[PATCH v5 2/7] block: allow blk-zoned devices to have non-power-of-2 zone size
Pankaj Raghav
p.raghav at samsung.com
Tue May 24 00:10:53 PDT 2022
On 5/24/22 07:19, Hannes Reinecke wrote:
>> diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h
>> index c4e4c7071b7b..f5c7a41032ba 100644
>> --- a/include/linux/blkdev.h
>> +++ b/include/linux/blkdev.h
>> @@ -676,6 +676,21 @@ static inline unsigned int
>> blk_queue_zone_no(struct request_queue *q,
>> return div64_u64(sector, zone_sectors);
>> }
>> +static inline bool blk_queue_is_zone_start(struct request_queue *q,
>> sector_t sec)
>> +{
>> + sector_t zone_sectors = blk_queue_zone_sectors(q);
>> + u64 remainder = 0;
>> +
>> + if (!blk_queue_is_zoned(q))
>> + return false;
>> +
>
> Not sure if we need this here; surely blk_queue_zone_sectors() will
> already barf, and none of the callers did this check.
>
I totally agree with you but all the other blk_queue_* functions had
this defensive check and I didn't want to break that pattern:
static inline unsigned int blk_queue_zone_no(struct request_queue *q,
sector_t sector)
{
....
if (!blk_queue_is_zoned(q))
return 0;
....
}
>> + if (is_power_of_2(zone_sectors))
>> + return IS_ALIGNED(sec, zone_sectors);
>> +
And, if the chunk sectors is 0, then we will do the npo2 calculation
resulting in a divide by zero even though chances of that happening is
very very low as you pointed out.
>> + div64_u64_rem(sec, zone_sectors, &remainder);
>> + return remainder == 0;
>> +}
>> +
>> static inline bool blk_queue_zone_is_seq(struct request_queue *q,
>> sector_t sector)
>> {
>> @@ -722,6 +737,12 @@ static inline unsigned int
>> blk_queue_zone_no(struct request_queue *q,
>> {
>> return 0;
>> }
>> +
>> +static inline bool blk_queue_is_zone_start(struct request_queue *q,
>> sector_t sec)
>> +{
>> + return false;
>> +}
>> +
>> static inline unsigned int queue_max_open_zones(const struct
>> request_queue *q)
>> {
>> return 0;
> Otherwise looks good.
>
Thanks!
> Cheers,
>
> Hannes
More information about the Linux-nvme
mailing list