Stop restricting the PCI search to a range of PCI domains fed to
pci_get_domain_bus_and_slot(). Instead, use for_each_pci_dev() and
look at all PCI domains in one pass.
On systems with more than 8 sockets, this avoids error messages like
"Information: Invalid level, Can't get TDP control information at
specified levels on cpu 480" from the intel speed select utility.
Fixes: aa2ddd2425 ("platform/x86: ISST: Use numa node id for cpu pci dev mapping")
Signed-off-by: Steve Wahl <steve.wahl@hpe.com>
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Link: https://lore.kernel.org/r/20230519160420.2588475-1-steve.wahl@hpe.com
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Call mutex_unlock(&isst_tpmi_dev_lock) before returning on this
error path.
Fixes: d805456c71 ("platform/x86: ISST: Enumerate TPMI SST and create framework")
Signed-off-by: Dan Carpenter <error27@gmail.com>
Acked-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Link: https://lore.kernel.org/r/dcdebbb7-7de6-4d04-8e7a-43d5ca043484@kili.mountain
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
To support S3/S4 with TPMI interface add suspend/resume callbacks.
Here HW state is stored in suspend callback and restored during
resume callback.
The hardware state which needs to be stored/restored:
- CLOS configuration
- CLOS Association
- SST-CP enable/disable status
- SST-PP perf level setting
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Suggested-by: Hans de Goede <hdegoede@redhat.com>
Reviewed-by: Zhang Rui <rui.zhang@intel.com>
Tested-by: Pragya Tanwar <pragya.tanwar@intel.com>
Link: https://lore.kernel.org/r/20230308070642.1727167-9-srinivas.pandruvada@linux.intel.com
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
The support of Intel Speed Select Technology - Turbo Frequency (SST-TF)
feature enables the ability to set different “All core turbo ratio
limits” to cores based on the priority. By using this feature, some cores
can be configured to get higher turbo frequency by designating them as
high priority at the cost of lower or no turbo frequency on the low
priority cores.
One new IOCTLs are added:
ISST_IF_GET_TURBO_FREQ_INFO : Get information about turbo frequency
buckets
Once an instance is identified, read or write from correct MMIO
offset for a given field as defined in the specification.
For details on SST-TF operations using intel-speed-selet utility,
refer to:
Documentation/admin-guide/pm/intel-speed-select.rst
under the kernel documentation
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Reviewed-by: Zhang Rui <rui.zhang@intel.com>
Tested-by: Pragya Tanwar <pragya.tanwar@intel.com>
Link: https://lore.kernel.org/r/20230308070642.1727167-8-srinivas.pandruvada@linux.intel.com
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
The Intel Speed Select Technology - Base Frequency (SST-BF) feature lets
the user control base frequency. If some critical workload threads demand
constant high guaranteed performance, then this feature can be used to
execute the thread at higher base frequency on specific sets of CPUs
(high priority CPUs) at the cost of lower base frequency (low priority
CPUs) on other CPUs.
Two new IOCTLs are added:
ISST_IF_GET_BASE_FREQ_INFO : Get frequency information for high and
low priority CPUs
ISST_IF_GET_BASE_FREQ_CPU_MASK : CPUs capable of higher frequency
Once an instance is identified, read or write from correct MMIO
offset for a given field as defined in the specification.
For details on SST-BF operations using intel-speed-selet utility,
refer to:
Documentation/admin-guide/pm/intel-speed-select.rst
under the kernel documentation
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Reviewed-by: Zhang Rui <rui.zhang@intel.com>
Tested-by: Pragya Tanwar <pragya.tanwar@intel.com>
Link: https://lore.kernel.org/r/20230308070642.1727167-7-srinivas.pandruvada@linux.intel.com
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
This Intel Speed Select Technology - Performance Profile (SST-PP) feature
introduces a mechanism that allows multiple optimized performance profiles
per system. Each profile defines a set of CPUs that need to be online and
rest offline to sustain a guaranteed base frequency.
Five new IOCTLs are added:
ISST_IF_PERF_LEVELS : Get number of performance levels
ISST_IF_PERF_SET_LEVEL : Set to a new performance level
ISST_IF_PERF_SET_FEATURE : Activate SST-BF/SST-TF for a performance level
ISST_IF_GET_PERF_LEVEL_INFO : Get parameters for a performance level
ISST_IF_GET_PERF_LEVEL_CPU_MASK : Get CPU mask for a performance level
Once an instance is identified, read or write from correct MMIO
offset for a given field as defined in the specification.
For details on SST PP operations using intel-speed-selet utility,
refer to:
Documentation/admin-guide/pm/intel-speed-select.rst
under the kernel documentation
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Reviewed-by: Zhang Rui <rui.zhang@intel.com>
Tested-by: Pragya Tanwar <pragya.tanwar@intel.com>
Link: https://lore.kernel.org/r/20230308070642.1727167-6-srinivas.pandruvada@linux.intel.com
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Intel Speed Select Technology Core Power (SST-CP) is an interface that
allows users to define per core priority. This defines a mechanism to
distribute power among cores when there is a power constrained
scenario. This defines a class of service (CLOS) configuration.
Three new IOCTLs are added:
ISST_IF_CORE_POWER_STATE : Enable/Disable SST-CP
ISST_IF_CLOS_PARAM : Configure CLOS parameters
ISST_IF_CLOS_ASSOC : Associate CPUs to a CLOS
To associate CPUs to CLOS, either Linux CPU numbering or PUNIT numbering
scheme can be used, using parameter punit_cpu_map (1: for PUNIT numbering
0 for Linux CPU number).
There is no change to IOCTL to get PUNIT CPU number for a CPU.
Introduce get_instance() function, which is used by majority of IOCTLs
processing to convert a socket and power domain to
tpmi_per_power_domain_info * instance. This instance has all the MMIO
offsets stored to read a particular field.
Once an instance is identified, read or write from correct MMIO
offset for a given field as defined in the specification.
For details on SST CP operations using intel-speed-selet utility,
refer to:
Documentation/admin-guide/pm/intel-speed-select.rst
under the kernel documentation
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Reviewed-by: Zhang Rui <rui.zhang@intel.com>
Tested-by: Pragya Tanwar <pragya.tanwar@intel.com>
Link: https://lore.kernel.org/r/20230308070642.1727167-5-srinivas.pandruvada@linux.intel.com
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
SST registers are presented to OS in multi-layer structures starting
with a SST header showing version information freezing current
definition.
For details on SST terminology refer to
Documentation/admin-guide/pm/intel-speed-select.rst
under the kernel documentation
SST TPMI details are published in the following document:
https://github.com/intel/tpmi_power_management/blob/main/SST_TPMI_public_disclosure_FINAL.docx
SST MMIO structure layout follows:
SST-HEADER
SST-CP Header
SST-CP CONTROL
SST-CP STATUS
SST-CP CONFIG0
SST-CP CONFIG1
...
...
SST-PP Header
SST-PP OFFSET_0
SST-PP OFFSET_1
SST_PP_0_INFO
SST_PP_1_INFO
SST_PP_2_INFO
SST_PP_3_INFO
SST-PP CONTROL
SST-PP STATUS
Each register bank contains information to get to next lower level
information. This information is parsed and stored in the struct
tpmi_per_power_domain_info for each domain. This information is
used to process each SST requests.
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Reviewed-by: Zhang Rui <rui.zhang@intel.com>
Tested-by: Pragya Tanwar <pragya.tanwar@intel.com>
Link: https://lore.kernel.org/r/20230308070642.1727167-4-srinivas.pandruvada@linux.intel.com
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Enumerate TPMI SST driver and create basic framework to add more
features.
The basic user space interface is still same as the legacy using
/dev/isst_interface. Users of "intel-speed-select" utility should
be able to use same commands as prior gens without being aware
of new underlying hardware interface.
TPMI SST driver enumerates on device "intel_vsec.tpmi-sst". Since there
can be multiple instances and there is one common SST core, split
implementation into two parts: A common core part and an enumeration
part. The enumeration driver is loaded for each device instance and
register with the TPMI SST core driver.
On very first enumeration the TPMI SST core driver register with SST
core driver to get IOCTL callbacks. The api_version is incremented
for IOCTL ISST_IF_GET_PLATFORM_INFO, so that user space can issue
new IOCTLs.
Each TPMI package contains multiple power domains. Each power domain
has its own set of SST controls. For each domain map the MMIO memory
and update per domain struct tpmi_per_power_domain_info. This information
will be used to implement other SST interfaces.
Implement first IOCTL commands to get number of TPMI SST instances
and instance mask as some of the power domains may not have any
SST controls.
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Reviewed-by: Zhang Rui <rui.zhang@intel.com>
Tested-by: Pragya Tanwar <pragya.tanwar@intel.com>
Link: https://lore.kernel.org/r/20230308070642.1727167-3-srinivas.pandruvada@linux.intel.com
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
To map Linux CPU numbering scheme to hardware CPU numbering scheme
MSR 0x53 is getting used. But for new generation of CPUs, this MSR
is not valid. Since this is model specific MSR, this is possible.
A new MSR 0x54 is defined for this purpose. User space can use the
API version to distinguish format from MSR 0x53.
Intel speed select utility is updated to use the new format based
on the API version.
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Reviewed-by: Zhang Rui <rui.zhang@intel.com>
Tested-by: Pragya Tanwar <pragya.tanwar@intel.com>
Link: https://lore.kernel.org/r/20230308070642.1727167-2-srinivas.pandruvada@linux.intel.com
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
User space can get the API version using IOCTL
ISST_IF_GET_PLATFORM_INFO. This information can be used to get IOCTLs
supported by the kernel driver. This version is hardcoded in the driver.
Allow the registered client to specify the supported API version. In
this way a registered client can specify a higher API version to extend
IOCTL set.
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Link: https://lore.kernel.org/r/20230211063257.311746-5-srinivas.pandruvada@linux.intel.com
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
The common IOCTL handler has a predefined list of IOCTLs it can
handle. There is no default handler, if there is no match.
Allow a client driver to define their own version of default IOCTL
callback. In this way the default handling is passed to the client
drivers to handle.
With the introduction of TPMI target, IOCTL list is extended. The
additional TPMI specific IOCTLs will be passed to the TPMI client
driver default IOCTL handler.
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Link: https://lore.kernel.org/r/20230211063257.311746-4-srinivas.pandruvada@linux.intel.com
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Add TPMI as one of the device type which can be registered with ISST
common driver.
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Link: https://lore.kernel.org/r/20230211063257.311746-3-srinivas.pandruvada@linux.intel.com
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
A new command CONFIG_TDP_GET_RATIO_INFO is added, with sub command type
of 0x0C. The previous range of valid sub commands was from 0x00 to 0x0B.
Change the valid range from 0x00 to 0x0C.
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Link: https://lore.kernel.org/r/20230227053504.2734214-1-srinivas.pandruvada@linux.intel.com
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
On a multiple package system using Sub-NUMA clustering, there is an issue
in mapping Linux CPU number to PUNIT PCI device when manufacturer decided
to reuse the PCI bus number across packages. Bus number can be reused as
long as they are in different domain or segment. In this case some CPU
will fail to find a PCI device to issue SST requests.
When bus numbers are reused across CPU packages, we are using proximity
information by matching CPU numa node id to PUNIT PCI device numa node
id. But on a package there can be only one PUNIT PCI device, but multiple
numa nodes (one for each sub cluster). So, the numa node ID of the PUNIT
PCI device can only match with one numa node id of CPUs in a sub cluster
in the package.
Since there can be only one PUNIT PCI device per package, if we match
with numa node id of any sub cluster in that package, we can use that
mapping for any CPU in that package. So, store the match information
in a per package data structure and return the information when there
is no match.
While here, use defines for max bus number instead of hardcoding.
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Link: https://lore.kernel.org/r/20220629194817.2418240-1-srinivas.pandruvada@linux.intel.com
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
As reported:
[ 256.104522] ======================================================
[ 256.113783] WARNING: possible circular locking dependency detected
[ 256.120093] 5.16.0-rc6-yocto-standard+ #99 Not tainted
[ 256.125362] ------------------------------------------------------
[ 256.131673] intel-speed-sel/844 is trying to acquire lock:
[ 256.137290] ffffffffc036f0d0 (punit_misc_dev_lock){+.+.}-{3:3}, at: isst_if_open+0x18/0x90 [isst_if_common]
[ 256.147171]
[ 256.147171] but task is already holding lock:
[ 256.153135] ffffffff8ee7cb50 (misc_mtx){+.+.}-{3:3}, at: misc_open+0x2a/0x170
[ 256.160407]
[ 256.160407] which lock already depends on the new lock.
[ 256.160407]
[ 256.168712]
[ 256.168712] the existing dependency chain (in reverse order) is:
[ 256.176327]
[ 256.176327] -> #1 (misc_mtx){+.+.}-{3:3}:
[ 256.181946] lock_acquire+0x1e6/0x330
[ 256.186265] __mutex_lock+0x9b/0x9b0
[ 256.190497] mutex_lock_nested+0x1b/0x20
[ 256.195075] misc_register+0x32/0x1a0
[ 256.199390] isst_if_cdev_register+0x65/0x180 [isst_if_common]
[ 256.205878] isst_if_probe+0x144/0x16e [isst_if_mmio]
...
[ 256.241976]
[ 256.241976] -> #0 (punit_misc_dev_lock){+.+.}-{3:3}:
[ 256.248552] validate_chain+0xbc6/0x1750
[ 256.253131] __lock_acquire+0x88c/0xc10
[ 256.257618] lock_acquire+0x1e6/0x330
[ 256.261933] __mutex_lock+0x9b/0x9b0
[ 256.266165] mutex_lock_nested+0x1b/0x20
[ 256.270739] isst_if_open+0x18/0x90 [isst_if_common]
[ 256.276356] misc_open+0x100/0x170
[ 256.280409] chrdev_open+0xa5/0x1e0
...
The call sequence suggested that misc_device /dev file can be opened
before misc device is yet to be registered, which is done only once.
Here punit_misc_dev_lock was used as common lock, to protect the
registration by multiple ISST HW drivers, one time setup, prevent
duplicate registry of misc device and prevent load/unload when device
is open.
We can split into locks:
- One which just prevent duplicate call to misc_register() and one
time setup. Also never call again if the misc_register() failed or
required one time setup is failed. This lock is not shared with
any misc device callbacks.
- The other lock protects registry, load and unload of HW drivers.
Sequence in isst_if_cdev_register()
- Register callbacks under punit_misc_dev_open_lock
- Call isst_misc_reg() which registers misc_device on the first
registry which is under punit_misc_dev_reg_lock, which is not
shared with callbacks.
Sequence in isst_if_cdev_unregister
Just opposite of isst_if_cdev_register
Reported-and-tested-by: Liwei Song <liwei.song@windriver.com>
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Link: https://lore.kernel.org/r/20220112022521.54669-1-srinivas.pandruvada@linux.intel.com
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
The code works the same either way, but it's better to use semi-colons
to separate statements.
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Link: https://lore.kernel.org/r/20210825072357.GA12957@kili
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Move Intel Speed Select interface driver to intel sub-directory to improve
readability and rename it from intel_speed_select_if to speed_select_if.
Signed-off-by: Kate Hsuan <hpa@redhat.com>
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Link: https://lore.kernel.org/r/20210820110458.73018-14-andriy.shevchenko@linux.intel.com
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
