Date: Mon, 24 Feb 2020 08:16:38 -0600 From: Jason Bacon <bacon4000@gmail.com> To: freebsd-infiniband@freebsd.org Subject: Re: [PATCH]: ipoib with mlx4 initialisation ordering Message-ID: <3684a21c-e26c-abcf-b443-aa0ddccc1338@gmail.com> In-Reply-To: <9d76992b-6ba4-2419-61ff-5035aa45e597@selasky.org> References: <20200222004838.GA22659@moira.hest-guild.se> <9d76992b-6ba4-2419-61ff-5035aa45e597@selasky.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On 2020-02-24 03:05, Hans Petter Selasky wrote: > On 2020-02-22 01:48, Andreas Kempe wrote: >> This issue can be remedied by changing the initialisation of the IPoIB= >> module to happen after the mlx4 driver is initialised. By doing this, >> all multicast groups will be cleaned up before the ib_multicast client= >> is destroyed. > > Hi Andreas, > > Are you sure it is not the module exit that should be ordered instead? > > =C2=A0module_exit(ipoib_cleanup_module); > > Because from the description this issue happen on shutdown and not load= =2E > > I'm currently trying to reproduce the issue. > > Dropping freebsd-net @ > > --HPS > _______________________________________________ > freebsd-infiniband@freebsd.org mailing list > https://lists.freebsd.org/mailman/listinfo/freebsd-infiniband > To unsubscribe, send any mail to=20 > "freebsd-infiniband-unsubscribe@freebsd.org" Side-note: I was seeing the same symptom occasionally on our CentOS=20 clusters for years.=C2=A0 I frequently had to go to the data center and c= heck=20 on hung systems after routine updates, which would often be stuck at=20 "Unloading IB modules".=C2=A0 I wonder if it was a similar issue. --=20 Earth is a beta site.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?3684a21c-e26c-abcf-b443-aa0ddccc1338>