hulk inclusion
category: bugfix
bugzilla: 48265
CVE: NA

--------------------------------

This function is called only when we mount resctrl sysfs, for error
handling we need to destroy schemata list when next few steps failed
after creation of schemata list.

Fixes: 7e9b5caeefff ("arm64/mpam: resctrl: Add helpers for init and destroy schemata list")
Signed-off-by: NWang ShaoBo <bobo.shaobowang@huawei.com>
Reviewed-by: NCheng Jian <cj.chengjian@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

hulk inclusion
category: bugfix
bugzilla: 48265
CVE: NA

--------------------------------

Use fs_context to parse mount options, this old process parsing from
parse_rdtgroupfs_options() will be obsoleted and removed.
Signed-off-by: NWang ShaoBo <bobo.shaobowang@huawei.com>
Reviewed-by: NCheng Jian <cj.chengjian@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

hulk inclusion
category: feature
feature: ARM MPAM support
bugzilla: 48265
CVE: NA

--------------------------------

Based on 61fa56e1dd8a ("arm64/mpam: Add resctrl_ctrl_feature structure to manage
ctrl features"), we add several ctrl features and supply corresponding
mount options, including mbPbm, mbMax, mbMin, mbPrio, caMax, caPrio, caPbm,
if MPAM system supports relevant features, we can mount resctrl like this:

e.g.
   > mount -t resctrl resctrl /sys/fs/resctrl -o mbMax,mbMin,caPrio
   > cd /sys/fs/resctrl && cat schemata
     L3:0=0x7fff;1=0x7fff;2=0x7fff;3=0x7fff #default select cpbm as basic ctrl feature
     L3PRI:0=3;1=3;2=3;3=3
     MBMAX:0=100;1=100;2=100;3=100
     MBMIN:0=0;1=0;2=0;3=0

   > mount -t resctrl resctrl /sys/fs/resctrl
   > cd /sys/fs/resctrl && cat schemata
     L3:0=0x7fff;1=0x7fff;2=0x7fff;3=0x7fff #default select cpbm as basic ctrl feature
     MB:0=100;1=100;2=100;3=100  #default select mbw max as basic ctrl feature

   > mount -t resctrl resctrl /sys/fs/resctrl -o caMax
   > cd /sys/fs/resctrl && cat schemata
     L3:0=33554432;1=33554432;2=33554432;3=33554432 #use cmax ctrl feature
     MB:0=100;1=100;2=100;3=100  #default select mbw max as basic ctrl feature

For Cache MSCs, basic ctrl features include cmax(Cache Maximum Capacity)
and cpbm(Cache protion bitmap) partition, if mount options are not specified,
default cpbm will be selected.

For Memory MSCs, basic ctrl features include max(Memory Bandwidth Maximum)
and pbm(Memory Bandwidth Portion Bitmap) partition, if mount options are
not specified, default max will be selected.

Above mount options also can be used accompany with cdp options.

e.g.
   > mount -t resctrl resctrl /sys/fs/resctrl -o caMax,caPrio,cdpl3
   > cd /sys/fs/resctrl && cat schemata
     L3CODE:0=33554432;1=33554432;2=33554432;3=33554432 #code use cmax ctrl feature
     L3DATA:0=33554432;1=33554432;2=33554432;3=33554432 #data use cmax ctrl feature
     L3CODEPRI:0=3;1=3;2=3;3=3 #code use intpriority ctrl feature
     L3DATAPRI:0=3;1=3;2=3;3=3 #data use intpriority ctrl feature
     MB:0=100;1=100;2=100;3=100  #default select mbw max as basic ctrl feature

By combining these mount parameters can we use MPAM more powerfully.
Signed-off-by: NWang ShaoBo <bobo.shaobowang@huawei.com>
Reviewed-by: NXiongfeng Wang <wangxiongfeng2@huawei.com>
Reviewed-by: NCheng Jian <cj.chengjian@huawei.com>
Signed-off-by: NYang Yingliang <yangyingliang@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

hulk inclusion
category: feature
feature: ARM MPAM support
bugzilla: 48265
CVE: NA

--------------------------------

Sometimes monitoring will have such anomalies:

e.g.
    > cd /sys/fs/resctrl/ && grep . mon_data/*
      mon_data/mon_L3CODE_00:14336
      mon_data/mon_L3CODE_01:344064
      mon_data/mon_L3CODE_02:2048
      mon_data/mon_L3CODE_03:27648
      mon_data/mon_L3DATA_00:0  #L3DATA's monitoring data always be 0
      mon_data/mon_L3DATA_01:0
      mon_data/mon_L3DATA_02:0
      mon_data/mon_L3DATA_03:0
      mon_data/mon_MB_00:392
      mon_data/mon_MB_01:552
      mon_data/mon_MB_02:160
      mon_data/mon_MB_03:0

If cdp on, tasks in resctrl default group with closid=0 and rmid=0 don't
know how to fill proper partid_i/pmg_i and partid_d/pmg_d into MPAMx_ELx
sysregs by mpam_sched_in() called by __switch_to(), it's because current
cpu's default closid and rmid are also equal to 0 and to make the operation
modifying configuration passed.

Update per cpu default closid of none-zero value, call update_closid_rmid()
to update each cpu's mpam proper MPAMx_ELx sysregs for setting partid and
pmg when mounting resctrl sysfs, it looks like a practical method.
Signed-off-by: NWang ShaoBo <bobo.shaobowang@huawei.com>
Reviewed-by: NXiongfeng Wang <wangxiongfeng2@huawei.com>
Reviewed-by: NCheng Jian <cj.chengjian@huawei.com>
Signed-off-by: NYang Yingliang <yangyingliang@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

hulk inclusion
category: feature
feature: ARM MPAM support
bugzilla: 48265
CVE: NA

--------------------------------

MPAM includes partid, pmg, monitor, all of these we collectively call
mpam id, if cdp on, we would allocate a new mpamid_new which equals to
mpamid + 1, and at some places mpamid may not need to be encapsulated
into struct { u16 val; } for simplicity, So we use a simpler macro
resctrl_cdp_mpamid_map_val() to complete this cdp mapping process.
Signed-off-by: NWang ShaoBo <bobo.shaobowang@huawei.com>
Reviewed-by: NXiongfeng Wang <wangxiongfeng2@huawei.com>
Reviewed-by: NCheng Jian <cj.chengjian@huawei.com>
Signed-off-by: NYang Yingliang <yangyingliang@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

hulk inclusion
category: feature
feature: ARM MPAM support
bugzilla: 48265
CVE: NA

--------------------------------

ctrl_features array, introduced by 61fa56e1dd8a ("arm64/mpam: Add
resctrl_ctrl_feature structure to manage ctrl features"), which lives
in raw_resctrl_resource structure for listing ctrl features's type do
we support in total for this resource, this filters illegal parameters
outside from mount options and provides useful info for add_schema()
for registering a new control type node in schema list.

This action helps us to add new ctrl feature easier later.
Signed-off-by: NWang ShaoBo <bobo.shaobowang@huawei.com>
Reviewed-by: NXiongfeng Wang <wangxiongfeng2@huawei.com>
Reviewed-by: NCheng Jian <cj.chengjian@huawei.com>
Signed-off-by: NYang Yingliang <yangyingliang@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

hulk inclusion
category: feature
feature: ARM MPAM support
bugzilla: 48265
CVE: NA

--------------------------------

rmid is used to mark each resctrl group for monitoring, anyhow, also
following corresponding resctrl group's configuration, we export rmid
sysfile to resctrl sysfs for any usage elsewhere such as SMMU io, user
can get rmid from a resctrl group and set this rmid to a target io
through SMMU driver if SMMU MPAM implemented, so make related io devices
can be monitored or accomplish aimed configuration for resource's usage.
Signed-off-by: NWang ShaoBo <bobo.shaobowang@huawei.com>
Reviewed-by: NXiongfeng Wang <wangxiongfeng2@huawei.com>
Reviewed-by: NCheng Jian <cj.chengjian@huawei.com>
Signed-off-by: NYang Yingliang <yangyingliang@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

hulk inclusion
category: feature
feature: ARM MPAM support
bugzilla: 48265
CVE: NA

--------------------------------

So far there are some declarations shared by resctrlfs.c and mpam
core module files under kernel/mpam directory scattered in mpam.h
and resctrl.h, this is organized like this:

-- asm/
   +-- resctrl.h        +
   +-- mpam.h           |    +
   +-- mpam_resource.h  |    |    +
                        |    |    |
-- fs/                  |    |    +-> mpam/
   +-- resctrlfs.c <----+----+------> +-- mpam_resctrl.c ...

We move this declarations shared by resctrlfs.c and mpam/ to resctrl.h
and split another declarations into mpam_internal.h, also including
moving mpam_resource.h to mpam/ directory, currently this is organized
like this:

-- asm/
   +-- mpam.h           +----> export to other modules(e.g. SMMU master io)
   +-- resctrl.h        +
                        |
-- mpam/                |
   +-- mpam_internal.h  |    +
   +-- mpam_resource.h  |    |    +
                        |    |    |
-- fs/                  |    +----+-> mpam/
   +-- resctrlfs.c <----+-----------> +-- mpam_resctrl.c ...

In this way can we build a clearer framework for MPAM usage.
Signed-off-by: NWang ShaoBo <bobo.shaobowang@huawei.com>
Reviewed-by: NXiongfeng Wang <wangxiongfeng2@huawei.com>
Reviewed-by: NCheng Jian <cj.chengjian@huawei.com>
Signed-off-by: NYang Yingliang <yangyingliang@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

hulk inclusion
category: feature
feature: ARM MPAM support
bugzilla: 48265
CVE: NA

--------------------------------

Some resource's properities such as closid and rmid are exported like
Intel-RDT in our resctrl design, but there also has two main differences,
one is MB(Memory Bandwidth), for we MB is also divided into two directories
MB and MB_MON to show respective properties about control and monitor type
as same as LxCache, another is we adopt features sysfile under resources'
directories, which indicates the properties of control type of corresponding
resource, for instance MB hardlimit.

e.g.
    > mount -t resctrl resctrl /sys/fs/resctrl -o mbHdl
    > cd /sys/fs/resctrl/ && cat info/MB/features
      mbHdl@1  #indicate MBHDL setting's upper bound is 1
    > cat schemata
      L3:0=7fff;1=7fff;2=7fff;3=7fff
      MB:0=100;1=100;2=100;3=100
      MBHDL:0=1;1=1;2=1;3=1
Signed-off-by: NWang ShaoBo <bobo.shaobowang@huawei.com>
Reviewed-by: NXiongfeng Wang <wangxiongfeng2@huawei.com>
Reviewed-by: NCheng Jian <cj.chengjian@huawei.com>
Signed-off-by: NYang Yingliang <yangyingliang@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

hulk inclusion
category: feature
feature: ARM MPAM support
bugzilla: 48265
CVE: NA

--------------------------------

Structure resctrl_ctrl_feature taken by resources is introduced to manage
ctrl features, of which characteristic like max width from outer input
and the base we parse from.

Now it is more practical for declaring a new ctrl feature, such as SCHEMA_PRI
feature, only associated with internal priority setting exported by mpam
devices, where informations is collected from mpam_resctrl_resource_init(),
and next be chosen open or close by user options.

ctrl_ctrl_feature structure contains a flags field to avoid duplicated
control type, for instance, SCHEMA_COMM feature selectes cpbm (Cache
portion bitmap) as resource Cache default control type, so we should not
enable this feature no longer if user manually selectes cpbm control
type through mount options.

This field evt in ctrl_ctrl_feature structure is enum rdt_event_id type
variable which works like eee4ad2a36e6 ("arm64/mpam: Add hook-events id
for ctrl features") illustrates.
Signed-off-by: NWang ShaoBo <bobo.shaobowang@huawei.com>
Reviewed-by: NXiongfeng Wang <wangxiongfeng2@huawei.com>
Reviewed-by: NCheng Jian <cj.chengjian@huawei.com>
Signed-off-by: NYang Yingliang <yangyingliang@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

hulk inclusion
category: feature
feature: ARM MPAM support
bugzilla: 48265
CVE: NA

--------------------------------

For MPAM, a rmid can do monitoring work only with a monitor resource
allocated, we adopt a mechanism for monitor resource dynamic allocation
and recycling, it is different from Intel-RDT operation who creates a
kworker thread for dynamically monitoring Cache usage and checks if it
is below a threshold adjustable for rmid free, for we have detected that
this method will affect the cpu utilization in many cases, sometimes this
influence cannot be accepted.

Our method is simple, as different resource's monitor number varies, we
deliever two list, one for storing rmids which has exclusive monitor
resource and another for storing this rmids which have monitor resource
shared, this shared monitor id always be 0. it works like this, if a new
rmid apply for a resource monitor which is in used, then we put this rmid
to the tail of latter list and temporarily give a default monitor id 0
util someone releases available monitor resource, if this new rmid has
all resources' monitor resource needed, then it will be put into exclusive
list.

This implements the LRU allocation of monitor resources and give users
part control rights of allocation and release, if resctrl group's quantity
can be guaranteed or user don't need monitoring too many groups
synchronously, this is a more appropriate way for user deployment, not
only that, also can it avoid the risk of inaccuracy in monitoring when
monitoring operation happen to too many groups at the same time.
Signed-off-by: NWang ShaoBo <bobo.shaobowang@huawei.com>
Reviewed-by: NXiongfeng Wang <wangxiongfeng2@huawei.com>
Reviewed-by: NCheng Jian <cj.chengjian@huawei.com>
Signed-off-by: NYang Yingliang <yangyingliang@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

hulk inclusion
category: feature
feature: ARM MPAM support
bugzilla: 48265
CVE: NA

--------------------------------

So far we use sd_closid, including {reqpartid, intpartid}, to label each
resctrl group including ctrlgroup and mongroup, This can perfectly handle
this case where number of reqpartid exceeds intpartid, this always happen
when intpartid narrowing supported, otherwise their two are of same number.
So we use excessive reqpartid to indicate (1)- how configurations can be
synchronized from the configuration indexed by intpartid, not only that,
(2)- take part of monitor role.

But reqpartid in (2) with pmg still be scattered, So far we have not yet
a right way to explain how can we use their two properly. In order to
ensure their resources can be fully utilized, and given this idea from
Intel-RDT's design which uses rmid for monitoring, a rmid remap matrix is
delivered for transforming partid and pmg to rmid, this matrix is organized
like this:

                 [bitmap entry indexed by partid]
                       [col pos is partid]

                     [0]  [1]  [2]  [3]  [4]  [5]
   occ->bitmap[:0]    1    0    0    1    1    1
        bitmap[:1]    1    0    0    1    1    1
        bitmap[:2]    1    1    1    1    1    1
        bitmap[:3]    1    1    1    1    1    1
[row pos-1 is pmg]

Calculate rmid = partid + NR_partid * pmg

occ represents if this bitmap has been used by a partid, it is because
a certain partid should not be accompany with a duplicated pmg for
monitoring, this design easily saves a lot of space, and can also decrease
time complexity of allocating and free rmid process from O(NR_partid)*
O(NR_pmg) to O(NR_partid) + O(log(NR_pmg)) compared with using list.

By this way, we get a continuous rmid set with upper bound(NR_pmg *
NR_partid - 1), given an rmid we can assume that if it's a valid rmid
by judging whether it falls within this range or not.

rmid implicts the reqpartid info, so we can use relevant helpers to get
this reqpartid for sd_closid@reqpartid and perfectly accomplish this
configuration sync mission, this also makes closid simpler which can be
consists of intpartid index only, also each resctrl group is happy to own
consecutive rmid.

This also has some profound influences, for instance for MPAM there
also support SMMU io using partid and pmg, we can use a single helper
mpam_rmid_to_partid_pmg() in SMMU driver to complete this remap process
for rmid input from outside user space.
Signed-off-by: NWang ShaoBo <bobo.shaobowang@huawei.com>
Reviewed-by: NXiongfeng Wang <wangxiongfeng2@huawei.com>
Reviewed-by: NCheng Jian <cj.chengjian@huawei.com>
Signed-off-by: NYang Yingliang <yangyingliang@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

hulk inclusion
category: feature
feature: ARM MPAM support
bugzilla: 48265
CVE: NA

--------------------------------

There are two aspects involved:

  - Getting configuration

    We divide event QOS_XX_PRI_EVENT_ID into QOS_XX_INTPRI_EVENT_ID and
    QOS_XX_DSPRI_EVENT_ID, in spite of having attempted to set same value
    of filling dspri and intpti in mpam_config structure but exactly we
    need read seperately to ensure their independence.

    Besides, an event such as QOS_CAT_INTPRI_EVENT_ID is not necessary to
    be read from MSC's register but set to be 0 directly if corresponding
    feature doesn't support.

  - Applying configuration

    When applying downstream or internal priority configuration, given
    the independence of their two, we should check if feature mpam_feat_
    xxpri_part supported first and next check mpam_feat_xxpri_part_0_low,
    and convert dspri and intpri into a proper value according to it's max
    width.
Signed-off-by: NWang ShaoBo <bobo.shaobowang@huawei.com>
Reviewed-by: NXiongfeng Wang <wangxiongfeng2@huawei.com>
Reviewed-by: NCheng Jian <cj.chengjian@huawei.com>
Signed-off-by: NYang Yingliang <yangyingliang@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

hulk inclusion
category: feature
feature: ARM MPAM support
bugzilla: 48265
CVE: NA

--------------------------------

The MPAM MSC error interrupt tells us how we misconfigured the MSC.
We don't expect to to this. If the interrupt fires, print a
summary, and mark MPAM as broken. Eventually we will try and cleanly
teardown when we see this.

Now we can register from a helper mpam_register_device_irq() to
register overflow and error interrupt from mpam device, When devices
come and go we want to make sure the error irq is enabled. We disable
the error irq when cpus are taken offline in case the component remains
online even when the associated CPUs are offline.

Code of this patch are borrowed from james <james.morse@arm.com>.

[Wang ShaoBo: few version adaptation changes]
Signed-off-by: NJames Morse <james.morse@arm.com>
Link: http://www.linux-arm.org/git?p=linux-jm.git;a=patch;h=6d1ceca3eb5953fc16a524c9aad933519aa3f64c
Link: http://www.linux-arm.org/git?p=linux-jm.git;a=patch;h=81d178c198165fd557431d6879135d2e03ea92c0Signed-off-by: NWang ShaoBo <bobo.shaobowang@huawei.com>
Reviewed-by: NXiongfeng Wang <wangxiongfeng2@huawei.com>
Reviewed-by: NCheng Jian <cj.chengjian@huawei.com>
Signed-off-by: NYang Yingliang <yangyingliang@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

hulk inclusion
category: feature
feature: ARM MPAM support
bugzilla: 48265
CVE: NA

--------------------------------

MPAM spec says, when an MPAMCFG register other than MPAMCFG_INTPARTID
is read or written, if the value of MPAMCFG_PART_SEL.INTERNAL is not 1,
MPAMF_ESR is set to indicate an intPARTID_Range error. So we should
set MPAMCFG_PART_SEL.INTERNAL to 1 before reading MPAMCFG_PRI register.
Signed-off-by: NWang ShaoBo <bobo.shaobowang@huawei.com>
Reviewed-by: NXiongfeng Wang <wangxiongfeng2@huawei.com>
Reviewed-by: NCheng Jian <cj.chengjian@huawei.com>
Signed-off-by: NYang Yingliang <yangyingliang@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

hulk inclusion
category: feature
feature: ARM MPAM support
bugzilla: 48265
CVE: NA

--------------------------------

If cdp enabled, LxCODE and LxDATA are assigned two different partid
each occupies a monitor, but because not all features use cdp mode,
for instance MB(Memory Bandwidth), we should make sure this two partid/
monitor be operated simultaneously for display.

e.g.

    +- code stream (partid = 0, monitor = 0)   ----+--->  L3CODE
cpu-+                                              +
    +- data stream (partid = 1, monitor = 1)   ----+--->  L3DATA
                                                   |
                                                   +--->  MB
Signed-off-by: NWang ShaoBo <bobo.shaobowang@huawei.com>
Reviewed-by: NXiongfeng Wang <wangxiongfeng2@huawei.com>
Reviewed-by: NCheng Jian <cj.chengjian@huawei.com>
Signed-off-by: NYang Yingliang <yangyingliang@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

hulk inclusion
category: feature
feature: ARM MPAM support
bugzilla: 48265
CVE: NA

--------------------------------

Reading/Writing registers directly for getting or putting configuration
is not friendly with expansion and legibility, multiple types of schemata
ctrls is supported, of which value should be converted to a proper value
based on specific definition and range in corresponding register according
to MPAM spec, Using event id instead to indicate which type configuration
we want to get looks easier for us.

Besides, different hook-events have different setting bound such as bwa_wd
for adaptive range conversion when writing configuration, this can be
associated with specific event for conversion.
Signed-off-by: NWang ShaoBo <bobo.shaobowang@huawei.com>
Reviewed-by: NXiongfeng Wang <wangxiongfeng2@huawei.com>
Reviewed-by: NCheng Jian <cj.chengjian@huawei.com>
Signed-off-by: NYang Yingliang <yangyingliang@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

hulk inclusion
category: feature
feature: ARM MPAM support
bugzilla: 48265
CVE: NA

--------------------------------

MPAMCFG_INTPARTID.INTERNAL must be set when narrowing reqpartid to
intpartid according to MPAM spec definitions, and this action must be
done before writing MPAMCFG_PART_SEL if narrowing implemented. So we
plan this work that do narrowing unifiedly when narrowing is supported.
Signed-off-by: NWang ShaoBo <bobo.shaobowang@huawei.com>
Reviewed-by: NXiongfeng Wang <wangxiongfeng2@huawei.com>
Reviewed-by: NCheng Jian <cj.chengjian@huawei.com>
Signed-off-by: NYang Yingliang <yangyingliang@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

hulk inclusion
category: feature
feature: ARM MPAM support
bugzilla: 48265
CVE: NA

--------------------------------

Use an array for storing extend ctrls' max width, on purpose, checking
each input value from schemata.

Note the useful value of each ctrls' max width is at least 1, 0 means
meaningless, and greater than 1 means the choices can be selected.
Signed-off-by: NWang ShaoBo <bobo.shaobowang@huawei.com>
Reviewed-by: NXiongfeng Wang <wangxiongfeng2@huawei.com>
Reviewed-by: NCheng Jian <cj.chengjian@huawei.com>
Signed-off-by: NYang Yingliang <yangyingliang@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

hulk inclusion
category: feature
feature: ARM MPAM support
bugzilla: 48265
CVE: NA

--------------------------------

Store default priority in mpam class structure from reading devices'
intpri_wd and dspri_wd.

intpri_wd and dspri_wd represent the number of implemented bits in the
internal/downstream priority field in MPAMCFG_PRI, when INTPRI_0_IS_LOW
/DSPRI_0_IS_LOW is not set, we need to rotate input priority(higher value
higher priority) from user space to target priority (higher value lower
priority) and this is restricted by implemented bits.
Signed-off-by: NWang ShaoBo <bobo.shaobowang@huawei.com>
Reviewed-by: NXiongfeng Wang <wangxiongfeng2@huawei.com>
Reviewed-by: NCheng Jian <cj.chengjian@huawei.com>
Signed-off-by: NYang Yingliang <yangyingliang@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

hulk inclusion
category: feature
feature: ARM MPAM support
bugzilla: 48265
CVE: NA

--------------------------------

Register MPAMCFG_PRI's default value is also used for software default
usage after probing resources, two fields hwdef_intpri and hwdef_dspri are
placed into mpam_device structure to store the default priority setting.
Signed-off-by: NWang ShaoBo <bobo.shaobowang@huawei.com>
Reviewed-by: NXiongfeng Wang <wangxiongfeng2@huawei.com>
Reviewed-by: NCheng Jian <cj.chengjian@huawei.com>
Signed-off-by: NYang Yingliang <yangyingliang@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

hulk inclusion
category: feature
feature: ARM MPAM support
bugzilla: 48265
CVE: NA

--------------------------------

Currently configuration control type is devided into three classes: COMMON,
PRIORITY and HARDLIMIT, capacities' features for mount options is stored
in ctrl_extend_bits field live in resctrl resource structure to figure
out which configuration type is allowed to apply, when writing schemata
sysfile, all related configurations' content from corresponding configuration
array will be updated and applied once time.

we can set configuration like this:
  e.g.
     > mount -t resctrl resctrl /sys/fs/resctrl && cd /sys/fs/resctrl -o hardlimit
     > cat schemata
       L3:0=7fff;1=7fff;2=7fff;3=7fff
       MB:0=100;1=100;2=100;3=100
       MBHDL:0=1;1=1;2=1;3=1
     > echo 'MB:0=10' > schemata && echo 'MBHDL:0=0' > schemata # no hardlimit

This also deletes opt_list no longer needed that used as organizing
different control types, now we can check supports from ctrl_extend_bits
and do extended control-type works by schema list.
Signed-off-by: NWang ShaoBo <bobo.shaobowang@huawei.com>
Reviewed-by: NXiongfeng Wang <wangxiongfeng2@huawei.com>
Reviewed-by: NCheng Jian <cj.chengjian@huawei.com>
Signed-off-by: NYang Yingliang <yangyingliang@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

hulk inclusion
category: feature
feature: ARM MPAM support
bugzilla: 48265
CVE: NA

--------------------------------

So far each mon group is tagged by sd_closid, we can monitor data by
switching cpus' sd_closid.reqpartid and pmg, and ensuring consistent
configuration for mon groups by following it's parent ctrl group through
sd_closid.intpartid.

Most of this code is borrowed from Intel-RDT.
Signed-off-by: NWang ShaoBo <bobo.shaobowang@huawei.com>
Reviewed-by: NXiongfeng Wang <wangxiongfeng2@huawei.com>
Reviewed-by: NCheng Jian <cj.chengjian@huawei.com>
Signed-off-by: NYang Yingliang <yangyingliang@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

hulk inclusion
category: feature
feature: ARM MPAM support
bugzilla: 48265
CVE: NA

--------------------------------

For each ctrl group, it's monitoring data should include all it's child
mon groups' monitoring data, these code is borrowed from Intel-RDT for
facilitating users to configure different monitoring strategies.
Signed-off-by: NWang ShaoBo <bobo.shaobowang@huawei.com>
Reviewed-by: NXiongfeng Wang <wangxiongfeng2@huawei.com>
Reviewed-by: NCheng Jian <cj.chengjian@huawei.com>
Signed-off-by: NYang Yingliang <yangyingliang@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

hulk inclusion
category: feature
feature: ARM MPAM support
bugzilla: 48265
CVE: NA

--------------------------------

Currently we use partid and pmg (Performance Monitoring Group) to filter
some performance events so that the performance of a particular partid
and pmg can be monitored, but pmg looks useless except for making a
filter with partid, especially when pmg varies in different MPAM resources,
it makes difficult to allocate pmg resource when creating new mon group
in resctrl sysfs, even causes a lot of waste.

So we use a software-defined sd_closid instead of 32-bit integer to label
each rdtgroup (including mon group), sd_closid include intpartid for
allocation and reqpartid for synchronizing configuration and monitoring,
Given MPAM has narrowing feature, also includes the concept (hw_reqpartid,
hw_intpartid we named), when narrowing is not supported, number of intpartid
and reqpartid equals to hw_reqpartid, otherwise intpartid and reqpartid is
related to minimum number of both hw_reqpartid and hw_intpartid supported
across different resources, by using this way, not only we solve above
problem but also use relax reqpartid for creating new mon group. additionally,
pmg is also preferred when it is available.

e.g.
     hw_intpartid:  0  1  2  3  4  5  6  7
     hw_reqpartid:  0  1  2  3  4  5  6  7  8  9  10 11 12 13 14 15

                    |  |  |  |  |  |  |  |  |  |  |  |  |  |  |  |
                                            |  |  |  |  |  |  |  |
resctrl ctrl group: p0 p1 p2 p3 p4 p5 p6 p7 |  |  |  |  |  |  |  |
                          |     |
resctrl mon group:        |     +-----------------------m4 m5 m6 m7
                          +-----------------m0 m1 m2 m3
                          In this case, use extra reqpartid to create
                          m0, m1, m2, m3 mon group for p2 ctrl group,
                          and m4, m5, m6, m7 for p4.

As we know reqpartid both supports allocating and monitoring filter, we
should synchronize config of ctrl group with child mon groups under this
design, each mon group's configuration indexed by a reqpartid that called
slave is closely following it's father ctrl group that called master
whenever configuration changes. not only that, we let task_struct keep
both intpartid and reqpartid so we can know if tasks belong to a same
ctrl group through intpartid and change cpu's partid by writing MPAMx_ELx
through reqpartid when tasks switching.
Signed-off-by: NWang ShaoBo <bobo.shaobowang@huawei.com>
Reviewed-by: NXiongfeng Wang <wangxiongfeng2@huawei.com>
Reviewed-by: NCheng Jian <cj.chengjian@huawei.com>
Signed-off-by: NYang Yingliang <yangyingliang@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

hulk inclusion
category: feature
feature: ARM MPAM support
bugzilla: 48265
CVE: NA

--------------------------------

According to Arm MPAM spec definition, register MPAMCFG_PART_SEL's high
16 bit should be set to 0 and MPAMCFG_INTPARTID's high 16 bit should be
set to 1 when establishing intpartid association, and we should use
intpartid in MPAMCFG_PART_SEL instead of reqpartid as long as intpartid
narrowing is implemented.
Signed-off-by: NWang ShaoBo <bobo.shaobowang@huawei.com>
Reviewed-by: NXiongfeng Wang <wangxiongfeng2@huawei.com>
Reviewed-by: NCheng Jian <cj.chengjian@huawei.com>
Signed-off-by: NYang Yingliang <yangyingliang@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

hulk inclusion
category: feature
feature: ARM MPAM support
bugzilla: 48265
CVE: NA

--------------------------------

Code in resctrlfs.c is not shared with x86 RDT currently, but may be
updated to support both in the future, so remove unrelated CONFIG for
now to make code clearer.
Signed-off-by: NWang ShaoBo <bobo.shaobowang@huawei.com>
Reviewed-by: NXiongfeng Wang <wangxiongfeng2@huawei.com>
Reviewed-by: NCheng Jian <cj.chengjian@huawei.com>
Signed-off-by: NYang Yingliang <yangyingliang@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

hulk inclusion
category: feature
feature: ARM MPAM support
bugzilla: 48265
CVE: NA

--------------------------------

We redesign monitoring process for user, as following illustrates:

  e.g. before rewriting:
     mount /sys/fs/resctrl && cd /sys/fs/resctrl
	 mkdir p1 && cd p1
     echo 1 > ctrlmon  # this allocates a monitor resource for this group
     ...               # associating task/cpu with this group
     grep . mon_data/* # get monitor data from mon_data directory
  e.g. after rewriting:
	 mount /sys/fs/resctrl && cd /sys/fs/resctrl
 	 mkdir p1 && cd p1 # automically allocating a monitoring resource
     ...               # associate task/cpu with this group
     grep . mon_data/* # directly get monitor data

ctrlmon is used for manually allocating a monitor resource for monitoring
a specified group (labeled by partid and pmg), we delete ctrlmon because
this action is redundant.

User should know which group has been allocated a available monitor resource
and only this monitor resource is released then this monitor resource can
be reallocated to a new group after, this action is redundant and unnecessary,
as monitor resource is used only when monitoring process happens, so a
relax monitor resource can be allocated to multiple groups and take effect
when monitoring process happened.

But should some restrictions be known, a monitor resource for monitoring
Cache-occupancy might be kept for a long time until it doesn't need to
be use anymore, or below a threshold as like intel-RDT limbo list works,
otherwise you may see that the monitoring result is very small beyond
exception when you force switch one mon resource from one group to another.

We deliver a simple LRU mon resource allocation mechanism, but so far
it just assign a monitor according to the order in which groups was created,
this is incomplete and needs subsequent improvement.
Signed-off-by: NWang ShaoBo <bobo.shaobowang@huawei.com>
Reviewed-by: NXiongfeng Wang <wangxiongfeng2@huawei.com>
Reviewed-by: NCheng Jian <cj.chengjian@huawei.com>
Signed-off-by: NYang Yingliang <yangyingliang@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

hulk inclusion
category: feature
feature: ARM MPAM support
bugzilla: 48265
CVE: NA

--------------------------------

Rearrange helpers' declaration place for resctrlfs and clean up
header files included, this make code more clear.
Signed-off-by: NWang ShaoBo <bobo.shaobowang@huawei.com>
Reviewed-by: NXiongfeng Wang <wangxiongfeng2@huawei.com>
Reviewed-by: NCheng Jian <cj.chengjian@huawei.com>
Signed-off-by: NYang Yingliang <yangyingliang@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

hulk inclusion
category: feature
feature: ARM MPAM support
bugzilla: 48265
CVE: NA

--------------------------------

commit 43be0de7be8f ("arm64/mpam: Support cdp on allocating monitors")
allows us to allocate two monitor once, we apply this two monitors to
different monitor sysfile under mon_data directory according to its'
closid, as following illustrates.

-- resctrl/
         +-- schemata
             L3CODE:0=xx  # closid
             L3DATA:1=xx  # closid+1
             MB:0=xx      # closid
         +-- mon_data/
                    +-- mon_L3CODE_00 # monitor
                    +-- mon_L3DATA_00 # monitor+1
                    +-- mon_MB_00     # monitor

When monitoring happens, we read the private data of each monitor
sysfile which contains closid, monitor and pmg, this is used for
obtaining monitor data.
Signed-off-by: NWang ShaoBo <bobo.shaobowang@huawei.com>
Reviewed-by: NXiongfeng Wang <wangxiongfeng2@huawei.com>
Reviewed-by: NCheng Jian <cj.chengjian@huawei.com>
Signed-off-by: NYang Yingliang <yangyingliang@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

hulk inclusion
category: feature
feature: ARM MPAM support
bugzilla: 48265
CVE: NA

--------------------------------

This prepares for simultaneously monitoring LxDATA and LxCODE when cdp
is open, under our implementation, LxDATA and LxCODE is allocated closid
and closid+1, so we should keep two monitor once time for each.

Why there needs one monitors for each closid when cdp is open, but not
switch one between the two LxDATA and LxCODE is because this monitor kept
by target closid maybe busy for a long time, it would cause inaccuracy if
we force switching.
Signed-off-by: NWang ShaoBo <bobo.shaobowang@huawei.com>
Reviewed-by: NXiongfeng Wang <wangxiongfeng2@huawei.com>
Reviewed-by: NCheng Jian <cj.chengjian@huawei.com>
Signed-off-by: NYang Yingliang <yangyingliang@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

hulk inclusion
category: feature
feature: ARM MPAM support
bugzilla: 48265
CVE: NA

--------------------------------

Resctrl ctrlmon write/read functions should be moved to mpam_ctrlmon.c
to make code clear.
Signed-off-by: NWang ShaoBo <bobo.shaobowang@huawei.com>
Reviewed-by: NXiongfeng Wang <wangxiongfeng2@huawei.com>
Reviewed-by: NCheng Jian <cj.chengjian@huawei.com>
Signed-off-by: NYang Yingliang <yangyingliang@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

hulk inclusion
category: feature
feature: ARM MPAM support
bugzilla: 48265
CVE: NA

--------------------------------

Replace u32 bitmask with bitmap for closid allocation, it's because closid
may be too large to use 32 bits.

This also support cdp, when cdp is enabled, closid will be assigned twice
once time, giving closid to code LxCache and closid+1 to data LxDATA, so
do free process.
Signed-off-by: NWang ShaoBo <bobo.shaobowang@huawei.com>
Reviewed-by: NXiongfeng Wang <wangxiongfeng2@huawei.com>
Reviewed-by: NCheng Jian <cj.chengjian@huawei.com>
Signed-off-by: NYang Yingliang <yangyingliang@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

hulk inclusion
category: feature
feature: ARM MPAM support
bugzilla: 48265
CVE: NA

--------------------------------

resctrl_resource_reset() would be excuted if resctrl sysfs umount, this
help to reset all settings stored in related structures (such as mpam_cfg)
and put MSCs back to default state.

This is similar to 6ab0b81f2c18 ("arm64/mpam: Fix unreset resources when mkdir
ctrl group or umount resctrl") but using helpers from mpam devices module.
Signed-off-by: NWang ShaoBo <bobo.shaobowang@huawei.com>
Reviewed-by: NXiongfeng Wang <wangxiongfeng2@huawei.com>
Reviewed-by: NCheng Jian <cj.chengjian@huawei.com>
Signed-off-by: NYang Yingliang <yangyingliang@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

hulk inclusion
category: feature
feature: ARM MPAM support
bugzilla: 48265
CVE: NA

--------------------------------

For MPAM, Processing elements (PEs) issue memory-system requests, PEs must
implement the MPAMn_ELx registers and their behaviors to generate the
PARTID and PMG fields of memory-system requests.

So far schemata supports cdp writing and reading, to grab MPAM info from
cpu for downstream MSCs, SYS_MPAMx_ELx registers should be filled in both
partid_d and partid_i (mapped from closids) of LxDATA and LxCODE and pmg_d
and pmg_i (mapped from rmid).
Signed-off-by: NWang ShaoBo <bobo.shaobowang@huawei.com>
Reviewed-by: NXiongfeng Wang <wangxiongfeng2@huawei.com>
Reviewed-by: NCheng Jian <cj.chengjian@huawei.com>
Signed-off-by: NYang Yingliang <yangyingliang@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

hulk inclusion
category: feature
feature: ARM MPAM support
bugzilla: 48265
CVE: NA

--------------------------------

schemata labels each row with hw_closid, which can be parsed into closid
according to fixed rules (LxCODE and MBA are given closid, LxDATA is
given to closid + 1), so the maximum number of rdtgroup can be created
is also restricted by half if cdp enabled.

The length of Lx Cache domains displayed in schemata is compressed, this
is because for specified hardware platform, domains of each resource may
be too many to be easily operated for user interaction.

This patch also move parse_cbm() and parse_bw() to mpam_resctrl.c for clarity.
Signed-off-by: NWang ShaoBo <bobo.shaobowang@huawei.com>
Reviewed-by: NXiongfeng Wang <wangxiongfeng2@huawei.com>
Reviewed-by: NCheng Jian <cj.chengjian@huawei.com>
Signed-off-by: NYang Yingliang <yangyingliang@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

hulk inclusion
category: feature
feature: ARM MPAM support
bugzilla: 48265
CVE: NA

--------------------------------

Add a schema list for each rdt domain, we use this list to store changes
from schemata row instead of previous ctrlval array live in resctrl
resource structure, when mounting resctrl sysfs happened, we would reset
all resource's configuration into default by resctrl_group_update_domains().

Currently each row in schemata sysfile occupy a list node, this may be
extended for perfecting control types.
Signed-off-by: NWang ShaoBo <bobo.shaobowang@huawei.com>
Reviewed-by: NXiongfeng Wang <wangxiongfeng2@huawei.com>
Reviewed-by: NCheng Jian <cj.chengjian@huawei.com>
Signed-off-by: NYang Yingliang <yangyingliang@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

hulk inclusion
category: feature
feature: ARM MPAM support
bugzilla: 48265
CVE: NA

--------------------------------

Initialize schemata list when mount resctrl sysfs and destroy it when
umount, each list node contains the value updated by schemata (in resctrl
sysfs) row.

Partial code is borrowed from 250656171d95 ("x86/resctrl: Stop using Lx
CODE/DATA resources"), as it illustrates:

  Now that CDP enable/disable is global, and the closid offset correction
  is based on the configuration being applied, we are using different
  hw_closid slots in the ctrl array for CODE/DATA schema. This lets
  us merge them using the same Lx resource twice for CDP's CODE/DATA
  schema. This keeps the illusion of separate caches in the resctrl code.

  When CDP is enabled for a cache, create two schema generating the names
  and setting the configuration type.

  We can now remove the initialisation of the illusionary hw_resources:
  'cdp_capable' just requires setting a flag, resctrl knows what to do
  from there.

Link: http://www.linux-arm.org/git?p=linux-jm.git;a=commit;h=250656171d95dea079cc661098a0984e7237aa25Signed-off-by: NWang ShaoBo <bobo.shaobowang@huawei.com>
Reviewed-by: NXiongfeng Wang <wangxiongfeng2@huawei.com>
Reviewed-by: NCheng Jian <cj.chengjian@huawei.com>
Signed-off-by: NYang Yingliang <yangyingliang@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

hulk inclusion
category: feature
feature: ARM MPAM support
bugzilla: 48265
CVE: NA

--------------------------------

This supports cdpl2,cdpl3 parameters for mount options, some other options
including capabilities' feature such as priority and hardlimit will be
supported in the future, a simple example like this.

e.g.
   > mount -t resctrl resctrl /sys/fs/resctrl -o cdpl3
   > cd /sys/fs/resctrl && cat schemata
     L3CODE:0=7fff;1=7fff;2=7fff;3=7fff
     L3CODE:0=7fff;1=7fff;2=7fff;3=7fff
     MB:0=100;1=100;2=100;3=100

Note that we only complete this part interface adaption, not mean cdp is
supported currently.
Signed-off-by: NWang ShaoBo <bobo.shaobowang@huawei.com>
Reviewed-by: NXiongfeng Wang <wangxiongfeng2@huawei.com>
Reviewed-by: NCheng Jian <cj.chengjian@huawei.com>
Signed-off-by: NYang Yingliang <yangyingliang@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

hulk inclusion
category: feature
feature: ARM MPAM support
bugzilla: 48265
CVE: NA

--------------------------------

CDP (Code and Data Prioritization) should also be supported, because
separate code and data caches is an illusion resctrl creates using
CDP, as James said, The L2 cache is controlled from one place regardless.
Arm doesn't specify a cache topology. Platforms may have separate L2
code and data caches, with independent controls. On such a system
we would need a unified L2 cache to be an illusion. To support Arm's
MPAM, we need CDP to not be implicit between the architecture code
and the file-system code. this add a series definitions independent
of resctrl resources.

To do this we make the code/data/both 'type' a property of the
configuration that comes from the schema. This lets us combined the
illusionary cache. Eventually we separate the architecture code and
file-system code's idea of closid, the architecture code can then
provide helpers to map one to the other.

Part of this code is borrowed to James's, See links.

Link: http://www.linux-arm.org/git?p=linux-jm.git;a=commit;h=57a6f6204f72e2afc116721b2b86451fa19a32a6
Link: http://www.linux-arm.org/git?p=linux-jm.git;a=commit;h=1385052cce87a8aed5dc0e96967cedd9e74a17e0Signed-off-by: NWang ShaoBo <bobo.shaobowang@huawei.com>
Reviewed-by: NXiongfeng Wang <wangxiongfeng2@huawei.com>
Reviewed-by: NCheng Jian <cj.chengjian@huawei.com>
Signed-off-by: NYang Yingliang <yangyingliang@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>