From: Tony Luck <tony.luck@intel.com>
To: tony.luck@intel.com
Cc: Dave.Martin@arm.com, babu.moger@amd.com, david.e.box@intel.com,
dfustini@baylibre.com, fenghuay@nvidia.com, hch@infradead.org,
james.morse@arm.com, linux-kernel@vger.kernel.org,
maciej.wieczor-retman@intel.com, patches@lists.linux.dev,
peternewman@google.com, reinette.chatre@intel.com,
x86@kernel.org, yu.c.chen@intel.com
Subject: [PATCH v14.1 18/25] x86/resctrl: Use registered function pointers for AET enumeration
Date: Tue, 29 Sep 2026 10:52:24 -0700 [thread overview]
Message-ID: <20260929175224.16565-1-tony.luck@intel.com> (raw)
In-Reply-To: <20260928221509.68002-19-tony.luck@intel.com>
The pmt_telemetry driver registers enumeration functions with resctrl
Application Energy Telemetry (AET) code.
Use the function pointers instead of direct function calls in preparation
for the telemetry driver to be a loadable module.
The pmt_telemetry is forced to be built-in to the kernel, so it does not matter
at this point that there is no matching module_put() for the try_module_get()
in intel_aet_pre_mount(). This must be resolved with additions in the file
system unmount path before pmt_telemetry can be configured as a module.
Signed-off-by: Tony Luck <tony.luck@intel.com>
---
v14:
Updates to resolve locking problems reported by Sashiko in v13.
AET is now going to use a single mutex for both registration
and in a later patch for device removal. So change the name
from aet_register_lock to just aet_lock.
Reverse locking hierarchy between AET and pmt_telemetry.
No longer OK to call into pmt_telemetry to enumerate
features with the aet_lock held. Instead, make use of
module pinning while holding aet_lock.
---
arch/x86/kernel/cpu/resctrl/internal.h | 4 +-
arch/x86/kernel/cpu/resctrl/core.c | 2 +-
arch/x86/kernel/cpu/resctrl/intel_aet.c | 59 ++++++++++++++++++++++---
3 files changed, 57 insertions(+), 8 deletions(-)
diff --git a/arch/x86/kernel/cpu/resctrl/internal.h b/arch/x86/kernel/cpu/resctrl/internal.h
index c038b7d80ce3..8406addc05f5 100644
--- a/arch/x86/kernel/cpu/resctrl/internal.h
+++ b/arch/x86/kernel/cpu/resctrl/internal.h
@@ -234,15 +234,15 @@ void rdt_domain_reconfigure_cdp(struct rdt_resource *r);
void resctrl_arch_mbm_cntr_assign_set_one(struct rdt_resource *r);
#ifdef CONFIG_X86_CPU_RESCTRL_INTEL_AET
-bool intel_aet_get_events(void);
void __exit intel_aet_exit(void);
+bool intel_aet_pre_mount(void);
int intel_aet_read_event(int domid, u32 rmid, void *arch_priv, u64 *val);
void intel_aet_mon_domain_setup(int cpu, int id, struct rdt_resource *r,
struct list_head *add_pos);
bool intel_handle_aet_option(bool force_off, char *tok);
#else
-static inline bool intel_aet_get_events(void) { return false; }
static inline void __exit intel_aet_exit(void) { }
+static inline bool intel_aet_pre_mount(void) { return false; }
static inline int intel_aet_read_event(int domid, u32 rmid, void *arch_priv, u64 *val)
{
return -EINVAL;
diff --git a/arch/x86/kernel/cpu/resctrl/core.c b/arch/x86/kernel/cpu/resctrl/core.c
index 0262174df7ad..ac67b4523b2d 100644
--- a/arch/x86/kernel/cpu/resctrl/core.c
+++ b/arch/x86/kernel/cpu/resctrl/core.c
@@ -795,7 +795,7 @@ void resctrl_arch_pre_mount(void)
struct rdt_resource *r = &rdt_resources_all[RDT_RESOURCE_PERF_PKG].r_resctrl;
int cpu;
- if (!intel_aet_get_events())
+ if (!intel_aet_pre_mount())
return;
/*
diff --git a/arch/x86/kernel/cpu/resctrl/intel_aet.c b/arch/x86/kernel/cpu/resctrl/intel_aet.c
index 6c4f0bf3b876..d79c459f4c54 100644
--- a/arch/x86/kernel/cpu/resctrl/intel_aet.c
+++ b/arch/x86/kernel/cpu/resctrl/intel_aet.c
@@ -12,6 +12,7 @@
#define pr_fmt(fmt) "resctrl: " fmt
#include <linux/bits.h>
+#include <linux/cleanup.h>
#include <linux/compiler_types.h>
#include <linux/container_of.h>
#include <linux/cpumask.h>
@@ -25,6 +26,7 @@
#include <linux/io.h>
#include <linux/minmax.h>
#include <linux/module.h>
+#include <linux/mutex.h>
#include <linux/printk.h>
#include <linux/rculist.h>
#include <linux/rcupdate.h>
@@ -293,6 +295,18 @@ static enum pmt_feature_id lookup_pfid(const char *pfname)
return FEATURE_INVALID;
}
+/*
+ * Serialises AET's view of pmt_telemetry:
+ * - pmt_module, get_feature, put_feature
+ * - every event_group's ->pfg
+ *
+ * Lock ordering with pmt/telemetry.c's ep_lock is ep_lock -> aet_lock.
+ * AET therefore must not call get_feature() (which takes ep_lock) while
+ * holding aet_lock. put_feature() does not take ep_lock and may be
+ * invoked with aet_lock held.
+ */
+static DEFINE_MUTEX(aet_lock);
+
static struct module *pmt_module;
static struct pmt_feature_group *(*get_feature)(enum pmt_feature_id id);
static void (*put_feature)(struct pmt_feature_group *p);
@@ -308,7 +322,8 @@ static void (*put_feature)(struct pmt_feature_group *p);
* struct pmt_feature_group to indicate that its events are successfully
* enabled.
*/
-bool intel_aet_get_events(void)
+static bool aet_get_events(struct pmt_feature_group *(*get)(enum pmt_feature_id id),
+ void (*put)(struct pmt_feature_group *p))
{
struct pmt_feature_group *p;
enum pmt_feature_id pfid;
@@ -317,14 +332,16 @@ bool intel_aet_get_events(void)
for_each_event_group(peg) {
pfid = lookup_pfid((*peg)->pfname);
- p = intel_pmt_get_regions_by_feature(pfid);
+ p = get(pfid);
if (IS_ERR_OR_NULL(p))
continue;
if (enable_events(*peg, p)) {
- (*peg)->pfg = p;
+ scoped_guard(mutex, &aet_lock) {
+ (*peg)->pfg = p;
+ }
ret = true;
} else {
- intel_pmt_put_feature_group(p);
+ put(p);
}
}
@@ -335,6 +352,7 @@ void intel_aet_register_enumeration(struct module *module,
struct pmt_feature_group *(*get)(enum pmt_feature_id id),
void (*put)(struct pmt_feature_group *p))
{
+ guard(mutex)(&aet_lock);
get_feature = get;
put_feature = put;
pmt_module = module;
@@ -343,19 +361,50 @@ EXPORT_SYMBOL_NS_GPL(intel_aet_register_enumeration, "INTEL_PMT");
void intel_aet_unregister_enumeration(void)
{
+ guard(mutex)(&aet_lock);
pmt_module = NULL;
get_feature = NULL;
put_feature = NULL;
}
EXPORT_SYMBOL_NS_GPL(intel_aet_unregister_enumeration, "INTEL_PMT");
+bool intel_aet_pre_mount(void)
+{
+ struct pmt_feature_group *(*get)(enum pmt_feature_id id);
+ void (*put)(struct pmt_feature_group *p);
+ struct module *mod;
+
+ /*
+ * Snapshot the callbacks and pin the pmt_telemetry module under
+ * aet_lock. The module reference keeps pmt_telem_exit() from
+ * running intel_aet_unregister_enumeration() and changing these
+ * pointers, so the snapshot remains valid after dropping aet_lock.
+ */
+ scoped_guard(mutex, &aet_lock) {
+ if (!get_feature || !put_feature)
+ return false;
+ if (!try_module_get(pmt_module))
+ return false;
+ mod = pmt_module;
+ get = get_feature;
+ put = put_feature;
+ }
+
+ if (!aet_get_events(get, put)) {
+ module_put(mod);
+ return false;
+ }
+
+ return true;
+}
+
void __exit intel_aet_exit(void)
{
struct event_group **peg;
for_each_event_group(peg) {
if ((*peg)->pfg) {
- intel_pmt_put_feature_group((*peg)->pfg);
+ put_feature((*peg)->pfg);
(*peg)->pfg = NULL;
}
}
--
2.55.0
next prev parent reply other threads:[~2026-09-29 17:52 UTC|newest]
Thread overview: 32+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-09-28 22:14 [PATCH v13 00/25] Allow AET to use PMT as loadable module Tony Luck
2026-09-28 22:14 ` [PATCH v13 01/25] fs/resctrl: Ensure default group reports tasks on monitor-only systems Tony Luck
2026-09-28 22:14 ` [PATCH v13 02/25] x86/cpufeatures: Add missing CQM feature dependency Tony Luck
2026-09-28 22:14 ` [PATCH v13 03/25] x86/resctrl: Check if monitoring features are supported Tony Luck
2026-09-28 22:14 ` [PATCH v13 04/25] x86/resctrl: Centralize monitoring feature enumeration Tony Luck
2026-09-28 22:14 ` [PATCH v13 05/25] x86/resctrl: Apply Intel MBM quirk from rdt_get_l3_mon_config() Tony Luck
2026-09-28 22:14 ` [PATCH v13 06/25] x86/resctrl: Delete resctrl_cpu_detect() Tony Luck
2026-09-28 22:14 ` [PATCH v13 07/25] arm,x86,fs/resctrl: Replace architecture resctrl_arch_{alloc,mon}_capable() Tony Luck
2026-09-28 22:14 ` [PATCH v13 08/25] x86/resctrl: Update special case for Intel Haswell enumeration Tony Luck
2026-09-28 22:14 ` [PATCH v13 09/25] x86/resctrl: Delete rdt_alloc_capable and rdt_mon_capable Tony Luck
2026-09-28 22:14 ` [PATCH v13 10/25] fs/resctrl: Remove redundant calls to resctrl_mon_capable() Tony Luck
2026-09-28 22:14 ` [PATCH v13 11/25] x86/resctrl: Honor rdt={perf|energy} options to force enable AET events Tony Luck
2026-09-28 22:14 ` [PATCH v13 12/25] fs/resctrl: Add interface to disable a monitor event Tony Luck
2026-09-28 22:14 ` [PATCH v13 13/25] arm,x86,fs/resctrl: Allocate maximum needed rmid_ptrs[] Tony Luck
2026-09-28 22:14 ` [PATCH v13 14/25] arm,x86,fs/resctrl: Use right size for L3 monitor data structures Tony Luck
2026-09-28 22:14 ` [PATCH v13 15/25] x86,fs/resctrl: Handle systems where AET is the only resource Tony Luck
2026-09-28 22:15 ` [PATCH v13 16/25] x86/resctrl: Add PMT registration API for AET enumeration callbacks Tony Luck
2026-09-28 22:15 ` [PATCH v13 17/25] platform/x86/intel/pmt: Register enumeration functions with resctrl Tony Luck
2026-09-28 22:15 ` [PATCH v13 18/25] x86/resctrl: Use registered function pointers for AET enumeration Tony Luck
2026-09-29 17:52 ` Tony Luck [this message]
2026-09-28 22:15 ` [PATCH v13 19/25] arm,x86,fs/resctrl: Enumerate AET on every resctrl mount Tony Luck
2026-09-29 17:52 ` [PATCH v14.1 " Tony Luck
2026-09-28 22:15 ` [PATCH v13 20/25] x86/resctrl: Enforce system RMID limit on AET Tony Luck
2026-09-29 17:52 ` [PATCH v14.1 " Tony Luck
2026-09-28 22:15 ` [PATCH v13 21/25] x86/resctrl: Export interface to report telemetry unbind/remove Tony Luck
2026-09-29 17:52 ` [PATCH v14.1 " Tony Luck
2026-09-28 22:15 ` [PATCH v13 22/25] platform/x86/intel/pmt: Inform resctrl when MMIO maps are being removed Tony Luck
2026-09-28 22:15 ` [PATCH v13 23/25] x86/resctrl: Require 64-bit x86 for resctrl support Tony Luck
2026-09-28 22:15 ` [PATCH v13 24/25] x86/resctrl: Simplify Kconfig options for resctrl Tony Luck
2026-09-28 22:15 ` [PATCH v13 25/25] x86,fs/resctrl: Document telemetry mount timing caveat Tony Luck
2026-09-29 0:29 ` [PATCH v13 00/25] Allow AET to use PMT as loadable module Luck, Tony
2026-09-29 19:43 ` Luck, Tony
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=20260929175224.16565-1-tony.luck@intel.com \
--to=tony.luck@intel.com \
--cc=Dave.Martin@arm.com \
--cc=babu.moger@amd.com \
--cc=david.e.box@intel.com \
--cc=dfustini@baylibre.com \
--cc=fenghuay@nvidia.com \
--cc=hch@infradead.org \
--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®