Re: [PATCH v2] uapi: futex: Add a futex syscall

On Tue, Oct 26, 2021 at 3:33 AM Arnd Bergmann [off-list ref] wrote:

On Mon, Oct 25, 2021 at 6:33 PM André Almeida [off-list ref] wrote:

quoted

Às 02:54 de 21/10/21, Alistair Francis escreveu:

quoted

From: Alistair Francis <redacted>

quoted

quoted

+#if defined(__NR_futex)
+     if (sizeof(*timeout) == sizeof(struct __kernel_old_timespec))
+             return syscall(__NR_futex, uaddr, op, val, timeout, uaddr2, val3);
+
+     if (timeout && timeout->tv_sec == (long)timeout->tv_sec) {
+             struct __kernel_old_timespec ts32;
+
+             ts32.tv_sec = (__kernel_long_t) timeout->tv_sec;> +             ts32.tv_nsec = (__kernel_long_t) timeout->tv_nsec;
+
+             return syscall(__NR_futex, uaddr, op, val, &ts32, uaddr2, val3);
+     } else if (!timeout) {
+             return syscall(__NR_futex, uaddr, op, val, NULL, uaddr2, val3);
+     }
+#endif

If I read this part right, you will always use ts32 for __NR_futex. I
know that it can be misleading, but __NR_futex uses ts64 in 64-bit
archs, so they shouldn't be converted to ts32 in those cases.

__kernel_old_timespec is the correct type for sys_futex() on all
architectures.

Maybe name the local variable 'ts' or 'ts_old' instead of 'ts32' then?

quoted

quoted

+{
+#if defined(__NR_futex_time64)
+     int ret =  syscall(__NR_futex_time64, uaddr, op, val, nr_requeue, uaddr2, val3);
+
+     if (ret == 0 || errno != ENOSYS)
+             return ret;
+#endif
+
+#if defined(__NR_futex)
+     return syscall(__NR_futex, uaddr, op, val, nr_requeue, uaddr2, val3);
+#endif
+
+     errno = ENOSYS;
+     return -1;
+}
+
+#endif /* _UAPI_LINUX_FUTEX_SYSCALL_H */

Sorry if this question was already asked but I didn't find it in the
thread: Should we go with wrappers for the most common op? Like:

__kernel_futex_wait(volatile uint32_t *uaddr, uint32_t val, struct
timespec *timeout)

__kernel_futex_wake(volatile uint32_t *uaddr, uint32_t nr_wake)

I had suggested having just a single function definition here, but having one
per argument type seems reasonable as well. Having one definition
per futex_op value would also be possible, but in that case I suppose
we need all 13 of them, not just two.

Sorry I have taken so long to look at this again. I have addressed
your other comments.

I don't have a strong preference here. I do like that by having a
generic __kernel_futex_syscall_timeout() it should be easier to
convert existing uses of SYSCALL() to the new function (the requeue is
it's own thing anyway).

Alistair

        Arnd