copy relevant parameters, ignore all others. */
+ priv->tx_coal_frames = ec->tx_max_coalesced_frames;
+ priv->tx_coal_timer = ec->tx_coalesce_usecs;
+ priv->rx_riwt = rx_riwt;
+ priv->hw->dma->rx_watchdog(priv->ioaddr, priv->rx_riwt);
+
+ return 0;
+}
+
static const struct ethtool_ops stmmac_ethtool_ops = {
.begin = stmmac_check_if_running,
.get_drvinfo = stmmac_ethtool_getdrvinfo,
@@ -542,6 +625,8 @@ static const struct ethtool_ops stmmac_ethtool_ops = {
.set_eee = stmmac_ethtool_op_set_eee,
.get_sset_count = stmmac_get_sset_count,
.get_ts_info = ethtool_op_get_ts_info,
+ .get_coalesce = stmmac_get_coalesce,
+ .set_coalesce = stmmac_set_coalesce,
};
void stmmac_set_ethtool_ops(struct net_device *netdev)
diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
index 8e610a1..b0731e6 100644
--- a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
+++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
@@ -77,8 +77,6 @@
#define STMMAC_ALIGN(x) L1_CACHE_ALIGN(x)
#define JUMBO_LEN 9000
-#define STMMAC_TX_TM 40000
-#define STMMAC_TX_MAX_FRAMES 32 /* Max coalesced frame */
/* Module parameters */
#define TX_TIMEO 5000 /* default 5 seconds */
@@ -140,6 +138,8 @@ static int stmmac_init_fs(struct net_device *dev);
static void stmmac_exit_fs(void);
#endif
+#define STMMAC_COAL_TIMER(x) (jiffies + usecs_to_jiffies(x))
+
/**
* stmmac_verify_args - verify the driver parameters.
* Description: it verifies if some wrong parameter is passed to the driver.
@@ -996,9 +996,9 @@ static void stmmac_init_tx_coalesce(struct stmmac_priv *priv)
{
/* Set Tx coalesce parameters and timers */
priv->tx_coal_frames = STMMAC_TX_MAX_FRAMES;
- priv->tx_coal_timer = jiffies + usecs_to_jiffies(STMMAC_TX_TM);
+ priv->tx_coal_timer = STMMAC_COAL_TX_TIMER;
init_timer(&priv->txtimer);
- priv->txtimer.expires = priv->tx_coal_timer;
+ priv->txtimer.expires = STMMAC_COAL_TIMER(priv->tx_coal_timer);
priv->txtimer.data = (unsigned long)priv;
priv->txtimer.function = stmmac_txtimer;
add_timer(&priv->txtimer);
@@ -1118,11 +1118,10 @@ static int stmmac_open(struct net_device *dev)
stmmac_init_tx_coalesce(priv);
- if ((priv->use_riwt) && (priv->hw->dma->rx_watchdog))
- /* Program RX Watchdog register to the default values
- * FIXME: provide user value for RIWT
- */
- priv->hw->dma->rx_watchdog(priv->ioaddr, DEFAULT_DMA_RIWT);
+ if ((priv->use_riwt) && (priv->hw->dma->rx_watchdog)) {
+ priv->rx_riwt = MAX_DMA_RIWT;
+ priv->hw->dma->rx_watchdog(priv->ioaddr, MAX_DMA_RIWT);
+ }
napi_enable(&priv->napi);
@@ -1299,7 +1298,8 @@ static netdev_tx_t stmmac_xmit(struct sk_buff *skb, struct net_device *dev)
priv->xstats.tx_reset_ic_bit++;
TX_DBG("\t[entry %d]: tx_count_frames %d\n", entry,
priv->tx_count_frames);
- mod_timer(&priv->txtimer, priv->tx_coal_timer);
+ mod_timer(&priv->txtimer,
+ STMMAC_COAL_TIMER(priv->tx_coal_timer));
} else
priv->tx_count_frames = 0;
--
1.7.4.4
^ permalink raw reply related
* [net-next.git 8/8] stmmac: update the driver version to Sept_2012
From: Giuseppe CAVALLARO @ 2012-09-11 6:55 UTC (permalink / raw)
To: netdev; +Cc: bhutchings, davem, Giuseppe Cavallaro
In-Reply-To: <1347346514-23411-1-git-send-email-peppe.cavallaro@st.com>
Many new feauture have been introduced in the driver:
sysFS, Rx HW watchdog... so this patch updates the
driver's version.
Signed-off-by: Giuseppe Cavallaro <peppe.cavallaro@st.com>
---
drivers/net/ethernet/stmicro/stmmac/stmmac.h | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac.h b/drivers/net/ethernet/stmicro/stmmac/stmmac.h
index ad4f6b9..38662e1 100644
--- a/drivers/net/ethernet/stmicro/stmmac/stmmac.h
+++ b/drivers/net/ethernet/stmicro/stmmac/stmmac.h
@@ -24,7 +24,7 @@
#define __STMMAC_H__
#define STMMAC_RESOURCE_NAME "stmmaceth"
-#define DRV_MODULE_VERSION "March_2012"
+#define DRV_MODULE_VERSION "Sept_2012"
#include <linux/clk.h>
#include <linux/stmmac.h>
--
1.7.4.4
^ permalink raw reply related
* [net-next.git 7/8] stmmac: update the doc with new IRQ mitigation
From: Giuseppe CAVALLARO @ 2012-09-11 6:55 UTC (permalink / raw)
To: netdev; +Cc: bhutchings, davem, Giuseppe Cavallaro
In-Reply-To: <1347346514-23411-1-git-send-email-peppe.cavallaro@st.com>
This patch updates the stmmac.txt adding some information
about the new rx/tx mitigation schema adopted in the driver.
Signed-off-by: Giuseppe Cavallaro <peppe.cavallaro@st.com>
---
Documentation/networking/stmmac.txt | 28 +++++++++++++++-------------
1 files changed, 15 insertions(+), 13 deletions(-)
diff --git a/Documentation/networking/stmmac.txt b/Documentation/networking/stmmac.txt
index ef9ee71..f9fa6db 100644
--- a/Documentation/networking/stmmac.txt
+++ b/Documentation/networking/stmmac.txt
@@ -29,11 +29,9 @@ The kernel configuration option is STMMAC_ETH:
dma_txsize: DMA tx ring size;
buf_sz: DMA buffer size;
tc: control the HW FIFO threshold;
- tx_coe: Enable/Disable Tx Checksum Offload engine;
watchdog: transmit timeout (in milliseconds);
flow_ctrl: Flow control ability [on/off];
pause: Flow Control Pause Time;
- tmrate: timer period (only if timer optimisation is configured).
3) Command line options
Driver parameters can be also passed in command line by using:
@@ -60,17 +58,19 @@ Then the poll method will be scheduled at some future point.
The incoming packets are stored, by the DMA, in a list of pre-allocated socket
buffers in order to avoid the memcpy (Zero-copy).
-4.3) Timer-Driver Interrupt
-Instead of having the device that asynchronously notifies the frame receptions,
-the driver configures a timer to generate an interrupt at regular intervals.
-Based on the granularity of the timer, the frames that are received by the
-device will experience different levels of latency. Some NICs have dedicated
-timer device to perform this task. STMMAC can use either the RTC device or the
-TMU channel 2 on STLinux platforms.
-The timers frequency can be passed to the driver as parameter; when change it,
-take care of both hardware capability and network stability/performance impact.
-Several performance tests on STM platforms showed this optimisation allows to
-spare the CPU while having the maximum throughput.
+4.3) Interrupt Mitigation
+The driver is able to mitigate the number of its DMA interrupts
+using NAPI for the reception on chips older than the 3.50.
+New chips have an HW RX-Watchdog used for this mitigation.
+
+On Tx-side, the mitigation schema is based on a SW timer that calls the
+tx function (stmmac_tx) to reclaim the resource after transmitting the
+frames.
+Also there is another parameter (like a threshold) used to program
+the descriptors avoiding to set the interrupt on completion bit in
+when the frame is sent (xmit).
+
+Mitigation parameters can be tuned by ethtool.
4.4) WOL
Wake up on Lan feature through Magic and Unicast frames are supported for the
@@ -121,6 +121,7 @@ struct plat_stmmacenet_data {
int bugged_jumbo;
int pmt;
int force_sf_dma_mode;
+ int riwt_off;
void (*fix_mac_speed)(void *priv, unsigned int speed);
void (*bus_setup)(void __iomem *ioaddr);
int (*init)(struct platform_device *pdev);
@@ -156,6 +157,7 @@ Where:
o pmt: core has the embedded power module (optional).
o force_sf_dma_mode: force DMA to use the Store and Forward mode
instead of the Threshold.
+ o riwt_off: force to disable the RX watchdog feature and switch to NAPI mode.
o fix_mac_speed: this callback is used for modifying some syscfg registers
(on ST SoCs) according to the link speed negotiated by the
physical layer .
--
1.7.4.4
^ permalink raw reply related
* [net-next.git 6/8] stmmac: fix and review the rx irq path after adding new mitigation
From: Giuseppe CAVALLARO @ 2012-09-11 6:55 UTC (permalink / raw)
To: netdev; +Cc: bhutchings, davem, Giuseppe Cavallaro
In-Reply-To: <1347346514-23411-1-git-send-email-peppe.cavallaro@st.com>
After adopting the new mitigation approach I have
found a couple of problems in the rx irq path.
The enable/disable_dma_irq functions now are used for
handling the DMA_INTR_ENA_RIE bit in the DMA register 7.
So this patch masks the specific bit in this register.
Also these function names have been changed (to make clear
their meaning).
A new extra statistic field has been added to show the early
receive status in the interrupt handler as well. This has been
useful on debugging stage, indeed.
In the end, this patch also adds an extra check to avoid to call
napi_schedule when the DMA_INTR_ENA_RIE bit is disabled in the
Interrupt Mask register.
Signed-off-by: Giuseppe Cavallaro <peppe.cavallaro@st.com>
---
drivers/net/ethernet/stmicro/stmmac/common.h | 12 +++++---
.../net/ethernet/stmicro/stmmac/dwmac1000_dma.c | 4 +-
drivers/net/ethernet/stmicro/stmmac/dwmac100_dma.c | 4 +-
drivers/net/ethernet/stmicro/stmmac/dwmac_dma.h | 4 +-
drivers/net/ethernet/stmicro/stmmac/dwmac_lib.c | 27 ++++++++++++++-----
.../net/ethernet/stmicro/stmmac/stmmac_ethtool.c | 9 ++++--
drivers/net/ethernet/stmicro/stmmac/stmmac_main.c | 17 ++++--------
7 files changed, 45 insertions(+), 32 deletions(-)
diff --git a/drivers/net/ethernet/stmicro/stmmac/common.h b/drivers/net/ethernet/stmicro/stmmac/common.h
index b9033cc..9daa9df 100644
--- a/drivers/net/ethernet/stmicro/stmmac/common.h
+++ b/drivers/net/ethernet/stmicro/stmmac/common.h
@@ -85,7 +85,7 @@ struct stmmac_extra_stats {
unsigned long rx_missed_cntr;
unsigned long rx_overflow_cntr;
unsigned long rx_vlan;
- /* Tx/Rx IRQ errors */
+ /* Tx/Rx IRQ error info */
unsigned long tx_undeflow_irq;
unsigned long tx_process_stopped_irq;
unsigned long tx_jabber_irq;
@@ -95,7 +95,8 @@ struct stmmac_extra_stats {
unsigned long rx_watchdog_irq;
unsigned long tx_early_irq;
unsigned long fatal_bus_error_irq;
- /* Extra info */
+ /* Tx/Rx IRQ Events */
+ unsigned long rx_early_irq;
unsigned long threshold;
unsigned long tx_pkt_n;
unsigned long rx_pkt_n;
@@ -106,11 +107,12 @@ struct stmmac_extra_stats {
unsigned long txtimer;
unsigned long tx_clean;
unsigned long tx_reset_ic_bit;
+ unsigned long irq_receive_pmt_irq_n;
+ /* MMC info */
unsigned long mmc_tx_irq_n;
unsigned long mmc_rx_irq_n;
unsigned long mmc_rx_csum_offload_irq_n;
/* EEE */
- unsigned long irq_receive_pmt_irq_n;
unsigned long irq_tx_path_in_lpi_mode_n;
unsigned long irq_tx_path_exit_lpi_mode_n;
unsigned long irq_rx_path_in_lpi_mode_n;
@@ -302,8 +304,8 @@ struct stmmac_dma_ops {
void (*dma_diagnostic_fr) (void *data, struct stmmac_extra_stats *x,
void __iomem *ioaddr);
void (*enable_dma_transmission) (void __iomem *ioaddr);
- void (*enable_dma_irq) (void __iomem *ioaddr);
- void (*disable_dma_irq) (void __iomem *ioaddr);
+ void (*enable_rx_dma_irq) (void __iomem *ioaddr);
+ void (*disable_rx_dma_irq) (void __iomem *ioaddr);
void (*start_tx) (void __iomem *ioaddr);
void (*stop_tx) (void __iomem *ioaddr);
void (*start_rx) (void __iomem *ioaddr);
diff --git a/drivers/net/ethernet/stmicro/stmmac/dwmac1000_dma.c b/drivers/net/ethernet/stmicro/stmmac/dwmac1000_dma.c
index bf83c03..9804e82 100644
--- a/drivers/net/ethernet/stmicro/stmmac/dwmac1000_dma.c
+++ b/drivers/net/ethernet/stmicro/stmmac/dwmac1000_dma.c
@@ -184,8 +184,8 @@ const struct stmmac_dma_ops dwmac1000_dma_ops = {
.dump_regs = dwmac1000_dump_dma_regs,
.dma_mode = dwmac1000_dma_operation_mode,
.enable_dma_transmission = dwmac_enable_dma_transmission,
- .enable_dma_irq = dwmac_enable_dma_irq,
- .disable_dma_irq = dwmac_disable_dma_irq,
+ .enable_rx_dma_irq = dwmac_enable_rx_dma_irq,
+ .disable_rx_dma_irq = dwmac_disable_rx_dma_irq,
.start_tx = dwmac_dma_start_tx,
.stop_tx = dwmac_dma_stop_tx,
.start_rx = dwmac_dma_start_rx,
diff --git a/drivers/net/ethernet/stmicro/stmmac/dwmac100_dma.c b/drivers/net/ethernet/stmicro/stmmac/dwmac100_dma.c
index c2b4d55..81630c3 100644
--- a/drivers/net/ethernet/stmicro/stmmac/dwmac100_dma.c
+++ b/drivers/net/ethernet/stmicro/stmmac/dwmac100_dma.c
@@ -134,8 +134,8 @@ const struct stmmac_dma_ops dwmac100_dma_ops = {
.dma_mode = dwmac100_dma_operation_mode,
.dma_diagnostic_fr = dwmac100_dma_diagnostic_fr,
.enable_dma_transmission = dwmac_enable_dma_transmission,
- .enable_dma_irq = dwmac_enable_dma_irq,
- .disable_dma_irq = dwmac_disable_dma_irq,
+ .enable_rx_dma_irq = dwmac_enable_rx_dma_irq,
+ .disable_rx_dma_irq = dwmac_disable_rx_dma_irq,
.start_tx = dwmac_dma_start_tx,
.stop_tx = dwmac_dma_stop_tx,
.start_rx = dwmac_dma_start_rx,
diff --git a/drivers/net/ethernet/stmicro/stmmac/dwmac_dma.h b/drivers/net/ethernet/stmicro/stmmac/dwmac_dma.h
index 4eeff5d..b91c8cf 100644
--- a/drivers/net/ethernet/stmicro/stmmac/dwmac_dma.h
+++ b/drivers/net/ethernet/stmicro/stmmac/dwmac_dma.h
@@ -105,8 +105,8 @@
#define DMA_CONTROL_FTF 0x00100000 /* Flush transmit FIFO */
extern void dwmac_enable_dma_transmission(void __iomem *ioaddr);
-extern void dwmac_enable_dma_irq(void __iomem *ioaddr);
-extern void dwmac_disable_dma_irq(void __iomem *ioaddr);
+extern void dwmac_enable_rx_dma_irq(void __iomem *ioaddr);
+extern void dwmac_disable_rx_dma_irq(void __iomem *ioaddr);
extern void dwmac_dma_start_tx(void __iomem *ioaddr);
extern void dwmac_dma_stop_tx(void __iomem *ioaddr);
extern void dwmac_dma_start_rx(void __iomem *ioaddr);
diff --git a/drivers/net/ethernet/stmicro/stmmac/dwmac_lib.c b/drivers/net/ethernet/stmicro/stmmac/dwmac_lib.c
index 73766e6..a650019 100644
--- a/drivers/net/ethernet/stmicro/stmmac/dwmac_lib.c
+++ b/drivers/net/ethernet/stmicro/stmmac/dwmac_lib.c
@@ -39,14 +39,20 @@ void dwmac_enable_dma_transmission(void __iomem *ioaddr)
writel(1, ioaddr + DMA_XMT_POLL_DEMAND);
}
-void dwmac_enable_dma_irq(void __iomem *ioaddr)
+void dwmac_enable_rx_dma_irq(void __iomem *ioaddr)
{
- writel(DMA_INTR_DEFAULT_MASK, ioaddr + DMA_INTR_ENA);
+ u32 value = readl(ioaddr + DMA_INTR_ENA);
+
+ value |= DMA_INTR_ENA_RIE;
+ writel(value, ioaddr + DMA_INTR_ENA);
}
-void dwmac_disable_dma_irq(void __iomem *ioaddr)
+void dwmac_disable_rx_dma_irq(void __iomem *ioaddr)
{
- writel(0, ioaddr + DMA_INTR_ENA);
+ u32 value = readl(ioaddr + DMA_INTR_ENA);
+
+ value &= ~DMA_INTR_ENA_RIE;
+ writel(value, ioaddr + DMA_INTR_ENA);
}
void dwmac_dma_start_tx(void __iomem *ioaddr)
@@ -206,15 +212,22 @@ int dwmac_dma_interrupt(void __iomem *ioaddr,
/* TX/RX NORMAL interrupts */
if (intr_status & DMA_STATUS_NIS) {
x->normal_irq_n++;
- if (likely(intr_status & DMA_STATUS_RI))
- ret |= handle_rx;
- if (intr_status & (DMA_STATUS_TI))
+ if (likely(intr_status & DMA_STATUS_RI)) {
+ u32 value = readl(ioaddr + DMA_INTR_ENA);
+ /* to schedule NAPI on real RIE event. */
+ if (likely(value & DMA_INTR_ENA_RIE))
+ ret |= handle_rx;
+ }
+ if (intr_status & DMA_STATUS_TI)
ret |= handle_tx;
+ if (intr_status & DMA_STATUS_ERI)
+ x->rx_early_irq++;
}
/* Optional hardware blocks, interrupts should be disabled */
if (unlikely(intr_status &
(DMA_STATUS_GPI | DMA_STATUS_GMI | DMA_STATUS_GLI)))
pr_info("%s: unexpected status %08x\n", __func__, intr_status);
+
/* Clear the interrupt by writing a logic 1 to the CSR5[15-0] */
writel((intr_status & 0x1ffff), ioaddr + DMA_STATUS);
diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_ethtool.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_ethtool.c
index d27cc18..1ea5520 100644
--- a/drivers/net/ethernet/stmicro/stmmac/stmmac_ethtool.c
+++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_ethtool.c
@@ -76,7 +76,7 @@ static const struct stmmac_stats stmmac_gstrings_stats[] = {
STMMAC_STAT(rx_missed_cntr),
STMMAC_STAT(rx_overflow_cntr),
STMMAC_STAT(rx_vlan),
- /* Tx/Rx IRQ errors */
+ /* Tx/Rx IRQ error info */
STMMAC_STAT(tx_undeflow_irq),
STMMAC_STAT(tx_process_stopped_irq),
STMMAC_STAT(tx_jabber_irq),
@@ -86,7 +86,8 @@ static const struct stmmac_stats stmmac_gstrings_stats[] = {
STMMAC_STAT(rx_watchdog_irq),
STMMAC_STAT(tx_early_irq),
STMMAC_STAT(fatal_bus_error_irq),
- /* Extra info */
+ /* Tx/Rx IRQ Events */
+ STMMAC_STAT(rx_early_irq),
STMMAC_STAT(threshold),
STMMAC_STAT(tx_pkt_n),
STMMAC_STAT(rx_pkt_n),
@@ -97,10 +98,12 @@ static const struct stmmac_stats stmmac_gstrings_stats[] = {
STMMAC_STAT(txtimer),
STMMAC_STAT(tx_clean),
STMMAC_STAT(tx_reset_ic_bit),
+ STMMAC_STAT(irq_receive_pmt_irq_n),
+ /* MMC info */
STMMAC_STAT(mmc_tx_irq_n),
STMMAC_STAT(mmc_rx_irq_n),
STMMAC_STAT(mmc_rx_csum_offload_irq_n),
- STMMAC_STAT(irq_receive_pmt_irq_n),
+ /* EEE */
STMMAC_STAT(irq_tx_path_in_lpi_mode_n),
STMMAC_STAT(irq_tx_path_exit_lpi_mode_n),
STMMAC_STAT(irq_rx_path_in_lpi_mode_n),
diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
index b0731e6..0b37c6d 100644
--- a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
+++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
@@ -778,12 +778,12 @@ static void stmmac_tx(struct stmmac_priv *priv)
static inline void stmmac_enable_irq(struct stmmac_priv *priv)
{
- priv->hw->dma->enable_dma_irq(priv->ioaddr);
+ priv->hw->dma->enable_rx_dma_irq(priv->ioaddr);
}
static inline void stmmac_disable_irq(struct stmmac_priv *priv)
{
- priv->hw->dma->disable_dma_irq(priv->ioaddr);
+ priv->hw->dma->disable_rx_dma_irq(priv->ioaddr);
}
static void stmmac_txtimer(unsigned long data)
@@ -815,14 +815,6 @@ static void stmmac_tx_err(struct stmmac_priv *priv)
netif_wake_queue(priv->dev);
}
-static void stmmac_rx_work(struct stmmac_priv *priv)
-{
- if (likely(napi_schedule_prep(&priv->napi))) {
- stmmac_disable_irq(priv);
- __napi_schedule(&priv->napi);
- }
-}
-
static void stmmac_dma_interrupt(struct stmmac_priv *priv)
{
int status;
@@ -830,7 +822,10 @@ static void stmmac_dma_interrupt(struct stmmac_priv *priv)
status = priv->hw->dma->dma_interrupt(priv->ioaddr, &priv->xstats);
if (likely(status & handle_rx)) {
priv->xstats.rx_normal_irq_n++;
- stmmac_rx_work(priv);
+ if (likely(napi_schedule_prep(&priv->napi))) {
+ stmmac_disable_irq(priv);
+ __napi_schedule(&priv->napi);
+ }
}
if (likely(status & handle_tx)) {
priv->xstats.tx_normal_irq_n++;
--
1.7.4.4
^ permalink raw reply related
* [net-next.git 4/8 (V3)] stmmac: add Rx watchdog support to mitigate the DMA irqs
From: Giuseppe CAVALLARO @ 2012-09-11 6:55 UTC (permalink / raw)
To: netdev; +Cc: bhutchings, davem, Giuseppe Cavallaro
In-Reply-To: <1347346514-23411-1-git-send-email-peppe.cavallaro@st.com>
GMAC devices newer than databook 3.40 has an embedded timer
that can be used for mitigating the number of interrupts.
So this patch adds this optimizations.
At any rate, the Rx watchdog can be disable (on bugged HW) by
passing from the platform the riwt_off field.
In this implementation the rx timer stored in the Reg9 is fixed
to the max value. Next step will be to tune it via ethtool.
V2: added a platform parameter to force to disable the rx-watchdog
for example on new core where it is bugged.
V3: do not disable NAPI when Rx watchdog is used.
Signed-off-by: Giuseppe Cavallaro <peppe.cavallaro@st.com>
---
drivers/net/ethernet/stmicro/stmmac/common.h | 7 ++++
drivers/net/ethernet/stmicro/stmmac/dwmac1000.h | 3 --
.../net/ethernet/stmicro/stmmac/dwmac1000_dma.c | 6 +++
drivers/net/ethernet/stmicro/stmmac/dwmac_dma.h | 3 +-
drivers/net/ethernet/stmicro/stmmac/stmmac.h | 1 +
drivers/net/ethernet/stmicro/stmmac/stmmac_main.c | 36 +++++++++++++++----
include/linux/stmmac.h | 1 +
7 files changed, 45 insertions(+), 12 deletions(-)
diff --git a/drivers/net/ethernet/stmicro/stmmac/common.h b/drivers/net/ethernet/stmicro/stmmac/common.h
index 1d6bd3e..02eb2da 100644
--- a/drivers/net/ethernet/stmicro/stmmac/common.h
+++ b/drivers/net/ethernet/stmicro/stmmac/common.h
@@ -48,6 +48,10 @@
#define CHIP_DBG(fmt, args...) do { } while (0)
#endif
+/* Synopsys Core versions */
+#define DWMAC_CORE_3_40 0x34
+#define DWMAC_CORE_3_50 0x35
+
#undef FRAME_FILTER_DEBUG
/* #define FRAME_FILTER_DEBUG */
@@ -165,6 +169,7 @@ struct stmmac_extra_stats {
#define DMA_HW_FEAT_SAVLANINS 0x08000000 /* Source Addr or VLAN Insertion */
#define DMA_HW_FEAT_ACTPHYIF 0x70000000 /* Active/selected PHY interface */
#define DEFAULT_DMA_PBL 8
+#define DEFAULT_DMA_RIWT 0xff /* Max RI Watchdog Timer count */
enum rx_frame_status { /* IPC status */
good_frame = 0,
@@ -301,6 +306,8 @@ struct stmmac_dma_ops {
struct stmmac_extra_stats *x);
/* If supported then get the optional core features */
unsigned int (*get_hw_feature) (void __iomem *ioaddr);
+ /* Program the HW RX Watchdog */
+ void (*rx_watchdog) (void __iomem *ioaddr, u32 riwt);
};
struct stmmac_ops {
diff --git a/drivers/net/ethernet/stmicro/stmmac/dwmac1000.h b/drivers/net/ethernet/stmicro/stmmac/dwmac1000.h
index 0e4cace..7ad56af 100644
--- a/drivers/net/ethernet/stmicro/stmmac/dwmac1000.h
+++ b/drivers/net/ethernet/stmicro/stmmac/dwmac1000.h
@@ -230,8 +230,5 @@ enum rtc_control {
#define GMAC_MMC_TX_INTR 0x108
#define GMAC_MMC_RX_CSUM_OFFLOAD 0x208
-/* Synopsys Core versions */
-#define DWMAC_CORE_3_40 0x34
-
extern const struct stmmac_dma_ops dwmac1000_dma_ops;
#endif /* __DWMAC1000_H__ */
diff --git a/drivers/net/ethernet/stmicro/stmmac/dwmac1000_dma.c b/drivers/net/ethernet/stmicro/stmmac/dwmac1000_dma.c
index 0335000..bf83c03 100644
--- a/drivers/net/ethernet/stmicro/stmmac/dwmac1000_dma.c
+++ b/drivers/net/ethernet/stmicro/stmmac/dwmac1000_dma.c
@@ -174,6 +174,11 @@ static unsigned int dwmac1000_get_hw_feature(void __iomem *ioaddr)
return readl(ioaddr + DMA_HW_FEATURE);
}
+static void dwmac1000_rx_watchdog(void __iomem *ioaddr, u32 riwt)
+{
+ writel(riwt, ioaddr + DMA_RX_WATCHDOG);
+}
+
const struct stmmac_dma_ops dwmac1000_dma_ops = {
.init = dwmac1000_dma_init,
.dump_regs = dwmac1000_dump_dma_regs,
@@ -187,4 +192,5 @@ const struct stmmac_dma_ops dwmac1000_dma_ops = {
.stop_rx = dwmac_dma_stop_rx,
.dma_interrupt = dwmac_dma_interrupt,
.get_hw_feature = dwmac1000_get_hw_feature,
+ .rx_watchdog = dwmac1000_rx_watchdog,
};
diff --git a/drivers/net/ethernet/stmicro/stmmac/dwmac_dma.h b/drivers/net/ethernet/stmicro/stmmac/dwmac_dma.h
index e49c9a0..4eeff5d 100644
--- a/drivers/net/ethernet/stmicro/stmmac/dwmac_dma.h
+++ b/drivers/net/ethernet/stmicro/stmmac/dwmac_dma.h
@@ -35,7 +35,8 @@
#define DMA_CONTROL 0x00001018 /* Ctrl (Operational Mode) */
#define DMA_INTR_ENA 0x0000101c /* Interrupt Enable */
#define DMA_MISSED_FRAME_CTR 0x00001020 /* Missed Frame Counter */
-#define DMA_AXI_BUS_MODE 0x00001028 /* AXI Bus Mode */
+#define DMA_RX_WATCHDOG 0x00001024 /* Receive Int Watchdog Timer */
+#define DMA_AXI_BUS_MODE 0x00001028 /* AXI Bus Mode */
#define DMA_CUR_TX_BUF_ADDR 0x00001050 /* Current Host Tx Buffer */
#define DMA_CUR_RX_BUF_ADDR 0x00001054 /* Current Host Rx Buffer */
#define DMA_HW_FEATURE 0x00001058 /* HW Feature Register */
diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac.h b/drivers/net/ethernet/stmicro/stmmac/stmmac.h
index 0f5ab28..6901e3c 100644
--- a/drivers/net/ethernet/stmicro/stmmac/stmmac.h
+++ b/drivers/net/ethernet/stmicro/stmmac/stmmac.h
@@ -89,6 +89,7 @@ struct stmmac_priv {
int eee_active;
int tx_lpi_timer;
struct timer_list txtimer;
+ int use_riwt;
u32 tx_count_frames;
u32 tx_coal_frames;
u32 tx_coal_timer;
diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
index 3df3c3b..8e610a1 100644
--- a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
+++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
@@ -133,6 +133,7 @@ MODULE_PARM_DESC(eee_timer, "LPI tx expiration time in msec");
#define STMMAC_LPI_TIMER(x) (jiffies + msecs_to_jiffies(x))
static irqreturn_t stmmac_interrupt(int irq, void *dev_id);
+static int stmmac_rx(struct stmmac_priv *priv, int limit);
#ifdef CONFIG_STMMAC_DEBUG_FS
static int stmmac_init_fs(struct net_device *dev);
@@ -603,6 +604,8 @@ static void init_dma_desc_rings(struct net_device *dev)
priv->dirty_tx = 0;
priv->cur_tx = 0;
+ if (priv->use_riwt)
+ dis_ic = 1;
/* Clear the Rx/Tx descriptors */
priv->hw->desc->init_rx_desc(priv->dma_rx, rxsize, dis_ic);
priv->hw->desc->init_tx_desc(priv->dma_tx, txsize);
@@ -812,7 +815,7 @@ static void stmmac_tx_err(struct stmmac_priv *priv)
netif_wake_queue(priv->dev);
}
-static void stmmac_rx_schedule(struct stmmac_priv *priv)
+static void stmmac_rx_work(struct stmmac_priv *priv)
{
if (likely(napi_schedule_prep(&priv->napi))) {
stmmac_disable_irq(priv);
@@ -827,7 +830,7 @@ static void stmmac_dma_interrupt(struct stmmac_priv *priv)
status = priv->hw->dma->dma_interrupt(priv->ioaddr, &priv->xstats);
if (likely(status & handle_rx)) {
priv->xstats.rx_normal_irq_n++;
- stmmac_rx_schedule(priv);
+ stmmac_rx_work(priv);
}
if (likely(status & handle_tx)) {
priv->xstats.tx_normal_irq_n++;
@@ -1115,7 +1118,14 @@ static int stmmac_open(struct net_device *dev)
stmmac_init_tx_coalesce(priv);
+ if ((priv->use_riwt) && (priv->hw->dma->rx_watchdog))
+ /* Program RX Watchdog register to the default values
+ * FIXME: provide user value for RIWT
+ */
+ priv->hw->dma->rx_watchdog(priv->ioaddr, DEFAULT_DMA_RIWT);
+
napi_enable(&priv->napi);
+
skb_queue_head_init(&priv->rx_recycle);
netif_start_queue(dev);
@@ -1436,14 +1446,12 @@ static int stmmac_rx(struct stmmac_priv *priv, int limit)
#endif
skb->protocol = eth_type_trans(skb, priv->dev);
- if (unlikely(!priv->plat->rx_coe)) {
- /* No RX COE for old mac10/100 devices */
+ if (unlikely(!priv->plat->rx_coe))
skb_checksum_none_assert(skb);
- netif_receive_skb(skb);
- } else {
+ else
skb->ip_summed = CHECKSUM_UNNECESSARY;
- napi_gro_receive(&priv->napi, skb);
- }
+
+ napi_gro_receive(&priv->napi, skb);
priv->dev->stats.rx_packets++;
priv->dev->stats.rx_bytes += frame_len;
@@ -2013,6 +2021,15 @@ struct stmmac_priv *stmmac_dvr_probe(struct device *device,
if (flow_ctrl)
priv->flow_ctrl = FLOW_AUTO; /* RX/TX pause on */
+ /* Rx Watchdog is available in the COREs newer than the 3.40.
+ * In some case, for example on bugged HW this feature
+ * has to be disable and this can be done by passing the
+ * riwt_off field from the platform. */
+ if ((priv->synopsys_id >= DWMAC_CORE_3_50) && (!priv->plat->riwt_off)) {
+ priv->use_riwt = 1;
+ pr_info(" Enable RX Mitigation via HW Watchdog Timer\n");
+ }
+
netif_napi_add(ndev, &priv->napi, stmmac_poll, 64);
spin_lock_init(&priv->lock);
@@ -2104,6 +2121,9 @@ int stmmac_suspend(struct net_device *ndev)
netif_device_detach(ndev);
netif_stop_queue(ndev);
+ if (priv->use_riwt)
+ dis_ic = 1;
+
napi_disable(&priv->napi);
/* Stop TX/RX DMA */
diff --git a/include/linux/stmmac.h b/include/linux/stmmac.h
index a1547ea..de5b2f8 100644
--- a/include/linux/stmmac.h
+++ b/include/linux/stmmac.h
@@ -104,6 +104,7 @@ struct plat_stmmacenet_data {
int bugged_jumbo;
int pmt;
int force_sf_dma_mode;
+ int riwt_off;
void (*fix_mac_speed)(void *priv, unsigned int speed);
void (*bus_setup)(void __iomem *ioaddr);
int (*init)(struct platform_device *pdev);
--
1.7.4.4
^ permalink raw reply related
* [net-next.git 3/8 (V2)] stmmac: add the initial tx coalesce schema
From: Giuseppe CAVALLARO @ 2012-09-11 6:55 UTC (permalink / raw)
To: netdev; +Cc: bhutchings, davem, Giuseppe Cavallaro
In-Reply-To: <1347346514-23411-1-git-send-email-peppe.cavallaro@st.com>
This patch adds a new schema used for mitigating the
number of transmit interrupts.
It is based on a sw timer and a threshold value.
The timer is used to periodically call the stmmac_tx
function that can be invoked by the ISR but only for
the descriptors where the interrupt on completion
field has been set. This is tuned by a threshold.
Next step is to add the ability to tune these coalesce
values by ethtool.
Till now I have put a default that showed a real gain
on all the platforms ARM/SH4 where I performed benchmarks.
V2: review the logic to manage the IC bit in the TDESC
that was bugged because it didn't take care about the
fragments. Also fix the tx_count_frames that has not to be
limited to TX DMA ring. Thanks to Ben Hutchings.
Signed-off-by: Giuseppe Cavallaro <peppe.cavallaro@st.com>
---
drivers/net/ethernet/stmicro/stmmac/common.h | 8 +-
drivers/net/ethernet/stmicro/stmmac/stmmac.h | 4 +
.../net/ethernet/stmicro/stmmac/stmmac_ethtool.c | 9 +-
drivers/net/ethernet/stmicro/stmmac/stmmac_main.c | 92 ++++++++++++--------
4 files changed, 69 insertions(+), 44 deletions(-)
diff --git a/drivers/net/ethernet/stmicro/stmmac/common.h b/drivers/net/ethernet/stmicro/stmmac/common.h
index bd32fe6..1d6bd3e 100644
--- a/drivers/net/ethernet/stmicro/stmmac/common.h
+++ b/drivers/net/ethernet/stmicro/stmmac/common.h
@@ -95,11 +95,13 @@ struct stmmac_extra_stats {
unsigned long threshold;
unsigned long tx_pkt_n;
unsigned long rx_pkt_n;
- unsigned long rx_napi_poll;
+ unsigned long normal_irq_n;
unsigned long rx_normal_irq_n;
+ unsigned long rx_napi_poll;
unsigned long tx_normal_irq_n;
- unsigned long sched_timer_n;
- unsigned long normal_irq_n;
+ unsigned long txtimer;
+ unsigned long tx_clean;
+ unsigned long tx_reset_ic_bit;
unsigned long mmc_tx_irq_n;
unsigned long mmc_rx_irq_n;
unsigned long mmc_rx_csum_offload_irq_n;
diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac.h b/drivers/net/ethernet/stmicro/stmmac/stmmac.h
index 9f35769..0f5ab28 100644
--- a/drivers/net/ethernet/stmicro/stmmac/stmmac.h
+++ b/drivers/net/ethernet/stmicro/stmmac/stmmac.h
@@ -88,6 +88,10 @@ struct stmmac_priv {
int eee_enabled;
int eee_active;
int tx_lpi_timer;
+ struct timer_list txtimer;
+ u32 tx_count_frames;
+ u32 tx_coal_frames;
+ u32 tx_coal_timer;
};
extern int phyaddr;
diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_ethtool.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_ethtool.c
index 505fe71..48ad0bc 100644
--- a/drivers/net/ethernet/stmicro/stmmac/stmmac_ethtool.c
+++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_ethtool.c
@@ -90,12 +90,13 @@ static const struct stmmac_stats stmmac_gstrings_stats[] = {
STMMAC_STAT(threshold),
STMMAC_STAT(tx_pkt_n),
STMMAC_STAT(rx_pkt_n),
- STMMAC_STAT(rx_napi_poll),
+ STMMAC_STAT(normal_irq_n),
STMMAC_STAT(rx_normal_irq_n),
+ STMMAC_STAT(rx_napi_poll),
STMMAC_STAT(tx_normal_irq_n),
- STMMAC_STAT(sched_timer_n),
- STMMAC_STAT(normal_irq_n),
- STMMAC_STAT(normal_irq_n),
+ STMMAC_STAT(txtimer),
+ STMMAC_STAT(tx_clean),
+ STMMAC_STAT(tx_reset_ic_bit),
STMMAC_STAT(mmc_tx_irq_n),
STMMAC_STAT(mmc_rx_irq_n),
STMMAC_STAT(mmc_rx_csum_offload_irq_n),
diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
index 8e1e53e..3df3c3b 100644
--- a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
+++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
@@ -77,6 +77,8 @@
#define STMMAC_ALIGN(x) L1_CACHE_ALIGN(x)
#define JUMBO_LEN 9000
+#define STMMAC_TX_TM 40000
+#define STMMAC_TX_MAX_FRAMES 32 /* Max coalesced frame */
/* Module parameters */
#define TX_TIMEO 5000 /* default 5 seconds */
@@ -695,8 +697,11 @@ static void stmmac_dma_operation_mode(struct stmmac_priv *priv)
static void stmmac_tx(struct stmmac_priv *priv)
{
unsigned int txsize = priv->dma_tx_size;
+ unsigned long flags;
+
+ spin_lock_irqsave(&priv->tx_lock, flags);
- spin_lock(&priv->tx_lock);
+ priv->xstats.tx_clean++;
while (priv->dirty_tx != priv->cur_tx) {
int last;
@@ -741,7 +746,7 @@ static void stmmac_tx(struct stmmac_priv *priv)
skb_recycle_check(skb, priv->dma_buf_sz))
__skb_queue_head(&priv->rx_recycle, skb);
else
- dev_kfree_skb(skb);
+ dev_kfree_skb_any(skb);
priv->tx_skbuff[entry] = NULL;
}
@@ -765,7 +770,7 @@ static void stmmac_tx(struct stmmac_priv *priv)
stmmac_enable_eee_mode(priv);
mod_timer(&priv->eee_ctrl_timer, STMMAC_LPI_TIMER(eee_timer));
}
- spin_unlock(&priv->tx_lock);
+ spin_unlock_irqrestore(&priv->tx_lock, flags);
}
static inline void stmmac_enable_irq(struct stmmac_priv *priv)
@@ -778,29 +783,12 @@ static inline void stmmac_disable_irq(struct stmmac_priv *priv)
priv->hw->dma->disable_dma_irq(priv->ioaddr);
}
-static int stmmac_has_work(struct stmmac_priv *priv)
+static void stmmac_txtimer(unsigned long data)
{
- unsigned int has_work = 0;
- int rxret, tx_work = 0;
-
- rxret = priv->hw->desc->get_rx_owner(priv->dma_rx +
- (priv->cur_rx % priv->dma_rx_size));
-
- if (priv->dirty_tx != priv->cur_tx)
- tx_work = 1;
-
- if (likely(!rxret || tx_work))
- has_work = 1;
+ struct stmmac_priv *priv = (struct stmmac_priv *)data;
- return has_work;
-}
-
-static inline void _stmmac_schedule(struct stmmac_priv *priv)
-{
- if (likely(stmmac_has_work(priv))) {
- stmmac_disable_irq(priv);
- napi_schedule(&priv->napi);
- }
+ priv->xstats.txtimer++;
+ stmmac_tx(priv);
}
/**
@@ -824,7 +812,7 @@ static void stmmac_tx_err(struct stmmac_priv *priv)
netif_wake_queue(priv->dev);
}
-static inline void stmmac_rx_schedule(struct stmmac_priv *priv)
+static void stmmac_rx_schedule(struct stmmac_priv *priv)
{
if (likely(napi_schedule_prep(&priv->napi))) {
stmmac_disable_irq(priv);
@@ -1001,6 +989,18 @@ static int stmmac_init_dma_engine(struct stmmac_priv *priv)
priv->dma_rx_phy);
}
+static void stmmac_init_tx_coalesce(struct stmmac_priv *priv)
+{
+ /* Set Tx coalesce parameters and timers */
+ priv->tx_coal_frames = STMMAC_TX_MAX_FRAMES;
+ priv->tx_coal_timer = jiffies + usecs_to_jiffies(STMMAC_TX_TM);
+ init_timer(&priv->txtimer);
+ priv->txtimer.expires = priv->tx_coal_timer;
+ priv->txtimer.data = (unsigned long)priv;
+ priv->txtimer.function = stmmac_txtimer;
+ add_timer(&priv->txtimer);
+}
+
/**
* stmmac_open - open entry point of the driver
* @dev : pointer to the device structure.
@@ -1113,6 +1113,8 @@ static int stmmac_open(struct net_device *dev)
priv->tx_lpi_timer = STMMAC_DEFAULT_TWT_LS_TIMER;
priv->eee_enabled = stmmac_eee_init(priv);
+ stmmac_init_tx_coalesce(priv);
+
napi_enable(&priv->napi);
skb_queue_head_init(&priv->rx_recycle);
netif_start_queue(dev);
@@ -1160,6 +1162,8 @@ static int stmmac_release(struct net_device *dev)
napi_disable(&priv->napi);
skb_queue_purge(&priv->rx_recycle);
+ del_timer_sync(&priv->txtimer);
+
/* Free the IRQ lines */
free_irq(dev->irq, dev);
if (priv->wol_irq != dev->irq)
@@ -1202,6 +1206,7 @@ static netdev_tx_t stmmac_xmit(struct sk_buff *skb, struct net_device *dev)
int nfrags = skb_shinfo(skb)->nr_frags;
struct dma_desc *desc, *first;
unsigned int nopaged_len = skb_headlen(skb);
+ unsigned long flags;
if (unlikely(stmmac_tx_avail(priv) < nfrags + 1)) {
if (!netif_queue_stopped(dev)) {
@@ -1213,7 +1218,7 @@ static netdev_tx_t stmmac_xmit(struct sk_buff *skb, struct net_device *dev)
return NETDEV_TX_BUSY;
}
- spin_lock(&priv->tx_lock);
+ spin_lock_irqsave(&priv->tx_lock, flags);
if (priv->tx_path_in_lpi_mode)
stmmac_disable_eee_mode(priv);
@@ -1222,11 +1227,13 @@ static netdev_tx_t stmmac_xmit(struct sk_buff *skb, struct net_device *dev)
#ifdef STMMAC_XMIT_DEBUG
if ((skb->len > ETH_FRAME_LEN) || nfrags)
- pr_info("stmmac xmit:\n"
- "\tskb addr %p - len: %d - nopaged_len: %d\n"
- "\tn_frags: %d - ip_summed: %d - %s gso\n",
- skb, skb->len, nopaged_len, nfrags, skb->ip_summed,
- !skb_is_gso(skb) ? "isn't" : "is");
+ pr_debug("stmmac xmit: [entry %d]\n"
+ "\tskb addr %p - len: %d - nopaged_len: %d\n"
+ "\tn_frags: %d - ip_summed: %d - %s gso\n"
+ "\ttx_count_frames %d\n", entry,
+ skb, skb->len, nopaged_len, nfrags, skb->ip_summed,
+ !skb_is_gso(skb) ? "isn't" : "is",
+ priv->tx_count_frames);
#endif
csum_insertion = (skb->ip_summed == CHECKSUM_PARTIAL);
@@ -1236,9 +1243,9 @@ static netdev_tx_t stmmac_xmit(struct sk_buff *skb, struct net_device *dev)
#ifdef STMMAC_XMIT_DEBUG
if ((nfrags > 0) || (skb->len > ETH_FRAME_LEN))
- pr_debug("stmmac xmit: skb len: %d, nopaged_len: %d,\n"
- "\t\tn_frags: %d, ip_summed: %d\n",
- skb->len, nopaged_len, nfrags, skb->ip_summed);
+ pr_debug("\tskb len: %d, nopaged_len: %d,\n"
+ "\t\tn_frags: %d, ip_summed: %d\n",
+ skb->len, nopaged_len, nfrags, skb->ip_summed);
#endif
priv->tx_skbuff[entry] = skb;
@@ -1269,10 +1276,22 @@ static netdev_tx_t stmmac_xmit(struct sk_buff *skb, struct net_device *dev)
wmb();
}
- /* Interrupt on completition only for the latest segment */
+ /* Finalize the latest segment. */
priv->hw->desc->close_tx_desc(desc);
wmb();
+ /* According to the coalesce parameter the IC bit for the latest
+ * segment could be reset and the timer re-started to invoke the
+ * stmmac_tx function. This approach takes care about the fragments. */
+ priv->tx_count_frames += nfrags + 1;
+ if (priv->tx_coal_frames > priv->tx_count_frames) {
+ priv->hw->desc->clear_tx_ic(desc);
+ priv->xstats.tx_reset_ic_bit++;
+ TX_DBG("\t[entry %d]: tx_count_frames %d\n", entry,
+ priv->tx_count_frames);
+ mod_timer(&priv->txtimer, priv->tx_coal_timer);
+ } else
+ priv->tx_count_frames = 0;
/* To avoid raise condition */
priv->hw->desc->set_tx_owner(first);
@@ -1302,7 +1321,7 @@ static netdev_tx_t stmmac_xmit(struct sk_buff *skb, struct net_device *dev)
priv->hw->dma->enable_dma_transmission(priv->ioaddr);
- spin_unlock(&priv->tx_lock);
+ spin_unlock_irqrestore(&priv->tx_lock, flags);
return NETDEV_TX_OK;
}
@@ -1447,7 +1466,6 @@ static int stmmac_rx(struct stmmac_priv *priv, int limit)
* all interfaces.
* Description :
* This function implements the the reception process.
- * Also it runs the TX completion thread
*/
static int stmmac_poll(struct napi_struct *napi, int budget)
{
--
1.7.4.4
^ permalink raw reply related
* [net-next.git 2/8 (V2)] stmmac: manage tx clean out of rx_poll
From: Giuseppe CAVALLARO @ 2012-09-11 6:55 UTC (permalink / raw)
To: netdev; +Cc: bhutchings, davem, Giuseppe Cavallaro
In-Reply-To: <1347346514-23411-1-git-send-email-peppe.cavallaro@st.com>
This patch is to invoke the stmmac_tx (tx handler)
out of the NAPI poll method.
This will make easier the next step to add the new
mitigation schema.
Also the patch enhances the ethtool to report some
stats for normal TX and RX IRQs.
V2: use status & handle_rx/tx
Signed-off-by: Giuseppe Cavallaro <peppe.cavallaro@st.com>
---
drivers/net/ethernet/stmicro/stmmac/common.h | 13 +++++++----
drivers/net/ethernet/stmicro/stmmac/dwmac_lib.c | 7 +++--
.../net/ethernet/stmicro/stmmac/stmmac_ethtool.c | 4 ++-
drivers/net/ethernet/stmicro/stmmac/stmmac_main.c | 22 ++++++++++++++-----
4 files changed, 31 insertions(+), 15 deletions(-)
diff --git a/drivers/net/ethernet/stmicro/stmmac/common.h b/drivers/net/ethernet/stmicro/stmmac/common.h
index 719be39..bd32fe6 100644
--- a/drivers/net/ethernet/stmicro/stmmac/common.h
+++ b/drivers/net/ethernet/stmicro/stmmac/common.h
@@ -95,7 +95,9 @@ struct stmmac_extra_stats {
unsigned long threshold;
unsigned long tx_pkt_n;
unsigned long rx_pkt_n;
- unsigned long poll_n;
+ unsigned long rx_napi_poll;
+ unsigned long rx_normal_irq_n;
+ unsigned long tx_normal_irq_n;
unsigned long sched_timer_n;
unsigned long normal_irq_n;
unsigned long mmc_tx_irq_n;
@@ -169,10 +171,11 @@ enum rx_frame_status { /* IPC status */
llc_snap = 4,
};
-enum tx_dma_irq_status {
- tx_hard_error = 1,
- tx_hard_error_bump_tc = 2,
- handle_tx_rx = 3,
+enum dma_irq_status {
+ tx_hard_error = 0x1,
+ tx_hard_error_bump_tc = 0x2,
+ handle_rx = 0x4,
+ handle_tx = 0x8,
};
enum core_specific_irq_mask {
diff --git a/drivers/net/ethernet/stmicro/stmmac/dwmac_lib.c b/drivers/net/ethernet/stmicro/stmmac/dwmac_lib.c
index 4e0e18a..73766e6 100644
--- a/drivers/net/ethernet/stmicro/stmmac/dwmac_lib.c
+++ b/drivers/net/ethernet/stmicro/stmmac/dwmac_lib.c
@@ -206,9 +206,10 @@ int dwmac_dma_interrupt(void __iomem *ioaddr,
/* TX/RX NORMAL interrupts */
if (intr_status & DMA_STATUS_NIS) {
x->normal_irq_n++;
- if (likely((intr_status & DMA_STATUS_RI) ||
- (intr_status & (DMA_STATUS_TI))))
- ret = handle_tx_rx;
+ if (likely(intr_status & DMA_STATUS_RI))
+ ret |= handle_rx;
+ if (intr_status & (DMA_STATUS_TI))
+ ret |= handle_tx;
}
/* Optional hardware blocks, interrupts should be disabled */
if (unlikely(intr_status &
diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_ethtool.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_ethtool.c
index 76fd61a..505fe71 100644
--- a/drivers/net/ethernet/stmicro/stmmac/stmmac_ethtool.c
+++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_ethtool.c
@@ -90,7 +90,9 @@ static const struct stmmac_stats stmmac_gstrings_stats[] = {
STMMAC_STAT(threshold),
STMMAC_STAT(tx_pkt_n),
STMMAC_STAT(rx_pkt_n),
- STMMAC_STAT(poll_n),
+ STMMAC_STAT(rx_napi_poll),
+ STMMAC_STAT(rx_normal_irq_n),
+ STMMAC_STAT(tx_normal_irq_n),
STMMAC_STAT(sched_timer_n),
STMMAC_STAT(normal_irq_n),
STMMAC_STAT(normal_irq_n),
diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
index c8985f3..8e1e53e 100644
--- a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
+++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
@@ -824,16 +824,27 @@ static void stmmac_tx_err(struct stmmac_priv *priv)
netif_wake_queue(priv->dev);
}
+static inline void stmmac_rx_schedule(struct stmmac_priv *priv)
+{
+ if (likely(napi_schedule_prep(&priv->napi))) {
+ stmmac_disable_irq(priv);
+ __napi_schedule(&priv->napi);
+ }
+}
static void stmmac_dma_interrupt(struct stmmac_priv *priv)
{
int status;
status = priv->hw->dma->dma_interrupt(priv->ioaddr, &priv->xstats);
- if (likely(status == handle_tx_rx))
- _stmmac_schedule(priv);
-
- else if (unlikely(status == tx_hard_error_bump_tc)) {
+ if (likely(status & handle_rx)) {
+ priv->xstats.rx_normal_irq_n++;
+ stmmac_rx_schedule(priv);
+ }
+ if (likely(status & handle_tx)) {
+ priv->xstats.tx_normal_irq_n++;
+ stmmac_tx(priv);
+ } else if (unlikely(status & tx_hard_error_bump_tc)) {
/* Try to bump up the dma threshold on this failure */
if (unlikely(tc != SF_DMA_MODE) && (tc <= 256)) {
tc += 64;
@@ -1443,8 +1454,7 @@ static int stmmac_poll(struct napi_struct *napi, int budget)
struct stmmac_priv *priv = container_of(napi, struct stmmac_priv, napi);
int work_done = 0;
- priv->xstats.poll_n++;
- stmmac_tx(priv);
+ priv->xstats.rx_napi_poll++;
work_done = stmmac_rx(priv, budget);
if (work_done < budget) {
--
1.7.4.4
^ permalink raw reply related
* [net-next.git 1/8] stmmac: remove dead code for TIMER
From: Giuseppe CAVALLARO @ 2012-09-11 6:55 UTC (permalink / raw)
To: netdev; +Cc: bhutchings, davem, Giuseppe Cavallaro
In-Reply-To: <1347346514-23411-1-git-send-email-peppe.cavallaro@st.com>
TIMER option is not longer supported and this
code can be considered dead for this driver in
the new kernel series.
In fact, It was not updated at all and never used.
Signed-off-by: Giuseppe Cavallaro <peppe.cavallaro@st.com>
---
drivers/net/ethernet/stmicro/stmmac/Kconfig | 25 ----
drivers/net/ethernet/stmicro/stmmac/Makefile | 1 -
drivers/net/ethernet/stmicro/stmmac/stmmac.h | 6 -
drivers/net/ethernet/stmicro/stmmac/stmmac_main.c | 101 +--------------
drivers/net/ethernet/stmicro/stmmac/stmmac_timer.c | 134 --------------------
drivers/net/ethernet/stmicro/stmmac/stmmac_timer.h | 46 -------
6 files changed, 3 insertions(+), 310 deletions(-)
delete mode 100644 drivers/net/ethernet/stmicro/stmmac/stmmac_timer.c
delete mode 100644 drivers/net/ethernet/stmicro/stmmac/stmmac_timer.h
diff --git a/drivers/net/ethernet/stmicro/stmmac/Kconfig b/drivers/net/ethernet/stmicro/stmmac/Kconfig
index 9f44827..1164930 100644
--- a/drivers/net/ethernet/stmicro/stmmac/Kconfig
+++ b/drivers/net/ethernet/stmicro/stmmac/Kconfig
@@ -54,31 +54,6 @@ config STMMAC_DA
By default, the DMA arbitration scheme is based on Round-robin
(rx:tx priority is 1:1).
-config STMMAC_TIMER
- bool "STMMAC Timer optimisation"
- default n
- depends on RTC_HCTOSYS_DEVICE
- ---help---
- Use an external timer for mitigating the number of network
- interrupts. Currently, for SH architectures, it is possible
- to use the TMU channel 2 and the SH-RTC device.
-
-choice
- prompt "Select Timer device"
- depends on STMMAC_TIMER
-
-config STMMAC_TMU_TIMER
- bool "TMU channel 2"
- depends on CPU_SH4
- ---help---
-
-config STMMAC_RTC_TIMER
- bool "Real time clock"
- depends on RTC_CLASS
- ---help---
-
-endchoice
-
choice
prompt "Select the DMA TX/RX descriptor operating modes"
depends on STMMAC_ETH
diff --git a/drivers/net/ethernet/stmicro/stmmac/Makefile b/drivers/net/ethernet/stmicro/stmmac/Makefile
index bc965ac..c8e8ea6 100644
--- a/drivers/net/ethernet/stmicro/stmmac/Makefile
+++ b/drivers/net/ethernet/stmicro/stmmac/Makefile
@@ -1,5 +1,4 @@
obj-$(CONFIG_STMMAC_ETH) += stmmac.o
-stmmac-$(CONFIG_STMMAC_TIMER) += stmmac_timer.o
stmmac-$(CONFIG_STMMAC_RING) += ring_mode.o
stmmac-$(CONFIG_STMMAC_CHAINED) += chain_mode.o
stmmac-$(CONFIG_STMMAC_PLATFORM) += stmmac_platform.o
diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac.h b/drivers/net/ethernet/stmicro/stmmac/stmmac.h
index e872e1d..9f35769 100644
--- a/drivers/net/ethernet/stmicro/stmmac/stmmac.h
+++ b/drivers/net/ethernet/stmicro/stmmac/stmmac.h
@@ -31,9 +31,6 @@
#include <linux/phy.h>
#include <linux/pci.h>
#include "common.h"
-#ifdef CONFIG_STMMAC_TIMER
-#include "stmmac_timer.h"
-#endif
struct stmmac_priv {
/* Frequently used values are kept adjacent for cache effect */
@@ -78,9 +75,6 @@ struct stmmac_priv {
spinlock_t tx_lock;
int wolopts;
int wol_irq;
-#ifdef CONFIG_STMMAC_TIMER
- struct stmmac_timer *tm;
-#endif
struct plat_stmmacenet_data *plat;
struct stmmac_counters mmc;
struct dma_features dma_cap;
diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
index c136162..c8985f3 100644
--- a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
+++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
@@ -115,16 +115,6 @@ static int tc = TC_DEFAULT;
module_param(tc, int, S_IRUGO | S_IWUSR);
MODULE_PARM_DESC(tc, "DMA threshold control value");
-/* Pay attention to tune this parameter; take care of both
- * hardware capability and network stabitily/performance impact.
- * Many tests showed that ~4ms latency seems to be good enough. */
-#ifdef CONFIG_STMMAC_TIMER
-#define DEFAULT_PERIODIC_RATE 256
-static int tmrate = DEFAULT_PERIODIC_RATE;
-module_param(tmrate, int, S_IRUGO | S_IWUSR);
-MODULE_PARM_DESC(tmrate, "External timer freq. (default: 256Hz)");
-#endif
-
#define DMA_BUFFER_SIZE BUF_SIZE_2KiB
static int buf_sz = DMA_BUFFER_SIZE;
module_param(buf_sz, int, S_IRUGO | S_IWUSR);
@@ -536,12 +526,6 @@ static void init_dma_desc_rings(struct net_device *dev)
else
bfsize = stmmac_set_bfsize(dev->mtu, priv->dma_buf_sz);
-#ifdef CONFIG_STMMAC_TIMER
- /* Disable interrupts on completion for the reception if timer is on */
- if (likely(priv->tm->enable))
- dis_ic = 1;
-#endif
-
DBG(probe, INFO, "stmmac: txsize %d, rxsize %d, bfsize %d\n",
txsize, rxsize, bfsize);
@@ -786,22 +770,12 @@ static void stmmac_tx(struct stmmac_priv *priv)
static inline void stmmac_enable_irq(struct stmmac_priv *priv)
{
-#ifdef CONFIG_STMMAC_TIMER
- if (likely(priv->tm->enable))
- priv->tm->timer_start(tmrate);
- else
-#endif
- priv->hw->dma->enable_dma_irq(priv->ioaddr);
+ priv->hw->dma->enable_dma_irq(priv->ioaddr);
}
static inline void stmmac_disable_irq(struct stmmac_priv *priv)
{
-#ifdef CONFIG_STMMAC_TIMER
- if (likely(priv->tm->enable))
- priv->tm->timer_stop();
- else
-#endif
- priv->hw->dma->disable_dma_irq(priv->ioaddr);
+ priv->hw->dma->disable_dma_irq(priv->ioaddr);
}
static int stmmac_has_work(struct stmmac_priv *priv)
@@ -829,25 +803,6 @@ static inline void _stmmac_schedule(struct stmmac_priv *priv)
}
}
-#ifdef CONFIG_STMMAC_TIMER
-void stmmac_schedule(struct net_device *dev)
-{
- struct stmmac_priv *priv = netdev_priv(dev);
-
- priv->xstats.sched_timer_n++;
-
- _stmmac_schedule(priv);
-}
-
-static void stmmac_no_timer_started(unsigned int x)
-{;
-};
-
-static void stmmac_no_timer_stopped(void)
-{;
-};
-#endif
-
/**
* stmmac_tx_err:
* @priv: pointer to the private device structure
@@ -1049,23 +1004,6 @@ static int stmmac_open(struct net_device *dev)
struct stmmac_priv *priv = netdev_priv(dev);
int ret;
-#ifdef CONFIG_STMMAC_TIMER
- priv->tm = kzalloc(sizeof(struct stmmac_timer *), GFP_KERNEL);
- if (unlikely(priv->tm == NULL))
- return -ENOMEM;
-
- priv->tm->freq = tmrate;
-
- /* Test if the external timer can be actually used.
- * In case of failure continue without timer. */
- if (unlikely((stmmac_open_ext_timer(dev, priv->tm)) < 0)) {
- pr_warning("stmmaceth: cannot attach the external timer.\n");
- priv->tm->freq = 0;
- priv->tm->timer_start = stmmac_no_timer_started;
- priv->tm->timer_stop = stmmac_no_timer_stopped;
- } else
- priv->tm->enable = 1;
-#endif
clk_enable(priv->stmmac_clk);
stmmac_check_ether_addr(priv);
@@ -1152,10 +1090,6 @@ static int stmmac_open(struct net_device *dev)
priv->hw->dma->start_tx(priv->ioaddr);
priv->hw->dma->start_rx(priv->ioaddr);
-#ifdef CONFIG_STMMAC_TIMER
- priv->tm->timer_start(tmrate);
-#endif
-
/* Dump DMA/MAC registers */
if (netif_msg_hw(priv)) {
priv->hw->mac->dump_regs(priv->ioaddr);
@@ -1182,9 +1116,6 @@ open_error_wolirq:
free_irq(dev->irq, dev);
open_error:
-#ifdef CONFIG_STMMAC_TIMER
- kfree(priv->tm);
-#endif
if (priv->phydev)
phy_disconnect(priv->phydev);
@@ -1215,12 +1146,6 @@ static int stmmac_release(struct net_device *dev)
netif_stop_queue(dev);
-#ifdef CONFIG_STMMAC_TIMER
- /* Stop and release the timer */
- stmmac_close_ext_timer();
- if (priv->tm != NULL)
- kfree(priv->tm);
-#endif
napi_disable(&priv->napi);
skb_queue_purge(&priv->rx_recycle);
@@ -1336,12 +1261,6 @@ static netdev_tx_t stmmac_xmit(struct sk_buff *skb, struct net_device *dev)
/* Interrupt on completition only for the latest segment */
priv->hw->desc->close_tx_desc(desc);
-#ifdef CONFIG_STMMAC_TIMER
- /* Clean IC while using timer */
- if (likely(priv->tm->enable))
- priv->hw->desc->clear_tx_ic(desc);
-#endif
-
wmb();
/* To avoid raise condition */
@@ -1539,7 +1458,7 @@ static int stmmac_poll(struct napi_struct *napi, int budget)
* stmmac_tx_timeout
* @dev : Pointer to net device structure
* Description: this function is called when a packet transmission fails to
- * complete within a reasonable tmrate. The driver will mark the error in the
+ * complete within a reasonable time. The driver will mark the error in the
* netdev structure and arrange for the device to be reset to a sane state
* in order to transmit a new packet.
*/
@@ -2157,11 +2076,6 @@ int stmmac_suspend(struct net_device *ndev)
netif_device_detach(ndev);
netif_stop_queue(ndev);
-#ifdef CONFIG_STMMAC_TIMER
- priv->tm->timer_stop();
- if (likely(priv->tm->enable))
- dis_ic = 1;
-#endif
napi_disable(&priv->napi);
/* Stop TX/RX DMA */
@@ -2212,10 +2126,6 @@ int stmmac_resume(struct net_device *ndev)
priv->hw->dma->start_tx(priv->ioaddr);
priv->hw->dma->start_rx(priv->ioaddr);
-#ifdef CONFIG_STMMAC_TIMER
- if (likely(priv->tm->enable))
- priv->tm->timer_start(tmrate);
-#endif
napi_enable(&priv->napi);
netif_start_queue(ndev);
@@ -2311,11 +2221,6 @@ static int __init stmmac_cmdline_opt(char *str)
} else if (!strncmp(opt, "eee_timer:", 6)) {
if (kstrtoint(opt + 10, 0, &eee_timer))
goto err;
-#ifdef CONFIG_STMMAC_TIMER
- } else if (!strncmp(opt, "tmrate:", 7)) {
- if (kstrtoint(opt + 7, 0, &tmrate))
- goto err;
-#endif
}
}
return 0;
diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_timer.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_timer.c
deleted file mode 100644
index 2a0e1ab..0000000
--- a/drivers/net/ethernet/stmicro/stmmac/stmmac_timer.c
+++ /dev/null
@@ -1,134 +0,0 @@
-/*******************************************************************************
- STMMAC external timer support.
-
- Copyright (C) 2007-2009 STMicroelectronics Ltd
-
- This program is free software; you can redistribute it and/or modify it
- under the terms and conditions of the GNU General Public License,
- version 2, as published by the Free Software Foundation.
-
- This program is distributed in the hope it will be useful, but WITHOUT
- ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for
- more details.
-
- You should have received a copy of the GNU General Public License along with
- this program; if not, write to the Free Software Foundation, Inc.,
- 51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA.
-
- The full GNU General Public License is included in this distribution in
- the file called "COPYING".
-
- Author: Giuseppe Cavallaro <peppe.cavallaro@st.com>
-*******************************************************************************/
-
-#include <linux/kernel.h>
-#include <linux/etherdevice.h>
-#include "stmmac_timer.h"
-
-static void stmmac_timer_handler(void *data)
-{
- struct net_device *dev = (struct net_device *)data;
-
- stmmac_schedule(dev);
-}
-
-#define STMMAC_TIMER_MSG(timer, freq) \
-printk(KERN_INFO "stmmac_timer: %s Timer ON (freq %dHz)\n", timer, freq);
-
-#if defined(CONFIG_STMMAC_RTC_TIMER)
-#include <linux/rtc.h>
-static struct rtc_device *stmmac_rtc;
-static rtc_task_t stmmac_task;
-
-static void stmmac_rtc_start(unsigned int new_freq)
-{
- rtc_irq_set_freq(stmmac_rtc, &stmmac_task, new_freq);
- rtc_irq_set_state(stmmac_rtc, &stmmac_task, 1);
-}
-
-static void stmmac_rtc_stop(void)
-{
- rtc_irq_set_state(stmmac_rtc, &stmmac_task, 0);
-}
-
-int stmmac_open_ext_timer(struct net_device *dev, struct stmmac_timer *tm)
-{
- stmmac_task.private_data = dev;
- stmmac_task.func = stmmac_timer_handler;
-
- stmmac_rtc = rtc_class_open(CONFIG_RTC_HCTOSYS_DEVICE);
- if (stmmac_rtc == NULL) {
- pr_err("open rtc device failed\n");
- return -ENODEV;
- }
-
- rtc_irq_register(stmmac_rtc, &stmmac_task);
-
- /* Periodic mode is not supported */
- if ((rtc_irq_set_freq(stmmac_rtc, &stmmac_task, tm->freq) < 0)) {
- pr_err("set periodic failed\n");
- rtc_irq_unregister(stmmac_rtc, &stmmac_task);
- rtc_class_close(stmmac_rtc);
- return -1;
- }
-
- STMMAC_TIMER_MSG(CONFIG_RTC_HCTOSYS_DEVICE, tm->freq);
-
- tm->timer_start = stmmac_rtc_start;
- tm->timer_stop = stmmac_rtc_stop;
-
- return 0;
-}
-
-int stmmac_close_ext_timer(void)
-{
- rtc_irq_set_state(stmmac_rtc, &stmmac_task, 0);
- rtc_irq_unregister(stmmac_rtc, &stmmac_task);
- rtc_class_close(stmmac_rtc);
- return 0;
-}
-
-#elif defined(CONFIG_STMMAC_TMU_TIMER)
-#include <linux/clk.h>
-#define TMU_CHANNEL "tmu2_clk"
-static struct clk *timer_clock;
-
-static void stmmac_tmu_start(unsigned int new_freq)
-{
- clk_set_rate(timer_clock, new_freq);
- clk_enable(timer_clock);
-}
-
-static void stmmac_tmu_stop(void)
-{
- clk_disable(timer_clock);
-}
-
-int stmmac_open_ext_timer(struct net_device *dev, struct stmmac_timer *tm)
-{
- timer_clock = clk_get(NULL, TMU_CHANNEL);
-
- if (timer_clock == NULL)
- return -1;
-
- if (tmu2_register_user(stmmac_timer_handler, (void *)dev) < 0) {
- timer_clock = NULL;
- return -1;
- }
-
- STMMAC_TIMER_MSG("TMU2", tm->freq);
- tm->timer_start = stmmac_tmu_start;
- tm->timer_stop = stmmac_tmu_stop;
-
- return 0;
-}
-
-int stmmac_close_ext_timer(void)
-{
- clk_disable(timer_clock);
- tmu2_unregister_user();
- clk_put(timer_clock);
- return 0;
-}
-#endif
diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_timer.h b/drivers/net/ethernet/stmicro/stmmac/stmmac_timer.h
deleted file mode 100644
index aea9b14..0000000
--- a/drivers/net/ethernet/stmicro/stmmac/stmmac_timer.h
+++ /dev/null
@@ -1,46 +0,0 @@
-/*******************************************************************************
- STMMAC external timer Header File.
-
- Copyright (C) 2007-2009 STMicroelectronics Ltd
-
- This program is free software; you can redistribute it and/or modify it
- under the terms and conditions of the GNU General Public License,
- version 2, as published by the Free Software Foundation.
-
- This program is distributed in the hope it will be useful, but WITHOUT
- ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for
- more details.
-
- You should have received a copy of the GNU General Public License along with
- this program; if not, write to the Free Software Foundation, Inc.,
- 51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA.
-
- The full GNU General Public License is included in this distribution in
- the file called "COPYING".
-
- Author: Giuseppe Cavallaro <peppe.cavallaro@st.com>
-*******************************************************************************/
-#ifndef __STMMAC_TIMER_H__
-#define __STMMAC_TIMER_H__
-
-struct stmmac_timer {
- void (*timer_start) (unsigned int new_freq);
- void (*timer_stop) (void);
- unsigned int freq;
- unsigned int enable;
-};
-
-/* Open the HW timer device and return 0 in case of success */
-int stmmac_open_ext_timer(struct net_device *dev, struct stmmac_timer *tm);
-/* Stop the timer and release it */
-int stmmac_close_ext_timer(void);
-/* Function used for scheduling task within the stmmac */
-void stmmac_schedule(struct net_device *dev);
-
-#if defined(CONFIG_STMMAC_TMU_TIMER)
-extern int tmu2_register_user(void *fnt, void *data);
-extern void tmu2_unregister_user(void);
-#endif
-
-#endif /* __STMMAC_TIMER_H__ */
--
1.7.4.4
^ permalink raw reply related
* [net-next.git 0/8 (V4)] stmmac: remove dead code for STMMAC_TIMER and add new mitigation schema
From: Giuseppe CAVALLARO @ 2012-09-11 6:55 UTC (permalink / raw)
To: netdev; +Cc: bhutchings, davem, Giuseppe Cavallaro
These patch series remove the STMMAC_TIMER option no longer updated
and never used and add a new mitigation schema.
Having removed the Timer opt, this has made the driver slim.
On top of this work, it has been easier to introduce the new
mitigation schema based on HW RX-watchdog (available in new cores).
In fact, 3.50 and newer cores have an HW RX-Watchdog that can be used for
mitigating the Rx-interrupts and first results look promising.
Running n-u-t-t-c-p with the following parameters:
Throughput: 500Mbps
UDP Buffer size: 1328bytes
TCP Buffer size: 65536bytes
for example, I got on ST box (arm-based) these improvements:
--------------------------------------------------------------------
Original | With New Mitigation patch
--------------------------------------------------------------------
Test CPU usage pkt/loss | CPU usage pkt/loss
Type Mbps % % |Mbps % %
--------------------------------------------------------------------
UDP-RX 395.5065 95 20.89 |499.9552 23 0.00
UDP-TX 499.5578 100 0.08915 |500.0152 100 0.00
TCP-RX 499.9221 77 |499.9217 27
TCP-TX 389.5719 99 |499.9171 80
--------------------------------------------------------------------
... no regression on ST boxes (SH based) I always test.
This is a brief explanation of the new mitigation schema although there
is a patch that updates the driver's documentation.
o On Rx-side I have:
New GMACs will use the RX-watchdog timer; old ones will continue to
use NAPI to mitigate the RX DMA interrupts.
For the RX-watchdog, there is a parameter that is the RI Watchdog
Timer count. It indicates the number of system clock cycles and can be
set via *ethtool*.
o On Tx-side, the mitigation schema is based on a SW timer
that calls the tx function (stmmac_tx) to reclaim the resource after
transmitting the frames.
Also there is another parameter (a threshold) used to program
the descriptors avoiding to set the interrupt on completion bit in
when the frame is sent (xmit). This means that the stmmac_tx can be
called by the ISR too. Also this parameter can be tuned via ethtool.
V2: these patches add the ethtool support to get/set coalesce parameters
and totally remove the sysFS support added in the first patches.
V3: added several fixes: for example NAPI and RX-watchdog work together
while in the previous implementation the HW RX-watchdog disabled NAPI.
On the tx side, erroneously the tx coalesce frame parameter was limited
to the ring size and the driver didn't take care of the segment numbers
when enable/disable the IC bit in the TDES.
V4: reject not supported coalesce settings in ethtool.
Giuseppe Cavallaro (8):
stmmac: remove dead code for TIMER
stmmac: manage tx clean out of rx_poll
stmmac: add the initial tx coalesce schema
stmmac: add Rx watchdog support to mitigate the DMA irqs
stmmac: get/set coalesce parameters via ethtool
stmmac: fix and review the rx irq path after adding new mitigation
stmmac: update the doc with new IRQ mitigation
stmmac: update the driver version to Sept_2012
Documentation/networking/stmmac.txt | 28 ++-
drivers/net/ethernet/stmicro/stmmac/Kconfig | 25 --
drivers/net/ethernet/stmicro/stmmac/Makefile | 1 -
drivers/net/ethernet/stmicro/stmmac/common.h | 42 +++-
drivers/net/ethernet/stmicro/stmmac/dwmac1000.h | 3 -
.../net/ethernet/stmicro/stmmac/dwmac1000_dma.c | 10 +-
drivers/net/ethernet/stmicro/stmmac/dwmac100_dma.c | 4 +-
drivers/net/ethernet/stmicro/stmmac/dwmac_dma.h | 7 +-
drivers/net/ethernet/stmicro/stmmac/dwmac_lib.c | 28 ++-
drivers/net/ethernet/stmicro/stmmac/stmmac.h | 14 +-
.../net/ethernet/stmicro/stmmac/stmmac_ethtool.c | 103 ++++++++-
drivers/net/ethernet/stmicro/stmmac/stmmac_main.c | 240 ++++++++------------
drivers/net/ethernet/stmicro/stmmac/stmmac_timer.c | 134 -----------
drivers/net/ethernet/stmicro/stmmac/stmmac_timer.h | 46 ----
include/linux/stmmac.h | 1 +
15 files changed, 280 insertions(+), 406 deletions(-)
delete mode 100644 drivers/net/ethernet/stmicro/stmmac/stmmac_timer.c
delete mode 100644 drivers/net/ethernet/stmicro/stmmac/stmmac_timer.h
--
1.7.4.4
^ permalink raw reply
* Re: [PATCH 01/12] qlcnic: Refactoring - template based hardware interface
From: Francois Romieu @ 2012-09-11 6:01 UTC (permalink / raw)
To: Sony Chacko; +Cc: davem, netdev, Dept_NX_Linux_NIC_Driver, Anirban Chakraborty
In-Reply-To: <1347326403-1705-2-git-send-email-sony.chacko@qlogic.com>
Sony Chacko <sony.chacko@qlogic.com> :
> Modify 82xx driver to support new adapter - Qlogic 83XX CNA
> Separate adapter specific hardware accesses routines
> Create template based hardware interface
>
> Signed-off-by: Anirban Chakraborty <anirban.chakraborty@qlogic.com>
> Signed-off-by: Sony Chacko <sony.chacko@qlogic.com>
> ---
> drivers/net/ethernet/qlogic/qlcnic/qlcnic.h | 67 ++++-
> drivers/net/ethernet/qlogic/qlcnic/qlcnic_ctx.c | 4 +-
> .../net/ethernet/qlogic/qlcnic/qlcnic_ethtool.c | 32 ++-
> drivers/net/ethernet/qlogic/qlcnic/qlcnic_hdr.h | 55 +---
> drivers/net/ethernet/qlogic/qlcnic/qlcnic_hw.c | 125 ++++++--
> drivers/net/ethernet/qlogic/qlcnic/qlcnic_hw.h | 36 +++
> drivers/net/ethernet/qlogic/qlcnic/qlcnic_init.c | 1 +
> drivers/net/ethernet/qlogic/qlcnic/qlcnic_main.c | 309 +++++++++-----------
> 8 files changed, 359 insertions(+), 270 deletions(-)
> create mode 100644 drivers/net/ethernet/qlogic/qlcnic/qlcnic_hw.h
>
> diff --git a/drivers/net/ethernet/qlogic/qlcnic/qlcnic.h b/drivers/net/ethernet/qlogic/qlcnic/qlcnic.h
> index eaa1db9..a4ae965 100644
> --- a/drivers/net/ethernet/qlogic/qlcnic/qlcnic.h
> +++ b/drivers/net/ethernet/qlogic/qlcnic/qlcnic.h
> @@ -439,6 +439,8 @@ struct qlcnic_fw_dump {
> struct qlcnic_dump_template_hdr *tmpl_hdr;
> };
>
> +struct qlcnic_hardware_ops;
> +
qlcnic_hardware_ops is defined in this same file and it only uses
pointers to yet unknown types. You can define it here.
[...]
> @@ -552,8 +556,9 @@ struct qlcnic_recv_context {
> /* HW context creation */
>
> #define QLCNIC_OS_CRB_RETRY_COUNT 4000
> -#define QLCNIC_CDRP_SIGNATURE_MAKE(pcifn, version) \
> - (((pcifn) & 0xff) | (((version) & 0xff) << 8) | (0xcafe << 16))
> +#define QLCNIC_CDRP_SIGNATURE_MAKE(ahw) \
> + (((ahw->pci_func) & 0xff) | (((ahw->fw_hal_version) & \
> + 0xff) << 8) | (0xcafe << 16))
It could be a proper function and it could belong to the (only) file where
it is actually used (once).
[...]
> +static inline void
> +qlcnic_read_crb(struct qlcnic_adapter *adapter, char *buf,
> + loff_t offset, size_t size)
?
static inline void qlcnic_read_crb(struct qlcnic_adapter *adapter, char *buf,
loff_t offset, size_t size)
[...]
> diff --git a/drivers/net/ethernet/qlogic/qlcnic/qlcnic_ethtool.c b/drivers/net/ethernet/qlogic/qlcnic/qlcnic_ethtool.c
> index 9e9e78a..c81fcf8 100644
> --- a/drivers/net/ethernet/qlogic/qlcnic/qlcnic_ethtool.c
> +++ b/drivers/net/ethernet/qlogic/qlcnic/qlcnic_ethtool.c
[...]
> @@ -603,8 +608,7 @@ qlcnic_set_pauseparam(struct net_device *netdev,
> else
> qlcnic_gb_unset_rx_flowctl(val);
>
> - QLCWR32(adapter, QLCNIC_NIU_GB_MAC_CONFIG_0(port),
> - val);
> + QLCWR32(adapter, QLCNIC_NIU_GB_MAC_CONFIG_0(port), val);
This change does not fit the patch description.
[...]
> diff --git a/drivers/net/ethernet/qlogic/qlcnic/qlcnic_hdr.h b/drivers/net/ethernet/qlogic/qlcnic/qlcnic_hdr.h
> index 28a6b28..0bee9f7 100644
> --- a/drivers/net/ethernet/qlogic/qlcnic/qlcnic_hdr.h
> +++ b/drivers/net/ethernet/qlogic/qlcnic/qlcnic_hdr.h
[...]
Most of the changes in this file does not fit the patch description.
You could as well make a separate patch dedicated to removing unused stuff.
5...]
> diff --git a/drivers/net/ethernet/qlogic/qlcnic/qlcnic_hw.c b/drivers/net/ethernet/qlogic/qlcnic/qlcnic_hw.c
> index b528e52..ee4cd43 100644
> --- a/drivers/net/ethernet/qlogic/qlcnic/qlcnic_hw.c
> +++ b/drivers/net/ethernet/qlogic/qlcnic/qlcnic_hw.c
[...]
> @@ -880,7 +889,7 @@ qlcnic_pci_get_crb_addr_2M(struct qlcnic_adapter *adapter,
> m = &crb_128M_2M_map[CRB_BLK(off)].sub_block[CRB_SUBBLK(off)];
>
> if (m->valid && (m->start_128M <= off) && (m->end_128M > off)) {
> - *addr = adapter->ahw->pci_base0 + m->start_2M +
> + *addr = ahw->pci_base0 + m->start_2M +
> (off - m->start_128M);
*addr = ahw->pci_base0 + m->start_2M + off - m->start_128M;
[...]
> @@ -1105,7 +1116,7 @@ correct:
> i = 0;
> writel(TA_CTL_ENABLE, (mem_crb + TEST_AGT_CTRL));
> writel((TA_CTL_START | TA_CTL_ENABLE),
> - (mem_crb + TEST_AGT_CTRL));
> + (mem_crb + TEST_AGT_CTRL));
writel(TA_CTL_START | TA_CTL_ENABLE, mem_crb + TEST_AGT_CTRL);
The topic of the patch does not ask for a change here and these lines
don't even appear in the context diff of such changes. It could be
isolated in a separate patch (where you would probably be suggested to
remove the useless parenthesis in "(TA_CTL_START | TA_CTL_ENABLE)" as well).
[...]
> @@ -1786,3 +1800,54 @@ error:
> vfree(fw_dump->data);
> return -EINVAL;
> }
> +
> +void qlcnic_get_func_no(struct qlcnic_adapter *adapter)
> +{
> + void __iomem *msix_base_addr;
> + u32 func;
> + u32 msix_base;
> +
> + pci_read_config_dword(adapter->pdev, QLCNIC_MSIX_TABLE_OFFSET, &func);
> + msix_base_addr = adapter->ahw->pci_base0 + QLCNIC_MSIX_BASE;
> + msix_base = readl(msix_base_addr);
> + func = (func - msix_base)/QLCNIC_MSIX_TBL_PGSIZE;
> + adapter->ahw->pci_func = func;
> +}
It is not used in this patch. #2 adds a declaration for it in qlcnic.h, a
function pointer for it in qlcnic_hardware_ops, then uses it. #5 renames
it to qlcnic_82xx_get_func_no.
Reviewers won't complain if you keep the patchset at the smallest size
for a given set of changes, especially as the qlcnic_82xx_ prefix is
used a few lines below.
> +
> +void qlcnic_get_ocm_win(struct qlcnic_hardware_context *ahw)
> +{
> + u32 addr;
> +
> + addr = QLCNIC_PCIX_PS_REG(PCIX_OCM_WINDOW_REG(ahw->pci_func));
> + ahw->ocm_win_crb = qlcnic_get_ioaddr(ahw, addr);
> +}
Either it's hidden behind some clever macro or it isn't used at all in
the whole patchset.
> +
> +void qlcnic_82xx_read_crb(struct qlcnic_adapter *adapter, char *buf,
> + loff_t offset, size_t size)
> +{
> + u32 data;
> + u64 qmdata;
> +
> + if (ADDR_IN_RANGE(offset, QLCNIC_PCI_CAMQM, QLCNIC_PCI_CAMQM_END)) {
> + qlcnic_pci_camqm_read_2M(adapter, offset, &qmdata);
> + memcpy(buf, &qmdata, size);
> + } else {
> + data = QLCRD32(adapter, offset);
> + memcpy(buf, &data, size);
> + }
> +}
Nit: reduce the scope of data and qmdata.
[...]
> diff --git a/drivers/net/ethernet/qlogic/qlcnic/qlcnic_hw.h b/drivers/net/ethernet/qlogic/qlcnic/qlcnic_hw.h
> new file mode 100644
> index 0000000..7dab9e2
> --- /dev/null
> +++ b/drivers/net/ethernet/qlogic/qlcnic/qlcnic_hw.h
> @@ -0,0 +1,36 @@
> +#ifndef __QLCNIC_HW_H
> +#define __QLCNIC_HW_H
> +
> +#include "qlcnic.h"
> +
> +/* List of PCI device IDs */
> +#define PCI_DEVICE_ID_QLOGIC_QLE824X 0x8020
> +#define QLCNIC_P3P_BAR0_LENGTH 0x00200000UL
> +
> +#define QLCNIC_BAR_LENGTH(dev_id, bar) \
> +do { \
> + if (dev_id == PCI_DEVICE_ID_QLOGIC_QLE824X)\
> + *bar = QLCNIC_P3P_BAR0_LENGTH; \
> + else \
> + *bar = 0; \
> +} while (0)
It is used only once in the whole patchset.
> diff --git a/drivers/net/ethernet/qlogic/qlcnic/qlcnic_init.c b/drivers/net/ethernet/qlogic/qlcnic/qlcnic_init.c
> index 0bcda9c..350bf79 100644
> --- a/drivers/net/ethernet/qlogic/qlcnic/qlcnic_init.c
> +++ b/drivers/net/ethernet/qlogic/qlcnic/qlcnic_init.c
[...]
> @@ -343,9 +312,9 @@ static const struct net_device_ops qlcnic_netdev_failed_ops = {
> };
>
> static struct qlcnic_nic_template qlcnic_ops = {
> - .config_bridged_mode = qlcnic_config_bridged_mode,
> - .config_led = qlcnic_config_led,
> - .start_firmware = qlcnic_start_firmware
> + .config_bridged_mode = qlcnic_82xx_config_bridged_mode,
> + .config_led = qlcnic_82xx_config_led,
> + .start_firmware = qlcnic_82xx_start_firmware
> };
You should really consider aligning as:
static struct qlcnic_nic_template qlcnic_ops = {
.config_bridged_mode = qlcnic_82xx_config_bridged_mode,
.config_led = qlcnic_82xx_config_led,
.start_firmware = qlcnic_82xx_start_firmware
};
(off to work)
--
Ueimor
^ permalink raw reply
* Re: [PATCH 0/12] qlcnic: patches for new adapter - Qlogic 83XX CNA
From: Francois Romieu @ 2012-09-11 6:01 UTC (permalink / raw)
To: Sony Chacko; +Cc: davem, netdev, Dept_NX_Linux_NIC_Driver
In-Reply-To: <1347326403-1705-1-git-send-email-sony.chacko@qlogic.com>
Sony Chacko <sony.chacko@qlogic.com> :
[...]
> We have updated the patches.
> Please apply the updated 12 patch series to net-next.
Please consider including a submission count in the subject. A formal
changelog in message #0 will be welcome too.
<my life>
Spliting the patches further would help review.
</my life>
--
Ueimor
^ permalink raw reply
* Re: [PATCH v3 0/7] mv643xx.c: Add basic device tree support.
From: Benjamin Herrenschmidt @ 2012-09-11 6:03 UTC (permalink / raw)
To: Arnd Bergmann
Cc: thomas.petazzoni-wi1+55ScJUtKEb57/3fJTNBPR1lH4CV8,
andrew-g2DYL2Zd6BY, netdev-u79uwXL29TY76Z2rM5mHXA,
devicetree-discuss-uLR06cmDAlY/bJ5BZ2RsiQ,
linuxppc-dev-uLR06cmDAlY/bJ5BZ2RsiQ,
ben.dooks-4yDnlxn2s6sWdaTGBSpHTA, dale-1viX+2+OPRFcxvNqPlePQg,
Ian Molton, David Miller,
linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r
In-Reply-To: <201209101422.13875.arnd-r2nGTMty4D4@public.gmane.org>
On Mon, 2012-09-10 at 14:22 +0000, Arnd Bergmann wrote:
> Following up on the old discussion, I talked briefly about this
> issue with BenH at the kernel summit. The outcome basically is that
> it's a bit sad to have incompatible bindings, but it's not the end
> of the world,and it's more important to do it right this time.
>
> Just make sure that you use different values for the 'compatible'
> strings and then do what you need to get the ARM hardware working.
>
> Ideally, the new binding should be written in a way that powerpc
> machines can use the same one, but the existing ones all use
> an version of Open Firmware that is not going to get updated
> and it's also not too likely that we are going to see new
> powerpc machines based on this chip.
Right, mostly these machines where the Pegasos. Those came with a fairly
busted variant of Open Firmware which generated a pretty gross
device-tree.
For some reason, the manufacturer of those things was never willing to
fix anything in their firmware (despite the distributor providing
patches etc...), seemingly on the assumption that whatever they were
doing was perfect and operating system people like us didn't matter one
little bit :-)
So I don't care much about it. It would be nice to keep them working
since people in the community still have them but if it goes through
some "compat" code that detects old/broken device-trees and eventually
disappears when we finally drop support, then so be it.
Cheers,
Ben.
^ permalink raw reply
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox