Re: [PATCH v2 0/4] x86: fix get_nr_restart_syscall()

From: Thomas Gleixner
Date: Mon Feb 17 2020 - 10:54:09 EST

Andy Lutomirski <luto@xxxxxxxxxx> writes:

> On Wed, Dec 18, 2019 at 12:02 PM Linus Torvalds
> <torvalds@xxxxxxxxxxxxxxxxxxxx> wrote:
>> On Wed, Dec 18, 2019 at 7:19 AM Oleg Nesterov <oleg@xxxxxxxxxx> wrote:
>> >
>> > Andy, Linus, do you have any objections?
>> It's ok by me, no objections. I still don't love your "hide the bit in
>> thread flags over return to user space", and would still prefer it in
>> the restart block, but I don't care _that_ deeply.
> I'd rather stick it in restart_block. I'd also like to see the kernel
> *verify* that the variant of restart_syscall() that's invoked is the
> same as the variant that should be invoked. In my mind, very few
> syscalls say "I can't believe there are no major bugs in here" like
> restart_syscall(), and being conservative is nice.

Just mopping up my backlog. What happened to this?