On Fri, Mar 06, 2020 at 05:15:19PM +0100, Jean-Philippe Brucker wrote: > On Fri, Mar 06, 2020 at 10:52:45AM -0400, Jason Gunthorpe wrote: > > On Fri, Mar 06, 2020 at 03:35:56PM +0100, Jean-Philippe Brucker wrote: > > > On Fri, Mar 06, 2020 at 09:09:19AM -0400, Jason Gunthorpe wrote: > > > > On Fri, Mar 06, 2020 at 10:56:14AM +0100, Jean-Philippe Brucker wrote: > > > > > I tried to keep it simple like that: normally mmu_notifier_get() is called > > > > > in bind(), and mmu_notifier_put() is called in unbind(). > > > > > > > > > > Multiple device drivers may call bind() with the same mm. Each bind() > > > > > calls mmu_notifier_get(), obtains the same io_mm, and returns a new bond > > > > > (a device<->mm link). Each bond is freed by calling unbind(), which calls > > > > > mmu_notifier_put(). > > > > > > > > > > That's the most common case. Now if the process is killed and the mm > > > > > disappears, we do need to avoid use-after-free caused by DMA of the > > > > > mappings and the page tables. > > > > > > > > This is why release must do invalidate all - but it doesn't need to do > > > > any more - as no SPTE can be established without a mmget() - and > > > > mmget() is no longer possible past release. > > > > > > In our case we don't have SPTEs, the whole pgd is shared between MMU and > > > IOMMU (isolated using PASID tables). > > > > Okay, but this just means that 'invalidate all' also requires > > switching the PASID to use some pgd that is permanently 'all fail'. > > > > > At this point no one told the device to stop working on this queue, > > > it may still be doing DMA on this address space. > > > > Sure, but there are lots of cases where a defective user space can > > cause pages under active DMA to disappear, like munmap for > > instance. Process exit is really no different, the PASID should take > > errors and the device & driver should do whatever error flow it has. > > We do have the possibility to shut things down in order, so to me this > feels like a band-aid. ->release() is called by exit_mmap which is called by mmput. There are over a 100 callsites to mmput() and I'm not totally sure what the rules are for release(). We've run into problems before with things like this. IMHO, due to this, it is best for release to be simple and have conservative requirements on context like all the other notifier callbacks. It is is not a good place to put complex HW fencing driver code. In particular that link you referenced is suggesting the driver tear down could take minutes - IMHO it is not OK to block mmput() for minutes. > The idea has come up before though [1], and I'm not strongly opposed > to this model, but I'm still not convinced it's necessary. It does > add more complexity to IOMMU drivers, to avoid printing out the > errors that we wouldn't otherwise see, whereas device drivers need > in any case to implement the logic that forces stop DMA. Errors should not be printed to the kernel log for PASID cases anyhow. PASID will be used by unpriv user, and unpriv user should not be able to trigger kernel prints at will, eg by doing dma to nmap VA or whatever. Process exit is just another case of this, and should not be treated specially. Jason
