Re: Runtime trouble with commit dbd952127d (seccomp: introduce writer locking)

[Guenter Roeck <linux@roeck-us.net>]

On 08/10/2014 02:10 PM, Linus Torvalds wrote:
> On Sun, Aug 10, 2014 at 1:51 PM, Guenter Roeck <linux@roeck-us.net> wrote:
>> That means that the broken configuration is (CONFIG_DEBUG_SPINLOCK=n,
>> CONFIG_SMP=n).
>> It also means that the BUG_ON checks introduced with the seccomp commit
>> will cause this configuration to fail hard at least for architectures where
>> CONFIG_SMP
>> can be disabled, and if those architectures use
>> include/linux/spinlock_types_up.h.
>
> Yes. This is why we have "assert_spin_locked()". You can't use
> BUG_ON(spin_is_locked()), and !spin_is_locked() tends to be even worse
> unless you can prove that nobody else can get the lock simultaneously.
>

git grep 'BUG_ON.*!spin_is_locked'

suggests that this may be a spreading sickness ... unless the other users
know for sure that SMP will always be configured.

What is the better fix ? Add NR_CPUS != 1 to the test, as mm does,
or remove it entirely ? Or something else ?

Guenter