Re: [PATCH] IB/core: Make ib_mad_client_id atomic

[jackm]

On Mon, 30 Apr 2018 13:10:49 -0400
Doug Ledford <dledford@xxxxxxxxxx> wrote:

Looks good!

-Jack

> On Mon, 2018-04-30 at 08:49 -0600, Jason Gunthorpe wrote:
> > On Mon, Apr 23, 2018 at 10:16:18PM +0300, jackm wrote:
> >   
> > > > > TIDs need to be globally unique on the entire machine.    
> > > Jason, that is not exactly correct.  
> > 
> > The expecation for /dev/umad users is that they all receive locally
> > unique TID prefixes. The kernel may be OK to keep things
> > port-specific but it is slightly breaking the API we are presenting
> > to userspace to allow them to alias..
> > 
> > Jason  
> 
> Would people be happier with this commit message then:
> 
> IB/core: Make ib_mad_client_id atomic
>   
> Currently, the kernel protects access to the agent ID allocator on a
> per port basis using a spinlock, so it is impossible for two
> apps/threads on the same port to get the same TID, but it is entirely
> possible for two threads on different ports to end up with the same
> TID.  
> 
> As this can be confusing (regardless of it being legal according to
> the IB Spec 1.3, C13-18.1.1, in section 13.4.6.4 - TransactionID
> usage), and as the rdma-core user space API for /dev/umad devices
> implies unique TIDs even across ports, make the TID an atomic type so
> that no two allocations, regardless of port number, will be the same.
> 
> Signed-off-by: Håkon Bugge <haakon.bugge@xxxxxxxxxx>
> Reviewed-by: Jack Morgenstein <jackm@xxxxxxxxxxxxxxxxxx>
> Reviewed-by: Ira Weiny <ira.weiny@xxxxxxxxx>
> Reviewed-by: Zhu Yanjun <yanjun.zhu@xxxxxxxxxx>
> Signed-off-by: Doug Ledford <dledford@xxxxxxxxxx>
> 
>