Re: [PATCH] infiniband/mthca : Fix userland mapping of mthca infiniband cards in Xen dom0

[Roland Dreier <rdreier-FYB4Gu1CFyUAvxtiuMwx3w@public.gmane.org>]

 > > Finally as a stylistic thing I would probably prefer to see the
 > > vm_page_prot manipulation written as
 > >
 > >        vma->vm_page_prot = pgprot_noncached(vm_get_page_prot(vma->vm_flags));
 > >
 > 
 > Makes sense.
 > While digging, I've also seen things like :
 > 
 > vma->vm_page_prot = pgprot_noncached(vm_get_page_prot(vma->vm_flags | VM_IO));

This isn't the same, since it doesn't set VM_IO in vm_flags.

 > > wrapped up in a helper function; presumably every instance of
 > >
 > >        vma->vm_page_prot = pgprot_FOO(vma->vm_page_prot)
 > >
 > > under drivers/ would need the same fix for Xen.
 > 
 > The issue is actually only related to mappings of bus memory to
 > userland. In kernel space, ioremap() is generaly used, which I
 > understand sets the _PAGE_IOMAP flag.

Right, but setting vm_page_prot means a driver is doing userspace
mapping of something, and if it's setting pgprot, that would generally
mean it's not mapping RAM.

 - R.
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html