[PATCH RFC 3/5] nvme: pci: use admin queue timeout over NVME_ADMIN_TIMEOUT

[Maurizio Lombardi]

On Wed Feb 18, 2026 at 2:10 PM CET, Maximilian Heyne wrote:
> On Thu, Feb 12, 2026 at 01:09:49PM +0100, Maurizio Lombardi wrote:
>> From: David Jeffery <djeffery at redhat.com>
>> 
>> While tearing down its queues, nvme-pci uses NVME_ADMIN_TIMEOUT as its
>> timeout target. Instead, use the configured admin queue's timeout value
>> when available to match the device's existing timeout setting.
>> 
>> Signed-off-by: David Jeffery <djeffery at redhat.com>
>> ---
>>  drivers/nvme/host/pci.c | 3 ++-
>>  1 file changed, 2 insertions(+), 1 deletion(-)
>> 
>> diff --git a/drivers/nvme/host/pci.c b/drivers/nvme/host/pci.c
>> index 58f3097888a7..853cd57e4480 100644
>> --- a/drivers/nvme/host/pci.c
>> +++ b/drivers/nvme/host/pci.c
>> @@ -2911,9 +2911,10 @@ static bool __nvme_delete_io_queues(struct nvme_dev *dev, u8 opcode)
>>  {
>>  	int nr_queues = dev->online_queues - 1, sent = 0;
>>  	unsigned long timeout;
>> +	struct request_queue *q = dev->ctrl.admin_q;
>>  
>>   retry:
>> -	timeout = NVME_ADMIN_TIMEOUT;
>> +	timeout = q ? q->rq_timeout : NVME_ADMIN_TIMEOUT;
>
> Skimming over the code I think the check is unnecessary since
> ctrl.admin_q is only changed in nvme_alloc_admin_tag_set. Only when this
> succeeds, we'll setup io queues. When they get deleted we will still
> have the ctrl.admin_q being valid.

Yes, afaict you are right and this check is unnecessary, I will remove it.

Maurizio