On Fri, Apr 28, 2023 at 11:27:53AM +0200, Claudio Imbrenda wrote: > Fix a potential race in gmap_make_secure() and remove the last user of > follow_page() without FOLL_GET. > > The old code is locking something it doesn't have a reference to, and > as explained by Jason and David in this discussion: > https://lore.kernel.org/linux-mm/Y9J4P%2FRNvY1Ztn0Q@xxxxxxxxxx/ > it can lead to all kind of bad things, including the page getting > unmapped (MADV_DONTNEED), freed, reallocated as a larger folio and the > unlock_page() would target the wrong bit. > There is also another race with the FOLL_WRITE, which could race > between the follow_page() and the get_locked_pte(). > > The main point is to remove the last use of follow_page() without > FOLL_GET or FOLL_PIN, removing the races can be considered a nice > bonus. > > Signed-off-by: Claudio Imbrenda <imbrenda@xxxxxxxxxxxxx> > Link: https://lore.kernel.org/linux-mm/Y9J4P%2FRNvY1Ztn0Q@xxxxxxxxxx/ > Suggested-by: Jason Gunthorpe <jgg@xxxxxxxxxx> > Fixes: 214d9bbcd3a6 ("s390/mm: provide memory management functions for protected KVM guests") > --- > arch/s390/kernel/uv.c | 32 +++++++++++--------------------- > 1 file changed, 11 insertions(+), 21 deletions(-) Reviewed-by: Jason Gunthorpe <jgg@xxxxxxxxxx> Jason
