[PATCHv2] nvme-multipath: split bios with the ns_head bio_set before submitting

[PATCHv2] nvme-multipath: split bios with the ns_head bio_set before submitting

[Hannes Reinecke]

If the bio is moved to a different queue via blk_steal_bios() and
the original queue is destroyed in nvme_remove_ns() we'll be ending
with a crash in bio_endio() as the mempool for the split bio bvecs
had already been destroyed.
So split the bio using the original queue (which will remain during the
lifetime of the bio) before sending it down to the underlying device.

Signed-off-by: Hannes Reinecke <hare at suse.com>
---
 drivers/nvme/host/multipath.c | 8 ++++++++
 1 file changed, 8 insertions(+)

diff --git a/drivers/nvme/host/multipath.c b/drivers/nvme/host/multipath.c
index 2551264ef2b5..e13e960a7139 100644
--- a/drivers/nvme/host/multipath.c
+++ b/drivers/nvme/host/multipath.c
@@ -232,6 +232,14 @@ static blk_qc_t nvme_ns_head_make_request(struct request_queue *q,
 	blk_qc_t ret = BLK_QC_T_NONE;
 	int srcu_idx;
 
+	/*
+	 * The namespace might be going away and the bio might
+	 * be moved to a different queue via blk_steal_bios(),
+	 * so we need to use the bio_split pool from the original
+	 * queue to allocate the bvecs from.
+	 */
+	blk_queue_split(q, *bio);
+
 	srcu_idx = srcu_read_lock(&head->srcu);
 	ns = nvme_find_path(head);
 	if (likely(ns)) {
-- 
2.16.4

[PATCHv2] nvme-multipath: split bios with the ns_head bio_set before submitting

[Keith Busch]

On Tue, Apr 30, 2019@06:27:45PM +0200, Hannes Reinecke wrote:
> +	/*
> +	 * The namespace might be going away and the bio might
> +	 * be moved to a different queue via blk_steal_bios(),
> +	 * so we need to use the bio_split pool from the original
> +	 * queue to allocate the bvecs from.
> +	 */
> +	blk_queue_split(q, *bio);
                           ^^^^

Shouldn't that be '&bio'?

[PATCHv2] nvme-multipath: split bios with the ns_head bio_set before submitting

[Hannes Reinecke]

On 4/30/19 6:33 PM, Keith Busch wrote:
> On Tue, Apr 30, 2019@06:27:45PM +0200, Hannes Reinecke wrote:
>> +	/*
>> +	 * The namespace might be going away and the bio might
>> +	 * be moved to a different queue via blk_steal_bios(),
>> +	 * so we need to use the bio_split pool from the original
>> +	 * queue to allocate the bvecs from.
>> +	 */
>> +	blk_queue_split(q, *bio);
>                             ^^^^
> 
> Shouldn't that be '&bio'?
> 
Lemme check ...

/me cross-checks the build logs

and indeed, you are correct.

Will be resending.

Cheers,

Hannes
-- 
Dr. Hannes Reinecke            Teamlead Storage & Networking
hare at suse.de                              +49 911 74053 688
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 N?rnberg
GF: Felix Imend?rffer, Mary Higgins, Sri Rasiah
HRB 21284 (AG N?rnberg)