From: Reinette Chatre <reinette.chatre@intel.com>
To: Ben Horgan <ben.horgan@arm.com>
Cc: <Dave.Martin@arm.com>, <babu.moger@amd.com>, <bp@alien8.de>,
<dave.hansen@linux.intel.com>, <fenghuay@nvidia.com>,
<fustini@kernel.org>, <james.morse@arm.com>,
<linux-kernel@vger.kernel.org>, <peternewman@google.com>,
<tglx@linutronix.de>, <tony.luck@intel.com>, <x86@kernel.org>,
<yu.c.chen@intel.com>
Subject: Re: [RFC PATCH 1/3] fs/resctrl: Add schema_name to struct resctrl_ctrl
Date: Wed, 18 Nov 2026 22:12:40 -0800 [thread overview]
Message-ID: <4aea42b8-991d-4f7d-832b-1c92d87bcda6@intel.com> (raw)
In-Reply-To: <20260921135702.2874931-2-ben.horgan@arm.com>
Hi Ben,
On 9/21/26 6:57 AM, Ben Horgan wrote:
> In order to properly align the schema in the schemata and size file the
> maximum of length of all the schema names needs to be
> known. resctrl_resource_ctrl_max_len() makes this calculation but stops
> short of actually constructing and storing the schema names. As the schema
> name is based on the resource name and the schema name with some special
> cases for legacy controls this leads to a lot of conditional formatting
> when displaying the schema name.
The schema name is based on the "final" resource name. This becomes significant
when CDP is enabled since there may then be two "final" resources,
one per CDP type, but they are backed by the same underlying resource.
The two "final" (CODE/DATA) resources thus share the same controls.
> Construct the schema name and save it to struct resctrl when mounting
> resctrl. This makes displaying the schema name easy whilst control of the
> naming stays within resctrl.
>
> Signed-off-by: Ben Horgan <ben.horgan@arm.com>
> ---
...
> +
> static int final_resources_list_add(struct rdt_resource *r, enum resctrl_conf_type type)
> {
> + struct resctrl_ctrl *ctrl, *em_ctrl;
> struct rdt_resource_final *f;
> const char *suffix = "";
> - int ret, cl;
> + int ret;
>
> f = kzalloc_obj(*f);
> if (!f)
> @@ -3409,24 +3420,32 @@ static int final_resources_list_add(struct rdt_resource *r, enum resctrl_conf_ty
> return -EINVAL;
> }
>
> - cl = strlen(f->name);
> -
> - /*
> - * Maintain tabular format by taking into account the names of all
> - * the resource's controls.
> - */
> - cl += resctrl_resource_ctrl_max_len(f->res);
> -
> - /*
> - * If CDP is supported by this resource, but not enabled,
> - * include the suffix. This ensures the tabular format of the
> - * schemata file does not change between mounts of the filesystem.
> - */
> - if (r->cdp_capable && !resctrl_arch_get_cdp_enabled(r))
> - cl += 4;
> + for_each_resource_ctrl(ctrl, r) {
> + if (resctrl_ctrl_is_default(ctrl)) {
> + ret = snprintf(ctrl->schema_name, sizeof(ctrl->schema_name),
> + "%s", f->name);
> + } else {
> + ret = snprintf(ctrl->schema_name, sizeof(ctrl->schema_name),
> + "%s_%s", f->name, resctrl_ctrl_name_str(ctrl->name));
> + }
> + if (ret >= sizeof(ctrl->schema_name)) {
> + kfree(f);
> + return -EINVAL;
> + }
> + update_max_name_width(r, ctrl->schema_name);
When CDP is enabled final_resources_list_add() is called once per CDP type. Since the CDP types
have the same underlying resource the two calls iterate over the same control objects with
the second iteration overwriting the resctrl_ctrl::schema_name generated during the first iteration.
I am reworking this area to address the conclusions about control naming reached during RFC v2
discussion. These changes that you have not seen yet make it difficult to discuss new ways to
make the schema name easier to display. I am working on addressing this issue as part of v3. We
can revisit then.
Reinette
next prev parent reply other threads:[~2026-11-19 6:45 UTC|newest]
Thread overview: 56+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-08-04 5:26 [RFC v2] arm,x86,fs/resctrl: Generic schema description Proof of Concept Reinette Chatre
2026-08-05 16:59 ` Ben Horgan
2026-08-05 23:57 ` Reinette Chatre
2026-08-06 17:04 ` Luck, Tony
2026-08-06 17:11 ` Reinette Chatre
2026-08-07 14:14 ` Moger, Babu
2026-08-07 15:35 ` Reinette Chatre
2026-08-07 18:33 ` Moger, Babu
2026-08-07 22:53 ` Reinette Chatre
2026-08-10 18:52 ` Babu Moger
2026-08-11 4:06 ` Reinette Chatre
2026-08-11 14:59 ` Babu Moger
2026-08-11 15:30 ` Reinette Chatre
2026-08-11 16:09 ` Babu Moger
2026-08-24 14:40 ` Ben Horgan
2026-08-24 18:07 ` Reinette Chatre
2026-08-25 15:20 ` Ben Horgan
2026-08-07 15:35 ` Chen, Yu C
2026-08-07 17:57 ` Moger, Babu
2026-08-10 2:05 ` Chen, Yu C
2026-08-10 15:53 ` Reinette Chatre
2026-08-10 17:23 ` Chen, Yu C
2026-08-10 18:09 ` Reinette Chatre
2026-08-11 3:49 ` Chen, Yu C
2026-08-13 13:59 ` Chen Yu
2026-08-13 15:19 ` Reinette Chatre
2026-08-18 16:02 ` Chen Yu
2026-08-18 17:10 ` Reinette Chatre
2026-09-14 2:49 ` Chen Yu
2026-09-18 23:11 ` Reinette Chatre
2026-09-21 5:13 ` Chen, Yu C
2026-09-21 15:01 ` Reinette Chatre
2026-09-22 3:19 ` Chen, Yu C
2026-09-21 16:21 ` Luck, Tony
2026-08-10 2:55 ` Fenghua Yu
2026-08-10 3:02 ` Chen, Yu C
2026-08-10 3:25 ` Fenghua Yu
2026-08-10 5:24 ` Chen, Yu C
2026-08-10 15:53 ` Reinette Chatre
2026-08-17 15:09 ` Reinette Chatre
2026-09-11 10:09 ` Ben Horgan
2026-09-18 22:47 ` Reinette Chatre
2026-09-21 10:22 ` Ben Horgan
2026-09-21 16:06 ` Reinette Chatre
2026-09-21 16:53 ` Ben Horgan
2026-09-23 15:28 ` Reinette Chatre
2026-09-24 14:38 ` Ben Horgan
2026-09-21 13:56 ` [RFC PATCH 0/3] Suggestions for cache scalar controls Ben Horgan
2026-09-21 13:57 ` [RFC PATCH 1/3] fs/resctrl: Add schema_name to struct resctrl_ctrl Ben Horgan
2026-09-22 20:15 ` Drew Fustini
2026-09-23 9:05 ` Ben Horgan
2026-11-19 6:12 ` Reinette Chatre [this message]
2026-09-21 13:57 ` [RFC PATCH 2/3] fs/resctrl: Use correct schema name in bw_validate() last_cmd_status messages Ben Horgan
2026-09-21 13:57 ` [RFC PATCH 3/3] fs/resctrl: display size for scalar cache controls Ben Horgan
2026-09-21 14:08 ` [RFC PATCH 0/3] Suggestions for cache scalar controls Ben Horgan
2026-09-22 20:26 ` Drew Fustini
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=4aea42b8-991d-4f7d-832b-1c92d87bcda6@intel.com \
--to=reinette.chatre@intel.com \
--cc=Dave.Martin@arm.com \
--cc=babu.moger@amd.com \
--cc=ben.horgan@arm.com \
--cc=bp@alien8.de \
--cc=dave.hansen@linux.intel.com \
--cc=fenghuay@nvidia.com \
--cc=fustini@kernel.org \
--cc=james.morse@arm.com \
--cc=linux-kernel@vger.kernel.org \
--cc=peternewman@google.com \
--cc=tglx@linutronix.de \
--cc=tony.luck@intel.com \
--cc=x86@kernel.org \
--cc=yu.c.chen@intel.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
all inboxes | Powered by JetHome®