2020-06-07 11:48:47

by Niklas Cassel

Subject: [PATCH] nvme: do not call del_gendisk() on a disk that was never added

device_add_disk() is negated by del_gendisk().
alloc_disk_node() is negated by put_disk().

In nvme_alloc_ns(), device_add_disk() is one of the last things being
called in the success case, and only void functions are being called
after this. Therefore this call should not be negated in the error path.

The superfluous call to del_gendisk() leads to the following prints:
[ 7.839975] kobject: '(null)' (000000001ff73734): is not initialized, yet kobject_put() is being called.
[ 7.840865] WARNING: CPU: 2 PID: 361 at lib/kobject.c:736 kobject_put+0x70/0x120

Fixes: 33cfdc2aa696 ("nvme: enforce extended LBA format for fabrics metadata")
Signed-off-by: Niklas Cassel <[email protected]>
---
An alternative would be to do like nvme_ns_remove(), i.e. in the error
path; check if ns->disk->flags & GENHD_FL_UP is set, and only then call
del_gendisk(). However, that seems unnecessary, since as nvme_alloc_ns()
is currently written, we know that device_add_disk() does not need to be
negated.

drivers/nvme/host/core.c | 4 +---
1 file changed, 1 insertion(+), 3 deletions(-)

diff --git a/drivers/nvme/host/core.c b/drivers/nvme/host/core.c
index 0585efa47d8f..c2c5bc4fb702 100644
--- a/drivers/nvme/host/core.c
+++ b/drivers/nvme/host/core.c
@@ -3669,7 +3669,7 @@ static void nvme_alloc_ns(struct nvme_ctrl *ctrl, unsigned nsid)
ns->disk = disk;

if (__nvme_revalidate_disk(disk, id))
- goto out_free_disk;
+ goto out_put_disk;

if ((ctrl->quirks & NVME_QUIRK_LIGHTNVM) && id->vs[0] == 0x1) {
ret = nvme_nvm_register(ns, disk_name, node);
@@ -3696,8 +3696,6 @@ static void nvme_alloc_ns(struct nvme_ctrl *ctrl, unsigned nsid)
/* prevent double queue cleanup */
ns->disk->queue = NULL;
put_disk(ns->disk);
- out_free_disk:
- del_gendisk(ns->disk);
out_unlink_ns:
mutex_lock(&ctrl->subsys->lock);
list_del_rcu(&ns->siblings);
--
2.26.2

2020-06-07 13:18:00

by Max Gurtovoy

Subject: Re: [PATCH] nvme: do not call del_gendisk() on a disk that was never added

On 6/7/2020 2:45 PM, Niklas Cassel wrote:
> device_add_disk() is negated by del_gendisk().
> alloc_disk_node() is negated by put_disk().
>
> In nvme_alloc_ns(), device_add_disk() is one of the last things being
> called in the success case, and only void functions are being called
> after this. Therefore this call should not be negated in the error path.
>
> The superfluous call to del_gendisk() leads to the following prints:
> [ 7.839975] kobject: '(null)' (000000001ff73734): is not initialized, yet kobject_put() is being called.
> [ 7.840865] WARNING: CPU: 2 PID: 361 at lib/kobject.c:736 kobject_put+0x70/0x120
>
> Fixes: 33cfdc2aa696 ("nvme: enforce extended LBA format for fabrics metadata")
> Signed-off-by: Niklas Cassel <[email protected]>
> ---
> An alternative would be to do like nvme_ns_remove(), i.e. in the error
> path; check if ns->disk->flags & GENHD_FL_UP is set, and only then call
> del_gendisk(). However, that seems unnecessary, since as nvme_alloc_ns()
> is currently written, we know that device_add_disk() does not need to be
> negated.
>
> drivers/nvme/host/core.c | 4 +---
> 1 file changed, 1 insertion(+), 3 deletions(-)

Looks good,

Reviewed-by: Max Gurtovoy <[email protected]>

2020-06-08 04:48:23

by Sagi Grimberg

Subject: Re: [PATCH] nvme: do not call del_gendisk() on a disk that was never added

Reviewed-by: Sagi Grimberg <[email protected]>

2020-06-09 13:47:27

by Christoph Hellwig

Subject: Re: [PATCH] nvme: do not call del_gendisk() on a disk that was never added

Thanks,

applied to nvme-5.8.