ACK: [Vivid][SRU][PATCH 0/2] Fixes for LP: #1473883

Seth Forshee seth.forshee at canonical.com
Mon Jul 13 20:24:29 UTC 2015


On Mon, Jul 13, 2015 at 09:20:58AM -0500, Chris J Arges wrote:
> BugLink: http://bugs.launchpad.net/bugs/1473883
> 
> [Impact]
> 
> While load/unload mlx4_core twice with SR-IOV mode enabled in host with
> multiple Mellanox devices (some of them support SR-IOV and other don't) this
> will lead to kernel panic.
> 
> [Fix]
> 
> commit 5114a04e6c73a0c6e74737e801b8a3b3d40c7e36
> commit ed3d2276ef72be23c6367358d80004130d8c797d
> 
> $ git describe 5114a04e6c73a0c6e74737e801b8a3b3d40c7e36
> ed3d2276ef72be23c6367358d80004130d8c797d
> v4.1-rc6-1067-g5114a04
> v4.1-rc6-1068-ged3d227
> 
> [Test Case]
> 
> 1- add the "options mlx4_core num_vfs=60 port_type_array=2,2" to /etc/modprobe.d/mlx4_core.conf file.
> 2- unload mlx4_* kernel modules: modprobe -rv mlx4_en; modprobe -rv mlx4_ib; modprobe -rv mlx4_core;
> 3- load mlx4_en kernel module: modprobe -v mlx4_en
> 4- edit /etc/modprobe.d/mlx4_core.conf file and put "options mlx4_core num_vfs=60 port_type_array=2,2" in comment.
> 5 -repeat 2 and 3
> 6- will get the following call trace.

Fixes are simple and clean cherry picks. The first one is whitespace
damaged, but it's that way upstream too.




More information about the kernel-team mailing list