Re: [PATCH] dm: Validate zoned model of devices built with zoned targets

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On 12/1/23 20:09, Johannes Thumshirn wrote:
> On 01.12.23 03:03, Damien Le Moal wrote:
>> +	/* Count conventional and sequential zones */
>> +	noio_flag = memalloc_noio_save();
>> +	ret = dm_blk_do_report_zones(md, t, 0, nr_zones,
>> +				     dm_validate_zoned_model_cb, &zc);
>> +	memalloc_noio_restore(noio_flag);
> 
> Why do we need the memalloc_noio context here? I don't see an allocation 
> in dm_validate_zoned_model_cb.

It is not for the callback but rather for the underlying target type and from
there the underlying device drivers, which may all need to allocate memory for
the report zone (e.g. there is at least a command buffer allocated for SCSI
disks). So I prefer staying on the safe side of things here and use GFP_NOIO.

-- 
Damien Le Moal
Western Digital Research





[Index of Archives]     [DM Crypt]     [Fedora Desktop]     [ATA RAID]     [Fedora Marketing]     [Fedora Packaging]     [Fedora SELinux]     [Yosemite Discussion]     [KDE Users]     [Fedora Docs]

  Powered by Linux