mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: "Luck, Tony" <tony.luck@intel.com>
To: Reinette Chatre <reinette.chatre@intel.com>
Cc: 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>,
	Anil Keshavamurthy <anil.s.keshavamurthy@intel.com>,
	Chen Yu <yu.c.chen@intel.com>,
	x86@kernel.org, linux-kernel@vger.kernel.org,
	patches@lists.linux.dev
Subject: Re: [PATCH v6 00/30] x86,fs/resctrl telemetry monitoring
Date: Tue, 8 Jul 2025 12:08:24 -0700	[thread overview]
Message-ID: <aG1sqKBJSfHydDsx@agluck-desk3> (raw)
In-Reply-To: <aGa8Pg9pSCJ3XjtY@agluck-desk3>

On Thu, Jul 03, 2025 at 10:22:06AM -0700, Luck, Tony wrote:
> On Thu, Jul 03, 2025 at 09:45:15AM -0700, Reinette Chatre wrote:
> > Hi Tony and Dave,
> > 
> > On 6/26/25 9:49 AM, Tony Luck wrote:
> > >  --- 14 ---
> > > Add mon_evt::is_floating_point set by resctrl file system code to limit
> > > which events architecture code can request be displayed in floating point.
> > > 
> > > Simplified the fixed-point to floating point algorithm. Reinette is
> > > correct that the additional "lshift" and "rshift" operations are not
> > > required. All that is needed is to multiply the fixed point fractional
> > > part by 10**decimal_places, add a rounding amount equivalent to a "1"
> > > in the binary place after those supplied. Finally divide by 2**binary_places
> > > (with a right shift).
> > > 
> > > Explained in commit comment how I chose the number of decimal places to
> > > use for each binary places value.
> > > 
> > > N.B. Dave Martin expressed an opinion that the kernel should not do
> > > this conversion. Instead it should enumerate the scaling factor for
> > > each event where hardware reported a fixed point value. This patch
> > > could be dropped and replaced with one to enumerate scaling factors
> > > per event if others agree with Dave.
> > 
> > Could resctrl accommodate both usages? For example, it does not
> > look too invasive to add a second file <mon_evt::name>.raw for the
> > mon_evt::is_floating_point events that can output something like Dave
> > suggested in [1]:
> > 
> > .raw file format could be:
> > 	#format:<output that depends on format>
> > 	#fixed-point:<value>/<scaling factor>
> > 
> > Example output:
> > 	fixed-point:0x60000/0x40000
> 
> Dave: Is that what you want in the ".raw" file? An alternative would be
> to put the format information for non-integer events into an
> "info" file ("info/{RESOURCE_NAME}_MON/monfeatures.raw.formats"?)
> and just put the raw value into the ".raw" file under mon_data.

Note that I thought it easier for users to keep the raw file to just
showing a value, rather than including the formatting details in
Reinette's proposal.

Patch to implement my alternative suggestion below. To the user things
look like this:

$ cd /sys/fs/resctrl/mon_data/mon_PERF_PKG_01
$ cat core_energy
0.02203
$ cat core_energy.raw
5775
$ cat /sys/fs/resctrl/info/PERF_PKG_MON/mon_features_raw_scale
core_energy 262144
activity 262144
$ bc -ql
5775 / 262144
.02202987670898437500

If this seems useful I can write up a commit message and include
as its own patch in v7. Suggestions for better names?

-Tony

---

diff --git a/fs/resctrl/internal.h b/fs/resctrl/internal.h
index 4704ea7228ca..5ac4e3c98f23 100644
--- a/fs/resctrl/internal.h
+++ b/fs/resctrl/internal.h
@@ -90,6 +90,8 @@ extern struct mon_evt mon_event_all[QOS_NUM_EVENTS];
  *                   the event file belongs. When @sum is one this
  *                   is the id of the L3 cache that all domains to be
  *                   summed share.
+ * @raw:             Set for ".raw" files that directly show hardware
+ *                   provided counts with no interpretation.
  *
  * Pointed to by the kernfs kn->priv field of monitoring event files.
  * Readers and writers must hold rdtgroup_mutex.
@@ -100,6 +102,7 @@ struct mon_data {
 	struct mon_evt		*evt;
 	int			domid;
 	bool			sum;
+	bool			raw;
 };
 
 /**
diff --git a/fs/resctrl/ctrlmondata.c b/fs/resctrl/ctrlmondata.c
index 29de0e380ccc..78e7af296d5a 100644
--- a/fs/resctrl/ctrlmondata.c
+++ b/fs/resctrl/ctrlmondata.c
@@ -753,7 +753,7 @@ int rdtgroup_mondata_show(struct seq_file *m, void *arg)
 		seq_puts(m, "Error\n");
 	else if (rr.err == -EINVAL)
 		seq_puts(m, "Unavailable\n");
-	else if (evt->binary_bits == 0)
+	else if (md->raw || evt->binary_bits == 0)
 		seq_printf(m, "%llu\n", rr.val);
 	else
 		print_event_value(m, evt->binary_bits, rr.val);
diff --git a/fs/resctrl/rdtgroup.c b/fs/resctrl/rdtgroup.c
index 511362a67532..97786831722a 100644
--- a/fs/resctrl/rdtgroup.c
+++ b/fs/resctrl/rdtgroup.c
@@ -1158,6 +1158,21 @@ static int rdt_mon_features_show(struct kernfs_open_file *of,
 	return 0;
 }
 
+static int rdt_mon_features_raw_scale_show(struct kernfs_open_file *of,
+					   struct seq_file *seq, void *v)
+{
+	struct rdt_resource *r = rdt_kn_parent_priv(of->kn);
+	struct mon_evt *mevt;
+
+	for_each_mon_event(mevt) {
+		if (mevt->rid != r->rid || !mevt->enabled || !mevt->binary_bits)
+			continue;
+		seq_printf(seq, "%s %u\n", mevt->name, 1 << mevt->binary_bits);
+	}
+
+	return 0;
+}
+
 static int rdt_bw_gran_show(struct kernfs_open_file *of,
 			    struct seq_file *seq, void *v)
 {
@@ -1823,6 +1838,13 @@ static struct rftype res_common_files[] = {
 		.seq_show	= rdt_mon_features_show,
 		.fflags		= RFTYPE_MON_INFO,
 	},
+	{
+		.name		= "mon_features_raw_scale",
+		.mode		= 0444,
+		.kf_ops		= &rdtgroup_kf_single_ops,
+		.seq_show	= rdt_mon_features_raw_scale_show,
+		.fflags		= RFTYPE_MON_INFO,
+	},
 	{
 		.name		= "num_rmids",
 		.mode		= 0444,
@@ -2905,7 +2927,7 @@ static void rmdir_all_sub(void)
  */
 static struct mon_data *mon_get_kn_priv(enum resctrl_res_level rid, int domid,
 					struct mon_evt *mevt,
-					bool do_sum)
+					bool do_sum, bool rawfile)
 {
 	struct mon_data *priv;
 
@@ -2916,7 +2938,8 @@ static struct mon_data *mon_get_kn_priv(enum resctrl_res_level rid, int domid,
 
 	list_for_each_entry(priv, &mon_data_kn_priv_list, list) {
 		if (priv->rid == rid && priv->domid == domid &&
-		    priv->sum == do_sum && priv->evt == mevt)
+		    priv->sum == do_sum && priv->evt == mevt &&
+		    priv->raw == rawfile)
 			return priv;
 	}
 
@@ -2928,6 +2951,7 @@ static struct mon_data *mon_get_kn_priv(enum resctrl_res_level rid, int domid,
 	priv->domid = domid;
 	priv->sum = do_sum;
 	priv->evt = mevt;
+	priv->raw = rawfile;
 	list_add_tail(&priv->list, &mon_data_kn_priv_list);
 
 	return priv;
@@ -3078,12 +3102,13 @@ static int mon_add_all_files(struct kernfs_node *kn, struct rdt_domain_hdr *hdr,
 	struct rmid_read rr = {0};
 	struct mon_data *priv;
 	struct mon_evt *mevt;
+	char rawname[64];
 	int ret;
 
 	for_each_mon_event(mevt) {
 		if (mevt->rid != r->rid || !mevt->enabled)
 			continue;
-		priv = mon_get_kn_priv(r->rid, domid, mevt, do_sum);
+		priv = mon_get_kn_priv(r->rid, domid, mevt, do_sum, false);
 		if (WARN_ON_ONCE(!priv))
 			return -EINVAL;
 
@@ -3093,6 +3118,18 @@ static int mon_add_all_files(struct kernfs_node *kn, struct rdt_domain_hdr *hdr,
 
 		if (r->rid == RDT_RESOURCE_L3 && !do_sum && resctrl_is_mbm_event(mevt->evtid))
 			mon_event_read(&rr, r, hdr, prgrp, &hdr->cpu_mask, mevt, true);
+
+		if (!mevt->binary_bits)
+			continue;
+
+		sprintf(rawname, "%s.raw", mevt->name);
+		priv = mon_get_kn_priv(r->rid, domid, mevt, do_sum, true);
+		if (WARN_ON_ONCE(!priv))
+			return -EINVAL;
+
+		ret = mon_addfile(kn, rawname, priv);
+		if (ret)
+			return ret;
 	}
 
 	return 0;

  reply	other threads:[~2025-07-08 19:08 UTC|newest]

Thread overview: 89+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-06-26 16:49 Tony Luck
2025-06-26 16:49 ` [PATCH v6 01/30] x86,fs/resctrl: Consolidate monitor event descriptions Tony Luck
2025-06-27 21:55   ` Fenghua Yu
2025-07-08 20:52   ` Reinette Chatre
2025-06-26 16:49 ` [PATCH v6 02/30] x86,fs/resctrl: Replace architecture event enabled checks Tony Luck
2025-06-27 22:15   ` Fenghua Yu
2025-07-08 20:52   ` Reinette Chatre
2025-06-26 16:49 ` [PATCH v6 03/30] x86/resctrl: Remove 'rdt_mon_features' global variable Tony Luck
2025-07-08 20:53   ` Reinette Chatre
2025-06-26 16:49 ` [PATCH v6 04/30] x86,fs/resctrl: Prepare for more monitor events Tony Luck
2025-07-08 20:55   ` Reinette Chatre
2025-06-26 16:49 ` [PATCH v6 05/30] x86,fs/resctrl: Improve domain type checking Tony Luck
2025-07-08 21:01   ` Reinette Chatre
2025-06-26 16:49 ` [PATCH v6 06/30] x86/resctrl: Move L3 initialization out of domain_add_cpu_mon() Tony Luck
2025-07-08 20:56   ` Reinette Chatre
2025-06-26 16:49 ` [PATCH v6 07/30] x86,fs/resctrl: Refactor domain_remove_cpu_mon() ready for new domain types Tony Luck
2025-07-08 20:57   ` Reinette Chatre
2025-06-26 16:49 ` [PATCH v6 08/30] x86/resctrl: Clean up domain_remove_cpu_ctrl() Tony Luck
2025-06-26 16:49 ` [PATCH v6 09/30] x86,fs/resctrl: Use struct rdt_domain_hdr instead of struct rdt_mon_domain Tony Luck
2025-07-08 21:04   ` Reinette Chatre
2025-06-26 16:49 ` [PATCH v6 10/30] x86,fs/resctrl: Rename struct rdt_mon_domain and rdt_hw_mon_domain Tony Luck
2025-07-08 21:06   ` Reinette Chatre
2025-06-26 16:49 ` [PATCH v6 11/30] x86,fs/resctrl: Rename some L3 specific functions Tony Luck
2025-07-08 21:08   ` Reinette Chatre
2025-06-26 16:49 ` [PATCH v6 12/30] fs/resctrl: Make event details accessible to functions when reading events Tony Luck
2025-07-09 22:12   ` Reinette Chatre
2025-06-26 16:49 ` [PATCH v6 13/30] x86,fs/resctrl: Handle events that can be read from any CPU Tony Luck
2025-07-08 21:15   ` Reinette Chatre
2025-06-26 16:49 ` [PATCH v6 14/30] x86,fs/resctrl: Support binary fixed point event counters Tony Luck
2025-06-27 21:22   ` Fenghua Yu
2025-06-27 22:28     ` Luck, Tony
2025-06-27 21:49   ` Fenghua Yu
2025-07-08 21:46   ` Reinette Chatre
2025-07-09 16:52     ` Luck, Tony
2025-06-26 16:49 ` [PATCH v6 15/30] x86,fs/resctrl: Add an architectural hook called for each mount Tony Luck
2025-06-26 16:49 ` [PATCH v6 16/30] x86,fs/resctrl: Add and initialize rdt_resource for package scope core monitor Tony Luck
2025-07-08 22:05   ` Reinette Chatre
2025-06-26 16:49 ` [PATCH v6 17/30] x86/resctrl: Discover hardware telemetry events Tony Luck
2025-06-27 18:06   ` Luck, Tony
2025-07-03 18:27     ` Reinette Chatre
2025-07-03 20:17       ` Luck, Tony
2025-07-03 20:31         ` Reinette Chatre
2025-07-03 21:11           ` Luck, Tony
2025-07-03 22:00             ` Reinette Chatre
2025-07-03 23:29               ` Luck, Tony
2025-07-08 23:51   ` Reinette Chatre
2025-06-26 16:49 ` [PATCH v6 18/30] x86/resctrl: Count valid telemetry aggregators per package Tony Luck
2025-07-09  2:20   ` Reinette Chatre
2025-07-09 18:12     ` Luck, Tony
2025-07-09 22:13       ` Reinette Chatre
2025-07-09 22:48         ` Luck, Tony
2025-07-09 22:59           ` Reinette Chatre
2025-06-26 16:49 ` [PATCH v6 19/30] x86/resctrl: Complete telemetry event enumeration Tony Luck
2025-07-09  2:38   ` Reinette Chatre
2025-06-26 16:49 ` [PATCH v6 20/30] x86,fs/resctrl: Fill in details of Clearwater Forest events Tony Luck
2025-07-09  3:00   ` Reinette Chatre
2025-06-26 16:49 ` [PATCH v6 21/30] x86,fs/resctrl: Add architectural event pointer Tony Luck
2025-07-09  3:21   ` Reinette Chatre
2025-07-09 21:16     ` Reinette Chatre
2025-06-26 16:49 ` [PATCH v6 22/30] x86/resctrl: Read core telemetry events Tony Luck
2025-07-09 15:48   ` Reinette Chatre
2025-07-09 21:57     ` Luck, Tony
2025-07-09 22:13       ` Reinette Chatre
2025-06-26 16:49 ` [PATCH v6 23/30] x86/resctrl: Handle domain creation/deletion for RDT_RESOURCE_PERF_PKG Tony Luck
2025-07-09 22:13   ` Reinette Chatre
2025-06-26 16:49 ` [PATCH v6 24/30] x86/resctrl: Add energy/perf choices to rdt boot option Tony Luck
2025-07-09 22:14   ` Reinette Chatre
2025-06-26 16:49 ` [PATCH v6 25/30] x86/resctrl: Handle number of RMIDs supported by telemetry resources Tony Luck
2025-07-09 22:17   ` Reinette Chatre
2025-06-26 16:49 ` [PATCH v6 26/30] x86,fs/resctrl: Move RMID initialization to first mount Tony Luck
2025-07-09 22:18   ` Reinette Chatre
2025-06-26 16:49 ` [PATCH v6 27/30] x86/resctrl: Enable RDT_RESOURCE_PERF_PKG Tony Luck
2025-06-26 16:49 ` [PATCH v6 28/30] fs/resctrl: Provide interface to create a debugfs info directory Tony Luck
2025-07-09 22:19   ` Reinette Chatre
2025-06-26 16:49 ` [PATCH v6 29/30] x86/resctrl: Add debug info/PERF_PKG_MON/status files Tony Luck
2025-07-09 22:22   ` Reinette Chatre
2025-06-26 16:49 ` [PATCH v6 30/30] x86,fs/resctrl: Update Documentation for package events Tony Luck
2025-07-09 22:24   ` Reinette Chatre
2025-06-27  0:26 ` [PATCH v6 00/30] x86,fs/resctrl telemetry monitoring Luck, Tony
2025-06-27 18:09   ` Luck, Tony
2025-06-30 17:51 ` Reinette Chatre
2025-06-30 22:46   ` Luck, Tony
2025-07-08 20:50     ` Reinette Chatre
2025-07-03 16:45 ` Reinette Chatre
2025-07-03 17:22   ` Luck, Tony
2025-07-08 19:08     ` Luck, Tony [this message]
2025-07-08 20:49       ` Reinette Chatre
2025-07-08 22:43         ` Luck, Tony
2025-07-08 23:26           ` Reinette Chatre

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=aG1sqKBJSfHydDsx@agluck-desk3 \
    --to=tony.luck@intel.com \
    --cc=Dave.Martin@arm.com \
    --cc=anil.s.keshavamurthy@intel.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=reinette.chatre@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®