2021-03-28 03:28:00

by Max Filippov

[permalink] [raw]
Subject: [PATCH] xtensa: fix uaccess-related livelock in do_page_fault

If a uaccess (e.g. get_user()) triggers a fault and there's a
fault signal pending, the handler will return to the uaccess without
having performed a uaccess fault fixup, and so the CPU will immediately
execute the uaccess instruction again, whereupon it will livelock
bouncing between that instruction and the fault handler.

https://lore.kernel.org/lkml/[email protected]/

Cc: [email protected]
Reported-by: Mark Rutland <[email protected]>
Signed-off-by: Max Filippov <[email protected]>
---
arch/xtensa/mm/fault.c | 5 ++++-
1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/arch/xtensa/mm/fault.c b/arch/xtensa/mm/fault.c
index 7666408ce12a..95a74890c7e9 100644
--- a/arch/xtensa/mm/fault.c
+++ b/arch/xtensa/mm/fault.c
@@ -112,8 +112,11 @@ void do_page_fault(struct pt_regs *regs)
*/
fault = handle_mm_fault(vma, address, flags, regs);

- if (fault_signal_pending(fault, regs))
+ if (fault_signal_pending(fault, regs)) {
+ if (!user_mode(regs))
+ goto bad_page_fault;
return;
+ }

if (unlikely(fault & VM_FAULT_ERROR)) {
if (fault & VM_FAULT_OOM)
--
2.20.1