Re: [PATCH v2 14/17] resource: Move devmem revoke code to resource framework

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Thu, Oct 15, 2020 at 2:09 AM Jason Gunthorpe <jgg@xxxxxxxx> wrote:
>
> On Fri, Oct 09, 2020 at 11:28:54AM -0700, Dan Williams wrote:
> > On Fri, Oct 9, 2020 at 7:32 AM Jason Gunthorpe <jgg@xxxxxxxx> wrote:
> > >
> > > On Fri, Oct 09, 2020 at 04:24:45PM +0200, Daniel Vetter wrote:
> > > > On Fri, Oct 9, 2020 at 2:31 PM Jason Gunthorpe <jgg@xxxxxxxx> wrote:
> > > > >
> > > > > On Fri, Oct 09, 2020 at 09:59:31AM +0200, Daniel Vetter wrote:
> > > > >
> > > > > > +struct address_space *iomem_get_mapping(void)
> > > > > > +{
> > > > > > +     return iomem_inode->i_mapping;
> > > > >
> > > > > This should pair an acquire with the release below
> > > > >
> > > > > > +     /*
> > > > > > +      * Publish /dev/mem initialized.
> > > > > > +      * Pairs with smp_load_acquire() in revoke_iomem().
> > > > > > +      */
> > > > > > +     smp_store_release(&iomem_inode, inode);
> > > > >
> > > > > However, this seems abnormal, initcalls rarely do this kind of stuff
> > > > > with global data..
> > > > >
> > > > > The kernel crashes if this fs_initcall is raced with
> > > > > iomem_get_mapping() due to the unconditional dereference, so I think
> > > > > it can be safely switched to a simple assignment.
> > > >
> > > > Ah yes I checked this all, but forgot to correctly annotate the
> > > > iomem_get_mapping access. For reference, see b34e7e298d7a ("/dev/mem:
> > > > Add missing memory barriers for devmem_inode").
> > >
> > > Oh yikes, so revoke_iomem can run concurrently during early boot,
> > > tricky.
> >
> > It runs early because request_mem_region() can run before fs_initcall.
> > Rather than add an unnecessary lock just arrange for the revoke to be
> > skipped before the inode is initialized. The expectation is that any
> > early resource reservations will block future userspace mapping
> > attempts.
>
> Actually, on this point a simple WRITE_ONCE/READ_ONCE pairing is OK,
> Paul once explained that the pointer chase on the READ_ONCE side is
> required to be like an acquire - this is why rcu_dereference is just
> READ_ONCE

Hm so WRITE_ONCE doesn't have any barriers, and we'd need that for
updating the pointer. That would leave things rather inconsistent, so
I think I'll just leave it as-is for symmetry reasons. None of this
code matters for performance anyway, so micro-optimizing barriers
seems a bit silly.
-Daniel
-- 
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch



[Index of Archives]     [KVM ARM]     [KVM ia64]     [KVM ppc]     [Virtualization Tools]     [Spice Development]     [Libvirt]     [Libvirt Users]     [Linux USB Devel]     [Linux Audio Users]     [Yosemite Questions]     [Linux Kernel]     [Linux SCSI]     [XFree86]

  Powered by Linux