> In debugging the failure of the glibc tst-cond18 test on parisc, I
> realized
> that futexes need to use the same locks the lws calls. This fixes all the
> pthread 'cond' tests. Sadly, there are still problems with thread
> cancellation.
>
> Signed-of-by: John David Anglin <dave.anglin@xxxxxxxx>
>
> diff --git a/arch/parisc/include/asm/futex.h
> b/arch/parisc/include/asm/futex.h
> index 2388bdb..a8d0586 100644
> --- a/arch/parisc/include/asm/futex.h
> +++ b/arch/parisc/include/asm/futex.h
> @@ -8,6 +8,29 @@
> #include <asm/atomic.h>
> #include <asm/errno.h>
>
> +/* The following has to match the LWS code in syscall.S. We have
> + sixteen four-word locks. */
> +
> +static inline void
> +_futex_spin_lock_irqsave (u32 __user *uaddr, unsigned long int *flags)
> +{
> + extern u32 lws_lock_start[];
> + long index = ((long)uaddr & 0xf0) >> 2;
> + arch_spinlock_t *s = (arch_spinlock_t *)&lws_lock_start[index];
> + local_irq_save(*flags);
> + arch_spin_lock(s);
> +}
> +
> +static inline void
> +_futex_spin_unlock_irqrestore (u32 __user *uaddr, unsigned long int
> *flags)
> +{
> + extern u32 lws_lock_start[];
> + long index = ((long)uaddr & 0xf0) >> 2;
> + arch_spinlock_t *s = (arch_spinlock_t *)&lws_lock_start[index];
> + arch_spin_unlock(s);
> + local_irq_restore(*flags);
> +}
> +
> static inline int
> futex_atomic_op_inuser (int encoded_op, u32 __user *uaddr)
> {
This needs tabs.
Greetings,
Eike
--
To unsubscribe from this list: send the line "unsubscribe linux-parisc" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
