From: "Gupta, Anshuman" <anshuman.gupta@intel.com>
To: Ramalingam C <ramalingam.c@intel.com>
Cc: igt-dev@lists.freedesktop.org
Subject: Re: [igt-dev] [PATCH i-g-t v3 1/2] tests/i915/i915_pm_rpm: Enable PC8+ residency test for all Gen9+
Date: Wed, 27 Mar 2019 19:15:01 +0530 [thread overview]
Message-ID: <e93cb578-80f4-2893-3fab-644251508fc2@intel.com> (raw)
In-Reply-To: <20190327122841.GA3905@intel.com>
On 3/27/2019 5:58 PM, Ramalingam C wrote:
> On 2019-03-27 at 15:25:22 +0530, Anshuman Gupta wrote:
>> Enabled has_pc8 global for ICL and Gen9+.
>> Modified PC8+ residency sub-test with all screen enabled.
>>
>> v2:Fixed the issue of skipped test on HSW.
>> Improved the code comment for MSR_PKG_CST_CONFIG_CONTROL mask and PC8
>> bits, it holds good for SKL/ICL and Goldmont microarchitecture.
>> Code readabilty improvement.
>> v3:Removed the connected_screens global. [Ram]
>> Removed pc8_needs_screen_off from mode_set_data structure,
>> made it global, aligning to has_pc8 and has_runtime_pm globals. [Ram]
>> Reuse connector lcoal variable in init_modeset_params_for_all_screen(). [Ram]
>> Addressed Coding guide lines comments. [Ram]
>>
>> Signed-off-by: Anshuman Gupta <anshuman.gupta@intel.com>
>> ---
>> tests/i915/i915_pm_rpm.c | 73 +++++++++++++++++++++++++++++++++++++++++++-----
>> 1 file changed, 66 insertions(+), 7 deletions(-)
>>
>> diff --git a/tests/i915/i915_pm_rpm.c b/tests/i915/i915_pm_rpm.c
>> index be296f5..7e0c370 100644
>> --- a/tests/i915/i915_pm_rpm.c
>> +++ b/tests/i915/i915_pm_rpm.c
>> @@ -52,7 +52,7 @@
>> #include "igt_device.h"
>>
>> #define MSR_PKG_CST_CONFIG_CONTROL 0xE2
>> -/* HSW/BDW: */
>> +/* HSW/BDW SKL/ICL and Goldmont */
> This is not HSW or BDW. I guess you want to list the platforms on which
> these bit mask will work. Please update the comment accordingly.
>> #define PKG_CST_LIMIT_MASK 0xF
>> #define PKG_CST_LIMIT_C8 0x6
>>
>> @@ -90,7 +90,7 @@ enum plane_type {
>>
>> int drm_fd, msr_fd, pc8_status_fd;
>> int debugfs;
>> -bool has_runtime_pm, has_pc8;
>> +bool has_runtime_pm, has_pc8, pc8_needs_screen_off;
>> struct mode_set_data ms_data;
>>
>> /* Stuff used when creating FBs and mode setting. */
>> @@ -121,6 +121,7 @@ struct modeset_params {
>> struct modeset_params lpsp_mode_params;
>> struct modeset_params non_lpsp_mode_params;
>> struct modeset_params *default_mode_params;
>> +struct modeset_params *screens_mode_params[MAX_CONNECTORS];
>>
>> static int8_t *pm_data = NULL;
>>
>> @@ -297,6 +298,37 @@ static bool init_modeset_params_for_type(struct mode_set_data *data,
>> return true;
>> }
>>
>> +static void init_modeset_params_for_all_screen(struct mode_set_data *data)
> BTW why do we need to pass the ms_data as a parameter when it is
> declared in global space? Am I missing something here?
I am not sure about it, exisitng code uses this global at so many places.
>> +{
>> + drmModeConnectorPtr connector = NULL;
>> + drmModeModeInfoPtr mode = NULL;
>> + int screen = 0;
>> +
>> + if (!data->res)
>> + return false;
>> +
>> + for (int i = 0; i < data->res->count_connectors; i++) {
>> + connector = data->connectors[i];
>> +
>> + if (connector->connection == DRM_MODE_CONNECTED
>> + && connector->count_modes) {
>> + screens_mode_params[screen] =
>> + malloc(sizeof(struct modeset_params));
>> + mode = &connector->modes[0];
>> + igt_create_pattern_fb(drm_fd, mode->hdisplay, mode->vdisplay,
>> + DRM_FORMAT_XRGB8888, LOCAL_DRM_FORMAT_MOD_NONE,
>> + &screens_mode_params[screen]->fb);
> Line length exceeds 80chars. Please wrap it. Applicable for elsewhere
> too.
here it will not be possible, it will not be readable, need to create a
separate function and move all these chunk of big lines, hope that will
be ok.
>> + screens_mode_params[screen]->crtc_id =
>> + kmstest_find_crtc_for_connector(drm_fd, data->res, connector, 0);
>> + screens_mode_params[screen]->connector_id = connector->connector_id;
>> + screens_mode_params[screen]->mode = mode;
>> + screen++;
>> + }
>> + }
>> +
>> + return;
>> +}
>> +
>> static void init_modeset_cached_params(struct mode_set_data *data)
>> {
>> bool lpsp, non_lpsp;
>> @@ -305,6 +337,7 @@ static void init_modeset_cached_params(struct mode_set_data *data)
>> SCREEN_TYPE_LPSP);
>> non_lpsp = init_modeset_params_for_type(data, &non_lpsp_mode_params,
>> SCREEN_TYPE_NON_LPSP);
>> + init_modeset_params_for_all_screen(data);
>>
>> if (lpsp)
>> default_mode_params = &lpsp_mode_params;
>> @@ -353,6 +386,22 @@ static bool enable_one_screen_with_type(struct mode_set_data *data,
>> return set_mode_for_params(params);
>> }
>>
>> +static void enable_all_screens(struct mode_set_data *data)
>> +{
>> + struct modeset_params *params = NULL;
>> +
>> + /* SKIP if there are no connected screens. */
>> + igt_require(screens_mode_params[0]);
>> +
>> + for (int i = 0; i < MAX_CONNECTORS ; i++) {
>> + params = screens_mode_params[i];
>> + if (params)
>> + set_mode_for_params(params);
>> + else
>> + break;
>> + }
>> +}
>> +
>> static void enable_one_screen(struct mode_set_data *data)
>> {
>> /* SKIP if there are no connected screens. */
>> @@ -685,8 +734,12 @@ static void setup_pc8(void)
>> {
>> has_pc8 = false;
>>
>> - /* Only Haswell supports the PC8 feature. */
>> - if (!IS_HASWELL(ms_data.devid) && !IS_BROADWELL(ms_data.devid))
>> + if (IS_HASWELL(ms_data.devid) || IS_BROADWELL(ms_data.devid))
>> + pc8_needs_screen_off = true;
>> + else if (AT_LEAST_GEN(ms_data.devid, 9))
>> + pc8_needs_screen_off = false;
>> + /* Only Haswell supports the PC8 feature on lesser than GEN9. */
>> + else
>> return;
>>
>> /* Make sure our Kernel supports MSR and the module is loaded. */
>> @@ -808,9 +861,15 @@ static void pc8_residency_subtest(void)
>> "configuration.\n");
>>
>> /* Make sure PC8+ residencies stop! */
>> - enable_one_screen(&ms_data);
>> - igt_assert_f(!pc8_plus_residency_changed(10),
>> - "PC8+ residency didn't stop with screen enabled.\n");
>> + if (pc8_needs_screen_off) {
>> + enable_one_screen(&ms_data);
>> + igt_assert_f(!pc8_plus_residency_changed(10),
> Could we use a macro for this timeout "10" ?
yes we can but there are other existing instances too are using
different timeut values, so there will be pleanty of timeout
macros.
>> + "PC8+ residency didn't stop with screen enabled.\n");
>> + } else {
>> + enable_all_screens(&ms_data);
>> + igt_assert_f(pc8_plus_residency_changed(10),
> macro should be good here.
>> + "Machine is not reaching PC8+ states with all screen enabled.\n");
> Please consider to make the log crisp to meet the 80 char limit.
>
> -Ram
>> + }
>> }
>>
>> static void modeset_subtest(enum screen_type type, int rounds, int wait_flags)
>> --
>> 2.7.4
>>
_______________________________________________
igt-dev mailing list
igt-dev@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/igt-dev
next prev parent reply other threads:[~2019-03-27 13:45 UTC|newest]
Thread overview: 9+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-03-27 9:55 [igt-dev] [PATCH i-g-t v3 0/2] Enabling PC8+ residency for all GEN9+ platforms v3 Anshuman Gupta
2019-03-27 9:55 ` [igt-dev] [PATCH i-g-t v3 1/2] tests/i915/i915_pm_rpm: Enable PC8+ residency test for all Gen9+ Anshuman Gupta
2019-03-27 12:28 ` Ramalingam C
2019-03-27 13:45 ` Gupta, Anshuman [this message]
2019-03-27 15:36 ` Ramalingam C
2019-03-27 9:55 ` [igt-dev] [PATCH i-g-t v3 2/2] tests/i915/i915_pm_rpm: modeset-pc8-residency-stress Anshuman Gupta
2019-03-27 12:32 ` Ramalingam C
2019-03-27 11:20 ` [igt-dev] ✓ Fi.CI.BAT: success for Enabling PC8+ residency for all GEN9+ platforms (rev3) Patchwork
2019-03-27 20:57 ` [igt-dev] ✓ Fi.CI.IGT: " Patchwork
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=e93cb578-80f4-2893-3fab-644251508fc2@intel.com \
--to=anshuman.gupta@intel.com \
--cc=igt-dev@lists.freedesktop.org \
--cc=ramalingam.c@intel.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox