NVMe pci driver unable to create queues

NVMe pci driver unable to create queues

[Johannes Thumshirn]

Hi,

I have the follwing errors on a ARM 64 device:

$ dmesg  | grep nvme
[   14.490744] nvme nvme0: pci function 0005:90:00.0
[   15.913973] nvme 0005:90:00.0: Could not set queue count (6)
[   15.913986] nvme nvme0: IO queues not created

but cannot really debug why the driver isn't able to create it's IO
queues.

Has anybody seen a similar error? All I've found was
https://lkml.org/lkml/2016/4/26/1557 and that wasn't really helpfull.

Thanks in advance,
       Johannes

-- 
Johannes Thumshirn                                          Storage
jthumshirn at suse.de                                +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 N?rnberg
GF: Felix Imend?rffer, Jane Smithard, Graham Norton
HRB 21284 (AG N?rnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850

NVMe pci driver unable to create queues

[Keith Busch]

On Tue, Aug 23, 2016@11:26:41AM +0200, Johannes Thumshirn wrote:
> Hi,
> 
> I have the follwing errors on a ARM 64 device:
> 
> $ dmesg  | grep nvme
> [   14.490744] nvme nvme0: pci function 0005:90:00.0
> [   15.913973] nvme 0005:90:00.0: Could not set queue count (6)
> [   15.913986] nvme nvme0: IO queues not created
> 
> but cannot really debug why the driver isn't able to create it's IO
> queues.
> 
> Has anybody seen a similar error? All I've found was
> https://lkml.org/lkml/2016/4/26/1557 and that wasn't really helpfull.

The return code "6" means "internal device error" and indicates the
device is in a bad state incapable of IO.

The driver should have provided a character device handle at /dev/nvme0
that you can issue admin commands to query and potentially fix the
device. You may need to check with the device vendor for those steps.

NVMe pci driver unable to create queues

[Johannes Thumshirn]

On Tue, Aug 23, 2016@10:46:53AM -0400, Keith Busch wrote:
> On Tue, Aug 23, 2016@11:26:41AM +0200, Johannes Thumshirn wrote:
> > Hi,
> > 
> > I have the follwing errors on a ARM 64 device:
> > 
> > $ dmesg  | grep nvme
> > [   14.490744] nvme nvme0: pci function 0005:90:00.0
> > [   15.913973] nvme 0005:90:00.0: Could not set queue count (6)
> > [   15.913986] nvme nvme0: IO queues not created
> > 
> > but cannot really debug why the driver isn't able to create it's IO
> > queues.
> > 
> > Has anybody seen a similar error? All I've found was
> > https://lkml.org/lkml/2016/4/26/1557 and that wasn't really helpfull.
> 
> The return code "6" means "internal device error" and indicates the
> device is in a bad state incapable of IO.
> 
> The driver should have provided a character device handle at /dev/nvme0
> that you can issue admin commands to query and potentially fix the
> device. You may need to check with the device vendor for those steps.

Thanks for the answer.

As nvme get-log and nvme error-log don't return something different
than INTERNAL(6) I think "it's dead jim".

Thanks anyways.

-- 
Johannes Thumshirn                                          Storage
jthumshirn at suse.de                                +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 N?rnberg
GF: Felix Imend?rffer, Jane Smithard, Graham Norton
HRB 21284 (AG N?rnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850