From: Chuck Lever <chuck.lever@oracle.com>
To: Josef Bacik <josef@toxicpanda.com>
Cc: linux-nfs@vger.kernel.org, kernel-team@fb.com
Subject: Re: [PATCH v2 10/13] nfsd: move th_cnt into nfsd_net
Date: Thu, 25 Jan 2024 16:01:27 -0500 [thread overview]
Message-ID: <ZbLMJxLWIvomQIzO@tissot.1015granger.net> (raw)
In-Reply-To: <0fa7bf5b5bbc863180e50363435b5a56c43dc5e3.1706212208.git.josef@toxicpanda.com>
On Thu, Jan 25, 2024 at 02:53:20PM -0500, Josef Bacik wrote:
> This is the last global stat, move it into nfsd_net and adjust all the
> users to use that variant instead of the global one.
Hm. I thought nfsd threads were a global resource -- they service
all network namespaces. So, shouldn't the same thread count be
surfaced to all containers? Won't they all see all of the nfsd
processes?
> Signed-off-by: Josef Bacik <josef@toxicpanda.com>
> ---
> fs/nfsd/netns.h | 3 +++
> fs/nfsd/nfssvc.c | 4 ++--
> fs/nfsd/stats.c | 4 ++--
> fs/nfsd/stats.h | 6 ------
> 4 files changed, 7 insertions(+), 10 deletions(-)
>
> diff --git a/fs/nfsd/netns.h b/fs/nfsd/netns.h
> index 0cef4bb407a9..8d3f4cb7cab4 100644
> --- a/fs/nfsd/netns.h
> +++ b/fs/nfsd/netns.h
> @@ -179,6 +179,9 @@ struct nfsd_net {
> /* Per-netns stats counters */
> struct percpu_counter counter[NFSD_STATS_COUNTERS_NUM];
>
> + /* number of available threads */
> + atomic_t th_cnt;
> +
> /* longest hash chain seen */
> unsigned int longest_chain;
>
> diff --git a/fs/nfsd/nfssvc.c b/fs/nfsd/nfssvc.c
> index d98a6abad990..0961b95dcef6 100644
> --- a/fs/nfsd/nfssvc.c
> +++ b/fs/nfsd/nfssvc.c
> @@ -924,7 +924,7 @@ nfsd(void *vrqstp)
>
> current->fs->umask = 0;
>
> - atomic_inc(&nfsdstats.th_cnt);
> + atomic_inc(&nn->th_cnt);
>
> set_freezable();
>
> @@ -940,7 +940,7 @@ nfsd(void *vrqstp)
> nfsd_file_net_dispose(nn);
> }
>
> - atomic_dec(&nfsdstats.th_cnt);
> + atomic_dec(&nn->th_cnt);
>
> out:
> /* Release the thread */
> diff --git a/fs/nfsd/stats.c b/fs/nfsd/stats.c
> index 44e275324b06..360e6dbf4e5c 100644
> --- a/fs/nfsd/stats.c
> +++ b/fs/nfsd/stats.c
> @@ -27,7 +27,6 @@
>
> #include "nfsd.h"
>
> -struct nfsd_stats nfsdstats;
> struct svc_stat nfsd_svcstats = {
> .program = &nfsd_program,
> };
> @@ -47,7 +46,7 @@ static int nfsd_show(struct seq_file *seq, void *v)
> percpu_counter_sum_positive(&nn->counter[NFSD_STATS_IO_WRITE]));
>
> /* thread usage: */
> - seq_printf(seq, "th %u 0", atomic_read(&nfsdstats.th_cnt));
> + seq_printf(seq, "th %u 0", atomic_read(&nn->th_cnt));
>
> /* deprecated thread usage histogram stats */
> for (i = 0; i < 10; i++)
> @@ -112,6 +111,7 @@ void nfsd_percpu_counters_destroy(struct percpu_counter counters[], int num)
>
> int nfsd_stat_counters_init(struct nfsd_net *nn)
> {
> + atomic_set(&nn->th_cnt, 0);
> return nfsd_percpu_counters_init(nn->counter, NFSD_STATS_COUNTERS_NUM);
> }
>
> diff --git a/fs/nfsd/stats.h b/fs/nfsd/stats.h
> index c24be4ddbe7d..5675d283a537 100644
> --- a/fs/nfsd/stats.h
> +++ b/fs/nfsd/stats.h
> @@ -10,12 +10,6 @@
> #include <uapi/linux/nfsd/stats.h>
> #include <linux/percpu_counter.h>
>
> -struct nfsd_stats {
> - atomic_t th_cnt; /* number of available threads */
> -};
> -
> -extern struct nfsd_stats nfsdstats;
> -
> extern struct svc_stat nfsd_svcstats;
>
> int nfsd_percpu_counters_init(struct percpu_counter *counters, int num);
> --
> 2.43.0
>
>
--
Chuck Lever
next prev parent reply other threads:[~2024-01-25 21:01 UTC|newest]
Thread overview: 30+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-01-25 19:53 [PATCH v2 00/13] Make nfs and nfsd stats visible in network ns Josef Bacik
2024-01-25 19:53 ` [PATCH v2 01/13] sunrpc: don't change ->sv_stats if it doesn't exist Josef Bacik
2024-01-25 19:53 ` [PATCH v2 02/13] nfs: stop setting ->pg_stats for unused stats Josef Bacik
2024-01-25 19:53 ` [PATCH v2 03/13] sunrpc: pass in the sv_stats struct through svc_create* Josef Bacik
2024-01-25 20:56 ` Chuck Lever
2024-01-25 21:56 ` Josef Bacik
2024-01-25 19:53 ` [PATCH v2 04/13] sunrpc: remove ->pg_stats from svc_program Josef Bacik
2024-01-25 19:53 ` [PATCH v2 05/13] sunrpc: add a struct rpc_stats arg to rpc_create_args Josef Bacik
2024-01-25 20:53 ` Chuck Lever
2024-01-25 21:54 ` Josef Bacik
2024-01-25 22:30 ` Jeff Layton
2024-01-26 13:49 ` Chuck Lever III
2024-01-25 19:53 ` [PATCH v2 06/13] sunrpc: use the struct net as the svc proc private Josef Bacik
2024-01-25 19:53 ` [PATCH v2 07/13] nfsd: rename NFSD_NET_* to NFSD_STATS_* Josef Bacik
2024-01-25 19:53 ` [PATCH v2 08/13] nfsd: expose /proc/net/sunrpc/nfsd in net namespaces Josef Bacik
2024-01-25 19:53 ` [PATCH v2 09/13] nfsd: make all of the nfsd stats per-network namespace Josef Bacik
2024-01-25 19:53 ` [PATCH v2 10/13] nfsd: move th_cnt into nfsd_net Josef Bacik
2024-01-25 21:01 ` Chuck Lever [this message]
2024-01-25 21:56 ` Josef Bacik
2024-01-26 13:01 ` Jeff Layton
2024-01-26 13:48 ` Chuck Lever III
2024-01-26 14:08 ` Jeff Layton
2024-01-26 14:27 ` Chuck Lever III
2024-01-26 15:03 ` Jeff Layton
2024-01-26 15:16 ` Chuck Lever III
2024-01-26 15:35 ` Jeff Layton
2024-01-25 19:53 ` [PATCH v2 11/13] nfsd: make svc_stat per-network namespace instead of global Josef Bacik
2024-01-25 19:53 ` [PATCH v2 12/13] nfs: expose /proc/net/sunrpc/nfs in net namespaces Josef Bacik
2024-01-25 19:53 ` [PATCH v2 13/13] nfs: make the rpc_stat per net namespace Josef Bacik
2024-01-26 13:12 ` [PATCH v2 00/13] Make nfs and nfsd stats visible in network ns Jeff Layton
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=ZbLMJxLWIvomQIzO@tissot.1015granger.net \
--to=chuck.lever@oracle.com \
--cc=josef@toxicpanda.com \
--cc=kernel-team@fb.com \
--cc=linux-nfs@vger.kernel.org \
/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