From: Reinette Chatre <reinette.chatre@intel.com>
To: Tony Luck <tony.luck@intel.com>, Fenghua Yu <fenghuay@nvidia.com>,
"Maciej Wieczor-Retman" <maciej.wieczor-retman@intel.com>,
Peter Newman <peternewman@google.com>,
James Morse <james.morse@arm.com>,
Babu Moger <babu.moger@amd.com>,
Drew Fustini <dfustini@baylibre.com>,
Dave Martin <Dave.Martin@arm.com>, Chen Yu <yu.c.chen@intel.com>
Cc: <x86@kernel.org>, <linux-kernel@vger.kernel.org>,
<patches@lists.linux.dev>
Subject: Re: [PATCH v7 19/31] x86/resctrl: Complete telemetry event enumeration
Date: Fri, 25 Jul 2025 16:41:19 -0700 [thread overview]
Message-ID: <765d3f16-6ecd-4581-940f-d062d3340c14@intel.com> (raw)
In-Reply-To: <20250711235341.113933-20-tony.luck@intel.com>
Hi Tony,
On 7/11/25 4:53 PM, Tony Luck wrote:
> Counters for telemetry events are in MMIO space. Each telemetry_region
> structure returned in the pmt_feature_group returned from OOBMSM contains
> the base MMIO address for the counters.
>
> There may be multiple aggregators per package. Scan all the
> telemetry_region structures again and save the number of regions together
> with a flex array of the MMIO addresses for each aggregator indexed by
> package id.
I do not see why it is needed to switch back and forth between interchangeable
regions and aggregators. Why not just stick with telemetry regions? It is
confusing when, for example, saying "number of regions" followed by "for each
aggregator"? Why not just say "number of regions" followed by "for each
region"?
>
> Completed structure for each event group looks like this:
>
> +---------------------+---------------------+
> pkginfo** -->|pkginfo[package ID 0]|pkginfo[package ID 1]|
> +---------------------+---------------------+
> | |
> v v
> +----------------+ +----------------+
> |struct mmio_info| |struct mmio_info|
mmio_info -> pkg_mmio_info
> +----------------+ +----------------+
> |num_regions = N | |num_regions = N |
The above "There may be multiple aggregators (telemetry regions?) per
package." could add that the number of telemetry regions per package may
be different and supported by an example where one package has "N"
regions and the other "M".
> | addrs[0] | | addrs[0] |
> | addrs[1] | | addrs[1] |
> | ... | | ... |
> | addrs[N-1] | | addrs[N-1] |
> +----------------+ +----------------+
>
> Signed-off-by: Tony Luck <tony.luck@intel.com>
> ---
> arch/x86/kernel/cpu/resctrl/intel_aet.c | 64 ++++++++++++++++++++++++-
> 1 file changed, 63 insertions(+), 1 deletion(-)
>
> diff --git a/arch/x86/kernel/cpu/resctrl/intel_aet.c b/arch/x86/kernel/cpu/resctrl/intel_aet.c
> index 7cd6c06f9205..3f383f0a9d08 100644
> --- a/arch/x86/kernel/cpu/resctrl/intel_aet.c
> +++ b/arch/x86/kernel/cpu/resctrl/intel_aet.c
> @@ -19,12 +19,26 @@
>
> #include "internal.h"
>
> +/**
> + * struct pkg_mmio_info - MMIO address information for one event group of a package.
> + * @num_regions: Number of telemetry regions on this package.
> + * @addrs: Array of MMIO addresses, one per telemetry region on this package.
> + *
> + * Provides convenient access to all MMIO addresses of one event group
> + * for one package. Used when reading event data on a package.
> + */
> +struct pkg_mmio_info {
> + unsigned int num_regions;
> + void __iomem *addrs[] __counted_by(num_regions);
> +};
> +
> /**
> * struct event_group - All information about a group of telemetry events.
> * @pfg: Points to the aggregated telemetry space information
> * within the OOBMSM driver that contains data for all
> * telemetry regions.
> * @list: List of active event groups.
> + * @pkginfo: Per-package MMIO addresses of telemetry regions belonging to this group.
> * @guid: Unique number per XML description file.
> * @mmio_size: Number of bytes of MMIO registers for this group.
> */
> @@ -32,6 +46,7 @@ struct event_group {
> /* Data fields for additional structures to manage this group. */
> struct pmt_feature_group *pfg;
> struct list_head list;
> + struct pkg_mmio_info **pkginfo;
>
> /* Remaining fields initialized from XML file. */
> u32 guid;
> @@ -90,15 +105,32 @@ static bool skip_this_region(struct telemetry_region *tr, struct event_group *e)
> return false;
> }
>
> +static void free_pkg_mmio_info(struct pkg_mmio_info **mmi)
> +{
> + int num_pkgs = topology_max_packages();
> +
> + if (!mmi)
> + return;
> +
> + for (int i = 0; i < num_pkgs; i++)
> + kfree(mmi[i]);
> + kfree(mmi);
> +}
> +
> +DEFINE_FREE(pkg_mmio_info, struct pkg_mmio_info **, free_pkg_mmio_info(_T))
> +
> /*
> * Discover events from one pmt_feature_group.
> * 1) Count how many usable telemetry regions per package.
> - * 2...) To be continued.
> + * 2) Allocate per-package structures and populate with MMIO
> + * addresses of the telemetry regions used by each aggregator.
"the telemetry regions used by each aggregator" does not sound right. "telemetry region == aggregator", no?
> */
> static int discover_events(struct event_group *e, struct pmt_feature_group *p)
> {
> + struct pkg_mmio_info **pkginfo __free(pkg_mmio_info) = NULL;
> int *pkgcounts __free(kfree) = NULL;
> struct telemetry_region *tr;
> + struct pkg_mmio_info *mmi;
> int num_pkgs;
>
> num_pkgs = topology_max_packages();
> @@ -108,6 +140,7 @@ static int discover_events(struct event_group *e, struct pmt_feature_group *p)
> tr = &p->regions[i];
> if (skip_this_region(tr, e))
> continue;
> +
> if (!pkgcounts) {
> pkgcounts = kcalloc(num_pkgs, sizeof(*pkgcounts), GFP_KERNEL);
> if (!pkgcounts)
squash with previous patch.
> @@ -119,6 +152,32 @@ static int discover_events(struct event_group *e, struct pmt_feature_group *p)
> if (!pkgcounts)
> return -ENODEV;
>
> + /* Allocate array for per-package struct pkg_mmio_info data */
> + pkginfo = kcalloc(num_pkgs, sizeof(*pkginfo), GFP_KERNEL);
> + if (!pkginfo)
> + return -ENOMEM;
> +
> + /*
> + * Allocate per-package pkg_mmio_info structures and initialize
> + * count of telemetry_regions in each one.
> + */
> + for (int i = 0; i < num_pkgs; i++) {
> + pkginfo[i] = kzalloc(struct_size(pkginfo[i], addrs, pkgcounts[i]), GFP_KERNEL);
> + if (!pkginfo[i])
> + return -ENOMEM;
> + pkginfo[i]->num_regions = pkgcounts[i];
> + }
> +
> + /* Save MMIO address(es) for each telemetry region in per-package structures */
> + for (int i = 0; i < p->count; i++) {
> + tr = &p->regions[i];
> + if (skip_this_region(tr, e))
> + continue;
> + mmi = pkginfo[tr->plat_info.package_id];
> + mmi->addrs[--pkgcounts[tr->plat_info.package_id]] = tr->addr;
> + }
> + e->pkginfo = no_free_ptr(pkginfo);
> +
> return 0;
> }
>
> @@ -151,6 +210,7 @@ static bool get_pmt_feature(enum pmt_feature_id feature, struct event_group **ev
> (*peg)->pfg = no_free_ptr(p);
> return true;
> }
> + free_pkg_mmio_info((*peg)->pkginfo);
Is this necessary? pkginfo will only be set on success, no?
> }
>
> return false;
> @@ -179,6 +239,8 @@ void __exit intel_aet_exit(void)
> list_for_each_entry_safe(evg, tmp, &active_event_groups, list) {
> intel_pmt_put_feature_group(evg->pfg);
> evg->pfg = NULL;
> + free_pkg_mmio_info(evg->pkginfo);
> + evg->pkginfo = NULL;
> list_del(&evg->list);
> }
> }
Reinette
next prev parent reply other threads:[~2025-07-25 23:41 UTC|newest]
Thread overview: 61+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-07-11 23:53 [PATCH v7 00/31] x86,fs/resctrl telemetry monitoring Tony Luck
2025-07-11 23:53 ` [PATCH v7 01/31] x86,fs/resctrl: Consolidate monitor event descriptions Tony Luck
2025-07-17 17:51 ` Reinette Chatre
2025-07-11 23:53 ` [PATCH v7 02/31] x86,fs/resctrl: Replace architecture event enabled checks Tony Luck
2025-07-11 23:53 ` [PATCH v7 03/31] x86/resctrl: Remove 'rdt_mon_features' global variable Tony Luck
2025-07-11 23:53 ` [PATCH v7 04/31] x86,fs/resctrl: Prepare for more monitor events Tony Luck
2025-07-11 23:53 ` [PATCH v7 05/31] x86,fs/resctrl: Improve domain type checking Tony Luck
2025-07-25 23:17 ` Reinette Chatre
2025-07-11 23:53 ` [PATCH v7 06/31] x86/resctrl: Move L3 initialization into new helper function Tony Luck
2025-07-25 23:21 ` Reinette Chatre
2025-07-11 23:53 ` [PATCH v7 07/31] x86,fs/resctrl: Refactor domain_remove_cpu_mon() ready for new domain types Tony Luck
2025-07-25 23:29 ` Reinette Chatre
2025-07-11 23:53 ` [PATCH v7 08/31] x86/resctrl: Clean up domain_remove_cpu_ctrl() Tony Luck
2025-07-25 23:22 ` Reinette Chatre
2025-07-11 23:53 ` [PATCH v7 09/31] x86,fs/resctrl: Use struct rdt_domain_hdr instead of struct rdt_mon_domain Tony Luck
2025-07-25 23:25 ` Reinette Chatre
2025-07-11 23:53 ` [PATCH v7 10/31] x86,fs/resctrl: Rename struct rdt_mon_domain and rdt_hw_mon_domain Tony Luck
2025-07-25 23:26 ` Reinette Chatre
2025-07-11 23:53 ` [PATCH v7 11/31] x86,fs/resctrl: Rename some L3 specific functions Tony Luck
2025-07-25 23:26 ` Reinette Chatre
2025-07-11 23:53 ` [PATCH v7 12/31] fs/resctrl: Make event details accessible to functions when reading events Tony Luck
2025-07-25 23:27 ` Reinette Chatre
2025-07-11 23:53 ` [PATCH v7 13/31] x86,fs/resctrl: Handle events that can be read from any CPU Tony Luck
2025-07-25 23:32 ` Reinette Chatre
2025-07-11 23:53 ` [PATCH v7 14/31] x86,fs/resctrl: Support binary fixed point event counters Tony Luck
2025-07-25 23:34 ` Reinette Chatre
2025-07-11 23:53 ` [PATCH v7 15/31] x86,fs/resctrl: Add an architectural hook called for each mount Tony Luck
2025-07-25 23:35 ` Reinette Chatre
2025-07-11 23:53 ` [PATCH v7 16/31] x86,fs/resctrl: Add and initialize rdt_resource for package scope core monitor Tony Luck
2025-07-25 23:36 ` Reinette Chatre
2025-07-11 23:53 ` [PATCH v7 17/31] x86/resctrl: Discover hardware telemetry events Tony Luck
2025-07-25 23:39 ` Reinette Chatre
2025-07-11 23:53 ` [PATCH v7 18/31] x86/resctrl: Count valid telemetry aggregators per package Tony Luck
2025-07-25 23:40 ` Reinette Chatre
2025-07-11 23:53 ` [PATCH v7 19/31] x86/resctrl: Complete telemetry event enumeration Tony Luck
2025-07-25 23:41 ` Reinette Chatre [this message]
2025-07-11 23:53 ` [PATCH v7 20/31] x86,fs/resctrl: Fill in details of events for guid 0x26696143 and 0x26557651 Tony Luck
2025-07-25 23:43 ` Reinette Chatre
2025-07-11 23:53 ` [PATCH v7 21/31] x86,fs/resctrl: Add architectural event pointer Tony Luck
2025-07-25 23:43 ` Reinette Chatre
2025-07-11 23:53 ` [PATCH v7 22/31] x86/resctrl: Read telemetry events Tony Luck
2025-07-25 23:45 ` Reinette Chatre
2025-07-11 23:53 ` [PATCH v7 23/31] x86/resctrl: Handle domain creation/deletion for RDT_RESOURCE_PERF_PKG Tony Luck
2025-07-25 23:46 ` Reinette Chatre
2025-07-11 23:53 ` [PATCH v7 24/31] x86/resctrl: Add energy/perf choices to rdt boot option Tony Luck
2025-07-25 23:46 ` Reinette Chatre
2025-07-11 23:53 ` [PATCH v7 25/31] x86/resctrl: Handle number of RMIDs supported by telemetry resources Tony Luck
2025-07-25 23:49 ` Reinette Chatre
2025-07-11 23:53 ` [PATCH v7 26/31] fs/resctrl: Fix life-cycle of closid_num_dirty_rmid Tony Luck
2025-07-25 23:51 ` Reinette Chatre
2025-07-11 23:53 ` [PATCH v7 27/31] x86,fs/resctrl: Move RMID initialization to first mount Tony Luck
2025-07-25 23:53 ` Reinette Chatre
2025-07-11 23:53 ` [PATCH v7 28/31] x86/resctrl: Enable RDT_RESOURCE_PERF_PKG Tony Luck
2025-07-25 23:54 ` Reinette Chatre
2025-07-11 23:53 ` [PATCH v7 29/31] fs/resctrl: Provide interface to create architecture specific debugfs area Tony Luck
2025-07-25 23:55 ` Reinette Chatre
2025-07-11 23:53 ` [PATCH v7 30/31] x86/resctrl: Add debugfs files to show telemetry aggregator status Tony Luck
2025-07-11 23:53 ` [PATCH v7 31/31] x86,fs/resctrl: Update Documentation for package events Tony Luck
2025-07-30 18:42 ` [PATCH v7 00/31] x86,fs/resctrl telemetry monitoring Moger, Babu
2025-07-30 20:27 ` Reinette Chatre
2025-07-30 22:05 ` Moger, Babu
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=765d3f16-6ecd-4581-940f-d062d3340c14@intel.com \
--to=reinette.chatre@intel.com \
--cc=Dave.Martin@arm.com \
--cc=babu.moger@amd.com \
--cc=dfustini@baylibre.com \
--cc=fenghuay@nvidia.com \
--cc=james.morse@arm.com \
--cc=linux-kernel@vger.kernel.org \
--cc=maciej.wieczor-retman@intel.com \
--cc=patches@lists.linux.dev \
--cc=peternewman@google.com \
--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®