/*
 * This is the Fusion MPT base driver providing common API layer interface
 * for access to MPT (Message Passing Technology) firmware.
 *
 * This code is based on drivers/scsi/mpt3sas/mpt3sas_base.c
 * Copyright (C) 2012-2014  LSI Corporation
 * Copyright (C) 2013-2014 Avago Technologies
 *  (mailto: MPT-FusionLinux.pdl@avagotech.com)
 *
 * This program is free software; you can redistribute it and/or
 * modify it under the terms of the GNU General Public License
 * as published by the Free Software Foundation; either version 2
 * of the License, or (at your option) any later version.
 *
 * This program is distributed in the hope that 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.
 *
 * NO WARRANTY
 * THE PROGRAM IS PROVIDED ON AN "AS IS" BASIS, WITHOUT WARRANTIES OR
 * CONDITIONS OF ANY KIND, EITHER EXPRESS OR IMPLIED INCLUDING, WITHOUT
 * LIMITATION, ANY WARRANTIES OR CONDITIONS OF TITLE, NON-INFRINGEMENT,
 * MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Each Recipient is
 * solely responsible for determining the appropriateness of using and
 * distributing the Program and assumes all risks associated with its
 * exercise of rights under this Agreement, including but not limited to
 * the risks and costs of program errors, damage to or loss of data,
 * programs or equipment, and unavailability or interruption of operations.

 * DISCLAIMER OF LIABILITY
 * NEITHER RECIPIENT NOR ANY CONTRIBUTORS SHALL HAVE ANY LIABILITY FOR ANY
 * DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
 * DAMAGES (INCLUDING WITHOUT LIMITATION LOST PROFITS), HOWEVER CAUSED AND
 * ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR
 * TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE
 * USE OR DISTRIBUTION OF THE PROGRAM OR THE EXERCISE OF ANY RIGHTS GRANTED
 * HEREUNDER, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGES

 * 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 Street, Fifth Floor, Boston, MA  02110-1301,
 * USA.
 */

#include <linux/kernel.h>
#include <linux/module.h>
#include <linux/errno.h>
#include <linux/init.h>
#include <linux/slab.h>
#include <linux/types.h>
#include <linux/pci.h>
#include <linux/kdev_t.h>
#include <linux/blkdev.h>
#include <linux/delay.h>
#include <linux/interrupt.h>
#include <linux/dma-mapping.h>
#include <linux/io.h>
#include <linux/time.h>
#include <linux/ktime.h>
#include <linux/kthread.h>
#include <asm/page.h>        /* To get host page size per arch */
#include <linux/aer.h>


#include "mpt3sas_base.h"

static MPT_CALLBACK     mpt_callbacks[MPT_MAX_CALLBACKS];


#define FAULT_POLLING_INTERVAL 1000 /* in milliseconds */

 /* maximum controller queue depth */
#define MAX_HBA_QUEUE_DEPTH     30000
#define MAX_CHAIN_DEPTH         100000
static int max_queue_depth = -1;
module_param(max_queue_depth, int, 0444);
MODULE_PARM_DESC(max_queue_depth, " max controller queue depth ");

static int max_sgl_entries = -1;
module_param(max_sgl_entries, int, 0444);
MODULE_PARM_DESC(max_sgl_entries, " max sg entries ");

static int msix_disable = -1;
module_param(msix_disable, int, 0444);
MODULE_PARM_DESC(msix_disable, " disable msix routed interrupts (default=0)");

static int smp_affinity_enable = 1;
module_param(smp_affinity_enable, int, 0444);
MODULE_PARM_DESC(smp_affinity_enable, "SMP affinity feature enable/disable Default: enable(1)");

static int max_msix_vectors = -1;
module_param(max_msix_vectors, int, 0444);
MODULE_PARM_DESC(max_msix_vectors,
        " max msix vectors");

static int irqpoll_weight = -1;
module_param(irqpoll_weight, int, 0444);
MODULE_PARM_DESC(irqpoll_weight,
        "irq poll weight (default= one fourth of HBA queue depth)");

static int mpt3sas_fwfault_debug;
MODULE_PARM_DESC(mpt3sas_fwfault_debug,
        " enable detection of firmware fault and halt firmware - (default=0)");

static int perf_mode = -1;
module_param(perf_mode, int, 0444);
MODULE_PARM_DESC(perf_mode,
        "Performance mode (only for Aero/Sea Generation), options:\n\t\t"
        "0 - balanced: high iops mode is enabled &\n\t\t"
        "interrupt coalescing is enabled only on high iops queues,\n\t\t"
        "1 - iops: high iops mode is disabled &\n\t\t"
        "interrupt coalescing is enabled on all queues,\n\t\t"
        "2 - latency: high iops mode is disabled &\n\t\t"
        "interrupt coalescing is enabled on all queues with timeout value 0xA,\n"
        "\t\tdefault - default perf_mode is 'balanced'"
        );

enum mpt3sas_perf_mode {
        MPT_PERF_MODE_DEFAULT   = -1,
        MPT_PERF_MODE_BALANCED  = 0,
        MPT_PERF_MODE_IOPS      = 1,
        MPT_PERF_MODE_LATENCY   = 2,
};

static int
_base_wait_on_iocstate(struct MPT3SAS_ADAPTER *ioc,
                u32 ioc_state, int timeout);
static int
_base_get_ioc_facts(struct MPT3SAS_ADAPTER *ioc);
static void
_base_clear_outstanding_commands(struct MPT3SAS_ADAPTER *ioc);

/**
 * mpt3sas_base_check_cmd_timeout - Function
 *              to check timeout and command termination due
 *              to Host reset.
 *
 * @ioc:        per adapter object.
 * @status:     Status of issued command.
 * @mpi_request:mf request pointer.
 * @sz:         size of buffer.
 *
 * @Returns - 1/0 Reset to be done or Not
 */
u8
mpt3sas_base_check_cmd_timeout(struct MPT3SAS_ADAPTER *ioc,
                u8 status, void *mpi_request, int sz)
{
        u8 issue_reset = 0;

        if (!(status & MPT3_CMD_RESET))
                issue_reset = 1;

        ioc_err(ioc, "Command %s\n",
                issue_reset == 0 ? "terminated due to Host Reset" : "Timeout");
        _debug_dump_mf(mpi_request, sz);

        return issue_reset;
}

/**
 * _scsih_set_fwfault_debug - global setting of ioc->fwfault_debug.
 * @val: ?
 * @kp: ?
 *
 * Return: ?
 */
static int
_scsih_set_fwfault_debug(const char *val, const struct kernel_param *kp)
{
        int ret = param_set_int(val, kp);
        struct MPT3SAS_ADAPTER *ioc;

        if (ret)
                return ret;

        /* global ioc spinlock to protect controller list on list operations */
        pr_info("setting fwfault_debug(%d)\n", mpt3sas_fwfault_debug);
        spin_lock(&gioc_lock);
        list_for_each_entry(ioc, &mpt3sas_ioc_list, list)
                ioc->fwfault_debug = mpt3sas_fwfault_debug;
        spin_unlock(&gioc_lock);
        return 0;
}
module_param_call(mpt3sas_fwfault_debug, _scsih_set_fwfault_debug,
        param_get_int, &mpt3sas_fwfault_debug, 0644);

/**
 * _base_readl_aero - retry readl for max three times.
 * @addr: MPT Fusion system interface register address
 *
 * Retry the readl() for max three times if it gets zero value
 * while reading the system interface register.
 */
static inline u32
_base_readl_aero(const volatile void __iomem *addr)
{
        u32 i = 0, ret_val;

        do {
                ret_val = readl(addr);
                i++;
        } while (ret_val == 0 && i < 3);

        return ret_val;
}

static inline u32
_base_readl(const volatile void __iomem *addr)
{
        return readl(addr);
}

/**
 * _base_clone_reply_to_sys_mem - copies reply to reply free iomem
 *                                in BAR0 space.
 *
 * @ioc: per adapter object
 * @reply: reply message frame(lower 32bit addr)
 * @index: System request message index.
 */
static void
_base_clone_reply_to_sys_mem(struct MPT3SAS_ADAPTER *ioc, u32 reply,
                u32 index)
{
        /*
         * 256 is offset within sys register.
         * 256 offset MPI frame starts. Max MPI frame supported is 32.
         * 32 * 128 = 4K. From here, Clone of reply free for mcpu starts
         */
        u16 cmd_credit = ioc->facts.RequestCredit + 1;
        void __iomem *reply_free_iomem = (void __iomem *)ioc->chip +
                        MPI_FRAME_START_OFFSET +
                        (cmd_credit * ioc->request_sz) + (index * sizeof(u32));

        writel(reply, reply_free_iomem);
}

/**
 * _base_clone_mpi_to_sys_mem - Writes/copies MPI frames
 *                              to system/BAR0 region.
 *
 * @dst_iomem: Pointer to the destination location in BAR0 space.
 * @src: Pointer to the Source data.
 * @size: Size of data to be copied.
 */
static void
_base_clone_mpi_to_sys_mem(void *dst_iomem, void *src, u32 size)
{
        int i;
        u32 *src_virt_mem = (u32 *)src;

        for (i = 0; i < size/4; i++)
                writel((u32)src_virt_mem[i],
                                (void __iomem *)dst_iomem + (i * 4));
}

/**
 * _base_clone_to_sys_mem - Writes/copies data to system/BAR0 region
 *
 * @dst_iomem: Pointer to the destination location in BAR0 space.
 * @src: Pointer to the Source data.
 * @size: Size of data to be copied.
 */
static void
_base_clone_to_sys_mem(void __iomem *dst_iomem, void *src, u32 size)
{
        int i;
        u32 *src_virt_mem = (u32 *)(src);

        for (i = 0; i < size/4; i++)
                writel((u32)src_virt_mem[i],
                        (void __iomem *)dst_iomem + (i * 4));
}

/**
 * _base_get_chain - Calculates and Returns virtual chain address
 *                       for the provided smid in BAR0 space.
 *
 * @ioc: per adapter object
 * @smid: system request message index
 * @sge_chain_count: Scatter gather chain count.
 *
 * Return: the chain address.
 */
static inline void __iomem*
_base_get_chain(struct MPT3SAS_ADAPTER *ioc, u16 smid,
                u8 sge_chain_count)
{
        void __iomem *base_chain, *chain_virt;
        u16 cmd_credit = ioc->facts.RequestCredit + 1;

        base_chain  = (void __iomem *)ioc->chip + MPI_FRAME_START_OFFSET +
                (cmd_credit * ioc->request_sz) +
                REPLY_FREE_POOL_SIZE;
        chain_virt = base_chain + (smid * ioc->facts.MaxChainDepth *
                        ioc->request_sz) + (sge_chain_count * ioc->request_sz);
        return chain_virt;
}

/**
 * _base_get_chain_phys - Calculates and Returns physical address
 *                      in BAR0 for scatter gather chains, for
 *                      the provided smid.
 *
 * @ioc: per adapter object
 * @smid: system request message index
 * @sge_chain_count: Scatter gather chain count.
 *
 * Return: Physical chain address.
 */
static inline phys_addr_t
_base_get_chain_phys(struct MPT3SAS_ADAPTER *ioc, u16 smid,
                u8 sge_chain_count)
{
        phys_addr_t base_chain_phys, chain_phys;
        u16 cmd_credit = ioc->facts.RequestCredit + 1;

        base_chain_phys  = ioc->chip_phys + MPI_FRAME_START_OFFSET +
                (cmd_credit * ioc->request_sz) +
                REPLY_FREE_POOL_SIZE;
        chain_phys = base_chain_phys + (smid * ioc->facts.MaxChainDepth *
                        ioc->request_sz) + (sge_chain_count * ioc->request_sz);
        return chain_phys;
}

/**
 * _base_get_buffer_bar0 - Calculates and Returns BAR0 mapped Host
 *                      buffer address for the provided smid.
 *                      (Each smid can have 64K starts from 17024)
 *
 * @ioc: per adapter object
 * @smid: system request message index
 *
 * Return: Pointer to buffer location in BAR0.
 */

static void __iomem *
_base_get_buffer_bar0(struct MPT3SAS_ADAPTER *ioc, u16 smid)
{
        u16 cmd_credit = ioc->facts.RequestCredit + 1;
        // Added extra 1 to reach end of chain.
        void __iomem *chain_end = _base_get_chain(ioc,
                        cmd_credit + 1,
                        ioc->facts.MaxChainDepth);
        return chain_end + (smid * 64 * 1024);
}

/**
 * _base_get_buffer_phys_bar0 - Calculates and Returns BAR0 mapped
 *              Host buffer Physical address for the provided smid.
 *              (Each smid can have 64K starts from 17024)
 *
 * @ioc: per adapter object
 * @smid: system request message index
 *
 * Return: Pointer to buffer location in BAR0.
 */
static phys_addr_t
_base_get_buffer_phys_bar0(struct MPT3SAS_ADAPTER *ioc, u16 smid)
{
        u16 cmd_credit = ioc->facts.RequestCredit + 1;
        phys_addr_t chain_end_phys = _base_get_chain_phys(ioc,
                        cmd_credit + 1,
                        ioc->facts.MaxChainDepth);
        return chain_end_phys + (smid * 64 * 1024);
}

/**
 * _base_get_chain_buffer_dma_to_chain_buffer - Iterates chain
 *                      lookup list and Provides chain_buffer
 *                      address for the matching dma address.
 *                      (Each smid can have 64K starts from 17024)
 *
 * @ioc: per adapter object
 * @chain_buffer_dma: Chain buffer dma address.
 *
 * Return: Pointer to chain buffer. Or Null on Failure.
 */
static void *
_base_get_chain_buffer_dma_to_chain_buffer(struct MPT3SAS_ADAPTER *ioc,
                dma_addr_t chain_buffer_dma)
{
        u16 index, j;
        struct chain_tracker *ct;

        for (index = 0; index < ioc->scsiio_depth; index++) {
                for (j = 0; j < ioc->chains_needed_per_io; j++) {
                        ct = &ioc->chain_lookup[index].chains_per_smid[j];
                        if (ct && ct->chain_buffer_dma == chain_buffer_dma)
                                return ct->chain_buffer;
                }
        }
        ioc_info(ioc, "Provided chain_buffer_dma address is not in the lookup list\n");
        return NULL;
}

/**
 * _clone_sg_entries -  MPI EP's scsiio and config requests
 *                      are handled here. Base function for
 *                      double buffering, before submitting
 *                      the requests.
 *
 * @ioc: per adapter object.
 * @mpi_request: mf request pointer.
 * @smid: system request message index.
 */
static void _clone_sg_entries(struct MPT3SAS_ADAPTER *ioc,
                void *mpi_request, u16 smid)
{
        Mpi2SGESimple32_t *sgel, *sgel_next;
        u32  sgl_flags, sge_chain_count = 0;
        bool is_write = false;
        u16 i = 0;
        void __iomem *buffer_iomem;
        phys_addr_t buffer_iomem_phys;
        void __iomem *buff_ptr;
        phys_addr_t buff_ptr_phys;
        void __iomem *dst_chain_addr[MCPU_MAX_CHAINS_PER_IO];
        void *src_chain_addr[MCPU_MAX_CHAINS_PER_IO];
        phys_addr_t dst_addr_phys;
        MPI2RequestHeader_t *request_hdr;
        struct scsi_cmnd *scmd;
        struct scatterlist *sg_scmd = NULL;
        int is_scsiio_req = 0;

        request_hdr = (MPI2RequestHeader_t *) mpi_request;

        if (request_hdr->Function == MPI2_FUNCTION_SCSI_IO_REQUEST) {
                Mpi25SCSIIORequest_t *scsiio_request =
                        (Mpi25SCSIIORequest_t *)mpi_request;
                sgel = (Mpi2SGESimple32_t *) &scsiio_request->SGL;
                is_scsiio_req = 1;
        } else if (request_hdr->Function == MPI2_FUNCTION_CONFIG) {
                Mpi2ConfigRequest_t  *config_req =
                        (Mpi2ConfigRequest_t *)mpi_request;
                sgel = (Mpi2SGESimple32_t *) &config_req->PageBufferSGE;
        } else
                return;

        /* From smid we can get scsi_cmd, once we have sg_scmd,
         * we just need to get sg_virt and sg_next to get virual
         * address associated with sgel->Address.
         */

        if (is_scsiio_req) {
                /* Get scsi_cmd using smid */
                scmd = mpt3sas_scsih_scsi_lookup_get(ioc, smid);
                if (scmd == NULL) {
                        ioc_err(ioc, "scmd is NULL\n");
                        return;
                }

                /* Get sg_scmd from scmd provided */
                sg_scmd = scsi_sglist(scmd);
        }

        /*
         * 0 - 255      System register
         * 256 - 4352   MPI Frame. (This is based on maxCredit 32)
         * 4352 - 4864  Reply_free pool (512 byte is reserved
         *              considering maxCredit 32. Reply need extra
         *              room, for mCPU case kept four times of
         *              maxCredit).
         * 4864 - 17152 SGE chain element. (32cmd * 3 chain of
         *              128 byte size = 12288)
         * 17152 - x    Host buffer mapped with smid.
         *              (Each smid can have 64K Max IO.)
         * BAR0+Last 1K MSIX Addr and Data
         * Total size in use 2113664 bytes of 4MB BAR0
         */

        buffer_iomem = _base_get_buffer_bar0(ioc, smid);
        buffer_iomem_phys = _base_get_buffer_phys_bar0(ioc, smid);

        buff_ptr = buffer_iomem;
        buff_ptr_phys = buffer_iomem_phys;
        WARN_ON(buff_ptr_phys > U32_MAX);

        if (le32_to_cpu(sgel->FlagsLength) &
                        (MPI2_SGE_FLAGS_HOST_TO_IOC << MPI2_SGE_FLAGS_SHIFT))
                is_write = true;

        for (i = 0; i < MPT_MIN_PHYS_SEGMENTS + ioc->facts.MaxChainDepth; i++) {

                sgl_flags =
                    (le32_to_cpu(sgel->FlagsLength) >> MPI2_SGE_FLAGS_SHIFT);

                switch (sgl_flags & MPI2_SGE_FLAGS_ELEMENT_MASK) {
                case MPI2_SGE_FLAGS_CHAIN_ELEMENT:
                        /*
                         * Helper function which on passing
                         * chain_buffer_dma returns chain_buffer. Get
                         * the virtual address for sgel->Address
                         */
                        sgel_next =
                                _base_get_chain_buffer_dma_to_chain_buffer(ioc,
                                                le32_to_cpu(sgel->Address));
                        if (sgel_next == NULL)
                                return;
                        /*
                         * This is coping 128 byte chain
                         * frame (not a host buffer)
                         */
                        dst_chain_addr[sge_chain_count] =
                                _base_get_chain(ioc,
                                        smid, sge_chain_count);
                        src_chain_addr[sge_chain_count] =
                                                (void *) sgel_next;
                        dst_addr_phys = _base_get_chain_phys(ioc,
                                                smid, sge_chain_count);
                        WARN_ON(dst_addr_phys > U32_MAX);
                        sgel->Address =
                                cpu_to_le32(lower_32_bits(dst_addr_phys));
                        sgel = sgel_next;
                        sge_chain_count++;
                        break;
                case MPI2_SGE_FLAGS_SIMPLE_ELEMENT:
                        if (is_write) {
                                if (is_scsiio_req) {
                                        _base_clone_to_sys_mem(buff_ptr,
                                            sg_virt(sg_scmd),
                                            (le32_to_cpu(sgel->FlagsLength) &
                                            0x00ffffff));
                                        /*
                                         * FIXME: this relies on a a zero
                                         * PCI mem_offset.
                                         */
                                        sgel->Address =
                                            cpu_to_le32((u32)buff_ptr_phys);
                                } else {
                                        _base_clone_to_sys_mem(buff_ptr,
                                            ioc->config_vaddr,
                                            (le32_to_cpu(sgel->FlagsLength) &
                                            0x00ffffff));
                                        sgel->Address =
                                            cpu_to_le32((u32)buff_ptr_phys);
                                }
                        }
                        buff_ptr += (le32_to_cpu(sgel->FlagsLength) &
                            0x00ffffff);
                        buff_ptr_phys += (le32_to_cpu(sgel->FlagsLength) &
                            0x00ffffff);
                        if ((le32_to_cpu(sgel->FlagsLength) &
                            (MPI2_SGE_FLAGS_END_OF_BUFFER
                                        << MPI2_SGE_FLAGS_SHIFT)))
                                goto eob_clone_chain;
                        else {
                                /*
                                 * Every single element in MPT will have
                                 * associated sg_next. Better to sanity that
                                 * sg_next is not NULL, but it will be a bug
                                 * if it is null.
                                 */
                                if (is_scsiio_req) {
                                        sg_scmd = sg_next(sg_scmd);
                                        if (sg_scmd)
                                                sgel++;
                                        else
                                                goto eob_clone_chain;
                                }
                        }
                        break;
                }
        }

eob_clone_chain:
        for (i = 0; i < sge_chain_count; i++) {
                if (is_scsiio_req)
                        _base_clone_to_sys_mem(dst_chain_addr[i],
                                src_chain_addr[i], ioc->request_sz);
        }
}

/**
 *  mpt3sas_remove_dead_ioc_func - kthread context to remove dead ioc
 * @arg: input argument, used to derive ioc
 *
 * Return:
 * 0 if controller is removed from pci subsystem.
 * -1 for other case.
 */
static int mpt3sas_remove_dead_ioc_func(void *arg)
{
        struct MPT3SAS_ADAPTER *ioc = (struct MPT3SAS_ADAPTER *)arg;
        struct pci_dev *pdev;

        if (!ioc)
                return -1;

        pdev = ioc->pdev;
        if (!pdev)
                return -1;
        pci_stop_and_remove_bus_device_locked(pdev);
        return 0;
}

/**
 * _base_fault_reset_work - workq handling ioc fault conditions
 * @work: input argument, used to derive ioc
 *
 * Context: sleep.
 */
static void
_base_fault_reset_work(struct work_struct *work)
{
        struct MPT3SAS_ADAPTER *ioc =
            container_of(work, struct MPT3SAS_ADAPTER, fault_reset_work.work);
        unsigned long    flags;
        u32 doorbell;
        int rc;
        struct task_struct *p;


        spin_lock_irqsave(&ioc->ioc_reset_in_progress_lock, flags);
        if ((ioc->shost_recovery && (ioc->ioc_coredump_loop == 0)) ||
                        ioc->pci_error_recovery)
                goto rearm_timer;
        spin_unlock_irqrestore(&ioc->ioc_reset_in_progress_lock, flags);

        doorbell = mpt3sas_base_get_iocstate(ioc, 0);
        if ((doorbell & MPI2_IOC_STATE_MASK) == MPI2_IOC_STATE_MASK) {
                ioc_err(ioc, "SAS host is non-operational !!!!\n");

                /* It may be possible that EEH recovery can resolve some of
                 * pci bus failure issues rather removing the dead ioc function
                 * by considering controller is in a non-operational state. So
                 * here priority is given to the EEH recovery. If it doesn't
                 * not resolve this issue, mpt3sas driver will consider this
                 * controller to non-operational state and remove the dead ioc
                 * function.
                 */
                if (ioc->non_operational_loop++ < 5) {
                        spin_lock_irqsave(&ioc->ioc_reset_in_progress_lock,
                                                         flags);
                        goto rearm_timer;
                }

                /*
                 * Call _scsih_flush_pending_cmds callback so that we flush all
                 * pending commands back to OS. This call is required to aovid
                 * deadlock at block layer. Dead IOC will fail to do diag reset,
                 * and this call is safe since dead ioc will never return any
                 * command back from HW.
                 */
                ioc->schedule_dead_ioc_flush_running_cmds(ioc);
                /*
                 * Set remove_host flag early since kernel thread will
                 * take some time to execute.
                 */
                ioc->remove_host = 1;
                /*Remove the Dead Host */
                p = kthread_run(mpt3sas_remove_dead_ioc_func, ioc,
                    "%s_dead_ioc_%d", ioc->driver_name, ioc->id);
                if (IS_ERR(p))
                        ioc_err(ioc, "%s: Running mpt3sas_dead_ioc thread failed !!!!\n",
                                __func__);
                else
                        ioc_err(ioc, "%s: Running mpt3sas_dead_ioc thread success !!!!\n",
                                __func__);
                return; /* don't rearm timer */
        }

        if ((doorbell & MPI2_IOC_STATE_MASK) == MPI2_IOC_STATE_COREDUMP) {
                u8 timeout = (ioc->manu_pg11.CoreDumpTOSec) ?
                    ioc->manu_pg11.CoreDumpTOSec :
                    MPT3SAS_DEFAULT_COREDUMP_TIMEOUT_SECONDS;

                timeout /= (FAULT_POLLING_INTERVAL/1000);

                if (ioc->ioc_coredump_loop == 0) {
                        mpt3sas_print_coredump_info(ioc,
                            doorbell & MPI2_DOORBELL_DATA_MASK);
                        /* do not accept any IOs and disable the interrupts */
                        spin_lock_irqsave(
                            &ioc->ioc_reset_in_progress_lock, flags);
                        ioc->shost_recovery = 1;
                        spin_unlock_irqrestore(
                            &ioc->ioc_reset_in_progress_lock, flags);
                        mpt3sas_base_mask_interrupts(ioc);
                        _base_clear_outstanding_commands(ioc);
                }

                ioc_info(ioc, "%s: CoreDump loop %d.",
                    __func__, ioc->ioc_coredump_loop);

                /* Wait until CoreDump completes or times out */
                if (ioc->ioc_coredump_loop++ < timeout) {
                        spin_lock_irqsave(
                            &ioc->ioc_reset_in_progress_lock, flags);
                        goto rearm_timer;
                }
        }

        if (ioc->ioc_coredump_loop) {
                if ((doorbell & MPI2_IOC_STATE_MASK) != MPI2_IOC_STATE_COREDUMP)
                        ioc_err(ioc, "%s: CoreDump completed. LoopCount: %d",
                            __func__, ioc->ioc_coredump_loop);
                else
                        ioc_err(ioc, "%s: CoreDump Timed out. LoopCount: %d",
                            __func__, ioc->ioc_coredump_loop);
                ioc->ioc_coredump_loop = MPT3SAS_COREDUMP_LOOP_DONE;
        }
        ioc->non_operational_loop = 0;
        if ((doorbell & MPI2_IOC_STATE_MASK) != MPI2_IOC_STATE_OPERATIONAL) {
                rc = mpt3sas_base_hard_reset_handler(ioc, FORCE_BIG_HAMMER);
                ioc_warn(ioc, "%s: hard reset: %s\n",
                         __func__, rc == 0 ? "success" : "failed");
                doorbell = mpt3sas_base_get_iocstate(ioc, 0);
                if ((doorbell & MPI2_IOC_STATE_MASK) == MPI2_IOC_STATE_FAULT) {
                        mpt3sas_print_fault_code(ioc, doorbell &
                            MPI2_DOORBELL_DATA_MASK);
                } else if ((doorbell & MPI2_IOC_STATE_MASK) ==
                    MPI2_IOC_STATE_COREDUMP)
                        mpt3sas_print_coredump_info(ioc, doorbell &
                            MPI2_DOORBELL_DATA_MASK);
                if (rc && (doorbell & MPI2_IOC_STATE_MASK) !=
                    MPI2_IOC_STATE_OPERATIONAL)
                        return; /* don't rearm timer */
        }
        ioc->ioc_coredump_loop = 0;

        spin_lock_irqsave(&ioc->ioc_reset_in_progress_lock, flags);
 rearm_timer:
        if (ioc->fault_reset_work_q)
                queue_delayed_work(ioc->fault_reset_work_q,
                    &ioc->fault_reset_work,
                    msecs_to_jiffies(FAULT_POLLING_INTERVAL));
        spin_unlock_irqrestore(&ioc->ioc_reset_in_progress_lock, flags);
}

/**
 * mpt3sas_base_start_watchdog - start the fault_reset_work_q
 * @ioc: per adapter object
 *
 * Context: sleep.
 */
void
mpt3sas_base_start_watchdog(struct MPT3SAS_ADAPTER *ioc)
{
        unsigned long    flags;

        if (ioc->fault_reset_work_q)
                return;

        /* initialize fault polling */

        INIT_DELAYED_WORK(&ioc->fault_reset_work, _base_fault_reset_work);
        snprintf(ioc->fault_reset_work_q_name,
            sizeof(ioc->fault_reset_work_q_name), "poll_%s%d_status",
            ioc->driver_name, ioc->id);
        ioc->fault_reset_work_q =
                create_singlethread_workqueue(ioc->fault_reset_work_q_name);
        if (!ioc->fault_reset_work_q) {
                ioc_err(ioc, "%s: failed (line=%d)\n", __func__, __LINE__);
                return;
        }
        spin_lock_irqsave(&ioc->ioc_reset_in_progress_lock, flags);
        if (ioc->fault_reset_work_q)
                queue_delayed_work(ioc->fault_reset_work_q,
                    &ioc->fault_reset_work,
                    msecs_to_jiffies(FAULT_POLLING_INTERVAL));
        spin_unlock_irqrestore(&ioc->ioc_reset_in_progress_lock, flags);
}

/**
 * mpt3sas_base_stop_watchdog - stop the fault_reset_work_q
 * @ioc: per adapter object
 *
 * Context: sleep.
 */
void
mpt3sas_base_stop_watchdog(struct MPT3SAS_ADAPTER *ioc)
{
        unsigned long flags;
        struct workqueue_struct *wq;

        spin_lock_irqsave(&ioc->ioc_reset_in_progress_lock, flags);
        wq = ioc->fault_reset_work_q;
        ioc->fault_reset_work_q = NULL;
        spin_unlock_irqrestore(&ioc->ioc_reset_in_progress_lock, flags);
        if (wq) {
                if (!cancel_delayed_work_sync(&ioc->fault_reset_work))
                        flush_workqueue(wq);
                destroy_workqueue(wq);
        }
}

/**
 * mpt3sas_base_fault_info - verbose translation of firmware FAULT code
 * @ioc: per adapter object
 * @fault_code: fault code
 */
void
mpt3sas_base_fault_info(struct MPT3SAS_ADAPTER *ioc , u16 fault_code)
{
        ioc_err(ioc, "fault_state(0x%04x)!\n", fault_code);
}

/**
 * mpt3sas_base_coredump_info - verbose translation of firmware CoreDump state
 * @ioc: per adapter object
 * @fault_code: fault code
 *
 * Return nothing.
 */
void
mpt3sas_base_coredump_info(struct MPT3SAS_ADAPTER *ioc, u16 fault_code)
{
        ioc_err(ioc, "coredump_state(0x%04x)!\n", fault_code);
}

/**
 * mpt3sas_base_wait_for_coredump_completion - Wait until coredump
 * completes or times out
 * @ioc: per adapter object
 * @caller: caller function name
 *
 * Returns 0 for success, non-zero for failure.
 */
int
mpt3sas_base_wait_for_coredump_completion(struct MPT3SAS_ADAPTER *ioc,
                const char *caller)
{
        u8 timeout = (ioc->manu_pg11.CoreDumpTOSec) ?
                        ioc->manu_pg11.CoreDumpTOSec :
                        MPT3SAS_DEFAULT_COREDUMP_TIMEOUT_SECONDS;

        int ioc_state = _base_wait_on_iocstate(ioc, MPI2_IOC_STATE_FAULT,
                                        timeout);

        if (ioc_state)
                ioc_err(ioc,
                    "%s: CoreDump timed out. (ioc_state=0x%x)\n",
                    caller, ioc_state);
        else
                ioc_info(ioc,
                    "%s: CoreDump completed. (ioc_state=0x%x)\n",
                    caller, ioc_state);

        return ioc_state;
}

/**
 * mpt3sas_halt_firmware - halt's mpt controller firmware
 * @ioc: per adapter object
 *
 * For debugging timeout related issues.  Writing 0xCOFFEE00
 * to the doorbell register will halt controller firmware. With
 * the purpose to stop both driver and firmware, the enduser can
 * obtain a ring buffer from controller UART.
 */
void
mpt3sas_halt_firmware(struct MPT3SAS_ADAPTER *ioc)
{
        u32 doorbell;

        if (!ioc->fwfault_debug)
                return;

        dump_stack();

        doorbell = ioc->base_readl(&ioc->chip->Doorbell);
        if ((doorbell & MPI2_IOC_STATE_MASK) == MPI2_IOC_STATE_FAULT) {
                mpt3sas_print_fault_code(ioc, doorbell &
                    MPI2_DOORBELL_DATA_MASK);
        } else if ((doorbell & MPI2_IOC_STATE_MASK) ==
            MPI2_IOC_STATE_COREDUMP) {
                mpt3sas_print_coredump_info(ioc, doorbell &
                    MPI2_DOORBELL_DATA_MASK);
        } else {
                writel(0xC0FFEE00, &ioc->chip->Doorbell);
                ioc_err(ioc, "Firmware is halted due to command timeout\n");
        }

        if (ioc->fwfault_debug == 2)
                for (;;)
                        ;
        else
                panic("panic in %s\n", __func__);
}

/**
 * _base_sas_ioc_info - verbose translation of the ioc status
 * @ioc: per adapter object
 * @mpi_reply: reply mf payload returned from firmware
 * @request_hdr: request mf
 */
static void
_base_sas_ioc_info(struct MPT3SAS_ADAPTER *ioc, MPI2DefaultReply_t *mpi_reply,
        MPI2RequestHeader_t *request_hdr)
{
        u16 ioc_status = le16_to_cpu(mpi_reply->IOCStatus) &
            MPI2_IOCSTATUS_MASK;
        char *desc = NULL;
        u16 frame_sz;
        char *func_str = NULL;

        /* SCSI_IO, RAID_PASS are handled from _scsih_scsi_ioc_info */
        if (request_hdr->Function == MPI2_FUNCTION_SCSI_IO_REQUEST ||
            request_hdr->Function == MPI2_FUNCTION_RAID_SCSI_IO_PASSTHROUGH ||
            request_hdr->Function == MPI2_FUNCTION_EVENT_NOTIFICATION)
                return;

        if (ioc_status == MPI2_IOCSTATUS_CONFIG_INVALID_PAGE)
                return;

        switch (ioc_status) {

/****************************************************************************
*  Common IOCStatus values for all replies
****************************************************************************/

        case MPI2_IOCSTATUS_INVALID_FUNCTION:
                desc = "invalid function";
                break;
        case MPI2_IOCSTATUS_BUSY:
                desc = "busy";
                break;
        case MPI2_IOCSTATUS_INVALID_SGL:
                desc = "invalid sgl";
                break;
        case MPI2_IOCSTATUS_INTERNAL_ERROR:
                desc = "internal error";
                break;
        case MPI2_IOCSTATUS_INVALID_VPID:
                desc = "invalid vpid";
                break;
        case MPI2_IOCSTATUS_INSUFFICIENT_RESOURCES:
                desc = "insufficient resources";
                break;
        case MPI2_IOCSTATUS_INSUFFICIENT_POWER:
                desc = "insufficient power";
                break;
        case MPI2_IOCSTATUS_INVALID_FIELD:
                desc = "invalid field";
                break;
        case MPI2_IOCSTATUS_INVALID_STATE:
                desc = "invalid state";
                break;
        case MPI2_IOCSTATUS_OP_STATE_NOT_SUPPORTED:
                desc = "op state not supported";
                break;

/****************************************************************************
*  Config IOCStatus values
****************************************************************************/

        case MPI2_IOCSTATUS_CONFIG_INVALID_ACTION:
                desc = "config invalid action";
                break;
        case MPI2_IOCSTATUS_CONFIG_INVALID_TYPE:
                desc = "config invalid type";
                break;
        case MPI2_IOCSTATUS_CONFIG_INVALID_PAGE:
                desc = "config invalid page";
                break;
        case MPI2_IOCSTATUS_CONFIG_INVALID_DATA:
                desc = "config invalid data";
                break;
        case MPI2_IOCSTATUS_CONFIG_NO_DEFAULTS:
                desc = "config no defaults";
                break;
        case MPI2_IOCSTATUS_CONFIG_CANT_COMMIT:
                desc = "config cant commit";
                break;

/****************************************************************************
*  SCSI IO Reply
****************************************************************************/

        case MPI2_IOCSTATUS_SCSI_RECOVERED_ERROR:
        case MPI2_IOCSTATUS_SCSI_INVALID_DEVHANDLE:
        case MPI2_IOCSTATUS_SCSI_DEVICE_NOT_THERE:
        case MPI2_IOCSTATUS_SCSI_DATA_OVERRUN:
        case MPI2_IOCSTATUS_SCSI_DATA_UNDERRUN:
        case MPI2_IOCSTATUS_SCSI_IO_DATA_ERROR:
        case MPI2_IOCSTATUS_SCSI_PROTOCOL_ERROR:
        case MPI2_IOCSTATUS_SCSI_TASK_TERMINATED:
        case MPI2_IOCSTATUS_SCSI_RESIDUAL_MISMATCH:
        case MPI2_IOCSTATUS_SCSI_TASK_MGMT_FAILED:
        case MPI2_IOCSTATUS_SCSI_IOC_TERMINATED:
        case MPI2_IOCSTATUS_SCSI_EXT_TERMINATED:
                break;

/****************************************************************************
*  For use by SCSI Initiator and SCSI Target end-to-end data protection
****************************************************************************/

        case MPI2_IOCSTATUS_EEDP_GUARD_ERROR:
                desc = "eedp guard error";
                break;
        case MPI2_IOCSTATUS_EEDP_REF_TAG_ERROR:
                desc = "eedp ref tag error";
                break;
        case MPI2_IOCSTATUS_EEDP_APP_TAG_ERROR:
                desc = "eedp app tag error";
                break;

/****************************************************************************
*  SCSI Target values
****************************************************************************/

        case MPI2_IOCSTATUS_TARGET_INVALID_IO_INDEX:
                desc = "target invalid io index";
                break;
        case MPI2_IOCSTATUS_TARGET_ABORTED:
                desc = "target aborted";
                break;
        case MPI2_IOCSTATUS_TARGET_NO_CONN_RETRYABLE:
                desc = "target no conn retryable";
                break;
        case MPI2_IOCSTATUS_TARGET_NO_CONNECTION:
                desc = "target no connection";
                break;
        case MPI2_IOCSTATUS_TARGET_XFER_COUNT_MISMATCH:
                desc = "target xfer count mismatch";
                break;
        case MPI2_IOCSTATUS_TARGET_DATA_OFFSET_ERROR:
                desc = "target data offset error";
                break;
        case MPI2_IOCSTATUS_TARGET_TOO_MUCH_WRITE_DATA:
                desc = "target too much write data";
                break;
        case MPI2_IOCSTATUS_TARGET_IU_TOO_SHORT:
                desc = "target iu too short";
                break;
        case MPI2_IOCSTATUS_TARGET_ACK_NAK_TIMEOUT:
                desc = "target ack nak timeout";
                break;
        case MPI2_IOCSTATUS_TARGET_NAK_RECEIVED:
                desc = "target nak received";
                break;

/****************************************************************************
*  Serial Attached SCSI values
****************************************************************************/

        case MPI2_IOCSTATUS_SAS_SMP_REQUEST_FAILED:
                desc = "smp request failed";
                break;
        case MPI2_IOCSTATUS_SAS_SMP_DATA_OVERRUN:
                desc = "smp data overrun";
                break;

/****************************************************************************
*  Diagnostic Buffer Post / Diagnostic Release values
****************************************************************************/

        case MPI2_IOCSTATUS_DIAGNOSTIC_RELEASED:
                desc = "diagnostic released";
                break;
        default:
                break;
        }

        if (!desc)
                return;

        switch (request_hdr->Function) {
        case MPI2_FUNCTION_CONFIG:
                frame_sz = sizeof(Mpi2ConfigRequest_t) + ioc->sge_size;
                func_str = "config_page";
                break;
        case MPI2_FUNCTION_SCSI_TASK_MGMT:
                frame_sz = sizeof(Mpi2SCSITaskManagementRequest_t);
                func_str = "task_mgmt";
                break;
        case MPI2_FUNCTION_SAS_IO_UNIT_CONTROL:
                frame_sz = sizeof(Mpi2SasIoUnitControlRequest_t);
                func_str = "sas_iounit_ctl";
                break;
        case MPI2_FUNCTION_SCSI_ENCLOSURE_PROCESSOR:
                frame_sz = sizeof(Mpi2SepRequest_t);
                func_str = "enclosure";
                break;
        case MPI2_FUNCTION_IOC_INIT:
                frame_sz = sizeof(Mpi2IOCInitRequest_t);
                func_str = "ioc_init";
                break;
        case MPI2_FUNCTION_PORT_ENABLE:
                frame_sz = sizeof(Mpi2PortEnableRequest_t);
                func_str = "port_enable";
                break;
        case MPI2_FUNCTION_SMP_PASSTHROUGH:
                frame_sz = sizeof(Mpi2SmpPassthroughRequest_t) + ioc->sge_size;
                func_str = "smp_passthru";
                break;
        case MPI2_FUNCTION_NVME_ENCAPSULATED:
                frame_sz = sizeof(Mpi26NVMeEncapsulatedRequest_t) +
                    ioc->sge_size;
                func_str = "nvme_encapsulated";
                break;
        default:
                frame_sz = 32;
                func_str = "unknown";
                break;
        }

        ioc_warn(ioc, "ioc_status: %s(0x%04x), request(0x%p),(%s)\n",
                 desc, ioc_status, request_hdr, func_str);

        _debug_dump_mf(request_hdr, frame_sz/4);
}

/**
 * _base_display_event_data - verbose translation of firmware asyn events
 * @ioc: per adapter object
 * @mpi_reply: reply mf payload returned from firmware
 */
static void
_base_display_event_data(struct MPT3SAS_ADAPTER *ioc,
        Mpi2EventNotificationReply_t *mpi_reply)
{
        char *desc = NULL;
        u16 event;

        if (!(ioc->logging_level & MPT_DEBUG_EVENTS))
                return;

        event = le16_to_cpu(mpi_reply->Event);

        switch (event) {
        case MPI2_EVENT_LOG_DATA:
                desc = "Log Data";
                break;
        case MPI2_EVENT_STATE_CHANGE:
                desc = "Status Change";
                break;
        case MPI2_EVENT_HARD_RESET_RECEIVED:
                desc = "Hard Reset Received";
                break;
        case MPI2_EVENT_EVENT_CHANGE:
                desc = "Event Change";
                break;
        case MPI2_EVENT_SAS_DEVICE_STATUS_CHANGE:
                desc = "Device Status Change";
                break;
        case MPI2_EVENT_IR_OPERATION_STATUS:
                if (!ioc->hide_ir_msg)
                        desc = "IR Operation Status";
                break;
        case MPI2_EVENT_SAS_DISCOVERY:
        {
                Mpi2EventDataSasDiscovery_t *event_data =
                    (Mpi2EventDataSasDiscovery_t *)mpi_reply->EventData;
                ioc_info(ioc, "Discovery: (%s)",
                         event_data->ReasonCode == MPI2_EVENT_SAS_DISC_RC_STARTED ?
                         "start" : "stop");
                if (event_data->DiscoveryStatus)
                        pr_cont(" discovery_status(0x%08x)",
                            le32_to_cpu(event_data->DiscoveryStatus));
                pr_cont("\n");
                return;
        }
        case MPI2_EVENT_SAS_BROADCAST_PRIMITIVE:
                desc = "SAS Broadcast Primitive";
                break;
        case MPI2_EVENT_SAS_INIT_DEVICE_STATUS_CHANGE:
                desc = "SAS Init Device Status Change";
                break;
        case MPI2_EVENT_SAS_INIT_TABLE_OVERFLOW:
                desc = "SAS Init Table Overflow";
                break;
        case MPI2_EVENT_SAS_TOPOLOGY_CHANGE_LIST:
                desc = "SAS Topology Change List";
                break;
        case MPI2_EVENT_SAS_ENCL_DEVICE_STATUS_CHANGE:
                desc = "SAS Enclosure Device Status Change";
                break;
        case MPI2_EVENT_IR_VOLUME:
                if (!ioc->hide_ir_msg)
                        desc = "IR Volume";
                break;
        case MPI2_EVENT_IR_PHYSICAL_DISK:
                if (!ioc->hide_ir_msg)
                        desc = "IR Physical Disk";
                break;
        case MPI2_EVENT_IR_CONFIGURATION_CHANGE_LIST:
                if (!ioc->hide_ir_msg)
                        desc = "IR Configuration Change List";
                break;
        case MPI2_EVENT_LOG_ENTRY_ADDED:
                if (!ioc->hide_ir_msg)
                        desc = "Log Entry Added";
                break;
        case MPI2_EVENT_TEMP_THRESHOLD:
                desc = "Temperature Threshold";
                break;
        case MPI2_EVENT_ACTIVE_CABLE_EXCEPTION:
                desc = "Cable Event";
                break;
        case MPI2_EVENT_SAS_DEVICE_DISCOVERY_ERROR:
                desc = "SAS Device Discovery Error";
                break;
        case MPI2_EVENT_PCIE_DEVICE_STATUS_CHANGE:
                desc = "PCIE Device Status Change";
                break;
        case MPI2_EVENT_PCIE_ENUMERATION:
        {
                Mpi26EventDataPCIeEnumeration_t *event_data =
                        (Mpi26EventDataPCIeEnumeration_t *)mpi_reply->EventData;
                ioc_info(ioc, "PCIE Enumeration: (%s)",
                         event_data->ReasonCode == MPI26_EVENT_PCIE_ENUM_RC_STARTED ?
                         "start" : "stop");
                if (event_data->EnumerationStatus)
                        pr_cont("enumeration_status(0x%08x)",
                                le32_to_cpu(event_data->EnumerationStatus));
                pr_cont("\n");
                return;
        }
        case MPI2_EVENT_PCIE_TOPOLOGY_CHANGE_LIST:
                desc = "PCIE Topology Change List";
                break;
        }

        if (!desc)
                return;

        ioc_info(ioc, "%s\n", desc);
}

/**
 * _base_sas_log_info - verbose translation of firmware log info
 * @ioc: per adapter object
 * @log_info: log info
 */
static void
_base_sas_log_info(struct MPT3SAS_ADAPTER *ioc , u32 log_info)
{
        union loginfo_type {
                u32     loginfo;
                struct {
                        u32     subcode:16;
                        u32     code:8;
                        u32     originator:4;
                        u32     bus_type:4;
                } dw;
        };
        union loginfo_type sas_loginfo;
        char *originator_str = NULL;

        sas_loginfo.loginfo = log_info;
        if (sas_loginfo.dw.bus_type != 3 /*SAS*/)
                return;

        /* each nexus loss loginfo */
        if (log_info == 0x31170000)
                return;

        /* eat the loginfos associated with task aborts */
        if (ioc->ignore_loginfos && (log_info == 0x30050000 || log_info ==
            0x31140000 || log_info == 0x31130000))
                return;

        switch (sas_loginfo.dw.originator) {
        case 0:
                originator_str = "IOP";
                break;
        case 1:
                originator_str = "PL";
                break;
        case 2:
                if (!ioc->hide_ir_msg)
                        originator_str = "IR";
                else
                        originator_str = "WarpDrive";
                break;
        }

        ioc_warn(ioc, "log_info(0x%08x): originator(%s), code(0x%02x), sub_code(0x%04x)\n",
                 log_info,
                 originator_str, sas_loginfo.dw.code, sas_loginfo.dw.subcode);
}

/**
 * _base_display_reply_info -
 * @ioc: per adapter object
 * @smid: system request message index
 * @msix_index: MSIX table index supplied by the OS
 * @reply: reply message frame(lower 32bit addr)
 */
static void
_base_display_reply_info(struct MPT3SAS_ADAPTER *ioc, u16 smid, u8 msix_index,
        u32 reply)
{
        MPI2DefaultReply_t *mpi_reply;
        u16 ioc_status;
        u32 loginfo = 0;

        mpi_reply = mpt3sas_base_get_reply_virt_addr(ioc, reply);
        if (unlikely(!mpi_reply)) {
                ioc_err(ioc, "mpi_reply not valid at %s:%d/%s()!\n",
                        __FILE__, __LINE__, __func__);
                return;
        }
        ioc_status = le16_to_cpu(mpi_reply->IOCStatus);

        if ((ioc_status & MPI2_IOCSTATUS_MASK) &&
            (ioc->logging_level & MPT_DEBUG_REPLY)) {
                _base_sas_ioc_info(ioc , mpi_reply,
                   mpt3sas_base_get_msg_frame(ioc, smid));
        }

        if (ioc_status & MPI2_IOCSTATUS_FLAG_LOG_INFO_AVAILABLE) {
                loginfo = le32_to_cpu(mpi_reply->IOCLogInfo);
                _base_sas_log_info(ioc, loginfo);
        }

        if (ioc_status || loginfo) {
                ioc_status &= MPI2_IOCSTATUS_MASK;
                mpt3sas_trigger_mpi(ioc, ioc_status, loginfo);
        }
}

/**
 * mpt3sas_base_done - base internal command completion routine
 * @ioc: per adapter object
 * @smid: system request message index
 * @msix_index: MSIX table index supplied by the OS
 * @reply: reply message frame(lower 32bit addr)
 *
 * Return:
 * 1 meaning mf should be freed from _base_interrupt
 * 0 means the mf is freed from this function.
 */
u8
mpt3sas_base_done(struct MPT3SAS_ADAPTER *ioc, u16 smid, u8 msix_index,
        u32 reply)
{
        MPI2DefaultReply_t *mpi_reply;

        mpi_reply = mpt3sas_base_get_reply_virt_addr(ioc, reply);
        if (mpi_reply && mpi_reply->Function == MPI2_FUNCTION_EVENT_ACK)
                return mpt3sas_check_for_pending_internal_cmds(ioc, smid);

        if (ioc->base_cmds.status == MPT3_CMD_NOT_USED)
                return 1;

        ioc->base_cmds.status |= MPT3_CMD_COMPLETE;
        if (mpi_reply) {
                ioc->base_cmds.status |= MPT3_CMD_REPLY_VALID;
                memcpy(ioc->base_cmds.reply, mpi_reply, mpi_reply->MsgLength*4);
        }
        ioc->base_cmds.status &= ~MPT3_CMD_PENDING;

        complete(&ioc->base_cmds.done);
        return 1;
}

/**
 * _base_async_event - main callback handler for firmware asyn events
 * @ioc: per adapter object
 * @msix_index: MSIX table index supplied by the OS
 * @reply: reply message frame(lower 32bit addr)
 *
 * Return:
 * 1 meaning mf should be freed from _base_interrupt
 * 0 means the mf is freed from this function.
 */
static u8
_base_async_event(struct MPT3SAS_ADAPTER *ioc, u8 msix_index, u32 reply)
{
        Mpi2EventNotificationReply_t *mpi_reply;
        Mpi2EventAckRequest_t *ack_request;
        u16 smid;
        struct _event_ack_list *delayed_event_ack;

        mpi_reply = mpt3sas_base_get_reply_virt_addr(ioc, reply);
        if (!mpi_reply)
                return 1;
        if (mpi_reply->Function != MPI2_FUNCTION_EVENT_NOTIFICATION)
                return 1;

        _base_display_event_data(ioc, mpi_reply);

        if (!(mpi_reply->AckRequired & MPI2_EVENT_NOTIFICATION_ACK_REQUIRED))
                goto out;
        smid = mpt3sas_base_get_smid(ioc, ioc->base_cb_idx);
        if (!smid) {
                delayed_event_ack = kzalloc(sizeof(*delayed_event_ack),
                                        GFP_ATOMIC);
                if (!delayed_event_ack)
                        goto out;
                INIT_LIST_HEAD(&delayed_event_ack->list);
                delayed_event_ack->Event = mpi_reply->Event;
                delayed_event_ack->EventContext = mpi_reply->EventContext;
                list_add_tail(&delayed_event_ack->list,
                                &ioc->delayed_event_ack_list);
                dewtprintk(ioc,
                           ioc_info(ioc, "DELAYED: EVENT ACK: event (0x%04x)\n",
                                    le16_to_cpu(mpi_reply->Event)));
                goto out;
        }

        ack_request = mpt3sas_base_get_msg_frame(ioc, smid);
        memset(ack_request, 0, sizeof(Mpi2EventAckRequest_t));
        ack_request->Function = MPI2_FUNCTION_EVENT_ACK;
        ack_request->Event = mpi_reply->Event;
        ack_request->EventContext = mpi_reply->EventContext;
        ack_request->VF_ID = 0;  /* TODO */
        ack_request->VP_ID = 0;
        ioc->put_smid_default(ioc, smid);

 out:

        /* scsih callback handler */
        mpt3sas_scsih_event_callback(ioc, msix_index, reply);

        /* ctl callback handler */
        mpt3sas_ctl_event_callback(ioc, msix_index, reply);

        return 1;
}

static struct scsiio_tracker *
_get_st_from_smid(struct MPT3SAS_ADAPTER *ioc, u16 smid)
{
        struct scsi_cmnd *cmd;

        if (WARN_ON(!smid) ||
            WARN_ON(smid >= ioc->hi_priority_smid))
                return NULL;

        cmd = mpt3sas_scsih_scsi_lookup_get(ioc, smid);
        if (cmd)
                return scsi_cmd_priv(cmd);

        return NULL;
}

/**
 * _base_get_cb_idx - obtain the callback index
 * @ioc: per adapter object
 * @smid: system request message index
 *
 * Return: callback index.
 */
static u8
_base_get_cb_idx(struct MPT3SAS_ADAPTER *ioc, u16 smid)
{
        int i;
        u16 ctl_smid = ioc->scsiio_depth - INTERNAL_SCSIIO_CMDS_COUNT + 1;
        u8 cb_idx = 0xFF;

        if (smid < ioc->hi_priority_smid) {
                struct scsiio_tracker *st;

                if (smid < ctl_smid) {
                        st = _get_st_from_smid(ioc, smid);
                        if (st)
                                cb_idx = st->cb_idx;
                } else if (smid == ctl_smid)
                        cb_idx = ioc->ctl_cb_idx;
        } else if (smid < ioc->internal_smid) {
                i = smid - ioc->hi_priority_smid;
                cb_idx = ioc->hpr_lookup[i].cb_idx;
        } else if (smid <= ioc->hba_queue_depth) {
                i = smid - ioc->internal_smid;
                cb_idx = ioc->internal_lookup[i].cb_idx;
        }
        return cb_idx;
}

/**
 * mpt3sas_base_mask_interrupts - disable interrupts
 * @ioc: per adapter object
 *
 * Disabling ResetIRQ, Reply and Doorbell Interrupts
 */
void
mpt3sas_base_mask_interrupts(struct MPT3SAS_ADAPTER *ioc)
{
        u32 him_register;

        ioc->mask_interrupts = 1;
        him_register = ioc->base_readl(&ioc->chip->HostInterruptMask);
        him_register |= MPI2_HIM_DIM + MPI2_HIM_RIM + MPI2_HIM_RESET_IRQ_MASK;
        writel(him_register, &ioc->chip->HostInterruptMask);
        ioc->base_readl(&ioc->chip->HostInterruptMask);
}

/**
 * mpt3sas_base_unmask_interrupts - enable interrupts
 * @ioc: per adapter object
 *
 * Enabling only Reply Interrupts
 */
void
mpt3sas_base_unmask_interrupts(struct MPT3SAS_ADAPTER *ioc)
{
        u32 him_register;

        him_register = ioc->base_readl(&ioc->chip->HostInterruptMask);
        him_register &= ~MPI2_HIM_RIM;
        writel(him_register, &ioc->chip->HostInterruptMask);
        ioc->mask_interrupts = 0;
}

union reply_descriptor {
        u64 word;
        struct {
                u32 low;
                u32 high;
        } u;
};

static u32 base_mod64(u64 dividend, u32 divisor)
{
        u32 remainder;

        if (!divisor)
                pr_err("mpt3sas: DIVISOR is zero, in div fn\n");
        remainder = do_div(dividend, divisor);
        return remainder;
}

/**
 * _base_process_reply_queue - Process reply descriptors from reply
 *              descriptor post queue.
 * @reply_q: per IRQ's reply queue object.
 *
 * Return: number of reply descriptors processed from reply
 *              descriptor queue.
 */
static int
_base_process_reply_queue(struct adapter_reply_queue *reply_q)
{
        union reply_descriptor rd;
        u64 completed_cmds;
        u8 request_descript_type;
        u16 smid;
        u8 cb_idx;
        u32 reply;
        u8 msix_index = reply_q->msix_index;
        struct MPT3SAS_ADAPTER *ioc = reply_q->ioc;
        Mpi2ReplyDescriptorsUnion_t *rpf;
        u8 rc;

        completed_cmds = 0;
        if (!atomic_add_unless(&reply_q->busy, 1, 1))
                return completed_cmds;

        rpf = &reply_q->reply_post_free[reply_q->reply_post_host_index];
        request_descript_type = rpf->Default.ReplyFlags
             & MPI2_RPY_DESCRIPT_FLAGS_TYPE_MASK;
        if (request_descript_type == MPI2_RPY_DESCRIPT_FLAGS_UNUSED) {
                atomic_dec(&reply_q->busy);
                return completed_cmds;
        }

        cb_idx = 0xFF;
        do {
                rd.word = le64_to_cpu(rpf->Words);
                if (rd.u.low == UINT_MAX || rd.u.high == UINT_MAX)
                        goto out;
                reply = 0;
                smid = le16_to_cpu(rpf->Default.DescriptorTypeDependent1);
                if (request_descript_type ==
                    MPI25_RPY_DESCRIPT_FLAGS_FAST_PATH_SCSI_IO_SUCCESS ||
                    request_descript_type ==
                    MPI2_RPY_DESCRIPT_FLAGS_SCSI_IO_SUCCESS ||
                    request_descript_type ==
                    MPI26_RPY_DESCRIPT_FLAGS_PCIE_ENCAPSULATED_SUCCESS) {
                        cb_idx = _base_get_cb_idx(ioc, smid);
                        if ((likely(cb_idx < MPT_MAX_CALLBACKS)) &&
                            (likely(mpt_callbacks[cb_idx] != NULL))) {
                                rc = mpt_callbacks[cb_idx](ioc, smid,
                                    msix_index, 0);
                                if (rc)
                                        mpt3sas_base_free_smid(ioc, smid);
                        }
                } else if (request_descript_type ==
                    MPI2_RPY_DESCRIPT_FLAGS_ADDRESS_REPLY) {
                        reply = le32_to_cpu(
                            rpf->AddressReply.ReplyFrameAddress);
                        if (reply > ioc->reply_dma_max_address ||
                            reply < ioc->reply_dma_min_address)
                                reply = 0;
                        if (smid) {
                                cb_idx = _base_get_cb_idx(ioc, smid);
                                if ((likely(cb_idx < MPT_MAX_CALLBACKS)) &&
                                    (likely(mpt_callbacks[cb_idx] != NULL))) {
                                        rc = mpt_callbacks[cb_idx](ioc, smid,
                                            msix_index, reply);
                                        if (reply)
                                                _base_display_reply_info(ioc,
                                                    smid, msix_index, reply);
                                        if (rc)
                                                mpt3sas_base_free_smid(ioc,
                                                    smid);
                                }
                        } else {
                                _base_async_event(ioc, msix_index, reply);
                        }

                        /* reply free queue handling */
                        if (reply) {
                                ioc->reply_free_host_index =
                                    (ioc->reply_free_host_index ==
                                    (ioc->reply_free_queue_depth - 1)) ?
                                    0 : ioc->reply_free_host_index + 1;
                                ioc->reply_free[ioc->reply_free_host_index] =
                                    cpu_to_le32(reply);
                                if (ioc->is_mcpu_endpoint)
                                        _base_clone_reply_to_sys_mem(ioc,
                                                reply,
                                                ioc->reply_free_host_index);
                                writel(ioc->reply_free_host_index,
                                    &ioc->chip->ReplyFreeHostIndex);
                        }
                }

                rpf->Words = cpu_to_le64(ULLONG_MAX);
                reply_q->reply_post_host_index =
                    (reply_q->reply_post_host_index ==
                    (ioc->reply_post_queue_depth - 1)) ? 0 :
                    reply_q->reply_post_host_index + 1;
                request_descript_type =
                    reply_q->reply_post_free[reply_q->reply_post_host_index].
                    Default.ReplyFlags & MPI2_RPY_DESCRIPT_FLAGS_TYPE_MASK;
                completed_cmds++;
                /* Update the reply post host index after continuously
                 * processing the threshold number of Reply Descriptors.
                 * So that FW can find enough entries to post the Reply
                 * Descriptors in the reply descriptor post queue.
                 */
                if (completed_cmds >= ioc->thresh_hold) {
                        if (ioc->combined_reply_queue) {
                                writel(reply_q->reply_post_host_index |
                                                ((msix_index  & 7) <<
                                                 MPI2_RPHI_MSIX_INDEX_SHIFT),
                                    ioc->replyPostRegisterIndex[msix_index/8]);
                        } else {
                                writel(reply_q->reply_post_host_index |
                                                (msix_index <<
                                                 MPI2_RPHI_MSIX_INDEX_SHIFT),
                                                &ioc->chip->ReplyPostHostIndex);
                        }
                        if (!reply_q->irq_poll_scheduled) {
                                reply_q->irq_poll_scheduled = true;
                                irq_poll_sched(&reply_q->irqpoll);
                        }
                        atomic_dec(&reply_q->busy);
                        return completed_cmds;
                }
                if (request_descript_type == MPI2_RPY_DESCRIPT_FLAGS_UNUSED)
                        goto out;
                if (!reply_q->reply_post_host_index)
                        rpf = reply_q->reply_post_free;
                else
                        rpf++;
        } while (1);

 out:

        if (!completed_cmds) {
                atomic_dec(&reply_q->busy);
                return completed_cmds;
        }

        if (ioc->is_warpdrive) {
                writel(reply_q->reply_post_host_index,
                ioc->reply_post_host_index[msix_index]);
                atomic_dec(&reply_q->busy);
                return completed_cmds;
        }

        /* Update Reply Post Host Index.
         * For those HBA's which support combined reply queue feature
         * 1. Get the correct Supplemental Reply Post Host Index Register.
         *    i.e. (msix_index / 8)th entry from Supplemental Reply Post Host
         *    Index Register address bank i.e replyPostRegisterIndex[],
         * 2. Then update this register with new reply host index value
         *    in ReplyPostIndex field and the MSIxIndex field with
         *    msix_index value reduced to a value between 0 and 7,
         *    using a modulo 8 operation. Since each Supplemental Reply Post
         *    Host Index Register supports 8 MSI-X vectors.
         *
         * For other HBA's just update the Reply Post Host Index register with
         * new reply host index value in ReplyPostIndex Field and msix_index
         * value in MSIxIndex field.
         */
        if (ioc->combined_reply_queue)
                writel(reply_q->reply_post_host_index | ((msix_index  & 7) <<
                        MPI2_RPHI_MSIX_INDEX_SHIFT),
                        ioc->replyPostRegisterIndex[msix_index/8]);
        else
                writel(reply_q->reply_post_host_index | (msix_index <<
                        MPI2_RPHI_MSIX_INDEX_SHIFT),
                        &ioc->chip->ReplyPostHostIndex);
        atomic_dec(&reply_q->busy);
        return completed_cmds;
}

/**
 * _base_interrupt - MPT adapter (IOC) specific interrupt handler.
 * @irq: irq number (not used)
 * @bus_id: bus identifier cookie == pointer to MPT_ADAPTER structure
 *
 * Return: IRQ_HANDLED if processed, else IRQ_NONE.
 */
static irqreturn_t
_base_interrupt(int irq, void *bus_id)
{
        struct adapter_reply_queue *reply_q = bus_id;
        struct MPT3SAS_ADAPTER *ioc = reply_q->ioc;

        if (ioc->mask_interrupts)
                return IRQ_NONE;
        if (reply_q->irq_poll_scheduled)
                return IRQ_HANDLED;
        return ((_base_process_reply_queue(reply_q) > 0) ?
                        IRQ_HANDLED : IRQ_NONE);
}

/**
 * _base_irqpoll - IRQ poll callback handler
 * @irqpoll: irq_poll object
 * @budget: irq poll weight
 *
 * returns number of reply descriptors processed
 */
static int
_base_irqpoll(struct irq_poll *irqpoll, int budget)
{
        struct adapter_reply_queue *reply_q;
        int num_entries = 0;

        reply_q = container_of(irqpoll, struct adapter_reply_queue,
                        irqpoll);
        if (reply_q->irq_line_enable) {
                disable_irq_nosync(reply_q->os_irq);
                reply_q->irq_line_enable = false;
        }
        num_entries = _base_process_reply_queue(reply_q);
        if (num_entries < budget) {
                irq_poll_complete(irqpoll);
                reply_q->irq_poll_scheduled = false;
                reply_q->irq_line_enable = true;
                enable_irq(reply_q->os_irq);
                /*
                 * Go for one more round of processing the
                 * reply descriptor post queue incase if HBA
                 * Firmware has posted some reply descriptors
                 * while reenabling the IRQ.
                 */
                _base_process_reply_queue(reply_q);
        }

        return num_entries;
}

/**
 * _base_init_irqpolls - initliaze IRQ polls
 * @ioc: per adapter object
 *
 * returns nothing
 */
static void
_base_init_irqpolls(struct MPT3SAS_ADAPTER *ioc)
{
        struct adapter_reply_queue *reply_q, *next;

        if (list_empty(&ioc->reply_queue_list))
                return;

        list_for_each_entry_safe(reply_q, next, &ioc->reply_queue_list, list) {
                irq_poll_init(&reply_q->irqpoll,
                        ioc->hba_queue_depth/4, _base_irqpoll);
                reply_q->irq_poll_scheduled = false;
                reply_q->irq_line_enable = true;
                reply_q->os_irq = pci_irq_vector(ioc->pdev,
                    reply_q->msix_index);
        }
}

/**
 * _base_is_controller_msix_enabled - is controller support muli-reply queues
 * @ioc: per adapter object
 *
 * Return: Whether or not MSI/X is enabled.
 */
static inline int
_base_is_controller_msix_enabled(struct MPT3SAS_ADAPTER *ioc)
{
        return (ioc->facts.IOCCapabilities &
            MPI2_IOCFACTS_CAPABILITY_MSI_X_INDEX) && ioc->msix_enable;
}

/**
 * mpt3sas_base_sync_reply_irqs - flush pending MSIX interrupts
 * @ioc: per adapter object
 * @poll: poll over reply descriptor pools incase interrupt for
 *              timed-out SCSI command got delayed
 * Context: non ISR conext
 *
 * Called when a Task Management request has completed.
 */
void
mpt3sas_base_sync_reply_irqs(struct MPT3SAS_ADAPTER *ioc, u8 poll)
{
        struct adapter_reply_queue *reply_q;

        /* If MSIX capability is turned off
         * then multi-queues are not enabled
         */
        if (!_base_is_controller_msix_enabled(ioc))
                return;

        list_for_each_entry(reply_q, &ioc->reply_queue_list, list) {
                if (ioc->shost_recovery || ioc->remove_host ||
                                ioc->pci_error_recovery)
                        return;
                /* TMs are on msix_index == 0 */
                if (reply_q->msix_index == 0)
                        continue;
                synchronize_irq(pci_irq_vector(ioc->pdev, reply_q->msix_index));
                if (reply_q->irq_poll_scheduled) {
                        /* Calling irq_poll_disable will wait for any pending
                         * callbacks to have completed.
                         */
                        irq_poll_disable(&reply_q->irqpoll);
                        irq_poll_enable(&reply_q->irqpoll);
                        /* check how the scheduled poll has ended,
                         * clean up only if necessary
                         */
                        if (reply_q->irq_poll_scheduled) {
                                reply_q->irq_poll_scheduled = false;
                                reply_q->irq_line_enable = true;
                                enable_irq(reply_q->os_irq);
                        }
                }
        }
        if (poll)
                _base_process_reply_queue(reply_q);
}

/**
 * mpt3sas_base_release_callback_handler - clear interrupt callback handler
 * @cb_idx: callback index
 */
void
mpt3sas_base_release_callback_handler(u8 cb_idx)
{
        mpt_callbacks[cb_idx] = NULL;
}

/**
 * mpt3sas_base_register_callback_handler - obtain index for the interrupt callback handler
 * @cb_func: callback function
 *
 * Return: Index of @cb_func.
 */
u8
mpt3sas_base_register_callback_handler(MPT_CALLBACK cb_func)
{
        u8 cb_idx;

        for (cb_idx = MPT_MAX_CALLBACKS-1; cb_idx; cb_idx--)
                if (mpt_callbacks[cb_idx] == NULL)
                        break;

        mpt_callbacks[cb_idx] = cb_func;
        return cb_idx;
}

/**
 * mpt3sas_base_initialize_callback_handler - initialize the interrupt callback handler
 */
void
mpt3sas_base_initialize_callback_handler(void)
{
        u8 cb_idx;

        for (cb_idx = 0; cb_idx < MPT_MAX_CALLBACKS; cb_idx++)
                mpt3sas_base_release_callback_handler(cb_idx);
}


/**
 * _base_build_zero_len_sge - build zero length sg entry
 * @ioc: per adapter object
 * @paddr: virtual address for SGE
 *
 * Create a zero length scatter gather entry to insure the IOCs hardware has
 * something to use if the target device goes brain dead and tries
 * to send data even when none is asked for.
 */
static void
_base_build_zero_len_sge(struct MPT3SAS_ADAPTER *ioc, void *paddr)
{
        u32 flags_length = (u32)((MPI2_SGE_FLAGS_LAST_ELEMENT |
            MPI2_SGE_FLAGS_END_OF_BUFFER | MPI2_SGE_FLAGS_END_OF_LIST |
            MPI2_SGE_FLAGS_SIMPLE_ELEMENT) <<
            MPI2_SGE_FLAGS_SHIFT);
        ioc->base_add_sg_single(paddr, flags_length, -1);
}

/**
 * _base_add_sg_single_32 - Place a simple 32 bit SGE at address pAddr.
 * @paddr: virtual address for SGE
 * @flags_length: SGE flags and data transfer length
 * @dma_addr: Physical address
 */
static void
_base_add_sg_single_32(void *paddr, u32 flags_length, dma_addr_t dma_addr)
{
        Mpi2SGESimple32_t *sgel = paddr;

        flags_length |= (MPI2_SGE_FLAGS_32_BIT_ADDRESSING |
            MPI2_SGE_FLAGS_SYSTEM_ADDRESS) << MPI2_SGE_FLAGS_SHIFT;
        sgel->FlagsLength = cpu_to_le32(flags_length);
        sgel->Address = cpu_to_le32(dma_addr);
}


/**
 * _base_add_sg_single_64 - Place a simple 64 bit SGE at address pAddr.
 * @paddr: virtual address for SGE
 * @flags_length: SGE flags and data transfer length
 * @dma_addr: Physical address
 */
static void
_base_add_sg_single_64(void *paddr, u32 flags_length, dma_addr_t dma_addr)
{
        Mpi2SGESimple64_t *sgel = paddr;

        flags_length |= (MPI2_SGE_FLAGS_64_BIT_ADDRESSING |
            MPI2_SGE_FLAGS_SYSTEM_ADDRESS) << MPI2_SGE_FLAGS_SHIFT;
        sgel->FlagsLength = cpu_to_le32(flags_length);
        sgel->Address = cpu_to_le64(dma_addr);
}

/**
 * _base_get_chain_buffer_tracker - obtain chain tracker
 * @ioc: per adapter object
 * @scmd: SCSI commands of the IO request
 *
 * Return: chain tracker from chain_lookup table using key as
 * smid and smid's chain_offset.
 */
static struct chain_tracker *
_base_get_chain_buffer_tracker(struct MPT3SAS_ADAPTER *ioc,
                               struct scsi_cmnd *scmd)
{
        struct chain_tracker *chain_req;
        struct scsiio_tracker *st = scsi_cmd_priv(scmd);
        u16 smid = st->smid;
        u8 chain_offset =
           atomic_read(&ioc->chain_lookup[smid - 1].chain_offset);

        if (chain_offset == ioc->chains_needed_per_io)
                return NULL;

        chain_req = &ioc->chain_lookup[smid - 1].chains_per_smid[chain_offset];
        atomic_inc(&ioc->chain_lookup[smid - 1].chain_offset);
        return chain_req;
}


/**
 * _base_build_sg - build generic sg
 * @ioc: per adapter object
 * @psge: virtual address for SGE
 * @data_out_dma: physical address for WRITES
 * @data_out_sz: data xfer size for WRITES
 * @data_in_dma: physical address for READS
 * @data_in_sz: data xfer size for READS
 */
static void
_base_build_sg(struct MPT3SAS_ADAPTER *ioc, void *psge,
        dma_addr_t data_out_dma, size_t data_out_sz, dma_addr_t data_in_dma,
        size_t data_in_sz)
{
        u32 sgl_flags;

        if (!data_out_sz && !data_in_sz) {
                _base_build_zero_len_sge(ioc, psge);
                return;
        }

        if (data_out_sz && data_in_sz) {
                /* WRITE sgel first */
                sgl_flags = (MPI2_SGE_FLAGS_SIMPLE_ELEMENT |
                    MPI2_SGE_FLAGS_END_OF_BUFFER | MPI2_SGE_FLAGS_HOST_TO_IOC);
                sgl_flags = sgl_flags << MPI2_SGE_FLAGS_SHIFT;
                ioc->base_add_sg_single(psge, sgl_flags |
                    data_out_sz, data_out_dma);

                /* incr sgel */
                psge += ioc->sge_size;

                /* READ sgel last */
                sgl_flags = (MPI2_SGE_FLAGS_SIMPLE_ELEMENT |
                    MPI2_SGE_FLAGS_LAST_ELEMENT | MPI2_SGE_FLAGS_END_OF_BUFFER |
                    MPI2_SGE_FLAGS_END_OF_LIST);
                sgl_flags = sgl_flags << MPI2_SGE_FLAGS_SHIFT;
                ioc->base_add_sg_single(psge, sgl_flags |

                    data_in_sz, data_in_dma);
        } else if (data_out_sz) /* WRITE */ {
                sgl_flags = (MPI2_SGE_FLAGS_SIMPLE_ELEMENT |
                    MPI2_SGE_FLAGS_LAST_ELEMENT | MPI2_SGE_FLAGS_END_OF_BUFFER |
                    MPI2_SGE_FLAGS_END_OF_LIST | MPI2_SGE_FLAGS_HOST_TO_IOC);
                sgl_flags = sgl_flags << MPI2_SGE_FLAGS_SHIFT;
                ioc->base_add_sg_single(psge, sgl_flags |
                    data_out_sz, data_out_dma);
        } else if (data_in_sz) /* READ */ {
                sgl_flags = (MPI2_SGE_FLAGS_SIMPLE_ELEMENT |
                    MPI2_SGE_FLAGS_LAST_ELEMENT | MPI2_SGE_FLAGS_END_OF_BUFFER |
                    MPI2_SGE_FLAGS_END_OF_LIST);
                sgl_flags = sgl_flags << MPI2_SGE_FLAGS_SHIFT;
                ioc->base_add_sg_single(psge, sgl_flags |
                    data_in_sz, data_in_dma);
        }
}

/* IEEE format sgls */

/**
 * _base_build_nvme_prp - This function is called for NVMe end devices to build
 * a native SGL (NVMe PRP). The native SGL is built starting in the first PRP
 * entry of the NVMe message (PRP1).  If the data buffer is small enough to be
 * described entirely using PRP1, then PRP2 is not used.  If needed, PRP2 is
 * used to describe a larger data buffer.  If the data buffer is too large to
 * describe using the two PRP entriess inside the NVMe message, then PRP1
 * describes the first data memory segment, and PRP2 contains a pointer to a PRP
 * list located elsewhere in memory to describe the remaining data memory
 * segments.  The PRP list will be contiguous.
 *
 * The native SGL for NVMe devices is a Physical Region Page (PRP).  A PRP
 * consists of a list of PRP entries to describe a number of noncontigous
 * physical memory segments as a single memory buffer, just as a SGL does.  Note
 * however, that this function is only used by the IOCTL call, so the memory
 * given will be guaranteed to be contiguous.  There is no need to translate
 * non-contiguous SGL into a PRP in this case.  All PRPs will describe
 * contiguous space that is one page size each.
 *
 * Each NVMe message contains two PRP entries.  The first (PRP1) either contains
 * a PRP list pointer or a PRP element, depending upon the command.  PRP2
 * contains the second PRP element if the memory being described fits within 2
 * PRP entries, or a PRP list pointer if the PRP spans more than two entries.
 *
 * A PRP list pointer contains the address of a PRP list, structured as a linear
 * array of PRP entries.  Each PRP entry in this list describes a segment of
 * physical memory.
 *
 * Each 64-bit PRP entry comprises an address and an offset field.  The address
 * always points at the beginning of a 4KB physical memory page, and the offset
 * describes where within that 4KB page the memory segment begins.  Only the
 * first element in a PRP list may contain a non-zero offest, implying that all
 * memory segments following the first begin at the start of a 4KB page.
 *
 * Each PRP element normally describes 4KB of physical memory, with exceptions
 * for the first and last elements in the list.  If the memory being described
 * by the list begins at a non-zero offset within the first 4KB page, then the
 * first PRP element will contain a non-zero offset indicating where the region
 * begins within the 4KB page.  The last memory segment may end before the end
 * of the 4KB segment, depending upon the overall size of the memory being
 * described by the PRP list.
 *
 * Since PRP entries lack any indication of size, the overall data buffer length
 * is used to determine where the end of the data memory buffer is located, and
 * how many PRP entries are required to describe it.
 *
 * @ioc: per adapter object
 * @smid: system request message index for getting asscociated SGL
 * @nvme_encap_request: the NVMe request msg frame pointer
 * @data_out_dma: physical address for WRITES
 * @data_out_sz: data xfer size for WRITES
 * @data_in_dma: physical address for READS
 * @data_in_sz: data xfer size for READS
 */
static void
_base_build_nvme_prp(struct MPT3SAS_ADAPTER *ioc, u16 smid,
        Mpi26NVMeEncapsulatedRequest_t *nvme_encap_request,
        dma_addr_t data_out_dma, size_t data_out_sz, dma_addr_t data_in_dma,
        size_t data_in_sz)
{
        int             prp_size = NVME_PRP_SIZE;
        __le64          *prp_entry, *prp1_entry, *prp2_entry;
        __le64          *prp_page;
        dma_addr_t      prp_entry_dma, prp_page_dma, dma_addr;
        u32             offset, entry_len;
        u32             page_mask_result, page_mask;
        size_t          length;
        struct mpt3sas_nvme_cmd *nvme_cmd =
                (void *)nvme_encap_request->NVMe_Command;

        /*
         * Not all commands require a data transfer. If no data, just return
         * without constructing any PRP.
         */
        if (!data_in_sz && !data_out_sz)
                return;
        prp1_entry = &nvme_cmd->prp1;
        prp2_entry = &nvme_cmd->prp2;
        prp_entry = prp1_entry;
        /*
         * For the PRP entries, use the specially allocated buffer of
         * contiguous memory.
         */
        prp_page = (__le64 *)mpt3sas_base_get_pcie_sgl(ioc, smid);
        prp_page_dma = mpt3sas_base_get_pcie_sgl_dma(ioc, smid);

        /*
         * Check if we are within 1 entry of a page boundary we don't
         * want our first entry to be a PRP List entry.
         */
        page_mask = ioc->page_size - 1;
        page_mask_result = (uintptr_t)((u8 *)prp_page + prp_size) & page_mask;
        if (!page_mask_result) {
                /* Bump up to next page boundary. */
                prp_page = (__le64 *)((u8 *)prp_page + prp_size);
                prp_page_dma = prp_page_dma + prp_size;
        }

        /*
         * Set PRP physical pointer, which initially points to the current PRP
         * DMA memory page.
         */
        prp_entry_dma = prp_page_dma;

        /* Get physical address and length of the data buffer. */
        if (data_in_sz) {
                dma_addr = data_in_dma;
                length = data_in_sz;
        } else {
                dma_addr = data_out_dma;
                length = data_out_sz;
        }

        /* Loop while the length is not zero. */
        while (length) {
                /*
                 * Check if we need to put a list pointer here if we are at
                 * page boundary - prp_size (8 bytes).
                 */
                page_mask_result = (prp_entry_dma + prp_size) & page_mask;
                if (!page_mask_result) {
                        /*
                         * This is the last entry in a PRP List, so we need to
                         * put a PRP list pointer here.  What this does is:
                         *   - bump the current memory pointer to the next
                         *     address, which will be the next full page.
                         *   - set the PRP Entry to point to that page.  This
                         *     is now the PRP List pointer.
                         *   - bump the PRP Entry pointer the start of the
                         *     next page.  Since all of this PRP memory is
                         *     contiguous, no need to get a new page - it's
                         *     just the next address.
                         */
                        prp_entry_dma++;
                        *prp_entry = cpu_to_le64(prp_entry_dma);
                        prp_entry++;
                }

                /* Need to handle if entry will be part of a page. */
                offset = dma_addr & page_mask;
                entry_len = ioc->page_size - offset;

                if (prp_entry == prp1_entry) {
                        /*
                         * Must fill in the first PRP pointer (PRP1) before
                         * moving on.
                         */
                        *prp1_entry = cpu_to_le64(dma_addr);

                        /*
                         * Now point to the second PRP entry within the
                         * command (PRP2).
                         */
                        prp_entry = prp2_entry;
                } else if (prp_entry == prp2_entry) {
                        /*
                         * Should the PRP2 entry be a PRP List pointer or just
                         * a regular PRP pointer?  If there is more than one
                         * more page of data, must use a PRP List pointer.
                         */
                        if (length > ioc->page_size) {
                                /*
                                 * PRP2 will contain a PRP List pointer because
                                 * more PRP's are needed with this command. The
                                 * list will start at the beginning of the
                                 * contiguous buffer.
                                 */
                                *prp2_entry = cpu_to_le64(prp_entry_dma);

                                /*
                                 * The next PRP Entry will be the start of the
                                 * first PRP List.
                                 */
                                prp_entry = prp_page;
                        } else {
                                /*
                                 * After this, the PRP Entries are complete.
                                 * This command uses 2 PRP's and no PRP list.
                                 */
                                *prp2_entry = cpu_to_le64(dma_addr);
                        }
                } else {
                        /*
                         * Put entry in list and bump the addresses.
                         *
                         * After PRP1 and PRP2 are filled in, this will fill in
                         * all remaining PRP entries in a PRP List, one per
                         * each time through the loop.
                         */
                        *prp_entry = cpu_to_le64(dma_addr);
                        prp_entry++;
                        prp_entry_dma++;
                }

                /*
                 * Bump the phys address of the command's data buffer by the
                 * entry_len.
                 */
                dma_addr += entry_len;

                /* Decrement length accounting for last partial page. */
                if (entry_len > length)
                        length = 0;
                else
                        length -= entry_len;
        }
}

/**
 * base_make_prp_nvme -
 * Prepare PRPs(Physical Region Page)- SGLs specific to NVMe drives only
 *
 * @ioc:                per adapter object
 * @scmd:               SCSI command from the mid-layer
 * @mpi_request:        mpi request
 * @smid:               msg Index
 * @sge_count:          scatter gather element count.
 *
 * Return:              true: PRPs are built
 *                      false: IEEE SGLs needs to be built
 */
static void
base_make_prp_nvme(struct MPT3SAS_ADAPTER *ioc,
                struct scsi_cmnd *scmd,
                Mpi25SCSIIORequest_t *mpi_request,
                u16 smid, int sge_count)
{
        int sge_len, num_prp_in_chain = 0;
        Mpi25IeeeSgeChain64_t *main_chain_element, *ptr_first_sgl;
        __le64 *curr_buff;
        dma_addr_t msg_dma, sge_addr, offset;
        u32 page_mask, page_mask_result;
        struct scatterlist *sg_scmd;
        u32 first_prp_len;
        int data_len = scsi_bufflen(scmd);
        u32 nvme_pg_size;

        nvme_pg_size = max_t(u32, ioc->page_size, NVME_PRP_PAGE_SIZE);
        /*
         * Nvme has a very convoluted prp format.  One prp is required
         * for each page or partial page. Driver need to split up OS sg_list
         * entries if it is longer than one page or cross a page
         * boundary.  Driver also have to insert a PRP list pointer entry as
         * the last entry in each physical page of the PRP list.
         *
         * NOTE: The first PRP "entry" is actually placed in the first
         * SGL entry in the main message as IEEE 64 format.  The 2nd
         * entry in the main message is the chain element, and the rest
         * of the PRP entries are built in the contiguous pcie buffer.
         */
        page_mask = nvme_pg_size - 1;

        /*
         * Native SGL is needed.
         * Put a chain element in main message frame that points to the first
         * chain buffer.
         *
         * NOTE:  The ChainOffset field must be 0 when using a chain pointer to
         *        a native SGL.
         */

        /* Set main message chain element pointer */
        main_chain_element = (pMpi25IeeeSgeChain64_t)&mpi_request->SGL;
        /*
         * For NVMe the chain element needs to be the 2nd SG entry in the main
         * message.
         */
        main_chain_element = (Mpi25IeeeSgeChain64_t *)
                ((u8 *)main_chain_element + sizeof(MPI25_IEEE_SGE_CHAIN64));

        /*
         * For the PRP entries, use the specially allocated buffer of
         * contiguous memory.  Normal chain buffers can't be used
         * because each chain buffer would need to be the size of an OS
         * page (4k).
         */
        curr_buff = mpt3sas_base_get_pcie_sgl(ioc, smid);
        msg_dma = mpt3sas_base_get_pcie_sgl_dma(ioc, smid);

        main_chain_element->Address = cpu_to_le64(msg_dma);
        main_chain_element->NextChainOffset = 0;
        main_chain_element->Flags = MPI2_IEEE_SGE_FLAGS_CHAIN_ELEMENT |
                        MPI2_IEEE_SGE_FLAGS_SYSTEM_ADDR |
                        MPI26_IEEE_SGE_FLAGS_NSF_NVME_PRP;

        /* Build first prp, sge need not to be page aligned*/
        ptr_first_sgl = (pMpi25IeeeSgeChain64_t)&mpi_request->SGL;
        sg_scmd = scsi_sglist(scmd);
        sge_addr = sg_dma_address(sg_scmd);
        sge_len = sg_dma_len(sg_scmd);

        offset = sge_addr & page_mask;
        first_prp_len = nvme_pg_size - offset;

        ptr_first_sgl->Address = cpu_to_le64(sge_addr);
        ptr_first_sgl->Length = cpu_to_le32(first_prp_len);

        data_len -= first_prp_len;

        if (sge_len > first_prp_len) {
                sge_addr += first_prp_len;
                sge_len -= first_prp_len;
        } else if (data_len && (sge_len == first_prp_len)) {
                sg_scmd = sg_next(sg_scmd);
                sge_addr = sg_dma_address(sg_scmd);
                sge_len = sg_dma_len(sg_scmd);
        }

        for (;;) {
                offset = sge_addr & page_mask;

                /* Put PRP pointer due to page boundary*/
                page_mask_result = (uintptr_t)(curr_buff + 1) & page_mask;
                if (unlikely(!page_mask_result)) {
                        scmd_printk(KERN_NOTICE,
                                scmd, "page boundary curr_buff: 0x%p\n",
                                curr_buff);
                        msg_dma += 8;
                        *curr_buff = cpu_to_le64(msg_dma);
                        curr_buff++;
                        num_prp_in_chain++;
                }

                *curr_buff = cpu_to_le64(sge_addr);
                curr_buff++;
                msg_dma += 8;
                num_prp_in_chain++;

                sge_addr += nvme_pg_size;
                sge_len -= nvme_pg_size;
                data_len -= nvme_pg_size;

                if (data_len <= 0)
                        break;

                if (sge_len > 0)
                        continue;

                sg_scmd = sg_next(sg_scmd);
                sge_addr = sg_dma_address(sg_scmd);
                sge_len = sg_dma_len(sg_scmd);
        }

        main_chain_element->Length =
                cpu_to_le32(num_prp_in_chain * sizeof(u64));
        return;
}

static bool
base_is_prp_possible(struct MPT3SAS_ADAPTER *ioc,
        struct _pcie_device *pcie_device, struct scsi_cmnd *scmd, int sge_count)
{
        u32 data_length = 0;
        bool build_prp = true;

        data_length = scsi_bufflen(scmd);
        if (pcie_device &&
            (mpt3sas_scsih_is_pcie_scsi_device(pcie_device->device_info))) {
                build_prp = false;
                return build_prp;
        }

        /* If Datalenth is <= 16K and number of SGE’s entries are <= 2
         * we built IEEE SGL
         */
        if ((data_length <= NVME_PRP_PAGE_SIZE*4) && (sge_count <= 2))
                build_prp = false;

        return build_prp;
}

/**
 * _base_check_pcie_native_sgl - This function is called for PCIe end devices to
 * determine if the driver needs to build a native SGL.  If so, that native
 * SGL is built in the special contiguous buffers allocated especially for
 * PCIe SGL creation.  If the driver will not build a native SGL, return
 * TRUE and a normal IEEE SGL will be built.  Currently this routine
 * supports NVMe.
 * @ioc: per adapter object
 * @mpi_request: mf request pointer
 * @smid: system request message index
 * @scmd: scsi command
 * @pcie_device: points to the PCIe device's info
 *
 * Return: 0 if native SGL was built, 1 if no SGL was built
 */
static int
_base_check_pcie_native_sgl(struct MPT3SAS_ADAPTER *ioc,
        Mpi25SCSIIORequest_t *mpi_request, u16 smid, struct scsi_cmnd *scmd,
        struct _pcie_device *pcie_device)
{
        int sges_left;

        /* Get the SG list pointer and info. */
        sges_left = scsi_dma_map(scmd);
        if (sges_left < 0) {
                sdev_printk(KERN_ERR, scmd->device,
                        "scsi_dma_map failed: request for %d bytes!\n",
                        scsi_bufflen(scmd));
                return 1;
        }

        /* Check if we need to build a native SG list. */
        if (base_is_prp_possible(ioc, pcie_device,
                                scmd, sges_left) == 0) {
                /* We built a native SG list, just return. */
                goto out;
        }

        /*
         * Build native NVMe PRP.
         */
        base_make_prp_nvme(ioc, scmd, mpi_request,
                        smid, sges_left);

        return 0;
out:
        scsi_dma_unmap(scmd);
        return 1;
}

/**
 * _base_add_sg_single_ieee - add sg element for IEEE format
 * @paddr: virtual address for SGE
 * @flags: SGE flags
 * @chain_offset: number of 128 byte elements from start of segment
 * @length: data transfer length
 * @dma_addr: Physical address
 */
static void
_base_add_sg_single_ieee(void *paddr, u8 flags, u8 chain_offset, u32 length,
        dma_addr_t dma_addr)
{
        Mpi25IeeeSgeChain64_t *sgel = paddr;

        sgel->Flags = flags;
        sgel->NextChainOffset = chain_offset;
        sgel->Length = cpu_to_le32(length);
        sgel->Address = cpu_to_le64(dma_addr);
}

/**
 * _base_build_zero_len_sge_ieee - build zero length sg entry for IEEE format
 * @ioc: per adapter object
 * @paddr: virtual address for SGE
 *
 * Create a zero length scatter gather entry to insure the IOCs hardware has
 * something to use if the target device goes brain dead and tries
 * to send data even when none is asked for.
 */
static void
_base_build_zero_len_sge_ieee(struct MPT3SAS_ADAPTER *ioc, void *paddr)
{
        u8 sgl_flags = (MPI2_IEEE_SGE_FLAGS_SIMPLE_ELEMENT |
                MPI2_IEEE_SGE_FLAGS_SYSTEM_ADDR |
                MPI25_IEEE_SGE_FLAGS_END_OF_LIST);

        _base_add_sg_single_ieee(paddr, sgl_flags, 0, 0, -1);
}

/**
 * _base_build_sg_scmd - main sg creation routine
 *              pcie_device is unused here!
 * @ioc: per adapter object
 * @scmd: scsi command
 * @smid: system request message index
 * @unused: unused pcie_device pointer
 * Context: none.
 *
 * The main routine that builds scatter gather table from a given
 * scsi request sent via the .queuecommand main handler.
 *
 * Return: 0 success, anything else error
 */
static int
_base_build_sg_scmd(struct MPT3SAS_ADAPTER *ioc,
        struct scsi_cmnd *scmd, u16 smid, struct _pcie_device *unused)
{
        Mpi2SCSIIORequest_t *mpi_request;
        dma_addr_t chain_dma;
        struct scatterlist *sg_scmd;
        void *sg_local, *chain;
        u32 chain_offset;
        u32 chain_length;
        u32 chain_flags;
        int sges_left;
        u32 sges_in_segment;
        u32 sgl_flags;
        u32 sgl_flags_last_element;
        u32 sgl_flags_end_buffer;
        struct chain_tracker *chain_req;

        mpi_request = mpt3sas_base_get_msg_frame(ioc, smid);

        /* init scatter gather flags */
        sgl_flags = MPI2_SGE_FLAGS_SIMPLE_ELEMENT;
        if (scmd->sc_data_direction == DMA_TO_DEVICE)
                sgl_flags |= MPI2_SGE_FLAGS_HOST_TO_IOC;
        sgl_flags_last_element = (sgl_flags | MPI2_SGE_FLAGS_LAST_ELEMENT)
            << MPI2_SGE_FLAGS_SHIFT;
        sgl_flags_end_buffer = (sgl_flags | MPI2_SGE_FLAGS_LAST_ELEMENT |
            MPI2_SGE_FLAGS_END_OF_BUFFER | MPI2_SGE_FLAGS_END_OF_LIST)
            << MPI2_SGE_FLAGS_SHIFT;
        sgl_flags = sgl_flags << MPI2_SGE_FLAGS_SHIFT;

        sg_scmd = scsi_sglist(scmd);
        sges_left = scsi_dma_map(scmd);
        if (sges_left < 0) {
                sdev_printk(KERN_ERR, scmd->device,
                 "scsi_dma_map failed: request for %d bytes!\n",
                 scsi_bufflen(scmd));
                return -ENOMEM;
        }

        sg_local = &mpi_request->SGL;
        sges_in_segment = ioc->max_sges_in_main_message;
        if (sges_left <= sges_in_segment)
                goto fill_in_last_segment;

        mpi_request->ChainOffset = (offsetof(Mpi2SCSIIORequest_t, SGL) +
            (sges_in_segment * ioc->sge_size))/4;

        /* fill in main message segment when there is a chain following */
        while (sges_in_segment) {
                if (sges_in_segment == 1)
                        ioc->base_add_sg_single(sg_local,
                            sgl_flags_last_element | sg_dma_len(sg_scmd),
                            sg_dma_address(sg_scmd));
                else
                        ioc->base_add_sg_single(sg_local, sgl_flags |
                            sg_dma_len(sg_scmd), sg_dma_address(sg_scmd));
                sg_scmd = sg_next(sg_scmd);
                sg_local += ioc->sge_size;
                sges_left--;
                sges_in_segment--;
        }

        /* initializing the chain flags and pointers */
        chain_flags = MPI2_SGE_FLAGS_CHAIN_ELEMENT << MPI2_SGE_FLAGS_SHIFT;
        chain_req = _base_get_chain_buffer_tracker(ioc, scmd);
        if (!chain_req)
                return -1;
        chain = chain_req->chain_buffer;
        chain_dma = chain_req->chain_buffer_dma;
        do {
                sges_in_segment = (sges_left <=
                    ioc->max_sges_in_chain_message) ? sges_left :
                    ioc->max_sges_in_chain_message;
                chain_offset = (sges_left == sges_in_segment) ?
                    0 : (sges_in_segment * ioc->sge_size)/4;
                chain_length = sges_in_segment * ioc->sge_size;
                if (chain_offset) {
                        chain_offset = chain_offset <<
                            MPI2_SGE_CHAIN_OFFSET_SHIFT;
                        chain_length += ioc->sge_size;
                }
                ioc->base_add_sg_single(sg_local, chain_flags | chain_offset |
                    chain_length, chain_dma);
                sg_local = chain;
                if (!chain_offset)
                        goto fill_in_last_segment;

                /* fill in chain segments */
                while (sges_in_segment) {
                        if (sges_in_segment == 1)
                                ioc->base_add_sg_single(sg_local,
                                    sgl_flags_last_element |
                                    sg_dma_len(sg_scmd),
                                    sg_dma_address(sg_scmd));
                        else
                                ioc->base_add_sg_single(sg_local, sgl_flags |
                                    sg_dma_len(sg_scmd),
                                    sg_dma_address(sg_scmd));
                        sg_scmd = sg_next(sg_scmd);
                        sg_local += ioc->sge_size;
                        sges_left--;
                        sges_in_segment--;
                }

                chain_req = _base_get_chain_buffer_tracker(ioc, scmd);
                if (!chain_req)
                        return -1;
                chain = chain_req->chain_buffer;
                chain_dma = chain_req->chain_buffer_dma;
        } while (1);


 fill_in_last_segment:

        /* fill the last segment */
        while (sges_left) {
                if (sges_left == 1)
                        ioc->base_add_sg_single(sg_local, sgl_flags_end_buffer |
                            sg_dma_len(sg_scmd), sg_dma_address(sg_scmd));
                else
                        ioc->base_add_sg_single(sg_local, sgl_flags |
                            sg_dma_len(sg_scmd), sg_dma_address(sg_scmd));
                sg_scmd = sg_next(sg_scmd);
                sg_local += ioc->sge_size;
                sges_left--;
        }

        return 0;
}

/**
 * _base_build_sg_scmd_ieee - main sg creation routine for IEEE format
 * @ioc: per adapter object
 * @scmd: scsi command
 * @smid: system request message index
 * @pcie_device: Pointer to pcie_device. If set, the pcie native sgl will be
 * constructed on need.
 * Context: none.
 *
 * The main routine that builds scatter gather table from a given
 * scsi request sent via the .queuecommand main handler.
 *
 * Return: 0 success, anything else error
 */
static int
_base_build_sg_scmd_ieee(struct MPT3SAS_ADAPTER *ioc,
        struct scsi_cmnd *scmd, u16 smid, struct _pcie_device *pcie_device)
{
        Mpi25SCSIIORequest_t *mpi_request;
        dma_addr_t chain_dma;
        struct scatterlist *sg_scmd;
        void *sg_local, *chain;
        u32 chain_offset;
        u32 chain_length;
        int sges_left;
        u32 sges_in_segment;
        u8 simple_sgl_flags;
        u8 simple_sgl_flags_last;
        u8 chain_sgl_flags;
        struct chain_tracker *chain_req;

        mpi_request = mpt3sas_base_get_msg_frame(ioc, smid);

        /* init scatter gather flags */
        simple_sgl_flags = MPI2_IEEE_SGE_FLAGS_SIMPLE_ELEMENT |
            MPI2_IEEE_SGE_FLAGS_SYSTEM_ADDR;
        simple_sgl_flags_last = simple_sgl_flags |
            MPI25_IEEE_SGE_FLAGS_END_OF_LIST;
        chain_sgl_flags = MPI2_IEEE_SGE_FLAGS_CHAIN_ELEMENT |
            MPI2_IEEE_SGE_FLAGS_SYSTEM_ADDR;

        /* Check if we need to build a native SG list. */
        if ((pcie_device) && (_base_check_pcie_native_sgl(ioc, mpi_request,
                        smid, scmd, pcie_device) == 0)) {
                /* We built a native SG list, just return. */
                return 0;
        }

        sg_scmd = scsi_sglist(scmd);
        sges_left = scsi_dma_map(scmd);
        if (sges_left < 0) {
                sdev_printk(KERN_ERR, scmd->device,
                        "scsi_dma_map failed: request for %d bytes!\n",
                        scsi_bufflen(scmd));
                return -ENOMEM;
        }

        sg_local = &mpi_request->SGL;
        sges_in_segment = (ioc->request_sz -
                   offsetof(Mpi25SCSIIORequest_t, SGL))/ioc->sge_size_ieee;
        if (sges_left <= sges_in_segment)
                goto fill_in_last_segment;

        mpi_request->ChainOffset = (sges_in_segment - 1 /* chain element */) +
            (offsetof(Mpi25SCSIIORequest_t, SGL)/ioc->sge_size_ieee);

        /* fill in main message segment when there is a chain following */
        while (sges_in_segment > 1) {
                _base_add_sg_single_ieee(sg_local, simple_sgl_flags, 0,
                    sg_dma_len(sg_scmd), sg_dma_address(sg_scmd));
                sg_scmd = sg_next(sg_scmd);
                sg_local += ioc->sge_size_ieee;
                sges_left--;
                sges_in_segment--;
        }

        /* initializing the pointers */
        chain_req = _base_get_chain_buffer_tracker(ioc, scmd);
        if (!chain_req)
                return -1;
        chain = chain_req->chain_buffer;
        chain_dma = chain_req->chain_buffer_dma;
        do {
                sges_in_segment = (sges_left <=
                    ioc->max_sges_in_chain_message) ? sges_left :
                    ioc->max_sges_in_chain_message;
                chain_offset = (sges_left == sges_in_segment) ?
                    0 : sges_in_segment;
                chain_length = sges_in_segment * ioc->sge_size_ieee;
                if (chain_offset)
                        chain_length += ioc->sge_size_ieee;
                _base_add_sg_single_ieee(sg_local, chain_sgl_flags,
                    chain_offset, chain_length, chain_dma);

                sg_local = chain;
                if (!chain_offset)
                        goto fill_in_last_segment;

                /* fill in chain segments */
                while (sges_in_segment) {
                        _base_add_sg_single_ieee(sg_local, simple_sgl_flags, 0,
                            sg_dma_len(sg_scmd), sg_dma_address(sg_scmd));
                        sg_scmd = sg_next(sg_scmd);
                        sg_local += ioc->sge_size_ieee;
                        sges_left--;
                        sges_in_segment--;
                }

                chain_req = _base_get_chain_buffer_tracker(ioc, scmd);
                if (!chain_req)
                        return -1;
                chain = chain_req->chain_buffer;
                chain_dma = chain_req->chain_buffer_dma;
        } while (1);


 fill_in_last_segment:

        /* fill the last segment */
        while (sges_left > 0) {
                if (sges_left == 1)
                        _base_add_sg_single_ieee(sg_local,
                            simple_sgl_flags_last, 0, sg_dma_len(sg_scmd),
                            sg_dma_address(sg_scmd));
                else
                        _base_add_sg_single_ieee(sg_local, simple_sgl_flags, 0,
                            sg_dma_len(sg_scmd), sg_dma_address(sg_scmd));
                sg_scmd = sg_next(sg_scmd);
                sg_local += ioc->sge_size_ieee;
                sges_left--;
        }

        return 0;
}

/**
 * _base_build_sg_ieee - build generic sg for IEEE format
 * @ioc: per adapter object
 * @psge: virtual address for SGE
 * @data_out_dma: physical address for WRITES
 * @data_out_sz: data xfer size for WRITES
 * @data_in_dma: physical address for READS
 * @data_in_sz: data xfer size for READS
 */
static void
_base_build_sg_ieee(struct MPT3SAS_ADAPTER *ioc, void *psge,
        dma_addr_t data_out_dma, size_t data_out_sz, dma_addr_t data_in_dma,
        size_t data_in_sz)
{
        u8 sgl_flags;

        if (!data_out_sz && !data_in_sz) {
                _base_build_zero_len_sge_ieee(ioc, psge);
                return;
        }

        if (data_out_sz && data_in_sz) {
                /* WRITE sgel first */
                sgl_flags = MPI2_IEEE_SGE_FLAGS_SIMPLE_ELEMENT |
                    MPI2_IEEE_SGE_FLAGS_SYSTEM_ADDR;
                _base_add_sg_single_ieee(psge, sgl_flags, 0, data_out_sz,
                    data_out_dma);

                /* incr sgel */
                psge += ioc->sge_size_ieee;

                /* READ sgel last */
                sgl_flags |= MPI25_IEEE_SGE_FLAGS_END_OF_LIST;
                _base_add_sg_single_ieee(psge, sgl_flags, 0, data_in_sz,
                    data_in_dma);
        } else if (data_out_sz) /* WRITE */ {
                sgl_flags = MPI2_IEEE_SGE_FLAGS_SIMPLE_ELEMENT |
                    MPI25_IEEE_SGE_FLAGS_END_OF_LIST |
                    MPI2_IEEE_SGE_FLAGS_SYSTEM_ADDR;
                _base_add_sg_single_ieee(psge, sgl_flags, 0, data_out_sz,
                    data_out_dma);
        } else if (data_in_sz) /* READ */ {
                sgl_flags = MPI2_IEEE_SGE_FLAGS_SIMPLE_ELEMENT |
                    MPI25_IEEE_SGE_FLAGS_END_OF_LIST |
                    MPI2_IEEE_SGE_FLAGS_SYSTEM_ADDR;
                _base_add_sg_single_ieee(psge, sgl_flags, 0, data_in_sz,
                    data_in_dma);
        }
}

#define convert_to_kb(x) ((x) << (PAGE_SHIFT - 10))

/**
 * _base_config_dma_addressing - set dma addressing
 * @ioc: per adapter object
 * @pdev: PCI device struct
 *
 * Return: 0 for success, non-zero for failure.
 */
static int
_base_config_dma_addressing(struct MPT3SAS_ADAPTER *ioc, struct pci_dev *pdev)
{
        struct sysinfo s;
        int dma_mask;

        if (ioc->is_mcpu_endpoint ||
            sizeof(dma_addr_t) == 4 || ioc->use_32bit_dma ||
            dma_get_required_mask(&pdev->dev) <= 32)
                dma_mask = 32;
        /* Set 63 bit DMA mask for all SAS3 and SAS35 controllers */
        else if (ioc->hba_mpi_version_belonged > MPI2_VERSION)
                dma_mask = 63;
        else
                dma_mask = 64;

        if (dma_set_mask(&pdev->dev, DMA_BIT_MASK(dma_mask)) ||
            dma_set_coherent_mask(&pdev->dev, DMA_BIT_MASK(dma_mask)))
                return -ENODEV;

        if (dma_mask > 32) {
                ioc->base_add_sg_single = &_base_add_sg_single_64;
                ioc->sge_size = sizeof(Mpi2SGESimple64_t);
        } else {
                ioc->base_add_sg_single = &_base_add_sg_single_32;
                ioc->sge_size = sizeof(Mpi2SGESimple32_t);
        }

        si_meminfo(&s);
        ioc_info(ioc, "%d BIT PCI BUS DMA ADDRESSING SUPPORTED, total mem (%ld kB)\n",
                dma_mask, convert_to_kb(s.totalram));

        return 0;
}

/**
 * _base_check_enable_msix - checks MSIX capabable.
 * @ioc: per adapter object
 *
 * Check to see if card is capable of MSIX, and set number
 * of available msix vectors
 */
static int
_base_check_enable_msix(struct MPT3SAS_ADAPTER *ioc)
{
        int base;
        u16 message_control;

        /* Check whether controller SAS2008 B0 controller,
         * if it is SAS2008 B0 controller use IO-APIC instead of MSIX
         */
        if (ioc->pdev->device == MPI2_MFGPAGE_DEVID_SAS2008 &&
            ioc->pdev->revision == SAS2_PCI_DEVICE_B0_REVISION) {
                return -EINVAL;
        }

        base = pci_find_capability(ioc->pdev, PCI_CAP_ID_MSIX);
        if (!base) {
                dfailprintk(ioc, ioc_info(ioc, "msix not supported\n"));
                return -EINVAL;
        }

        /* get msix vector count */
        /* NUMA_IO not supported for older controllers */
        if (ioc->pdev->device == MPI2_MFGPAGE_DEVID_SAS2004 ||
            ioc->pdev->device == MPI2_MFGPAGE_DEVID_SAS2008 ||
            ioc->pdev->device == MPI2_MFGPAGE_DEVID_SAS2108_1 ||
            ioc->pdev->device == MPI2_MFGPAGE_DEVID_SAS2108_2 ||
            ioc->pdev->device == MPI2_MFGPAGE_DEVID_SAS2108_3 ||
            ioc->pdev->device == MPI2_MFGPAGE_DEVID_SAS2116_1 ||
            ioc->pdev->device == MPI2_MFGPAGE_DEVID_SAS2116_2)
                ioc->msix_vector_count = 1;
        else {
                pci_read_config_word(ioc->pdev, base + 2, &message_control);
                ioc->msix_vector_count = (message_control & 0x3FF) + 1;
        }
        dinitprintk(ioc, ioc_info(ioc, "msix is supported, vector_count(%d)\n",
                                  ioc->msix_vector_count));
        return 0;
}

/**
 * _base_free_irq - free irq
 * @ioc: per adapter object
 *
 * Freeing respective reply_queue from the list.
 */
static void
_base_free_irq(struct MPT3SAS_ADAPTER *ioc)
{
        struct adapter_reply_queue *reply_q, *next;

        if (list_empty(&ioc->reply_queue_list))
                return;

        list_for_each_entry_safe(reply_q, next, &ioc->reply_queue_list, list) {
                list_del(&reply_q->list);
                if (ioc->smp_affinity_enable)
                        irq_set_affinity_hint(pci_irq_vector(ioc->pdev,
                            reply_q->msix_index), NULL);
                free_irq(pci_irq_vector(ioc->pdev, reply_q->msix_index),
                         reply_q);
                kfree(reply_q);
        }
}

/**
 * _base_request_irq - request irq
 * @ioc: per adapter object
 * @index: msix index into vector table
 *
 * Inserting respective reply_queue into the list.
 */
static int
_base_request_irq(struct MPT3SAS_ADAPTER *ioc, u8 index)
{
        struct pci_dev *pdev = ioc->pdev;
        struct adapter_reply_queue *reply_q;
        int r;

        reply_q =  kzalloc(sizeof(struct adapter_reply_queue), GFP_KERNEL);
        if (!reply_q) {
                ioc_err(ioc, "unable to allocate memory %zu!\n",
                        sizeof(struct adapter_reply_queue));
                return -ENOMEM;
        }
        reply_q->ioc = ioc;
        reply_q->msix_index = index;

        atomic_set(&reply_q->busy, 0);
        if (ioc->msix_enable)
                snprintf(reply_q->name, MPT_NAME_LENGTH, "%s%d-msix%d",
                    ioc->driver_name, ioc->id, index);
        else
                snprintf(reply_q->name, MPT_NAME_LENGTH, "%s%d",
                    ioc->driver_name, ioc->id);
        r = request_irq(pci_irq_vector(pdev, index), _base_interrupt,
                        IRQF_SHARED, reply_q->name, reply_q);
        if (r) {
                pr_err("%s: unable to allocate interrupt %d!\n",
                       reply_q->name, pci_irq_vector(pdev, index));
                kfree(reply_q);
                return -EBUSY;
        }

        INIT_LIST_HEAD(&reply_q->list);
        list_add_tail(&reply_q->list, &ioc->reply_queue_list);
        return 0;
}

/**
 * _base_assign_reply_queues - assigning msix index for each cpu
 * @ioc: per adapter object
 *
 * The enduser would need to set the affinity via /proc/irq/#/smp_affinity
 *
 * It would nice if we could call irq_set_affinity, however it is not
 * an exported symbol
 */
static void
_base_assign_reply_queues(struct MPT3SAS_ADAPTER *ioc)
{
        unsigned int cpu, nr_cpus, nr_msix, index = 0;
        struct adapter_reply_queue *reply_q;
        int local_numa_node;

        if (!_base_is_controller_msix_enabled(ioc))
                return;

        if (ioc->msix_load_balance)
                return;

        memset(ioc->cpu_msix_table, 0, ioc->cpu_msix_table_sz);

        nr_cpus = num_online_cpus();
        nr_msix = ioc->reply_queue_count = min(ioc->reply_queue_count,
                                               ioc->facts.MaxMSIxVectors);
        if (!nr_msix)
                return;

        if (ioc->smp_affinity_enable) {

                /*
                 * set irq affinity to local numa node for those irqs
                 * corresponding to high iops queues.
                 */
                if (ioc->high_iops_queues) {
                        local_numa_node = dev_to_node(&ioc->pdev->dev);
                        for (index = 0; index < ioc->high_iops_queues;
                            index++) {
                                irq_set_affinity_hint(pci_irq_vector(ioc->pdev,
                                    index), cpumask_of_node(local_numa_node));
                        }
                }

                list_for_each_entry(reply_q, &ioc->reply_queue_list, list) {
                        const cpumask_t *mask;

                        if (reply_q->msix_index < ioc->high_iops_queues)
                                continue;

                        mask = pci_irq_get_affinity(ioc->pdev,
                            reply_q->msix_index);
                        if (!mask) {
                                ioc_warn(ioc, "no affinity for msi %x\n",
                                         reply_q->msix_index);
                                goto fall_back;
                        }

                        for_each_cpu_and(cpu, mask, cpu_online_mask) {
                                if (cpu >= ioc->cpu_msix_table_sz)
                                        break;
                                ioc->cpu_msix_table[cpu] = reply_q->msix_index;
                        }
                }
                return;
        }

fall_back:
        cpu = cpumask_first(cpu_online_mask);
        nr_msix -= ioc->high_iops_queues;
        index = 0;

        list_for_each_entry(reply_q, &ioc->reply_queue_list, list) {
                unsigned int i, group = nr_cpus / nr_msix;

                if (reply_q->msix_index < ioc->high_iops_queues)
                        continue;

                if (cpu >= nr_cpus)
                        break;

                if (index < nr_cpus % nr_msix)
                        group++;

                for (i = 0 ; i < group ; i++) {
                        ioc->cpu_msix_table[cpu] = reply_q->msix_index;
                        cpu = cpumask_next(cpu, cpu_online_mask);
                }
                index++;
        }
}

/**
 * _base_check_and_enable_high_iops_queues - enable high iops mode
 * @ioc: per adapter object
 * @hba_msix_vector_count: msix vectors supported by HBA
 *
 * Enable high iops queues only if
 *  - HBA is a SEA/AERO controller and
 *  - MSI-Xs vector supported by the HBA is 128 and
 *  - total CPU count in the system >=16 and
 *  - loaded driver with default max_msix_vectors module parameter and
 *  - system booted in non kdump mode
 *
 * returns nothing.
 */
static void
_base_check_and_enable_high_iops_queues(struct MPT3SAS_ADAPTER *ioc,
                int hba_msix_vector_count)
{
        u16 lnksta, speed;

        if (perf_mode == MPT_PERF_MODE_IOPS ||
            perf_mode == MPT_PERF_MODE_LATENCY) {
                ioc->high_iops_queues = 0;
                return;
        }

        if (perf_mode == MPT_PERF_MODE_DEFAULT) {

                pcie_capability_read_word(ioc->pdev, PCI_EXP_LNKSTA, &lnksta);
                speed = lnksta & PCI_EXP_LNKSTA_CLS;

                if (speed < 0x4) {
                        ioc->high_iops_queues = 0;
                        return;
                }
        }

        if (!reset_devices && ioc->is_aero_ioc &&
            hba_msix_vector_count == MPT3SAS_GEN35_MAX_MSIX_QUEUES &&
            num_online_cpus() >= MPT3SAS_HIGH_IOPS_REPLY_QUEUES &&
            max_msix_vectors == -1)
                ioc->high_iops_queues = MPT3SAS_HIGH_IOPS_REPLY_QUEUES;
        else
                ioc->high_iops_queues = 0;
}

/**
 * _base_disable_msix - disables msix
 * @ioc: per adapter object
 *
 */
static void
_base_disable_msix(struct MPT3SAS_ADAPTER *ioc)
{
        if (!ioc->msix_enable)
                return;
        pci_free_irq_vectors(ioc->pdev);
        ioc->msix_enable = 0;
}

/**
 * _base_alloc_irq_vectors - allocate msix vectors
 * @ioc: per adapter object
 *
 */
static int
_base_alloc_irq_vectors(struct MPT3SAS_ADAPTER *ioc)
{
        int i, irq_flags = PCI_IRQ_MSIX;
        struct irq_affinity desc = { .pre_vectors = ioc->high_iops_queues };
        struct irq_affinity *descp = &desc;

        if (ioc->smp_affinity_enable)
                irq_flags |= PCI_IRQ_AFFINITY;
        else
                descp = NULL;

        ioc_info(ioc, " %d %d\n", ioc->high_iops_queues,
            ioc->reply_queue_count);

        i = pci_alloc_irq_vectors_affinity(ioc->pdev,
            ioc->high_iops_queues,
            ioc->reply_queue_count, irq_flags, descp);

        return i;
}

/**
 * _base_enable_msix - enables msix, failback to io_apic
 * @ioc: per adapter object
 *
 */
static int
_base_enable_msix(struct MPT3SAS_ADAPTER *ioc)
{
        int r;
        int i, local_max_msix_vectors;
        u8 try_msix = 0;

        ioc->msix_load_balance = false;

        if (msix_disable == -1 || msix_disable == 0)
                try_msix = 1;

        if (!try_msix)
                goto try_ioapic;

        if (_base_check_enable_msix(ioc) != 0)
                goto try_ioapic;

        ioc_info(ioc, "MSI-X vectors supported: %d\n", ioc->msix_vector_count);
        pr_info("\t no of cores: %d, max_msix_vectors: %d\n",
                ioc->cpu_count, max_msix_vectors);
        if (ioc->is_aero_ioc)
                _base_check_and_enable_high_iops_queues(ioc,
                        ioc->msix_vector_count);
        ioc->reply_queue_count =
                min_t(int, ioc->cpu_count + ioc->high_iops_queues,
                ioc->msix_vector_count);

        if (!ioc->rdpq_array_enable && max_msix_vectors == -1)
                local_max_msix_vectors = (reset_devices) ? 1 : 8;
        else
                local_max_msix_vectors = max_msix_vectors;

        if (local_max_msix_vectors > 0)
                ioc->reply_queue_count = min_t(int, local_max_msix_vectors,
                        ioc->reply_queue_count);
        else if (local_max_msix_vectors == 0)
                goto try_ioapic;

        /*
         * Enable msix_load_balance only if combined reply queue mode is
         * disabled on SAS3 & above generation HBA devices.
         */
        if (!ioc->combined_reply_queue &&
            ioc->hba_mpi_version_belonged != MPI2_VERSION) {
                ioc_info(ioc,
                    "combined ReplyQueue is off, Enabling msix load balance\n");
                ioc->msix_load_balance = true;
        }

        /*
         * smp affinity setting is not need when msix load balance
         * is enabled.
         */
        if (ioc->msix_load_balance)
                ioc->smp_affinity_enable = 0;

        r = _base_alloc_irq_vectors(ioc);
        if (r < 0) {
                ioc_info(ioc, "pci_alloc_irq_vectors failed (r=%d) !!!\n", r);
                goto try_ioapic;
        }

        ioc->msix_enable = 1;
        ioc->reply_queue_count = r;
        for (i = 0; i < ioc->reply_queue_count; i++) {
                r = _base_request_irq(ioc, i);
                if (r) {
                        _base_free_irq(ioc);
                        _base_disable_msix(ioc);
                        goto try_ioapic;
                }
        }

        ioc_info(ioc, "High IOPs queues : %s\n",
                        ioc->high_iops_queues ? "enabled" : "disabled");

        return 0;

/* failback to io_apic interrupt routing */
 try_ioapic:
        ioc->high_iops_queues = 0;
        ioc_info(ioc, "High IOPs queues : disabled\n");
        ioc->reply_queue_count = 1;
        r = pci_alloc_irq_vectors(ioc->pdev, 1, 1, PCI_IRQ_LEGACY);
        if (r < 0) {
                dfailprintk(ioc,
                            ioc_info(ioc, "pci_alloc_irq_vector(legacy) failed (r=%d) !!!\n",
                                     r));
        } else
                r = _base_request_irq(ioc, 0);

        return r;
}

/**
 * mpt3sas_base_unmap_resources - free controller resources
 * @ioc: per adapter object
 */
static void
mpt3sas_base_unmap_resources(struct MPT3SAS_ADAPTER *ioc)
{
        struct pci_dev *pdev = ioc->pdev;

        dexitprintk(ioc, ioc_info(ioc, "%s\n", __func__));

        _base_free_irq(ioc);
        _base_disable_msix(ioc);

        kfree(ioc->replyPostRegisterIndex);
        ioc->replyPostRegisterIndex = NULL;


        if (ioc->chip_phys) {
                iounmap(ioc->chip);
                ioc->chip_phys = 0;
        }

        if (pci_is_enabled(pdev)) {
                pci_release_selected_regions(ioc->pdev, ioc->bars);
                pci_disable_pcie_error_reporting(pdev);
                pci_disable_device(pdev);
        }
}

static int
_base_diag_reset(struct MPT3SAS_ADAPTER *ioc);

/**
 * _base_check_for_fault_and_issue_reset - check if IOC is in fault state
 *     and if it is in fault state then issue diag reset.
 * @ioc: per adapter object
 *
 * Returns: 0 for success, non-zero for failure.
 */
static int
_base_check_for_fault_and_issue_reset(struct MPT3SAS_ADAPTER *ioc)
{
        u32 ioc_state;
        int rc = -EFAULT;

        dinitprintk(ioc, pr_info("%s\n", __func__));
        if (ioc->pci_error_recovery)
                return 0;
        ioc_state = mpt3sas_base_get_iocstate(ioc, 0);
        dhsprintk(ioc, pr_info("%s: ioc_state(0x%08x)\n", __func__, ioc_state));

        if ((ioc_state & MPI2_IOC_STATE_MASK) == MPI2_IOC_STATE_FAULT) {
                mpt3sas_print_fault_code(ioc, ioc_state &
                    MPI2_DOORBELL_DATA_MASK);
                rc = _base_diag_reset(ioc);
        } else if ((ioc_state & MPI2_IOC_STATE_MASK) ==
            MPI2_IOC_STATE_COREDUMP) {
                mpt3sas_print_coredump_info(ioc, ioc_state &
                     MPI2_DOORBELL_DATA_MASK);
                mpt3sas_base_wait_for_coredump_completion(ioc, __func__);
                rc = _base_diag_reset(ioc);
        }

        return rc;
}

/**
 * mpt3sas_base_map_resources - map in controller resources (io/irq/memap)
 * @ioc: per adapter object
 *
 * Return: 0 for success, non-zero for failure.
 */
int
mpt3sas_base_map_resources(struct MPT3SAS_ADAPTER *ioc)
{
        struct pci_dev *pdev = ioc->pdev;
        u32 memap_sz;
        u32 pio_sz;
        int i, r = 0, rc;
        u64 pio_chip = 0;
        phys_addr_t chip_phys = 0;
        struct adapter_reply_queue *reply_q;

        dinitprintk(ioc, ioc_info(ioc, "%s\n", __func__));

        ioc->bars = pci_select_bars(pdev, IORESOURCE_MEM);
        if (pci_enable_device_mem(pdev)) {
                ioc_warn(ioc, "pci_enable_device_mem: failed\n");
                ioc->bars = 0;
                return -ENODEV;
        }


        if (pci_request_selected_regions(pdev, ioc->bars,
            ioc->driver_name)) {
                ioc_warn(ioc, "pci_request_selected_regions: failed\n");
                ioc->bars = 0;
                r = -ENODEV;
                goto out_fail;
        }

/* AER (Advanced Error Reporting) hooks */
        pci_enable_pcie_error_reporting(pdev);

        pci_set_master(pdev);


        if (_base_config_dma_addressing(ioc, pdev) != 0) {
                ioc_warn(ioc, "no suitable DMA mask for %s\n", pci_name(pdev));
                r = -ENODEV;
                goto out_fail;
        }

        for (i = 0, memap_sz = 0, pio_sz = 0; (i < DEVICE_COUNT_RESOURCE) &&
             (!memap_sz || !pio_sz); i++) {
                if (pci_resource_flags(pdev, i) & IORESOURCE_IO) {
                        if (pio_sz)
                                continue;
                        pio_chip = (u64)pci_resource_start(pdev, i);
                        pio_sz = pci_resource_len(pdev, i);
                } else if (pci_resource_flags(pdev, i) & IORESOURCE_MEM) {
                        if (memap_sz)
                                continue;
                        ioc->chip_phys = pci_resource_start(pdev, i);
                        chip_phys = ioc->chip_phys;
                        memap_sz = pci_resource_len(pdev, i);
                        ioc->chip = ioremap(ioc->chip_phys, memap_sz);
                }
        }

        if (ioc->chip == NULL) {
                ioc_err(ioc,
                    "unable to map adapter memory! or resource not found\n");
                r = -EINVAL;
                goto out_fail;
        }

        mpt3sas_base_mask_interrupts(ioc);

        r = _base_get_ioc_facts(ioc);
        if (r) {
                rc = _base_check_for_fault_and_issue_reset(ioc);
                if (rc || (_base_get_ioc_facts(ioc)))
                        goto out_fail;
        }

        if (!ioc->rdpq_array_enable_assigned) {
                ioc->rdpq_array_enable = ioc->rdpq_array_capable;
                ioc->rdpq_array_enable_assigned = 1;
        }

        r = _base_enable_msix(ioc);
        if (r)
                goto out_fail;

        if (!ioc->is_driver_loading)
                _base_init_irqpolls(ioc);
        /* Use the Combined reply queue feature only for SAS3 C0 & higher
         * revision HBAs and also only when reply queue count is greater than 8
         */
        if (ioc->combined_reply_queue) {
                /* Determine the Supplemental Reply Post Host Index Registers
                 * Addresse. Supplemental Reply Post Host Index Registers
                 * starts at offset MPI25_SUP_REPLY_POST_HOST_INDEX_OFFSET and
                 * each register is at offset bytes of
                 * MPT3_SUP_REPLY_POST_HOST_INDEX_REG_OFFSET from previous one.
                 */
                ioc->replyPostRegisterIndex = kcalloc(
                     ioc->combined_reply_index_count,
                     sizeof(resource_size_t *), GFP_KERNEL);
                if (!ioc->replyPostRegisterIndex) {
                        ioc_err(ioc,
                            "allocation for replyPostRegisterIndex failed!\n");
                        r = -ENOMEM;
                        goto out_fail;
                }

                for (i = 0; i < ioc->combined_reply_index_count; i++) {
                        ioc->replyPostRegisterIndex[i] = (resource_size_t *)
                             ((u8 __force *)&ioc->chip->Doorbell +
                             MPI25_SUP_REPLY_POST_HOST_INDEX_OFFSET +
                             (i * MPT3_SUP_REPLY_POST_HOST_INDEX_REG_OFFSET));
                }
        }

        if (ioc->is_warpdrive) {
                ioc->reply_post_host_index[0] = (resource_size_t __iomem *)
                    &ioc->chip->ReplyPostHostIndex;

                for (i = 1; i < ioc->cpu_msix_table_sz; i++)
                        ioc->reply_post_host_index[i] =
                        (resource_size_t __iomem *)
                        ((u8 __iomem *)&ioc->chip->Doorbell + (0x4000 + ((i - 1)
                        * 4)));
        }

        list_for_each_entry(reply_q, &ioc->reply_queue_list, list)
                pr_info("%s: %s enabled: IRQ %d\n",
                        reply_q->name,
                        ioc->msix_enable ? "PCI-MSI-X" : "IO-APIC",
                        pci_irq_vector(ioc->pdev, reply_q->msix_index));

        ioc_info(ioc, "iomem(%pap), mapped(0x%p), size(%d)\n",
                 &chip_phys, ioc->chip, memap_sz);
        ioc_info(ioc, "ioport(0x%016llx), size(%d)\n",
                 (unsigned long long)pio_chip, pio_sz);

        /* Save PCI configuration state for recovery from PCI AER/EEH errors */
        pci_save_state(pdev);
        return 0;

 out_fail:
        mpt3sas_base_unmap_resources(ioc);
        return r;
}

/**
 * mpt3sas_base_get_msg_frame - obtain request mf pointer
 * @ioc: per adapter object
 * @smid: system request message index(smid zero is invalid)
 *
 * Return: virt pointer to message frame.
 */
void *
mpt3sas_base_get_msg_frame(struct MPT3SAS_ADAPTER *ioc, u16 smid)
{
        return (void *)(ioc->request + (smid * ioc->request_sz));
}

/**
 * mpt3sas_base_get_sense_buffer - obtain a sense buffer virt addr
 * @ioc: per adapter object
 * @smid: system request message index
 *
 * Return: virt pointer to sense buffer.
 */
void *
mpt3sas_base_get_sense_buffer(struct MPT3SAS_ADAPTER *ioc, u16 smid)
{
        return (void *)(ioc->sense + ((smid - 1) * SCSI_SENSE_BUFFERSIZE));
}

/**
 * mpt3sas_base_get_sense_buffer_dma - obtain a sense buffer dma addr
 * @ioc: per adapter object
 * @smid: system request message index
 *
 * Return: phys pointer to the low 32bit address of the sense buffer.
 */
__le32
mpt3sas_base_get_sense_buffer_dma(struct MPT3SAS_ADAPTER *ioc, u16 smid)
{
        return cpu_to_le32(ioc->sense_dma + ((smid - 1) *
            SCSI_SENSE_BUFFERSIZE));
}

/**
 * mpt3sas_base_get_pcie_sgl - obtain a PCIe SGL virt addr
 * @ioc: per adapter object
 * @smid: system request message index
 *
 * Return: virt pointer to a PCIe SGL.
 */
void *
mpt3sas_base_get_pcie_sgl(struct MPT3SAS_ADAPTER *ioc, u16 smid)
{
        return (void *)(ioc->pcie_sg_lookup[smid - 1].pcie_sgl);
}

/**
 * mpt3sas_base_get_pcie_sgl_dma - obtain a PCIe SGL dma addr
 * @ioc: per adapter object
 * @smid: system request message index
 *
 * Return: phys pointer to the address of the PCIe buffer.
 */
dma_addr_t
mpt3sas_base_get_pcie_sgl_dma(struct MPT3SAS_ADAPTER *ioc, u16 smid)
{
        return ioc->pcie_sg_lookup[smid - 1].pcie_sgl_dma;
}

/**
 * mpt3sas_base_get_reply_virt_addr - obtain reply frames virt address
 * @ioc: per adapter object
 * @phys_addr: lower 32 physical addr of the reply
 *
 * Converts 32bit lower physical addr into a virt address.
 */
void *
mpt3sas_base_get_reply_virt_addr(struct MPT3SAS_ADAPTER *ioc, u32 phys_addr)
{
        if (!phys_addr)
                return NULL;
        return ioc->reply + (phys_addr - (u32)ioc->reply_dma);
}

/**
 * _base_get_msix_index - get the msix index
 * @ioc: per adapter object
 * @scmd: scsi_cmnd object
 *
 * returns msix index of general reply queues,
 * i.e. reply queue on which IO request's reply
 * should be posted by the HBA firmware.
 */
static inline u8
_base_get_msix_index(struct MPT3SAS_ADAPTER *ioc,
        struct scsi_cmnd *scmd)
{
        /* Enables reply_queue load balancing */
        if (ioc->msix_load_balance)
                return ioc->reply_queue_count ?
                    base_mod64(atomic64_add_return(1,
                    &ioc->total_io_cnt), ioc->reply_queue_count) : 0;

        return ioc->cpu_msix_table[raw_smp_processor_id()];
}

/**
 * _base_sdev_nr_inflight_request -get number of inflight requests
 *                                 of a request queue.
 * @q: request_queue object
 *
 * returns number of inflight request of a request queue.
 */
inline unsigned long
_base_sdev_nr_inflight_request(struct request_queue *q)
{
        struct blk_mq_hw_ctx *hctx = q->queue_hw_ctx[0];

        return atomic_read(&hctx->nr_active);
}


/**
 * _base_get_high_iops_msix_index - get the msix index of
 *                              high iops queues
 * @ioc: per adapter object
 * @scmd: scsi_cmnd object
 *
 * Returns: msix index of high iops reply queues.
 * i.e. high iops reply queue on which IO request's
 * reply should be posted by the HBA firmware.
 */
static inline u8
_base_get_high_iops_msix_index(struct MPT3SAS_ADAPTER *ioc,
        struct scsi_cmnd *scmd)
{
        /**
         * Round robin the IO interrupts among the high iops
         * reply queues in terms of batch count 16 when outstanding
         * IOs on the target device is >=8.
         */
        if (_base_sdev_nr_inflight_request(scmd->device->request_queue) >
            MPT3SAS_DEVICE_HIGH_IOPS_DEPTH)
                return base_mod64((
                    atomic64_add_return(1, &ioc->high_iops_outstanding) /
                    MPT3SAS_HIGH_IOPS_BATCH_COUNT),
                    MPT3SAS_HIGH_IOPS_REPLY_QUEUES);

        return _base_get_msix_index(ioc, scmd);
}

/**
 * mpt3sas_base_get_smid - obtain a free smid from internal queue
 * @ioc: per adapter object
 * @cb_idx: callback index
 *
 * Return: smid (zero is invalid)
 */
u16
mpt3sas_base_get_smid(struct MPT3SAS_ADAPTER *ioc, u8 cb_idx)
{
        unsigned long flags;
        struct request_tracker *request;
        u16 smid;

        spin_lock_irqsave(&ioc->scsi_lookup_lock, flags);
        if (list_empty(&ioc->internal_free_list)) {
                spin_unlock_irqrestore(&ioc->scsi_lookup_lock, flags);
                ioc_err(ioc, "%s: smid not available\n", __func__);
                return 0;
        }

        request = list_entry(ioc->internal_free_list.next,
            struct request_tracker, tracker_list);
        request->cb_idx = cb_idx;
        smid = request->smid;
        list_del(&request->tracker_list);
        spin_unlock_irqrestore(&ioc->scsi_lookup_lock, flags);
        return smid;
}

/**
 * mpt3sas_base_get_smid_scsiio - obtain a free smid from scsiio queue
 * @ioc: per adapter object
 * @cb_idx: callback index
 * @scmd: pointer to scsi command object
 *
 * Return: smid (zero is invalid)
 */
u16
mpt3sas_base_get_smid_scsiio(struct MPT3SAS_ADAPTER *ioc, u8 cb_idx,
        struct scsi_cmnd *scmd)
{
        struct scsiio_tracker *request = scsi_cmd_priv(scmd);
        unsigned int tag = scmd->request->tag;
        u16 smid;

        smid = tag + 1;
        request->cb_idx = cb_idx;
        request->smid = smid;
        request->scmd = scmd;
        INIT_LIST_HEAD(&request->chain_list);
        return smid;
}

/**
 * mpt3sas_base_get_smid_hpr - obtain a free smid from hi-priority queue
 * @ioc: per adapter object
 * @cb_idx: callback index
 *
 * Return: smid (zero is invalid)
 */
u16
mpt3sas_base_get_smid_hpr(struct MPT3SAS_ADAPTER *ioc, u8 cb_idx)
{
        unsigned long flags;
        struct request_tracker *request;
        u16 smid;

        spin_lock_irqsave(&ioc->scsi_lookup_lock, flags);
        if (list_empty(&ioc->hpr_free_list)) {
                spin_unlock_irqrestore(&ioc->scsi_lookup_lock, flags);
                return 0;
        }

        request = list_entry(ioc->hpr_free_list.next,
            struct request_tracker, tracker_list);
        request->cb_idx = cb_idx;
        smid = request->smid;
        list_del(&request->tracker_list);
        spin_unlock_irqrestore(&ioc->scsi_lookup_lock, flags);
        return smid;
}

static void
_base_recovery_check(struct MPT3SAS_ADAPTER *ioc)
{
        /*
         * See _wait_for_commands_to_complete() call with regards to this code.
         */
        if (ioc->shost_recovery && ioc->pending_io_count) {
                ioc->pending_io_count = scsi_host_busy(ioc->shost);
                if (ioc->pending_io_count == 0)
                        wake_up(&ioc->reset_wq);
        }
}

void mpt3sas_base_clear_st(struct MPT3SAS_ADAPTER *ioc,
                           struct scsiio_tracker *st)
{
        if (WARN_ON(st->smid == 0))
                return;
        st->cb_idx = 0xFF;
        st->direct_io = 0;
        st->scmd = NULL;
        atomic_set(&ioc->chain_lookup[st->smid - 1].chain_offset, 0);
        st->smid = 0;
}

/**
 * mpt3sas_base_free_smid - put smid back on free_list
 * @ioc: per adapter object
 * @smid: system request message index
 */
void
mpt3sas_base_free_smid(struct MPT3SAS_ADAPTER *ioc, u16 smid)
{
        unsigned long flags;
        int i;

        if (smid < ioc->hi_priority_smid) {
                struct scsiio_tracker *st;
                void *request;

                st = _get_st_from_smid(ioc, smid);
                if (!st) {
                        _base_recovery_check(ioc);
                        return;
                }

                /* Clear MPI request frame */
                request = mpt3sas_base_get_msg_frame(ioc, smid);
                memset(request, 0, ioc->request_sz);

                mpt3sas_base_clear_st(ioc, st);
                _base_recovery_check(ioc);
                return;
        }

        spin_lock_irqsave(&ioc->scsi_lookup_lock, flags);
        if (smid < ioc->internal_smid) {
                /* hi-priority */
                i = smid - ioc->hi_priority_smid;
                ioc->hpr_lookup[i].cb_idx = 0xFF;
                list_add(&ioc->hpr_lookup[i].tracker_list, &ioc->hpr_free_list);
        } else if (smid <= ioc->hba_queue_depth) {
                /* internal queue */
                i = smid - ioc->internal_smid;
                ioc->internal_lookup[i].cb_idx = 0xFF;
                list_add(&ioc->internal_lookup[i].tracker_list,
                    &ioc->internal_free_list);
        }
        spin_unlock_irqrestore(&ioc->scsi_lookup_lock, flags);
}

/**
 * _base_mpi_ep_writeq - 32 bit write to MMIO
 * @b: data payload
 * @addr: address in MMIO space
 * @writeq_lock: spin lock
 *
 * This special handling for MPI EP to take care of 32 bit
 * environment where its not quarenteed to send the entire word
 * in one transfer.
 */
static inline void
_base_mpi_ep_writeq(__u64 b, volatile void __iomem *addr,
                                        spinlock_t *writeq_lock)
{
        unsigned long flags;

        spin_lock_irqsave(writeq_lock, flags);
        __raw_writel((u32)(b), addr);
        __raw_writel((u32)(b >> 32), (addr + 4));
        spin_unlock_irqrestore(writeq_lock, flags);
}

/**
 * _base_writeq - 64 bit write to MMIO
 * @b: data payload
 * @addr: address in MMIO space
 * @writeq_lock: spin lock
 *
 * Glue for handling an atomic 64 bit word to MMIO. This special handling takes
 * care of 32 bit environment where its not quarenteed to send the entire word
 * in one transfer.
 */
#if defined(writeq) && defined(CONFIG_64BIT)
static inline void
_base_writeq(__u64 b, volatile void __iomem *addr, spinlock_t *writeq_lock)
{
        wmb();
        __raw_writeq(b, addr);
        barrier();
}
#else
static inline void
_base_writeq(__u64 b, volatile void __iomem *addr, spinlock_t *writeq_lock)
{
        _base_mpi_ep_writeq(b, addr, writeq_lock);
}
#endif

/**
 * _base_set_and_get_msix_index - get the msix index and assign to msix_io
 *                                variable of scsi tracker
 * @ioc: per adapter object
 * @smid: system request message index
 *
 * returns msix index.
 */
static u8
_base_set_and_get_msix_index(struct MPT3SAS_ADAPTER *ioc, u16 smid)
{
        struct scsiio_tracker *st = NULL;

        if (smid < ioc->hi_priority_smid)
                st = _get_st_from_smid(ioc, smid);

        if (st == NULL)
                return  _base_get_msix_index(ioc, NULL);

        st->msix_io = ioc->get_msix_index_for_smlio(ioc, st->scmd);
        return st->msix_io;
}

/**
 * _base_put_smid_mpi_ep_scsi_io - send SCSI_IO request to firmware
 * @ioc: per adapter object
 * @smid: system request message index
 * @handle: device handle
 */
static void
_base_put_smid_mpi_ep_scsi_io(struct MPT3SAS_ADAPTER *ioc,
        u16 smid, u16 handle)
{
        Mpi2RequestDescriptorUnion_t descriptor;
        u64 *request = (u64 *)&descriptor;
        void *mpi_req_iomem;
        __le32 *mfp = (__le32 *)mpt3sas_base_get_msg_frame(ioc, smid);

        _clone_sg_entries(ioc, (void *) mfp, smid);
        mpi_req_iomem = (void __force *)ioc->chip +
                        MPI_FRAME_START_OFFSET + (smid * ioc->request_sz);
        _base_clone_mpi_to_sys_mem(mpi_req_iomem, (void *)mfp,
                                        ioc->request_sz);
        descriptor.SCSIIO.RequestFlags = MPI2_REQ_DESCRIPT_FLAGS_SCSI_IO;
        descriptor.SCSIIO.MSIxIndex = _base_set_and_get_msix_index(ioc, smid);
        descriptor.SCSIIO.SMID = cpu_to_le16(smid);
        descriptor.SCSIIO.DevHandle = cpu_to_le16(handle);
        descriptor.SCSIIO.LMID = 0;
        _base_mpi_ep_writeq(*request, &ioc->chip->RequestDescriptorPostLow,
            &ioc->scsi_lookup_lock);
}

/**
 * _base_put_smid_scsi_io - send SCSI_IO request to firmware
 * @ioc: per adapter object
 * @smid: system request message index
 * @handle: device handle
 */
static void
_base_put_smid_scsi_io(struct MPT3SAS_ADAPTER *ioc, u16 smid, u16 handle)
{
        Mpi2RequestDescriptorUnion_t descriptor;
        u64 *request = (u64 *)&descriptor;


        descriptor.SCSIIO.RequestFlags = MPI2_REQ_DESCRIPT_FLAGS_SCSI_IO;
        descriptor.SCSIIO.MSIxIndex = _base_set_and_get_msix_index(ioc, smid);
        descriptor.SCSIIO.SMID = cpu_to_le16(smid);
        descriptor.SCSIIO.DevHandle = cpu_to_le16(handle);
        descriptor.SCSIIO.LMID = 0;
        _base_writeq(*request, &ioc->chip->RequestDescriptorPostLow,
            &ioc->scsi_lookup_lock);
}

/**
 * _base_put_smid_fast_path - send fast path request to firmware
 * @ioc: per adapter object
 * @smid: system request message index
 * @handle: device handle
 */
static void
_base_put_smid_fast_path(struct MPT3SAS_ADAPTER *ioc, u16 smid,
        u16 handle)
{
        Mpi2RequestDescriptorUnion_t descriptor;
        u64 *request = (u64 *)&descriptor;

        descriptor.SCSIIO.RequestFlags =
            MPI25_REQ_DESCRIPT_FLAGS_FAST_PATH_SCSI_IO;
        descriptor.SCSIIO.MSIxIndex = _base_set_and_get_msix_index(ioc, smid);
        descriptor.SCSIIO.SMID = cpu_to_le16(smid);
        descriptor.SCSIIO.DevHandle = cpu_to_le16(handle);
        descriptor.SCSIIO.LMID = 0;
        _base_writeq(*request, &ioc->chip->RequestDescriptorPostLow,
            &ioc->scsi_lookup_lock);
}

/**
 * _base_put_smid_hi_priority - send Task Management request to firmware
 * @ioc: per adapter object
 * @smid: system request message index
 * @msix_task: msix_task will be same as msix of IO incase of task abort else 0.
 */
static void
_base_put_smid_hi_priority(struct MPT3SAS_ADAPTER *ioc, u16 smid,
        u16 msix_task)
{
        Mpi2RequestDescriptorUnion_t descriptor;
        void *mpi_req_iomem;
        u64 *request;

        if (ioc->is_mcpu_endpoint) {
                __le32 *mfp = (__le32 *)mpt3sas_base_get_msg_frame(ioc, smid);

                /* TBD 256 is offset within sys register. */
                mpi_req_iomem = (void __force *)ioc->chip
                                        + MPI_FRAME_START_OFFSET
                                        + (smid * ioc->request_sz);
                _base_clone_mpi_to_sys_mem(mpi_req_iomem, (void *)mfp,
                                                        ioc->request_sz);
        }

        request = (u64 *)&descriptor;

        descriptor.HighPriority.RequestFlags =
            MPI2_REQ_DESCRIPT_FLAGS_HIGH_PRIORITY;
        descriptor.HighPriority.MSIxIndex =  msix_task;
        descriptor.HighPriority.SMID = cpu_to_le16(smid);
        descriptor.HighPriority.LMID = 0;
        descriptor.HighPriority.Reserved1 = 0;
        if (ioc->is_mcpu_endpoint)
                _base_mpi_ep_writeq(*request,
                                &ioc->chip->RequestDescriptorPostLow,
                                &ioc->scsi_lookup_lock);
        else
                _base_writeq(*request, &ioc->chip->RequestDescriptorPostLow,
                    &ioc->scsi_lookup_lock);
}

/**
 * mpt3sas_base_put_smid_nvme_encap - send NVMe encapsulated request to
 *  firmware
 * @ioc: per adapter object
 * @smid: system request message index
 */
void
mpt3sas_base_put_smid_nvme_encap(struct MPT3SAS_ADAPTER *ioc, u16 smid)
{
        Mpi2RequestDescriptorUnion_t descriptor;
        u64 *request = (u64 *)&descriptor;

        descriptor.Default.RequestFlags =
                MPI26_REQ_DESCRIPT_FLAGS_PCIE_ENCAPSULATED;
        descriptor.Default.MSIxIndex =  _base_set_and_get_msix_index(ioc, smid);
        descriptor.Default.SMID = cpu_to_le16(smid);
        descriptor.Default.LMID = 0;
        descriptor.Default.DescriptorTypeDependent = 0;
        _base_writeq(*request, &ioc->chip->RequestDescriptorPostLow,
            &ioc->scsi_lookup_lock);
}

/**
 * _base_put_smid_default - Default, primarily used for config pages
 * @ioc: per adapter object
 * @smid: system request message index
 */
static void
_base_put_smid_default(struct MPT3SAS_ADAPTER *ioc, u16 smid)
{
        Mpi2RequestDescriptorUnion_t descriptor;
        void *mpi_req_iomem;
        u64 *request;

        if (ioc->is_mcpu_endpoint) {
                __le32 *mfp = (__le32 *)mpt3sas_base_get_msg_frame(ioc, smid);

                _clone_sg_entries(ioc, (void *) mfp, smid);
                /* TBD 256 is offset within sys register */
                mpi_req_iomem = (void __force *)ioc->chip +
                        MPI_FRAME_START_OFFSET + (smid * ioc->request_sz);
                _base_clone_mpi_to_sys_mem(mpi_req_iomem, (void *)mfp,
                                                        ioc->request_sz);
        }
        request = (u64 *)&descriptor;