[PATCH 0/2] iommu: Remove iommu_sva_ops::mm_exit()

Jason Gunthorpe jgg at ziepe.ca
Fri Apr 10 02:58:58 AEST 2020


On Thu, Apr 09, 2020 at 09:21:34AM -0700, Jacob Pan wrote:
> On Thu, 9 Apr 2020 11:25:19 -0300
> Jason Gunthorpe <jgg at ziepe.ca> wrote:
> 
> > On Thu, Apr 09, 2020 at 07:14:24AM -0700, Jacob Pan wrote:
> > > > When the process is killed, mm release can happen before fds are
> > > > released. If you look at do_exit() in kernel/exit.c:
> > > > 
> > > > 	exit_mm()
> > > > 	  mmput()  
> > > > 	   -> mmu release notifier    
> > > > 	...
> > > > 	exit_files()
> > > > 	  close_files()
> > > > 	    fput()
> > > > 	exit_task_work()
> > > > 	  __fput()  
> > > > 	   -> unbind()    
> > > >   
> > > So unbind is coming anyway, the difference in handling in mmu
> > > release notifier is whether we silently drop DMA fault vs.
> > > reporting fault?  
> > 
> > Userspace can significantly delay the final fput triggering the
> > unbind, the above is only for the trivial case where the process
> > owning the mm_struct is the only process holding the fd.
> > 
> Are you talking about FDs owned buy children after fork() or FDs sent
> over to another process. I think, in either case SVA is not supported.

Supported or not a hostile user space can trigger these conditions and
it should not cause misbehavior from the kernel (eg log spamming)

Jason


More information about the Linux-accelerators mailing list