* Re: [PATCH V2 1/3] pm80xx : Support for get phy profile functionality.
From: kernel test robot @ 2020-06-24 14:18 UTC (permalink / raw)
To: Deepak Ukey, linux-scsi
Cc: kbuild-all, Vasanthalakshmi.Tharmarajan, Viswas.G, deepak.ukey,
jinpu.wang, martin.petersen, dpf, yuuzheng, auradkar, vishakhavc
In-Reply-To: <20200624120322.6265-2-deepak.ukey@microchip.com>
[-- Attachment #1: Type: text/plain, Size: 2893 bytes --]
Hi Deepak,
Thank you for the patch! Perhaps something to improve:
[auto build test WARNING on mkp-scsi/for-next]
[also build test WARNING on scsi/for-next v5.8-rc2 next-20200624]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use as documented in
https://git-scm.com/docs/git-format-patch]
url: https://github.com/0day-ci/linux/commits/Deepak-Ukey/pm80xx-Updates-for-the-driver-version-0-1-39/20200624-195528
base: https://git.kernel.org/pub/scm/linux/kernel/git/mkp/scsi.git for-next
config: x86_64-rhel-7.6 (attached as .config)
compiler: gcc-9 (Debian 9.3.0-13) 9.3.0
reproduce (this is a W=1 build):
# save the attached .config to linux build tree
make W=1 ARCH=x86_64
If you fix the issue, kindly add following tag as appropriate
Reported-by: kernel test robot <lkp@intel.com>
All warnings (new ones prefixed by >>):
drivers/scsi/pm8001/pm80xx_hwi.c:69:6: warning: no previous prototype for 'pm80xx_pci_mem_copy' [-Wmissing-prototypes]
69 | void pm80xx_pci_mem_copy(struct pm8001_hba_info *pm8001_ha, u32 soffset,
| ^~~~~~~~~~~~~~~~~~~
>> drivers/scsi/pm8001/pm80xx_hwi.c:5064:5: warning: no previous prototype for 'pm8001_chip_get_phy_profile' [-Wmissing-prototypes]
5064 | int pm8001_chip_get_phy_profile(struct pm8001_hba_info *pm8001_ha, int phy_id,
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~
drivers/scsi/pm8001/pm80xx_hwi.c:5094:6: warning: no previous prototype for 'mpi_set_phy_profile_req' [-Wmissing-prototypes]
5094 | void mpi_set_phy_profile_req(struct pm8001_hba_info *pm8001_ha,
| ^~~~~~~~~~~~~~~~~~~~~~~
vim +/pm8001_chip_get_phy_profile +5064 drivers/scsi/pm8001/pm80xx_hwi.c
5063
> 5064 int pm8001_chip_get_phy_profile(struct pm8001_hba_info *pm8001_ha, int phy_id,
5065 int page_code, struct completion *comp, void *buf)
5066 {
5067 u32 tag;
5068 struct get_phy_profile_req payload;
5069 struct inbound_queue_table *circularQ;
5070 struct pm8001_ccb_info *ccb;
5071 int rc, ppc_phyid;
5072 u32 opc = OPC_INB_GET_PHY_PROFILE;
5073
5074 memset(&payload, 0, sizeof(payload));
5075
5076 rc = pm8001_tag_alloc(pm8001_ha, &tag);
5077 if (rc)
5078 PM8001_FAIL_DBG(pm8001_ha, pm8001_printk("Invalid tag\n"));
5079
5080 ccb = &pm8001_ha->ccb_info[tag];
5081 ccb->completion = comp;
5082 ccb->resp_buf = buf;
5083 circularQ = &pm8001_ha->inbnd_q_tbl[0];
5084
5085 payload.tag = cpu_to_le32(tag);
5086 ppc_phyid = (page_code & 0xFF) << 8 | (phy_id & 0xFF);
5087 payload.ppc_phyid = cpu_to_le32(ppc_phyid);
5088
5089 pm8001_mpi_build_cmd(pm8001_ha, circularQ, opc, &payload,
5090 sizeof(payload), 0);
5091 return rc;
5092 }
5093
---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/kbuild-all@lists.01.org
[-- Attachment #2: .config.gz --]
[-- Type: application/gzip, Size: 49244 bytes --]
^ permalink raw reply
* [PATCH 1/1] scsi: storvsc: fix spelling mistake
From: Flavio Suligoi @ 2020-06-24 13:56 UTC (permalink / raw)
To: K . Y . Srinivasan, Haiyang Zhang, Stephen Hemminger, Wei Liu,
James E . J . Bottomley, Martin K . Petersen
Cc: linux-hyperv, linux-scsi, linux-kernel, Flavio Suligoi
Fix typo: "trigerred" --> "triggered"
Signed-off-by: Flavio Suligoi <f.suligoi@asem.it>
---
drivers/scsi/storvsc_drv.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/scsi/storvsc_drv.c b/drivers/scsi/storvsc_drv.c
index 624467e2590a..3040696c20dd 100644
--- a/drivers/scsi/storvsc_drv.c
+++ b/drivers/scsi/storvsc_drv.c
@@ -1038,7 +1038,7 @@ static void storvsc_handle_error(struct vmscsi_request *vm_srb,
do_work = true;
process_err_fn = storvsc_device_scan;
/*
- * Retry the I/O that trigerred this.
+ * Retry the I/O that triggered this.
*/
set_host_byte(scmnd, DID_REQUEUE);
}
--
2.17.1
^ permalink raw reply related
* Re: [PATCH] scsi: sd: add runtime pm to open / release
From: Bart Van Assche @ 2020-06-24 13:33 UTC (permalink / raw)
To: Martin Kepplinger, jejb, martin.petersen; +Cc: linux-scsi, linux-kernel, kernel
In-Reply-To: <20200623111018.31954-1-martin.kepplinger@puri.sm>
On 2020-06-23 04:10, Martin Kepplinger wrote:
> This add a very conservative but simple implementation for runtime PM
> to the sd scsi driver:
> Resume when opened (mounted) and suspend when released (unmounted).
>
> Improvements that allow suspending while a device is "open" can
> be added later, but now we save power when no filesystem is mounted
> and runtime PM is enabled.
>
> Signed-off-by: Martin Kepplinger <martin.kepplinger@puri.sm>
> ---
> drivers/scsi/sd.c | 6 ++++++
> 1 file changed, 6 insertions(+)
>
> diff --git a/drivers/scsi/sd.c b/drivers/scsi/sd.c
> index d90fefffe31b..fe4cb7c50ec1 100644
> --- a/drivers/scsi/sd.c
> +++ b/drivers/scsi/sd.c
> @@ -1372,6 +1372,7 @@ static int sd_open(struct block_device *bdev, fmode_t mode)
> SCSI_LOG_HLQUEUE(3, sd_printk(KERN_INFO, sdkp, "sd_open\n"));
>
> sdev = sdkp->device;
> + scsi_autopm_get_device(sdev);
>
> /*
> * If the device is in error recovery, wait until it is done.
> @@ -1418,6 +1419,9 @@ static int sd_open(struct block_device *bdev, fmode_t mode)
>
> error_out:
> scsi_disk_put(sdkp);
> +
> + scsi_autopm_put_device(sdev);
> +
> return retval;
> }
>
> @@ -1441,6 +1445,8 @@ static void sd_release(struct gendisk *disk, fmode_t mode)
>
> SCSI_LOG_HLQUEUE(3, sd_printk(KERN_INFO, sdkp, "sd_release\n"));
>
> + scsi_autopm_put_device(sdev);
> +
> if (atomic_dec_return(&sdkp->openers) == 0 && sdev->removable) {
> if (scsi_block_when_processing_errors(sdev))
> scsi_set_medium_removal(sdev, SCSI_REMOVAL_ALLOW);
My understanding of the above patch is that it introduces a regression,
namely by disabling runtime suspend as long as an sd device is held open.
Bart.
^ permalink raw reply
* [PATCH V2 3/3] pm80xx : Wait for PHY startup before draining libsas queue.
From: Deepak Ukey @ 2020-06-24 12:03 UTC (permalink / raw)
To: linux-scsi
Cc: Vasanthalakshmi.Tharmarajan, Viswas.G, deepak.ukey, jinpu.wang,
martin.petersen, dpf, yuuzheng, auradkar, vishakhavc, bjashnani,
radha, akshatzen
In-Reply-To: <20200624120322.6265-1-deepak.ukey@microchip.com>
From: peter chang <dpf@google.com>
The host's scan finish waits for the libsas queue to drain. However,
if the PHYs are still in the process of starting then the queue will
be empty. This means that we declare the scan finished before it has
even started. Here we wait for various events from the firmware-side.
The wait behavior can be configured using the module parameter
"wait_for_phy_startup", if the parameter is enabled, the driver will wait
for the phy event from the firmware, before proceeding with load.
Signed-off-by: Viswas G <Viswas.G@microchip.com>
Signed-off-by: peter chang <dpf@google.com>
Signed-off-by: Radha Ramachandran <radha@google.com>
Signed-off-by: Deepak Ukey <Deepak.Ukey@microchip.com>
---
drivers/scsi/pm8001/pm8001_ctl.c | 36 +++++++++++++++++++++
drivers/scsi/pm8001/pm8001_defs.h | 6 ++--
drivers/scsi/pm8001/pm8001_init.c | 22 +++++++++++++
drivers/scsi/pm8001/pm8001_sas.c | 67 +++++++++++++++++++++++++++++++++++++--
drivers/scsi/pm8001/pm8001_sas.h | 4 +++
drivers/scsi/pm8001/pm80xx_hwi.c | 67 ++++++++++++++++++++++++++++++++-------
6 files changed, 185 insertions(+), 17 deletions(-)
diff --git a/drivers/scsi/pm8001/pm8001_ctl.c b/drivers/scsi/pm8001/pm8001_ctl.c
index 3c9f42779dd0..eae629610a5f 100644
--- a/drivers/scsi/pm8001/pm8001_ctl.c
+++ b/drivers/scsi/pm8001/pm8001_ctl.c
@@ -88,6 +88,41 @@ static ssize_t controller_fatal_error_show(struct device *cdev,
}
static DEVICE_ATTR_RO(controller_fatal_error);
+/**
+ * phy_startup_timeout_show - per-phy discovery timeout
+ * @cdev: pointer to embedded class device
+ * @buf: the buffer returned
+ *
+ * A sysfs 'read/write' shost attribute.
+ */
+static ssize_t phy_startup_timeout_show(struct device *cdev,
+ struct device_attribute *attr, char *buf)
+{
+ struct Scsi_Host *shost = class_to_shost(cdev);
+ struct sas_ha_struct *sha = SHOST_TO_SAS_HA(shost);
+ struct pm8001_hba_info *pm8001_ha = sha->lldd_ha;
+
+ return snprintf(buf, PAGE_SIZE, "%08xh\n",
+ pm8001_ha->phy_startup_timeout / HZ);
+}
+
+static ssize_t phy_startup_timeout_store(struct device *cdev,
+ struct device_attribute *attr, const char *buf, size_t count)
+{
+ struct Scsi_Host *shost = class_to_shost(cdev);
+ struct sas_ha_struct *sha = SHOST_TO_SAS_HA(shost);
+ struct pm8001_hba_info *pm8001_ha = sha->lldd_ha;
+ int val = 0;
+
+ if (kstrtoint(buf, 0, &val) < 0)
+ return -EINVAL;
+
+ pm8001_ha->phy_startup_timeout = val * HZ;
+ return strlen(buf);
+}
+
+static DEVICE_ATTR_RW(phy_startup_timeout);
+
/**
* pm8001_ctl_fw_version_show - firmware version
* @cdev: pointer to embedded class device
@@ -867,6 +902,7 @@ static DEVICE_ATTR(update_fw, S_IRUGO|S_IWUSR|S_IWGRP,
struct device_attribute *pm8001_host_attrs[] = {
&dev_attr_interface_rev,
&dev_attr_controller_fatal_error,
+ &dev_attr_phy_startup_timeout,
&dev_attr_fw_version,
&dev_attr_update_fw,
&dev_attr_aap_log,
diff --git a/drivers/scsi/pm8001/pm8001_defs.h b/drivers/scsi/pm8001/pm8001_defs.h
index fd700ce5e80c..aaeabb2f2808 100644
--- a/drivers/scsi/pm8001/pm8001_defs.h
+++ b/drivers/scsi/pm8001/pm8001_defs.h
@@ -141,7 +141,9 @@ enum pm8001_hba_info_flags {
*/
#define PHY_LINK_DISABLE 0x00
#define PHY_LINK_DOWN 0x01
-#define PHY_STATE_LINK_UP_SPCV 0x2
-#define PHY_STATE_LINK_UP_SPC 0x1
+#define PHY_STATE_LINK_UP_SPCV 0x02
+#define PHY_STATE_LINK_UP_SPC 0x01
+#define PHY_STATE_LINK_RESET 0x03
+#define PHY_STATE_HARD_RESET 0x04
#endif
diff --git a/drivers/scsi/pm8001/pm8001_init.c b/drivers/scsi/pm8001/pm8001_init.c
index 46a2f86f3d9b..55b40e4663e6 100644
--- a/drivers/scsi/pm8001/pm8001_init.c
+++ b/drivers/scsi/pm8001/pm8001_init.c
@@ -61,6 +61,26 @@ MODULE_PARM_DESC(staggered_spinup, "enable the staggered spinup feature.\n"
" 0/N: false\n"
" 1/Y: true\n");
+static bool wait_for_phy_startup;
+module_param(wait_for_phy_startup, bool, 0644);
+MODULE_PARM_DESC(wait_for_phy_startup, "wait for phy start.\n"
+ " 0/N: disable\n"
+ " 1/Y: enable\n");
+
+/* if nothing is detected, the PHYs will reset continuously once they
+ * are started. we don't have a good way of differentiating a trained
+ * but waiting-on-signature from one that's never going to train
+ * (nothing attached or dead drive), so we wait an possibly
+ * unreasonable amount of time. this is stuck in start_timeout, and
+ * checked in the host's scan_finished callback for PHYs that haven't
+ * yet come up.
+ *
+ */
+static ulong phy_startup_timeout = 60;
+module_param(phy_startup_timeout, ulong, 0644);
+MODULE_PARM_DESC(phy_startup_timeout_s,
+ " seconds to wait for discovery, per-PHY.");
+
static struct scsi_transport_template *pm8001_stt;
/**
@@ -495,6 +515,8 @@ static struct pm8001_hba_info *pm8001_pci_alloc(struct pci_dev *pdev,
pm8001_ha->id = pm8001_id++;
pm8001_ha->logging_level = logging_level;
pm8001_ha->staggered_spinup = staggered_spinup;
+ pm8001_ha->phy_startup_timeout = phy_startup_timeout * HZ;
+ pm8001_ha->wait_for_phy_startup = wait_for_phy_startup;
pm8001_ha->non_fatal_count = 0;
if (link_rate >= 1 && link_rate <= 15)
pm8001_ha->link_rate = (link_rate << 8);
diff --git a/drivers/scsi/pm8001/pm8001_sas.c b/drivers/scsi/pm8001/pm8001_sas.c
index cf1c5e68953b..21086bdae6a9 100644
--- a/drivers/scsi/pm8001/pm8001_sas.c
+++ b/drivers/scsi/pm8001/pm8001_sas.c
@@ -39,6 +39,7 @@
*/
#include <linux/slab.h>
+#include "pm80xx_hwi.h"
#include "pm8001_sas.h"
/**
@@ -257,6 +258,59 @@ int pm8001_phy_control(struct asd_sas_phy *sas_phy, enum phy_func func,
return rc;
}
+static int pm8001_update_phy_mask(struct pm8001_hba_info *pm8001_ha)
+{
+ DECLARE_COMPLETION_ONSTACK(comp);
+ unsigned long timeout = msecs_to_jiffies(2000);
+ int ret = 0;
+ int i;
+ struct phy_status *phy_status =
+ kmalloc(sizeof(struct phy_status), GFP_KERNEL);
+ if (phy_status == NULL) {
+ pm8001_printk("Allocation failed\n");
+ return -ENOMEM;
+ }
+
+ for (i = 0; i < pm8001_ha->chip->n_phy; i++) {
+ struct pm8001_phy *phy = pm8001_ha->phy + i;
+
+ if (pm8001_ha->phy_mask & (1 << i) && phy->phy_state) {
+ ret = PM8001_CHIP_DISP->get_phy_profile_req(pm8001_ha,
+ i, SAS_PHY_GENERAL_STATUS_PAGE,
+ &comp, (void *)phy_status);
+ if (ret != 0) {
+ pm8001_printk("get_phy_profile_req:%d:%d\n",
+ i, ret);
+ return ret;
+ }
+ timeout = wait_for_completion_timeout(&comp, timeout);
+ if (timeout == 0)
+ return ret;
+ PM8001_DISC_DBG(pm8001_ha,
+ pm8001_printk("phy_id:%d state:%x nlr:%d plr:%d\n",
+ i,
+ phy_status->phy_state,
+ phy_status->nlr, phy_status->plr));
+ switch (phy_status->phy_state) {
+ case PHY_STATE_LINK_UP_SPCV:
+ case PHY_STATE_LINK_RESET:
+ case PHY_STATE_HARD_RESET:
+ /* these are handled by mpi_hw_event() */
+ break;
+ case PHY_LINK_DOWN:
+ if (time_is_after_jiffies(phy->start_timeout))
+ break;
+ clear_bit(i, &pm8001_ha->phy_mask);
+ break;
+ default:
+ clear_bit(i, &pm8001_ha->phy_mask);
+ break;
+ }
+ }
+ }
+ return ret;
+}
+
/**
* pm8001_scan_start - we should enable all HBA phys by sending the phy_start
* command to HBA.
@@ -274,6 +328,8 @@ void pm8001_scan_start(struct Scsi_Host *shost)
if (pm8001_ha->chip_id == chip_8001)
PM8001_CHIP_DISP->sas_re_init_req(pm8001_ha);
+ pm8001_ha->phy_mask = 0;
+
if (pm8001_ha->pdev->device == 0x8001 ||
pm8001_ha->pdev->device == 0x8081 ||
(pm8001_ha->spinup_interval != 0)) {
@@ -303,11 +359,16 @@ void pm8001_scan_start(struct Scsi_Host *shost)
int pm8001_scan_finished(struct Scsi_Host *shost, unsigned long time)
{
struct sas_ha_struct *ha = SHOST_TO_SAS_HA(shost);
+ struct pm8001_hba_info *pm8001_ha = ha->lldd_ha;
- /* give the phy enabling interrupt event time to come in (1s
- * is empirically about all it takes) */
- if (time < HZ)
+ /* Wait for PHY startup to finish */
+ PM8001_DISC_DBG(pm8001_ha,
+ pm8001_printk("mask:%lx\n", pm8001_ha->phy_mask));
+ if (pm8001_ha->phy_mask && pm8001_ha->wait_for_phy_startup) {
+ pm8001_update_phy_mask(pm8001_ha);
return 0;
+ }
+
/* Wait for discovery to finish */
sas_drain_work(ha);
return 1;
diff --git a/drivers/scsi/pm8001/pm8001_sas.h b/drivers/scsi/pm8001/pm8001_sas.h
index d2d73cba7f41..37c526be55bc 100644
--- a/drivers/scsi/pm8001/pm8001_sas.h
+++ b/drivers/scsi/pm8001/pm8001_sas.h
@@ -283,6 +283,7 @@ struct pm8001_phy {
struct completion *reset_completion;
bool port_reset_status;
bool reset_success;
+ unsigned long start_timeout;
};
/* port reset status */
@@ -578,6 +579,9 @@ struct pm8001_hba_info {
int phy_head;
int phy_tail;
spinlock_t phy_q_lock;
+ unsigned long phy_mask;
+ u32 phy_startup_timeout;
+ bool wait_for_phy_startup;
};
struct pm8001_work {
diff --git a/drivers/scsi/pm8001/pm80xx_hwi.c b/drivers/scsi/pm8001/pm80xx_hwi.c
index 748c7a06262f..91dc9d6d3657 100644
--- a/drivers/scsi/pm8001/pm80xx_hwi.c
+++ b/drivers/scsi/pm8001/pm80xx_hwi.c
@@ -80,18 +80,26 @@ void pm8001_spinup_timedout(struct timer_list *t)
if (i++ >= pm8001_ha->spinup_group && pm8001_ha->spinup_group)
break;
- if (pm8001_ha->phy_head == -1 || pm8001_ha->reset_in_progress)
- break; /* No phys to spinup */
-
- phy_id = pm8001_ha->phy_up[pm8001_ha->phy_head];
- /* Processed phy id, make it invalid 0xff for
- * checking repeated phy ups
- */
- pm8001_ha->phy_up[pm8001_ha->phy_head] = 0xff;
- if (pm8001_ha->phy_head == pm8001_ha->phy_tail) {
- pm8001_ha->phy_head = pm8001_ha->phy_tail = -1;
+ if (pm8001_ha->phy_head == -1 ||
+ pm8001_ha->reset_in_progress) {
+ /* No phys to spinup or timeout */
+ if (pm8001_ha->phy_mask) {
+ pm8001_printk("spinup timeout phy_mask:%lx\n",
+ pm8001_ha->phy_mask);
+ pm8001_ha->phy_mask = 0;
+ }
+ break;
} else {
- pm8001_ha->phy_head =
+ phy_id = pm8001_ha->phy_up[pm8001_ha->phy_head];
+ /* Processed phy id, make it invalid 0xff for
+ * checking repeated phy ups
+ */
+ pm8001_ha->phy_up[pm8001_ha->phy_head] = 0xff;
+ if (pm8001_ha->phy_head == pm8001_ha->phy_tail)
+ pm8001_ha->phy_head =
+ pm8001_ha->phy_tail = -1;
+ else
+ pm8001_ha->phy_head =
(pm8001_ha->phy_head+1) % PM8001_MAX_PHYS;
}
@@ -3584,8 +3592,10 @@ static int mpi_phy_start_resp(struct pm8001_hba_info *pm8001_ha, void *piomb)
phy->enable_completion = NULL;
}
- return 0;
+ if (!phy->phy_state)
+ clear_bit(phy_id, &pm8001_ha->phy_mask);
+ return 0;
}
/**
@@ -3646,6 +3656,36 @@ static int mpi_hw_event(struct pm8001_hba_info *pm8001_ha, void *piomb)
pm8001_printk("portid:%d phyid:%d event:0x%x status:0x%x\n",
port_id, phy_id, eventType, status));
+ switch (eventType) {
+ /* the PHY has completed startup and possibly queued
+ * stuff to the libsas side, so we no longer have to
+ * track that it is in the middle of doing something
+ * that libsas doesn't know about.
+ */
+ case HW_EVENT_SAS_PHY_UP:
+ case HW_EVENT_SATA_PHY_UP:
+ case HW_EVENT_PHY_DOWN:
+ case HW_EVENT_PORT_INVALID:
+ case HW_EVENT_BROADCAST_CHANGE:
+ case HW_EVENT_PHY_ERROR:
+ case HW_EVENT_HARD_RESET_RECEIVED:
+ case HW_EVENT_ID_FRAME_TIMEOUT:
+ case HW_EVENT_LINK_ERR_PHY_RESET_FAILED:
+ case HW_EVENT_PORT_RESET_TIMER_TMO:
+ case HW_EVENT_PORT_RECOVERY_TIMER_TMO:
+ case HW_EVENT_PORT_RECOVER:
+ clear_bit(phy_id, &pm8001_ha->phy_mask);
+ break;
+ /* we may have already gotten the response from the
+ * PHY, but if the PHY was powered off then this event
+ * signals that there is more in-flight and we have
+ * restarted the the spinup timer.
+ */
+ case HW_EVENT_SATA_SPINUP_HOLD:
+ set_bit(phy_id, &pm8001_ha->phy_mask);
+ break;
+ }
+
switch (eventType) {
case HW_EVENT_SAS_PHY_UP:
@@ -4975,6 +5015,9 @@ pm80xx_chip_phy_start_req(struct pm8001_hba_info *pm8001_ha, u8 phy_id)
PM8001_INIT_DBG(pm8001_ha,
pm8001_printk("PHY START REQ for phy_id %d\n", phy_id));
+ set_bit(phy_id, &pm8001_ha->phy_mask);
+ pm8001_ha->phy[phy_id].start_timeout =
+ jiffies + pm8001_ha->phy_startup_timeout;
payload.ase_sh_lm_slr_phyid = cpu_to_le32(SPINHOLD_ENABLE |
LINKMODE_AUTO | pm8001_ha->link_rate | phy_id);
--
2.16.3
^ permalink raw reply related
* [PATCH V2 2/3] pm80xx : Staggered spin up support.
From: Deepak Ukey @ 2020-06-24 12:03 UTC (permalink / raw)
To: linux-scsi
Cc: Vasanthalakshmi.Tharmarajan, Viswas.G, deepak.ukey, jinpu.wang,
martin.petersen, dpf, yuuzheng, auradkar, vishakhavc, bjashnani,
radha, akshatzen
In-Reply-To: <20200624120322.6265-1-deepak.ukey@microchip.com>
From: Viswas G <Viswas.G@microchip.com>
As a part of drive discovery, driver will initaite the drive spin up.
If all drives do spin up together, it will result in large power
consumption. To reduce the power consumption, driver provide an option
to make a small group of drives (say 3 or 4 drives together) to do the
spin up. The delay between two spin up group and no of drives to
spin up (group) can be programmed by the customer in seeprom and
driver will use it to control the spinup.
Signed-off-by: Viswas G <Viswas.G@microchip.com>
Signed-off-by: Radha Ramachandran <radha@google.com>
Signed-off-by: Deepak Ukey <Deepak.Ukey@microchip.com>
---
drivers/scsi/pm8001/pm8001_defs.h | 3 +
drivers/scsi/pm8001/pm8001_hwi.c | 14 ++++-
drivers/scsi/pm8001/pm8001_init.c | 53 +++++++++++++++-
drivers/scsi/pm8001/pm8001_sas.c | 36 ++++++++++-
drivers/scsi/pm8001/pm8001_sas.h | 13 ++++
drivers/scsi/pm8001/pm80xx_hwi.c | 125 +++++++++++++++++++++++++++++++++-----
6 files changed, 221 insertions(+), 23 deletions(-)
diff --git a/drivers/scsi/pm8001/pm8001_defs.h b/drivers/scsi/pm8001/pm8001_defs.h
index 1c7f15fd69ce..fd700ce5e80c 100644
--- a/drivers/scsi/pm8001/pm8001_defs.h
+++ b/drivers/scsi/pm8001/pm8001_defs.h
@@ -101,6 +101,9 @@ enum port_type {
#define USI_MAX_MEMCNT (PI + PM8001_MAX_SPCV_OUTB_NUM)
#define CONFIG_SCSI_PM8001_MAX_DMA_SG 528
#define PM8001_MAX_DMA_SG CONFIG_SCSI_PM8001_MAX_DMA_SG
+#define SPINUP_DELAY_OFFSET 0x890 /* 0x890 - delay */
+#define SPINUP_GROUP_OFFSET 0x892 /* 0x892 - group */
+#define PM80XX_MAX_SPINUP_DELAY 10000 /* 10000 ms */
enum memory_region_num {
AAP1 = 0x0, /* application acceleration processor */
IOP, /* IO processor */
diff --git a/drivers/scsi/pm8001/pm8001_hwi.c b/drivers/scsi/pm8001/pm8001_hwi.c
index fb9848e1d481..6378c8e8d6b2 100644
--- a/drivers/scsi/pm8001/pm8001_hwi.c
+++ b/drivers/scsi/pm8001/pm8001_hwi.c
@@ -3237,7 +3237,7 @@ int pm8001_mpi_local_phy_ctl(struct pm8001_hba_info *pm8001_ha, void *piomb)
(struct local_phy_ctl_resp *)(piomb + 4);
u32 status = le32_to_cpu(pPayload->status);
u32 phy_id = le32_to_cpu(pPayload->phyop_phyid) & ID_BITS;
- u32 phy_op = le32_to_cpu(pPayload->phyop_phyid) & OP_BITS;
+ u32 phy_op = (le32_to_cpu(pPayload->phyop_phyid) & OP_BITS) >> 8;
tag = le32_to_cpu(pPayload->tag);
if (status != 0) {
PM8001_MSG_DBG(pm8001_ha,
@@ -3248,6 +3248,13 @@ int pm8001_mpi_local_phy_ctl(struct pm8001_hba_info *pm8001_ha, void *piomb)
pm8001_printk("%x phy execute %x phy op success!\n",
phy_id, phy_op));
pm8001_ha->phy[phy_id].reset_success = true;
+ if (phy_op == PHY_NOTIFY_ENABLE_SPINUP &&
+ !pm8001_ha->reset_in_progress){
+ /* Notify the sas layer to discover
+ * the the whole sas domain
+ */
+ pm8001_bytes_dmaed(pm8001_ha, phy_id);
+ }
}
if (pm8001_ha->phy[phy_id].enable_completion) {
complete(pm8001_ha->phy[phy_id].enable_completion);
@@ -3643,7 +3650,10 @@ int pm8001_mpi_reg_resp(struct pm8001_hba_info *pm8001_ha, void *piomb)
pm8001_printk("DEVREG_FAILURE_DEVICE_TYPE_NOT_SUPPORTED\n"));
break;
}
- complete(pm8001_dev->dcompletion);
+ if (pm8001_dev->dcompletion) {
+ complete(pm8001_dev->dcompletion);
+ pm8001_dev->dcompletion = NULL;
+ }
ccb->task = NULL;
ccb->ccb_tag = 0xFFFFFFFF;
pm8001_tag_free(pm8001_ha, htag);
diff --git a/drivers/scsi/pm8001/pm8001_init.c b/drivers/scsi/pm8001/pm8001_init.c
index 38a65e1da823..46a2f86f3d9b 100644
--- a/drivers/scsi/pm8001/pm8001_init.c
+++ b/drivers/scsi/pm8001/pm8001_init.c
@@ -55,6 +55,12 @@ MODULE_PARM_DESC(link_rate, "Enable link rate.\n"
" 4: Link rate 6.0G\n"
" 8: Link rate 12.0G\n");
+bool staggered_spinup;
+module_param(staggered_spinup, bool, 0644);
+MODULE_PARM_DESC(staggered_spinup, "enable the staggered spinup feature.\n"
+ " 0/N: false\n"
+ " 1/Y: true\n");
+
static struct scsi_transport_template *pm8001_stt;
/**
@@ -164,7 +170,7 @@ static void pm8001_free(struct pm8001_hba_info *pm8001_ha)
if (!pm8001_ha)
return;
-
+ del_timer(&pm8001_ha->spinup_timer);
for (i = 0; i < USI_MAX_MEMCNT; i++) {
if (pm8001_ha->memoryMap.region[i].virt_ptr != NULL) {
dma_free_coherent(&pm8001_ha->pdev->dev,
@@ -488,6 +494,7 @@ static struct pm8001_hba_info *pm8001_pci_alloc(struct pci_dev *pdev,
pm8001_ha->shost = shost;
pm8001_ha->id = pm8001_id++;
pm8001_ha->logging_level = logging_level;
+ pm8001_ha->staggered_spinup = staggered_spinup;
pm8001_ha->non_fatal_count = 0;
if (link_rate >= 1 && link_rate <= 15)
pm8001_ha->link_rate = (link_rate << 8);
@@ -621,7 +628,8 @@ static void pm8001_post_sas_ha_init(struct Scsi_Host *shost,
* Currently we just set the fixed SAS address to our HBA,for manufacture,
* it should read from the EEPROM
*/
-static void pm8001_init_sas_add(struct pm8001_hba_info *pm8001_ha)
+static void pm8001_init_sas_add_and_spinup_config
+ (struct pm8001_hba_info *pm8001_ha)
{
u8 i, j;
u8 sas_add[8];
@@ -707,6 +715,45 @@ static void pm8001_init_sas_add(struct pm8001_hba_info *pm8001_ha)
memcpy(pm8001_ha->sas_addr, &pm8001_ha->phy[0].dev_sas_addr,
SAS_ADDR_SIZE);
#endif
+
+ /* For spinning up drives in group */
+ pm8001_ha->phy_head = -1;
+ pm8001_ha->phy_tail = -1;
+
+ for (i = 0; i < PM8001_MAX_PHYS; i++)
+ pm8001_ha->phy_up[i] = 0xff;
+
+ timer_setup(&pm8001_ha->spinup_timer,
+ (void *)pm8001_spinup_timedout, 0);
+
+ if (pm8001_ha->staggered_spinup == true) {
+ /* spinup interval in unit of 100 ms */
+ pm8001_ha->spinup_interval =
+ payload.func_specific[SPINUP_DELAY_OFFSET] * 100;
+ pm8001_ha->spinup_group =
+ payload.func_specific[SPINUP_GROUP_OFFSET];
+ } else {
+ pm8001_ha->spinup_interval = 0;
+ pm8001_ha->spinup_group = pm8001_ha->chip->n_phy;
+ }
+
+ if (pm8001_ha->spinup_interval > PM80XX_MAX_SPINUP_DELAY) {
+ PM8001_DISC_DBG(pm8001_ha, pm8001_printk(
+ "Spinup delay from Seeprom is %d ms, reset to %d ms\n",
+ pm8001_ha->spinup_interval * 100, PM80XX_MAX_SPINUP_DELAY));
+ pm8001_ha->spinup_interval = PM80XX_MAX_SPINUP_DELAY;
+ }
+
+ if (pm8001_ha->spinup_group > pm8001_ha->chip->n_phy) {
+ PM8001_DISC_DBG(pm8001_ha, pm8001_printk(
+ "Spinup group from Seeprom is %d, reset to %d\n",
+ pm8001_ha->spinup_group, pm8001_ha->chip->n_phy));
+ pm8001_ha->spinup_group = pm8001_ha->chip->n_phy;
+ }
+
+ PM8001_MSG_DBG(pm8001_ha, pm8001_printk(
+ "Spinup interval : %d Spinup group %d\n",
+ pm8001_ha->spinup_interval, pm8001_ha->spinup_group));
}
/*
@@ -1105,7 +1152,7 @@ static int pm8001_pci_probe(struct pci_dev *pdev,
pm80xx_set_thermal_config(pm8001_ha);
}
- pm8001_init_sas_add(pm8001_ha);
+ pm8001_init_sas_add_and_spinup_config(pm8001_ha);
/* phy setting support for motherboard controller */
if (pm8001_configure_phy_settings(pm8001_ha))
goto err_out_shost;
diff --git a/drivers/scsi/pm8001/pm8001_sas.c b/drivers/scsi/pm8001/pm8001_sas.c
index b7cbc312843e..cf1c5e68953b 100644
--- a/drivers/scsi/pm8001/pm8001_sas.c
+++ b/drivers/scsi/pm8001/pm8001_sas.c
@@ -265,14 +265,39 @@ int pm8001_phy_control(struct asd_sas_phy *sas_phy, enum phy_func func,
void pm8001_scan_start(struct Scsi_Host *shost)
{
int i;
+ unsigned long lock_flags;
struct pm8001_hba_info *pm8001_ha;
struct sas_ha_struct *sha = SHOST_TO_SAS_HA(shost);
+ DECLARE_COMPLETION(comp);
pm8001_ha = sha->lldd_ha;
/* SAS_RE_INITIALIZATION not available in SPCv/ve */
if (pm8001_ha->chip_id == chip_8001)
PM8001_CHIP_DISP->sas_re_init_req(pm8001_ha);
- for (i = 0; i < pm8001_ha->chip->n_phy; ++i)
- PM8001_CHIP_DISP->phy_start_req(pm8001_ha, i);
+
+ if (pm8001_ha->pdev->device == 0x8001 ||
+ pm8001_ha->pdev->device == 0x8081 ||
+ (pm8001_ha->spinup_interval != 0)) {
+ for (i = 0; i < pm8001_ha->chip->n_phy; ++i)
+ PM8001_CHIP_DISP->phy_start_req(pm8001_ha, i);
+ } else {
+ for (i = 0; i < pm8001_ha->chip->n_phy; ++i) {
+ spin_lock_irqsave(&pm8001_ha->lock, lock_flags);
+ pm8001_ha->phy_started = i;
+ pm8001_ha->scan_completion = ∁
+ pm8001_ha->phystart_timedout = 1;
+ spin_unlock_irqrestore(&pm8001_ha->lock, lock_flags);
+ PM8001_CHIP_DISP->phy_start_req(pm8001_ha, i);
+ wait_for_completion_timeout(&comp,
+ msecs_to_jiffies(500));
+ if (pm8001_ha->phystart_timedout)
+ PM8001_MSG_DBG(pm8001_ha, pm8001_printk(
+ "Timeout happened for phyid = %d\n", i));
+ }
+ spin_lock_irqsave(&pm8001_ha->lock, lock_flags);
+ pm8001_ha->phy_started = -1;
+ pm8001_ha->scan_completion = NULL;
+ spin_unlock_irqrestore(&pm8001_ha->lock, lock_flags);
+ }
}
int pm8001_scan_finished(struct Scsi_Host *shost, unsigned long time)
@@ -662,6 +687,13 @@ static int pm8001_dev_found_notify(struct domain_device *dev)
flag = 1; /* directly sata */
}
} /*register this device to HBA*/
+
+ if (pm8001_ha->phy_started == pm8001_device->attached_phy) {
+ if (pm8001_ha->scan_completion != NULL) {
+ pm8001_ha->phystart_timedout = 0;
+ complete(pm8001_ha->scan_completion);
+ }
+ }
PM8001_DISC_DBG(pm8001_ha, pm8001_printk("Found device\n"));
PM8001_CHIP_DISP->reg_dev_req(pm8001_ha, pm8001_device, flag);
spin_unlock_irqrestore(&pm8001_ha->lock, flags);
diff --git a/drivers/scsi/pm8001/pm8001_sas.h b/drivers/scsi/pm8001/pm8001_sas.h
index 488af79dec47..d2d73cba7f41 100644
--- a/drivers/scsi/pm8001/pm8001_sas.h
+++ b/drivers/scsi/pm8001/pm8001_sas.h
@@ -261,6 +261,7 @@ struct pm8001_port {
u8 port_attached;
u16 wide_port_phymap;
u8 port_state;
+ u8 port_id;
struct list_head list;
};
@@ -566,6 +567,17 @@ struct pm8001_hba_info {
u32 reset_in_progress;
u32 non_fatal_count;
u32 non_fatal_read_length;
+ bool staggered_spinup;
+ struct completion *scan_completion;
+ u32 phy_started;
+ u16 phystart_timedout;
+ int spinup_group;
+ int spinup_interval;
+ int phy_up[PM8001_MAX_PHYS];
+ struct timer_list spinup_timer;
+ int phy_head;
+ int phy_tail;
+ spinlock_t phy_q_lock;
};
struct pm8001_work {
@@ -684,6 +696,7 @@ void pm8001_open_reject_retry(
int pm8001_mem_alloc(struct pci_dev *pdev, void **virt_addr,
dma_addr_t *pphys_addr, u32 *pphys_addr_hi, u32 *pphys_addr_lo,
u32 mem_size, u32 align);
+void pm8001_spinup_timedout(struct timer_list *t);
void pm8001_chip_iounmap(struct pm8001_hba_info *pm8001_ha);
int pm8001_mpi_build_cmd(struct pm8001_hba_info *pm8001_ha,
diff --git a/drivers/scsi/pm8001/pm80xx_hwi.c b/drivers/scsi/pm8001/pm80xx_hwi.c
index a909d00c6897..748c7a06262f 100644
--- a/drivers/scsi/pm8001/pm80xx_hwi.c
+++ b/drivers/scsi/pm8001/pm80xx_hwi.c
@@ -46,6 +46,72 @@
#define SMP_DIRECT 1
#define SMP_INDIRECT 2
+static int pm80xx_chip_phy_ctl_req(struct pm8001_hba_info *pm8001_ha,
+ u32 phyId, u32 phy_op);
+
+void pm8001_queue_phyup(struct pm8001_hba_info *pm8001_ha, int phy_id)
+{
+ int i;
+
+ if (pm8001_ha->phy_head == -1) {
+ pm8001_ha->phy_head = pm8001_ha->phy_tail = 0;
+ } else {
+ /* If the phy id is already queued , discard the phy up */
+ for (i = 0; i < pm8001_ha->chip->n_phy; i++)
+ if (pm8001_ha->phy_up[i] == phy_id)
+ return;
+ pm8001_ha->phy_tail =
+ (pm8001_ha->phy_tail + 1) % PM8001_MAX_PHYS;
+ }
+ pm8001_ha->phy_up[pm8001_ha->phy_tail] = phy_id;
+}
+
+void pm8001_spinup_timedout(struct timer_list *t)
+{
+ struct pm8001_hba_info *pm8001_ha =
+ from_timer(pm8001_ha, t, spinup_timer);
+ struct pm8001_phy *phy;
+ unsigned long flags;
+ int i = 0, phy_id = 0xff;
+
+ spin_lock_irqsave(&pm8001_ha->phy_q_lock, flags);
+
+ do {
+ if (i++ >= pm8001_ha->spinup_group && pm8001_ha->spinup_group)
+ break;
+
+ if (pm8001_ha->phy_head == -1 || pm8001_ha->reset_in_progress)
+ break; /* No phys to spinup */
+
+ phy_id = pm8001_ha->phy_up[pm8001_ha->phy_head];
+ /* Processed phy id, make it invalid 0xff for
+ * checking repeated phy ups
+ */
+ pm8001_ha->phy_up[pm8001_ha->phy_head] = 0xff;
+ if (pm8001_ha->phy_head == pm8001_ha->phy_tail) {
+ pm8001_ha->phy_head = pm8001_ha->phy_tail = -1;
+ } else {
+ pm8001_ha->phy_head =
+ (pm8001_ha->phy_head+1) % PM8001_MAX_PHYS;
+ }
+
+ if (phy_id == 0xff)
+ break;
+ phy = &pm8001_ha->phy[phy_id];
+ if (phy->phy_type & PORT_TYPE_SAS) {
+ PM8001_CHIP_DISP->phy_ctl_req(pm8001_ha, phy_id,
+ PHY_NOTIFY_ENABLE_SPINUP);
+ } else {
+ PM8001_CHIP_DISP->phy_ctl_req(pm8001_ha, phy_id,
+ PHY_LINK_RESET);
+ }
+ } while (1);
+
+ if (pm8001_ha->phy_head != -1 && pm8001_ha->spinup_group)
+ mod_timer(&pm8001_ha->spinup_timer,
+ jiffies + msecs_to_jiffies(pm8001_ha->spinup_interval));
+ spin_unlock_irqrestore(&pm8001_ha->phy_q_lock, flags);
+}
int pm80xx_bar4_shift(struct pm8001_hba_info *pm8001_ha, u32 shift_value)
{
@@ -3302,11 +3368,12 @@ hw_event_sas_phy_up(struct pm8001_hba_info *pm8001_ha, void *piomb)
port->port_state = portstate;
port->wide_port_phymap |= (1U << phy_id);
phy->phy_state = PHY_STATE_LINK_UP_SPCV;
+ phy->port = port;
PM8001_MSG_DBG(pm8001_ha, pm8001_printk(
"portid:%d; phyid:%d; linkrate:%d; "
"portstate:%x; devicetype:%x\n",
port_id, phy_id, link_rate, portstate, deviceType));
-
+ port->port_id = port_id;
switch (deviceType) {
case SAS_PHY_UNUSED:
PM8001_MSG_DBG(pm8001_ha,
@@ -3314,8 +3381,12 @@ hw_event_sas_phy_up(struct pm8001_hba_info *pm8001_ha, void *piomb)
break;
case SAS_END_DEVICE:
PM8001_MSG_DBG(pm8001_ha, pm8001_printk("end device.\n"));
- pm80xx_chip_phy_ctl_req(pm8001_ha, phy_id,
- PHY_NOTIFY_ENABLE_SPINUP);
+ spin_lock_irqsave(&pm8001_ha->phy_q_lock, flags);
+ pm8001_queue_phyup(pm8001_ha, phy_id);
+ spin_unlock_irqrestore(&pm8001_ha->phy_q_lock, flags);
+ if (!timer_pending(&pm8001_ha->spinup_timer))
+ mod_timer(&pm8001_ha->spinup_timer,
+ jiffies + msecs_to_jiffies(pm8001_ha->spinup_interval));
port->port_attached = 1;
pm8001_get_lrate_mode(phy, link_rate);
break;
@@ -3351,9 +3422,10 @@ hw_event_sas_phy_up(struct pm8001_hba_info *pm8001_ha, void *piomb)
phy->frame_rcvd_size = sizeof(struct sas_identify_frame) - 4;
pm8001_get_attached_sas_addr(phy, phy->sas_phy.attached_sas_addr);
spin_unlock_irqrestore(&phy->sas_phy.frame_rcvd_lock, flags);
- if (pm8001_ha->flags == PM8001F_RUN_TIME)
- msleep(200);/*delay a moment to wait disk to spinup*/
- pm8001_bytes_dmaed(pm8001_ha, phy_id);
+ if (!pm8001_ha->reset_in_progress) {
+ if (deviceType != SAS_END_DEVICE)
+ pm8001_bytes_dmaed(pm8001_ha, phy_id);
+ }
}
/**
@@ -3388,11 +3460,17 @@ hw_event_sata_phy_up(struct pm8001_hba_info *pm8001_ha, void *piomb)
port->port_state = portstate;
phy->phy_state = PHY_STATE_LINK_UP_SPCV;
port->port_attached = 1;
+ phy->port = port;
+ port->port_id = port_id;
pm8001_get_lrate_mode(phy, link_rate);
phy->phy_type |= PORT_TYPE_SATA;
phy->phy_attached = 1;
phy->sas_phy.oob_mode = SATA_OOB_MODE;
- sas_ha->notify_phy_event(&phy->sas_phy, PHYE_OOB_DONE);
+ if (!pm8001_ha->reset_in_progress) {
+ sas_ha->notify_phy_event(&phy->sas_phy, PHYE_OOB_DONE);
+ } else
+ PM8001_MSG_DBG(pm8001_ha, pm8001_printk(
+ "HW_EVENT_PHY_UP: not notified to host\n"));
spin_lock_irqsave(&phy->sas_phy.frame_rcvd_lock, flags);
memcpy(phy->frame_rcvd, ((u8 *)&pPayload->sata_fis - 4),
sizeof(struct dev_to_host_fis));
@@ -3401,7 +3479,8 @@ hw_event_sata_phy_up(struct pm8001_hba_info *pm8001_ha, void *piomb)
phy->identify.device_type = SAS_SATA_DEV;
pm8001_get_attached_sas_addr(phy, phy->sas_phy.attached_sas_addr);
spin_unlock_irqrestore(&phy->sas_phy.frame_rcvd_lock, flags);
- pm8001_bytes_dmaed(pm8001_ha, phy_id);
+ if (!pm8001_ha->reset_in_progress)
+ pm8001_bytes_dmaed(pm8001_ha, phy_id);
}
/**
@@ -3497,12 +3576,14 @@ static int mpi_phy_start_resp(struct pm8001_hba_info *pm8001_ha, void *piomb)
status, phy_id));
if (status == 0) {
phy->phy_state = PHY_LINK_DOWN;
- if (pm8001_ha->flags == PM8001F_RUN_TIME &&
- phy->enable_completion != NULL) {
- complete(phy->enable_completion);
- phy->enable_completion = NULL;
- }
}
+
+ if (pm8001_ha->flags == PM8001F_RUN_TIME &&
+ phy->enable_completion != NULL) {
+ complete(phy->enable_completion);
+ phy->enable_completion = NULL;
+ }
+
return 0;
}
@@ -3580,7 +3661,14 @@ static int mpi_hw_event(struct pm8001_hba_info *pm8001_ha, void *piomb)
case HW_EVENT_SATA_SPINUP_HOLD:
PM8001_MSG_DBG(pm8001_ha,
pm8001_printk("HW_EVENT_SATA_SPINUP_HOLD\n"));
- sas_ha->notify_phy_event(&phy->sas_phy, PHYE_SPINUP_HOLD);
+ spin_lock_irqsave(&pm8001_ha->phy_q_lock, flags);
+ pm8001_queue_phyup(pm8001_ha, phy_id);
+ spin_unlock_irqrestore(&pm8001_ha->phy_q_lock, flags);
+
+ /* Start the timer if not started */
+ if (!timer_pending(&pm8001_ha->spinup_timer))
+ mod_timer(&pm8001_ha->spinup_timer,
+ jiffies + msecs_to_jiffies(pm8001_ha->spinup_interval));
break;
case HW_EVENT_PHY_DOWN:
PM8001_MSG_DBG(pm8001_ha,
@@ -4888,7 +4976,7 @@ pm80xx_chip_phy_start_req(struct pm8001_hba_info *pm8001_ha, u8 phy_id)
PM8001_INIT_DBG(pm8001_ha,
pm8001_printk("PHY START REQ for phy_id %d\n", phy_id));
- payload.ase_sh_lm_slr_phyid = cpu_to_le32(SPINHOLD_DISABLE |
+ payload.ase_sh_lm_slr_phyid = cpu_to_le32(SPINHOLD_ENABLE |
LINKMODE_AUTO | pm8001_ha->link_rate | phy_id);
/* SSC Disable and SAS Analog ST configuration */
/**
@@ -4950,6 +5038,8 @@ static int pm80xx_chip_reg_dev_req(struct pm8001_hba_info *pm8001_ha,
u16 ITNT = 2000;
struct domain_device *dev = pm8001_dev->sas_device;
struct domain_device *parent_dev = dev->parent;
+ struct pm8001_phy *phy;
+ struct pm8001_port *port;
circularQ = &pm8001_ha->inbnd_q_tbl[0];
memset(&payload, 0, sizeof(payload));
@@ -4981,8 +5071,11 @@ static int pm80xx_chip_reg_dev_req(struct pm8001_hba_info *pm8001_ha,
linkrate = (pm8001_dev->sas_device->linkrate < dev->port->linkrate) ?
pm8001_dev->sas_device->linkrate : dev->port->linkrate;
+ phy = &pm8001_ha->phy[phy_id];
+ port = phy->port;
+
payload.phyid_portid =
- cpu_to_le32(((pm8001_dev->sas_device->port->id) & 0xFF) |
+ cpu_to_le32(((port->port_id) & 0xFF) |
((phy_id & 0xFF) << 8));
payload.dtype_dlr_mcn_ir_retry = cpu_to_le32((retryFlag & 0x01) |
--
2.16.3
^ permalink raw reply related
* [PATCH V2 1/3] pm80xx : Support for get phy profile functionality.
From: Deepak Ukey @ 2020-06-24 12:03 UTC (permalink / raw)
To: linux-scsi
Cc: Vasanthalakshmi.Tharmarajan, Viswas.G, deepak.ukey, jinpu.wang,
martin.petersen, dpf, yuuzheng, auradkar, vishakhavc, bjashnani,
radha, akshatzen
In-Reply-To: <20200624120322.6265-1-deepak.ukey@microchip.com>
From: Viswas G <Viswas.G@microchip.com>
Added the support to get the phy profile which gives information
about the phy states, port and errors on phy.
Signed-off-by: Deepak Ukey <deepak.ukey@microchip.com>
Signed-off-by: Viswas G <Viswas.G@microchip.com>
Signed-off-by: Radha Ramachandran <radha@google.com>
---
drivers/scsi/pm8001/pm8001_ctl.h | 20 ++++++++++
drivers/scsi/pm8001/pm8001_init.c | 2 +
drivers/scsi/pm8001/pm8001_sas.h | 5 +++
drivers/scsi/pm8001/pm80xx_hwi.c | 83 ++++++++++++++++++++++++++++++++++++++-
drivers/scsi/pm8001/pm80xx_hwi.h | 2 +
5 files changed, 110 insertions(+), 2 deletions(-)
diff --git a/drivers/scsi/pm8001/pm8001_ctl.h b/drivers/scsi/pm8001/pm8001_ctl.h
index d0d43a250b9e..02dc04a1fbe8 100644
--- a/drivers/scsi/pm8001/pm8001_ctl.h
+++ b/drivers/scsi/pm8001/pm8001_ctl.h
@@ -41,6 +41,26 @@
#ifndef PM8001_CTL_H_INCLUDED
#define PM8001_CTL_H_INCLUDED
+struct phy_status {
+ char phy_id;
+ unsigned int phy_state:4;
+ unsigned int nlr:4;
+ unsigned int plr:4;
+ unsigned int reserved1:12;
+ unsigned char port_id;
+ unsigned int prts:4;
+ unsigned int reserved2:20;
+} __packed;
+
+struct phy_errcnt {
+ unsigned int InvalidDword;
+ unsigned int runningDisparityError;
+ unsigned int codeViolation;
+ unsigned int LossOfSyncDW;
+ unsigned int phyResetProblem;
+ unsigned int inboundCRCError;
+};
+
#define IOCTL_BUF_SIZE 4096
#define HEADER_LEN 28
#define SIZE_OFFSET 16
diff --git a/drivers/scsi/pm8001/pm8001_init.c b/drivers/scsi/pm8001/pm8001_init.c
index a8f5344fdfda..38a65e1da823 100644
--- a/drivers/scsi/pm8001/pm8001_init.c
+++ b/drivers/scsi/pm8001/pm8001_init.c
@@ -397,6 +397,8 @@ static int pm8001_alloc(struct pm8001_hba_info *pm8001_ha,
pm8001_ha->ccb_info[i].task = NULL;
pm8001_ha->ccb_info[i].ccb_tag = 0xffffffff;
pm8001_ha->ccb_info[i].device = NULL;
+ pm8001_ha->ccb_info[i].completion = NULL;
+ pm8001_ha->ccb_info[i].resp_buf = NULL;
++pm8001_ha->tags_num;
}
pm8001_ha->flags = PM8001F_INIT_TIME;
diff --git a/drivers/scsi/pm8001/pm8001_sas.h b/drivers/scsi/pm8001/pm8001_sas.h
index ae7ba9b3c4bc..488af79dec47 100644
--- a/drivers/scsi/pm8001/pm8001_sas.h
+++ b/drivers/scsi/pm8001/pm8001_sas.h
@@ -56,6 +56,7 @@
#include <scsi/sas_ata.h>
#include <linux/atomic.h>
#include "pm8001_defs.h"
+#include "pm8001_ctl.h"
#define DRV_NAME "pm80xx"
#define DRV_VERSION "0.1.39"
@@ -244,6 +245,8 @@ struct pm8001_dispatch {
int (*sas_diag_execute_req)(struct pm8001_hba_info *pm8001_ha,
u32 state);
int (*sas_re_init_req)(struct pm8001_hba_info *pm8001_ha);
+ int (*get_phy_profile_req)(struct pm8001_hba_info *pm8001_ha,
+ int phy, int page, struct completion *comp, void *buf);
};
struct pm8001_chip_info {
@@ -318,6 +321,8 @@ struct pm8001_ccb_info {
struct pm8001_prd buf_prd[PM8001_MAX_DMA_SG];
struct fw_control_ex *fw_control_context;
u8 open_retry;
+ struct completion *completion;
+ void *resp_buf;
};
struct mpi_mem {
diff --git a/drivers/scsi/pm8001/pm80xx_hwi.c b/drivers/scsi/pm8001/pm80xx_hwi.c
index 4d205ebaee87..a909d00c6897 100644
--- a/drivers/scsi/pm8001/pm80xx_hwi.c
+++ b/drivers/scsi/pm8001/pm80xx_hwi.c
@@ -3792,7 +3792,6 @@ static int mpi_set_controller_config_resp(struct pm8001_hba_info *pm8001_ha,
PM8001_MSG_DBG(pm8001_ha, pm8001_printk(
"SET CONTROLLER RESP: status 0x%x qlfr_pgcd 0x%x\n",
status, err_qlfr_pgcd));
-
return 0;
}
@@ -3818,9 +3817,58 @@ static int mpi_get_controller_config_resp(struct pm8001_hba_info *pm8001_ha,
static int mpi_get_phy_profile_resp(struct pm8001_hba_info *pm8001_ha,
void *piomb)
{
+ u32 tag, page_code;
+ struct phy_status *phy_status, *phy_stat;
+ struct phy_errcnt *phy_err, *phy_err_cnt;
+ struct pm8001_ccb_info *ccb;
+ struct get_phy_profile_resp *pPayload =
+ (struct get_phy_profile_resp *)(piomb + 4);
+ u32 status = le32_to_cpu(pPayload->status);
+
+ page_code = (u8)((pPayload->ppc_phyid & 0xFF00) >> 8);
+
PM8001_MSG_DBG(pm8001_ha,
- pm8001_printk(" pm80xx_addition_functionality\n"));
+ pm8001_printk(" pm80xx_addition_functionality\n"));
+ if (status) {
+ /* status is FAILED */
+ PM8001_FAIL_DBG(pm8001_ha, pm8001_printk(
+ "mpiGetPhyProfileReq failed with status 0x%08x\n",
+ status));
+ }
+ tag = le32_to_cpu(pPayload->tag);
+ ccb = &pm8001_ha->ccb_info[tag];
+ if (ccb->completion != NULL) {
+ if (status) {
+ /* signal fail status */
+ memset(&ccb->resp_buf, 0xff, sizeof(ccb->resp_buf));
+ } else if (page_code == SAS_PHY_GENERAL_STATUS_PAGE) {
+ phy_status = (struct phy_status *)ccb->resp_buf;
+ phy_stat =
+ (struct phy_status *)pPayload->ppc_specific_rsp;
+ phy_status->phy_id = le32_to_cpu(phy_stat->phy_id);
+ phy_status->phy_state =
+ le32_to_cpu(phy_stat->phy_state);
+ phy_status->plr = le32_to_cpu(phy_stat->plr);
+ phy_status->nlr = le32_to_cpu(phy_stat->nlr);
+ phy_status->port_id = le32_to_cpu(phy_stat->port_id);
+ phy_status->prts = le32_to_cpu(phy_stat->prts);
+ } else if (page_code == SAS_PHY_ERR_COUNTERS_PAGE) {
+ phy_err = (struct phy_errcnt *)ccb->resp_buf;
+ phy_err_cnt =
+ (struct phy_errcnt *)pPayload->ppc_specific_rsp;
+ phy_err->InvalidDword =
+ le32_to_cpu(phy_err_cnt->InvalidDword);
+ phy_err->runningDisparityError =
+ le32_to_cpu(phy_err_cnt->runningDisparityError);
+ phy_err->LossOfSyncDW =
+ le32_to_cpu(phy_err_cnt->LossOfSyncDW);
+ phy_err->phyResetProblem =
+ le32_to_cpu(phy_err_cnt->phyResetProblem);
+ }
+ complete(ccb->completion);
+ }
+ pm8001_tag_free(pm8001_ha, tag);
return 0;
}
@@ -5013,6 +5061,36 @@ pm80xx_chip_isr(struct pm8001_hba_info *pm8001_ha, u8 vec)
return IRQ_HANDLED;
}
+int pm8001_chip_get_phy_profile(struct pm8001_hba_info *pm8001_ha, int phy_id,
+ int page_code, struct completion *comp, void *buf)
+{
+ u32 tag;
+ struct get_phy_profile_req payload;
+ struct inbound_queue_table *circularQ;
+ struct pm8001_ccb_info *ccb;
+ int rc, ppc_phyid;
+ u32 opc = OPC_INB_GET_PHY_PROFILE;
+
+ memset(&payload, 0, sizeof(payload));
+
+ rc = pm8001_tag_alloc(pm8001_ha, &tag);
+ if (rc)
+ PM8001_FAIL_DBG(pm8001_ha, pm8001_printk("Invalid tag\n"));
+
+ ccb = &pm8001_ha->ccb_info[tag];
+ ccb->completion = comp;
+ ccb->resp_buf = buf;
+ circularQ = &pm8001_ha->inbnd_q_tbl[0];
+
+ payload.tag = cpu_to_le32(tag);
+ ppc_phyid = (page_code & 0xFF) << 8 | (phy_id & 0xFF);
+ payload.ppc_phyid = cpu_to_le32(ppc_phyid);
+
+ pm8001_mpi_build_cmd(pm8001_ha, circularQ, opc, &payload,
+ sizeof(payload), 0);
+ return rc;
+}
+
void mpi_set_phy_profile_req(struct pm8001_hba_info *pm8001_ha,
u32 operation, u32 phyid, u32 length, u32 *buf)
{
@@ -5113,4 +5191,5 @@ const struct pm8001_dispatch pm8001_80xx_dispatch = {
.set_nvmd_req = pm8001_chip_set_nvmd_req,
.fw_flash_update_req = pm8001_chip_fw_flash_update_req,
.set_dev_state_req = pm8001_chip_set_dev_state_req,
+ .get_phy_profile_req = pm8001_chip_get_phy_profile,
};
diff --git a/drivers/scsi/pm8001/pm80xx_hwi.h b/drivers/scsi/pm8001/pm80xx_hwi.h
index 701951a0f715..b5119c5479da 100644
--- a/drivers/scsi/pm8001/pm80xx_hwi.h
+++ b/drivers/scsi/pm8001/pm80xx_hwi.h
@@ -175,7 +175,9 @@
#define PHY_STOP_ERR_DEVICE_ATTACHED 0x1046
/* phy_profile */
+#define SAS_PHY_ERR_COUNTERS_PAGE 0x01
#define SAS_PHY_ANALOG_SETTINGS_PAGE 0x04
+#define SAS_PHY_GENERAL_STATUS_PAGE 0x05
#define PHY_DWORD_LENGTH 0xC
/* Thermal related */
--
2.16.3
^ permalink raw reply related
* [PATCH V2 0/3] pm80xx : Updates for the driver version 0.1.39.
From: Deepak Ukey @ 2020-06-24 12:03 UTC (permalink / raw)
To: linux-scsi
Cc: Vasanthalakshmi.Tharmarajan, Viswas.G, deepak.ukey, jinpu.wang,
martin.petersen, dpf, yuuzheng, auradkar, vishakhavc, bjashnani,
radha, akshatzen
From: Deepak Ukey <Deepak.Ukey@microchip.com>
This patch set includes some bug fixes and features for pm80xx driver.
Changes from V1:
For "Support for get phy profile functionality."
-Initalized the completion and removed the phy_prof_resp
structure.
For "Staggered spin up support."
-Removed the lock_flags from pm8001_hba_info structure.
For "Wait for PHY startup before draining libsas queue."
-Moved the completion initalization in first patch.
-Added module parameter to wait for phy startup.
Viswas G (2):
pm80xx : Support for get phy profile functionality.
pm80xx : Staggered spin up support.
peter chang (1):
pm80xx : Wait for PHY startup before draining libsas queue.
drivers/scsi/pm8001/pm8001_ctl.c | 36 ++++++
drivers/scsi/pm8001/pm8001_ctl.h | 20 +++
drivers/scsi/pm8001/pm8001_defs.h | 9 +-
drivers/scsi/pm8001/pm8001_hwi.c | 14 ++-
drivers/scsi/pm8001/pm8001_init.c | 77 +++++++++++-
drivers/scsi/pm8001/pm8001_sas.c | 103 +++++++++++++++-
drivers/scsi/pm8001/pm8001_sas.h | 22 ++++
drivers/scsi/pm8001/pm80xx_hwi.c | 253 +++++++++++++++++++++++++++++++++++---
drivers/scsi/pm8001/pm80xx_hwi.h | 2 +
9 files changed, 505 insertions(+), 31 deletions(-)
--
2.16.3
^ permalink raw reply
* Re: [RFC PATCH v3 5/8] ata_dev_printk: Use dev_printk
From: Bartlomiej Zolnierkiewicz @ 2020-06-24 10:35 UTC (permalink / raw)
To: Tony Asleson; +Cc: linux-scsi, linux-block, linux-ide
In-Reply-To: <20200623191749.115200-6-tasleson@redhat.com>
[ added linux-ide ML to Cc: ]
Hi,
On 6/23/20 9:17 PM, Tony Asleson wrote:
> Utilize the dev_printk function which will add structured data
> to the log message.
>
> Signed-off-by: Tony Asleson <tasleson@redhat.com>
> ---
> drivers/ata/libata-core.c | 10 +++++++---
> 1 file changed, 7 insertions(+), 3 deletions(-)
>
> diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c
> index beca5f91bb4c..44c874367fe3 100644
> --- a/drivers/ata/libata-core.c
> +++ b/drivers/ata/libata-core.c
> @@ -6475,6 +6475,7 @@ EXPORT_SYMBOL(ata_link_printk);
> void ata_dev_printk(const struct ata_device *dev, const char *level,
> const char *fmt, ...)
> {
> + const struct device *gendev;
> struct va_format vaf;
> va_list args;
>
> @@ -6483,9 +6484,12 @@ void ata_dev_printk(const struct ata_device *dev, const char *level,
> vaf.fmt = fmt;
> vaf.va = &args;
>
> - printk("%sata%u.%02u: %pV",
> - level, dev->link->ap->print_id, dev->link->pmp + dev->devno,
> - &vaf);
> + gendev = (dev->sdev) ? &dev->sdev->sdev_gendev : &dev->tdev;
> +
> + dev_printk(level, gendev, "ata%u.%02u: %pV",
> + dev->link->ap->print_id,
This duplicates the device information and breaks integrity of
libata logging functionality (ata_{dev,link,port}_printk() should
be all converted to use dev_printk() at the same time).
The root source of problem is that libata transport uses different
naming scheme for ->tdev devices (please see dev_set_name() in
ata_t{dev,link,port}_add()) than libata core for its logging
functionality (ata_{dev,link,port}_printk()).
Since libata transport is part of sysfs ABI we should be careful
to not break it so one idea for solving the issue is to convert
ata_t{dev,link,port}_add() to use libata logging naming scheme and
at the same time add sysfs symlinks for the old libata transport
naming scheme.
dev->sdev usage is not required for dev_printk() conversion and
should be considered as a separate change (since it also breaks
integrity of libata logging and can be considered as a mild
"layering violation" I don't think that it should be applied).
> + dev->link->pmp + dev->devno,
> + &vaf);
>
> va_end(args);
> }
>
Best regards,
--
Bartlomiej Zolnierkiewicz
Samsung R&D Institute Poland
Samsung Electronics
^ permalink raw reply
* Re: [PATCH v10 00/10] exynos-ufs: Add support for UFS HCI
From: Vinod Koul @ 2020-06-24 10:21 UTC (permalink / raw)
To: Alim Akhtar
Cc: 'Kishon Vijay Abraham I', robh, krzk, linux-samsung-soc,
avri.altman, stanley.chu, linux-scsi, linux-arm-kernel, cang,
devicetree, kwmad.kim, linux-kernel, 'Martin K. Petersen'
In-Reply-To: <000001d646a6$6cb5fd70$4621f850$@samsung.com>
On 20-06-20, 07:29, Alim Akhtar wrote:
> Hi Kishon,
>
> > -----Original Message-----
> > From: Alim Akhtar <alim.akhtar@samsung.com>
> > Sent: 11 June 2020 20:49
> > To: 'Kishon Vijay Abraham I' <kishon@ti.com>; 'Martin K. Petersen'
> > > >>
> > > >> Applied [1,2,3,4,5,9] to 5.9/scsi-queue. The series won't show up
> > > >> in my
> > > > public
> > > >> tree until shortly after -rc1 is released.
> > > >>
> > > > Thanks Martin,
> > > > Hi Rob and Kishon/Vinod
> > > > Can you please pickup dt-bindings and PHY driver respectively?
> > >
> > > You might have CC'ed me only for the PHY patch. I don't have the
> > > dt-bindings in my inbox. Care to re-send what's missing again? This
> > > will be merged after -rc1 is tagged.
> > >
>
> -rc1 is out, I do not see phy driver patch in your tree[1] yet, let me know if I am looking into right tree.
> [1] -> git://git.kernel.org/pub/scm/linux/kernel/git/phy/linux-phy.git
Right tree
>
> Thanks!
>
> > Sure, will re-send this series.
But patches have not been sent right, pls send and me/Kishon will review
Thanks
--
~Vinod
^ permalink raw reply
* RE: [RFC PATCH v3 4/5] scsi: ufs: L2P map management for HPB read
From: Avri Altman @ 2020-06-24 9:08 UTC (permalink / raw)
To: daejun7.park@samsung.com, jejb@linux.ibm.com,
martin.petersen@oracle.com, asutoshd@codeaurora.org,
stanley.chu@mediatek.com, cang@codeaurora.org, huobean@gmail.com,
bvanassche@acm.org, tomas.winkler@intel.com, ALIM AKHTAR
Cc: linux-scsi@vger.kernel.org, linux-kernel@vger.kernel.org,
Sang-yoon Oh, Sung-Jun Park, yongmyung lee, Jinyoung CHOI,
Adel Choi, BoRam Shin
In-Reply-To: <231786897.01592886181765.JavaMail.epsvc@epcpadp2>
> static struct ufshpb_driver ufshpb_drv;
> +unsigned int ufshpb_host_map_kbytes = 1 * 1024;
I think you've already declared this as a module parameter in 3/5.
No need to fix this now, unless there will be some more comments,
And you'll issue a v4.
Thanks,
Avri
^ permalink raw reply
* [Bug 208045] ARM ubuntu 18.04 as the iscsi server, using initiator login, the kernel crashes
From: bugzilla-daemon @ 2020-06-24 8:54 UTC (permalink / raw)
To: linux-scsi
In-Reply-To: <bug-208045-11613@https.bugzilla.kernel.org/>
https://bugzilla.kernel.org/show_bug.cgi?id=208045
--- Comment #15 from Bodo Stroesser (bstroesser@ts.fujitsu.com) ---
I just sent the patch to linux-scsi and target-devel.
--
You are receiving this mail because:
You are watching the assignee of the bug.
^ permalink raw reply
* [PATCH] scsi: target: tcmu: Fix crash on ARM during cmd completion
From: Bodo Stroesser @ 2020-06-24 8:53 UTC (permalink / raw)
To: Martin K. Petersen, Mike Christie, linux-scsi, target-devel
Cc: JiangYu, Daniel Meyerholt, Henry Willard, Bodo Stroesser
If tcmu_handle_completions() has to process a padding shorter
than sizeof(struct tcmu_cmd_entry), the current call to
tcmu_flush_dcache_range() with sizeof(struct tcmu_cmd_entry) as
length param is wrong and causes crashes on e.g. ARM, because
tcmu_flush_dcache_range() in this case calls
flush_dcache_page(vmalloc_to_page(start));
with start being an invalid address above the end of the
vmalloc'ed area.
The fix is to use the maximum of remaining ring space and
sizeof(struct tcmu_cmd_entry) as the length param.
The patch was tested on kernel 4.19.118.
See https://bugzilla.kernel.org/show_bug.cgi?id=208045#c10
Signed-off-by: Bodo Stroesser <bstroesser@ts.fujitsu.com>
Tested-by: JiangYu <lnsyyj@hotmail.com>
---
drivers/target/target_core_user.c | 9 ++++++++-
1 file changed, 8 insertions(+), 1 deletion(-)
diff --git a/drivers/target/target_core_user.c b/drivers/target/target_core_user.c
index 3885ca532f8f..82e476d48194 100644
--- a/drivers/target/target_core_user.c
+++ b/drivers/target/target_core_user.c
@@ -1221,7 +1221,14 @@ static unsigned int tcmu_handle_completions(struct tcmu_dev *udev)
struct tcmu_cmd_entry *entry = (void *) mb + CMDR_OFF + udev->cmdr_last_cleaned;
- tcmu_flush_dcache_range(entry, sizeof(*entry));
+ /*
+ * Flush max. up to end of cmd ring, since current entry might
+ * be a padding that is shorter than sizeof(*entry)
+ */
+ size_t ring_left = head_to_end(udev->cmdr_last_cleaned,
+ udev->cmdr_size);
+ tcmu_flush_dcache_range(entry, ring_left < sizeof(*entry) ?
+ ring_left : sizeof(*entry));
if (tcmu_hdr_get_op(entry->hdr.len_op) == TCMU_OP_PAD) {
UPDATE_HEAD(udev->cmdr_last_cleaned,
--
2.12.3
^ permalink raw reply related
* RE: [PATCH RFC v7 02/12] blk-mq: rename blk_mq_update_tag_set_depth()
From: Kashyap Desai @ 2020-06-24 8:13 UTC (permalink / raw)
To: Hannes Reinecke, John Garry, Ming Lei
Cc: axboe, jejb, martin.petersen, don.brace, Sumit Saxena, bvanassche,
hare, hch, Shivasharan Srikanteshwara, linux-block, linux-scsi,
esc.storagedev, chenxiang66
In-Reply-To: <eaf188d5-dac0-da44-1c83-31ff2860d8fa@suse.de>
>
> On 6/23/20 1:25 PM, John Garry wrote:
> > On 11/06/2020 09:26, John Garry wrote:
> >> On 11/06/2020 03:57, Ming Lei wrote:
> >>> On Thu, Jun 11, 2020 at 01:29:09AM +0800, John Garry wrote:
> >>>> From: Hannes Reinecke <hare@suse.de>
> >>>>
> >>>> The function does not set the depth, but rather transitions from
> >>>> shared to non-shared queues and vice versa.
> >>>> So rename it to blk_mq_update_tag_set_shared() to better reflect
> >>>> its purpose.
> >>>
> >>> It is fine to rename it for me, however:
> >>>
> >>> This patch claims to rename blk_mq_update_tag_set_shared(), but also
> >>> change blk_mq_init_bitmap_tags's signature.
> >>
> >> I was going to update the commit message here, but forgot again...
> >>
> >>>
> >>> So suggest to split this patch into two or add comment log on
> >>> changing blk_mq_init_bitmap_tags().
> >>
> >> I think I'll just split into 2x commits.
> >
> > Hi Hannes,
> >
> > Do you have any issue with splitting the undocumented changes into
> > another patch as so:
> >
> No, that's perfectly fine.
>
> Kashyap, I've also attached an updated patch for the elevator_count patch;
> if
> you agree John can include it in the next version.
Hannes - Patch looks good. Header does not include problem statement. How
about adding below in header ?
High CPU utilization on "native_queued_spin_lock_slowpath" due to lock
contention is possible in mq-deadline and bfq io scheduler when nr_hw_queues
is more than one.
It is because kblockd work queue can submit IO from all online CPUs (through
blk_mq_run_hw_queues) even though only one hctx has pending commands.
Elevator callback "has_work" for mq-deadline and bfq scheduler consider
pending work if there are any IOs on request queue and it does not account
hctx context.
I have not seen performance drop after this patch, but I will continue
further testing.
John - One more thing, I am working on megaraid_sas driver to provide both
host_tagset = 1 and 0 option through module parameter.
Kashyap
>
> Cheers,
>
> Hannes
> --
> Dr. Hannes Reinecke Teamlead Storage & Networking
> hare@suse.de +49 911 74053 688
> SUSE Software Solutions GmbH, Maxfeldstr. 5, 90409 Nürnberg HRB 36809
> (AG Nürnberg), Geschäftsführer: Felix Imendörffer
^ permalink raw reply
* RE: [PATCH v1] scsi: ufs: Disable WriteBooster capability in non-supported UFS device
From: Stanley Chu @ 2020-06-24 7:45 UTC (permalink / raw)
To: Avri Altman
Cc: linux-scsi@vger.kernel.org, martin.petersen@oracle.com,
alim.akhtar@samsung.com, jejb@linux.ibm.com, beanhuo@micron.com,
asutoshd@codeaurora.org, cang@codeaurora.org,
matthias.bgg@gmail.com, bvanassche@acm.org,
linux-mediatek@lists.infradead.org,
linux-arm-kernel@lists.infradead.org,
linux-kernel@vger.kernel.org, kuohong.wang@mediatek.com,
peter.wang@mediatek.com, chun-hung.wu@mediatek.com,
andy.teng@mediatek.com, chaotian.jing@mediatek.com,
cc.chou@mediatek.com
In-Reply-To: <SN6PR04MB4640974695F782566A83F2EFFC950@SN6PR04MB4640.namprd04.prod.outlook.com>
Hi Avri,
On Wed, 2020-06-24 at 07:26 +0000, Avri Altman wrote:
> >
> >
> > If UFS device is not qualified to enter the detection of WriteBooster
> > probing by disallowed UFS version or device quirks, then WriteBooster
> > capability in host shall be disabled to prevent any WriteBooster
> > operations in the future.
>
> Fixes: ?
I'll add it in v2, thanks!
Stanley Chu
^ permalink raw reply
* RE: [PATCH v2] scsi: ufs: Disable WriteBooster capability in non-supported UFS device
From: Avri Altman @ 2020-06-24 7:45 UTC (permalink / raw)
To: Stanley Chu, linux-scsi@vger.kernel.org,
martin.petersen@oracle.com, alim.akhtar@samsung.com,
jejb@linux.ibm.com
Cc: beanhuo@micron.com, asutoshd@codeaurora.org, cang@codeaurora.org,
matthias.bgg@gmail.com, bvanassche@acm.org,
linux-mediatek@lists.infradead.org,
linux-arm-kernel@lists.infradead.org,
linux-kernel@vger.kernel.org, kuohong.wang@mediatek.com,
peter.wang@mediatek.com, chun-hung.wu@mediatek.com,
andy.teng@mediatek.com, chaotian.jing@mediatek.com,
cc.chou@mediatek.com
In-Reply-To: <20200624074110.21919-1-stanley.chu@mediatek.com>
>
> If UFS device is not qualified to enter the detection of WriteBooster
> probing by disallowed UFS version or device quirks, then WriteBooster
> capability in host shall be disabled to prevent any WriteBooster
> operations in the future.
>
> Fixes: 3d17b9b5ab11 ("scsi: ufs: Add write booster feature support")
> Signed-off-by: Stanley Chu <stanley.chu@mediatek.com>
Reviewed-by: Avri Altman <avri.altman@wdc.com>
^ permalink raw reply
* [PATCH v2] scsi: ufs: Disable WriteBooster capability in non-supported UFS device
From: Stanley Chu @ 2020-06-24 7:41 UTC (permalink / raw)
To: linux-scsi, martin.petersen, avri.altman, alim.akhtar, jejb
Cc: beanhuo, asutoshd, cang, matthias.bgg, bvanassche, linux-mediatek,
linux-arm-kernel, linux-kernel, kuohong.wang, peter.wang,
chun-hung.wu, andy.teng, chaotian.jing, cc.chou, Stanley Chu
If UFS device is not qualified to enter the detection of WriteBooster
probing by disallowed UFS version or device quirks, then WriteBooster
capability in host shall be disabled to prevent any WriteBooster
operations in the future.
Fixes: 3d17b9b5ab11 ("scsi: ufs: Add write booster feature support")
Signed-off-by: Stanley Chu <stanley.chu@mediatek.com>
---
drivers/scsi/ufs/ufshcd.c | 35 +++++++++++++++++++----------------
1 file changed, 19 insertions(+), 16 deletions(-)
diff --git a/drivers/scsi/ufs/ufshcd.c b/drivers/scsi/ufs/ufshcd.c
index f173ad1bd79f..c62bd47beeaa 100644
--- a/drivers/scsi/ufs/ufshcd.c
+++ b/drivers/scsi/ufs/ufshcd.c
@@ -6847,21 +6847,31 @@ static int ufshcd_scsi_add_wlus(struct ufs_hba *hba)
static void ufshcd_wb_probe(struct ufs_hba *hba, u8 *desc_buf)
{
+ struct ufs_dev_info *dev_info = &hba->dev_info;
u8 lun;
u32 d_lu_wb_buf_alloc;
if (!ufshcd_is_wb_allowed(hba))
return;
+ /*
+ * Probe WB only for UFS-2.2 and UFS-3.1 (and later) devices or
+ * UFS devices with quirk UFS_DEVICE_QUIRK_SUPPORT_EXTENDED_FEATURES
+ * enabled
+ */
+ if (!(dev_info->wspecversion >= 0x310 ||
+ dev_info->wspecversion == 0x220 ||
+ (hba->dev_quirks & UFS_DEVICE_QUIRK_SUPPORT_EXTENDED_FEATURES)))
+ goto wb_disabled;
if (hba->desc_size[QUERY_DESC_IDN_DEVICE] <
DEVICE_DESC_PARAM_EXT_UFS_FEATURE_SUP + 4)
goto wb_disabled;
- hba->dev_info.d_ext_ufs_feature_sup =
+ dev_info->d_ext_ufs_feature_sup =
get_unaligned_be32(desc_buf +
DEVICE_DESC_PARAM_EXT_UFS_FEATURE_SUP);
- if (!(hba->dev_info.d_ext_ufs_feature_sup & UFS_DEV_WRITE_BOOSTER_SUP))
+ if (!(dev_info->d_ext_ufs_feature_sup & UFS_DEV_WRITE_BOOSTER_SUP))
goto wb_disabled;
/*
@@ -6870,17 +6880,17 @@ static void ufshcd_wb_probe(struct ufs_hba *hba, u8 *desc_buf)
* a max of 1 lun would have wb buffer configured.
* Now only shared buffer mode is supported.
*/
- hba->dev_info.b_wb_buffer_type =
+ dev_info->b_wb_buffer_type =
desc_buf[DEVICE_DESC_PARAM_WB_TYPE];
- hba->dev_info.b_presrv_uspc_en =
+ dev_info->b_presrv_uspc_en =
desc_buf[DEVICE_DESC_PARAM_WB_PRESRV_USRSPC_EN];
- if (hba->dev_info.b_wb_buffer_type == WB_BUF_MODE_SHARED) {
- hba->dev_info.d_wb_alloc_units =
+ if (dev_info->b_wb_buffer_type == WB_BUF_MODE_SHARED) {
+ dev_info->d_wb_alloc_units =
get_unaligned_be32(desc_buf +
DEVICE_DESC_PARAM_WB_SHARED_ALLOC_UNITS);
- if (!hba->dev_info.d_wb_alloc_units)
+ if (!dev_info->d_wb_alloc_units)
goto wb_disabled;
} else {
for (lun = 0; lun < UFS_UPIU_MAX_WB_LUN_ID; lun++) {
@@ -6891,7 +6901,7 @@ static void ufshcd_wb_probe(struct ufs_hba *hba, u8 *desc_buf)
(u8 *)&d_lu_wb_buf_alloc,
sizeof(d_lu_wb_buf_alloc));
if (d_lu_wb_buf_alloc) {
- hba->dev_info.wb_dedicated_lu = lun;
+ dev_info->wb_dedicated_lu = lun;
break;
}
}
@@ -6977,14 +6987,7 @@ static int ufs_get_device_desc(struct ufs_hba *hba)
ufs_fixup_device_setup(hba);
- /*
- * Probe WB only for UFS-3.1 devices or UFS devices with quirk
- * UFS_DEVICE_QUIRK_SUPPORT_EXTENDED_FEATURES enabled
- */
- if (dev_info->wspecversion >= 0x310 ||
- dev_info->wspecversion == 0x220 ||
- (hba->dev_quirks & UFS_DEVICE_QUIRK_SUPPORT_EXTENDED_FEATURES))
- ufshcd_wb_probe(hba, desc_buf);
+ ufshcd_wb_probe(hba, desc_buf);
/*
* ufshcd_read_string_desc returns size of the string
--
2.18.0
^ permalink raw reply related
* RE: [PATCH v1] scsi: ufs: Disable WriteBooster capability in non-supported UFS device
From: Avri Altman @ 2020-06-24 7:26 UTC (permalink / raw)
To: Stanley Chu, linux-scsi@vger.kernel.org,
martin.petersen@oracle.com, alim.akhtar@samsung.com,
jejb@linux.ibm.com
Cc: beanhuo@micron.com, asutoshd@codeaurora.org, cang@codeaurora.org,
matthias.bgg@gmail.com, bvanassche@acm.org,
linux-mediatek@lists.infradead.org,
linux-arm-kernel@lists.infradead.org,
linux-kernel@vger.kernel.org, kuohong.wang@mediatek.com,
peter.wang@mediatek.com, chun-hung.wu@mediatek.com,
andy.teng@mediatek.com, chaotian.jing@mediatek.com,
cc.chou@mediatek.com
In-Reply-To: <20200624025119.6509-1-stanley.chu@mediatek.com>
>
>
> If UFS device is not qualified to enter the detection of WriteBooster
> probing by disallowed UFS version or device quirks, then WriteBooster
> capability in host shall be disabled to prevent any WriteBooster
> operations in the future.
Fixes: ?
>
> Signed-off-by: Stanley Chu <stanley.chu@mediatek.com>
> ---
^ permalink raw reply
* [Bug 208045] ARM ubuntu 18.04 as the iscsi server, using initiator login, the kernel crashes
From: bugzilla-daemon @ 2020-06-24 4:56 UTC (permalink / raw)
To: linux-scsi
In-Reply-To: <bug-208045-11613@https.bugzilla.kernel.org/>
https://bugzilla.kernel.org/show_bug.cgi?id=208045
--- Comment #14 from JiangYu (lnsyyj@hotmail.com) ---
I upgraded patch4 on kernel 4.19.118 and tested the read and write for a day.
No problems encountered again.
Thank you!
--
You are receiving this mail because:
You are watching the assignee of the bug.
^ permalink raw reply
* Re: [PATCH -next] scsi: ufs: ufs-exynos: Fix return value check in exynos_ufs_init()
From: Martin K. Petersen @ 2020-06-24 4:30 UTC (permalink / raw)
To: Seungwon Jeon, James E.J. Bottomley, Avri Altman, Alim Akhtar,
Krzysztof Kozlowski, Kukjin Kim, Kiwoong Kim, Wei Yongjun
Cc: Martin K . Petersen, linux-samsung-soc, linux-scsi,
linux-arm-kernel, kernel-janitors, Hulk Robot
In-Reply-To: <20200618133837.127274-1-weiyongjun1@huawei.com>
On Thu, 18 Jun 2020 13:38:37 +0000, Wei Yongjun wrote:
> In case of error, the function devm_ioremap_resource() returns ERR_PTR()
> and never returns NULL. The NULL test in the return value check should
> be replaced with IS_ERR().
Applied to 5.9/scsi-queue, thanks!
[1/1] scsi: ufs: ufs-exynos: Fix return value check in exynos_ufs_init()
https://git.kernel.org/mkp/scsi/c/b2bc2200e89b
--
Martin K. Petersen Oracle Linux Engineering
^ permalink raw reply
* Re: [PATCH -next] scsi: ufs: allow exynos ufs driver to build as module
From: Martin K. Petersen @ 2020-06-24 4:30 UTC (permalink / raw)
To: linux-scsi, Alim Akhtar
Cc: Martin K . Petersen, linux-kernel, cang, stanley.chu, kwmad.kim,
avri.altman, jejb
In-Reply-To: <20200620173232.52521-1-alim.akhtar@samsung.com>
On Sat, 20 Jun 2020 23:02:32 +0530, Alim Akhtar wrote:
> Allow Exynos UFS driver to build as a module.
> This patch fix the below build issue reported by
> kernel build robot.
>
> drivers/scsi/ufs/ufs-exynos.o: in function `exynos_ufs_probe':
> drivers/scsi/ufs/ufs-exynos.c:1231: undefined reference to `ufshcd_pltfrm_init'
> drivers/scsi/ufs/ufs-exynos.o: in function `exynos_ufs_pre_pwr_mode':
> drivers/scsi/ufs/ufs-exynos.c:635: undefined reference to `ufshcd_get_pwr_dev_param'
> drivers/scsi/ufs/ufs-exynos.o:undefined reference to `ufshcd_pltfrm_shutdown'
> drivers/scsi/ufs/ufs-exynos.o:undefined reference to `ufshcd_pltfrm_suspend'
> drivers/scsi/ufs/ufs-exynos.o:undefined reference to `ufshcd_pltfrm_resume'
> drivers/scsi/ufs/ufs-exynos.o:undefined reference to `ufshcd_pltfrm_runtime_suspend'
> drivers/scsi/ufs/ufs-exynos.o:undefined reference to `ufshcd_pltfrm_runtime_resume'
> drivers/scsi/ufs/ufs-exynos.o:undefined reference to `ufshcd_pltfrm_runtime_idle'
Applied to 5.9/scsi-queue, thanks!
[1/1] scsi: ufs: Allow exynos ufs driver to build as module
https://git.kernel.org/mkp/scsi/c/d31503fe395d
--
Martin K. Petersen Oracle Linux Engineering
^ permalink raw reply
* Re: [PATCH v2] scsi: qla2xxx: Keep initiator ports after RSCN
From: Martin K. Petersen @ 2020-06-24 4:29 UTC (permalink / raw)
To: linux-scsi, Roman Bolshakov
Cc: Martin K . Petersen, GR-QLogic-Storage-Upstream, target-devel,
Nilesh Javali, Himanshu Madhani, Bart Van Assche, Arun Easi,
stable, Martin Wilck, Quinn Tran, linux, Daniel Wagner
In-Reply-To: <20200605144435.27023-1-r.bolshakov@yadro.com>
On Fri, 5 Jun 2020 17:44:37 +0300, Roman Bolshakov wrote:
> The driver performs SCR (state change registration) in all modes
> including pure target mode.
>
> For each RSCN, scan_needed flag is set in qla2x00_handle_rscn() for the
> port mentioned in the RSCN and fabric rescan is scheduled. During the
> rescan, GNN_FT handler, qla24xx_async_gnnft_done() deletes session of
> the port that caused the RSCN.
>
> [...]
Applied to 5.8/scsi-fixes, thanks!
[1/1] scsi: qla2xxx: Keep initiator ports after RSCN
https://git.kernel.org/mkp/scsi/c/632f24f09d5b
--
Martin K. Petersen Oracle Linux Engineering
^ permalink raw reply
* Re: [PATCH] libata: fix the ata_scsi_dma_need_drain stub
From: Martin K. Petersen @ 2020-06-24 4:29 UTC (permalink / raw)
To: Christoph Hellwig; +Cc: Martin K . Petersen, linux-scsi, linux-ide
In-Reply-To: <20200620071302.462974-2-hch@lst.de>
On Sat, 20 Jun 2020 09:13:02 +0200, Christoph Hellwig wrote:
> We don't only need the stub when libata is disable, but also if it
> is modular and there are built-in SAS drivers (which can happen when
> SCSI_SAS_ATA is disabled).
Applied to 5.8/scsi-fixes, thanks!
[1/1] scsi: libata: Fix the ata_scsi_dma_need_drain stub
https://git.kernel.org/mkp/scsi/c/aad4b4d15f30
--
Martin K. Petersen Oracle Linux Engineering
^ permalink raw reply
* Re: [PATCH] scsi: lpfc: Avoid another null dereference in lpfc_sli4_hba_unset()
From: Martin K. Petersen @ 2020-06-24 4:29 UTC (permalink / raw)
To: james.smart, jsmart2021, SeongJae Park, dick.kennedy
Cc: Martin K . Petersen, linux-scsi, jejb, SeongJae Park,
linux-kernel
In-Reply-To: <20200623084122.30633-1-sjpark@amazon.com>
On Tue, 23 Jun 2020 10:41:22 +0200, SeongJae Park wrote:
> Commit cdb42becdd40 ("scsi: lpfc: Replace io_channels for nvme and fcp
> with general hdw_queues per cpu") has introduced static checker warnings
> for potential null dereferences in 'lpfc_sli4_hba_unset()' and
> commit 1ffdd2c0440d ("scsi: lpfc: resolve static checker warning in
> lpfc_sli4_hba_unset") has tried to fix it. However, yet another
> potential null dereference is remaining. This commit fixes it.
>
> [...]
Applied to 5.8/scsi-fixes, thanks!
[1/1] scsi: lpfc: Avoid another null dereference in lpfc_sli4_hba_unset()
https://git.kernel.org/mkp/scsi/c/46da547e21d6
--
Martin K. Petersen Oracle Linux Engineering
^ permalink raw reply
* Re: [PATCH] mptscsih: fix read sense data size
From: Martin K. Petersen @ 2020-06-24 4:29 UTC (permalink / raw)
To: linux-scsi, Tomas Henzl; +Cc: Martin K . Petersen, sreekanth.reddy, ssaner
In-Reply-To: <20200616150446.4840-1-thenzl@redhat.com>
On Tue, 16 Jun 2020 17:04:46 +0200, Tomas Henzl wrote:
> The sense data buffer in sense_buf_pool is allocated with
> size of MPT_SENSE_BUFFER_ALLOC(64) (multiplied by req_depth)
> while SNS_LEN(sc)(96) is used when reading the data.
> That may lead to a read from unallocated area,
> sometimes from another (unallocated) page.
> To fix this limit the read size to MPT_SENSE_BUFFER_ALLOC.
Applied to 5.8/scsi-fixes, thanks!
[1/1] scsi: mptscsih: Fix read sense data size
https://git.kernel.org/mkp/scsi/c/afe89f115e84
--
Martin K. Petersen Oracle Linux Engineering
^ permalink raw reply
* Re: [PATCH] qla2xxx: Set NVME status code for failed NVME FCP request
From: Martin K. Petersen @ 2020-06-24 4:29 UTC (permalink / raw)
To: linux-scsi, Daniel Wagner
Cc: Martin K . Petersen, linux-kernel, GR-QLogic-Storage-Upstream,
Nilesh Javali
In-Reply-To: <20200604100745.89250-1-dwagner@suse.de>
On Thu, 4 Jun 2020 12:07:45 +0200, Daniel Wagner wrote:
> The qla2xxx driver knows when request was processed successfully or
> not. But it always sets the NVME status code to 0/NVME_SC_SUCCESS. The
> upper layer needs to figure out from the rcv_rsplen and
> transferred_length variables if the request was successfully. This is
> not always possible, e.g. when the request data length is 0, the
> transferred_length is also set 0 which is interpreted as success in
> nvme_fc_fcpio_done(). Let's inform the upper
> layer (nvme_fc_fcpio_done()) when something went wrong.
>
> [...]
Applied to 5.8/scsi-fixes, thanks!
[1/1] scsi: qla2xxx: Set NVMe status code for failed NVMe FCP request
https://git.kernel.org/mkp/scsi/c/ef2e3ec520a8
--
Martin K. Petersen Oracle Linux Engineering
^ permalink raw reply
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox