[PATCH net-next v5 1/8] __ptr_ring_full_next: Returns if ring will be full after next insertion
From: Simon Schippers <hidden>
Date: 2025-09-22 22:17:08
Also in:
kvm, lkml, virtualization
Subsystem:
the rest · Maintainer:
Linus Torvalds
Useful if the caller would like to act before the ptr_ring gets full after the next __ptr_ring_produce call. Because __ptr_ring_produce has a smp_wmb(), taking action before ensures memory ordering. Co-developed-by: Tim Gebauer <redacted> Signed-off-by: Tim Gebauer <redacted> Signed-off-by: Simon Schippers <redacted> --- include/linux/ptr_ring.h | 22 ++++++++++++++++++++++ 1 file changed, 22 insertions(+)
diff --git a/include/linux/ptr_ring.h b/include/linux/ptr_ring.h
index 551329220e4f..c45e95071d7e 100644
--- a/include/linux/ptr_ring.h
+++ b/include/linux/ptr_ring.h@@ -96,6 +96,28 @@ static inline bool ptr_ring_full_bh(struct ptr_ring *r) return ret; } +/* Returns if the ptr_ring will be full after inserting the next ptr. + */ +static inline bool __ptr_ring_full_next(struct ptr_ring *r) +{ + int p; + + /* Since __ptr_ring_discard_one invalidates in reverse order, the + * next producer entry might be NULL even though the current one + * is not. Therefore, also check the current producer entry with + * __ptr_ring_full. + */ + if (unlikely(r->size <= 1 || __ptr_ring_full(r))) + return true; + + p = r->producer + 1; + + if (unlikely(p >= r->size)) + p = 0; + + return r->queue[p]; +} + /* Note: callers invoking this in a loop must use a compiler barrier, * for example cpu_relax(). Callers must hold producer_lock. * Callers are responsible for making sure pointer that is being queued
--
2.43.0