From: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
To: allison.henderson@oracle.com, netdev@vger.kernel.org,
linux-rdma@vger.kernel.org, rds-devel@oss.oracle.com,
andrew@lunn.ch, guro@fb.com, tj@kernel.org, kernel-team@fb.com,
surenb@google.com, peterz@infradead.org, hannes@cmpxchg.org,
mkoutny@suse.com, cgroups@vger.kernel.org,
roman.gushchin@linux.dev
Subject: Re: [PATCH RFC v1] rds: Expose feature parameters via sysfs and ELF note
Date: Tue, 10 Jun 2025 16:51:45 -0400 [thread overview]
Message-ID: <aEia4cEFb0n9S_jj@char.us.oracle.com> (raw)
In-Reply-To: <20250610191144.422161-2-konrad.wilk@oracle.com>
On Tue, Jun 10, 2025 at 12:27:25PM -0400, Konrad Rzeszutek Wilk wrote:
Hi!
I've added some extra folks on the To: list to solicit feedback on this
idea since something similar was done via
3958e2d0c34e1 cgroup: make per-cgroup pressure stall tracking configurable
01ee6cfb1483f cgroup: export list of delegatable control files using sysfs
5f2e673405b74 cgroup: export list of cgroups v2 features using sysfs
Thank you for you time!
> We would like to have a programatic way for applications
> to query which of the features defined in include/uapi/linux/rds.h
> are actually implemented by the kernel.
>
> The problem is that applications can be built against newer
> kernel (or older) and they may have the feature implemented or not.
>
> The lack of a certain feature would signify that the kernel
> does not support it. The presence of it signifies the existence
> of it.
>
> This would provide the application to query the sysfs and figure
> out what is supported (and which ones are deprecated) and also
> what ioctl number to use for the specific feature (albeit that
> is already in include/uapi/linux/rds.h but this is an extra
> check if someone messed up).
>
> This patch would expose these extra sysfs values:
>
> /sys/module/rds/parameters/rds_ioctl_get_tos: 35297
> /sys/module/rds/parameters/rds_ioctl_set_tos: 35296
> /sys/module/rds/parameters/rds_socket_cancel_sent_to: 1
> /sys/module/rds/parameters/rds_socket_cong_monitor: 6
> /sys/module/rds/parameters/rds_socket_free_mr: 3
> /sys/module/rds/parameters/rds_socket_get_mr: 2
> /sys/module/rds/parameters/rds_socket_get_mr_for_dest: 7
> /sys/module/rds/parameters/rds_socket_recverr: 5
> /sys/module/rds/parameters/rds_socket_so_rxpath_latency: 9
> /sys/module/rds/parameters/rds_socket_so_transport: 8
> /sys/module/rds/parameters/rds_so_transport_ib: 0
> /sys/module/rds/parameters/rds_so_transport_tcp: 2
>
> Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
> ---
> Documentation/ABI/stable/sysfs-driver-rds | 92 +++++++++++++++++++++++
> net/rds/af_rds.c | 33 ++++++++
> 2 files changed, 125 insertions(+)
> create mode 100644 Documentation/ABI/stable/sysfs-driver-rds
>
> diff --git a/Documentation/ABI/stable/sysfs-driver-rds b/Documentation/ABI/stable/sysfs-driver-rds
> new file mode 100644
> index 000000000000..dcb1a335c5d6
> --- /dev/null
> +++ b/Documentation/ABI/stable/sysfs-driver-rds
> @@ -0,0 +1,92 @@
> +What: /sys/module/rds/parameters/rds_ioctl_set_tos
> +Date: Jun 2025
> +Contact: rds-devel@oss.oracle.com
> +Description:
> + The RDS driver supports the mechanism to set on a socket
> + the Quality of Service.
> +
> + The returned value is the socket ioctl number and this is read-only.
> +
> +What: /sys/module/rds/parameters/rds_ioctl_get_tos
> +Date: Jun 2025
> +Contact: rds-devel@oss.oracle.com
> +Description:
> + The RDS driver supports the mechanism to get on a socket
> + the Quality of Service.
> +
> + The returned value is the socket ioctl number and this is read-only.
> +
> +What: /sys/module/rds/parameters/rds_socket_cancel_sent_to
> +Date: Jun 2025
> +Contact: rds-devel@oss.oracle.com
> +Description:
> + The RDS driver supports the mechanism to cancel all pending
> + messages to a given destination.
> +
> + The returned value is the ioctl number and this is read-only.
> +
> +What: /sys/module/rds/parameters/rds_socket_get_mr
> +Date: Jun 2025
> +Contact: rds-devel@oss.oracle.com
> +Description:
> + The RDS driver supports the mechanism to retrieve the memory
> + ranges for the RDMA calls to setsockopt.
> +
> + The returned value is the ioctl number and this is read-only.
> +
> +What: /sys/module/rds/parameters/rds_socket_free_mr
> +Date: Jun 2025
> +Contact: rds-devel@oss.oracle.com
> +Description:
> + The RDS driver supports the mechanism to release the memory
> + ranges for the RDMA calls to setsockopt.
> +
> + The returned value is the ioctl number and this is read-only.
> +
> +What: /sys/module/rds/parameters/rds_socket_recverr
> +Date: Jun 2025
> +Contact: rds-devel@oss.oracle.com
> +Description:
> + The RDS driver supports the mechanism to send RDMA notifications
> + for any RDMA operation that fails.
> +
> + The returned value is the ioctl number and this is read-only.
> +
> +What: /sys/module/rds/parameters/rds_socket_cong_monitor
> +Date: Jun 2025
> +Contact: rds-devel@oss.oracle.com
> +Description:
> + The RDS driver supports mechanism to provide Congestion updates via
> + RDS_CMSG_CONG_UPDATE control messages.
> +
> + The returned value is the ioctl number and this is read-only.
> +
> +What: /sys/module/rds/parameters/rds_socket_get_mr_for_dest
> +Date: Jun 2025
> +Contact: rds-devel@oss.oracle.com
> +Description:
> + The returned value is the ioctl number and this is read-only.
> +
> +What: /sys/module/rds/parameters/rds_socket_so_transport
> +Date: Jun 2025
> +Contact: rds-devel@oss.oracle.com
> +Description:
> + The returned value is the ioctl number and this is read-only.
> +
> +What: /sys/module/rds/parameters/rds_socket_so_rxpath_latency
> +Date: Jun 2025
> +Contact: rds-devel@oss.oracle.com
> +Description:
> + The returned value is the ioctl number and this is read-only.
> +
> +What: /sys/module/rds/parameters/rds_so_transport_ib
> +Date: Jun 2025
> +Contact: rds-devel@oss.oracle.com
> +Description:
> + The returned value for the IB transport ioctl number and this is read-only.
> +
> +What: /sys/module/rds/parameters/rds_so_transport_tcp
> +Date: Jun 2025
> +Contact: rds-devel@oss.oracle.com
> +Description:
> + The returned value is the TCP transport number and this is read-only.
> diff --git a/net/rds/af_rds.c b/net/rds/af_rds.c
> index 8435a20968ef..15c8ded02dfb 100644
> --- a/net/rds/af_rds.c
> +++ b/net/rds/af_rds.c
> @@ -31,6 +31,7 @@
> *
> */
> #include <linux/module.h>
> +#include <linux/elfnote.h>
> #include <linux/errno.h>
> #include <linux/kernel.h>
> #include <linux/gfp.h>
> @@ -960,3 +961,35 @@ MODULE_DESCRIPTION("RDS: Reliable Datagram Sockets"
> MODULE_VERSION(DRV_VERSION);
> MODULE_LICENSE("Dual BSD/GPL");
> MODULE_ALIAS_NETPROTO(PF_RDS);
> +
> +#define RDS_IOCTL(feature, val) ELFNOTE64("rds.ioctl_" #feature, 0, val); \
> + unsigned int rds_ioctl_##feature = val; \
> + module_param(rds_ioctl_##feature, int, 0444)
> +
> +#define RDS_SOCKET(feature, val) ELFNOTE64("rds.socket_" #feature, 0, val); \
> + unsigned int rds_socket_##feature = val; \
> + module_param(rds_socket_##feature, int, 0444)
> +
> +#define RDS_SO_TRANSPORT(feature, val) ELFNOTE64("rds.so_transport_" #feature, 0, val); \
> + unsigned int rds_so_transport_##feature = val; \
> + module_param(rds_so_transport_##feature, int, 0444)
> +
> +/* The values used here correspond to include/uapi/linux/rds.h values */
> +
> +RDS_IOCTL(set_tos, SIOCRDSSETTOS);
> +RDS_IOCTL(get_tos, SIOCRDSGETTOS);
> +
> +/* Advertise setsocket/getsocket options. */
> +
> +RDS_SOCKET(cancel_sent_to, RDS_CANCEL_SENT_TO);
> +RDS_SOCKET(get_mr, RDS_GET_MR);
> +RDS_SOCKET(free_mr, RDS_FREE_MR);
> +RDS_SOCKET(recverr, RDS_RECVERR);
> +RDS_SOCKET(cong_monitor, RDS_CONG_MONITOR);
> +RDS_SOCKET(get_mr_for_dest, RDS_GET_MR_FOR_DEST);
> +RDS_SOCKET(so_transport, SO_RDS_TRANSPORT);
> +RDS_SOCKET(so_rxpath_latency, SO_RDS_MSG_RXPATH_LATENCY);
> +
> +/* The transport mechanisms. */
> +RDS_SO_TRANSPORT(ib, RDS_TRANS_IB);
> +RDS_SO_TRANSPORT(tcp, RDS_TRANS_TCP);
> --
> 2.43.5
>
prev parent reply other threads:[~2025-06-10 20:52 UTC|newest]
Thread overview: 5+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <20250610191144.422161-1-konrad.wilk@oracle.com>
2025-06-10 20:47 ` [rds-devel] [PATCH RFC v1] Feature reporting of RDS driver Konrad Rzeszutek Wilk
2025-06-10 21:32 ` Tejun Heo
2025-06-10 23:15 ` Konrad Rzeszutek Wilk
2025-06-16 17:45 ` Tejun Heo
[not found] ` <20250610191144.422161-2-konrad.wilk@oracle.com>
2025-06-10 20:51 ` Konrad Rzeszutek Wilk [this message]
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=aEia4cEFb0n9S_jj@char.us.oracle.com \
--to=konrad.wilk@oracle.com \
--cc=allison.henderson@oracle.com \
--cc=andrew@lunn.ch \
--cc=cgroups@vger.kernel.org \
--cc=guro@fb.com \
--cc=hannes@cmpxchg.org \
--cc=kernel-team@fb.com \
--cc=linux-rdma@vger.kernel.org \
--cc=mkoutny@suse.com \
--cc=netdev@vger.kernel.org \
--cc=peterz@infradead.org \
--cc=rds-devel@oss.oracle.com \
--cc=roman.gushchin@linux.dev \
--cc=surenb@google.com \
--cc=tj@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;
as well as URLs for NNTP newsgroup(s).