[PATCH] nvme: fixup crash on failed discovery

[PATCH] nvme: fixup crash on failed discovery

[Hannes Reinecke]

When the initial discovery fails the subsystem hasn't been setup,
and we'll encounter a kernel oops during shutdown.

Signed-off-by: Hannes Reinecke <hare at suse.com>
---
 drivers/nvme/host/multipath.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/nvme/host/multipath.c b/drivers/nvme/host/multipath.c
index 420efcda33d9..c1f41aa879c1 100644
--- a/drivers/nvme/host/multipath.c
+++ b/drivers/nvme/host/multipath.c
@@ -24,7 +24,7 @@ MODULE_PARM_DESC(multipath,
 
 inline bool nvme_ctrl_use_ana(struct nvme_ctrl *ctrl)
 {
-	return multipath && (ctrl->subsys->cmic & (1 << 3));
+	return multipath && ctrl->subsys && (ctrl->subsys->cmic & (1 << 3));
 }
 
 /*
-- 
2.12.3

[PATCH] nvme: fixup crash on failed discovery

[Christoph Hellwig]

On Tue, Aug 07, 2018@12:43:42PM +0200, Hannes Reinecke wrote:
> When the initial discovery fails the subsystem hasn't been setup,
> and we'll encounter a kernel oops during shutdown.

This looks ok, but I'm curious where we even call nvme_ctrl_use_ana
if we fail early.

[PATCH] nvme: fixup crash on failed discovery

[Hannes Reinecke]

On 08/07/2018 01:46 PM, Christoph Hellwig wrote:
> On Tue, Aug 07, 2018@12:43:42PM +0200, Hannes Reinecke wrote:
>> When the initial discovery fails the subsystem hasn't been setup,
>> and we'll encounter a kernel oops during shutdown.
> 
> This looks ok, but I'm curious where we even call nvme_ctrl_use_ana
> if we fail early.
> 
This is called via
nvme_sysfs_delete()->nvme_ctrl_delete()->nvme_stop_ctrl()->nvme_mpath_stop()
called from 'nvme discover' on a failing controller.

Cheers,

Hannes
-- 
Dr. Hannes Reinecke		               zSeries & Storage
hare at suse.com			               +49 911 74053 688
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 N?rnberg
GF: F. Imend?rffer, J. Smithard, D. Upmanyu, G. Norton
HRB 21284 (AG N?rnberg)

[PATCH] nvme: fixup crash on failed discovery

[Christoph Hellwig]

On Tue, Aug 07, 2018@02:01:09PM +0200, Hannes Reinecke wrote:
> On 08/07/2018 01:46 PM, Christoph Hellwig wrote:
> > On Tue, Aug 07, 2018@12:43:42PM +0200, Hannes Reinecke wrote:
> >> When the initial discovery fails the subsystem hasn't been setup,
> >> and we'll encounter a kernel oops during shutdown.
> > 
> > This looks ok, but I'm curious where we even call nvme_ctrl_use_ana
> > if we fail early.
> > 
> This is called via
> nvme_sysfs_delete()->nvme_ctrl_delete()->nvme_stop_ctrl()->nvme_mpath_stop()
> called from 'nvme discover' on a failing controller.

Thanks.  Applied with a slightly updated changelog.