Resctrl subsystem can support two monitoring modes, "mbm_event" or
"default". In mbm_event mode, monitoring event can only accumulate data
while it is backed by a hardware counter. In "default" mode, resctrl
assumes there is a hardware counter for each event within every CTRL_MON
and MON group.
Introduce interface to switch between mbm_event and default modes.
Example:
To list the MBM monitor modes supported:
$ cat /sys/fs/resctrl/info/L3_MON/mbm_assign_mode
[mbm_event]
default
To enable the "mbm_event" monitoring mode:
$ echo "mbm_event" > /sys/fs/resctrl/info/L3_MON/mbm_assign_mode
To enable the "default" monitoring mode:
$ echo "default" > /sys/fs/resctrl/info/L3_MON/mbm_assign_mode
MBM event counters are automatically reset as part of changing the mode.
Clear both architectural and non-architectural event states to prevent
overflow conditions during the next event read. Also clear assignable
counter configuration on all the domains.
Signed-off-by: Babu Moger <babu.moger@amd.com>
---
v14: Updated the changelog to reflect the change in monitor mode naming.
Added the call resctrl_bmec_files_show() to enable/disable files
related to BMEC.
Added resctrl_set_mon_evt_cfg() to reset event configuration values
when mode is changes.
v13: Resolved the conflicts due to FS/ARCH restructure.
Introduced the new resctrl_init_evt_configuration() to initialize
the event modes and configuration values.
Added the call to resctrl_bmec_files_show() hide/show BMEC related
files.
v12: Fixed the documentation for a consistency.
Introduced mbm_cntr_free_all() and resctrl_reset_rmid_all() to clear
counters and non-architectural states when monitor mode is changed.
https://lore.kernel.org/lkml/b60b4f72-6245-46db-a126-428fb13b6310@intel.com/
v11: Changed the name of the function rdtgroup_mbm_assign_mode_write() to
resctrl_mbm_assign_mode_write().
Rewrote the commit message with context.
Added few more details in resctrl.rst about mbm_cntr_assign mode.
Re-arranged the text in resctrl.rst file.
v10: The call mbm_cntr_reset() has been moved to earlier patch.
Minor documentation update.
v9: Fixed extra spaces in user documentation.
Fixed problem changing the mode to mbm_cntr_assign mode when it is
not supported. Added extra checks to detect if systems supports it.
Used the rdtgroup_cntr_id_init to initialize cntr_id.
v8: Reset the internal counters after mbm_cntr_assign mode is changed.
Renamed rdtgroup_mbm_cntr_reset() to mbm_cntr_reset()
Updated the documentation to make text generic.
v7: Changed the interface name to mbm_assign_mode.
Removed the references of ABMC.
Added the changes to reset global and domain bitmaps.
Added the changes to reset rmid.
v6: Changed the mode name to mbm_cntr_assign.
Moved all the FS related code here.
Added changes to reset mbm_cntr_map and resctrl group counters.
v5: Change log and mode description text correction.
v4: Minor commit text changes. Keep the default to ABMC when supported.
Fixed comments to reflect changed interface "mbm_mode".
v3: New patch to address the review comments from upstream.
---
Documentation/filesystems/resctrl.rst | 23 +++++++-
fs/resctrl/internal.h | 2 +
fs/resctrl/monitor.c | 27 ++++++++++
fs/resctrl/rdtgroup.c | 78 +++++++++++++++++++++++++--
4 files changed, 126 insertions(+), 4 deletions(-)
diff --git a/Documentation/filesystems/resctrl.rst b/Documentation/filesystems/resctrl.rst
index cd82c2966ed7..7e62c7fdcefa 100644
--- a/Documentation/filesystems/resctrl.rst
+++ b/Documentation/filesystems/resctrl.rst
@@ -259,7 +259,9 @@ with the following files:
"mbm_assign_mode":
The supported monitoring modes. The enclosed brackets indicate which mode
- is enabled.
+ is enabled. The MBM events (mbm_total_bytes and/or mbm_local_bytes) associated
+ with counters may reset when "mbm_assign_mode" is changed.
+
::
# cat /sys/fs/resctrl/info/L3_MON/mbm_assign_mode
@@ -279,6 +281,15 @@ with the following files:
of counters available is described in the "num_mbm_cntrs" file. Changing the
mode may cause all counters on the resource to reset.
+ Moving to mbm_event mode require users to assign the counters to the events.
+ Otherwise, the MBM event counters will return 'Unassigned' when read.
+
+ The mode is beneficial for AMD platforms that support more CTRL_MON
+ and MON groups than available hardware counters. By default, this
+ feature is enabled on AMD platforms with the ABMC (Assignable Bandwidth
+ Monitoring Counters) capability, ensuring counters remain assigned even
+ when the corresponding RMID is not actively used by any processor.
+
"default":
In default mode, resctrl assumes there is a hardware counter for each
@@ -288,6 +299,16 @@ with the following files:
result in misleading values or display "Unavailable" if no counter is assigned
to the event.
+ * To enable "mbm_event" monitoring mode:
+ ::
+
+ # echo "mbm_event" > /sys/fs/resctrl/info/L3_MON/mbm_assign_mode
+
+ * To enable "default" monitoring mode:
+ ::
+
+ # echo "default" > /sys/fs/resctrl/info/L3_MON/mbm_assign_mode
+
"num_mbm_cntrs":
The maximum number of counter IDs (total of available and assigned counters)
in each domain when the system supports mbm_event mode.
diff --git a/fs/resctrl/internal.h b/fs/resctrl/internal.h
index 14d99c723ea5..adc9ff3efdfd 100644
--- a/fs/resctrl/internal.h
+++ b/fs/resctrl/internal.h
@@ -414,6 +414,8 @@ void resctrl_unassign_cntr_event(struct rdt_resource *r, struct rdt_mon_domain *
struct rdtgroup *rdtgrp, struct mon_evt *mevt);
int mbm_cntr_get(struct rdt_resource *r, struct rdt_mon_domain *d,
struct rdtgroup *rdtgrp, enum resctrl_event_id evtid);
+void resctrl_reset_rmid_all(struct rdt_resource *r, struct rdt_mon_domain *d);
+void mbm_cntr_free_all(struct rdt_resource *r, struct rdt_mon_domain *d);
#ifdef CONFIG_RESCTRL_FS_PSEUDO_LOCK
int rdtgroup_locksetup_enter(struct rdtgroup *rdtgrp);
diff --git a/fs/resctrl/monitor.c b/fs/resctrl/monitor.c
index 618c94cd1ad8..504b869570e6 100644
--- a/fs/resctrl/monitor.c
+++ b/fs/resctrl/monitor.c
@@ -1045,6 +1045,33 @@ static void mbm_cntr_free(struct rdt_mon_domain *d, int cntr_id)
memset(&d->cntr_cfg[cntr_id], 0, sizeof(struct mbm_cntr_cfg));
}
+/**
+ * mbm_cntr_free_all() - Clear all the counter ID configuration details in the
+ * domain @d. Called when mbm_assign_mode is changed.
+ */
+void mbm_cntr_free_all(struct rdt_resource *r, struct rdt_mon_domain *d)
+{
+ memset(d->cntr_cfg, 0, sizeof(*d->cntr_cfg) * r->mon.num_mbm_cntrs);
+}
+
+/**
+ * resctrl_reset_rmid_all() - Reset all non-architecture states for all the
+ * supported RMIDs.
+ */
+void resctrl_reset_rmid_all(struct rdt_resource *r, struct rdt_mon_domain *d)
+{
+ u32 idx_limit = resctrl_arch_system_num_rmid_idx();
+ enum resctrl_event_id evt;
+ int idx;
+
+ for_each_mbm_event_id(evt) {
+ if (!resctrl_is_mon_event_enabled(evt))
+ continue;
+ idx = MBM_STATE_IDX(evt);
+ memset(d->mbm_states[idx], 0, sizeof(struct mbm_state) * idx_limit);
+ }
+}
+
/**
* resctrl_alloc_config_cntr() - Allocate a counter ID and configure it for the
* event pointed to by @mevt and the resctrl group @rdtgrp within the domain @d.
diff --git a/fs/resctrl/rdtgroup.c b/fs/resctrl/rdtgroup.c
index 8c67e0897f25..6bb61fcf8673 100644
--- a/fs/resctrl/rdtgroup.c
+++ b/fs/resctrl/rdtgroup.c
@@ -1876,6 +1876,77 @@ static int resctrl_mbm_assign_mode_show(struct kernfs_open_file *of,
return 0;
}
+static ssize_t resctrl_mbm_assign_mode_write(struct kernfs_open_file *of,
+ char *buf, size_t nbytes, loff_t off)
+{
+ struct rdt_resource *r = rdt_kn_parent_priv(of->kn);
+ struct rdt_mon_domain *d;
+ int ret = 0;
+ bool enable;
+
+ /* Valid input requires a trailing newline */
+ if (nbytes == 0 || buf[nbytes - 1] != '\n')
+ return -EINVAL;
+
+ buf[nbytes - 1] = '\0';
+
+ cpus_read_lock();
+ mutex_lock(&rdtgroup_mutex);
+
+ rdt_last_cmd_clear();
+
+ if (!strcmp(buf, "default")) {
+ enable = 0;
+ } else if (!strcmp(buf, "mbm_event")) {
+ if (r->mon.mbm_cntr_assignable) {
+ enable = 1;
+ } else {
+ ret = -EINVAL;
+ rdt_last_cmd_puts("mbm_event mode is not supported\n");
+ goto write_exit;
+ }
+ } else {
+ ret = -EINVAL;
+ rdt_last_cmd_puts("Unsupported assign mode\n");
+ goto write_exit;
+ }
+
+ if (enable != resctrl_arch_mbm_cntr_assign_enabled(r)) {
+ ret = resctrl_arch_mbm_cntr_assign_set(r, enable);
+ if (ret)
+ goto write_exit;
+
+ /* Update the visibility of BMEC related files */
+ resctrl_bmec_files_show(r, !enable);
+
+ /*
+ * Initialize the default memory transaction values for
+ * total and local events.
+ */
+ if (resctrl_is_mon_event_enabled(QOS_L3_MBM_TOTAL_EVENT_ID))
+ resctrl_set_mon_evt_cfg(QOS_L3_MBM_TOTAL_EVENT_ID,
+ MAX_EVT_CONFIG_BITS);
+ if (resctrl_is_mon_event_enabled(QOS_L3_MBM_LOCAL_EVENT_ID))
+ resctrl_set_mon_evt_cfg(QOS_L3_MBM_LOCAL_EVENT_ID,
+ READS_TO_LOCAL_MEM |
+ READS_TO_LOCAL_S_MEM |
+ NON_TEMP_WRITE_TO_LOCAL_MEM);
+ /*
+ * Reset all the non-achitectural RMID state and assignable counters.
+ */
+ list_for_each_entry(d, &r->mon_domains, hdr.list) {
+ mbm_cntr_free_all(r, d);
+ resctrl_reset_rmid_all(r, d);
+ }
+ }
+
+write_exit:
+ mutex_unlock(&rdtgroup_mutex);
+ cpus_read_unlock();
+
+ return ret ?: nbytes;
+}
+
static int resctrl_num_mbm_cntrs_show(struct kernfs_open_file *of,
struct seq_file *s, void *v)
{
@@ -2203,8 +2274,8 @@ static int resctrl_process_assign(struct rdt_resource *r, struct rdtgroup *rdtgr
struct mon_evt *mevt;
int assign_state;
char domain[10];
+ int ret = 0;
bool found;
- int ret;
mevt = mbm_get_mon_event_by_name(r, event);
if (!mevt) {
@@ -2249,7 +2320,7 @@ static int resctrl_process_assign(struct rdt_resource *r, struct rdtgroup *rdtgr
switch (assign_state) {
case ASSIGN_NONE:
- ret = resctrl_unassign_cntr_event(r, d, rdtgrp, mevt);
+ resctrl_unassign_cntr_event(r, d, rdtgrp, mevt);
break;
case ASSIGN_EXCLUSIVE:
ret = resctrl_assign_cntr_event(r, d, rdtgrp, mevt);
@@ -2463,9 +2534,10 @@ static struct rftype res_common_files[] = {
},
{
.name = "mbm_assign_mode",
- .mode = 0444,
+ .mode = 0644,
.kf_ops = &rdtgroup_kf_single_ops,
.seq_show = resctrl_mbm_assign_mode_show,
+ .write = resctrl_mbm_assign_mode_write,
.fflags = RFTYPE_MON_INFO | RFTYPE_RES_CACHE,
},
{
--
2.34.1Hi Babu,
On 6/13/25 2:05 PM, Babu Moger wrote:
> Resctrl subsystem can support two monitoring modes, "mbm_event" or
> "default". In mbm_event mode, monitoring event can only accumulate data
> while it is backed by a hardware counter. In "default" mode, resctrl
> assumes there is a hardware counter for each event within every CTRL_MON
> and MON group.
>
> Introduce interface to switch between mbm_event and default modes.
"Introduce interface" -> "Introduce mbm_assign_mode resctrl file"
>
> Example:
> To list the MBM monitor modes supported:
> $ cat /sys/fs/resctrl/info/L3_MON/mbm_assign_mode
> [mbm_event]
> default
>
> To enable the "mbm_event" monitoring mode:
> $ echo "mbm_event" > /sys/fs/resctrl/info/L3_MON/mbm_assign_mode
>
> To enable the "default" monitoring mode:
> $ echo "default" > /sys/fs/resctrl/info/L3_MON/mbm_assign_mode
>
> MBM event counters are automatically reset as part of changing the mode.
> Clear both architectural and non-architectural event states to prevent
> overflow conditions during the next event read. Also clear assignable
> counter configuration on all the domains.
>
> Signed-off-by: Babu Moger <babu.moger@amd.com>
> ---
...
> ---
> Documentation/filesystems/resctrl.rst | 23 +++++++-
> fs/resctrl/internal.h | 2 +
> fs/resctrl/monitor.c | 27 ++++++++++
> fs/resctrl/rdtgroup.c | 78 +++++++++++++++++++++++++--
> 4 files changed, 126 insertions(+), 4 deletions(-)
>
> diff --git a/Documentation/filesystems/resctrl.rst b/Documentation/filesystems/resctrl.rst
> index cd82c2966ed7..7e62c7fdcefa 100644
> --- a/Documentation/filesystems/resctrl.rst
> +++ b/Documentation/filesystems/resctrl.rst
> @@ -259,7 +259,9 @@ with the following files:
>
> "mbm_assign_mode":
> The supported monitoring modes. The enclosed brackets indicate which mode
> - is enabled.
> + is enabled. The MBM events (mbm_total_bytes and/or mbm_local_bytes) associated
Since there may be more events in future I think the "(mbm_total_bytes and/or
mbm_local_bytes)" can be dropped.
> + with counters may reset when "mbm_assign_mode" is changed.
> +
> ::
>
> # cat /sys/fs/resctrl/info/L3_MON/mbm_assign_mode
> @@ -279,6 +281,15 @@ with the following files:
> of counters available is described in the "num_mbm_cntrs" file. Changing the
> mode may cause all counters on the resource to reset.
>
> + Moving to mbm_event mode require users to assign the counters to the events.
"Moving to mbm_event mode require" -> "mbm_event counter assignment mode requires"
> + Otherwise, the MBM event counters will return 'Unassigned' when read.
> +
> + The mode is beneficial for AMD platforms that support more CTRL_MON
> + and MON groups than available hardware counters. By default, this
> + feature is enabled on AMD platforms with the ABMC (Assignable Bandwidth
> + Monitoring Counters) capability, ensuring counters remain assigned even
> + when the corresponding RMID is not actively used by any processor.
> +
> "default":
>
> In default mode, resctrl assumes there is a hardware counter for each
> @@ -288,6 +299,16 @@ with the following files:
> result in misleading values or display "Unavailable" if no counter is assigned
> to the event.
>
> + * To enable "mbm_event" monitoring mode:
> + ::
> +
> + # echo "mbm_event" > /sys/fs/resctrl/info/L3_MON/mbm_assign_mode
> +
> + * To enable "default" monitoring mode:
> + ::
> +
> + # echo "default" > /sys/fs/resctrl/info/L3_MON/mbm_assign_mode
> +
> "num_mbm_cntrs":
> The maximum number of counter IDs (total of available and assigned counters)
> in each domain when the system supports mbm_event mode.
> diff --git a/fs/resctrl/internal.h b/fs/resctrl/internal.h
> index 14d99c723ea5..adc9ff3efdfd 100644
> --- a/fs/resctrl/internal.h
> +++ b/fs/resctrl/internal.h
> @@ -414,6 +414,8 @@ void resctrl_unassign_cntr_event(struct rdt_resource *r, struct rdt_mon_domain *
> struct rdtgroup *rdtgrp, struct mon_evt *mevt);
> int mbm_cntr_get(struct rdt_resource *r, struct rdt_mon_domain *d,
> struct rdtgroup *rdtgrp, enum resctrl_event_id evtid);
> +void resctrl_reset_rmid_all(struct rdt_resource *r, struct rdt_mon_domain *d);
> +void mbm_cntr_free_all(struct rdt_resource *r, struct rdt_mon_domain *d);
>
> #ifdef CONFIG_RESCTRL_FS_PSEUDO_LOCK
> int rdtgroup_locksetup_enter(struct rdtgroup *rdtgrp);
> diff --git a/fs/resctrl/monitor.c b/fs/resctrl/monitor.c
> index 618c94cd1ad8..504b869570e6 100644
> --- a/fs/resctrl/monitor.c
> +++ b/fs/resctrl/monitor.c
> @@ -1045,6 +1045,33 @@ static void mbm_cntr_free(struct rdt_mon_domain *d, int cntr_id)
> memset(&d->cntr_cfg[cntr_id], 0, sizeof(struct mbm_cntr_cfg));
> }
>
> +/**
> + * mbm_cntr_free_all() - Clear all the counter ID configuration details in the
> + * domain @d. Called when mbm_assign_mode is changed.
> + */
> +void mbm_cntr_free_all(struct rdt_resource *r, struct rdt_mon_domain *d)
> +{
> + memset(d->cntr_cfg, 0, sizeof(*d->cntr_cfg) * r->mon.num_mbm_cntrs);
> +}
> +
> +/**
> + * resctrl_reset_rmid_all() - Reset all non-architecture states for all the
> + * supported RMIDs.
> + */
> +void resctrl_reset_rmid_all(struct rdt_resource *r, struct rdt_mon_domain *d)
struct rdt_resource *r is unused? At this time it seems unnecessary to check if
an MBM event belongs to particular resource since at this point I expect only L3
is possible. Even so, to be consistent and robust I think it would make flows
easier to understand by always ensureing that mon_evt::rid matches
expected resource.
> +{
> + u32 idx_limit = resctrl_arch_system_num_rmid_idx();
> + enum resctrl_event_id evt;
> + int idx;
> +
> + for_each_mbm_event_id(evt) {
> + if (!resctrl_is_mon_event_enabled(evt))
> + continue;
> + idx = MBM_STATE_IDX(evt);
> + memset(d->mbm_states[idx], 0, sizeof(struct mbm_state) * idx_limit);
sizeof(*d->mbm_states[0])?
> + }
> +}
> +
> /**
> * resctrl_alloc_config_cntr() - Allocate a counter ID and configure it for the
> * event pointed to by @mevt and the resctrl group @rdtgrp within the domain @d.
> diff --git a/fs/resctrl/rdtgroup.c b/fs/resctrl/rdtgroup.c
> index 8c67e0897f25..6bb61fcf8673 100644
> --- a/fs/resctrl/rdtgroup.c
> +++ b/fs/resctrl/rdtgroup.c
> @@ -1876,6 +1876,77 @@ static int resctrl_mbm_assign_mode_show(struct kernfs_open_file *of,
> return 0;
> }
>
> +static ssize_t resctrl_mbm_assign_mode_write(struct kernfs_open_file *of,
> + char *buf, size_t nbytes, loff_t off)
> +{
> + struct rdt_resource *r = rdt_kn_parent_priv(of->kn);
> + struct rdt_mon_domain *d;
> + int ret = 0;
> + bool enable;
> +
> + /* Valid input requires a trailing newline */
> + if (nbytes == 0 || buf[nbytes - 1] != '\n')
> + return -EINVAL;
> +
> + buf[nbytes - 1] = '\0';
> +
> + cpus_read_lock();
> + mutex_lock(&rdtgroup_mutex);
> +
> + rdt_last_cmd_clear();
> +
> + if (!strcmp(buf, "default")) {
> + enable = 0;
> + } else if (!strcmp(buf, "mbm_event")) {
> + if (r->mon.mbm_cntr_assignable) {
> + enable = 1;
> + } else {
> + ret = -EINVAL;
> + rdt_last_cmd_puts("mbm_event mode is not supported\n");
> + goto write_exit;
write_exit -> out_unlock
> + }
> + } else {
> + ret = -EINVAL;
> + rdt_last_cmd_puts("Unsupported assign mode\n");
> + goto write_exit;
> + }
> +
> + if (enable != resctrl_arch_mbm_cntr_assign_enabled(r)) {
> + ret = resctrl_arch_mbm_cntr_assign_set(r, enable);
> + if (ret)
> + goto write_exit;
> +
> + /* Update the visibility of BMEC related files */
> + resctrl_bmec_files_show(r, !enable);
> +
> + /*
> + * Initialize the default memory transaction values for
> + * total and local events.
> + */
> + if (resctrl_is_mon_event_enabled(QOS_L3_MBM_TOTAL_EVENT_ID))
> + resctrl_set_mon_evt_cfg(QOS_L3_MBM_TOTAL_EVENT_ID,
> + MAX_EVT_CONFIG_BITS);
> + if (resctrl_is_mon_event_enabled(QOS_L3_MBM_LOCAL_EVENT_ID))
> + resctrl_set_mon_evt_cfg(QOS_L3_MBM_LOCAL_EVENT_ID,
> + READS_TO_LOCAL_MEM |
> + READS_TO_LOCAL_S_MEM |
> + NON_TEMP_WRITE_TO_LOCAL_MEM);
Nice, yes, this belongs in resctrl fs.
> + /*
> + * Reset all the non-achitectural RMID state and assignable counters.
> + */
> + list_for_each_entry(d, &r->mon_domains, hdr.list) {
> + mbm_cntr_free_all(r, d);
> + resctrl_reset_rmid_all(r, d);
> + }
> + }
> +
> +write_exit:
> + mutex_unlock(&rdtgroup_mutex);
> + cpus_read_unlock();
> +
> + return ret ?: nbytes;
> +}
> +
> static int resctrl_num_mbm_cntrs_show(struct kernfs_open_file *of,
> struct seq_file *s, void *v)
> {
> @@ -2203,8 +2274,8 @@ static int resctrl_process_assign(struct rdt_resource *r, struct rdtgroup *rdtgr
> struct mon_evt *mevt;
> int assign_state;
> char domain[10];
> + int ret = 0;
> bool found;
> - int ret;
>
> mevt = mbm_get_mon_event_by_name(r, event);
> if (!mevt) {
> @@ -2249,7 +2320,7 @@ static int resctrl_process_assign(struct rdt_resource *r, struct rdtgroup *rdtgr
>
> switch (assign_state) {
> case ASSIGN_NONE:
> - ret = resctrl_unassign_cntr_event(r, d, rdtgrp, mevt);
> + resctrl_unassign_cntr_event(r, d, rdtgrp, mevt);
> break;
> case ASSIGN_EXCLUSIVE:
> ret = resctrl_assign_cntr_event(r, d, rdtgrp, mevt);
Two stray hunks?
> @@ -2463,9 +2534,10 @@ static struct rftype res_common_files[] = {
> },
> {
> .name = "mbm_assign_mode",
> - .mode = 0444,
> + .mode = 0644,
> .kf_ops = &rdtgroup_kf_single_ops,
> .seq_show = resctrl_mbm_assign_mode_show,
> + .write = resctrl_mbm_assign_mode_write,
> .fflags = RFTYPE_MON_INFO | RFTYPE_RES_CACHE,
> },
> {
Reinette
Hi Reinette,
On 6/25/25 18:40, Reinette Chatre wrote:
> Hi Babu,
>
> On 6/13/25 2:05 PM, Babu Moger wrote:
>> Resctrl subsystem can support two monitoring modes, "mbm_event" or
>> "default". In mbm_event mode, monitoring event can only accumulate data
>> while it is backed by a hardware counter. In "default" mode, resctrl
>> assumes there is a hardware counter for each event within every CTRL_MON
>> and MON group.
>>
>> Introduce interface to switch between mbm_event and default modes.
>
> "Introduce interface" -> "Introduce mbm_assign_mode resctrl file"
>
Sure.
>>
>> Example:
>> To list the MBM monitor modes supported:
>> $ cat /sys/fs/resctrl/info/L3_MON/mbm_assign_mode
>> [mbm_event]
>> default
>>
>> To enable the "mbm_event" monitoring mode:
>> $ echo "mbm_event" > /sys/fs/resctrl/info/L3_MON/mbm_assign_mode
>>
>> To enable the "default" monitoring mode:
>> $ echo "default" > /sys/fs/resctrl/info/L3_MON/mbm_assign_mode
>>
>> MBM event counters are automatically reset as part of changing the mode.
>> Clear both architectural and non-architectural event states to prevent
>> overflow conditions during the next event read. Also clear assignable
>> counter configuration on all the domains.
>>
>> Signed-off-by: Babu Moger <babu.moger@amd.com>
>> ---
>
> ...
>
>> ---
>> Documentation/filesystems/resctrl.rst | 23 +++++++-
>> fs/resctrl/internal.h | 2 +
>> fs/resctrl/monitor.c | 27 ++++++++++
>> fs/resctrl/rdtgroup.c | 78 +++++++++++++++++++++++++--
>> 4 files changed, 126 insertions(+), 4 deletions(-)
>>
>> diff --git a/Documentation/filesystems/resctrl.rst b/Documentation/filesystems/resctrl.rst
>> index cd82c2966ed7..7e62c7fdcefa 100644
>> --- a/Documentation/filesystems/resctrl.rst
>> +++ b/Documentation/filesystems/resctrl.rst
>> @@ -259,7 +259,9 @@ with the following files:
>>
>> "mbm_assign_mode":
>> The supported monitoring modes. The enclosed brackets indicate which mode
>> - is enabled.
>> + is enabled. The MBM events (mbm_total_bytes and/or mbm_local_bytes) associated
>
> Since there may be more events in future I think the "(mbm_total_bytes and/or
> mbm_local_bytes)" can be dropped.
Sure.
>
>> + with counters may reset when "mbm_assign_mode" is changed.
>> +
>> ::
>>
>> # cat /sys/fs/resctrl/info/L3_MON/mbm_assign_mode
>> @@ -279,6 +281,15 @@ with the following files:
>> of counters available is described in the "num_mbm_cntrs" file. Changing the
>> mode may cause all counters on the resource to reset.
>>
>> + Moving to mbm_event mode require users to assign the counters to the events.
>
> "Moving to mbm_event mode require" -> "mbm_event counter assignment mode requires"
Sure.
>
>> + Otherwise, the MBM event counters will return 'Unassigned' when read.
>> +
>> + The mode is beneficial for AMD platforms that support more CTRL_MON
>> + and MON groups than available hardware counters. By default, this
>> + feature is enabled on AMD platforms with the ABMC (Assignable Bandwidth
>> + Monitoring Counters) capability, ensuring counters remain assigned even
>> + when the corresponding RMID is not actively used by any processor.
>> +
>> "default":
>>
>> In default mode, resctrl assumes there is a hardware counter for each
>> @@ -288,6 +299,16 @@ with the following files:
>> result in misleading values or display "Unavailable" if no counter is assigned
>> to the event.
>>
>> + * To enable "mbm_event" monitoring mode:
>> + ::
>> +
>> + # echo "mbm_event" > /sys/fs/resctrl/info/L3_MON/mbm_assign_mode
>> +
>> + * To enable "default" monitoring mode:
>> + ::
>> +
>> + # echo "default" > /sys/fs/resctrl/info/L3_MON/mbm_assign_mode
>> +
>> "num_mbm_cntrs":
>> The maximum number of counter IDs (total of available and assigned counters)
>> in each domain when the system supports mbm_event mode.
>> diff --git a/fs/resctrl/internal.h b/fs/resctrl/internal.h
>> index 14d99c723ea5..adc9ff3efdfd 100644
>> --- a/fs/resctrl/internal.h
>> +++ b/fs/resctrl/internal.h
>> @@ -414,6 +414,8 @@ void resctrl_unassign_cntr_event(struct rdt_resource *r, struct rdt_mon_domain *
>> struct rdtgroup *rdtgrp, struct mon_evt *mevt);
>> int mbm_cntr_get(struct rdt_resource *r, struct rdt_mon_domain *d,
>> struct rdtgroup *rdtgrp, enum resctrl_event_id evtid);
>> +void resctrl_reset_rmid_all(struct rdt_resource *r, struct rdt_mon_domain *d);
>> +void mbm_cntr_free_all(struct rdt_resource *r, struct rdt_mon_domain *d);
>>
>> #ifdef CONFIG_RESCTRL_FS_PSEUDO_LOCK
>> int rdtgroup_locksetup_enter(struct rdtgroup *rdtgrp);
>> diff --git a/fs/resctrl/monitor.c b/fs/resctrl/monitor.c
>> index 618c94cd1ad8..504b869570e6 100644
>> --- a/fs/resctrl/monitor.c
>> +++ b/fs/resctrl/monitor.c
>> @@ -1045,6 +1045,33 @@ static void mbm_cntr_free(struct rdt_mon_domain *d, int cntr_id)
>> memset(&d->cntr_cfg[cntr_id], 0, sizeof(struct mbm_cntr_cfg));
>> }
>>
>> +/**
>> + * mbm_cntr_free_all() - Clear all the counter ID configuration details in the
>> + * domain @d. Called when mbm_assign_mode is changed.
>> + */
>> +void mbm_cntr_free_all(struct rdt_resource *r, struct rdt_mon_domain *d)
>> +{
>> + memset(d->cntr_cfg, 0, sizeof(*d->cntr_cfg) * r->mon.num_mbm_cntrs);
>> +}
>> +
>> +/**
>> + * resctrl_reset_rmid_all() - Reset all non-architecture states for all the
>> + * supported RMIDs.
>> + */
>> +void resctrl_reset_rmid_all(struct rdt_resource *r, struct rdt_mon_domain *d)
>
> struct rdt_resource *r is unused? At this time it seems unnecessary to check if
> an MBM event belongs to particular resource since at this point I expect only L3
> is possible. Even so, to be consistent and robust I think it would make flows
> easier to understand by always ensureing that mon_evt::rid matches
> expected resource.
Agree. But in this we dont access mon_evt structure here.
>
>> +{
>> + u32 idx_limit = resctrl_arch_system_num_rmid_idx();
>> + enum resctrl_event_id evt;
>> + int idx;
>> +
>> + for_each_mbm_event_id(evt) {
>> + if (!resctrl_is_mon_event_enabled(evt))
>> + continue;
>> + idx = MBM_STATE_IDX(evt);
>> + memset(d->mbm_states[idx], 0, sizeof(struct mbm_state) * idx_limit);
>
> sizeof(*d->mbm_states[0])?
>
Sure.
>> + }
>> +}
>> +
>> /**
>> * resctrl_alloc_config_cntr() - Allocate a counter ID and configure it for the
>> * event pointed to by @mevt and the resctrl group @rdtgrp within the domain @d.
>> diff --git a/fs/resctrl/rdtgroup.c b/fs/resctrl/rdtgroup.c
>> index 8c67e0897f25..6bb61fcf8673 100644
>> --- a/fs/resctrl/rdtgroup.c
>> +++ b/fs/resctrl/rdtgroup.c
>> @@ -1876,6 +1876,77 @@ static int resctrl_mbm_assign_mode_show(struct kernfs_open_file *of,
>> return 0;
>> }
>>
>> +static ssize_t resctrl_mbm_assign_mode_write(struct kernfs_open_file *of,
>> + char *buf, size_t nbytes, loff_t off)
>> +{
>> + struct rdt_resource *r = rdt_kn_parent_priv(of->kn);
>> + struct rdt_mon_domain *d;
>> + int ret = 0;
>> + bool enable;
>> +
>> + /* Valid input requires a trailing newline */
>> + if (nbytes == 0 || buf[nbytes - 1] != '\n')
>> + return -EINVAL;
>> +
>> + buf[nbytes - 1] = '\0';
>> +
>> + cpus_read_lock();
>> + mutex_lock(&rdtgroup_mutex);
>> +
>> + rdt_last_cmd_clear();
>> +
>> + if (!strcmp(buf, "default")) {
>> + enable = 0;
>> + } else if (!strcmp(buf, "mbm_event")) {
>> + if (r->mon.mbm_cntr_assignable) {
>> + enable = 1;
>> + } else {
>> + ret = -EINVAL;
>> + rdt_last_cmd_puts("mbm_event mode is not supported\n");
>> + goto write_exit;
>
> write_exit -> out_unlock
>
Sure.
>> + }
>> + } else {
>> + ret = -EINVAL;
>> + rdt_last_cmd_puts("Unsupported assign mode\n");
>> + goto write_exit;
>> + }
>> +
>> + if (enable != resctrl_arch_mbm_cntr_assign_enabled(r)) {
>> + ret = resctrl_arch_mbm_cntr_assign_set(r, enable);
>> + if (ret)
>> + goto write_exit;
>> +
>> + /* Update the visibility of BMEC related files */
>> + resctrl_bmec_files_show(r, !enable);
>> +
>> + /*
>> + * Initialize the default memory transaction values for
>> + * total and local events.
>> + */
>> + if (resctrl_is_mon_event_enabled(QOS_L3_MBM_TOTAL_EVENT_ID))
>> + resctrl_set_mon_evt_cfg(QOS_L3_MBM_TOTAL_EVENT_ID,
>> + MAX_EVT_CONFIG_BITS);
>> + if (resctrl_is_mon_event_enabled(QOS_L3_MBM_LOCAL_EVENT_ID))
>> + resctrl_set_mon_evt_cfg(QOS_L3_MBM_LOCAL_EVENT_ID,
>> + READS_TO_LOCAL_MEM |
>> + READS_TO_LOCAL_S_MEM |
>> + NON_TEMP_WRITE_TO_LOCAL_MEM);
>
> Nice, yes, this belongs in resctrl fs.
>
>> + /*
>> + * Reset all the non-achitectural RMID state and assignable counters.
>> + */
>> + list_for_each_entry(d, &r->mon_domains, hdr.list) {
>> + mbm_cntr_free_all(r, d);
>> + resctrl_reset_rmid_all(r, d);
>> + }
>> + }
>> +
>> +write_exit:
>> + mutex_unlock(&rdtgroup_mutex);
>> + cpus_read_unlock();
>> +
>> + return ret ?: nbytes;
>> +}
>> +
>> static int resctrl_num_mbm_cntrs_show(struct kernfs_open_file *of,
>> struct seq_file *s, void *v)
>> {
>> @@ -2203,8 +2274,8 @@ static int resctrl_process_assign(struct rdt_resource *r, struct rdtgroup *rdtgr
>> struct mon_evt *mevt;
>> int assign_state;
>> char domain[10];
>> + int ret = 0;
>> bool found;
>> - int ret;
>>
>> mevt = mbm_get_mon_event_by_name(r, event);
>> if (!mevt) {
>> @@ -2249,7 +2320,7 @@ static int resctrl_process_assign(struct rdt_resource *r, struct rdtgroup *rdtgr
>>
>> switch (assign_state) {
>> case ASSIGN_NONE:
>> - ret = resctrl_unassign_cntr_event(r, d, rdtgrp, mevt);
>> + resctrl_unassign_cntr_event(r, d, rdtgrp, mevt);
>> break;
>> case ASSIGN_EXCLUSIVE:
>> ret = resctrl_assign_cntr_event(r, d, rdtgrp, mevt);
>
> Two stray hunks?
>
Yes. Fixed it.
>> @@ -2463,9 +2534,10 @@ static struct rftype res_common_files[] = {
>> },
>> {
>> .name = "mbm_assign_mode",
>> - .mode = 0444,
>> + .mode = 0644,
>> .kf_ops = &rdtgroup_kf_single_ops,
>> .seq_show = resctrl_mbm_assign_mode_show,
>> + .write = resctrl_mbm_assign_mode_write,
>> .fflags = RFTYPE_MON_INFO | RFTYPE_RES_CACHE,
>> },
>> {
>
> Reinette
>
--
Thanks
Babu Moger
© 2016 - 2025 Red Hat, Inc.