Hi Tejun, Mike,

On 08/18/2010 01:51 AM +0900, Tejun Heo wrote:

On 08/17/2010 04:07 PM, Mike Snitzer wrote:

quoted

NOTE: NEC has already given some preliminary feedback to hch in the
"[PATCH, RFC 2/2] dm: support REQ_FLUSH directly" thread:
https://www.redhat.com/archives/dm-devel/2010-August/msg00026.html
https://www.redhat.com/archives/dm-devel/2010-August/msg00033.html

Hmmm... I think both issues don't exist in this incarnation of
conversion although I'm fairly sure there will be other issues.  :-)

The same issue is still there for request-based dm.  See below.

quoted

quoted

    A related question: Is dm_wait_for_completion() used in
    process_flush() safe against starvation under continuous influx of
    other commands?

As for your specific dm_wait_for_completion() concern -- I'll defer to
Mikulas.  But I'll add: we haven't had any reported starvation issues
with DM's existing barrier support.  DM uses a mempool for its clones,
so it should naturally throttle (without starvation) when memory gets
low.

I see but single pending flush and steady write streams w/o saturating
the mempool would be able to stall dm_wait_for_completeion(), no?  Eh
well, it's a separate issue, I guess.

Your understanding is correct, dm_wait_for_completion() for flush
will stall in such cases for request-based dm.
That's why I mentioned below in
https://www.redhat.com/archives/dm-devel/2010-August/msg00026.html.

    In other words, current request-based device-mapper can't handle
    other requests while a flush request is in progress.

In flush request handling, request-based dm uses dm_wait_for_completion()
to wait for the completion of cloned flush requests, depending on
the fact that there should be only flush requests in flight owning
to the block layer sequencing.

It's not a separate issue and we need to resolve it at least.
I'm still considering how I can fix the request-based dm.

Thanks,
Kiyoshi Ueda