Re: [PATCH for-rc v2] IB/hfi1: Move cached value of mm into handler

[Ira Weiny <ira.weiny@intel.com>]

On Fri, Nov 13, 2020 at 08:37:39AM -0500, Dennis Dalessandro wrote:
> On 11/12/2020 7:33 PM, Ira Weiny wrote:
> > So I think the final point is the key to fixing the bug.  Keeping any
> > current->mm which is not the one we opened the file with...  (or more
> > specifically the one which first registered memory).  In some ways this may be
> > worse than before because technically the parent could open the fd and hand it
> > to the child and have the child register with it's mm.  But that is ok
> > really...  May just be odd behavior for some users depending on what operations
> > they do and in what order.
> 
> I don't think that's worse than before.

'worse' was a bad word...  sorry.  I just meant that the new code opens up the
opportunity to get it to work with the child when this is really not the way I
the API was intended...  But now I am just bike shedding...  ;-)

> Before we were letting it operate on
> the wrong mm. That's so much worse. Yes, parent could open fd and hand it
> off, which is OK. The "odd" behavior is up to whoever wrote the user space
> code to do that in the first place.
> 
> > [1] Also, you probably should credit Jann for the idea with a suggested by tag.
> 
> Will change reported-by to suggested-by.

Ah yea both are appropriate.  I'm not an expert on if both should be tagged or
not...  But to me it seems suggested-by is a bit stronger because a solution
was provided beyond just a bug reported.

Ira

> 
> -Denny
>