From: Anthony Liguori <anthony@codemonkey.ws>
To: "Michael S. Tsirkin" <mst@redhat.com>
Cc: Gerhard Wiesinger <lists@wiesinger.com>,
qemu-devel@nongnu.org, Hannes Reinecke <hare@suse.de>,
Andreas Faerber <afaerber@suse.de>,
Alexander Graf <agraf@suse.de>
Subject: Re: [Qemu-devel] [PATCH][v14] megasas: LSI Megaraid SAS HBA emulation
Date: Fri, 24 Feb 2012 09:58:11 -0600 [thread overview]
Message-ID: <4F47B393.7070606@codemonkey.ws> (raw)
In-Reply-To: <20120223153419.GA20732@redhat.com>
On 02/23/2012 09:34 AM, Michael S. Tsirkin wrote:
> On Tue, Feb 21, 2012 at 10:36:43AM +0100, Hannes Reinecke wrote:
>> This patch adds an emulation for the LSI Megaraid SAS 8708EM2 HBA.
>> I've tested it to work with Linux, Windows Vista, and Windows7.
>> MSI-X support is currently broken; have to investigate.
>>
>> Changes since v13:
>> - Remove separate MSI-X BAR
>> - Simplify BAR allocation
>>
>> Changes since v12:
>> - Fixup flag setting via properties
>> - Fixup MSI-X handling
>> - Disable MSI-X per default
>>
>> Changes since v11:
>> - Remove unneeded variables
>>
>> Changes since v10:
>> - Port to new device type API
>> - Include suggestion from Alex Graf:
>> - Remove 'inline' function declaration
>> - Queue setup and interrupt enablement needs to be treated
>> independently
>> - Always read in 64 bit context and just mask out the top
>> bits if required
>>
>> Changes since v9:
>> - Split off trace events into a separate patch
>> - Do not check for max_luns in PD Info
>> - Update trace events
>> - Clarify license statement
>> - Fixup coding style issues
>>
>> Changes since v8:
>> - Remove 'disable' keyword from trace definitions
>> - Convert hand-crafted debugging statements with trace
>> definitions
>> - Treat 'context' tag as little endian
>>
>> Changes since v7:
>> - Port to new memory API
>> - Port to new PCI infrastructure
>> - Use fixed buffers for sense processing
>> - Update to updated SCSI infrastructure
>>
>> Changes since v6:
>> - Preliminary patches pushed to Kevins block tree
>> - Implement 64bit contexts, required for Windows7
>> - Use iovecs for DCMD processing
>> - Add MSI-X support
>> Latest Linux driver now happily uses MSI-X.
>> - Static iovec allocation
>> We have a fixed upper number of iovecs, so we can
>> save us the allocation. Suggested by Alex Graf.
>> - Update MFI header
>> Latest Linux driver has some more definitions,
>> add them
>> - Fixup AEN handling
>> - Update tracing details
>> - Remove sdev pointer from megasas_cmd_t
>>
>> Changes since v5:
>> - megasas: Use tracing infrastructure instead of DPRINTF
>> - megasas: Use new PCI infrastructure
>> - megasas: Check for iovec mapping failure
>> cpu_map_physical_memory() might fail, so we need to check for
>> it when mapping iovecs.
>> - megasas: Trace scsi buffer overflow
>> The transfer length as specified in the SCSI command might
>> disagree with the length of the iovec. We should be tracing
>> these issues.
>> - megasas: Reset frames after init firmware
>> When receiving an INIT FIRMWARE command we need reset all
>> frames, otherwise some frames might point to invalid memory.
>>
>> Chances since v4:
>> - megasas: checkpatch.pl fixes and update to work with the
>> changed interface in scsi_req_new(). Also included the
>> suggested fixes from Alex.
>>
>> Signed-off-by: Hannes Reinecke<hare@suse.de>
>> Cc: Alexander Graf<agraf@suse.de>
>> Cc: Andreas Faerber<afaerber@suse.de>
>> Cc: Gerhard Wiesinger<lists@wiesinger.com>
>
> So Alex asked whether I can merge this, which made me
> take a look. I don't know much about what this does
> so just general comments on all of the code.
This should come through Paolo's new SCSI tree.
And while I defended this before, I now think it really ought to be split after
trying (and failing) to trim the bits below in Thunderbird.
Regards,
Anthony Liguori
> Also, some issues related to msix - want to rip that code
> out for now since it does not work anyway?
>
> qemu has two styles for struct and enum types:
> 1. documented: typedef struct CamelCase CamelCase;
> 2. undocumented but still widely used: struct lower_case; (no typedef)
> *_t type typedef is used for numeric types such as target_phys_addr_t.
>
> This code mixes them in arbitrary ways. Pls don't do that,
> pls be consistent.
>
>
>
>
>> ---
>> Makefile.objs | 1 +
>> default-configs/pci.mak | 1 +
>> hw/megasas.c | 1865 +++++++++++++++++++++++++++++++++++++++++++++++
>> hw/mfi.h | 1281 ++++++++++++++++++++++++++++++++
>> hw/pci_ids.h | 3 +-
>> 5 files changed, 3150 insertions(+), 1 deletions(-)
>> create mode 100644 hw/megasas.c
>> create mode 100644 hw/mfi.h
>>
>> diff --git a/Makefile.objs b/Makefile.objs
>> index 391e524..5841998 100644
>> --- a/Makefile.objs
>> +++ b/Makefile.objs
>> @@ -283,6 +283,7 @@ hw-obj-$(CONFIG_AHCI) += ide/ich.o
>>
>> # SCSI layer
>> hw-obj-$(CONFIG_LSI_SCSI_PCI) += lsi53c895a.o
>> +hw-obj-$(CONFIG_MEGASAS_SCSI_PCI) += megasas.o
>> hw-obj-$(CONFIG_ESP) += esp.o
>>
>> hw-obj-y += dma-helpers.o sysbus.o isa-bus.o
>> diff --git a/default-configs/pci.mak b/default-configs/pci.mak
>> index 9d3e1db..4b49c00 100644
>> --- a/default-configs/pci.mak
>> +++ b/default-configs/pci.mak
>> @@ -10,6 +10,7 @@ CONFIG_EEPRO100_PCI=y
>> CONFIG_PCNET_PCI=y
>> CONFIG_PCNET_COMMON=y
>> CONFIG_LSI_SCSI_PCI=y
>> +CONFIG_MEGASAS_SCSI_PCI=y
>> CONFIG_RTL8139_PCI=y
>> CONFIG_E1000_PCI=y
>> CONFIG_IDE_CORE=y
>> diff --git a/hw/megasas.c b/hw/megasas.c
>> new file mode 100644
>> index 0000000..083c3d3
>> --- /dev/null
>> +++ b/hw/megasas.c
>> @@ -0,0 +1,1865 @@
>> +/*
>> + * QEMU MegaRAID SAS 8708EM2 Host Bus Adapter emulation
>
> Link to documentation/hardware spec/driver code?
>
>> + *
>> + * Copyright (c) 2009-2012 Hannes Reinecke, SUSE Labs
>> + *
>> + * This library is free software; you can redistribute it and/or
>> + * modify it under the terms of the GNU Lesser General Public
>> + * License as published by the Free Software Foundation; either
>> + * version 2 of the License, or (at your option) any later version.
>> + *
>> + * This library 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
>> + * Lesser General Public License for more details.
>> + *
>> + * You should have received a copy of the GNU Lesser General Public
>> + * License along with this library; if not, see<http://www.gnu.org/licenses/>.
>> + */
>> +
>> +#include "hw.h"
>> +#include "pci.h"
>> +#include "dma.h"
>> +#include "msix.h"
>> +#include "iov.h"
>> +#include "scsi.h"
>> +#include "scsi-defs.h"
>> +#include "block_int.h"
>> +
>> +#include "mfi.h"
>> +
>> +/* Static definitions */
>
> Remove above comment.
>
>> +#define MEGASAS_VERSION "1.60"
>> +#define MEGASAS_MAX_FRAMES 2048 /* Firmware limit at 65535 */
>> +#define MEGASAS_DEFAULT_FRAMES 1000 /* Windows requires this */
>> +#define MEGASAS_MAX_SGE 128 /* Firmware limit */
>> +#define MEGASAS_DEFAULT_SGE 80
>> +#define MEGASAS_MAX_SECTORS 0xFFFF /* No real limit */
>> +#define MEGASAS_MAX_ARRAYS 128
>> +
>> +#define MEGASAS_FLAG_USE_JBOD 0
>> +#define MEGASAS_MASK_USE_JBOD (1<< MEGASAS_FLAG_USE_JBOD)
>> +#define MEGASAS_FLAG_USE_MSIX 1
>> +#define MEGASAS_MASK_USE_MSIX (1<< MEGASAS_FLAG_USE_MSIX)
>> +#define MEGASAS_FLAG_USE_QUEUE64 2
>> +#define MEGASAS_MASK_USE_QUEUE64 (1<< MEGASAS_FLAG_USE_QUEUE64)
>> +
>> +const char *megasas_raid_modes[] = {
>> + "raid", "jbod"
>> +};
>> +
>> +const char *mfi_frame_desc[] = {
>> + "MFI init", "LD Read", "LD Write", "LD SCSI", "PD SCSI",
>> + "MFI Doorbell", "MFI Abort", "MFI SMP", "MFI Stop"};
>
> Can't the above go into mfi.h?
>
>> +
>> +struct megasas_cmd_t {
>> + uint32_t index;
>> + uint16_t flags;
>> + uint16_t count;
>> + uint64_t context;
>> +
>> + target_phys_addr_t pa;
>> + target_phys_addr_t pa_size;
>> + union mfi_frame *frame;
>> + SCSIRequest *req;
>> + struct iovec iov[MEGASAS_MAX_SGE];
>> + int iov_cnt;
>> + size_t iov_size;
>> + size_t iov_offset;
>> + struct megasas_state_t *state;
>> +};
>> +
>> +typedef struct megasas_state_t {
>> + PCIDevice dev;
>> + MemoryRegion mmio_io;
>> + MemoryRegion port_io;
>> + MemoryRegion queue_io;
>> + uint32_t frame_hi;
>> +
>> + int fw_state;
>> + uint32_t fw_sge;
>> + uint32_t fw_cmds;
>> + uint32_t flags;
>> + int fw_luns;
>> + int intr_mask;
>> + int doorbell;
>> + int busy;
>> + char *raid_mode_str;
>
> make it const char * and you will not need
> casts from const char * to char * which are just wrong.
>
>> +
>> + struct megasas_cmd_t *event_cmd;
>> + int event_locale;
>> + int event_class;
>> + int event_count;
>> + int shutdown_event;
>> + int boot_event;
>> +
>> + uint64_t reply_queue_pa;
>> + void *reply_queue;
>> + int reply_queue_len;
>> + int reply_queue_head;
>> + int reply_queue_tail;
>> + uint64_t consumer_pa;
>> + uint64_t producer_pa;
>> +
>> + struct megasas_cmd_t frames[MEGASAS_MAX_FRAMES];
>> +
>> + SCSIBus bus;
>> +} MPTState;
>
> Prefix with megasas or mfi, consistently.
>
>> +
>> +#define MEGASAS_INTR_DISABLED_MASK 0xFFFFFFFF
>> +
>> +static bool megasas_intr_enabled(MPTState *s)
>> +{
>> + if ((s->intr_mask& MEGASAS_INTR_DISABLED_MASK) !=
>> + MEGASAS_INTR_DISABLED_MASK) {
>> + return true;
>> + }
>> + return false;
>> +}
>> +
>> +static bool megasas_use_queue64(MPTState *s)
>> +{
>> + return s->flags& MEGASAS_MASK_USE_QUEUE64;
>> +}
>> +
>> +static bool megasas_use_msix(MPTState *s)
>> +{
>> + return s->flags& MEGASAS_MASK_USE_MSIX;
>> +}
>> +
>> +static bool megasas_is_jbod(MPTState *s)
>> +{
>> + return s->flags& MEGASAS_MASK_USE_JBOD;
>> +}
>> +
>> +static void megasas_frame_set_cmd_status(unsigned long frame, uint8_t v)
>> +{
>> + stb_phys(frame + offsetof(struct mfi_frame_header, cmd_status), v);
>> +}
>> +
>> +static void megasas_frame_set_scsi_status(unsigned long frame, uint8_t v)
>> +{
>> + stb_phys(frame + offsetof(struct mfi_frame_header, scsi_status), v);
>> +}
>> +
>> +/*
>> + * Context is considered opaque, but the HBA firmware is running
>> + * in little endian mode. So convert it to little endian, too.
>> + */
>> +static uint64_t megasas_frame_get_context(unsigned long frame)
>> +{
>> + return ldq_le_phys(frame + offsetof(struct mfi_frame_header, context));
>> +}
>> +
>> +static bool megasas_frame_is_ieee_sgl(struct megasas_cmd_t *cmd)
>> +{
>> + return cmd->flags& MFI_FRAME_IEEE_SGL;
>> +}
>> +
>> +static bool megasas_frame_is_sgl64(struct megasas_cmd_t *cmd)
>> +{
>> + return cmd->flags& MFI_FRAME_SGL64;
>> +}
>> +
>> +static bool megasas_frame_is_write(struct megasas_cmd_t *cmd)
>> +{
>> + return cmd->flags& MFI_FRAME_DIR_WRITE;
>> +}
>> +
>> +static bool megasas_frame_is_sense64(struct megasas_cmd_t *cmd)
>> +{
>> + return cmd->flags& MFI_FRAME_SENSE64;
>> +}
>> +
>> +static uint64_t megasas_sgl_get_addr(struct megasas_cmd_t *cmd,
>> + union mfi_sgl *sgl)
>> +{
>> + uint64_t addr;
>> +
>> + if (megasas_frame_is_ieee_sgl(cmd)) {
>> + addr = le64_to_cpu(sgl->sg_skinny->addr);
>> + } else if (megasas_frame_is_sgl64(cmd)) {
>> + addr = le64_to_cpu(sgl->sg64->addr);
>> + } else {
>> + addr = le32_to_cpu(sgl->sg32->addr);
>> + }
>> + return addr;
>> +}
>> +
>> +static uint32_t megasas_sgl_get_len(struct megasas_cmd_t *cmd,
>> + union mfi_sgl *sgl)
>> +{
>> + uint32_t len;
>> +
>> + if (megasas_frame_is_ieee_sgl(cmd)) {
>> + len = le32_to_cpu(sgl->sg_skinny->len);
>> + } else if (megasas_frame_is_sgl64(cmd)) {
>> + len = le32_to_cpu(sgl->sg64->len);
>> + } else {
>> + len = le32_to_cpu(sgl->sg32->len);
>> + }
>> + return len;
>> +}
>> +
>> +static union mfi_sgl *megasas_sgl_next(struct megasas_cmd_t *cmd,
>> + union mfi_sgl *sgl)
>> +{
>> + uint8_t *next = (uint8_t *)sgl;
>> +
>> + if (megasas_frame_is_ieee_sgl(cmd)) {
>> + next += sizeof(struct mfi_sg_skinny);
>> + } else if (megasas_frame_is_sgl64(cmd)) {
>> + next += sizeof(struct mfi_sg64);
>> + } else {
>> + next += sizeof(struct mfi_sg32);
>> + }
>> +
>> + if (next>= (uint8_t *)cmd->frame + cmd->pa_size) {
>> + return NULL;
>> + }
>> + return (union mfi_sgl *)next;
>> +}
>> +
>> +static void megasas_soft_reset(MPTState *s);
>> +
>> +static int megasas_map_sgl(struct megasas_cmd_t *cmd, union mfi_sgl *sgl)
>> +{
>> + int i;
>> + int is_write;
>> + int iov_count = 0;
>> + size_t iov_size = 0;
>> +
>> + cmd->flags = le16_to_cpu(cmd->frame->header.flags);
>> + cmd->iov_cnt = iov_count = cmd->frame->header.sge_count;
>> + if (cmd->iov_cnt> MEGASAS_MAX_SGE) {
>> + cmd->iov_cnt = 0;
>> + return iov_count;
>> + }
>> + is_write = megasas_frame_is_write(cmd);
>> + for (i = 0; i< cmd->iov_cnt; i++) {
>> + target_phys_addr_t iov_pa, iov_size_p;
>> +
>> + if (!sgl) {
>> + goto unmap;
>> + }
>> + iov_pa = megasas_sgl_get_addr(cmd, sgl);
>> + cmd->iov[i].iov_len = iov_size_p = megasas_sgl_get_len(cmd, sgl);
>> + if (!iov_pa || !iov_size_p) {
>> + goto unmap;
>> + }
>> + cmd->iov[i].iov_base = cpu_physical_memory_map(iov_pa,&iov_size_p,
>> + is_write);
>> + if (!iov_size_p || iov_size_p != cmd->iov[i].iov_len) {
>> + if (cmd->iov[i].iov_base) {
>> + cpu_physical_memory_unmap(cmd->iov[i].iov_base,
>> + iov_size_p, 0, 0);
>> + }
>> + goto unmap;
>> + }
>> + sgl = megasas_sgl_next(cmd, sgl);
>> + iov_size += cmd->iov[i].iov_len;
>> + iov_count--;
>> + }
>> + cmd->iov[i].iov_base = NULL;
>> + cmd->iov[i].iov_len = 0;
>> + cmd->iov_offset = 0;
>> + return 0;
>> +unmap:
>> + while (i> 0) {
>> + cpu_physical_memory_unmap(cmd->iov[i-1].iov_base,
>> + cmd->iov[i-1].iov_len, 0, 0);
>> + i--;
>> + }
>> + cmd->iov_cnt = 0;
>> + return iov_count;
>> +}
>> +
>> +static void megasas_unmap_sgl(struct megasas_cmd_t *cmd)
>> +{
>> + int i, is_write = megasas_frame_is_write(cmd);
>> +
>> + for (i = 0; i< cmd->iov_cnt; i++) {
>> + cpu_physical_memory_unmap(cmd->iov[i].iov_base, cmd->iov[i].iov_len,
>> + is_write, cmd->iov[i].iov_len);
>> + }
>> + cmd->iov_cnt = 0;
>> + cmd->iov_offset = 0;
>> +}
>> +
>> +/*
>> + * passthrough sense and io sense are at the same offset
>> + */
>> +static int megasas_build_sense(struct megasas_cmd_t *cmd, uint8_t *sense_ptr,
>> + uint8_t sense_len)
>> +{
>> + uint32_t pa_hi = 0, pa_lo;
>> + target_phys_addr_t pa;
>> +
>> + if (sense_len> cmd->frame->header.sense_len) {
>> + sense_len = cmd->frame->header.sense_len;
>> + }
>> +
>> + pa_lo = le32_to_cpu(cmd->frame->pass.sense_addr_lo);
>> + if (megasas_frame_is_sense64(cmd)) {
>> + pa_hi = le32_to_cpu(cmd->frame->pass.sense_addr_hi);
>> + }
>> + pa = ((uint64_t) pa_hi<< 32) | pa_lo;
>> + cpu_physical_memory_write(pa, sense_ptr, sense_len);
>> + cmd->frame->header.sense_len = sense_len;
>> + return sense_len;
>> +}
>> +
>> +static void megasas_write_sense(struct megasas_cmd_t *cmd, SCSISense sense)
>> +{
>> + uint8_t sense_buf[SCSI_SENSE_BUF_SIZE];
>> + uint8_t sense_len = 18;
>> +
>> + memset(sense_buf, 0, sense_len);
>> + sense_buf[0] = 0xf0;
>> + sense_buf[2] = sense.key;
>> + sense_buf[7] = 10;
>> + sense_buf[12] = sense.asc;
>> + sense_buf[13] = sense.ascq;
>> + megasas_build_sense(cmd, sense_buf, sense_len);
>> +}
>> +
>> +static void megasas_copy_sense(struct megasas_cmd_t *cmd)
>> +{
>> + uint8_t sense_buf[SCSI_SENSE_BUF_SIZE];
>> + uint8_t sense_len;
>> +
>> + sense_len = scsi_req_get_sense(cmd->req, sense_buf,
>> + cmd->frame->header.sense_len);
>> + megasas_build_sense(cmd, sense_buf, sense_len);
>> +}
>> +
>> +/*
>> + * Format an INQUIRY CDB
>> + */
>> +static int megasas_setup_inquiry(uint8_t *cdb, int pg,
>> + uint8_t *buf, int len)
>> +{
>> + memset(cdb, 0, 6);
>> + cdb[0] = INQUIRY;
>> + if (pg> 0) {
>> + cdb[1] = 0x1;
>> + cdb[2] = pg;
>> + }
>> + cdb[3] = (len>> 8)& 0xff;
>> + cdb[4] = (len& 0xff);
>> + return len;
>> +}
>> +
>> +/*
>> + * Encode lba and len into a READ_16/WRITE_16 CDB
>> + */
>> +static void megasas_encode_lba(uint8_t *cdb, uint64_t lba,
>> + uint32_t len, int is_write)
>> +{
>> + memset(cdb, 0x0, 16);
>> + if (is_write) {
>> + cdb[0] = WRITE_16;
>> + } else {
>> + cdb[0] = READ_16;
>> + }
>> + cdb[2] = (lba>> 56)& 0xff;
>> + cdb[3] = (lba>> 48)& 0xff;
>> + cdb[4] = (lba>> 40)& 0xff;
>> + cdb[5] = (lba>> 32)& 0xff;
>> + cdb[6] = (lba>> 24)& 0xff;
>> + cdb[7] = (lba>> 16)& 0xff;
>> + cdb[8] = (lba>> 8)& 0xff;
>> + cdb[9] = (lba)& 0xff;
>> + cdb[10] = (len>> 24)& 0xff;
>> + cdb[11] = (len>> 16)& 0xff;
>> + cdb[12] = (len>> 8)& 0xff;
>> + cdb[13] = (len)& 0xff;
>> +}
>> +
>> +/*
>> + * Utility functions
>> + */
>> +static uint64_t megasas_fw_time(void)
>> +{
>> + struct tm curtime;
>> + uint64_t bcd_time;
>> +
>> + qemu_get_timedate(&curtime, 0);
>> + bcd_time = ((uint64_t)curtime.tm_sec& 0xff)<< 48 |
>> + ((uint64_t)curtime.tm_min& 0xff)<< 40 |
>> + ((uint64_t)curtime.tm_hour& 0xff)<< 32 |
>> + ((uint64_t)curtime.tm_mday& 0xff)<< 24 |
>> + ((uint64_t)curtime.tm_mon& 0xff)<< 16 |
>> + ((uint64_t)(curtime.tm_year + 1900)& 0xffff);
>> +
>> + return bcd_time;
>> +}
>> +
>> +static uint64_t megasas_gen_sas_addr(uint64_t id)
>> +{
>> + uint64_t addr;
>> +
>> + addr = 0x5001a4aULL<< 36;
>> + addr |= id& 0xfffffffff;
>> +
>> + return addr;
>> +}
>> +
>> +/*
>> + * Frame handling
>> + */
>> +static int megasas_next_index(MPTState *s, int index, int limit)
>> +{
>> + index++;
>> + if (index == limit) {
>> + index = 0;
>> + }
>> + return index;
>> +}
>> +
>> +static struct megasas_cmd_t *megasas_lookup_frame(MPTState *s,
>> + target_phys_addr_t frame)
>> +{
>> + struct megasas_cmd_t *cmd = NULL;
>> + int num = 0, index;
>> +
>> + index = s->reply_queue_head;
>> +
>> + while (num< s->fw_cmds) {
>> + if (s->frames[index].pa&& s->frames[index].pa == frame) {
>> + cmd =&s->frames[index];
>> + break;
>> + }
>> + index = megasas_next_index(s, index, s->fw_cmds);
>> + num++;
>> + }
>> +
>> + return cmd;
>> +}
>> +
>> +static struct megasas_cmd_t *megasas_next_frame(MPTState *s,
>> + target_phys_addr_t frame)
>> +{
>> + struct megasas_cmd_t *cmd = NULL;
>> + int num = 0, index;
>> +
>> + cmd = megasas_lookup_frame(s, frame);
>> + if (cmd) {
>> + return cmd;
>> + }
>> + index = s->reply_queue_head;
>> + num = 0;
>> + while (num< s->fw_cmds) {
>> + if (!s->frames[index].pa) {
>> + cmd =&s->frames[index];
>> + break;
>> + }
>> + index = megasas_next_index(s, index, s->fw_cmds);
>> + num++;
>> + }
>> + return cmd;
>> +}
>> +
>> +static struct megasas_cmd_t *megasas_enqueue_frame(MPTState *s,
>> + target_phys_addr_t frame, uint64_t context, int count)
>> +{
>> + struct megasas_cmd_t *cmd = NULL;
>> + int frame_size = MFI_FRAME_SIZE * 16;
>> + target_phys_addr_t frame_size_p = frame_size;
>> +
>> + cmd = megasas_next_frame(s, frame);
>> + /* All frames busy */
>> + if (!cmd) {
>> + return NULL;
>> + }
>> + if (!cmd->pa) {
>> + cmd->pa = frame;
>> + /* Map all possible frames */
>> + cmd->frame = cpu_physical_memory_map(frame,&frame_size_p, 0);
>> + if (frame_size_p != frame_size) {
>> + if (cmd->frame) {
>> + cpu_physical_memory_unmap(cmd->frame, frame_size_p, 0, 0);
>> + cmd->frame = NULL;
>> + cmd->pa = 0;
>> + }
>> + s->event_count++;
>> + return NULL;
>> + }
>> + cmd->pa_size = frame_size_p;
>> + cmd->context = context;
>> + if (!megasas_use_queue64(s)) {
>> + cmd->context&= (uint64_t)0xFFFFFFFF;
>> + }
>> + }
>> + cmd->count = count;
>> + s->busy++;
>> +
>> + return cmd;
>> +}
>> +
>> +static void megasas_complete_frame(MPTState *s, uint64_t context)
>> +{
>> + int tail, queue_offset;
>> +
>> + /* Decrement busy count */
>> + s->busy--;
>> +
>> + if (s->reply_queue_pa) {
>> + /*
>> + * Put command on the reply queue.
>> + * Context is opaque, but emulation is running in
>> + * little endian. So convert it.
>> + */
>> + tail = s->reply_queue_head;
>> + if (megasas_use_queue64(s)) {
>> + queue_offset = tail * sizeof(uint64_t);
>> + stq_le_phys(s->reply_queue_pa + queue_offset, context);
>> + } else {
>> + queue_offset = tail * sizeof(uint32_t);
>> + stl_le_phys(s->reply_queue_pa + queue_offset, context);
>> + }
>> +
>> + s->reply_queue_head = megasas_next_index(s, tail, s->fw_cmds);
>> + }
>> +
>> + if (megasas_intr_enabled(s)) {
>> + /* Notify HBA */
>> + s->doorbell++;
>> + if (s->doorbell == 1) {
>> + if (msix_enabled(&s->dev)) {
>> + msix_notify(&s->dev, 0);
>> + } else {
>> + qemu_irq_raise(s->dev.irq[0]);
>> + }
>> + }
>> + }
>> +}
>> +
>> +static void megasas_reset_frames(MPTState *s)
>> +{
>> + int i;
>> + struct megasas_cmd_t *cmd;
>> +
>> + for (i = 0; i< s->fw_cmds; i++) {
>> + cmd =&s->frames[i];
>> + if (cmd->pa) {
>> + cpu_physical_memory_unmap(cmd->frame, cmd->pa_size, 0, 0);
>> + cmd->frame = NULL;
>> + cmd->pa = 0;
>> + }
>> + }
>> +}
>> +
>> +static void megasas_abort_command(struct megasas_cmd_t *cmd)
>> +{
>> + if (cmd->req) {
>> + scsi_req_abort(cmd->req, ABORTED_COMMAND);
>> + cmd->req = NULL;
>> + }
>> +}
>> +
>> +static int megasas_init_firmware(MPTState *s, struct megasas_cmd_t *cmd)
>> +{
>> + uint32_t pa_hi, pa_lo;
>> + target_phys_addr_t iq_pa, initq_size;
>> + struct mfi_init_qinfo *initq;
>> + uint32_t flags;
>> + int ret = MFI_STAT_OK;
>> +
>> + pa_lo = le32_to_cpu(cmd->frame->init.qinfo_new_addr_lo);
>> + pa_hi = le32_to_cpu(cmd->frame->init.qinfo_new_addr_hi);
>> + iq_pa = (((uint64_t) pa_hi<< 32) | pa_lo);
>> + initq_size = sizeof(*initq);
>> + initq = cpu_physical_memory_map(iq_pa,&initq_size, 0);
>> + if (!initq || initq_size != sizeof(*initq)) {
>> + s->event_count++;
>> + ret = MFI_STAT_MEMORY_NOT_AVAILABLE;
>> + goto out;
>> + }
>> + s->reply_queue_len = le32_to_cpu(initq->rq_entries)& 0xFFFF;
>> + if (s->reply_queue_len> s->fw_cmds) {
>> + s->event_count++;
>> + ret = MFI_STAT_INVALID_PARAMETER;
>> + goto out;
>> + }
>> + pa_lo = le32_to_cpu(initq->rq_addr_lo);
>> + pa_hi = le32_to_cpu(initq->rq_addr_hi);
>> + s->reply_queue_pa = ((uint64_t) pa_hi<< 32) | pa_lo;
>> + pa_lo = le32_to_cpu(initq->ci_addr_lo);
>> + pa_hi = le32_to_cpu(initq->ci_addr_hi);
>> + s->consumer_pa = ((uint64_t) pa_hi<< 32) | pa_lo;
>> + pa_lo = le32_to_cpu(initq->pi_addr_lo);
>> + pa_hi = le32_to_cpu(initq->pi_addr_hi);
>> + s->producer_pa = ((uint64_t) pa_hi<< 32) | pa_lo;
>> + s->reply_queue_head = ldl_le_phys(s->producer_pa);
>> + s->reply_queue_tail = ldl_le_phys(s->consumer_pa);
>> + flags = le32_to_cpu(initq->flags);
>> + if (flags& MFI_QUEUE_FLAG_CONTEXT64) {
>> + s->flags |= MEGASAS_MASK_USE_QUEUE64;
>> + }
>> + megasas_reset_frames(s);
>> + s->fw_state = MFI_FWSTATE_OPERATIONAL;
>> +out:
>> + if (initq) {
>> + cpu_physical_memory_unmap(initq, initq_size, 0, 0);
>> + }
>> + return ret;
>> +}
>> +
>> +static int megasas_map_dcmd(struct megasas_cmd_t *cmd)
>> +{
>> + target_phys_addr_t iov_pa, iov_size;
>> +
>> + cmd->flags = le16_to_cpu(cmd->frame->header.flags);
>> + if (!cmd->frame->header.sge_count) {
>> + cmd->iov_size = 0;
>> + cmd->iov_cnt = 0;
>> + return 0;
>> + } else if (cmd->frame->header.sge_count> 1) {
>> + cmd->iov_size = 0;
>> + cmd->iov_cnt = 0;
>> + return -1;
>> + }
>> + iov_pa = megasas_sgl_get_addr(cmd,&cmd->frame->dcmd.sgl);
>> + iov_size = megasas_sgl_get_len(cmd,&cmd->frame->dcmd.sgl);
>> + cmd->iov[0].iov_len = iov_size;
>> + cmd->iov[0].iov_base = cpu_physical_memory_map(iov_pa,&iov_size, 1);
>> + if (cmd->iov[0].iov_len != iov_size) {
>> + if (cmd->iov[0].iov_base) {
>> + cpu_physical_memory_unmap(cmd->iov[0].iov_base, iov_size, 1, 0);
>> + }
>> + cmd->iov_size = 0;
>> + cmd->iov_cnt = 0;
>> + return -1;
>> + }
>> + cmd->iov_cnt = 1;
>> + cmd->iov_size = iov_size;
>> + return cmd->iov_size;
>> +}
>> +
>> +static void megasas_finish_dcmd(struct megasas_cmd_t *cmd)
>> +{
>> + uint32_t iov_size = cmd->iov[0].iov_len;
>> +
>> + if (!cmd->iov_cnt) {
>> + return;
>> + }
>> +
>> + cpu_physical_memory_unmap(cmd->iov[0].iov_base, iov_size,
>> + 1, cmd->iov_size);
>> +
>> + cmd->iov_cnt = 0;
>> + if (iov_size> cmd->iov_size) {
>> + if (megasas_frame_is_ieee_sgl(cmd)) {
>> + cmd->frame->dcmd.sgl.sg_skinny->len = cpu_to_le32(iov_size);
>> + } else if (megasas_frame_is_sgl64(cmd)) {
>> + cmd->frame->dcmd.sgl.sg64->len = cpu_to_le32(iov_size);
>> + } else {
>> + cmd->frame->dcmd.sgl.sg32->len = cpu_to_le32(iov_size);
>> + }
>> + }
>> + cmd->iov_size = 0;
>> + return;
>> +}
>> +
>> +static int megasas_ctrl_get_info(MPTState *s, struct megasas_cmd_t *cmd)
>> +{
>> + struct mfi_ctrl_info *info = cmd->iov[0].iov_base;
>> + DeviceState *qdev;
>> + int num_ld_disks = 0;
>> +
>> + QTAILQ_FOREACH(qdev,&s->bus.qbus.children, sibling) {
>> + num_ld_disks++;
>> + }
>> +
>> + memset(info, 0x0, cmd->iov_size);
>> + if (cmd->iov_size< sizeof(struct mfi_ctrl_info)) {
>> + return MFI_STAT_INVALID_PARAMETER;
>> + }
>> +
>> + info->pci.vendor = cpu_to_le16(PCI_VENDOR_ID_LSI_LOGIC);
>> + info->pci.device = cpu_to_le16(PCI_DEVICE_ID_LSI_SAS1078);
>> + info->pci.subvendor = cpu_to_le16(PCI_VENDOR_ID_LSI_LOGIC);
>> + info->pci.subdevice = cpu_to_le16(0x1013);
>> +
>> + info->host.type = MFI_INFO_HOST_PCIX;
>> + info->device.type = MFI_INFO_DEV_SAS3G;
>> + info->device.port_count = 2;
>> + info->device.port_addr[0] = cpu_to_le64(megasas_gen_sas_addr((uint64_t)s));
>> +
>> + memcpy(info->product_name, "MegaRAID SAS 8708EM2", 20);
>> + snprintf(info->serial_number, 32, "QEMU%08lx",
>> + (unsigned long)s& 0xFFFFFFFF);
>> + snprintf(info->package_version, 0x60, "%s-QEMU", QEMU_VERSION);
>> + memcpy(info->image_component[0].name, "APP", 8);
>> + memcpy(info->image_component[0].version, MEGASAS_VERSION "-QEMU", 32);
>> + memcpy(info->image_component[0].build_date, __DATE__, 16);
>> + memcpy(info->image_component[0].build_time, __TIME__, 16);
>> + info->image_component_count = 1;
>> + info->current_fw_time = cpu_to_le32(megasas_fw_time());
>> + info->max_arms = 32;
>> + info->max_spans = 8;
>> + info->max_arrays = MEGASAS_MAX_ARRAYS;
>> + info->max_lds = s->fw_luns;
>> + info->max_cmds = cpu_to_le16(s->fw_cmds);
>> + info->max_sg_elements = cpu_to_le16(s->fw_sge);
>> + info->max_request_size = cpu_to_le32(MEGASAS_MAX_SECTORS);
>> + info->lds_present = cpu_to_le16(num_ld_disks);
>> + info->pd_present = cpu_to_le16(num_ld_disks + 1);
>> + info->pd_disks_present = cpu_to_le16(num_ld_disks);
>> + info->hw_present = cpu_to_le32(MFI_INFO_HW_NVRAM |
>> + MFI_INFO_HW_MEM |
>> + MFI_INFO_HW_FLASH);
>> + info->memory_size = cpu_to_le16(512);
>> + info->nvram_size = cpu_to_le16(32);
>> + info->flash_size = cpu_to_le16(16);
>> + info->raid_levels = cpu_to_le32(MFI_INFO_RAID_0);
>> + info->adapter_ops = cpu_to_le32(MFI_INFO_AOPS_RBLD_RATE |
>> + MFI_INFO_AOPS_SELF_DIAGNOSTIC |
>> + MFI_INFO_AOPS_MIXED_ARRAY);
>> + info->ld_ops = cpu_to_le32(MFI_INFO_LDOPS_DISK_CACHE_POLICY |
>> + MFI_INFO_LDOPS_ACCESS_POLICY |
>> + MFI_INFO_LDOPS_IO_POLICY |
>> + MFI_INFO_LDOPS_WRITE_POLICY |
>> + MFI_INFO_LDOPS_READ_POLICY);
>> + info->max_strips_per_io = cpu_to_le16(s->fw_sge);
>> + info->stripe_sz_ops.min = 3;
>> + info->stripe_sz_ops.max = ffs(MEGASAS_MAX_SECTORS + 1) - 1;
>> + info->properties.pred_fail_poll_interval = cpu_to_le16(300);
>> + info->properties.intr_throttle_cnt = cpu_to_le16(16);
>> + info->properties.intr_throttle_timeout = cpu_to_le16(50);
>> + info->properties.rebuild_rate = 30;
>> + info->properties.patrol_read_rate = 30;
>> + info->properties.bgi_rate = 30;
>> + info->properties.cc_rate = 30;
>> + info->properties.recon_rate = 30;
>> + info->properties.cache_flush_interval = 4;
>> + info->properties.spinup_drv_cnt = 2;
>> + info->properties.spinup_delay = 6;
>> + info->properties.ecc_bucket_size = 15;
>> + info->properties.ecc_bucket_leak_rate = cpu_to_le16(1440);
>> + info->properties.expose_encl_devices = 1;
>> + info->properties.OnOffProperties.enableJBOD = 1;
>> + info->pd_ops = cpu_to_le32(MFI_INFO_PDOPS_FORCE_ONLINE |
>> + MFI_INFO_PDOPS_FORCE_OFFLINE);
>> + info->pd_mix_support = cpu_to_le32(MFI_INFO_PDMIX_SAS |
>> + MFI_INFO_PDMIX_SATA |
>> + MFI_INFO_PDMIX_LD);
>> + cmd->iov_size = sizeof(struct mfi_ctrl_info);
>> + return MFI_STAT_OK;
>> +}
>> +
>> +static int megasas_mfc_get_defaults(MPTState *s, struct megasas_cmd_t *cmd)
>> +{
>> + struct mfi_defaults *info = cmd->iov[0].iov_base;
>> +
>> + memset(info, 0x0, cmd->iov_size);
>> + if (cmd->iov_size< sizeof(struct mfi_defaults)) {
>> + return MFI_STAT_INVALID_PARAMETER;
>> + }
>> +
>> + info->sas_addr = cpu_to_le64(megasas_gen_sas_addr((uint64_t)s));
>> + info->stripe_size = 3;
>> + info->flush_time = 4;
>> + info->background_rate = 30;
>> + info->allow_mix_in_enclosure = 1;
>> + info->allow_mix_in_ld = 1;
>> + info->direct_pd_mapping = 1;
>> + /* Enable for BIOS support */
>> + info->bios_enumerate_lds = 1;
>> + info->disable_ctrl_r = 1;
>> + info->expose_enclosure_devices = 1;
>> + info->disable_preboot_cli = 1;
>> + info->cluster_disable = 1;
>> +
>> + cmd->iov_size = sizeof(struct mfi_defaults);
>> + return MFI_STAT_OK;
>> +}
>> +
>> +static int megasas_dcmd_get_bios_info(MPTState *s, struct megasas_cmd_t *cmd)
>> +{
>> + struct mfi_bios_data *info = cmd->iov[0].iov_base;
>> +
>> + memset(info, 0x0, cmd->iov_size);
>> + if (cmd->iov_size< sizeof(struct mfi_bios_data)) {
>> + return MFI_STAT_INVALID_PARAMETER;
>> + }
>> + info->continue_on_error = 1;
>> + info->verbose = 1;
>> + if (megasas_is_jbod(s)) {
>> + info->expose_all_drives = 1;
>> + }
>> +
>> + cmd->iov_size = sizeof(struct mfi_bios_data);
>> + return MFI_STAT_OK;
>> +}
>> +
>> +static int megasas_dcmd_get_fw_time(MPTState *s, struct megasas_cmd_t *cmd)
>> +{
>> + uint64_t fw_time;
>> +
>> + fw_time = cpu_to_le64(megasas_fw_time());
>> +
>> + memcpy(cmd->iov[0].iov_base,&fw_time, sizeof(fw_time));
>> + cmd->iov_size = sizeof(fw_time);
>> + return MFI_STAT_OK;
>> +}
>> +
>> +static int megasas_dcmd_set_fw_time(MPTState *s, struct megasas_cmd_t *cmd)
>> +{
>> + uint64_t fw_time;
>> +
>> + /* This is a dummy; setting of firmware time is not allowed */
>> + memset(cmd->iov[0].iov_base, 0x0, cmd->iov_size);
>> + memcpy(&fw_time, cmd->frame->dcmd.mbox, sizeof(fw_time));
>> +
>> + fw_time = cpu_to_le64(megasas_fw_time());
>> + memcpy(cmd->iov[0].iov_base,&fw_time, sizeof(fw_time));
>> + cmd->iov_size = sizeof(fw_time);
>> + return MFI_STAT_OK;
>> +}
>> +
>> +static int megasas_event_info(MPTState *s, struct megasas_cmd_t *cmd)
>> +{
>> + struct mfi_evt_log_state *info = cmd->iov[0].iov_base;
>> +
>> + memset(info, 0, cmd->iov_size);
>> +
>> + info->newest_seq_num = cpu_to_le32(s->event_count);
>> + info->shutdown_seq_num = cpu_to_le32(s->shutdown_event);
>> + info->boot_seq_num = cpu_to_le32(s->boot_event);
>> +
>> + cmd->iov_size = sizeof(struct mfi_evt_log_state);
>> + return MFI_STAT_OK;
>> +}
>> +
>> +static int megasas_event_wait(MPTState *s, struct megasas_cmd_t *cmd)
>> +{
>> + union mfi_evt event;
>> +
>> + if (cmd->iov_size< sizeof(struct mfi_evt_detail)) {
>> + return MFI_STAT_INVALID_PARAMETER;
>> + }
>> + s->event_count = cpu_to_le32(cmd->frame->dcmd.mbox[0]);
>> + event.word = cpu_to_le32(cmd->frame->dcmd.mbox[4]);
>> + s->event_locale = event.members.locale;
>> + s->event_class = event.members.class;
>> + s->event_cmd = cmd;
>> + /* Decrease busy count; event frame doesn't count here */
>> + s->busy--;
>> + cmd->iov_size = sizeof(struct mfi_evt_detail);
>> + return MFI_STAT_INVALID_STATUS;
>> +}
>> +
>> +static int megasas_dcmd_pd_get_list(MPTState *s, struct megasas_cmd_t *cmd)
>> +{
>> + struct mfi_pd_list *info = cmd->iov[0].iov_base;
>> + DeviceState *qdev;
>> + uint32_t offset, num_pd_disks = 0, max_luns;
>> + uint16_t dev_id;
>> +
>> + memset(info, 0, cmd->iov_size);
>> + offset = 8;
>> + if (cmd->iov_size< (offset + sizeof(struct mfi_pd_address))) {
>> + return MFI_STAT_INVALID_PARAMETER;
>> + }
>> +
>> + max_luns = (cmd->iov_size - offset) / sizeof(struct mfi_pd_address);
>> + if (max_luns> s->fw_luns) {
>> + max_luns = s->fw_luns;
>> + }
>> +
>> + QTAILQ_FOREACH(qdev,&s->bus.qbus.children, sibling) {
>> + SCSIDevice *sdev = DO_UPCAST(SCSIDevice, qdev, qdev);
>> +
>> + dev_id = sdev->id;
>> + info->addr[num_pd_disks].device_id = cpu_to_le16(dev_id);
>> + info->addr[num_pd_disks].encl_device_id = cpu_to_le16(dev_id);
>> + info->addr[num_pd_disks].sas_addr[0] =
>> + cpu_to_le64(megasas_gen_sas_addr((uint64_t)sdev));
>> + num_pd_disks++;
>> + offset += sizeof(struct mfi_pd_address);
>> + }
>> +
>> + info->size = cpu_to_le32(offset);
>> + info->count = cpu_to_le32(num_pd_disks);
>> +
>> + return MFI_STAT_OK;
>> +}
>> +
>> +static int megasas_dcmd_pd_list_query(MPTState *s, struct megasas_cmd_t *cmd)
>> +{
>> + uint16_t flags;
>> +
>> + /* mbox0 contains flags */
>> + flags = le16_to_cpu(cmd->frame->dcmd.mbox[0]);
>> +
>> + if (flags == MR_PD_QUERY_TYPE_ALL || megasas_is_jbod(s)) {
>> + return megasas_dcmd_pd_get_list(s, cmd);
>> + }
>> +
>> + return MFI_STAT_OK;
>> +}
>> +
>> +static int megasas_pd_get_info_submit(SCSIDevice *sdev, int lun,
>> + struct megasas_cmd_t *cmd)
>> +{
>> + struct mfi_pd_info *info = cmd->iov[0].iov_base;
>> + uint8_t cmdbuf[6];
>> + SCSIRequest *req;
>> + size_t len;
>> +
>> + if (info->inquiry_data[4] == 0) {
>> + /* Additional length is zero, resubmit */
>> + megasas_setup_inquiry(cmdbuf, 0, info->inquiry_data,
>> + sizeof(info->inquiry_data));
>> + req = scsi_req_new(sdev, cmd->index, lun, cmdbuf, cmd);
>> + if (!req) {
>> + return MFI_STAT_FLASH_ALLOC_FAIL;
>> + }
>> + len = scsi_req_enqueue(req);
>> + if (len> 0) {
>> + cmd->iov_size = len;
>> + scsi_req_continue(req);
>> + }
>> + return MFI_STAT_INVALID_STATUS;
>> + } else if (info->vpd_page83[3] == 0) {
>> + /* Additional length is zero, resubmit */
>> + megasas_setup_inquiry(cmdbuf, 0x83, (uint8_t *)info->vpd_page83,
>> + sizeof(info->vpd_page83));
>> + req = scsi_req_new(sdev, cmd->index, lun, cmdbuf, cmd);
>> + if (!req) {
>> + return MFI_STAT_FLASH_ALLOC_FAIL;
>> + }
>> + len = scsi_req_enqueue(req);
>> + if (len> 0) {
>> + cmd->iov_size = len;
>> + scsi_req_continue(req);
>> + }
>> + return MFI_STAT_INVALID_STATUS;
>> + }
>> +
>> + /* Finished, set FW state */
>> + if (megasas_is_jbod(cmd->state)) {
>> + info->fw_state = cpu_to_le16(MFI_PD_STATE_SYSTEM);
>> + } else {
>> + info->fw_state = cpu_to_le16(MFI_PD_STATE_ONLINE);
>> + }
>> + return MFI_STAT_OK;
>> +}
>> +
>> +static int megasas_dcmd_pd_get_info(MPTState *s, struct megasas_cmd_t *cmd)
>> +{
>> + struct mfi_pd_info *info = cmd->iov[0].iov_base;
>> + uint64_t pd_size;
>> + uint16_t pd_id;
>> + SCSIDevice *sdev = NULL;
>> + int retval = MFI_STAT_DEVICE_NOT_FOUND;
>> +
>> + memset(info, 0, cmd->iov_size);
>> + if (cmd->iov_size< sizeof(struct mfi_pd_info)) {
>> + return MFI_STAT_INVALID_PARAMETER;
>> + }
>> +
>> + /* mbox0 has the ID */
>> + pd_id = le16_to_cpu(cmd->frame->dcmd.mbox[0]);
>> + sdev = scsi_device_find(&s->bus, 0, pd_id, 0);
>> +
>> + if (sdev) {
>> + BlockConf *conf =&sdev->conf;
>> +
>> + info->ref.v.device_id = pd_id;
>> + info->state.ddf.v.pd_type.in_vd = 1;
>> + info->state.ddf.v.pd_type.intf = 0x2;
>> + bdrv_get_geometry(conf->bs,&pd_size);
>> + info->raw_size = cpu_to_le64(pd_size);
>> + info->non_coerced_size = cpu_to_le64(pd_size);
>> + info->coerced_size = cpu_to_le64(pd_size);
>> + info->fw_state = cpu_to_le16(MFI_PD_STATE_OFFLINE);
>> + info->path_info.count = 1;
>> + info->path_info.sas_addr[0] =
>> + cpu_to_le64(megasas_gen_sas_addr((uint64_t)sdev));
>> + /* Submit inquiry */
>> + retval = megasas_pd_get_info_submit(sdev, pd_id, cmd);
>> + }
>> +
>> + return retval;
>> +}
>> +
>> +static int megasas_dcmd_ld_get_list(MPTState *s, struct megasas_cmd_t *cmd)
>> +{
>> + struct mfi_ld_list *info = cmd->iov[0].iov_base;
>> + uint32_t num_ld_disks = 0, max_ld_disks = s->fw_luns;
>> + uint64_t ld_size;
>> + int offset;
>> + DeviceState *qdev;
>> +
>> + memset(info, 0, cmd->iov_size);
>> + if (cmd->iov_size< sizeof(struct mfi_ld_list)) {
>> + return MFI_STAT_INVALID_PARAMETER;
>> + }
>> +
>> + if (megasas_is_jbod(s)) {
>> + max_ld_disks = 0;
>> + }
>> + QTAILQ_FOREACH(qdev,&s->bus.qbus.children, sibling) {
>> + SCSIDevice *sdev = DO_UPCAST(SCSIDevice, qdev, qdev);
>> + BlockConf *conf =&sdev->conf;
>> +
>> + if (num_ld_disks>= max_ld_disks)
>> + break;
>> + bdrv_get_geometry(conf->bs,&ld_size);
>> + ld_size *= 512;
>> + info->ld_list[num_ld_disks].ld.v.target_id = sdev->id;
>> + info->ld_list[num_ld_disks].state = MFI_LD_STATE_OPTIMAL;
>> + info->ld_list[num_ld_disks].size = cpu_to_le64(ld_size);
>> + num_ld_disks++;
>> + offset += 18;
>> + }
>> + info->ld_count = cpu_to_le32(num_ld_disks);
>> +
>> + return MFI_STAT_OK;
>> +}
>> +
>> +static int megasas_ld_get_info_submit(SCSIDevice *sdev, int lun,
>> + struct megasas_cmd_t *cmd)
>> +{
>> + struct mfi_ld_info *info = cmd->iov[0].iov_base;
>> + uint8_t cdb[6];
>> + SCSIRequest *req;
>> + ssize_t len;
>> +
>> + if (info->vpd_page83[3] == 0) {
>> + megasas_setup_inquiry(cdb, 0x83, (uint8_t *)info->vpd_page83,
>> + sizeof(info->vpd_page83));
>> + req = scsi_req_new(sdev, cmd->index, lun, cdb, cmd);
>> + if (!req) {
>> + return MFI_STAT_FLASH_ALLOC_FAIL;
>> + }
>> + len = scsi_req_enqueue(req);
>> + if (len> 0) {
>> + cmd->iov_size = len;
>> + scsi_req_continue(req);
>> + }
>> + return MFI_STAT_INVALID_STATUS;
>> + }
>> + info->ld_config.params.state = MFI_LD_STATE_OPTIMAL;
>> + return MFI_STAT_OK;
>> +}
>> +
>> +static int megasas_dcmd_ld_get_info(MPTState *s, struct megasas_cmd_t *cmd)
>> +{
>> + struct mfi_ld_info *info = cmd->iov[0].iov_base;
>> + uint64_t ld_size;
>> + uint16_t ld_id;
>> + uint32_t max_ld_disks = s->fw_luns;
>> + SCSIDevice *sdev = NULL;
>> + int retval = MFI_STAT_DEVICE_NOT_FOUND;
>> +
>> + memset(info, 0, cmd->iov_size);
>> + if (cmd->iov_size< sizeof(struct mfi_ld_info)) {
>> + return MFI_STAT_INVALID_PARAMETER;
>> + }
>> +
>> + /* mbox0 has the ID */
>> + ld_id = le16_to_cpu(cmd->frame->dcmd.mbox[0]);
>> +
>> + if (megasas_is_jbod(s)) {
>> + return MFI_STAT_DEVICE_NOT_FOUND;
>> + }
>> +
>> + if (ld_id< max_ld_disks) {
>> + sdev = scsi_device_find(&s->bus, 0, ld_id, 0);
>> + }
>> +
>> + if (sdev) {
>> + BlockConf *conf =&sdev->conf;
>> +
>> + info->ld_config.properties.ld.v.target_id = ld_id;
>> + info->ld_config.params.stripe_size = 3;
>> + info->ld_config.params.num_drives = 1;
>> + info->ld_config.params.state = MFI_LD_STATE_OFFLINE;
>> + info->ld_config.params.is_consistent = 1;
>> + bdrv_get_geometry(conf->bs,&ld_size);
>> + ld_size *= 512;
>> + info->size = cpu_to_le64(ld_size);
>> + retval = megasas_ld_get_info_submit(sdev, ld_id, cmd);
>> + }
>> +
>> + return retval;
>> +}
>> +
>> +static int megasas_dcmd_get_properties(MPTState *s, struct megasas_cmd_t *cmd)
>> +{
>> + struct mfi_ctrl_props *info = cmd->iov[0].iov_base;
>> +
>> + memset(info, 0x0, cmd->iov_size);
>> + if (cmd->iov_size< sizeof(struct mfi_ctrl_props)) {
>> + return MFI_STAT_INVALID_PARAMETER;
>> + }
>> + info->pred_fail_poll_interval = cpu_to_le16(300);
>> + info->intr_throttle_cnt = cpu_to_le16(16);
>> + info->intr_throttle_timeout = cpu_to_le16(50);
>> + info->rebuild_rate = 30;
>> + info->patrol_read_rate = 30;
>> + info->bgi_rate = 30;
>> + info->cc_rate = 30;
>> + info->recon_rate = 30;
>> + info->cache_flush_interval = 4;
>> + info->spinup_drv_cnt = 2;
>> + info->spinup_delay = 6;
>> + info->ecc_bucket_size = 15;
>> + info->ecc_bucket_leak_rate = cpu_to_le16(1440);
>> + info->expose_encl_devices = 1;
>> +
>> + return MFI_STAT_OK;
>> +}
>> +
>> +static int megasas_cache_flush(MPTState *s, struct megasas_cmd_t *cmd)
>> +{
>> + qemu_aio_flush();
>> + return MFI_STAT_OK;
>> +}
>> +
>> +static int megasas_ctrl_shutdown(MPTState *s, struct megasas_cmd_t *cmd)
>> +{
>> + s->fw_state = MFI_FWSTATE_READY;
>> + return MFI_STAT_OK;
>> +}
>> +
>> +static int megasas_cluster_reset_ld(MPTState *s, struct megasas_cmd_t *cmd)
>> +{
>> + return MFI_STAT_INVALID_DCMD;
>> +}
>> +
>> +static int megasas_dcmd_set_properties(MPTState *s, struct megasas_cmd_t *cmd)
>> +{
>> + return MFI_STAT_OK;
>> +}
>> +
>> +static int megasas_dcmd_dummy(MPTState *s, struct megasas_cmd_t *cmd)
>> +{
>> + if (cmd->iov_size) {
>> + memset(cmd->iov[0].iov_base, 0, cmd->iov_size);
>> + }
>> + return MFI_STAT_OK;
>> +}
>> +
>> +static const struct dcmd_cmd_tbl_t {
>> + int opcode;
>> + int (*func)(MPTState *s, struct megasas_cmd_t *cmd);
>> +} dcmd_cmd_tbl[] = {
>> + { MFI_DCMD_CTRL_MFI_HOST_MEM_ALLOC, megasas_dcmd_dummy },
>> + { MFI_DCMD_CTRL_GET_INFO, megasas_ctrl_get_info },
>> + { MFI_DCMD_CTRL_GET_PROPERTIES, megasas_dcmd_get_properties },
>> + { MFI_DCMD_CTRL_SET_PROPERTIES, megasas_dcmd_set_properties },
>> + { MFI_DCMD_CTRL_ALARM_GET, megasas_dcmd_dummy },
>> + { MFI_DCMD_CTRL_ALARM_ENABLE, megasas_dcmd_dummy },
>> + { MFI_DCMD_CTRL_ALARM_DISABLE, megasas_dcmd_dummy },
>> + { MFI_DCMD_CTRL_ALARM_SILENCE, megasas_dcmd_dummy },
>> + { MFI_DCMD_CTRL_ALARM_TEST, megasas_dcmd_dummy },
>> + { MFI_DCMD_CTRL_EVENT_GETINFO, megasas_event_info },
>> + { MFI_DCMD_CTRL_EVENT_GET, megasas_dcmd_dummy },
>> + { MFI_DCMD_CTRL_EVENT_WAIT, megasas_event_wait },
>> + { MFI_DCMD_CTRL_SHUTDOWN, megasas_ctrl_shutdown },
>> + { MFI_DCMD_HIBERNATE_STANDBY, megasas_dcmd_dummy },
>> + { MFI_DCMD_CTRL_GET_TIME, megasas_dcmd_get_fw_time },
>> + { MFI_DCMD_CTRL_SET_TIME, megasas_dcmd_set_fw_time },
>> + { MFI_DCMD_CTRL_BIOS_DATA_GET, megasas_dcmd_get_bios_info },
>> + { MFI_DCMD_CTRL_FACTORY_DEFAULTS, megasas_dcmd_dummy },
>> + { MFI_DCMD_CTRL_MFC_DEFAULTS_GET, megasas_mfc_get_defaults },
>> + { MFI_DCMD_CTRL_MFC_DEFAULTS_SET, megasas_dcmd_dummy },
>> + { MFI_DCMD_CTRL_CACHE_FLUSH, megasas_cache_flush },
>> + { MFI_DCMD_PD_GET_LIST, megasas_dcmd_pd_get_list },
>> + { MFI_DCMD_PD_LIST_QUERY, megasas_dcmd_pd_list_query },
>> + { MFI_DCMD_PD_GET_INFO, megasas_dcmd_pd_get_info },
>> + { MFI_DCMD_PD_STATE_SET, megasas_dcmd_dummy },
>> + { MFI_DCMD_PD_REBUILD, megasas_dcmd_dummy },
>> + { MFI_DCMD_PD_BLINK, megasas_dcmd_dummy },
>> + { MFI_DCMD_PD_UNBLINK, megasas_dcmd_dummy },
>> + { MFI_DCMD_LD_GET_LIST, megasas_dcmd_ld_get_list},
>> + { MFI_DCMD_LD_GET_INFO, megasas_dcmd_ld_get_info },
>> + { MFI_DCMD_LD_GET_PROP, megasas_dcmd_dummy },
>> + { MFI_DCMD_LD_SET_PROP, megasas_dcmd_dummy },
>> + { MFI_DCMD_LD_DELETE, megasas_dcmd_dummy },
>> + { MFI_DCMD_CFG_READ, megasas_dcmd_dummy },
>> + { MFI_DCMD_CFG_ADD, megasas_dcmd_dummy },
>> + { MFI_DCMD_CFG_CLEAR, megasas_dcmd_dummy },
>> + { MFI_DCMD_CFG_FOREIGN_READ, megasas_dcmd_dummy },
>> + { MFI_DCMD_CFG_FOREIGN_IMPORT, megasas_dcmd_dummy },
>> + { MFI_DCMD_BBU_STATUS, megasas_dcmd_dummy },
>> + { MFI_DCMD_BBU_CAPACITY_INFO, megasas_dcmd_dummy },
>> + { MFI_DCMD_BBU_DESIGN_INFO, megasas_dcmd_dummy },
>> + { MFI_DCMD_BBU_PROP_GET, megasas_dcmd_dummy },
>> + { MFI_DCMD_CLUSTER, megasas_dcmd_dummy },
>> + { MFI_DCMD_CLUSTER_RESET_ALL, megasas_dcmd_dummy },
>> + { MFI_DCMD_CLUSTER_RESET_LD, megasas_cluster_reset_ld },
>> + { -1, NULL }
>> +};
>> +
>> +static int megasas_handle_dcmd(MPTState *s, struct megasas_cmd_t *cmd)
>> +{
>> + int opcode, len;
>> + int retval = 0;
>> + const struct dcmd_cmd_tbl_t *cmdptr = dcmd_cmd_tbl;
>> +
>> + opcode = le32_to_cpu(cmd->frame->dcmd.opcode);
>> + len = megasas_map_dcmd(cmd);
>> + if (len< 0) {
>> + return MFI_STAT_MEMORY_NOT_AVAILABLE;
>> + }
>> + while (cmdptr->opcode != -1&& cmdptr->opcode != opcode) {
>> + cmdptr++;
>> + }
>> + if (cmdptr->opcode == -1) {
>> + retval = megasas_dcmd_dummy(s, cmd);
>> + } else {
>> + retval = cmdptr->func(s, cmd);
>> + }
>> + if (retval != MFI_STAT_INVALID_STATUS) {
>> + megasas_finish_dcmd(cmd);
>> + }
>> + return retval;
>> +}
>> +
>> +static int megasas_finish_internal_dcmd(struct megasas_cmd_t *cmd,
>> + SCSIRequest *req)
>> +{
>> + int opcode;
>> + int retval = MFI_STAT_OK;
>> + int lun = req->lun;
>> +
>> + opcode = le32_to_cpu(cmd->frame->dcmd.opcode);
>> + scsi_req_unref(req);
>> + switch (opcode) {
>> + case MFI_DCMD_PD_GET_INFO:
>> + retval = megasas_pd_get_info_submit(req->dev, lun, cmd);
>> + break;
>> + case MFI_DCMD_LD_GET_INFO:
>> + retval = megasas_ld_get_info_submit(req->dev, lun, cmd);
>> + break;
>> + default:
>> + retval = MFI_STAT_INVALID_DCMD;
>> + break;
>> + }
>> + if (retval != MFI_STAT_INVALID_STATUS) {
>> + megasas_finish_dcmd(cmd);
>> + }
>> + return retval;
>> +}
>> +
>> +static int megasas_handle_scsi(MPTState *s, struct megasas_cmd_t *cmd,
>> + int is_logical)
>> +{
>> + uint8_t *cdb;
>> + int len;
>> + struct SCSIDevice *sdev = NULL;
>> +
>> + cdb = cmd->frame->pass.cdb;
>> +
>> + if (cmd->frame->header.target_id< s->fw_luns) {
>> + sdev = scsi_device_find(&s->bus, 0, cmd->frame->header.target_id,
>> + cmd->frame->header.lun_id);
>> + }
>> + cmd->iov_size = le32_to_cpu(cmd->frame->header.data_len);
>> +
>> + if (!sdev || (megasas_is_jbod(s)&& is_logical)) {
>> + return MFI_STAT_DEVICE_NOT_FOUND;
>> + }
>> +
>> + if (cmd->frame->header.cdb_len> 16) {
>> + megasas_write_sense(cmd, SENSE_CODE(INVALID_OPCODE));
>> + cmd->frame->header.scsi_status = CHECK_CONDITION;
>> + s->event_count++;
>> + return MFI_STAT_SCSI_DONE_WITH_ERROR;
>> + }
>> +
>> + if (megasas_map_sgl(cmd,&cmd->frame->pass.sgl)) {
>> + megasas_write_sense(cmd, SENSE_CODE(TARGET_FAILURE));
>> + cmd->frame->header.scsi_status = CHECK_CONDITION;
>> + s->event_count++;
>> + return MFI_STAT_SCSI_DONE_WITH_ERROR;
>> + }
>> +
>> + cmd->req = scsi_req_new(sdev, cmd->index,
>> + cmd->frame->header.lun_id, cdb, cmd);
>> + if (!cmd->req) {
>> + megasas_write_sense(cmd, SENSE_CODE(NO_SENSE));
>> + cmd->frame->header.scsi_status = BUSY;
>> + s->event_count++;
>> + return MFI_STAT_SCSI_DONE_WITH_ERROR;
>> + }
>> +
>> + len = scsi_req_enqueue(cmd->req);
>> + if (len> 0) {
>> + if (len< cmd->iov_size) {
>> + cmd->iov_size = len;
>> + }
>> + scsi_req_continue(cmd->req);
>> + } else if (len< 0) {
>> + if (-len< cmd->iov_size) {
>> + cmd->iov_size = -len;
>> + }
>> + scsi_req_continue(cmd->req);
>> + }
>
> Shorter:
>
> if (len< 0) {
> len = -len;
> }
>
> if (len) {
> cmd->iov_size = MIN(len, cmd->iov_size);
> scsi_req_continue(cmd->req);
> }
>
>
>> + cmd->iov_size = -len;
>> + }
>
>> + return MFI_STAT_INVALID_STATUS;
>> +}
>> +
>> +static int megasas_handle_io(MPTState *s, struct megasas_cmd_t *cmd)
>> +{
>> + uint32_t lba_count, lba_start_hi, lba_start_lo;
>> + uint64_t lba_start;
>> + int write = cmd->frame->header.frame_cmd == MFI_CMD_LD_WRITE ? 1 : 0;
>> + uint8_t cdb[16];
>> + int len;
>> + struct SCSIDevice *sdev = NULL;
>> +
>> + lba_count = le32_to_cpu(cmd->frame->io.header.data_len);
>> + lba_start_lo = le32_to_cpu(cmd->frame->io.lba_lo);
>> + lba_start_hi = le32_to_cpu(cmd->frame->io.lba_hi);
>> + lba_start = ((uint64_t)lba_start_hi<< 32) | lba_start_lo;
>> +
>> + if (cmd->frame->header.target_id< s->fw_luns) {
>> + sdev = scsi_device_find(&s->bus, 0, cmd->frame->header.target_id,
>> + cmd->frame->header.lun_id);
>> + }
>> +
>> + if (!sdev) {
>> + return MFI_STAT_DEVICE_NOT_FOUND;
>> + }
>> +
>> + if (cmd->frame->header.cdb_len> 16) {
>> + megasas_write_sense(cmd, SENSE_CODE(INVALID_OPCODE));
>> + cmd->frame->header.scsi_status = CHECK_CONDITION;
>> + s->event_count++;
>> + return MFI_STAT_SCSI_DONE_WITH_ERROR;
>> + }
>> +
>> + cmd->iov_size = lba_count * sdev->blocksize;
>> + if (megasas_map_sgl(cmd,&cmd->frame->io.sgl)) {
>> + megasas_write_sense(cmd, SENSE_CODE(TARGET_FAILURE));
>> + cmd->frame->header.scsi_status = CHECK_CONDITION;
>> + s->event_count++;
>> + return MFI_STAT_SCSI_DONE_WITH_ERROR;
>> + }
>> +
>> + megasas_encode_lba(cdb, lba_start, lba_count, write);
>> + cmd->req = scsi_req_new(sdev, cmd->index,
>> + cmd->frame->header.lun_id, cdb, cmd);
>> + if (!cmd->req) {
>> + megasas_write_sense(cmd, SENSE_CODE(NO_SENSE));
>> + cmd->frame->header.scsi_status = BUSY;
>> + s->event_count++;
>> + return MFI_STAT_SCSI_DONE_WITH_ERROR;
>> + }
>> +
>> + len = scsi_req_enqueue(cmd->req);
>> + if (len> 0) {
>> + if (len< cmd->iov_size) {
>> + cmd->iov_size = len;
>> + }
>> + scsi_req_continue(cmd->req);
>> + } else if (len< 0) {
>> + if (-len< cmd->iov_size) {
>> + cmd->iov_size = -len;
>> + }
>> + scsi_req_continue(cmd->req);
>> + }
>> + return MFI_STAT_INVALID_STATUS;
>
> code duplicated from above. make it a function?
>
>> +}
>> +
>> +static int megasas_finish_internal_command(struct megasas_cmd_t *cmd,
>> + SCSIRequest *req)
>> +{
>> + int retval = MFI_STAT_INVALID_CMD;
>> +
>> + if (cmd->frame->header.frame_cmd == MFI_CMD_DCMD) {
>> + retval = megasas_finish_internal_dcmd(cmd, req);
>> + }
>> + return retval;
>> +}
>> +
>> +static void megasas_xfer_complete(SCSIRequest *req, uint32_t len)
>> +{
>> + struct megasas_cmd_t *cmd = req->hba_private;
>> + uint8_t *buf;
>> +
>> + if (len) {
>> + int is_write = megasas_frame_is_write(cmd);
>> + size_t bytes;
>> +
>> + buf = scsi_req_get_buf(req);
>> + if (is_write) {
>> + bytes = iov_to_buf(cmd->iov, cmd->iov_cnt, buf,
>> + cmd->iov_offset, len);
>> + if (bytes != len) {
>> + len = bytes;
>> + }
>
> as len is unused below this is dead code?
>
>> + cmd->iov_offset += bytes;
>> + } else {
>> + bytes = iov_from_buf(cmd->iov, cmd->iov_cnt, buf,
>> + cmd->iov_offset, len);
>> + if (bytes != len) {
>> + len = bytes;
>> + }
>
> as len is unused below this is dead code?
>
>> + cmd->iov_offset += bytes;
>> + }
>> + }
>
> so do we discard 0 length or continue?
> previous functions discard this one continues.
> Intentional?
>
>> + scsi_req_continue(req);
>> +}
>> +
>> +static void megasas_command_complete(SCSIRequest *req, uint32_t status)
>> +{
>> + struct megasas_cmd_t *cmd = req->hba_private;
>> + uint8_t cmd_status = MFI_STAT_OK;
>> +
>> + if (cmd->req != req) {
>> + /*
>> + * Internal command complete
>> + */
>> + cmd_status = megasas_finish_internal_command(cmd, req);
>> + if (cmd_status == MFI_STAT_INVALID_STATUS) {
>> + return;
>> + }
>> + } else {
>> + req->status = status;
>> + if (req->status != GOOD) {
>> + cmd_status = MFI_STAT_SCSI_DONE_WITH_ERROR;
>> + }
>> + if (req->status == CHECK_CONDITION) {
>> + megasas_copy_sense(cmd);
>> + }
>> +
>> + megasas_unmap_sgl(cmd);
>> + cmd->frame->header.scsi_status = req->status;
>> + scsi_req_unref(cmd->req);
>> + cmd->req = NULL;
>> + }
>> + cmd->frame->header.cmd_status = cmd_status;
>> + megasas_complete_frame(cmd->state, cmd->context);
>> +}
>> +
>> +static void megasas_command_cancel(SCSIRequest *req)
>> +{
>> + struct megasas_cmd_t *cmd = req->hba_private;
>> +
>> + if (cmd) {
>> + megasas_abort_command(cmd);
>> + } else {
>> + scsi_req_unref(req);
>> + }
>> + return;
>
> return here is useless
>
>> +}
>> +
>> +static int megasas_handle_abort(MPTState *s, struct megasas_cmd_t *cmd)
>> +{
>> + uint64_t abort_ctx = le64_to_cpu(cmd->frame->abort.abort_context);
>> + target_phys_addr_t abort_addr, addr_hi, addr_lo;
>> + struct megasas_cmd_t *abort_cmd;
>> +
>> + addr_hi = le32_to_cpu(cmd->frame->abort.abort_mfi_addr_hi);
>> + addr_lo = le32_to_cpu(cmd->frame->abort.abort_mfi_addr_lo);
>> + abort_addr = ((uint64_t)addr_hi<< 32) | addr_lo;
>> +
>> + abort_cmd = megasas_lookup_frame(s, abort_addr);
>> + if (!abort_cmd) {
>> + s->event_count++;
>> + return MFI_STAT_OK;
>> + }
>> + if (!megasas_use_queue64(s)) {
>> + abort_ctx&= (uint64_t)0xFFFFFFFF;
>> + }
>> + if (abort_cmd->context != abort_ctx) {
>> + s->event_count++;
>> + return MFI_STAT_ABORT_NOT_POSSIBLE;
>> + }
>> + megasas_abort_command(abort_cmd);
>> + if (!s->event_cmd || abort_cmd != s->event_cmd) {
>> + s->event_cmd = NULL;
>> + }
>> + s->event_count++;
>> + return MFI_STAT_OK;
>> +}
>> +
>> +static void megasas_handle_frame(MPTState *s, uint64_t frame_addr,
>> + uint32_t frame_count)
>> +{
>> + uint8_t frame_status = MFI_STAT_INVALID_CMD;
>> + uint64_t frame_context;
>> + struct megasas_cmd_t *cmd;
>> +
>> + /*
>> + * Always read 64bit context, top bits will be
>> + * masked out if required in megasas_enqueue_frame()
>> + */
>> + frame_context = megasas_frame_get_context(frame_addr);
>> +
>> + cmd = megasas_enqueue_frame(s, frame_addr, frame_context, frame_count);
>> + if (!cmd) {
>> + /* reply queue full */
>> + megasas_frame_set_scsi_status(frame_addr, BUSY);
>> + megasas_frame_set_cmd_status(frame_addr, MFI_STAT_SCSI_DONE_WITH_ERROR);
>> + megasas_complete_frame(s, frame_context);
>> + s->event_count++;
>> + return;
>> + }
>> + switch (cmd->frame->header.frame_cmd) {
>> + case MFI_CMD_INIT:
>> + frame_status = megasas_init_firmware(s, cmd);
>> + break;
>> + case MFI_CMD_DCMD:
>> + frame_status = megasas_handle_dcmd(s, cmd);
>> + break;
>> + case MFI_CMD_ABORT:
>> + frame_status = megasas_handle_abort(s, cmd);
>> + break;
>> + case MFI_CMD_PD_SCSI_IO:
>> + frame_status = megasas_handle_scsi(s, cmd, 0);
>> + break;
>> + case MFI_CMD_LD_SCSI_IO:
>> + frame_status = megasas_handle_scsi(s, cmd, 1);
>> + break;
>> + case MFI_CMD_LD_READ:
>> + case MFI_CMD_LD_WRITE:
>> + frame_status = megasas_handle_io(s, cmd);
>> + break;
>> + default:
>> + s->event_count++;
>> + break;
>> + }
>> + if (frame_status != MFI_STAT_INVALID_STATUS) {
>> + if (cmd->frame) {
>> + cmd->frame->header.cmd_status = frame_status;
>> + } else {
>> + megasas_frame_set_cmd_status(frame_addr, frame_status);
>> + }
>> + megasas_complete_frame(s, cmd->context);
>> + }
>> +}
>> +
>> +static uint64_t megasas_mmio_read(void *opaque, target_phys_addr_t addr,
>> + unsigned size)
>> +{
>> + MPTState *s = opaque;
>> + uint32_t retval = 0;
>> +
>> + switch (addr) {
>> + case MFI_IDB:
>> + retval = 0;
>> + break;
>> + case MFI_OMSG0:
>> + case MFI_OSP0:
>> + retval = (megasas_use_msix(s) ? MFI_FWSTATE_MSIX_SUPPORTED : 0) |
>> + (s->fw_state& MFI_FWSTATE_MASK) |
>> + ((s->fw_sge& 0xff)<< 16) |
>> + (s->fw_cmds& 0xFFFF);
>> + break;
>> + case MFI_OSTS:
>> + if (megasas_intr_enabled(s)&& s->doorbell) {
>> + retval = MFI_1078_RM | 1;
>> + }
>> + break;
>> + case MFI_OMSK:
>> + retval = s->intr_mask;
>> + break;
>> + case MFI_ODCR0:
>> + retval = s->doorbell;
>> + break;
>> + default:
>> + break;
>> + }
>> + return retval;
>> +}
>> +
>> +static void megasas_mmio_write(void *opaque, target_phys_addr_t addr,
>> + uint64_t val, unsigned size)
>> +{
>> + MPTState *s = opaque;
>> + uint64_t frame_addr;
>> + uint32_t frame_count;
>> + int i;
>> +
>> + switch (addr) {
>> + case MFI_IDB:
>> + if (val& MFI_FWINIT_ABORT) {
>> + /* Abort all pending cmds */
>> + for (i = 0; i< s->fw_cmds; i++) {
>> + megasas_abort_command(&s->frames[i]);
>> + }
>> + }
>> + if (val& MFI_FWINIT_READY) {
>> + /* move to FW READY */
>> + megasas_soft_reset(s);
>> + }
>> + if (val& MFI_FWINIT_MFIMODE) {
>> + /* discard MFIs */
>> + }
>> + break;
>> + case MFI_OMSK:
>> + s->intr_mask = val;
>> + if (!megasas_intr_enabled(s)&& !msix_enabled(&s->dev)) {
>> + qemu_irq_lower(s->dev.irq[0]);
>> + }
>> + break;
>> + case MFI_ODCR0:
>> + s->doorbell = 0;
>> + if (s->producer_pa&& megasas_intr_enabled(s)) {
>> + /* Update reply queue pointer */
>> + stl_le_phys(s->producer_pa, s->reply_queue_head);
>> + if (!msix_enabled(&s->dev)) {
>> + qemu_irq_lower(s->dev.irq[0]);
>> + }
>> + }
>> + break;
>> + case MFI_IQPH:
>> + /* Received high 32 bits of a 64 bit MFI frame address */
>> + s->frame_hi = val;
>> + break;
>> + case MFI_IQPL:
>> + /* Received low 32 bits of a 64 bit MFI frame address */
>> + case MFI_IQP:
>> + /* Received 32 bit MFI frame address */
>> + frame_addr = (val& ~0x1F);
>> + /* Add possible 64 bit offset */
>> + frame_addr |= ((uint64_t)s->frame_hi<< 32);
>> + s->frame_hi = 0;
>> + frame_count = (val>> 1)& 0xF;
>> + megasas_handle_frame(s, frame_addr, frame_count);
>> + break;
>> + default:
>> + break;
>> + }
>> +}
>> +
>> +static const MemoryRegionOps megasas_mmio_ops = {
>> + .read = megasas_mmio_read,
>> + .write = megasas_mmio_write,
>> + .endianness = DEVICE_LITTLE_ENDIAN,
>> + .impl = {
>> + .min_access_size = 4,
>> + .max_access_size = 4,
>> + }
>> +};
>> +
>> +static uint64_t megasas_port_read(void *opaque, target_phys_addr_t addr,
>> + unsigned size)
>> +{
>> + return megasas_mmio_read(opaque, addr& 0xff, size);
>> +}
>> +
>> +static void megasas_port_write(void *opaque, target_phys_addr_t addr,
>> + uint64_t val, unsigned size)
>> +{
>> + megasas_mmio_write(opaque, addr& 0xff, val, size);
>> +}
>> +
>> +static const MemoryRegionOps megasas_port_ops = {
>> + .read = megasas_port_read,
>> + .write = megasas_port_write,
>> + .endianness = DEVICE_LITTLE_ENDIAN,
>> + .impl = {
>> + .min_access_size = 4,
>> + .max_access_size = 4,
>> + }
>> +};
>> +
>> +static uint64_t megasas_queue_read(void *opaque, target_phys_addr_t addr,
>> + unsigned size)
>> +{
>> + return 0;
>> +}
>> +
>> +static void megasas_queue_write(void *opaque, target_phys_addr_t addr,
>> + uint64_t val, unsigned size)
>> +{
>> +}
>> +
>> +static const MemoryRegionOps megasas_queue_ops = {
>> + .read = megasas_queue_read,
>> + .write = megasas_queue_write,
>> + .endianness = DEVICE_LITTLE_ENDIAN,
>> + .impl = {
>> + .min_access_size = 4,
>> + .max_access_size = 4,
>> + }
>> +};
>> +
>> +static void megasas_soft_reset(MPTState *s)
>> +{
>> + int i;
>> + struct megasas_cmd_t *cmd;
>> +
>> + for (i = 0; i< s->fw_cmds; i++) {
>> + cmd =&s->frames[i];
>> + megasas_abort_command(cmd);
>> + }
>> + megasas_reset_frames(s);
>> + s->reply_queue_len = s->fw_cmds;
>> + s->reply_queue_pa = 0;
>> + s->consumer_pa = 0;
>> + s->producer_pa = 0;
>> + s->fw_state = MFI_FWSTATE_READY;
>> + s->doorbell = 0;
>> + s->intr_mask = MEGASAS_INTR_DISABLED_MASK;
>> + s->frame_hi = 0;
>> + s->flags&= ~MEGASAS_MASK_USE_QUEUE64;
>> + s->event_count++;
>> + s->boot_event = s->event_count;
>> +}
>> +
>> +static void megasas_scsi_reset(DeviceState *dev)
>> +{
>> + MPTState *s = DO_UPCAST(MPTState, dev.qdev, dev);
>> +
>> + megasas_soft_reset(s);
>> +}
>> +
>> +static const VMStateDescription vmstate_megasas = {
>> + .name = "megasas",
>> + .version_id = 0,
>> + .minimum_version_id = 0,
>> + .minimum_version_id_old = 0,
>> + .fields = (VMStateField[]) {
>> + VMSTATE_PCI_DEVICE(dev, MPTState),
>> +
>> + VMSTATE_INT32(fw_state, MPTState),
>> + VMSTATE_INT32(intr_mask, MPTState),
>> + VMSTATE_INT32(doorbell, MPTState),
>> + VMSTATE_UINT64(reply_queue_pa, MPTState),
>> + VMSTATE_UINT64(consumer_pa, MPTState),
>> + VMSTATE_UINT64(producer_pa, MPTState),
>> + VMSTATE_END_OF_LIST()
>> + }
>> +};
>> +
>> +static int megasas_scsi_uninit(PCIDevice *d)
>> +{
>> + MPTState *s = DO_UPCAST(MPTState, dev, d);
>
> You must also uinit msix.
> It is harmless to do this uncoditionally.
>
>
>> +
>> + memory_region_destroy(&s->mmio_io);
>> + memory_region_destroy(&s->port_io);
>> + memory_region_destroy(&s->queue_io);
>> + return 0;
>> +}
>> +
>> +static const struct SCSIBusInfo megasas_scsi_info = {
>> + .tcq = true,
>> + .max_target = MFI_MAX_LD,
>> + .max_lun = 255,
>> +
>> + .transfer_data = megasas_xfer_complete,
>> + .complete = megasas_command_complete,
>> + .cancel = megasas_command_cancel,
>> +};
>> +
>> +static int megasas_scsi_init(PCIDevice *dev)
>> +{
>> + MPTState *s = DO_UPCAST(MPTState, dev, dev);
>> + uint8_t *pci_conf;
>> + int i;
>> +
>> + pci_conf = s->dev.config;
>> +
>> + /* PCI latency timer = 0 */
>> + pci_conf[PCI_LATENCY_TIMER] = 0;
>> + /* Interrupt pin 1 */
>> + pci_conf[PCI_INTERRUPT_PIN] = 0x01;
>> +
>> + memory_region_init_io(&s->mmio_io,&megasas_mmio_ops, s,
>> + "megasas-mmio", 0x4000);
>> + memory_region_init_io(&s->port_io,&megasas_port_ops, s,
>> + "megasas-io", 256);
>> + memory_region_init_io(&s->queue_io,&megasas_queue_ops, s,
>> + "megasas-queue", 0x40000);
>> +
>> + if (megasas_use_msix(s)&&
>> + msix_init(&s->dev, 15,&s->mmio_io, 0, 0x2000)) {
>> + s->flags&= ~MEGASAS_MASK_USE_MSIX;
>
> You'd want an error message here. maybe even fail init.
>
>> + }
>> +
>> + pci_register_bar(&s->dev, 0, PCI_BASE_ADDRESS_SPACE_MEMORY,&s->mmio_io);
>> + pci_register_bar(&s->dev, 2, PCI_BASE_ADDRESS_SPACE_IO,&s->port_io);
>> + pci_register_bar(&s->dev, 3, PCI_BASE_ADDRESS_SPACE_MEMORY,&s->queue_io);
>> +
>> + if (megasas_use_msix(s)) {
>> + msix_vector_use(&s->dev, 0);
>
> You can do this unconditionally. But I have a question: are you using a
> single vector? I note that you request 15 vectors from the OS.
> The vector allocator
>
>> + }
>> +
>> + if (s->fw_sge>= MEGASAS_MAX_SGE - MFI_PASS_FRAME_SIZE) {
>> + s->fw_sge = MEGASAS_MAX_SGE - MFI_PASS_FRAME_SIZE;
>> + } else if (s->fw_sge>= 128 - MFI_PASS_FRAME_SIZE) {
>> + s->fw_sge = 128 - MFI_PASS_FRAME_SIZE;
>> + } else {
>> + s->fw_sge = 64 - MFI_PASS_FRAME_SIZE;
>> + }
>> + if (s->fw_cmds> MEGASAS_MAX_FRAMES) {
>> + s->fw_cmds = MEGASAS_MAX_FRAMES;
>> + }
>> + if (s->raid_mode_str) {
>> + if (!strcmp(s->raid_mode_str, "jbod")) {
>> + s->flags |= MEGASAS_MASK_USE_JBOD;
>> + } else {
>> + s->flags&= ~MEGASAS_MASK_USE_JBOD;
>> + s->raid_mode_str = (char *)megasas_raid_modes[0];
>> + }
>
> validate mode value. Anything that is not raid or jbod
> should fail.
>
>> + } else {
>> + s->raid_mode_str = (char *)megasas_raid_modes[0];
>> + }
>
> Don't cast, use consisten styles.
>
>> + s->fw_luns = (MFI_MAX_LD> MAX_SCSI_DEVS) ?
>> + MAX_SCSI_DEVS : MFI_MAX_LD;
>> + s->producer_pa = 0;
>> + s->consumer_pa = 0;
>> + for (i = 0; i< s->fw_cmds; i++) {
>> + s->frames[i].index = i;
>> + s->frames[i].context = -1;
>> + s->frames[i].pa = 0;
>> + s->frames[i].state = s;
>> + }
>> +
>> + scsi_bus_new(&s->bus,&dev->qdev,&megasas_scsi_info);
>> + scsi_bus_legacy_handle_cmdline(&s->bus);
>> + return 0;
>> +}
>> +
>> +static Property megasas_properties[] = {
>> + DEFINE_PROP_UINT32("max_sge", MPTState, fw_sge,
>> + MEGASAS_DEFAULT_SGE),
>> + DEFINE_PROP_UINT32("max_cmds", MPTState, fw_cmds,
>> + MEGASAS_DEFAULT_FRAMES),
>> + DEFINE_PROP_STRING("mode", MPTState, raid_mode_str),
>> + DEFINE_PROP_BIT("use_msix", MPTState, flags,
>> + MEGASAS_FLAG_USE_MSIX, false),
>
> This is just a workaround for debugging, right?
> So either just remove all msix code for now,
> if 0 all msix code, or name property x-use_msix
>
>> + DEFINE_PROP_END_OF_LIST(),
>> +};
>> +
>> +static void megasas_class_init(ObjectClass *oc, void *data)
>> +{
>> + DeviceClass *dc = DEVICE_CLASS(oc);
>> + PCIDeviceClass *pc = PCI_DEVICE_CLASS(oc);
>> +
>> + pc->init = megasas_scsi_init;
>> + pc->exit = megasas_scsi_uninit;
>> + pc->vendor_id = PCI_VENDOR_ID_LSI_LOGIC;
>> + pc->device_id = PCI_DEVICE_ID_LSI_SAS1078;
>> + pc->subsystem_vendor_id = PCI_VENDOR_ID_LSI_LOGIC;
>> + pc->subsystem_id = 0x1013;
>> + pc->class_id = PCI_CLASS_STORAGE_RAID;
>> + dc->props = megasas_properties;
>> + dc->reset = megasas_scsi_reset;
>> + dc->vmsd =&vmstate_megasas;
>> + dc->desc = "LSI MegaRAID SAS 1078";
>> +}
>> +
>> +static TypeInfo megasas_info = {
>> + .name = "megasas",
>> + .parent = TYPE_PCI_DEVICE,
>> + .instance_size = sizeof(MPTState),
>> + .class_init = megasas_class_init,
>> +};
>> +
>> +static void megaraid1078_register_types(void)
>> +{
>> + type_register_static(&megasas_info);
>> +}
>> +
>> +type_init(megaraid1078_register_types);
>
>
> why not megasas_ ?
>
>> diff --git a/hw/mfi.h b/hw/mfi.h
>> new file mode 100644
>> index 0000000..4790c7c
>> --- /dev/null
>> +++ b/hw/mfi.h
>
> Sorry if this was discussed already, where is this
> code from? freebsd? it seems to have this:
> http://gitorious.org/freebsd/freebsd/blobs/HEAD/sys/dev/mfi/mfireg.h
> Want to name the file the same and add a link?
> This would be an explanation why we keep the
> file in a weird style incompatible with qemu.
>
> Still some things I think are better off fixed.
> Noted below.
>
>> @@ -0,0 +1,1281 @@
>> +/*-
>> + * Copyright (c) 2006 IronPort Systems
>> + * All rights reserved.
>> + *
>> + * Redistribution and use in source and binary forms, with or without
>> + * modification, are permitted provided that the following conditions
>> + * are met:
>> + * 1. Redistributions of source code must retain the above copyright
>> + * notice, this list of conditions and the following disclaimer.
>> + * 2. Redistributions in binary form must reproduce the above copyright
>> + * notice, this list of conditions and the following disclaimer in the
>> + * documentation and/or other materials provided with the distribution.
>> + *
>> + * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
>> + * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
>> + * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
>> + * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
>> + * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
>> + * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
>> + * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
>> + * 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 OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
>> + * SUCH DAMAGE.
>> + */
>> +/*-
>> + * Copyright (c) 2007 LSI Corp.
>> + * Copyright (c) 2007 Rajesh Prabhakaran.
>> + * All rights reserved.
>> + *
>> + * Redistribution and use in source and binary forms, with or without
>> + * modification, are permitted provided that the following conditions
>> + * are met:
>> + * 1. Redistributions of source code must retain the above copyright
>> + * notice, this list of conditions and the following disclaimer.
>> + * 2. Redistributions in binary form must reproduce the above copyright
>> + * notice, this list of conditions and the following disclaimer in the
>> + * documentation and/or other materials provided with the distribution.
>> + *
>> + * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
>> + * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
>> + * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
>> + * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
>> + * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
>> + * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
>> + * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
>> + * 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 OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
>> + * SUCH DAMAGE.
>> + */
>
> Why do we need two of these? They appear identical.
> Just combine all copyrights.
>
>> +
>> +#ifndef _MFI_H
>> +#define _MFI_H
>
> Don't start identifiers with a single _ followed
> by an upper case letter. MFI_REG_H would be a better name.
>
>> +
>> +/*
>> + * MegaRAID SAS MFI firmware definitions
>> + *
>> + * Calling this driver 'MegaRAID SAS' is a bit misleading. It's a completely
>> + * new firmware interface from the old AMI MegaRAID one, and there is no
>> + * reason why this interface should be limited to just SAS. In any case, LSI
>> + * seems to also call this interface 'MFI', so that will be used here.
>> + */
>
> *Which* driver name is misleading?
> I'd suggest droping this comment that argues with itself,
> and explaining what this header *is*. E.g. is the below right?
> /*
> MFI is a common firmware interface used by MegaRAID
> family of controllers by SAS and LSI
> */
>> +
>> +/*
>> + * Start with the register set.
>> All registers are 32 bits wide.
>> + * The usual Intel IOP style setup.
>> + */
>> +#define MFI_IMSG0 0x10 /* Inbound message 0 */
>> +#define MFI_IMSG1 0x14 /* Inbound message 1 */
>> +#define MFI_OMSG0 0x18 /* Outbound message 0 */
>> +#define MFI_OMSG1 0x1c /* Outbound message 1 */
>> +#define MFI_IDB 0x20 /* Inbound doorbell */
>> +#define MFI_ISTS 0x24 /* Inbound interrupt status */
>> +#define MFI_IMSK 0x28 /* Inbound interrupt mask */
>> +#define MFI_ODB 0x2c /* Outbound doorbell */
>> +#define MFI_OSTS 0x30 /* Outbound interrupt status */
>> +#define MFI_OMSK 0x34 /* Outbound interrupt mask */
>> +#define MFI_IQP 0x40 /* Inbound queue port */
>> +#define MFI_OQP 0x44 /* Outbound queue port */
>> +
>> +/*
>> + * 1078 specific related register
>> + */
>> +#define MFI_ODR0 0x9c /* outbound doorbell register0 */
>> +#define MFI_ODCR0 0xa0 /* outbound doorbell clear register0 */
>> +#define MFI_OSP0 0xb0 /* outbound scratch pad0 */
>> +#define MFI_IQPL 0xc0 /* Inbound queue port (low bytes) */
>> +#define MFI_IQPH 0xc4 /* Inbound queue port (high bytes) */
>> +#define MFI_DIAG 0xf8 /* Host diag */
>> +#define MFI_SEQ 0xfc /* Sequencer offset */
>> +#define MFI_1078_EIM 0x80000004 /* 1078 enable intrrupt mask */
>> +#define MFI_RMI 0x2 /* reply message interrupt */
>> +#define MFI_1078_RM 0x80000000 /* reply 1078 message interrupt */
>> +#define MFI_ODC 0x4 /* outbound doorbell change interrupt */
>> +
>> +/*
>> + * gen2 specific changes
>> + */
>> +#define MFI_GEN2_EIM 0x00000005 /* gen2 enable interrupt mask */
>> +#define MFI_GEN2_RM 0x00000001 /* reply gen2 message interrupt */
>> +
>> +/*
>> + * skinny specific changes
>> + */
>> +#define MFI_SKINNY_IDB 0x00 /* Inbound doorbell is at 0x00 for skinny */
>> +#define MFI_SKINNY_RM 0x00000001 /* reply skinny message interrupt */
>> +
>> +/* Bits for MFI_OSTS */
>> +#define MFI_OSTS_INTR_VALID 0x00000002
>> +
>> +/*
>> + * Firmware state values. Found in OMSG0 during initialization.
>> + */
>> +#define MFI_FWSTATE_MASK 0xf0000000
>> +#define MFI_FWSTATE_UNDEFINED 0x00000000
>> +#define MFI_FWSTATE_BB_INIT 0x10000000
>> +#define MFI_FWSTATE_FW_INIT 0x40000000
>> +#define MFI_FWSTATE_WAIT_HANDSHAKE 0x60000000
>> +#define MFI_FWSTATE_FW_INIT_2 0x70000000
>> +#define MFI_FWSTATE_DEVICE_SCAN 0x80000000
>> +#define MFI_FWSTATE_BOOT_MSG_PENDING 0x90000000
>> +#define MFI_FWSTATE_FLUSH_CACHE 0xa0000000
>> +#define MFI_FWSTATE_READY 0xb0000000
>> +#define MFI_FWSTATE_OPERATIONAL 0xc0000000
>> +#define MFI_FWSTATE_FAULT 0xf0000000
>> +#define MFI_FWSTATE_MAXSGL_MASK 0x00ff0000
>> +#define MFI_FWSTATE_MAXCMD_MASK 0x0000ffff
>> +#define MFI_FWSTATE_MSIX_SUPPORTED 0x04000000
>> +#define MFI_FWSTATE_HOSTMEMREQD_MASK 0x08000000
>> +
>> +/*
>> + * Control bits to drive the card to ready state. These go into the IDB
>> + * register.
>> + */
>> +#define MFI_FWINIT_ABORT 0x00000001 /* Abort all pending commands */
>> +#define MFI_FWINIT_READY 0x00000002 /* Move from operational to ready */
>> +#define MFI_FWINIT_MFIMODE 0x00000004 /* unknown */
>> +#define MFI_FWINIT_CLEAR_HANDSHAKE 0x00000008 /* Respond to WAIT_HANDSHAKE */
>> +#define MFI_FWINIT_HOTPLUG 0x00000010
>> +#define MFI_FWINIT_STOP_ADP 0x00000020 /* Move to operational, stop */
>> +#define MFI_FWINIT_ADP_RESET 0x00000040 /* Reset ADP */
>> +
>> +/* MFI Commands */
>> +typedef enum {
>> + MFI_CMD_INIT = 0x00,
>> + MFI_CMD_LD_READ,
>> + MFI_CMD_LD_WRITE,
>> + MFI_CMD_LD_SCSI_IO,
>> + MFI_CMD_PD_SCSI_IO,
>> + MFI_CMD_DCMD,
>> + MFI_CMD_ABORT,
>> + MFI_CMD_SMP,
>> + MFI_CMD_STP
>> +} mfi_cmd_t ;
>
> space before ;, here and elsewhere.
>
>
>> +
>> +/* Direct commands */
>> +typedef enum {
>> + MFI_DCMD_CTRL_MFI_HOST_MEM_ALLOC = 0x0100e100,
>> + MFI_DCMD_CTRL_GET_INFO = 0x01010000,
>> + MFI_DCMD_CTRL_GET_PROPERTIES = 0x01020100,
>> + MFI_DCMD_CTRL_SET_PROPERTIES = 0x01020200,
>> + MFI_DCMD_CTRL_ALARM = 0x01030000,
>> + MFI_DCMD_CTRL_ALARM_GET = 0x01030100,
>> + MFI_DCMD_CTRL_ALARM_ENABLE = 0x01030200,
>> + MFI_DCMD_CTRL_ALARM_DISABLE = 0x01030300,
>> + MFI_DCMD_CTRL_ALARM_SILENCE = 0x01030400,
>> + MFI_DCMD_CTRL_ALARM_TEST = 0x01030500,
>> + MFI_DCMD_CTRL_EVENT_GETINFO = 0x01040100,
>> + MFI_DCMD_CTRL_EVENT_CLEAR = 0x01040200,
>> + MFI_DCMD_CTRL_EVENT_GET = 0x01040300,
>> + MFI_DCMD_CTRL_EVENT_COUNT = 0x01040400,
>> + MFI_DCMD_CTRL_EVENT_WAIT = 0x01040500,
>> + MFI_DCMD_CTRL_SHUTDOWN = 0x01050000,
>> + MFI_DCMD_HIBERNATE_STANDBY = 0x01060000,
>> + MFI_DCMD_CTRL_GET_TIME = 0x01080101,
>> + MFI_DCMD_CTRL_SET_TIME = 0x01080102,
>> + MFI_DCMD_CTRL_BIOS_DATA_GET = 0x010c0100,
>> + MFI_DCMD_CTRL_BIOS_DATA_SET = 0x010c0200,
>> + MFI_DCMD_CTRL_FACTORY_DEFAULTS = 0x010d0000,
>> + MFI_DCMD_CTRL_MFC_DEFAULTS_GET = 0x010e0201,
>> + MFI_DCMD_CTRL_MFC_DEFAULTS_SET = 0x010e0202,
>> + MFI_DCMD_CTRL_CACHE_FLUSH = 0x01101000,
>> + MFI_DCMD_PD_GET_LIST = 0x02010000,
>> + MFI_DCMD_PD_LIST_QUERY = 0x02010100,
>> + MFI_DCMD_PD_GET_INFO = 0x02020000,
>> + MFI_DCMD_PD_STATE_SET = 0x02030100,
>> + MFI_DCMD_PD_REBUILD = 0x02040100,
>> + MFI_DCMD_PD_BLINK = 0x02070100,
>> + MFI_DCMD_PD_UNBLINK = 0x02070200,
>> + MFI_DCMD_LD_GET_LIST = 0x03010000,
>> + MFI_DCMD_LD_GET_INFO = 0x03020000,
>> + MFI_DCMD_LD_GET_PROP = 0x03030000,
>> + MFI_DCMD_LD_SET_PROP = 0x03040000,
>> + MFI_DCMD_LD_DELETE = 0x03090000,
>> + MFI_DCMD_CFG_READ = 0x04010000,
>> + MFI_DCMD_CFG_ADD = 0x04020000,
>> + MFI_DCMD_CFG_CLEAR = 0x04030000,
>> + MFI_DCMD_CFG_FOREIGN_READ = 0x04060100,
>> + MFI_DCMD_CFG_FOREIGN_IMPORT = 0x04060400,
>> + MFI_DCMD_BBU_STATUS = 0x05010000,
>> + MFI_DCMD_BBU_CAPACITY_INFO = 0x05020000,
>> + MFI_DCMD_BBU_DESIGN_INFO = 0x05030000,
>> + MFI_DCMD_BBU_PROP_GET = 0x05050100,
>> + MFI_DCMD_CLUSTER = 0x08000000,
>> + MFI_DCMD_CLUSTER_RESET_ALL = 0x08010100,
>> + MFI_DCMD_CLUSTER_RESET_LD = 0x08010200
>> +} mfi_dcmd_t ;
>
>
> space before ;
>
>> +
>> +/* Modifiers for MFI_DCMD_CTRL_FLUSHCACHE */
>> +#define MFI_FLUSHCACHE_CTRL 0x01
>> +#define MFI_FLUSHCACHE_DISK 0x02
>> +
>> +/* Modifiers for MFI_DCMD_CTRL_SHUTDOWN */
>> +#define MFI_SHUTDOWN_SPINDOWN 0x01
>> +
>> +/*
>> + * MFI Frame flags
>> + */
>> +typedef enum {
>> + MFI_FRAME_DONT_POST_IN_REPLY_QUEUE = 0x0001,
>> + MFI_FRAME_SGL64 = 0x0002,
>> + MFI_FRAME_SENSE64 = 0x0004,
>> + MFI_FRAME_DIR_WRITE = 0x0008,
>> + MFI_FRAME_DIR_READ = 0x0010,
>> + MFI_FRAME_IEEE_SGL = 0x0020,
>> +} mfi_frame_flags;
>> +
>> +/* MFI Status codes */
>> +typedef enum {
>> + MFI_STAT_OK = 0x00,
>> + MFI_STAT_INVALID_CMD,
>> + MFI_STAT_INVALID_DCMD,
>> + MFI_STAT_INVALID_PARAMETER,
>> + MFI_STAT_INVALID_SEQUENCE_NUMBER,
>> + MFI_STAT_ABORT_NOT_POSSIBLE,
>> + MFI_STAT_APP_HOST_CODE_NOT_FOUND,
>> + MFI_STAT_APP_IN_USE,
>> + MFI_STAT_APP_NOT_INITIALIZED,
>> + MFI_STAT_ARRAY_INDEX_INVALID,
>> + MFI_STAT_ARRAY_ROW_NOT_EMPTY,
>> + MFI_STAT_CONFIG_RESOURCE_CONFLICT,
>> + MFI_STAT_DEVICE_NOT_FOUND,
>> + MFI_STAT_DRIVE_TOO_SMALL,
>> + MFI_STAT_FLASH_ALLOC_FAIL,
>> + MFI_STAT_FLASH_BUSY,
>> + MFI_STAT_FLASH_ERROR = 0x10,
>> + MFI_STAT_FLASH_IMAGE_BAD,
>> + MFI_STAT_FLASH_IMAGE_INCOMPLETE,
>> + MFI_STAT_FLASH_NOT_OPEN,
>> + MFI_STAT_FLASH_NOT_STARTED,
>> + MFI_STAT_FLUSH_FAILED,
>> + MFI_STAT_HOST_CODE_NOT_FOUNT,
>> + MFI_STAT_LD_CC_IN_PROGRESS,
>> + MFI_STAT_LD_INIT_IN_PROGRESS,
>> + MFI_STAT_LD_LBA_OUT_OF_RANGE,
>> + MFI_STAT_LD_MAX_CONFIGURED,
>> + MFI_STAT_LD_NOT_OPTIMAL,
>> + MFI_STAT_LD_RBLD_IN_PROGRESS,
>> + MFI_STAT_LD_RECON_IN_PROGRESS,
>> + MFI_STAT_LD_WRONG_RAID_LEVEL,
>> + MFI_STAT_MAX_SPARES_EXCEEDED,
>> + MFI_STAT_MEMORY_NOT_AVAILABLE = 0x20,
>> + MFI_STAT_MFC_HW_ERROR,
>> + MFI_STAT_NO_HW_PRESENT,
>> + MFI_STAT_NOT_FOUND,
>> + MFI_STAT_NOT_IN_ENCL,
>> + MFI_STAT_PD_CLEAR_IN_PROGRESS,
>> + MFI_STAT_PD_TYPE_WRONG,
>> + MFI_STAT_PR_DISABLED,
>> + MFI_STAT_ROW_INDEX_INVALID,
>> + MFI_STAT_SAS_CONFIG_INVALID_ACTION,
>> + MFI_STAT_SAS_CONFIG_INVALID_DATA,
>> + MFI_STAT_SAS_CONFIG_INVALID_PAGE,
>> + MFI_STAT_SAS_CONFIG_INVALID_TYPE,
>> + MFI_STAT_SCSI_DONE_WITH_ERROR,
>> + MFI_STAT_SCSI_IO_FAILED,
>> + MFI_STAT_SCSI_RESERVATION_CONFLICT,
>> + MFI_STAT_SHUTDOWN_FAILED = 0x30,
>> + MFI_STAT_TIME_NOT_SET,
>> + MFI_STAT_WRONG_STATE,
>> + MFI_STAT_LD_OFFLINE,
>> + MFI_STAT_PEER_NOTIFICATION_REJECTED,
>> + MFI_STAT_PEER_NOTIFICATION_FAILED,
>> + MFI_STAT_RESERVATION_IN_PROGRESS,
>> + MFI_STAT_I2C_ERRORS_DETECTED,
>> + MFI_STAT_PCI_ERRORS_DETECTED,
>> + MFI_STAT_DIAG_FAILED,
>> + MFI_STAT_BOOT_MSG_PENDING,
>> + MFI_STAT_FOREIGN_CONFIG_INCOMPLETE,
>> + MFI_STAT_INVALID_SGL,
>> + MFI_STAT_UNSUPPORTED_HW,
>> + MFI_STAT_CC_SCHEDULE_DISABLED,
>> + MFI_STAT_PD_COPYBACK_IN_PROGRESS,
>> + MFI_STAT_MULTIPLE_PDS_IN_ARRAY = 0x40,
>> + MFI_STAT_FW_DOWNLOAD_ERROR,
>> + MFI_STAT_FEATURE_SECURITY_NOT_ENABLED,
>> + MFI_STAT_LOCK_KEY_ALREADY_EXISTS,
>> + MFI_STAT_LOCK_KEY_BACKUP_NOT_ALLOWED,
>> + MFI_STAT_LOCK_KEY_VERIFY_NOT_ALLOWED,
>> + MFI_STAT_LOCK_KEY_VERIFY_FAILED,
>> + MFI_STAT_LOCK_KEY_REKEY_NOT_ALLOWED,
>> + MFI_STAT_LOCK_KEY_INVALID,
>> + MFI_STAT_LOCK_KEY_ESCROW_INVALID,
>> + MFI_STAT_LOCK_KEY_BACKUP_REQUIRED,
>> + MFI_STAT_SECURE_LD_EXISTS,
>> + MFI_STAT_LD_SECURE_NOT_ALLOWED,
>> + MFI_STAT_REPROVISION_NOT_ALLOWED,
>> + MFI_STAT_PD_SECURITY_TYPE_WRONG,
>> + MFI_STAT_LD_ENCRYPTION_TYPE_INVALID,
>> + MFI_STAT_CONFIG_FDE_NON_FDE_MIX_NOT_ALLOWED = 0x50,
>> + MFI_STAT_CONFIG_LD_ENCRYPTION_TYPE_MIX_NOT_ALLOWED,
>> + MFI_STAT_SECRET_KEY_NOT_ALLOWED,
>> + MFI_STAT_PD_HW_ERRORS_DETECTED,
>> + MFI_STAT_LD_CACHE_PINNED,
>> + MFI_STAT_POWER_STATE_SET_IN_PROGRESS,
>> + MFI_STAT_POWER_STATE_SET_BUSY,
>> + MFI_STAT_POWER_STATE_WRONG,
>> + MFI_STAT_PR_NO_AVAILABLE_PD_FOUND,
>> + MFI_STAT_CTRL_RESET_REQUIRED,
>> + MFI_STAT_LOCK_KEY_EKM_NO_BOOT_AGENT,
>> + MFI_STAT_SNAP_NO_SPACE,
>> + MFI_STAT_SNAP_PARTIAL_FAILURE,
>> + MFI_STAT_UPGRADE_KEY_INCOMPATIBLE,
>> + MFI_STAT_PFK_INCOMPATIBLE,
>> + MFI_STAT_PD_MAX_UNCONFIGURED,
>> + MFI_STAT_IO_METRICS_DISABLED = 0x60,
>> + MFI_STAT_AEC_NOT_STOPPED,
>> + MFI_STAT_PI_TYPE_WRONG,
>> + MFI_STAT_LD_PD_PI_INCOMPATIBLE,
>> + MFI_STAT_PI_NOT_ENABLED,
>> + MFI_STAT_LD_BLOCK_SIZE_MISMATCH,
>> + MFI_STAT_INVALID_STATUS = 0xFF
>> +} mfi_status_t ;
>> +
>> +/* Event classes */
>> +typedef enum {
>> + MFI_EVT_CLASS_DEBUG = -2,
>> + MFI_EVT_CLASS_PROGRESS = -1,
>> + MFI_EVT_CLASS_INFO = 0,
>> + MFI_EVT_CLASS_WARNING = 1,
>> + MFI_EVT_CLASS_CRITICAL = 2,
>> + MFI_EVT_CLASS_FATAL = 3,
>> + MFI_EVT_CLASS_DEAD = 4
>> +} mfi_evt_class_t ;
>> +
>> +/* Event locales */
>> +typedef enum {
>> + MFI_EVT_LOCALE_LD = 0x0001,
>> + MFI_EVT_LOCALE_PD = 0x0002,
>> + MFI_EVT_LOCALE_ENCL = 0x0004,
>> + MFI_EVT_LOCALE_BBU = 0x0008,
>> + MFI_EVT_LOCALE_SAS = 0x0010,
>> + MFI_EVT_LOCALE_CTRL = 0x0020,
>> + MFI_EVT_LOCALE_CONFIG = 0x0040,
>> + MFI_EVT_LOCALE_CLUSTER = 0x0080,
>> + MFI_EVT_LOCALE_ALL = 0xffff
>> +} mfi_evt_locale_t;
>> +
>> +/* Event args */
>> +typedef enum {
>> + MR_EVT_ARGS_NONE = 0x00,
>> + MR_EVT_ARGS_CDB_SENSE,
>> + MR_EVT_ARGS_LD,
>> + MR_EVT_ARGS_LD_COUNT,
>> + MR_EVT_ARGS_LD_LBA,
>> + MR_EVT_ARGS_LD_OWNER,
>> + MR_EVT_ARGS_LD_LBA_PD_LBA,
>> + MR_EVT_ARGS_LD_PROG,
>> + MR_EVT_ARGS_LD_STATE,
>> + MR_EVT_ARGS_LD_STRIP,
>> + MR_EVT_ARGS_PD,
>> + MR_EVT_ARGS_PD_ERR,
>> + MR_EVT_ARGS_PD_LBA,
>> + MR_EVT_ARGS_PD_LBA_LD,
>> + MR_EVT_ARGS_PD_PROG,
>> + MR_EVT_ARGS_PD_STATE,
>> + MR_EVT_ARGS_PCI,
>> + MR_EVT_ARGS_RATE,
>> + MR_EVT_ARGS_STR,
>> + MR_EVT_ARGS_TIME,
>> + MR_EVT_ARGS_ECC,
>> + MR_EVT_ARGS_LD_PROP,
>> + MR_EVT_ARGS_PD_SPARE,
>> + MR_EVT_ARGS_PD_INDEX,
>> + MR_EVT_ARGS_DIAG_PASS,
>> + MR_EVT_ARGS_DIAG_FAIL,
>> + MR_EVT_ARGS_PD_LBA_LBA,
>> + MR_EVT_ARGS_PORT_PHY,
>> + MR_EVT_ARGS_PD_MISSING,
>> + MR_EVT_ARGS_PD_ADDRESS,
>> + MR_EVT_ARGS_BITMAP,
>> + MR_EVT_ARGS_CONNECTOR,
>> + MR_EVT_ARGS_PD_PD,
>> + MR_EVT_ARGS_PD_FRU,
>> + MR_EVT_ARGS_PD_PATHINFO,
>> + MR_EVT_ARGS_PD_POWER_STATE,
>> + MR_EVT_ARGS_GENERIC,
>> +} mfi_evt_args;
>> +
>> +/* Event codes */
>> +#define MR_EVT_CFG_CLEARED 0x0004
>> +#define MR_EVT_CTRL_SHUTDOWN 0x002a
>> +#define MR_EVT_LD_STATE_CHANGE 0x0051
>> +#define MR_EVT_PD_INSERTED 0x005b
>> +#define MR_EVT_PD_REMOVED 0x0070
>> +#define MR_EVT_PD_STATE_CHANGED 0x0072
>> +#define MR_EVT_LD_CREATED 0x008a
>> +#define MR_EVT_LD_DELETED 0x008b
>> +#define MR_EVT_FOREIGN_CFG_IMPORTED 0x00db
>> +#define MR_EVT_LD_OFFLINE 0x00fc
>> +#define MR_EVT_CTRL_HOST_BUS_SCAN_REQUESTED 0x0152
>> +
>> +typedef enum {
>> + MR_LD_CACHE_WRITE_BACK = 0x01,
>> + MR_LD_CACHE_WRITE_ADAPTIVE = 0x02,
>> + MR_LD_CACHE_READ_AHEAD = 0x04,
>> + MR_LD_CACHE_READ_ADAPTIVE = 0x08,
>> + MR_LD_CACHE_WRITE_CACHE_BAD_BBU = 0x10,
>> + MR_LD_CACHE_ALLOW_WRITE_CACHE = 0x20,
>> + MR_LD_CACHE_ALLOW_READ_CACHE = 0x40
>> +} mfi_ld_cache;
>> +
>> +typedef enum {
>> + MR_PD_CACHE_UNCHANGED = 0,
>> + MR_PD_CACHE_ENABLE = 1,
>> + MR_PD_CACHE_DISABLE = 2
>> +} mfi_pd_cache;
>> +
>> +typedef enum {
>> + MR_PD_QUERY_TYPE_ALL = 0,
>> + MR_PD_QUERY_TYPE_STATE = 1,
>> + MR_PD_QUERY_TYPE_POWER_STATE = 2,
>> + MR_PD_QUERY_TYPE_MEDIA_TYPE = 3,
>> + MR_PD_QUERY_TYPE_SPEED = 4,
>> + MR_PD_QUERY_TYPE_EXPOSED_TO_HOST = 5, /*query for system drives */
>> +} mfi_pd_query_type;
>> +
>> +/*
>> + * Other propertities and definitions
>> + */
>> +#define MFI_MAX_PD_CHANNELS 2
>> +#define MFI_MAX_LD_CHANNELS 2
>> +#define MFI_MAX_CHANNELS (MFI_MAX_PD_CHANNELS + MFI_MAX_LD_CHANNELS)
>> +#define MFI_MAX_CHANNEL_DEVS 128
>> +#define MFI_DEFAULT_ID -1
>> +#define MFI_MAX_LUN 8
>> +#define MFI_MAX_LD 64
>> +
>> +#define MFI_FRAME_SIZE 64
>> +#define MFI_MBOX_SIZE 12
>> +
>> +/* Firmware flashing can take 40s */
>> +#define MFI_POLL_TIMEOUT_SECS 50
>> +
>> +/* Allow for speedier math calculations */
>> +#define MFI_SECTOR_LEN 512
>> +
>> +/* Scatter Gather elements */
>> +struct mfi_sg32 {
>> + uint32_t addr;
>> + uint32_t len;
>> +} __attribute__ ((packed));
>> +
>> +struct mfi_sg64 {
>> + uint64_t addr;
>> + uint32_t len;
>> +} __attribute__ ((packed));
>> +
>> +struct mfi_sg_skinny {
>> + uint64_t addr;
>> + uint32_t len;
>> + uint32_t flag;
>> +} __attribute__ ((packed));
>> +
>> +union mfi_sgl {
>> + struct mfi_sg32 sg32[1];
>> + struct mfi_sg64 sg64[1];
>> + struct mfi_sg_skinny sg_skinny[1];
>> +} __attribute__ ((packed));
>> +
>> +/* Message frames. All messages have a common header */
>> +struct mfi_frame_header {
>> + uint8_t frame_cmd;
>> + uint8_t sense_len;
>> + uint8_t cmd_status;
>> + uint8_t scsi_status;
>> + uint8_t target_id;
>> + uint8_t lun_id;
>> + uint8_t cdb_len;
>> + uint8_t sge_count;
>> + uint64_t context;
>> + uint16_t flags;
>> + uint16_t timeout;
>> + uint32_t data_len;
>> +} __attribute__ ((packed));
>> +
>> +struct mfi_init_frame {
>> + struct mfi_frame_header header;
>> + uint32_t qinfo_new_addr_lo;
>> + uint32_t qinfo_new_addr_hi;
>> + uint32_t qinfo_old_addr_lo;
>> + uint32_t qinfo_old_addr_hi;
>> + uint32_t reserved[6];
>> +} __attribute__ ((packed));
>> +
>> +#define MFI_IO_FRAME_SIZE 40
>> +struct mfi_io_frame {
>> + struct mfi_frame_header header;
>> + uint32_t sense_addr_lo;
>> + uint32_t sense_addr_hi;
>> + uint32_t lba_lo;
>> + uint32_t lba_hi;
>> + union mfi_sgl sgl;
>> +} __attribute__ ((packed));
>> +
>> +#define MFI_PASS_FRAME_SIZE 48
>> +struct mfi_pass_frame {
>> + struct mfi_frame_header header;
>> + uint32_t sense_addr_lo;
>> + uint32_t sense_addr_hi;
>> + uint8_t cdb[16];
>> + union mfi_sgl sgl;
>> +} __attribute__ ((packed));
>> +
>> +#define MFI_DCMD_FRAME_SIZE 40
>> +struct mfi_dcmd_frame {
>> + struct mfi_frame_header header;
>> + uint32_t opcode;
>> + uint8_t mbox[MFI_MBOX_SIZE];
>> + union mfi_sgl sgl;
>> +} __attribute__ ((packed));
>> +
>> +struct mfi_abort_frame {
>> + struct mfi_frame_header header;
>> + uint64_t abort_context;
>> + uint32_t abort_mfi_addr_lo;
>> + uint32_t abort_mfi_addr_hi;
>> + uint32_t reserved1[6];
>> +} __attribute__ ((packed));
>> +
>> +struct mfi_smp_frame {
>> + struct mfi_frame_header header;
>> + uint64_t sas_addr;
>> + union {
>> + struct mfi_sg32 sg32[2];
>> + struct mfi_sg64 sg64[2];
>> + } sgl;
>> +} __attribute__ ((packed));
>> +
>> +struct mfi_stp_frame {
>> + struct mfi_frame_header header;
>> + uint16_t fis[10];
>> + uint32_t stp_flags;
>> + union {
>> + struct mfi_sg32 sg32[2];
>> + struct mfi_sg64 sg64[2];
>> + } sgl;
>> +} __attribute__ ((packed));
>> +
>> +union mfi_frame {
>> + struct mfi_frame_header header;
>> + struct mfi_init_frame init;
>> + struct mfi_io_frame io;
>> + struct mfi_pass_frame pass;
>> + struct mfi_dcmd_frame dcmd;
>> + struct mfi_abort_frame abort;
>> + struct mfi_smp_frame smp;
>> + struct mfi_stp_frame stp;
>> + uint64_t raw[8];
>> + uint8_t bytes[MFI_FRAME_SIZE];
>> +};
>> +
>> +#define MFI_SENSE_LEN 128
>> +struct mfi_sense {
>> + uint8_t data[MFI_SENSE_LEN];
>> +};
>> +
>> +#define MFI_QUEUE_FLAG_CONTEXT64 0x00000002
>> +
>> +/* The queue init structure that is passed with the init message */
>> +struct mfi_init_qinfo {
>> + uint32_t flags;
>> + uint32_t rq_entries;
>> + uint32_t rq_addr_lo;
>> + uint32_t rq_addr_hi;
>> + uint32_t pi_addr_lo;
>> + uint32_t pi_addr_hi;
>> + uint32_t ci_addr_lo;
>> + uint32_t ci_addr_hi;
>> +} __attribute__ ((packed));
>
> Don't use a packed attribute - it's not necessary here and creates
> much worse code as gcc must then assume unaligned struct address.
> You had to tweak the attribute anyway, just drop it.
>
>> +
>> +/* Controller properties */
>> +struct mfi_ctrl_props {
>> + uint16_t seq_num;
>> + uint16_t pred_fail_poll_interval;
>> + uint16_t intr_throttle_cnt;
>> + uint16_t intr_throttle_timeout;
>> + uint8_t rebuild_rate;
>> + uint8_t patrol_read_rate;
>> + uint8_t bgi_rate;
>> + uint8_t cc_rate;
>> + uint8_t recon_rate;
>> + uint8_t cache_flush_interval;
>> + uint8_t spinup_drv_cnt;
>> + uint8_t spinup_delay;
>> + uint8_t cluster_enable;
>> + uint8_t coercion_mode;
>> + uint8_t alarm_enable;
>> + uint8_t disable_auto_rebuild;
>> + uint8_t disable_battery_warn;
>> + uint8_t ecc_bucket_size;
>> + uint16_t ecc_bucket_leak_rate;
>> + uint8_t restore_hotspare_on_insertion;
>> + uint8_t expose_encl_devices;
>> + uint8_t maintainPdFailHistory;
>> + uint8_t disallowHostRequestReordering;
>> + uint8_t abortCCOnError;
>> + uint8_t loadBalanceMode;
>> + uint8_t disableAutoDetectBackplane;
>> + uint8_t snapVDSpace;
>> + struct {
>> + /* set TRUE to disable copyBack (0=copyback enabled) */
>> + uint32_t copyBackDisabled:1,
>> + SMARTerEnabled:1,
>> + prCorrectUnconfiguredAreas:1,
>> + useFdeOnly:1,
>> + disableNCQ:1,
>> + SSDSMARTerEnabled:1,
>> + SSDPatrolReadEnabled:1,
>> + enableSpinDownUnconfigured:1,
>> + autoEnhancedImport:1,
>> + enableSecretKeyControl:1,
>> + disableOnlineCtrlReset:1,
>> + allowBootWithPinnedCache:1,
>> + disableSpinDownHS:1,
>> + enableJBOD:1,
>> + reserved:18;
>> + } OnOffProperties;
>
> Using bitfields for anything where you care about endian-ness
> is IMO wrong, and you normally do care for BE host + LE guest.
> No idea what bsd does to handle this.
>
>> + uint8_t autoSnapVDSpace; /* % of source LD to be
>> + * reserved for auto snapshot
>> + * in snapshot repository, for
>> + * metadata and user data
>> + * 1=5%, 2=10%, 3=15% and so on
>> + */
>> + uint8_t viewSpace; /* snapshot writeable VIEWs
>> + * capacity as a % of source LD
>> + * capacity. 0=READ only
>> + * 1=5%, 2=10%, 3=15% and so on
>> + */
>> + uint16_t spinDownTime; /* # of idle minutes before device
>> + * is spun down (0=use FW defaults)
>> + */
>> + uint8_t reserved[24];
>> +} __attribute__ ((packed));
>> +
>> +/* PCI information about the card. */
>> +struct mfi_info_pci {
>> + uint16_t vendor;
>> + uint16_t device;
>> + uint16_t subvendor;
>> + uint16_t subdevice;
>> + uint8_t reserved[24];
>> +} __attribute__ ((packed));
>> +
>> +/* Host (front end) interface information */
>> +struct mfi_info_host {
>> + uint8_t type;
>> +#define MFI_INFO_HOST_PCIX 0x01
>> +#define MFI_INFO_HOST_PCIE 0x02
>> +#define MFI_INFO_HOST_ISCSI 0x04
>> +#define MFI_INFO_HOST_SAS3G 0x08
>> + uint8_t reserved[6];
>> + uint8_t port_count;
>> + uint64_t port_addr[8];
>> +} __attribute__ ((packed));
>> +
>> +/* Device (back end) interface information */
>> +struct mfi_info_device {
>> + uint8_t type;
>> +#define MFI_INFO_DEV_SPI 0x01
>> +#define MFI_INFO_DEV_SAS3G 0x02
>> +#define MFI_INFO_DEV_SATA1 0x04
>> +#define MFI_INFO_DEV_SATA3G 0x08
>> + uint8_t reserved[6];
>> + uint8_t port_count;
>> + uint64_t port_addr[8];
>> +} __attribute__ ((packed));
>> +
>> +/* Firmware component information */
>> +struct mfi_info_component {
>> + char name[8];
>> + char version[32];
>> + char build_date[16];
>> + char build_time[16];
>> +} __attribute__ ((packed));
>> +
>> +/* Controller default settings */
>> +struct mfi_defaults {
>> + uint64_t sas_addr;
>> + uint8_t phy_polarity;
>> + uint8_t background_rate;
>> + uint8_t stripe_size;
>> + uint8_t flush_time;
>> + uint8_t write_back;
>> + uint8_t read_ahead;
>> + uint8_t cache_when_bbu_bad;
>> + uint8_t cached_io;
>> + uint8_t smart_mode;
>> + uint8_t alarm_disable;
>> + uint8_t coercion;
>> + uint8_t zrc_config;
>> + uint8_t dirty_led_shows_drive_activity;
>> + uint8_t bios_continue_on_error;
>> + uint8_t spindown_mode;
>> + uint8_t allowed_device_types;
>> + uint8_t allow_mix_in_enclosure;
>> + uint8_t allow_mix_in_ld;
>> + uint8_t allow_sata_in_cluster;
>> + uint8_t max_chained_enclosures;
>> + uint8_t disable_ctrl_r;
>> + uint8_t enable_web_bios;
>> + uint8_t phy_polarity_split;
>> + uint8_t direct_pd_mapping;
>> + uint8_t bios_enumerate_lds;
>> + uint8_t restored_hot_spare_on_insertion;
>> + uint8_t expose_enclosure_devices;
>> + uint8_t maintain_pd_fail_history;
>> + uint8_t disable_puncture;
>> + uint8_t zero_based_enumeration;
>> + uint8_t disable_preboot_cli;
>> + uint8_t show_drive_led_on_activity;
>> + uint8_t cluster_disable;
>> + uint8_t sas_disable;
>> + uint8_t auto_detect_backplane;
>> + uint8_t fde_only;
>> + uint8_t delay_during_post;
>> + uint8_t resv[19];
>> +} __attribute__ ((packed));
>> +
>> +/* Controller default settings */
>> +struct mfi_bios_data {
>> + uint16_t boot_target_id;
>> + uint8_t do_not_int_13;
>> + uint8_t continue_on_error;
>> + uint8_t verbose;
>> + uint8_t geometry;
>> + uint8_t expose_all_drives;
>> + uint8_t reserved[56];
>> + uint8_t check_sum;
>> +} __attribute__ ((packed));
>> +
>> +/* SAS (?) controller info, returned from MFI_DCMD_CTRL_GETINFO. */
>> +struct mfi_ctrl_info {
>> + struct mfi_info_pci pci;
>> + struct mfi_info_host host;
>> + struct mfi_info_device device;
>> +
>> + /* Firmware components that are present and active. */
>> + uint32_t image_check_word;
>> + uint32_t image_component_count;
>> + struct mfi_info_component image_component[8];
>> +
>> + /* Firmware components that have been flashed but are inactive */
>> + uint32_t pending_image_component_count;
>> + struct mfi_info_component pending_image_component[8];
>> +
>> + uint8_t max_arms;
>> + uint8_t max_spans;
>> + uint8_t max_arrays;
>> + uint8_t max_lds;
>> + char product_name[80];
>> + char serial_number[32];
>> + uint32_t hw_present;
>> +#define MFI_INFO_HW_BBU 0x01
>> +#define MFI_INFO_HW_ALARM 0x02
>> +#define MFI_INFO_HW_NVRAM 0x04
>> +#define MFI_INFO_HW_UART 0x08
>> +#define MFI_INFO_HW_MEM 0x10
>> +#define MFI_INFO_HW_FLASH 0x20
>> + uint32_t current_fw_time;
>> + uint16_t max_cmds;
>> + uint16_t max_sg_elements;
>> + uint32_t max_request_size;
>> + uint16_t lds_present;
>> + uint16_t lds_degraded;
>> + uint16_t lds_offline;
>> + uint16_t pd_present;
>> + uint16_t pd_disks_present;
>> + uint16_t pd_disks_pred_failure;
>> + uint16_t pd_disks_failed;
>> + uint16_t nvram_size;
>> + uint16_t memory_size;
>> + uint16_t flash_size;
>> + uint16_t ram_correctable_errors;
>> + uint16_t ram_uncorrectable_errors;
>> + uint8_t cluster_allowed;
>> + uint8_t cluster_active;
>> + uint16_t max_strips_per_io;
>> +
>> + uint32_t raid_levels;
>> +#define MFI_INFO_RAID_0 0x01
>> +#define MFI_INFO_RAID_1 0x02
>> +#define MFI_INFO_RAID_5 0x04
>> +#define MFI_INFO_RAID_1E 0x08
>> +#define MFI_INFO_RAID_6 0x10
>> +
>> + uint32_t adapter_ops;
>> +#define MFI_INFO_AOPS_RBLD_RATE 0x0001
>> +#define MFI_INFO_AOPS_CC_RATE 0x0002
>> +#define MFI_INFO_AOPS_BGI_RATE 0x0004
>> +#define MFI_INFO_AOPS_RECON_RATE 0x0008
>> +#define MFI_INFO_AOPS_PATROL_RATE 0x0010
>> +#define MFI_INFO_AOPS_ALARM_CONTROL 0x0020
>> +#define MFI_INFO_AOPS_CLUSTER_SUPPORTED 0x0040
>> +#define MFI_INFO_AOPS_BBU 0x0080
>> +#define MFI_INFO_AOPS_SPANNING_ALLOWED 0x0100
>> +#define MFI_INFO_AOPS_DEDICATED_SPARES 0x0200
>> +#define MFI_INFO_AOPS_REVERTIBLE_SPARES 0x0400
>> +#define MFI_INFO_AOPS_FOREIGN_IMPORT 0x0800
>> +#define MFI_INFO_AOPS_SELF_DIAGNOSTIC 0x1000
>> +#define MFI_INFO_AOPS_MIXED_ARRAY 0x2000
>> +#define MFI_INFO_AOPS_GLOBAL_SPARES 0x4000
>> +
>> + uint32_t ld_ops;
>> +#define MFI_INFO_LDOPS_READ_POLICY 0x01
>> +#define MFI_INFO_LDOPS_WRITE_POLICY 0x02
>> +#define MFI_INFO_LDOPS_IO_POLICY 0x04
>> +#define MFI_INFO_LDOPS_ACCESS_POLICY 0x08
>> +#define MFI_INFO_LDOPS_DISK_CACHE_POLICY 0x10
>> +
>> + struct {
>> + uint8_t min;
>> + uint8_t max;
>> + uint8_t reserved[2];
>> + } __attribute__ ((packed)) stripe_sz_ops;
>> +
>> + uint32_t pd_ops;
>> +#define MFI_INFO_PDOPS_FORCE_ONLINE 0x01
>> +#define MFI_INFO_PDOPS_FORCE_OFFLINE 0x02
>> +#define MFI_INFO_PDOPS_FORCE_REBUILD 0x04
>> +
>> + uint32_t pd_mix_support;
>> +#define MFI_INFO_PDMIX_SAS 0x01
>> +#define MFI_INFO_PDMIX_SATA 0x02
>> +#define MFI_INFO_PDMIX_ENCL 0x04
>> +#define MFI_INFO_PDMIX_LD 0x08
>> +#define MFI_INFO_PDMIX_SATA_CLUSTER 0x10
>> +
>> + uint8_t ecc_bucket_count;
>> + uint8_t reserved2[11];
>> + struct mfi_ctrl_props properties;
>> + char package_version[0x60];
>> + uint8_t pad[0x800 - 0x6a0];
>> +} __attribute__ ((packed));
>> +
>> +/* keep track of an event. */
>> +union mfi_evt {
>> + struct {
>> + uint16_t locale;
>> + uint8_t reserved;
>> + int8_t class;
>> + } members;
>> + uint32_t word;
>> +} __attribute__ ((packed));
>> +
>> +/* event log state. */
>> +struct mfi_evt_log_state {
>> + uint32_t newest_seq_num;
>> + uint32_t oldest_seq_num;
>> + uint32_t clear_seq_num;
>> + uint32_t shutdown_seq_num;
>> + uint32_t boot_seq_num;
>> +} __attribute__ ((packed));
>> +
>> +struct mfi_progress {
>> + uint16_t progress;
>> + uint16_t elapsed_seconds;
>> +} __attribute__ ((packed));
>> +
>> +struct mfi_evt_ld {
>> + uint16_t target_id;
>> + uint8_t ld_index;
>> + uint8_t reserved;
>> +} __attribute__ ((packed));
>> +
>> +struct mfi_evt_pd {
>> + uint16_t device_id;
>> + uint8_t enclosure_index;
>> + uint8_t slot_number;
>> +} __attribute__ ((packed));
>> +
>> +/* event detail, returned from MFI_DCMD_CTRL_EVENT_WAIT. */
>> +struct mfi_evt_detail {
>> + uint32_t seq;
>> + uint32_t time;
>> + uint32_t code;
>> + union mfi_evt class;
>> + uint8_t arg_type;
>> + uint8_t reserved1[15];
>> +
>> + union {
>> + struct {
>> + struct mfi_evt_pd pd;
>> + uint8_t cdb_len;
>> + uint8_t sense_len;
>> + uint8_t reserved[2];
>> + uint8_t cdb[16];
>> + uint8_t sense[64];
>> + } cdb_sense;
>> +
>> + struct mfi_evt_ld ld;
>> +
>> + struct {
>> + struct mfi_evt_ld ld;
>> + uint64_t count;
>> + } ld_count;
>> +
>> + struct {
>> + uint64_t lba;
>> + struct mfi_evt_ld ld;
>> + } ld_lba;
>> +
>> + struct {
>> + struct mfi_evt_ld ld;
>> + uint32_t pre_owner;
>> + uint32_t new_owner;
>> + } ld_owner;
>> +
>> + struct {
>> + uint64_t ld_lba;
>> + uint64_t pd_lba;
>> + struct mfi_evt_ld ld;
>> + struct mfi_evt_pd pd;
>> + } ld_lba_pd_lba;
>> +
>> + struct {
>> + struct mfi_evt_ld ld;
>> + struct mfi_progress prog;
>> + } ld_prog;
>> +
>> + struct {
>> + struct mfi_evt_ld ld;
>> + uint32_t prev_state;
>> + uint32_t new_state;
>> + } ld_state;
>> +
>> + struct {
>> + uint64_t strip;
>> + struct mfi_evt_ld ld;
>> + } ld_strip;
>> +
>> + struct mfi_evt_pd pd;
>> +
>> + struct {
>> + struct mfi_evt_pd pd;
>> + uint32_t err;
>> + } pd_err;
>> +
>> + struct {
>> + uint64_t lba;
>> + struct mfi_evt_pd pd;
>> + } pd_lba;
>> +
>> + struct {
>> + uint64_t lba;
>> + struct mfi_evt_pd pd;
>> + struct mfi_evt_ld ld;
>> + } pd_lba_ld;
>> +
>> + struct {
>> + struct mfi_evt_pd pd;
>> + struct mfi_progress prog;
>> + } pd_prog;
>> +
>> + struct {
>> + struct mfi_evt_pd ld;
>> + uint32_t prev_state;
>> + uint32_t new_state;
>> + } pd_state;
>> +
>> + struct {
>> + uint16_t venderId;
>> + uint16_t deviceId;
>> + uint16_t subVenderId;
>> + uint16_t subDeviceId;
>> + } pci;
>> +
>> + uint32_t rate;
>> +
>> + char str[96];
>> +
>> + struct {
>> + uint32_t rtc;
>> + uint16_t elapsedSeconds;
>> + } time;
>> +
>> + struct {
>> + uint32_t ecar;
>> + uint32_t elog;
>> + char str[64];
>> + } ecc;
>> +
>> + uint8_t b[96];
>> + uint16_t s[48];
>> + uint32_t w[24];
>> + uint64_t d[12];
>> + } args;
>> +
>> + char description[128];
>> +} __attribute__ ((packed));
>> +
>> +struct mfi_evt_list {
>> + uint32_t count;
>> + uint32_t reserved;
>> + struct mfi_evt_detail event[1];
>> +} __attribute__ ((packed));
>> +
>> +union mfi_pd_ref {
>> + struct {
>> + uint16_t device_id;
>> + uint16_t seq_num;
>> + } v;
>> + uint32_t ref;
>> +} __attribute__ ((packed));
>> +
>> +union mfi_pd_ddf_type {
>> + struct {
>> + union {
>> + struct {
>> + uint16_t forced_pd_guid:1,
>> + in_vd:1,
>> + is_global_spare:1,
>> + is_spare:1,
>> + is_foreign:1,
>> + reserved:7,
>> + intf:4;
>> + } pd_type;
>> + uint16_t type;
>> + } v;
>> + uint16_t reserved;
>> + } ddf;
>> + struct {
>> + uint32_t reserved;
>> + } non_disk;
>> + uint32_t type;
>> +} __attribute__ ((packed));
>> +
>> +struct mfi_pd_progress {
>> + struct {
>> + uint32_t rbld:1,
>> + patrol:1 ,
>> + clear:1,
>> + reserved:29;
>> + } active;
>> + struct mfi_progress rbld;
>> + struct mfi_progress patrol;
>> + struct mfi_progress clear;
>> + struct mfi_progress reserved[4];
>> +} __attribute__ ((packed));
>> +
>> +struct mfi_pd_info {
>> + union mfi_pd_ref ref;
>> + uint8_t inquiry_data[96];
>> + uint8_t vpd_page83[64];
>> + uint8_t not_supported;
>> + uint8_t scsi_dev_type;
>> + uint8_t connected_port_bitmap;
>> + uint8_t device_speed;
>> + uint32_t media_err_count;
>> + uint32_t other_err_count;
>> + uint32_t pred_fail_count;
>> + uint32_t last_pred_fail_event_seq_num;
>> + uint16_t fw_state;
>> + uint8_t disable_for_removal;
>> + uint8_t link_speed;
>> + union mfi_pd_ddf_type state;
>> + struct {
>> + uint8_t count;
>> + uint8_t is_path_broken;
>> + uint8_t reserved[6];
>> + uint64_t sas_addr[4];
>> + } path_info;
>> + uint64_t raw_size;
>> + uint64_t non_coerced_size;
>> + uint64_t coerced_size;
>> + uint16_t encl_device_id;
>> + uint8_t encl_index;
>> + uint8_t slot_number;
>> + struct mfi_pd_progress prog_info;
>> + uint8_t bad_block_table_full;
>> + uint8_t unusable_in_current_config;
>> + uint8_t vpd_page83_ext[64];
>> + uint8_t reserved[512-358];
>> +} __attribute__ ((packed));
>> +
>> +struct mfi_pd_address {
>> + uint16_t device_id;
>> + uint16_t encl_device_id;
>> + uint8_t encl_index;
>> + uint8_t slot_number;
>> + uint8_t scsi_dev_type;
>> + uint8_t connect_port_bitmap;
>> + uint64_t sas_addr[2];
>> +} __attribute__ ((packed));
>> +
>> +#define MAX_SYS_PDS 240
>> +struct mfi_pd_list {
>> + uint32_t size;
>> + uint32_t count;
>> + struct mfi_pd_address addr[MAX_SYS_PDS];
>> +} __attribute__ ((packed));
>> +
>> +union mfi_ld_ref {
>> + struct {
>> + uint8_t target_id;
>> + uint8_t reserved;
>> + uint16_t seq;
>> + } v;
>> + uint32_t ref;
>> +} __attribute__ ((packed));
>> +
>> +struct mfi_ld_list {
>> + uint32_t ld_count;
>> + uint32_t reserved1;
>> + struct {
>> + union mfi_ld_ref ld;
>> + uint8_t state;
>> + uint8_t reserved2[3];
>> + uint64_t size;
>> + } ld_list[MFI_MAX_LD];
>> +} __attribute__ ((packed));
>> +
>> +enum mfi_ld_access {
>> + MFI_LD_ACCESS_RW = 0,
>> + MFI_LD_ACCSSS_RO = 2,
>> + MFI_LD_ACCESS_BLOCKED = 3,
>> +};
>> +#define MFI_LD_ACCESS_MASK 3
>> +
>> +enum mfi_ld_state {
>> + MFI_LD_STATE_OFFLINE = 0,
>> + MFI_LD_STATE_PARTIALLY_DEGRADED = 1,
>> + MFI_LD_STATE_DEGRADED = 2,
>> + MFI_LD_STATE_OPTIMAL = 3
>> +};
>> +
>> +enum mfi_syspd_state {
>> + MFI_PD_STATE_UNCONFIGURED_GOOD = 0x00,
>> + MFI_PD_STATE_UNCONFIGURED_BAD = 0x01,
>> + MFI_PD_STATE_HOT_SPARE = 0x02,
>> + MFI_PD_STATE_OFFLINE = 0x10,
>> + MFI_PD_STATE_FAILED = 0x11,
>> + MFI_PD_STATE_REBUILD = 0x14,
>> + MFI_PD_STATE_ONLINE = 0x18,
>> + MFI_PD_STATE_COPYBACK = 0x20,
>> + MFI_PD_STATE_SYSTEM = 0x40
>> +};
>> +
>> +struct mfi_ld_props {
>> + union mfi_ld_ref ld;
>> + char name[16];
>> + uint8_t default_cache_policy;
>> + uint8_t access_policy;
>> + uint8_t disk_cache_policy;
>> + uint8_t current_cache_policy;
>> + uint8_t no_bgi;
>> + uint8_t reserved[7];
>> +} __attribute__ ((packed));
>> +
>> +struct mfi_ld_params {
>> + uint8_t primary_raid_level;
>> + uint8_t raid_level_qualifier;
>> + uint8_t secondary_raid_level;
>> + uint8_t stripe_size;
>> + uint8_t num_drives;
>> + uint8_t span_depth;
>> + uint8_t state;
>> + uint8_t init_state;
>> + uint8_t is_consistent;
>> + uint8_t reserved[23];
>> +} __attribute__ ((packed));
>> +
>> +struct mfi_ld_progress {
>> + uint32_t active;
>> +#define MFI_LD_PROGRESS_CC (1<<0)
>> +#define MFI_LD_PROGRESS_BGI (1<<1)
>> +#define MFI_LD_PROGRESS_FGI (1<<2)
>> +#define MFI_LD_PORGRESS_RECON (1<<3)
>> + struct mfi_progress cc;
>> + struct mfi_progress bgi;
>> + struct mfi_progress fgi;
>> + struct mfi_progress recon;
>> + struct mfi_progress reserved[4];
>> +} __attribute__ ((packed));
>> +
>> +struct mfi_span {
>> + uint64_t start_block;
>> + uint64_t num_blocks;
>> + uint16_t array_ref;
>> + uint8_t reserved[6];
>> +} __attribute__ ((packed));
>> +
>> +#define MFI_MAX_SPAN_DEPTH 8
>> +struct mfi_ld_config {
>> + struct mfi_ld_props properties;
>> + struct mfi_ld_params params;
>> + struct mfi_span span[MFI_MAX_SPAN_DEPTH];
>> +} __attribute__ ((packed));
>> +
>> +struct mfi_ld_info {
>> + struct mfi_ld_config ld_config;
>> + uint64_t size;
>> + struct mfi_ld_progress progress;
>> + uint16_t cluster_owner;
>> + uint8_t reconstruct_active;
>> + uint8_t reserved1[1];
>> + uint8_t vpd_page83[64];
>> + uint8_t reserved2[16];
>> +} __attribute__ ((packed));
>> +
>> +union mfi_spare_type {
>> + struct {
>> + uint8_t is_dedicate:1,
>> + is_revertable:1,
>> + is_encl_affinity:1,
>> + reserved:5;
>> + } v;
>> + uint8_t type;
>> +} __attribute__ ((packed));
>> +
>> +#define MAX_ARRAYS 16
>
> Use a prefix to avoid global namespace pollution.
>
>> +struct mfi_spare {
>> + union mfi_pd_ref ref;
>> + union mfi_spare_type spare_type;
>> + uint8_t reserved[2];
>> + uint8_t array_count;
>> + uint16_t array_refd[MAX_ARRAYS];
>> +} __attribute__ ((packed));
>> +
>> +#define MAX_ROW_SIZE 32
>
> Use a prefix to avoid global namespace pollution.
>
>> +struct mfi_array {
>> + uint64_t size;
>> + uint8_t num_drives;
>> + uint8_t reserved;
>> + uint16_t array_ref;
>> + uint8_t pad[20];
>> + struct {
>> + union mfi_pd_ref ref;
>> + uint16_t fw_state;
>> + struct {
>> + uint8_t pd;
>> + uint8_t slot;
>> + } encl;
>> + } pd[MAX_ROW_SIZE];
>> +} __attribute__ ((packed));
>> +
>> +struct mfi_config_data {
>> + uint32_t size;
>> + uint16_t array_count;
>> + uint16_t array_size;
>> + uint16_t log_drv_count;
>> + uint16_t log_drv_size;
>> + uint16_t spares_count;
>> + uint16_t spares_size;
>> + uint8_t reserved[16];
>> + uint8_t data;
>> + /*
>> + struct mfi_array array[];
>> + struct mfi_ld_config ld[];
>> + struct mfi_spare spare[];
>> + */
>> +} __attribute__ ((packed));
>> +
>> +#define MFI_SCSI_MAX_t ARGETS 128
>
> What's this?
>
>> +#define MFI_SCSI_MAX_LUNS 8
>> +#define MFI_SCSI_INITIATOR_ID 255
>> +#define MFI_SCSI_MAX_CMDS 8
>> +#define MFI_SCSI_MAX_CDB_LEN 16
>> +
>> +#endif /* _MFI_H */
>
> make below a separate patch pls.
>
>> diff --git a/hw/pci_ids.h b/hw/pci_ids.h
>> index e8235a7..0306255 100644
>> --- a/hw/pci_ids.h
>> +++ b/hw/pci_ids.h
>> @@ -12,9 +12,9 @@
>>
>> #define PCI_BASE_CLASS_STORAGE 0x01
>> #define PCI_BASE_CLASS_NETWORK 0x02
>> -
>
> Why? This separates base class list from storage
> subclasses.
>
>> #define PCI_CLASS_STORAGE_SCSI 0x0100
>> #define PCI_CLASS_STORAGE_IDE 0x0101
>> +#define PCI_CLASS_STORAGE_RAID 0x0104
>> #define PCI_CLASS_STORAGE_SATA 0x0106
>> #define PCI_CLASS_STORAGE_OTHER 0x0180
>>
>> @@ -47,6 +47,7 @@
>>
>> #define PCI_VENDOR_ID_LSI_LOGIC 0x1000
>> #define PCI_DEVICE_ID_LSI_53C895A 0x0012
>> +#define PCI_DEVICE_ID_LSI_SAS1078 0x0060
>>
>> #define PCI_VENDOR_ID_DEC 0x1011
>> #define PCI_DEVICE_ID_DEC_21154 0x0026
>> --
>> 1.7.3.4
>>
>
>
next prev parent reply other threads:[~2012-02-24 15:58 UTC|newest]
Thread overview: 15+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-02-21 9:36 [Qemu-devel] [PATCH][v14] megasas: LSI Megaraid SAS HBA emulation Hannes Reinecke
2012-02-21 18:54 ` Gerhard Wiesinger
2012-02-23 7:03 ` Gerhard Wiesinger
2012-02-23 7:14 ` Andreas Färber
2012-02-23 7:20 ` Gerhard Wiesinger
2012-02-23 7:12 ` Alexander Graf
2012-02-23 15:34 ` Michael S. Tsirkin
2012-02-24 15:58 ` Anthony Liguori [this message]
2012-02-24 16:05 ` Alexander Graf
2012-02-24 16:13 ` Anthony Liguori
2012-02-27 9:17 ` Hannes Reinecke
2012-02-27 10:31 ` Michael S. Tsirkin
2012-02-27 15:24 ` Hannes Reinecke
2012-03-02 7:20 ` Gerhard Wiesinger
2012-02-27 13:47 ` Andreas Färber
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=4F47B393.7070606@codemonkey.ws \
--to=anthony@codemonkey.ws \
--cc=afaerber@suse.de \
--cc=agraf@suse.de \
--cc=hare@suse.de \
--cc=lists@wiesinger.com \
--cc=mst@redhat.com \
--cc=qemu-devel@nongnu.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).