From: hare@kernel.org
To: Christoph Hellwig <hch@lst.de>
Cc: Keith Busch <kbusch@kernel.org>, Sagi Grimberg <sagi@grimberg.me>,
linux-nvme@lists.infradead.org, Hannes Reinecke <hare@suse.de>,
Hannes Reinecke <hare@kernel.org>
Subject: [PATCH 1/6] nvme-multipath: do not assign ->current_path in __nvme_find_path()
Date: Tue, 29 Jul 2025 09:06:48 +0200 [thread overview]
Message-ID: <20250729070653.125258-2-hare@kernel.org> (raw)
In-Reply-To: <20250729070653.125258-1-hare@kernel.org>
From: Hannes Reinecke <hare@suse.de>
__nvme_find_path() is the fallback if no paths are selected, but
setting ->current_path[] is only relevant for 'numa' and 'round-robin'
iopolicies. So move it out of __nvme_find_path() to avoid a pointless
assignment for other path checkers.
Signed-off-by: Hannes Reinecke <hare@kernel.org>
---
drivers/nvme/host/multipath.c | 40 +++++++++++++++++++++++------------
1 file changed, 27 insertions(+), 13 deletions(-)
diff --git a/drivers/nvme/host/multipath.c b/drivers/nvme/host/multipath.c
index 3da980dc60d9..116b2e71d339 100644
--- a/drivers/nvme/host/multipath.c
+++ b/drivers/nvme/host/multipath.c
@@ -342,8 +342,6 @@ static struct nvme_ns *__nvme_find_path(struct nvme_ns_head *head, int node)
if (!found)
found = fallback;
- if (found)
- rcu_assign_pointer(head->current_path[node], found);
return found;
}
@@ -364,8 +362,13 @@ static struct nvme_ns *nvme_round_robin_path(struct nvme_ns_head *head)
struct nvme_ns *old = srcu_dereference(head->current_path[node],
&head->srcu);
- if (unlikely(!old))
- return __nvme_find_path(head, node);
+ if (unlikely(!old)) {
+ ns = __nvme_find_path(head, node);
+ if (ns)
+ rcu_assign_pointer(head->current_path[node], found);
+ return ns;
+ }
+
if (list_is_singular(&head->list)) {
if (nvme_path_is_disabled(old))
@@ -455,9 +458,11 @@ static struct nvme_ns *nvme_numa_path(struct nvme_ns_head *head)
ns = srcu_dereference(head->current_path[node], &head->srcu);
if (unlikely(!ns))
- return __nvme_find_path(head, node);
- if (unlikely(!nvme_path_is_optimized(ns)))
- return __nvme_find_path(head, node);
+ ns = __nvme_find_path(head, node);
+ else if (unlikely(!nvme_path_is_optimized(ns)))
+ ns = __nvme_find_path(head, node);
+ if (ns)
+ rcu_assign_pointer(head->current_path[node], found);
return ns;
}
@@ -798,12 +803,21 @@ static void nvme_mpath_set_live(struct nvme_ns *ns)
mutex_lock(&head->lock);
if (nvme_path_is_optimized(ns)) {
- int node, srcu_idx;
-
- srcu_idx = srcu_read_lock(&head->srcu);
- for_each_online_node(node)
- __nvme_find_path(head, node);
- srcu_read_unlock(&head->srcu, srcu_idx);
+ int srcu_idx, iopolicy;
+
+ iopolicy = READ_ONCE(head->subsys->iopolicy);
+ if (iopolicy == NVME_IOPOLICY_NUMA ||
+ iopolicy == NVME_IOPOLICY_RR) {
+ int node;
+
+ srcu_idx = srcu_read_lock(&head->srcu);
+ for_each_online_node(node) {
+ struct nvme_ns *ns = __nvme_find_path(head, node);
+ if (ns)
+ rcu_assign_pointer(head->current_path[node], ns);
+ }
+ srcu_read_unlock(&head->srcu, srcu_idx);
+ }
}
mutex_unlock(&head->lock);
--
2.43.0
next prev parent reply other threads:[~2025-07-29 7:07 UTC|newest]
Thread overview: 13+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-07-29 7:06 [RFC PATCH 0/6] nvme multipath eBPF path selector hare
2025-07-29 7:06 ` hare [this message]
2025-07-29 7:06 ` [PATCH 2/6] nvme: export nvme_find_get_subsystem()/nvme_put_subsystem() hare
2025-07-29 7:06 ` [PATCH 3/6] nvme: add per-namespace iopolicy sysfs attribute hare
2025-07-29 7:06 ` [PATCH 4/6] nvme: add 'sector' parameter to nvme_find_path() hare
2025-07-29 7:06 ` [PATCH 5/6] nvme-bpf: eBPF struct_ops path selectors hare
2025-07-29 7:06 ` [PATCH 6/6] tools/testing/selftests: add sample nvme bpf path selector hare
2025-07-30 2:03 ` Geliang Tang
2025-07-30 5:56 ` Hannes Reinecke
2025-07-29 7:54 ` [RFC PATCH 0/6] nvme multipath eBPF " Christoph Hellwig
2025-07-29 14:53 ` Mike Christie
2025-07-30 14:06 ` Christoph Hellwig
2025-07-30 2:03 ` Geliang Tang
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20250729070653.125258-2-hare@kernel.org \
--to=hare@kernel.org \
--cc=hare@suse.de \
--cc=hch@lst.de \
--cc=kbusch@kernel.org \
--cc=linux-nvme@lists.infradead.org \
--cc=sagi@grimberg.me \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).