Re: [PATCH net-next] net: use ktime_t in struct scm_timestamping_internal | netdev | Ratatoskr

[PATCH net-next] net: use ktime_t in struct scm_timestamping_internal · Eric Dumazet <edumazet@google.com> · 2026-03-04
Re: [PATCH net-next] net: use ktime_t in struct scm_timestamping_internal · Willem de Bruijn <willemdebruijn.kernel@gmail.com> · 2026-03-04
Re: [PATCH net-next] net: use ktime_t in struct scm_timestamping_internal · Jason Xing <hidden> · 2026-03-04
Re: [PATCH net-next] net: use ktime_t in struct scm_timestamping_internal · Eric Dumazet <edumazet@google.com> · 2026-03-05
Re: [PATCH net-next] net: use ktime_t in struct scm_timestamping_internal · patchwork-bot+netdevbpf@kernel.org · 2026-03-05

Re: [PATCH net-next] net: use ktime_t in struct scm_timestamping_internal

From: Eric Dumazet <edumazet@google.com>
Date: 2026-03-05 02:56:18

On Wed, Mar 4, 2026 at 8:12 AM Jason Xing [off-list ref] wrote:

On Wed, Mar 4, 2026 at 9:30 AM Eric Dumazet [off-list ref] wrote:

quoted

-       if (has_timestamping) {
-               tss->ts[1] = (struct timespec64) {0};
+       if (tss->ts[0] | tss->ts[2]) {

Maybe use tss->ts[0] || tss->ts[2]? It's a bit weird to use a single |
as to readability.

This was done on purpose.
Code is more efficient, with a single conditional.

Thanks.