MIPS: Fix endless loop when processing signals for kernel tasks
authorDmitry Adamushko <dmitry.adamushka_ext@softathome.com>
Thu, 5 Apr 2012 18:37:34 +0000 (20:37 +0200)
committerRalf Baechle <ralf@linux-mips.org>
Wed, 5 Dec 2012 18:59:00 +0000 (19:59 +0100)
commitc90e6fbb220d44988cb65af3707367c57cdb65a8
treeb007883a08b72d2a3a22c12c91ceaf139e014373
parent2d33976fb395a5b9cb00f2c24132b3f272eb9a9d
MIPS: Fix endless loop when processing signals for kernel tasks

The problem occurs [1] when a kernel-mode task returns from a system
call with a pending signal.

A real-life scenario is a child of 'khelper' returning from a failed
kernel_execve() in ____call_usermodehelper() [ kernel/kmod.c ].
kernel_execve() fails due to a pending SIGKILL, which is the result of
"kill -9 -1" (at least, busybox's init does it upon reboot).

The loop is as follows:

* syscall_exit_work:
 - work_pending:            // start_of_the_loop
 - work_notifysig:
   - do_notify_resume()
     - do_signal()
       - if (!user_mode(regs)) return;
 - resume_userspace         // TIF_SIGPENDING is still set
 - work_pending             // so we call work_pending => goto
                            // start_of_the_loop

More information can be found in another LKML thread:
http://www.serverphorums.com/read.php?12,457826

[1] The problem was also reproduced on !CONFIG_VM86 x86, and the
following fix was accepted.

http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=29a2e2836ff9ea65a603c89df217f4198973a74f

Signed-off-by: Dmitry Adamushko <dmitry.adamushko@gmail.com>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/3571/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
arch/mips/kernel/entry.S