On Wed, 2021-06-23 at 14:03 +0200, Hannes Reinecke wrote:

On 6/23/21 12:58 PM, Matteo Croce wrote:

quoted

From: Matteo Croce <redacted>

With this series a monotonically increasing number is added to disks,
precisely in the genhd struct, and it's exported in sysfs and uevent.

This helps the userspace correlate events for devices that reuse the
same device, like loop.

I'm failing to see the point here.
Apparently you are assuming that there is a userspace tool tracking 
events, and has a need to correlate events related to different 
instances of the disk.
But if you have an userspace application tracking events, why can't the 
same application track the 'add' and 'remove' events to track the 
lifetime of the devices, and implement its own numbering based on that?

Why do we need to burden the kernel with this?

Cheers,

Hannes

Hi,

It is not an assumption, such tool does exist, and manually tracking
does not work because of the impossibility of reliably correlating
events to devices (we've tried, again and again and again), which is
the purpose of this series - to solve this long standing issue, which
has been causing problems both in testing and production for a long
time now, despite our best efforts to add workaround after workaround.

For more info please see the discussion on the v1:

https://lore.kernel.org/linux-fsdevel/20210315201331.GA2577561@casper.infradead.org/t/#m5b03e48013de14b4a080c90afdc4a8b8c94c30d4 (local)

and the bug linked in the cover letter:

https://github.com/systemd/systemd/issues/17469#issuecomment-762919781

-- 
Kind regards,
Luca Boccassi