task #29499913

commit 1aa7177cdcb38d114e0a80dc27482e645d8068f2 upstream

Add functionality for base-freq info|enable|disable info on CascadeLake-N.

Sample output:
Intel(R) Speed Select Technology
Executing on CPU model:85[0x55]
 package-0
  die-0
    cpu-0
      speed-select-base-freq
        high-priority-base-frequency(MHz):2700000
        high-priority-cpu-mask:00000000,0000e8c0
        high-priority-cpu-list:6,7,11,13,14,15
        low-priority-base-frequency(MHz):2100000
 package-1
  die-0
    cpu-20
      speed-select-base-freq
        high-priority-base-frequency(MHz):2700000
        high-priority-cpu-mask:0000000e,8c000000
        high-priority-cpu-list:26,27,31,33,34,35
        low-priority-base-frequency(MHz):2100000

The enable command always returns success, and the disable command always
returns failed because SST-BF cannot be enabled or disabled from the OS on
CascadeLake-N.

Enable command also have support for --auto|-a option, which sets cpufreq
scaling_min to max, so that the high priority base frequency can be the
required minimum for high priority cores. Disable command with -a/--auto
option reset the setting back to the min frequency.
Signed-off-by: NPrarit Bhargava <prarit@redhat.com>
Signed-off-by: NSrinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NErwei Deng <erwei@linux.alibaba.com>
Reviewed-by: NArtie Ding <artie.ding@linux.alibaba.com>

task #29499913

commit 062e4aac92e8acbab2fd135e90b580b903c3724b upstream

Add functionality for "perf-profile info" on CascadeLake-N.

Sample output:
intel-speed-select perf-profile info
Intel(R) Speed Select Technology
Executing on CPU model:85[0x55]
 package-0
  die-0
    cpu-0
      perf-profile-level-0
        cpu-count:20
        enable-cpu-mask:00000000,000fffff
        enable-cpu-list:0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19
        thermal-design-power-ratio:23
        base-frequency(MHz):2300
        speed-select-turbo-freq:unsupported
        speed-select-base-freq:enabled
        speed-select-base-freq
          high-priority-base-frequency(MHz):2700000
          high-priority-cpu-mask:00000000,0000e8c0
          high-priority-cpu-list:6,7,11,13,14,15
          low-priority-base-frequency(MHz):2100000
 package-1
  die-0
    cpu-20
      perf-profile-level-0
        cpu-count:20
        enable-cpu-mask:000000ff,fff00000
        enable-cpu-list:20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39
        thermal-design-power-ratio:23
        base-frequency(MHz):2300
        speed-select-turbo-freq:unsupported
        speed-select-base-freq:enabled
        speed-select-base-freq
          high-priority-base-frequency(MHz):2700000
          high-priority-cpu-mask:0000000e,8c000000
          high-priority-cpu-list:26,27,31,33,34,35
          low-priority-base-frequency(MHz):2100000
Signed-off-by: NPrarit Bhargava <prarit@redhat.com>
Signed-off-by: NSrinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NErwei Deng <erwei@linux.alibaba.com>
Reviewed-by: NArtie Ding <artie.ding@linux.alibaba.com>

task #29499913

commit c829f0ef7bfc4b294e67506779853547e8b52c68 upstream

CascadeLake-N only supports SST-BF and needs some of the perf-profile
commands, and the base-freq commands.

Add help functions, and create an empty command structures (the functions
will be implemented later in this patchset).  Call these functions
when running on CascadeLake-N.
Signed-off-by: NPrarit Bhargava <prarit@redhat.com>
Signed-off-by: NSrinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NErwei Deng <erwei@linux.alibaba.com>
Reviewed-by: NArtie Ding <artie.ding@linux.alibaba.com>

task #29499913

commit 1c1d935c8418a63d110b2b31c57ba3d75f94f49d upstream

Three CascadeLake-N models (6252N, 6230N, and 5218N) have SST-PBF support.

Return an error if the CascadeLake processor is not one of these specific
models.
Signed-off-by: NPrarit Bhargava <prarit@redhat.com>
Signed-off-by: NSrinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NErwei Deng <erwei@linux.alibaba.com>
Reviewed-by: NArtie Ding <artie.ding@linux.alibaba.com>

task #29499913

commit 210369dc73b23fab1b9727fba35c2f83d00b6277 upstream

Make the process_command take any help command and command list.  This
will make it easier to help commands and a command list for CascadeLake-N.
Signed-off-by: NPrarit Bhargava <prarit@redhat.com>
Signed-off-by: NSrinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NErwei Deng <erwei@linux.alibaba.com>
Reviewed-by: NArtie Ding <artie.ding@linux.alibaba.com>

task #29499913

commit ce1326a2f9f7bf8920bc7f07e7b13ce8737b4681 upstream

The current code structure has similar but separate command functions for
the enable and disable operations.  This can be improved by adding an int
argument to the command function structure, and interpreting 1 as enable
and 0 as disable.  This change results in the removal of the disable
command functions.

Add int argument to the command function structure.
Signed-off-by: NPrarit Bhargava <prarit@redhat.com>
Signed-off-by: NSrinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NErwei Deng <erwei@linux.alibaba.com>
Reviewed-by: NArtie Ding <artie.ding@linux.alibaba.com>

task #29499913

commit 4e26fabfe15ca67125f1b6984671c2e9048d64a5 upstream

The turbo-freq feature is dependent on the core-power feature. If the
core-power feature is disabled while the turbo-freq feature is enabled,
this will break the turbo-freq feature. This is a firmware limitation,
where they can't return error under this scenario.

So when trying to disable core-power, make sure that the turbo-freq
feature is not enabled. If it enabled, return error if user is trying to
disable the core-power feature.
Signed-off-by: NSrinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NErwei Deng <erwei@linux.alibaba.com>
Reviewed-by: NArtie Ding <artie.ding@linux.alibaba.com>

task #29499913

commit a6a82f9bcd2ad3fab5f20fc406727dc62775db92 upstream

Introduce --auto|-a option to turbo-freq enable feature, so that it
does in one step for users who are OK by setting all passed target cores
as high priority and set in CLOS 0 and remaining in CLOS 3. In this way,
users don't have to take multiple steps to enable turbo-freq feature. For
users who want more fine grain control, they can always use core-power
feature to set custom CLOS configuration and assignment.

While here also print the error to output when clos configuration fails.

For example
intel-speed-select -c 0-4 turbo-freq enable --auto

The above command will enable turbo-freq and core-power feature. Also
mark CPU 0 to CPU 4 as high priority.
Signed-off-by: NSrinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NErwei Deng <erwei@linux.alibaba.com>
Reviewed-by: NArtie Ding <artie.ding@linux.alibaba.com>

task #29499913

commit 354bd06f40c4ba7b2f12d9f0f119dff62a2f922f upstream

Introduce --auto|-a option to base-freq enable feature, so that it
does in one step for users who are OK by setting all cores with higher
base frequency to be set in CLOS 0 and remaining in CLOS 3. This option
also sets corresponding clos.min to CLOS 0 and CLOS3. In this way, users
don't have to take multiple steps to enable base-freq feature. For users
who want more fine grain control, they can always use core-power feature
to set custom CLOS configuration and assignment.

Also adjust cpufreq/scaling_min_freq for higher and lower priority cores.

For example user can use:
intel-speed-select base-freq enable --auto
Signed-off-by: NSrinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NErwei Deng <erwei@linux.alibaba.com>
Reviewed-by: NArtie Ding <artie.ding@linux.alibaba.com>

task #29499913

commit abd120e3bdf3dd72ba1ed9ac077a861e0e3dc43a upstream

Fix warning for:
isst-config.c: In function ‘set_cpu_online_offline’:
isst-config.c:221:3: warning: ignoring return value of ‘write’,
declared with attribute warn_unused_result [-Wunused-result]
   write(fd, "1\n", 2);
Signed-off-by: NSrinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NErwei Deng <erwei@linux.alibaba.com>
Reviewed-by: NArtie Ding <artie.ding@linux.alibaba.com>

task #29499913

commit b3abfd778bf1dbdd96f70bd7d00671d027f67c62 upstream

commit "c016ae8f9fa04d361efc8629de49ad3af12b5262
"tools/power/x86/intel-speed-select: Output success/failed for command
output" introduced a regression in perf-profile outputs. With this
the result field is changed to string interpreting every non zero
value as errors. But these commands display on zero (>0) result.
For example before this commit the display was:

package-1
  die-0
    cpu-14
        get-config-levels:4

Here the get-config-levels is interpreted as error and displayed
as error with the above commit:

package-1
  die-0
    cpu-14
        get-config-levels:failed(error 4)

Fix this issue by not using isst_display_result() to display such results,
but define a new function which formats this data and prints.
Signed-off-by: NSrinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NErwei Deng <erwei@linux.alibaba.com>
Reviewed-by: NArtie Ding <artie.ding@linux.alibaba.com>

task #29499913

commit 188afed9db7db3aefc8c9c33150be2f8f398da9a upstream

Add additional command to get the clos enable and priority type. The
current info option is actually dumping per clos QOS config, so name
the command appropriately to get-config.
Signed-off-by: NSrinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NErwei Deng <erwei@linux.alibaba.com>
Reviewed-by: NArtie Ding <artie.ding@linux.alibaba.com>

task #29499913

commit d2d1f304dc965e6a06e7f105b09bffceb477fccc upstream

Fix wrong debug print for cpu, which is displayed as CLOS. Also
avoid printing clos id, when user is specify clos as parameter.
Signed-off-by: NSrinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NErwei Deng <erwei@linux.alibaba.com>
Reviewed-by: NArtie Ding <artie.ding@linux.alibaba.com>

task #29499913

commit e118fbe366d817175b2c47bfa338959bbff5bd37 upstream

Format the get-assoc command output consistant with other commands.
For example:

Intel(R) Speed Select Technology
Executing on CPU model:142[0x8e]
 package-0
  die-0
    cpu-0
      get-assoc
        clos:0
Signed-off-by: NSrinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NErwei Deng <erwei@linux.alibaba.com>
Reviewed-by: NArtie Ding <artie.ding@linux.alibaba.com>

task #29499913

commit 3c64c81ad1f06823b603e9143193dcb4f3121a2f upstream

Using enable core mask, do online offline CPUs. There is a new option
--online|-o for set-config-level.
Signed-off-by: NSrinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NErwei Deng <erwei@linux.alibaba.com>
Reviewed-by: NArtie Ding <artie.ding@linux.alibaba.com>

task #29499913

commit 44460efe44e05eae2f21e57d06d542bbbb792e65 upstream

If the CPU package has the less logical CPU than topo_max_cpus, but un-present
CPU's punit_cpu_core will be initiated to 0 and they will be count to core 0

Like below, there are only 10 high priority cores (20 logical CPUs) in the CPU
package, but it count to 27 logic CPUs.

./intel-speed-select base-freq info -l 0 | grep mask
        high-priority-cpu-mask:7f000179,f000179f

With the fix patch:
./intel-speed-select base-freq info -l 0
        high-priority-cpu-mask:00000179,f000179f
Signed-off-by: NYouquan Song <youquan.song@intel.com>
Signed-off-by: NSrinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NErwei Deng <erwei@linux.alibaba.com>
Reviewed-by: NArtie Ding <artie.ding@linux.alibaba.com>

task #29499913

commit 1233c7b95c7045905e40c11484493f20ab521d21 upstream

Read the bucket and core count relationship via MSR and display
when displaying turbo ratio limits.
Signed-off-by: NSrinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NErwei Deng <erwei@linux.alibaba.com>
Reviewed-by: NArtie Ding <artie.ding@linux.alibaba.com>

task #29499913

commit 3bc3d30ca324bfc3045a1a7fe1f5fe5ad5d92fd9 upstream

cpumasks are allocated by calling the alloc_cpu_mask() function and are
never free'd.  They should be free'd after the commands have run.

Fix the memory leaks by calling free_cpu_set().
Signed-off-by: NPrarit Bhargava <prarit@redhat.com>
Acked-by: NSrinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Cc: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Cc: David Arcari <darcari@redhat.com>
Cc: linux-kernel@vger.kernel.org
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NErwei Deng <erwei@linux.alibaba.com>
Reviewed-by: NArtie Ding <artie.ding@linux.alibaba.com>

task #29499913

commit 522586a9d5c32b5d7569237eebbfd8241f196419 upstream

Command output has confusing data, returning "0" on success.  For example

|# ./intel-speed-select -c 14 turbo-freq enable
Intel(R) Speed Select Technology
Executing on CPU model:106[0x6a]
 package-1
   die-0
     cpu-14
       turbo-freq
         enable:0

To avoid confusion change the command output to 'success' or 'failed'.

v2: Remove help output line.
Signed-off-by: NPrarit Bhargava <prarit@redhat.com>
Acked-by: NSrinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Cc: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Cc: David Arcari <darcari@redhat.com>
Cc: linux-kernel@vger.kernel.org
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NErwei Deng <erwei@linux.alibaba.com>
Reviewed-by: NArtie Ding <artie.ding@linux.alibaba.com>

task #29499913

commit 49aed155ec1b07f8820cceec7317c597a6f69c6e upstream

The intel-speed-select tool currently only outputs a hexidecimal CPU mask,
which requires translation for use with kernel parameters such as
isolcpus.

Along with the CPU mask, output a human readable CPU list.
Signed-off-by: NPrarit Bhargava <prarit@redhat.com>
Acked-by: NSrinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Cc: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Cc: David Arcari <darcari@redhat.com>
Cc: linux-kernel@vger.kernel.org
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NErwei Deng <erwei@linux.alibaba.com>
Reviewed-by: NArtie Ding <artie.ding@linux.alibaba.com>

task #29499913

commit 76c2ef35f73913380f32d473a181d68decfca4f4 upstream

The intel-speed-select tool currently outputs the turbo ratio for every
bucket.  Make the output more user-friendly by changing the output to the
maximum turbo frequency.
Signed-off-by: NPrarit Bhargava <prarit@redhat.com>
Acked-by: NSrinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Cc: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Cc: David Arcari <darcari@redhat.com>
Cc: linux-kernel@vger.kernel.org
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NErwei Deng <erwei@linux.alibaba.com>
Reviewed-by: NArtie Ding <artie.ding@linux.alibaba.com>

task #29499913

commit 808088e4a1fb2d27a347db2890d43ae73c75a72f upstream

These features are introduced on new processors that will never operate
in the KHz range.

Save some zeros and switch the output to MHz.
Signed-off-by: NPrarit Bhargava <prarit@redhat.com>
Acked-by: NSrinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Cc: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Cc: David Arcari <darcari@redhat.com>
Cc: linux-kernel@vger.kernel.org
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NErwei Deng <erwei@linux.alibaba.com>
Reviewed-by: NArtie Ding <artie.ding@linux.alibaba.com>

task #29499913

commit dece22a2d5e37cf7c4cda9a7d5f036c9b3fdaccb upstream

The current output of 'intel-speed-select -c 53 perf-profile info -l 0'
shows

        speed-select-turbo-freq-support:1
        speed-select-base-freq-support:1
        speed-select-base-freq-enabled:0
        speed-select-turbo-freq-enabled:0

Simplify the output to single lines displaying status of disabled,
enabled, and unsupported.
Signed-off-by: NPrarit Bhargava <prarit@redhat.com>
Acked-by: NSrinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Cc: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Cc: David Arcari <darcari@redhat.com>
Cc: linux-kernel@vger.kernel.org
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NErwei Deng <erwei@linux.alibaba.com>
Reviewed-by: NArtie Ding <artie.ding@linux.alibaba.com>

task #29499913

commit 3ec2aef1b03eefeffe93817bd04068d3805d4a98 upstream

I have a system with 28 threads/socket but intel-speed-select reports
a cpu-count of 29.

Fix an off-by-one error in the cpu_count() function.
Signed-off-by: NPrarit Bhargava <prarit@redhat.com>
Acked-by: NSrinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Cc: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Cc: David Arcari <darcari@redhat.com>
Cc: linux-kernel@vger.kernel.org
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NErwei Deng <erwei@linux.alibaba.com>
Reviewed-by: NArtie Ding <artie.ding@linux.alibaba.com>

task #29499913

commit 43774c0dccb405a0b63a10829d45db4b74c29e40 upstream

Help is -h, not --h.
Signed-off-by: NPrarit Bhargava <prarit@redhat.com>
Acked-by: NSrinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Cc: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Cc: David Arcari <darcari@redhat.com>
Cc: linux-kernel@vger.kernel.org
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NErwei Deng <erwei@linux.alibaba.com>
Reviewed-by: NArtie Ding <artie.ding@linux.alibaba.com>

task #29499913

commit a8dc07448177dddd3947e89fe3904cb3169935c2 upstream

packag_ should be package_.
Signed-off-by: NPrarit Bhargava <prarit@redhat.com>
Acked-by: NSrinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Cc: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Cc: David Arcari <darcari@redhat.com>
Cc: linux-kernel@vger.kernel.org
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NErwei Deng <erwei@linux.alibaba.com>
Reviewed-by: NArtie Ding <artie.ding@linux.alibaba.com>

task #29499913

commit 010764b8856e5ee5056113704dc4b914ebc88f1d upstream

The isst_send_msr_command() function will read 8 bytes but we are
passing an address to an int (4 bytes) so it results in a read overflow.

Fixes: 3fb4f7cd472c ("tools/power/x86: A tool to validate Intel Speed Select commands")
Signed-off-by: NDan Carpenter <dan.carpenter@oracle.com>
Acked-by: NSrinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NErwei Deng <erwei@linux.alibaba.com>
Reviewed-by: NArtie Ding <artie.ding@linux.alibaba.com>

task #29499913

commit 2b4f3dab09b3a2853c49c866a383ffda8ee54a6f upstream

Add a .gitignore file for build include/ and final binary.
Signed-off-by: NPrarit Bhargava <prarit@redhat.com>
Acked-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: NSrinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Cc: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Cc: Andy Shevchenko <andy.shevchenko@gmail.com>
Cc: David Arcari <darcari@redhat.com>
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NErwei Deng <erwei@linux.alibaba.com>
Reviewed-by: NArtie Ding <artie.ding@linux.alibaba.com>

to #29593669

commit b115df076d337a727017538d11d7d46f5bcbff15 upstream

Add a Intel event file for perf.
Signed-off-by: NHaiyan Song <haiyanx.song@intel.com>
Reviewed-by: NKan Liang <kan.liang@linux.intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Jin Yao <yao.jin@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://lkml.kernel.org/r/8859095e-5b02-d6b7-fbdc-3f42b714bae0@intel.comSigned-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: NPeng Wang <rocking@linux.alibaba.com>
Reviewed-by: NArtie Ding <artie.ding@linux.alibaba.com>

fix #29564148

Global swapout can happen even with swappiness set to 0. In some
scenario, we do want OOM instead of unexpected global swapout.

This extends value range of vm_swappiness, and disables swapout
completely by setting swappiness to -1.
Signed-off-by: NXu Yu <xuyu@linux.alibaba.com>
Reviewed-by: NXunlei Pang <xlpang@linux.alibaba.com>

task #29239886

commit	7f7b2e358b1baa106afe6dec6dacb539d520dc6b linux-pm

On ICX platform, the CPU frequency will slowly ramp up
when woken up from C-states deeper than/equals to C1E,
and since C1E auto-promotion is enabled by default,
this might cause unexpected result.

Thus this patch disables C1E auto-promotion and expose
C1E as a separate idle state, so that the C1E and C6 can
be disabled via sysfs when necessary.

Besides C1 and C1E, the exit latency of C6 was measured
by a dedicated tool. However the exit latency(41us) exposed
by _CST is much smaller than the one measured(128us). This
is probably due to the _CST uses the exit latency when woken
up from PC0+C6, rather than PC6+C6 when C6 was measured.

erwei:
	This patch is coming from linux-pm, which provides the
	static C-state table for ICX. If not use this patch,
	the C-state information whould come from the ACPI which
	result in the cpu cannot enter the C1E and C6 state.
	And there are some differences between this patch and
	our kernel. The macro X86_MATCH_INTEL_FAM6_MODEL is
	used in this patch, but ICPU in source code. So I
	change them here.
Tested-by: NArtem Bityutskiy <artem.bityutskiy@linux.intel.com>
Acked-by: NArtem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: NZhang Rui <rui.zhang@intel.com>
Signed-off-by: NChen Yu <yu.c.chen@intel.com>
Signed-off-by: NErwei Deng <erwei@linux.alibaba.com>
Reviewed-by: NArtie Ding <artie.ding@linux.alibaba.com>

task #29239886

commit a3299182216397a0b943d2549d1997f4eba2bdd2 upstream

Add an admin-guide document for the intel_idle driver to describe
how it works: how it enumerates idle states, what happens during the
initialization of it, how it can be controlled via the kernel command
line and so on.
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: NRandy Dunlap <rdunlap@infradead.org>
Signed-off-by: Nyjia <yingbao.jia@intel.com>
Signed-off-by: NErwei Deng <erwei@linux.alibaba.com>
Reviewed-by: NArtie Ding <artie.ding@linux.alibaba.com>

task #29239886

commit 239ed06d0eefc4ee351af69bd64742ada56c4bdb upstream

To avoid build errors when CONFIG_ACPI_PROCESSOR_CSTATE is set and
CONFIG_ACPI_PROCESSOR is not (that may appear in randconfig builds),
make the former depend on the latter.
Acked-by: NLen Brown <len.brown@intel.com>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Nyjia <yingbao.jia@intel.com>
Signed-off-by: NErwei Deng <erwei@linux.alibaba.com>
Reviewed-by: NArtie Ding <artie.ding@linux.alibaba.com>

task #29239886

commit e6d4f08a677654385869ba0c39d7c9ceec47e5c5 upstream

In many cases, especially on server systems, it is desirable to avoid
enabling C-states that have been disabled in the platform firmware
(BIOS) setup, except for C1E.

As a rule, the C-states disabled this way are not listed by ACPI
_CST, so if that is used by intel_idle along with the specific
table of C-states that it has for the given processor, the C-states
disabled through the platform firmware will not be enabled by default
by intel_idle.

Accordingly, set the use_acpi flag (introduced previously) in all
server processor profiles defined in intel_idle (so as to make it use
ACPI _CST to decide which C-states to enable by default) and set
the CPUIDLE_FLAG_ALWAYS_ENABLE flag (also introduced previously)
for C1E in all C-states tables in intel_idle that contain C1 too
(so that C1E is enabled regardless of whether or not it is listed
by ACPI _CST).
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Nyjia <yingbao.jia@intel.com>
Signed-off-by: NErwei Deng <erwei@linux.alibaba.com>
Reviewed-by: NArtie Ding <artie.ding@linux.alibaba.com>

task #29239886

commit 4ec32d9e8e5b6d6eb491eeee3938665d8a2388fa upstream

Add a new module parameter called "no_acpi" to the intel_idle driver
to allow the driver to be prevented from using ACPI _CST via kernel
command line.
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Nyjia <yingbao.jia@intel.com>
Signed-off-by: NErwei Deng <erwei@linux.alibaba.com>
Reviewed-by: NArtie Ding <artie.ding@linux.alibaba.com>

task #29239886

commit bff8e60a86f4960133f90ad9add9adeb082b8154 upstream

Update the intel_idle driver to get the C-states information from ACPI
_CST in some cases in which the processor is known to the driver, as long as
that information is available and the new use_acpi flag is set in the
profile of the processor in question.

In the cases when there is a specific table of C-states for the given
processor in the driver, that table is used as the primary source of
information on the available C-states, but if ACPI _CST is present,
the C-states that are not listed by it will not be enabled by default
(they still can be enabled later by user space via sysfs, though).

The new CPUIDLE_FLAG_ALWAYS_ENABLE flag can be used for marking
C-states that should be enabled by default even if they are not
listed by ACPI _CST.
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Nyjia <yingbao.jia@intel.com>
Signed-off-by: NErwei Deng <erwei@linux.alibaba.com>
Reviewed-by: NArtie Ding <artie.ding@linux.alibaba.com>

task #29239886

commit 75a80267410e38ab76c4ceb39753f96d72113781 upstream

In certain situations it may be useful to prevent some idle states
from being used by default while allowing user space to enable them
later on.

For this purpose, introduce a new state flag, CPUIDLE_FLAG_OFF, to
mark idle states that should be disabled by default, make the core
set CPUIDLE_STATE_DISABLED_BY_USER for those states at the
initialization time and add a new state attribute in sysfs,
"default_status", to inform user space of the initial status of
the given idle state ("disabled" if CPUIDLE_FLAG_OFF is set for it,
"enabled" otherwise).
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Nyjia <yingbao.jia@intel.com>
Signed-off-by: NErwei Deng <erwei@linux.alibaba.com>
Reviewed-by: NArtie Ding <artie.ding@linux.alibaba.com>

task #29239886

commit aa5eee355b466cb33f97f79bed9740a472c4ab73 upstream

Important information is missing from user/admin cpuidle documentation
available today, so add a new user/admin document for cpuidle containing
current and comprehensive information to admin-guide and drop the old
.txt documents it is replacing.
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: NViresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: NUlf Hansson <ulf.hansson@linaro.org>
Signed-off-by: Nyjia <yingbao.jia@intel.com>
Signed-off-by: NErwei Deng <erwei@linux.alibaba.com>
Reviewed-by: NArtie Ding <artie.ding@linux.alibaba.com>

task #29239886

commit 44021606298870e4adc641ef3927e7bb47ca8236 upstream

Use BIT() macro to do a small tidy-up.

CPUIDLE_DRIVER_FLAGS_MASK is not used, so remove it.
Signed-off-by: NYangtao Li <tiny.windzz@gmail.com>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Nyjia <yingbao.jia@intel.com>
Signed-off-by: NErwei Deng <erwei@linux.alibaba.com>
Reviewed-by: NArtie Ding <artie.ding@linux.alibaba.com>

task #29239886

commit 18734958e9bfbc055805d110a38dc76307eba742 upstream

Modify the intel_idle driver to get the C-states information from ACPI
_CST if the processor model is not recognized by it.

The processor is still required to support MWAIT and the information
from ACPI _CST will only be used if all of the C-states listed by
_CST are of the ACPI_CSTATE_FFH type (which means that they are
expected to be entered via MWAIT).

Moreover, the driver assumes that the _CST information is the same
for all CPUs in the system, so it is sufficient to evaluate _CST for
one of them and extract the common list of C-states from there.
Also _CST is evaluated once at the system initialization time and
the driver does not respond to _CST change notifications (that can
be changed in the future).

The main functional difference between intel_idle with this change
and the ACPI processor driver is that the former sets the target
residency to be equal to the exit latency (provided by _CST) for
C1-type C-states and to 3 times the exit latency value for the other
C-state types, whereas the latter obtains the target residency by
multiplying the exit latency by the same number (2 by default) for
all C-state types.  Therefore it is expected that in general using
the former instead of the latter on the same system will lead to
improved energy-efficiency.
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Nyjia <yingbao.jia@intel.com>
Signed-off-by: NErwei Deng <erwei@linux.alibaba.com>
Reviewed-by: NArtie Ding <artie.ding@linux.alibaba.com>