Re: [PATCH v15 31/34] fs/resctrl: Introduce the interface to modify assignments in a group

["Moger, Babu" <babu.moger@amd.com>]

Hi Reinette,

On 7/17/25 23:01, Reinette Chatre wrote:
> Hi Babu,
> 
> On 7/8/25 3:17 PM, Babu Moger wrote:
>> Enable the mbm_l3_assignments resctrl file to be used to modify counter
>> assignments of CTRL_MON and MON groups when the "mbm_event" counter
>> assignment mode is enabled.
>>
>> The assignment modifications are done in the following format:
>> <Event>:<Domain id>=<Assignment state>
>>
>> Event: A valid MBM event in the
>>        /sys/fs/resctrl/info/L3_MON/event_configs directory.
>>
>> Domain ID: A valid domain ID. When writing, '*' applies the changes
>> 	   to all domains.
>>
>> Assignment states:
>>
>>     _ : Unassign the counter.
> 
> "the counter" -> "a counter" (user cannot control which counter is assigned)

Sure.

> 
>>
>>     e : Assign the counter exclusively.
> 
> "the counter" -> "a counter"

Sure.

> 
>>
>> Examples:
>>
>> $ cd /sys/fs/resctrl
>> $ cat /sys/fs/resctrl/mbm_L3_assignments
>>   mbm_total_bytes:0=e;1=e
>>   mbm_local_bytes:0=e;1=e
>>
>> To unassign the counter associated with the mbm_total_bytes event on
>> domain 0:
>>
>> $ echo "mbm_total_bytes:0=_" > mbm_L3_assignments
>> $ cat /sys/fs/resctrl/mbm_L3_assignments
>>   mbm_total_bytes:0=_;1=e
>>   mbm_local_bytes:0=e;1=e
>>
>> To unassign the counter associated with the mbm_total_bytes event on
>> all the domains:
>>
>> $ echo "mbm_total_bytes:*=_" > mbm_L3_assignments
>> $ cat /sys/fs/resctrl/mbm_L3_assignments
>>   mbm_total_bytes:0=_;1=_
>>   mbm_local_bytes:0=e;1=e
>>
>> Signed-off-by: Babu Moger <babu.moger@amd.com>
>> ---
> 
> ...
> 
>> ---
>>  Documentation/filesystems/resctrl.rst | 146 +++++++++++++++++++++++++-
>>  fs/resctrl/internal.h                 |   2 +
>>  fs/resctrl/monitor.c                  |  18 ++++
>>  fs/resctrl/rdtgroup.c                 | 123 +++++++++++++++++++++-
>>  4 files changed, 287 insertions(+), 2 deletions(-)
>>
>> diff --git a/Documentation/filesystems/resctrl.rst b/Documentation/filesystems/resctrl.rst
>> index 4dd1d605d6ad..2b2acb55d8b1 100644
>> --- a/Documentation/filesystems/resctrl.rst
>> +++ b/Documentation/filesystems/resctrl.rst
>> @@ -527,7 +527,8 @@ When the "mba_MBps" mount option is used all CTRL_MON groups will also contain:
>>  	Event: A valid MBM event in the
>>  	       /sys/fs/resctrl/info/L3_MON/event_configs directory.
>>  
>> -	Domain ID: A valid domain ID.
>> +	Domain ID: A valid domain ID. When writing, '*' applies the changes
>> +		   to all the domains.
>>  
>>  	Assignment states:
>>  
>> @@ -544,6 +545,34 @@ When the "mba_MBps" mount option is used all CTRL_MON groups will also contain:
>>  	   mbm_total_bytes:0=e;1=e
>>  	   mbm_local_bytes:0=e;1=e
>>  
>> +	Assignments can be modified by writing to the interface.
>> +
>> +	Example:
>> +	To unassign the counter associated with the mbm_total_bytes event on domain 0:
>> +	::
>> +
>> +	 # echo "mbm_total_bytes:0=_" > /sys/fs/resctrl/mbm_L3_assignments
>> +	 # cat /sys/fs/resctrl/mbm_L3_assignments
>> +	   mbm_total_bytes:0=_;1=e
>> +	   mbm_local_bytes:0=e;1=e
>> +
>> +	To unassign the counter associated with the mbm_total_bytes event on all the domains:
>> +	::
>> +
>> +	 # echo "mbm_total_bytes:*=_" > /sys/fs/resctrl/mbm_L3_assignments
>> +	 # cat /sys/fs/resctrl/mbm_L3_assignments
>> +	   mbm_total_bytes:0=_;1=_
>> +	   mbm_local_bytes:0=e;1=e
>> +
>> +	To assign the counter associated with the mbm_total_bytes event on all domains in
> 
> "the counter" -> "a counter" (motivation: user space cannot control which counter is assigned)

Sure.

> 
>> +	exclusive mode:
>> +	::
>> +
>> +	 # echo "mbm_total_bytes:*=e" > /sys/fs/resctrl/mbm_L3_assignments
>> +	 # cat /sys/fs/resctrl/mbm_L3_assignments
>> +	   mbm_total_bytes:0=e;1=e
>> +	   mbm_local_bytes:0=e;1=e
>> +
>>  Resource allocation rules
>>  -------------------------
>>  
>> @@ -1579,6 +1608,121 @@ View the llc occupancy snapshot::
>>    # cat /sys/fs/resctrl/p1/mon_data/mon_L3_00/llc_occupancy
>>    11234000
>>  
>> +
>> +Examples on working with mbm_assign_mode
>> +========================================
>> +
>> +a. Check if MBM counter assignment mode is supported.
>> +::
>> +
>> +  #mount -t resctrl resctrl /sys/fs/resctrl/
> 
> "#mount" -> "# mount" (please build htmldocs to confirm everything
> looks consistent)

Sure.

> 
>> +
>> +  # cat /sys/fs/resctrl/info/L3_MON/mbm_assign_mode
>> +  [mbm_event]
>> +  default
>> +
>> +The "mbm_event" mode is detected and enabled.
>> +
>> +b. Check how many assignable counters are supported.
>> +::
>> +
>> +  # cat /sys/fs/resctrl/info/L3_MON/num_mbm_cntrs
>> +  0=32;1=32
>> +
>> +c. Check how many assignable counters are available for assignment in each domain.
>> +::
>> +
>> +  # cat /sys/fs/resctrl/info/L3_MON/available_mbm_cntrs
>> +  0=30;1=30
>> +
>> +d. To list the default group's assign states:
> 
> Please note that these steps are inconsistent. Some end with "." and some end with ":"

Fixed it.

> 
>> +::
>> +
>> +  # cat /sys/fs/resctrl/mbm_L3_assignments
>> +  mbm_total_bytes:0=e;1=e
>> +  mbm_local_bytes:0=e;1=e
>> +
>> +e.  To unassign the counter associated with the mbm_total_bytes event on domain 0:
>> +::
>> +
>> +  # echo "mbm_total_bytes:0=_" > /sys/fs/resctrl/mbm_L3_assignments
>> +  # cat /sys/fs/resctrl/mbm_L3_assignments
>> +  mbm_total_bytes:0=_;1=e
>> +  mbm_local_bytes:0=e;1=e
>> +
>> +f. To unassign the counter associated with the mbm_total_bytes event on all domains:
>> +::
>> +
>> +  # echo "mbm_total_bytes:*=_" > /sys/fs/resctrl/mbm_L3_assignments
>> +  # cat /sys/fs/resctrl/mbm_L3_assignment
>> +  mbm_total_bytes:0=_;1=_
>> +  mbm_local_bytes:0=e;1=e
>> +
>> +g. To assign a counter associated with the mbm_total_bytes event on all domains in
> 
> ah ... here "a counter" is used.
> 
>> +exclusive mode:
>> +::
>> +
>> +  # echo "mbm_total_bytes:*=e" > /sys/fs/resctrl/mbm_L3_assignments
>> +  # cat /sys/fs/resctrl/mbm_L3_assignments
>> +  mbm_total_bytes:0=e;1=e
>> +  mbm_local_bytes:0=e;1=e
>> +
>> +h. Read the events mbm_total_bytes and mbm_local_bytes of the default group. There is
>> +no change in reading the events with the assignment.  If the event is unassigned when
>> +reading, then the read will come back as "Unassigned".
> 
> Since returning "Unassigned" can be complicated, the associated description should
> probably just be done once in the doc. I expect this will be in the "mon_data" section.

Addressed with
https://lore.kernel.org/lkml/2c107793-1258-4276-8b29-03359fa605f2@amd.com/

> 
>> +::
>> +
>> +  # cat /sys/fs/resctrl/mon_data/mon_L3_00/mbm_total_bytes
>> +  779247936
>> +  # cat /sys/fs/resctrl/mon_data/mon_L3_00/mbm_local_bytes
>> +  765207488
>> +
>> +i. Check the default event configurations.
> 
> Drop "default" (there are only these event configurations)
> 

Sure.

>> +::
>> +
>> +  # cat /sys/fs/resctrl/info/L3_MON/event_configs/mbm_total_bytes/event_filter
>> +  local_reads,remote_reads,local_non_temporal_writes,remote_non_temporal_writes,
>> +  local_reads_slow_memory,remote_reads_slow_memory,dirty_victim_writes_all
>> +
>> +  # cat /sys/fs/resctrl/info/L3_MON/event_configs/mbm_local_bytes/event_filter
>> +  local_reads,local_non_temporal_writes,local_reads_slow_memory
>> +
>> +j. Change the event configuration for mbm_local_bytes.
>> +::
>> +
>> +  # echo "local_reads, local_non_temporal_writes, local_reads_slow_memory, remote_reads" >
>> +  /sys/fs/resctrl/info/L3_MON/counter_configs/mbm_local_bytes/event_filter
> 
> "counter_configs" -> "event_configs" (this was also an issue in previous series, please
> use grep to ensure all are fixed)

Fixed.

> 
>> +
>> +  # cat /sys/fs/resctrl/info/L3_MON/counter_configs/mbm_local_bytes/event_filter
> 
> "counter_configs" -> "event_configs"
> 
Sure.

>> +  local_reads,local_non_temporal_writes,local_reads_slow_memory,remote_reads
>> +
>> +This will update all (across all domains of all monitor groups) counter assignments
>> +associated with the mbm_local_bytes event.
>> +
>> +k. Now read the local event again. The first read may come back with "Unavailable"
>> +status. The subsequent read of mbm_local_bytes will display the current value.
>> +::
>> +
>> +  # cat /sys/fs/resctrl/mon_data/mon_L3_00/mbm_local_bytes
>> +  Unavailable
>> +  # cat /sys/fs/resctrl/mon_data/mon_L3_00/mbm_local_bytes
>> +  314101
>> +
>> +l. Users have the option to go back to 'default' mbm_assign_mode if required. This can be
>> +done using the following command. Note that switching the mbm_assign_mode may reset all
>> +the MBM counters (and thus all MBM events) of all the resctrl groups.
>> +::
>> +
>> +  # echo "default" > /sys/fs/resctrl/info/L3_MON/mbm_assign_mode
>> +  # cat /sys/fs/resctrl/info/L3_MON/mbm_assign_mode
>> +  mbm_event
>> +  [default]
>> +
>> +m. Unmount the resctrl
> 
> "Unmount the resctrl" -> "Unmount resctrl" or "Unmount the resctrl filesystem"?

Changed to "Unmount the resctrl filesystem."
> 
>> +::
>> +
>> +  #umount /sys/fs/resctrl/
> 
> "#umount" -> "# umount"
> 

Sure.

>> +
>>  Intel RDT Errata
>>  ================
>>  
>> diff --git a/fs/resctrl/internal.h b/fs/resctrl/internal.h
>> index b42890fd937a..9881674909f2 100644
>> --- a/fs/resctrl/internal.h
>> +++ b/fs/resctrl/internal.h
>> @@ -403,6 +403,8 @@ bool closid_allocated(unsigned int closid);
>>  
>>  int resctrl_find_cleanest_closid(void);
>>  
>> +struct mon_evt *mbm_get_mon_event_by_name(struct rdt_resource *r, char *name);
>> +
>>  int rdtgroup_assign_cntr_event(struct rdt_mon_domain *d, struct rdtgroup *rdtgrp,
>>  			       struct mon_evt *mevt);
>>  void rdtgroup_unassign_cntr_event(struct rdt_mon_domain *d, struct rdtgroup *rdtgrp,
>> diff --git a/fs/resctrl/monitor.c b/fs/resctrl/monitor.c
>> index e30a867c00bb..61419849b257 100644
>> --- a/fs/resctrl/monitor.c
>> +++ b/fs/resctrl/monitor.c
>> @@ -1220,3 +1220,21 @@ void resctrl_update_cntr_allrdtgrp(struct mon_evt *mevt)
>>  			rdtgroup_update_cntr_event(r, crgrp, mevt->evtid);
>>  	}
>>  }
>> +
>> +/*
>> + * mbm_get_mon_event_by_name() - Return the mon_evt entry for the matching
>> + * event name.
>> + */
>> +struct mon_evt *mbm_get_mon_event_by_name(struct rdt_resource *r, char *name)
>> +{
>> +	struct mon_evt *mevt;
>> +
>> +	for_each_mon_event(mevt) {
>> +		if (mevt->rid == r->rid && mevt->enabled &&
>> +		    resctrl_is_mbm_event(mevt->evtid) &&
>> +		    !strcmp(mevt->name, name))
>> +			return mevt;
>> +	}
>> +
>> +	return NULL;
>> +}
>> diff --git a/fs/resctrl/rdtgroup.c b/fs/resctrl/rdtgroup.c
>> index dc108c5db298..b26baca389bb 100644
>> --- a/fs/resctrl/rdtgroup.c
>> +++ b/fs/resctrl/rdtgroup.c
>> @@ -2071,6 +2071,126 @@ static int mbm_L3_assignments_show(struct kernfs_open_file *of, struct seq_file
>>  	return ret;
>>  }
>>  
>> +static int rdtgroup_modify_assign_state(char *assign, struct rdt_mon_domain *d,
> 
> Please move to monitor.c

Sure.

> 
>> +					struct rdtgroup *rdtgrp, struct mon_evt *mevt)
>> +{
>> +	int ret = 0;
>> +
>> +	if (!assign || strlen(assign) != 1)
>> +		return -EINVAL;
>> +
>> +	switch (*assign) {
>> +	case 'e':
>> +		ret = rdtgroup_assign_cntr_event(d, rdtgrp, mevt);
> 
> This patch enables the user to assign counters to multiple events in a
> single write. Looking at this flow:
> rdtgroup_assign_cntr_event()->rdtgroup_alloc_config_cntr() may thus
> be called on the same domain but for different events in a single write from
> user space.
> I thus think that the error message in rdtgroup_alloc_config_cntr()
> should not just contain the domain ID, but also the event *name*
> (not the ID since user space cannot be expect to understand if).
> considering this it seems that @mevt can be passed through
> to rdtgroup_alloc_config_cntr() instead of just the event ID to support
> this?

Yes. Taken care.

> 
>> +		break;
>> +	case '_':
>> +		rdtgroup_unassign_cntr_event(d, rdtgrp, mevt);
>> +		break;
>> +	default:
>> +		ret = -EINVAL;
> 
> Please add a "break".
> 

Sure.

>> +	}
>> +
>> +	return ret;
>> +}
>> +
>> +static int resctrl_process_assign(struct rdt_resource *r, struct rdtgroup *rdtgrp,
> 
> Please move to monitor.c and rename to be more specific. For example, 
> resctrl_parse_mbm_assignment()? Open to ideas.

Sure. Fine with resctrl_parse_mbm_assignment().

> 
>> +				  char *event, char *tok)
>> +{
>> +	struct rdt_mon_domain *d;
>> +	unsigned long dom_id = 0;
>> +	char *dom_str, *id_str;
>> +	struct mon_evt *mevt;
>> +	int ret;
>> +
>> +	mevt = mbm_get_mon_event_by_name(r, event);
>> +	if (!mevt) {
>> +		rdt_last_cmd_printf("Invalid event %s\n", event);
>> +		return  -ENOENT;
>> +	}
>> +
>> +next:
>> +	if (!tok || tok[0] == '\0')
>> +		return 0;
>> +
>> +	/* Start processing the strings for each domain */
>> +	dom_str = strim(strsep(&tok, ";"));
>> +
>> +	id_str = strsep(&dom_str, "=");
>> +
>> +	/* Check for domain id '*' which means all domains */
>> +	if (id_str && *id_str == '*') {
>> +		ret = rdtgroup_modify_assign_state(dom_str, NULL, rdtgrp, mevt);
>> +		if (ret)
>> +			rdt_last_cmd_printf("Assign operation '%s:*=%s' failed\n",
>> +					    event, dom_str);
>> +		return ret;
>> +	} else if (!id_str || kstrtoul(id_str, 10, &dom_id)) {
>> +		rdt_last_cmd_puts("Missing domain id\n");
>> +		return -EINVAL;
>> +	}
>> +
>> +	/* Verify if the dom_id is valid */
>> +	list_for_each_entry(d, &r->mon_domains, hdr.list) {
>> +		if (d->hdr.id == dom_id) {
>> +			ret = rdtgroup_modify_assign_state(dom_str, d, rdtgrp, mevt);
>> +			if (ret) {
>> +				rdt_last_cmd_printf("Assign operation '%s:%ld=%s' failed\n",
>> +						    event, dom_id, dom_str);
>> +				return ret;
>> +			}
>> +			goto next;
>> +		}
>> +	}
>> +
>> +	rdt_last_cmd_printf("Invalid domain id %ld\n", dom_id);
>> +	return -EINVAL;
>> +}
>> +
> Reinette
> 

-- 
Thanks
Babu Moger