Linux Virtualization - Mar 2023 - [PATCH v2 0/7] vhost-scsi: Fix crashes and management op hangs

On 3/20/23 9:06 PM, Mike Christie wrote:
> The following patches were made over Linus tree.
Hi Michael, I see you merged my first version of the patchset in your
vhost branch.

Do you want me to just send a followup patchset?

The major diff between the 2 versions:

1. I added the first 2 patches which fix some bugs in the existing code
I found while doing some code review and testing another LIO patchset
plus v1.

Note: The other day I posted that I thought the 3rd patch in v1 caused
the bugs but they were already in the code.

2. In v2 I made one of the patches not need the vhost device lock when
unmapping/mapping LUNs, so you can add new LUNs even if one LUN on the same
vhost_scsi device was hung.

Since it's not regressions with the existing patches, I can just send those
as a followup patchset if that's preferred.

On Mon, Mar 20, 2023 at 09:29:50PM -0500, michael.christie at oracle.com
wrote:
> On 3/20/23 9:06 PM, Mike Christie wrote:
> > The following patches were made over Linus tree.
> 
> Hi Michael, I see you merged my first version of the patchset in your
> vhost branch.
> 
> Do you want me to just send a followup patchset?
> 
> The major diff between the 2 versions:
> 
> 1. I added the first 2 patches which fix some bugs in the existing code
> I found while doing some code review and testing another LIO patchset
> plus v1.
> 
> Note: The other day I posted that I thought the 3rd patch in v1 caused
> the bugs but they were already in the code.
> 
> 2. In v2 I made one of the patches not need the vhost device lock when
> unmapping/mapping LUNs, so you can add new LUNs even if one LUN on the same
> vhost_scsi device was hung.
> 
> Since it's not regressions with the existing patches, I can just send
those
> as a followup patchset if that's preferred.
It's ok, I will drop v1 and replace it with v2.
Do you feel any of this is needed in this release?

-- 
MST