Latest firmware adds a mechanism to save firmware logs just before
controller reset on pre-allocated internal controller DRAM. This feature
is called snapdump which will help debugging firmware issues.
This feature requires extra time and firmware reports these values through
new driver interface. Before initiating an OCR, driver needs to inform FW
to save a snapdump and then wait for a specified time for the snapdump
to complete.
Signed-off-by: Sumit Saxena <sumit.saxena@xxxxxxxxxxxx>
Signed-off-by: Shivasharan S <shivasharan.srikanteshwara@xxxxxxxxxxxx>
---
Changes in V2:
Replace PCI DMA APIs with generic DMA APIs
drivers/scsi/megaraid/megaraid_sas.h | 25 +++++-
drivers/scsi/megaraid/megaraid_sas_base.c | 118 +++++++++++++++++++++++++++-
drivers/scsi/megaraid/megaraid_sas_fusion.c | 66 ++++++++++++++--
drivers/scsi/megaraid/megaraid_sas_fusion.h | 12 +++
4 files changed, 208 insertions(+), 13 deletions(-)
diff --git a/drivers/scsi/megaraid/megaraid_sas.h b/drivers/scsi/megaraid/megaraid_sas.h
index 8c0f74a2740a..a2df9827e932 100644
--- a/drivers/scsi/megaraid/megaraid_sas.h
+++ b/drivers/scsi/megaraid/megaraid_sas.h
@@ -142,6 +142,7 @@
* CLR_HANDSHAKE: FW is waiting for HANDSHAKE from BIOS or Driver
* HOTPLUG : Resume from Hotplug
* MFI_STOP_ADP : Send signal to FW to stop processing
+ * MFI_ADP_TRIGGER_SNAP_DUMP: Inform firmware to initiate snap dump
*/
#define WRITE_SEQUENCE_OFFSET (0x0000000FC) /* I20 */
#define HOST_DIAGNOSTIC_OFFSET (0x000000F8) /* I20 */
@@ -158,6 +159,7 @@
#define MFI_RESET_FLAGS MFI_INIT_READY| \
MFI_INIT_MFIMODE| \
MFI_INIT_ABORT
+#define MFI_ADP_TRIGGER_SNAP_DUMP 0x00000100
#define MPI2_IOCINIT_MSGFLAG_RDPQ_ARRAY_MODE (0x01)
/*
@@ -860,8 +862,22 @@ struct megasas_ctrl_prop {
u32 reserved:18;
#endif
} OnOffProperties;
- u8 autoSnapVDSpace;
- u8 viewSpace;
+
+ union {
+ u8 autoSnapVDSpace;
+ u8 viewSpace;
+ struct {
+#if defined(__BIG_ENDIAN_BITFIELD)
+ u16 reserved2:11;
+ u16 enable_snap_dump:1;
+ u16 reserved1:4;
+#else
+ u16 reserved1:4;
+ u16 enable_snap_dump:1;
+ u16 reserved2:11;
+#endif
+ } on_off_properties2;
+ };
__le16 spinDownTime;
u8 reserved[24];
} __packed;
@@ -2185,6 +2201,9 @@ struct megasas_instance {
struct MR_LD_TARGETID_LIST *ld_targetid_list_buf;
dma_addr_t ld_targetid_list_buf_h;
+ struct MR_SNAPDUMP_PROPERTIES *snapdump_prop;
+ dma_addr_t snapdump_prop_h;
+
void *crash_buf[MAX_CRASH_DUMP_SIZE];
unsigned int fw_crash_buffer_size;
unsigned int fw_crash_state;
@@ -2316,6 +2335,7 @@ struct megasas_instance {
bool support_nvme_passthru;
u8 task_abort_tmo;
u8 max_reset_tmo;
+ u8 snapdump_wait_time;
};
struct MR_LD_VF_MAP {
u32 size;
@@ -2541,6 +2561,7 @@ void megasas_set_dynamic_target_properties(struct scsi_device *sdev,
bool is_target_prop);
int megasas_get_target_prop(struct megasas_instance *instance,
struct scsi_device *sdev);
+void megasas_get_snapdump_properties(struct megasas_instance *instance);
int megasas_set_crash_dump_params(struct megasas_instance *instance,
u8 crash_buf_state);
diff --git a/drivers/scsi/megaraid/megaraid_sas_base.c b/drivers/scsi/megaraid/megaraid_sas_base.c
index 4dc29e055461..3ddf71b8c38d 100644
--- a/drivers/scsi/megaraid/megaraid_sas_base.c
+++ b/drivers/scsi/megaraid/megaraid_sas_base.c
@@ -4661,6 +4661,87 @@ static void megasas_update_ext_vd_details(struct megasas_instance *instance)
fusion->drv_map_sz = sizeof(struct MR_DRV_RAID_MAP_ALL);
}
+/*
+ * dcmd.opcode - MR_DCMD_CTRL_SNAPDUMP_GET_PROPERTIES
+ * dcmd.hdr.length - number of bytes to read
+ * dcmd.sge - Ptr to MR_SNAPDUMP_PROPERTIES
+ * Desc: Fill in snapdump properties
+ * Status: MFI_STAT_OK- Command successful
+ */
+void megasas_get_snapdump_properties(struct megasas_instance *instance)
+{
+ int ret = 0;
+ struct megasas_cmd *cmd;
+ struct megasas_dcmd_frame *dcmd;
+ struct MR_SNAPDUMP_PROPERTIES *ci;
+ dma_addr_t ci_h = 0;
+
+ ci = instance->snapdump_prop;
+ ci_h = instance->snapdump_prop_h;
+
+ if (!ci)
+ return;
+
+ cmd = megasas_get_cmd(instance);
+
+ if (!cmd) {
+ dev_dbg(&instance->pdev->dev, "Failed to get a free cmd\n");
+ return;
+ }
+
+ dcmd = &cmd->frame->dcmd;
+
+ memset(ci, 0, sizeof(*ci));
+ memset(dcmd->mbox.b, 0, MFI_MBOX_SIZE);
+
+ dcmd->cmd = MFI_CMD_DCMD;
+ dcmd->cmd_status = MFI_STAT_INVALID_STATUS;
+ dcmd->sge_count = 1;
+ dcmd->flags = MFI_FRAME_DIR_READ;
+ dcmd->timeout = 0;
+ dcmd->pad_0 = 0;
+ dcmd->data_xfer_len = cpu_to_le32(sizeof(struct MR_SNAPDUMP_PROPERTIES));
+ dcmd->opcode = cpu_to_le32(MR_DCMD_CTRL_SNAPDUMP_GET_PROPERTIES);
+
+ megasas_set_dma_settings(instance, dcmd, ci_h,
+ sizeof(struct MR_SNAPDUMP_PROPERTIES));
+
+ if (!instance->mask_interrupts) {
+ ret = megasas_issue_blocked_cmd(instance, cmd,
+ MFI_IO_TIMEOUT_SECS);
+ } else {
+ ret = megasas_issue_polled(instance, cmd);
+ cmd->flags |= DRV_DCMD_SKIP_REFIRE;
+ }
+
+ switch (ret) {
+ case DCMD_SUCCESS:
+ instance->snapdump_wait_time =
+ min_t(u8, ci->trigger_min_num_sec_before_ocr,
+ MEGASAS_MAX_SNAP_DUMP_WAIT_TIME);
+ break;
+
+ case DCMD_TIMEOUT:
+ switch (dcmd_timeout_ocr_possible(instance)) {
+ case INITIATE_OCR:
+ cmd->flags |= DRV_DCMD_SKIP_REFIRE;
+ megasas_reset_fusion(instance->host,
+ MFI_IO_TIMEOUT_OCR);
+ break;
+ case KILL_ADAPTER:
+ megaraid_sas_kill_hba(instance);
+ break;
+ case IGNORE_TIMEOUT:
+ dev_info(&instance->pdev->dev, "Ignore DCMD timeout: %s %d\n",
+ __func__, __LINE__);
+ break;
+ }
+ }
+
+ if (ret != DCMD_TIMEOUT)
+ megasas_return_cmd(instance, cmd);
+}
+
/**
* megasas_get_controller_info - Returns FW's controller structure
* @instance: Adapter soft state
@@ -4720,6 +4801,7 @@ megasas_get_ctrl_info(struct megasas_instance *instance)
* CPU endianness format.
*/
le32_to_cpus((u32 *)&ci->properties.OnOffProperties);
+ le16_to_cpus((u16 *)&ci->properties.on_off_properties2);
le32_to_cpus((u32 *)&ci->adapterOperations2);
le32_to_cpus((u32 *)&ci->adapterOperations3);
le16_to_cpus((u16 *)&ci->adapter_operations4);
@@ -4741,6 +4823,11 @@ megasas_get_ctrl_info(struct megasas_instance *instance)
/*Check whether controller is iMR or MR */
instance->is_imr = (ci->memory_size ? 0 : 1);
+
+ instance->snapdump_wait_time =
+ (ci->properties.on_off_properties2.enable_snap_dump ?
+ MEGASAS_DEFAULT_SNAP_DUMP_WAIT_TIME : 0);
+
dev_info(&instance->pdev->dev,
"controller type\t: %s(%dMB)\n",
instance->is_imr ? "iMR" : "MR",
@@ -5539,6 +5626,11 @@ static int megasas_init_fw(struct megasas_instance *instance)
instance->crash_dump_buf = NULL;
}
+ if (instance->snapdump_wait_time) {
+ megasas_get_snapdump_properties(instance);
+ dev_info(&instance->pdev->dev, "Snap dump wait time\t: %d\n",
+ instance->snapdump_wait_time);
+ }
dev_info(&instance->pdev->dev,
"pci id\t\t: (0x%04x)/(0x%04x)/(0x%04x)/(0x%04x)\n",
@@ -5553,7 +5645,6 @@ static int megasas_init_fw(struct megasas_instance *instance)
dev_info(&instance->pdev->dev, "jbod sync map : %s\n",
instance->use_seqnum_jbod_fp ? "yes" : "no");
-
instance->max_sectors_per_req = instance->max_num_sge *
SGE_BUFFER_SIZE / 512;
if (tmp_sectors && (instance->max_sectors_per_req > tmp_sectors))
@@ -6257,6 +6348,14 @@ int megasas_alloc_ctrl_dma_buffers(struct megasas_instance *instance)
"Failed to allocate PD list buffer\n");
return -ENOMEM;
}
+
+ instance->snapdump_prop = dma_alloc_coherent(&pdev->dev,
+ sizeof(struct MR_SNAPDUMP_PROPERTIES),
+ &instance->snapdump_prop_h, GFP_KERNEL);
+
+ if (!instance->snapdump_prop)
+ dev_err(&pdev->dev,
+ "Failed to allocate snapdump properties buffer\n");
}
instance->pd_list_buf =
@@ -6400,6 +6499,12 @@ void megasas_free_ctrl_dma_buffers(struct megasas_instance *instance)
dma_free_coherent(&pdev->dev, CRASH_DMA_BUF_SIZE,
instance->crash_dump_buf,
instance->crash_dump_h);
+
+ if (instance->snapdump_prop)
+ dma_free_coherent(&pdev->dev,
+ sizeof(struct MR_SNAPDUMP_PROPERTIES),
+ instance->snapdump_prop,
+ instance->snapdump_prop_h);
}
/*
@@ -7763,8 +7868,15 @@ megasas_aen_polling(struct work_struct *work)
break;
case MR_EVT_CTRL_PROP_CHANGED:
- dcmd_ret = megasas_get_ctrl_info(instance);
- break;
+ dcmd_ret = megasas_get_ctrl_info(instance);
+ if (dcmd_ret == DCMD_SUCCESS &&
+ instance->snapdump_wait_time) {
+ megasas_get_snapdump_properties(instance);
+ dev_info(&instance->pdev->dev,
+ "Snap dump wait time\t: %d\n",
+ instance->snapdump_wait_time);
+ }
+ break;
default:
doscan = 0;
break;
diff --git a/drivers/scsi/megaraid/megaraid_sas_fusion.c b/drivers/scsi/megaraid/megaraid_sas_fusion.c
index 9ca4a52164bd..2868b71add2d 100644
--- a/drivers/scsi/megaraid/megaraid_sas_fusion.c
+++ b/drivers/scsi/megaraid/megaraid_sas_fusion.c
@@ -3884,14 +3884,57 @@ megasas_check_reset_fusion(struct megasas_instance *instance,
return 0;
}
+/**
+ * megasas_trigger_snap_dump - Trigger snap dump in FW
+ * @instance: Soft instance of adapter
+ */
+static inline void megasas_trigger_snap_dump(struct megasas_instance *instance)
+{
+ int j;
+ u32 fw_state;
+
+ if (!instance->disableOnlineCtrlReset) {
+ dev_info(&instance->pdev->dev, "Trigger snap dump\n");
+ writel(MFI_ADP_TRIGGER_SNAP_DUMP,
+ &instance->reg_set->doorbell);
+ readl(&instance->reg_set->doorbell);
+ }
+
+ for (j = 0; j < instance->snapdump_wait_time; j++) {
+ fw_state = instance->instancet->read_fw_status_reg(
+ instance->reg_set) & MFI_STATE_MASK;
+ if (fw_state == MFI_STATE_FAULT) {
+ dev_err(&instance->pdev->dev,
+ "Found FW in FAULT state, after snap dump trigger\n");
+ return;
+ }
+ msleep(1000);
+ }
+}
+
/* This function waits for outstanding commands on fusion to complete */
int megasas_wait_for_outstanding_fusion(struct megasas_instance *instance,
int reason, int *convert)
{
int i, outstanding, retval = 0, hb_seconds_missed = 0;
u32 fw_state;
+ u32 waittime_for_io_completion;
- for (i = 0; i < resetwaittime; i++) {
+ waittime_for_io_completion =
+ min_t(u32, resetwaittime,
+ (resetwaittime - instance->snapdump_wait_time));
+
+ if (reason == MFI_IO_TIMEOUT_OCR) {
+ dev_info(&instance->pdev->dev,
+ "MFI command is timed out\n");
+ megasas_complete_cmd_dpc_fusion((unsigned long)instance);
+ if (instance->snapdump_wait_time)
+ megasas_trigger_snap_dump(instance);
+ retval = 1;
+ goto out;
+ }
+
+ for (i = 0; i < waittime_for_io_completion; i++) {
/* Check if firmware is in fault state */
fw_state = instance->instancet->read_fw_status_reg(
instance->reg_set) & MFI_STATE_MASK;
@@ -3912,13 +3955,6 @@ int megasas_wait_for_outstanding_fusion(struct megasas_instance *instance,
goto out;
}
- if (reason == MFI_IO_TIMEOUT_OCR) {
- dev_info(&instance->pdev->dev,
- "MFI IO is timed out, initiating OCR\n");
- megasas_complete_cmd_dpc_fusion((unsigned long)instance);
- retval = 1;
- goto out;
- }
/* If SR-IOV VF mode & heartbeat timeout, don't wait */
if (instance->requestorId && !reason) {
@@ -3963,6 +3999,12 @@ int megasas_wait_for_outstanding_fusion(struct megasas_instance *instance,
msleep(1000);
}
+ if (instance->snapdump_wait_time) {
+ megasas_trigger_snap_dump(instance);
+ retval = 1;
+ goto out;
+ }
+
if (atomic_read(&instance->fw_outstanding)) {
dev_err(&instance->pdev->dev, "pending commands remain after waiting, "
"will reset adapter scsi%d.\n",
@@ -3970,6 +4012,7 @@ int megasas_wait_for_outstanding_fusion(struct megasas_instance *instance,
*convert = 1;
retval = 1;
}
+
out:
return retval;
}
@@ -4783,6 +4826,13 @@ int megasas_reset_fusion(struct Scsi_Host *shost, int reason)
megasas_set_crash_dump_params(instance,
MR_CRASH_BUF_TURN_OFF);
+ if (instance->snapdump_wait_time) {
+ megasas_get_snapdump_properties(instance);
+ dev_info(&instance->pdev->dev,
+ "Snap dump wait time\t: %d\n",
+ instance->snapdump_wait_time);
+ }
+
retval = SUCCESS;
/* Adapter reset completed successfully */
diff --git a/drivers/scsi/megaraid/megaraid_sas_fusion.h b/drivers/scsi/megaraid/megaraid_sas_fusion.h
index 8e5ebee6517f..14d8e409832c 100644
--- a/drivers/scsi/megaraid/megaraid_sas_fusion.h
+++ b/drivers/scsi/megaraid/megaraid_sas_fusion.h
@@ -725,6 +725,7 @@ struct MPI2_IOC_INIT_REQUEST {
#define MR_DCMD_CTRL_SHARED_HOST_MEM_ALLOC 0x010e8485 /* SR-IOV HB alloc*/
#define MR_DCMD_LD_VF_MAP_GET_ALL_LDS_111 0x03200200
#define MR_DCMD_LD_VF_MAP_GET_ALL_LDS 0x03150200
+#define MR_DCMD_CTRL_SNAPDUMP_GET_PROPERTIES 0x01200100
struct MR_DEV_HANDLE_INFO {
__le16 curDevHdl;
@@ -1063,6 +1064,9 @@ struct MR_FW_RAID_MAP_DYNAMIC {
#define MPI26_IEEE_SGE_FLAGS_NSF_NVME_PRP (0x08)
#define MPI26_IEEE_SGE_FLAGS_NSF_NVME_SGL (0x10)
+#define MEGASAS_DEFAULT_SNAP_DUMP_WAIT_TIME 15
+#define MEGASAS_MAX_SNAP_DUMP_WAIT_TIME 60
+
struct megasas_register_set;
struct megasas_instance;
@@ -1350,6 +1354,14 @@ enum CMD_RET_VALUES {
RETURN_CMD = 3,
};
+struct MR_SNAPDUMP_PROPERTIES {
+ u8 offload_num;
+ u8 max_num_supported;
+ u8 cur_num_supported;
+ u8 trigger_min_num_sec_before_ocr;
+ u8 reserved[12];
+};
+
void megasas_free_cmds_fusion(struct megasas_instance *instance);
int megasas_ioc_init_fusion(struct megasas_instance *instance);
u8 megasas_get_map_info(struct megasas_instance *instance);
--
2.16.1
