From: Leon Romanovsky <leon@kernel.org>
To: Doug Ledford <dledford@redhat.com>, Jason Gunthorpe <jgg@nvidia.com>
Cc: Aharon Landau <aharonl@nvidia.com>,
"David S. Miller" <davem@davemloft.net>,
Dennis Dalessandro <dennis.dalessandro@cornelisnetworks.com>,
Gal Pressman <galpress@amazon.com>,
Jakub Kicinski <kuba@kernel.org>,
linux-kernel@vger.kernel.org, linux-rdma@vger.kernel.org,
Maor Gottlieb <maorg@nvidia.com>,
Mark Zhang <markzhang@nvidia.com>,
Mike Marciniszyn <mike.marciniszyn@cornelisnetworks.com>,
Mustafa Ismail <mustafa.ismail@intel.com>,
netdev@vger.kernel.org, Potnuri Bharat Teja <bharat@chelsio.com>,
Saeed Mahameed <saeedm@nvidia.com>,
Selvin Xavier <selvin.xavier@broadcom.com>,
Shiraz Saleem <shiraz.saleem@intel.com>,
Yishai Hadas <yishaih@nvidia.com>,
Zhu Yanjun <zyjzyj2000@gmail.com>
Subject: Re: [PATCH rdma-next v3 05/13] RDMA/counter: Add an is_disabled field in struct rdma_hw_stats
Date: Wed, 6 Oct 2021 14:49:07 +0300 [thread overview]
Message-ID: <YV2NMzu0izn2vWrJ@unreal> (raw)
In-Reply-To: <09596ef74f4c213cb236e057e20e98264b67f16e.1633513239.git.leonro@nvidia.com>
On Wed, Oct 06, 2021 at 12:52:08PM +0300, Leon Romanovsky wrote:
> From: Aharon Landau <aharonl@nvidia.com>
>
> Add a bitmap in rdma_hw_stat structure, with each bit indicates whether
> the corresponding counter is currently disabled or not. By default
> hwcounters are enabled.
>
> Signed-off-by: Aharon Landau <aharonl@nvidia.com>
> Reviewed-by: Mark Zhang <markzhang@nvidia.com>
> Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
> ---
> drivers/infiniband/core/nldev.c | 11 ++++++++++-
> drivers/infiniband/core/verbs.c | 10 ++++++++++
> include/rdma/ib_verbs.h | 3 +++
> 3 files changed, 23 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/infiniband/core/nldev.c b/drivers/infiniband/core/nldev.c
> index 3f6b98a87566..67519730b1ac 100644
> --- a/drivers/infiniband/core/nldev.c
> +++ b/drivers/infiniband/core/nldev.c
> @@ -968,15 +968,21 @@ static int fill_stat_counter_hwcounters(struct sk_buff *msg,
> if (!table_attr)
> return -EMSGSIZE;
>
> - for (i = 0; i < st->num_counters; i++)
> + mutex_lock(&st->lock);
> + for (i = 0; i < st->num_counters; i++) {
> + if (test_bit(i, st->is_disabled))
> + continue;
> if (rdma_nl_stat_hwcounter_entry(msg, st->descs[i].name,
> st->value[i]))
> goto err;
> + }
> + mutex_unlock(&st->lock);
>
> nla_nest_end(msg, table_attr);
> return 0;
>
> err:
> + mutex_unlock(&st->lock);
> nla_nest_cancel(msg, table_attr);
> return -EMSGSIZE;
> }
> @@ -2104,6 +2110,9 @@ static int stat_get_doit_default_counter(struct sk_buff *skb,
> goto err_stats;
> }
> for (i = 0; i < num_cnts; i++) {
> + if (test_bit(i, stats->is_disabled))
> + continue;
> +
> v = stats->value[i] +
> rdma_counter_get_hwstat_value(device, port, i);
> if (rdma_nl_stat_hwcounter_entry(msg,
> diff --git a/drivers/infiniband/core/verbs.c b/drivers/infiniband/core/verbs.c
> index c3319a7584a2..d957b23a0e23 100644
> --- a/drivers/infiniband/core/verbs.c
> +++ b/drivers/infiniband/core/verbs.c
> @@ -2994,11 +2994,20 @@ struct rdma_hw_stats *rdma_alloc_hw_stats_struct(
> if (!stats)
> return NULL;
>
> + stats->is_disabled = kcalloc(BITS_TO_LONGS(num_counters),
> + sizeof(*stats->is_disabled), GFP_KERNEL);
> + if (!stats->is_disabled)
> + goto err;
> +
> stats->descs = descs;
> stats->num_counters = num_counters;
> stats->lifespan = msecs_to_jiffies(lifespan);
>
> return stats;
> +
> +err:
> + kfree(stats);
> + return NULL;
> }
> EXPORT_SYMBOL(rdma_alloc_hw_stats_struct);
>
> @@ -3008,6 +3017,7 @@ EXPORT_SYMBOL(rdma_alloc_hw_stats_struct);
> */
> void rdma_free_hw_stats_struct(struct rdma_hw_stats *stats)
> {
> + kfree(stats->is_disabled);
> kfree(stats);
Jason,
This hunk needs to be the following for the representors case. Can you
please fix it locally?
diff --git a/drivers/infiniband/core/verbs.c b/drivers/infiniband/core/verbs.c
index d957b23a0e23..e9e042b31386 100644
--- a/drivers/infiniband/core/verbs.c
+++ b/drivers/infiniband/core/verbs.c
@@ -3017,6 +3017,9 @@ EXPORT_SYMBOL(rdma_alloc_hw_stats_struct);
*/
void rdma_free_hw_stats_struct(struct rdma_hw_stats *stats)
{
+ if (!stats)
+ return;
+
kfree(stats->is_disabled);
kfree(stats);
}
Thanks
> }
> EXPORT_SYMBOL(rdma_free_hw_stats_struct);
> diff --git a/include/rdma/ib_verbs.h b/include/rdma/ib_verbs.h
> index 938c0c0a1c19..ae467365706b 100644
> --- a/include/rdma/ib_verbs.h
> +++ b/include/rdma/ib_verbs.h
> @@ -565,6 +565,8 @@ struct rdma_stat_desc {
> * their own value during their allocation routine.
> * @descs - Array of pointers to static descriptors used for the counters
> * in directory.
> + * @is_disabled - A bitmap to indicate each counter is currently disabled
> + * or not.
> * @num_counters - How many hardware counters there are. If name is
> * shorter than this number, a kernel oops will result. Driver authors
> * are encouraged to leave BUILD_BUG_ON(ARRAY_SIZE(@name) < num_counters)
> @@ -577,6 +579,7 @@ struct rdma_hw_stats {
> unsigned long timestamp;
> unsigned long lifespan;
> const struct rdma_stat_desc *descs;
> + unsigned long *is_disabled;
> int num_counters;
> u64 value[];
> };
> --
> 2.31.1
>
next prev parent reply other threads:[~2021-10-06 11:49 UTC|newest]
Thread overview: 16+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-10-06 9:52 [PATCH rdma-next v3 00/13] Optional counter statistics support Leon Romanovsky
2021-10-06 9:52 ` [PATCH mlx5-next v3 01/13] net/mlx5: Add ifc bits to support optional counters Leon Romanovsky
2021-10-06 9:52 ` [PATCH mlx5-next v3 02/13] net/mlx5: Add priorities for counters in RDMA namespaces Leon Romanovsky
2021-10-06 9:52 ` [PATCH rdma-next v3 03/13] RDMA/counter: Add a descriptor in struct rdma_hw_stats Leon Romanovsky
2021-10-06 9:52 ` [PATCH rdma-next v3 04/13] RDMA/core: Add a helper API rdma_free_hw_stats_struct Leon Romanovsky
2021-10-06 9:52 ` [PATCH rdma-next v3 05/13] RDMA/counter: Add an is_disabled field in struct rdma_hw_stats Leon Romanovsky
2021-10-06 11:49 ` Leon Romanovsky [this message]
2021-10-06 9:52 ` [PATCH rdma-next v3 06/13] RDMA/counter: Add optional counter support Leon Romanovsky
2021-10-06 9:52 ` [PATCH rdma-next v3 07/13] RDMA/nldev: Add support to get status of all counters Leon Romanovsky
2021-10-06 9:52 ` [PATCH rdma-next v3 08/13] RDMA/nldev: Split nldev_stat_set_mode_doit out of nldev_stat_set_doit Leon Romanovsky
2021-10-06 9:52 ` [PATCH rdma-next v3 09/13] RDMA/nldev: Allow optional-counter status configuration through RDMA netlink Leon Romanovsky
2021-10-07 13:07 ` Jason Gunthorpe
2021-10-06 9:52 ` [PATCH rdma-next v3 10/13] RDMA/mlx5: Support optional counters in hw_stats initialization Leon Romanovsky
2021-10-06 9:52 ` [PATCH rdma-next v3 11/13] RDMA/mlx5: Add steering support in optional flow counters Leon Romanovsky
2021-10-06 9:52 ` [PATCH rdma-next v3 12/13] RDMA/mlx5: Add modify_op_stat() support Leon Romanovsky
2021-10-06 9:52 ` [PATCH rdma-next v3 13/13] RDMA/mlx5: Add optional counter support in get_hw_stats callback Leon Romanovsky
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=YV2NMzu0izn2vWrJ@unreal \
--to=leon@kernel.org \
--cc=aharonl@nvidia.com \
--cc=bharat@chelsio.com \
--cc=davem@davemloft.net \
--cc=dennis.dalessandro@cornelisnetworks.com \
--cc=dledford@redhat.com \
--cc=galpress@amazon.com \
--cc=jgg@nvidia.com \
--cc=kuba@kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-rdma@vger.kernel.org \
--cc=maorg@nvidia.com \
--cc=markzhang@nvidia.com \
--cc=mike.marciniszyn@cornelisnetworks.com \
--cc=mustafa.ismail@intel.com \
--cc=netdev@vger.kernel.org \
--cc=saeedm@nvidia.com \
--cc=selvin.xavier@broadcom.com \
--cc=shiraz.saleem@intel.com \
--cc=yishaih@nvidia.com \
--cc=zyjzyj2000@gmail.com \
/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).