[v3,0/2] mm/memory-failure: Poison read receives SIGKILL instead of SIGBUS issue

Message

Jane Chu July 25, 2019, 10:01 p.m. UTC

Changes in v3:
 - move **tk cleanup to its own patch

Changes in v2:
 - move 'tk' allocations internal to add_to_kill(), suggested by Dan;
 - ran checkpatch.pl check, pointed out by Matthew;
 - Noaya pointed out that v1 would have missed the SIGKILL
   if "tk->addr == -EFAULT", since the code returns early.
   Incorporated Noaya's suggestion, also, skip VMAs where
   "tk->size_shift == 0" for zone device page, and deliver SIGBUS
   when "tk->size_shift != 0" so the payload is helpful;
 - added Suggested-by: Naoya Horiguchi <[e-mail]>

Jane Chu (2):
  mm/memory-failure.c clean up around tk pre-allocation
  mm/memory-failure: Poison read receives SIGKILL instead of SIGBUS if
    mmaped more than once

 mm/memory-failure.c | 62 ++++++++++++++++++++++-------------------------------
 1 file changed, 26 insertions(+), 36 deletions(-)