Hi
There is an issue with the SCSI UFS driver when the optional
BOOT well-known LUN fails to probe, which is not a fatal error.
The issue is that the device and its "managed" device link do not
then get deleted. The device because the device link has a
reference to it. The device link because it can only be deleted
by device_del(), but device_add() was never called, so device_del()
never will be either.
These V2 patches fix the issue by amending device link removal to accept
removal of a link with an unregistered consumer device, as suggested
by Rafael.
Changes in V2:
Take approach suggested by Rafael
Adrian Hunter (2):
driver core: Add ability to delete device links of unregistered devices
scsi: ufshcd: Fix device links when BOOT WLUN fails to probe
drivers/base/core.c | 11 ++++++++---
drivers/scsi/ufs/ufshcd.c | 23 +++++++++++++++++++++--
2 files changed, 29 insertions(+), 5 deletions(-)
Regards
Adrian
Managed device links are deleted by device_del(). However it is possible to
add a device link to a consumer before device_add(), and then discover an
error prevents the device from being used. In that case normally references
to the device would be dropped and the device would be deleted. However the
device link holds a reference to the device, so the device link and device
remain indefinitely.
Amend device link removal to accept removal of a link with an
unregistered consumer device.
To make that work nicely, the devlink_remove_symlinks() function must be
amended to cope with the absence of the consumer's sysfs presence,
otherwise sysfs_remove_link() will generate a warning.
Suggested-by: Rafael J. Wysocki <[email protected]>
Fixes: b294ff3e34490 ("scsi: ufs: core: Enable power management for wlun")
Signed-off-by: Adrian Hunter <[email protected]>
---
drivers/base/core.c | 11 ++++++++---
1 file changed, 8 insertions(+), 3 deletions(-)
diff --git a/drivers/base/core.c b/drivers/base/core.c
index ea5b85354526..24bacdb315c6 100644
--- a/drivers/base/core.c
+++ b/drivers/base/core.c
@@ -562,7 +562,8 @@ static void devlink_remove_symlinks(struct device *dev,
struct device *con = link->consumer;
char *buf;
- sysfs_remove_link(&link->link_dev.kobj, "consumer");
+ if (device_is_registered(con))
+ sysfs_remove_link(&link->link_dev.kobj, "consumer");
sysfs_remove_link(&link->link_dev.kobj, "supplier");
len = max(strlen(dev_bus_name(sup)) + strlen(dev_name(sup)),
@@ -575,8 +576,10 @@ static void devlink_remove_symlinks(struct device *dev,
return;
}
- snprintf(buf, len, "supplier:%s:%s", dev_bus_name(sup), dev_name(sup));
- sysfs_remove_link(&con->kobj, buf);
+ if (device_is_registered(con)) {
+ snprintf(buf, len, "supplier:%s:%s", dev_bus_name(sup), dev_name(sup));
+ sysfs_remove_link(&con->kobj, buf);
+ }
snprintf(buf, len, "consumer:%s:%s", dev_bus_name(con), dev_name(con));
sysfs_remove_link(&sup->kobj, buf);
kfree(buf);
@@ -885,6 +888,8 @@ static void device_link_put_kref(struct device_link *link)
{
if (link->flags & DL_FLAG_STATELESS)
kref_put(&link->kref, __device_link_del);
+ else if (!device_is_registered(link->consumer))
+ __device_link_del(&link->kref);
else
WARN(1, "Unable to drop a managed device link reference\n");
}
--
2.17.1
If a LUN fails to probe (e.g. absent BOOT WLUN), the device will not have
been registered but can still have a device link holding a reference to the
device. The unwanted device link will prevent runtime suspend indefinitely,
and cause some warnings if the supplier is ever deleted (e.g. by unbinding
the UFS host controller). Fix by explicitly deleting the device link when
SCSI destroys the SCSI device.
Fixes: b294ff3e34490 ("scsi: ufs: core: Enable power management for wlun")
Signed-off-by: Adrian Hunter <[email protected]>
---
drivers/scsi/ufs/ufshcd.c | 23 +++++++++++++++++++++--
1 file changed, 21 insertions(+), 2 deletions(-)
diff --git a/drivers/scsi/ufs/ufshcd.c b/drivers/scsi/ufs/ufshcd.c
index 708b3b62fc4d..9864a8ee0263 100644
--- a/drivers/scsi/ufs/ufshcd.c
+++ b/drivers/scsi/ufs/ufshcd.c
@@ -5020,15 +5020,34 @@ static int ufshcd_slave_configure(struct scsi_device *sdev)
static void ufshcd_slave_destroy(struct scsi_device *sdev)
{
struct ufs_hba *hba;
+ unsigned long flags;
hba = shost_priv(sdev->host);
/* Drop the reference as it won't be needed anymore */
if (ufshcd_scsi_to_upiu_lun(sdev->lun) == UFS_UPIU_UFS_DEVICE_WLUN) {
- unsigned long flags;
-
spin_lock_irqsave(hba->host->host_lock, flags);
hba->sdev_ufs_device = NULL;
spin_unlock_irqrestore(hba->host->host_lock, flags);
+ } else if (hba->sdev_ufs_device) {
+ struct device *supplier = NULL;
+
+ /* Ensure UFS Device WLUN exists and does not disappear */
+ spin_lock_irqsave(hba->host->host_lock, flags);
+ if (hba->sdev_ufs_device) {
+ supplier = &hba->sdev_ufs_device->sdev_gendev;
+ get_device(supplier);
+ }
+ spin_unlock_irqrestore(hba->host->host_lock, flags);
+
+ if (supplier) {
+ /*
+ * If a LUN fails to probe (e.g. absent BOOT WLUN), the
+ * device will not have been registered but can still
+ * have a device link holding a reference to the device.
+ */
+ device_link_remove(&sdev->sdev_gendev, supplier);
+ put_device(supplier);
+ }
}
}
--
2.17.1
On Fri, Jul 09, 2021 at 09:43:40AM +0300, Adrian Hunter wrote:
> Managed device links are deleted by device_del(). However it is possible to
> add a device link to a consumer before device_add(), and then discover an
> error prevents the device from being used. In that case normally references
> to the device would be dropped and the device would be deleted. However the
> device link holds a reference to the device, so the device link and device
> remain indefinitely.
Why are you not just manually removing the link you just created? You
manually added it, you know something failed so you need to clean up, so
why not clean this up too?
thanks,
greg k-h
On Fri, Jul 9, 2021 at 11:55 AM Greg Kroah-Hartman
<[email protected]> wrote:
>
> On Fri, Jul 09, 2021 at 09:43:40AM +0300, Adrian Hunter wrote:
> > Managed device links are deleted by device_del(). However it is possible to
> > add a device link to a consumer before device_add(), and then discover an
> > error prevents the device from being used. In that case normally references
> > to the device would be dropped and the device would be deleted. However the
> > device link holds a reference to the device, so the device link and device
> > remain indefinitely.
>
> Why are you not just manually removing the link you just created? You
> manually added it, you know something failed so you need to clean up, so
> why not clean this up too?
His changes are needed to do just that.
On Fri, Jul 9, 2021 at 8:43 AM Adrian Hunter <[email protected]> wrote:
>
> Managed device links are deleted by device_del(). However it is possible to
> add a device link to a consumer before device_add(), and then discover an
> error prevents the device from being used. In that case normally references
> to the device would be dropped and the device would be deleted. However the
> device link holds a reference to the device, so the device link and device
> remain indefinitely.
>
> Amend device link removal to accept removal of a link with an
> unregistered consumer device.
>
> To make that work nicely, the devlink_remove_symlinks() function must be
> amended to cope with the absence of the consumer's sysfs presence,
> otherwise sysfs_remove_link() will generate a warning.
>
> Suggested-by: Rafael J. Wysocki <[email protected]>
> Fixes: b294ff3e34490 ("scsi: ufs: core: Enable power management for wlun")
> Signed-off-by: Adrian Hunter <[email protected]>
> ---
> drivers/base/core.c | 11 ++++++++---
> 1 file changed, 8 insertions(+), 3 deletions(-)
>
> diff --git a/drivers/base/core.c b/drivers/base/core.c
> index ea5b85354526..24bacdb315c6 100644
> --- a/drivers/base/core.c
> +++ b/drivers/base/core.c
> @@ -562,7 +562,8 @@ static void devlink_remove_symlinks(struct device *dev,
> struct device *con = link->consumer;
> char *buf;
>
> - sysfs_remove_link(&link->link_dev.kobj, "consumer");
> + if (device_is_registered(con))
> + sysfs_remove_link(&link->link_dev.kobj, "consumer");
I think that this is needed regardless of the changes in
device_link_put_kref(), because if somebody decides to delete a
stateless device link before registering the consumer device,
sysfs_remove_link() will still complain, won't it?
> sysfs_remove_link(&link->link_dev.kobj, "supplier");
>
> len = max(strlen(dev_bus_name(sup)) + strlen(dev_name(sup)),
> @@ -575,8 +576,10 @@ static void devlink_remove_symlinks(struct device *dev,
> return;
> }
>
> - snprintf(buf, len, "supplier:%s:%s", dev_bus_name(sup), dev_name(sup));
> - sysfs_remove_link(&con->kobj, buf);
> + if (device_is_registered(con)) {
> + snprintf(buf, len, "supplier:%s:%s", dev_bus_name(sup), dev_name(sup));
> + sysfs_remove_link(&con->kobj, buf);
> + }
And here too, if I'm not mistaken.
So in that case it would be better to put the above changes into a
separate patch and add a Fixes tag to it.
> snprintf(buf, len, "consumer:%s:%s", dev_bus_name(con), dev_name(con));
> sysfs_remove_link(&sup->kobj, buf);
> kfree(buf);
> @@ -885,6 +888,8 @@ static void device_link_put_kref(struct device_link *link)
> {
> if (link->flags & DL_FLAG_STATELESS)
> kref_put(&link->kref, __device_link_del);
> + else if (!device_is_registered(link->consumer))
> + __device_link_del(&link->kref);
> else
> WARN(1, "Unable to drop a managed device link reference\n");
> }
> --
> 2.17.1
>
On 9/07/21 2:28 pm, Rafael J. Wysocki wrote:
> On Fri, Jul 9, 2021 at 8:43 AM Adrian Hunter <[email protected]> wrote:
>>
>> Managed device links are deleted by device_del(). However it is possible to
>> add a device link to a consumer before device_add(), and then discover an
>> error prevents the device from being used. In that case normally references
>> to the device would be dropped and the device would be deleted. However the
>> device link holds a reference to the device, so the device link and device
>> remain indefinitely.
>>
>> Amend device link removal to accept removal of a link with an
>> unregistered consumer device.
>>
>> To make that work nicely, the devlink_remove_symlinks() function must be
>> amended to cope with the absence of the consumer's sysfs presence,
>> otherwise sysfs_remove_link() will generate a warning.
>>
>> Suggested-by: Rafael J. Wysocki <[email protected]>
>> Fixes: b294ff3e34490 ("scsi: ufs: core: Enable power management for wlun")
>> Signed-off-by: Adrian Hunter <[email protected]>
>> ---
>> drivers/base/core.c | 11 ++++++++---
>> 1 file changed, 8 insertions(+), 3 deletions(-)
>>
>> diff --git a/drivers/base/core.c b/drivers/base/core.c
>> index ea5b85354526..24bacdb315c6 100644
>> --- a/drivers/base/core.c
>> +++ b/drivers/base/core.c
>> @@ -562,7 +562,8 @@ static void devlink_remove_symlinks(struct device *dev,
>> struct device *con = link->consumer;
>> char *buf;
>>
>> - sysfs_remove_link(&link->link_dev.kobj, "consumer");
>> + if (device_is_registered(con))
>> + sysfs_remove_link(&link->link_dev.kobj, "consumer");
>
> I think that this is needed regardless of the changes in
> device_link_put_kref(), because if somebody decides to delete a
> stateless device link before registering the consumer device,
> sysfs_remove_link() will still complain, won't it?
I would think so.
>
>> sysfs_remove_link(&link->link_dev.kobj, "supplier");
>>
>> len = max(strlen(dev_bus_name(sup)) + strlen(dev_name(sup)),
>> @@ -575,8 +576,10 @@ static void devlink_remove_symlinks(struct device *dev,
>> return;
>> }
>>
>> - snprintf(buf, len, "supplier:%s:%s", dev_bus_name(sup), dev_name(sup));
>> - sysfs_remove_link(&con->kobj, buf);
>> + if (device_is_registered(con)) {
>> + snprintf(buf, len, "supplier:%s:%s", dev_bus_name(sup), dev_name(sup));
>> + sysfs_remove_link(&con->kobj, buf);
>> + }
>
> And here too, if I'm not mistaken.
>
> So in that case it would be better to put the above changes into a
> separate patch and add a Fixes tag to it.
Yes, that makes sense. I'll send a V3
>
>> snprintf(buf, len, "consumer:%s:%s", dev_bus_name(con), dev_name(con));
>> sysfs_remove_link(&sup->kobj, buf);
>> kfree(buf);
>> @@ -885,6 +888,8 @@ static void device_link_put_kref(struct device_link *link)
>> {
>> if (link->flags & DL_FLAG_STATELESS)
>> kref_put(&link->kref, __device_link_del);
>> + else if (!device_is_registered(link->consumer))
>> + __device_link_del(&link->kref);
>> else
>> WARN(1, "Unable to drop a managed device link reference\n");
>> }
>> --
>> 2.17.1
>>