DPDK patches and discussions
 help / color / mirror / Atom feed
* [dpdk-dev] rte_eth_dev_stop and rte_eth_dev_close are still void
@ 2019-11-19 13:44 Andrew Rybchenko
  0 siblings, 0 replies; only message in thread
From: Andrew Rybchenko @ 2019-11-19 13:44 UTC (permalink / raw)
  To: Thomas Monjalon, Ferruh Yigit; +Cc: dev

Thomas, Ferruh,

the following deprecation notice is still in place:

* ethdev: Update API functions returning ``void`` to return ``int`` with
  negative errno values to indicate various error conditions (e.g.
  invalid port ID, unsupported operation, failed operation):

  - ``rte_eth_dev_stop``
  - ``rte_eth_dev_close``


I think the pair is the most complex from return value point of view.
Typically stop and close simply move on even if there are errors
on the way except trivial cases, when there is no way to do the job,
like invalid port ID or no dev_ops callback which should be mandatory.

The main question is what applications should do if either stop or
close return error. Panic? Just scream about it in logs?

I tend to say that functions still should be updated, but now we
should wait for 20.11 and make final decision at start of the
release cycle.

Should prototypes of callbacks be changed to allow drivers
provide operation status information (but still do as much as
it is possible)? I would say yes.

Thoughts?

Thanks,
Andrew.

^ permalink raw reply	[flat|nested] only message in thread

only message in thread, other threads:[~2019-11-19 13:45 UTC | newest]

Thread overview: (only message) (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2019-11-19 13:44 [dpdk-dev] rte_eth_dev_stop and rte_eth_dev_close are still void Andrew Rybchenko

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).