From: "Naladala, Ramanaidu" <Ramanaidu.naladala@intel.com>
To: <intel-gfx@lists.freedesktop.org>
Subject: Re: [PATCH v5] drm/i915/dmc: Create debugfs entry for dc6 counter
Date: Wed, 12 Mar 2025 23:44:20 +0530 [thread overview]
Message-ID: <29cd2d2d-5c10-4188-82e7-be83590426a1@intel.com> (raw)
In-Reply-To: <Z9GjYgbLPQA0ZM9b@ideak-desk.fi.intel.com>
Hi Imre/Thasleem,
On 3/12/2025 8:38 PM, Imre Deak wrote:
> On Wed, Mar 12, 2025 at 08:13:10PM +0530, Mohammed Thasleem wrote:
>> Starting from MTL we don't have a platform agnostic way to validate
>> DC6 state due to dc6 counter has been removed to validate DC state.
>>
>> The goal is to validate that the display HW can reach the DC6 power
>> state. There is no HW DC6 residency counter (and there wasn't such
>> a counter earlier either), so an alternative way is required. According
>> to the HW team the display driver has programmed everything correctly in
>> order to allow the DC6 power state if the DC5 power state is reached
>> (indicated by the HW DC5 residency counter incrementing) and DC6 is
>> enabled by the driver.
>>
>> Driver could take a snapshot of the DC5 residency counter right
>> after it enables DC6 (dc5_residency_start) and increment the SW
>> DC6 residency counter right before it disables DC6 or when user space
>> reads the DC6 counter. So the driver would update the counter at these
>> two points in the following way:
>> dc6_residency_counter += dc5_current_count - dc5_start_count
>>
>> v2: Update the discription. (Imre)
>> Read dc5 count during dc6 enable and disable then and update
>> dc6 residency counter. (Imre)
>> Remove variable from dmc structure. (Jani)
>> Updated the subject title.
>> v3: Add i915_power_domains lock to updated dc6 count in debugfs. (Imre)
>> Use flags to check dc6 enable/disable states. (Imre)
>> Move the display version check and counter read/update to
>> a helper. (Imre)
>> Resize the variable length. (Rodrigo)
>> Use old dc6 debugfs entry for every platform. (Rodrigo)
>> v4: Remove superfluous whitespace. (Jani)
>> Read DMC registers in intel_dmc.c (Jani)
>> Rename dc6_en_dis to dc6_enabled and change its type to bool. (Jani)
>> Rename update_dc6_count and move it to intel_dmc.c (Jani)
>> Rename dc6_en_dis to start_tracking. (Imre)
>> Have lock for dc6 state read aswelll. (Imre)
>> Keep the existing way print 'DC5 -> DC6 count' along with
>> new 'DC6 Allowed Count' print. (Imre)
>> Add counters in intel_dmc struct. (Imre)
>> Have interface to return dc6 allowed count. (Imre)
>> Rename dc6_count to dc6_allowed_count. (Rodrigo)
>> v5: Rename counters and move in to dc6_allowed structure. (Imre)
>> Order declaration lines in decreasing line length. (Imre)
>> Update start_tacking logic. (Imre)
>> Move get couner inside lock and DISPLAY_VER code to helper. (Imre)
>>
>> Signed-off-by: Mohammed Thasleem <mohammed.thasleem@intel.com>
>> ---
>> .../i915/display/intel_display_power_well.c | 11 +++-
>> drivers/gpu/drm/i915/display/intel_dmc.c | 54 +++++++++++++++++--
>> drivers/gpu/drm/i915/display/intel_dmc.h | 1 +
>> 3 files changed, 62 insertions(+), 4 deletions(-)
>>
>> diff --git a/drivers/gpu/drm/i915/display/intel_display_power_well.c b/drivers/gpu/drm/i915/display/intel_display_power_well.c
>> index 8ec87ffd87d2..574352439573 100644
>> --- a/drivers/gpu/drm/i915/display/intel_display_power_well.c
>> +++ b/drivers/gpu/drm/i915/display/intel_display_power_well.c
>> @@ -752,8 +752,9 @@ void gen9_sanitize_dc_state(struct intel_display *display)
>> void gen9_set_dc_state(struct intel_display *display, u32 state)
>> {
>> struct i915_power_domains *power_domains = &display->power.domains;
>> - u32 val;
>> + bool dc6_was_enabled, enable_dc6;
>> u32 mask;
>> + u32 val;
>>
>> if (!HAS_DISPLAY(display))
>> return;
>> @@ -772,11 +773,19 @@ void gen9_set_dc_state(struct intel_display *display, u32 state)
>> drm_err(display->drm, "DC state mismatch (0x%x -> 0x%x)\n",
>> power_domains->dc_state, val & mask);
>>
>> + enable_dc6 = state & DC_STATE_EN_UPTO_DC6;
>> + dc6_was_enabled = val & DC_STATE_EN_UPTO_DC6;
>> + if (!dc6_was_enabled && enable_dc6)
>> + intel_dmc_update_dc6_allowed_count(display, true);
>> +
>> val &= ~mask;
>> val |= state;
>>
>> gen9_write_dc_state(display, val);
>>
>> + if (!enable_dc6 && dc6_was_enabled)
>> + intel_dmc_update_dc6_allowed_count(display, false);
>> +
>> power_domains->dc_state = val & mask;
>> }
>>
>> diff --git a/drivers/gpu/drm/i915/display/intel_dmc.c b/drivers/gpu/drm/i915/display/intel_dmc.c
>> index fa6944e55d95..95abeec0586a 100644
>> --- a/drivers/gpu/drm/i915/display/intel_dmc.c
>> +++ b/drivers/gpu/drm/i915/display/intel_dmc.c
>> @@ -28,6 +28,7 @@
>> #include "i915_drv.h"
>> #include "i915_reg.h"
>> #include "intel_de.h"
>> +#include "intel_display_power_well.h"
>> #include "intel_dmc.h"
>> #include "intel_dmc_regs.h"
>> #include "intel_step.h"
>> @@ -57,6 +58,10 @@ struct intel_dmc {
>> const char *fw_path;
>> u32 max_fw_size; /* bytes */
>> u32 version;
>> + struct {
>> + u32 dc5_start;
>> + u32 count;
>> + } dc6_allowed;
>> struct dmc_fw_info {
>> u32 mmio_count;
>> i915_reg_t mmioaddr[20];
>> @@ -1232,6 +1237,43 @@ void intel_dmc_snapshot_print(const struct intel_dmc_snapshot *snapshot, struct
>> DMC_VERSION_MINOR(snapshot->version));
>> }
>>
>> +void intel_dmc_update_dc6_allowed_count(struct intel_display *display,
>> + bool start_tracking)
>> +{
>> + struct intel_dmc *dmc = display_to_dmc(display);
>> + u32 dc5_cur_count;
>> +
>> + if (DISPLAY_VER(dmc->display) < 14)
>> + return;
>> +
>> + dc5_cur_count = intel_de_read(dmc->display, DG1_DMC_DEBUG_DC5_COUNT);
>> +
>> + if (!start_tracking)
>> + dmc->dc6_allowed.count += dc5_cur_count - dmc->dc6_allowed.dc5_start;
>> +
>> + dmc->dc6_allowed.dc5_start = dc5_cur_count;
>> +}
>> +
>> +static u32 intel_dmc_get_dc6_allowed_count(struct intel_display *display, u32 *count)
> The return type isn't compatible with the -ENODEV returned value. I'd
> just return a bool, since the reason for an error is always the same.
>
>> +{
>> + struct i915_power_domains *power_domains = &display->power.domains;
>> + struct intel_dmc *dmc = display_to_dmc(display);
>> +
>> + if (DISPLAY_VER(display) < 14)
>> + return -ENODEV;
>> +
>> + mutex_lock(&power_domains->lock);
>> + bool dc6_enabled = DC_STATE_EN_UPTO_DC6 &
>> + intel_de_read(display, DC_STATE_EN);
The dc6_enabled flag indicates only the DC state limit. If all
conditions are met, the DMC can entry/exits DC6.
However, if the DC6 conditions are not met, the DMC can perform
entry/exits up to DC5. Entry/exits from
DC5 to DC3 can also change the DG1_DMC_DEBUG_DC5_COUNT counter values.
It is better to add a pc10 check
along with the dc6_enabled flag.
Correct me if my understanding is wrong.
>> + if (dc6_enabled)
>> + intel_dmc_update_dc6_allowed_count(display, false);
>> +
>> + *count = dmc->dc6_allowed.count;
>> + mutex_unlock(&power_domains->lock);
>> +
>> + return 0;
>> +}
>> +
>> static int intel_dmc_debugfs_status_show(struct seq_file *m, void *unused)
>> {
>> struct intel_display *display = m->private;
>> @@ -1239,6 +1281,7 @@ static int intel_dmc_debugfs_status_show(struct seq_file *m, void *unused)
>> struct intel_dmc *dmc = display_to_dmc(display);
>> intel_wakeref_t wakeref;
>> i915_reg_t dc5_reg, dc6_reg = INVALID_MMIO_REG;
>> + u32 dc6_allowed_count;
>>
>> if (!HAS_DMC(display))
>> return -ENODEV;
>> @@ -1287,9 +1330,14 @@ static int intel_dmc_debugfs_status_show(struct seq_file *m, void *unused)
>> }
>>
>> seq_printf(m, "DC3 -> DC5 count: %d\n", intel_de_read(display, dc5_reg));
>> - if (i915_mmio_reg_valid(dc6_reg))
>> - seq_printf(m, "DC5 -> DC6 count: %d\n",
>> - intel_de_read(display, dc6_reg));
>> +
>> + if (intel_dmc_get_dc6_allowed_count(display, &dc6_allowed_count) == 0) {
>> + seq_printf(m, "DC6 Allowed Count : %d\n", dc6_allowed_count);
> "DC5 -> DC6 allowed count: %d\n" would align better with the other
> entries printed.
>
>> + } else {
>> + if (i915_mmio_reg_valid(dc6_reg)) {
> The above else/if could be combined to 'else if (i915_mmio_reg_valid(...))'
> and the {} could be dropped from around both resultant if branches.
>
> With the above fixed the patch looks ok to me:
>
> Reviewed-by: Imre Deak <imre.deak@intel.com>
>
>> + seq_printf(m, "DC5 -> DC6 count: %d\n",
>> + intel_de_read(display, dc6_reg)); }
>> + }
>>
>> seq_printf(m, "program base: 0x%08x\n",
>> intel_de_read(display, DMC_PROGRAM(dmc->dmc_info[DMC_FW_MAIN].start_mmioaddr, 0)));
>> diff --git a/drivers/gpu/drm/i915/display/intel_dmc.h b/drivers/gpu/drm/i915/display/intel_dmc.h
>> index 44cecef98e73..c78426eb4cd5 100644
>> --- a/drivers/gpu/drm/i915/display/intel_dmc.h
>> +++ b/drivers/gpu/drm/i915/display/intel_dmc.h
>> @@ -26,6 +26,7 @@ void intel_dmc_debugfs_register(struct intel_display *display);
>>
>> struct intel_dmc_snapshot *intel_dmc_snapshot_capture(struct intel_display *display);
>> void intel_dmc_snapshot_print(const struct intel_dmc_snapshot *snapshot, struct drm_printer *p);
>> +void intel_dmc_update_dc6_allowed_count(struct intel_display *display, bool start_tracking);
>>
>> void assert_dmc_loaded(struct intel_display *display);
>>
>> --
>> 2.43.0
>>
next prev parent reply other threads:[~2025-03-12 18:14 UTC|newest]
Thread overview: 68+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-02-03 8:56 [PATCH] drm/i915/dmc: Add debugfs for dc6 counter Mohammed Thasleem
2025-02-03 9:23 ` Jani Nikula
2025-02-03 15:46 ` Rodrigo Vivi
2025-02-03 12:43 ` Imre Deak
2025-02-03 13:39 ` Gustavo Sousa
2025-02-03 14:26 ` Imre Deak
2025-02-03 14:59 ` Gustavo Sousa
2025-02-03 15:14 ` Imre Deak
2025-02-03 15:45 ` Rodrigo Vivi
2025-02-03 16:01 ` Imre Deak
2025-02-03 16:12 ` Rodrigo Vivi
2025-02-03 16:27 ` Imre Deak
2025-02-03 16:42 ` Rodrigo Vivi
2025-02-03 16:51 ` Imre Deak
2025-02-03 17:15 ` Rodrigo Vivi
2025-02-03 19:22 ` Imre Deak
2025-02-03 20:19 ` Gustavo Sousa
2025-02-03 20:23 ` Vivi, Rodrigo
2025-02-03 20:40 ` Gustavo Sousa
2025-02-03 20:59 ` Vivi, Rodrigo
2025-02-03 21:18 ` Gustavo Sousa
2025-02-04 18:10 ` Imre Deak
2025-02-04 17:15 ` Imre Deak
2025-02-03 16:37 ` Gustavo Sousa
2025-02-03 16:49 ` Imre Deak
2025-02-03 17:15 ` ✗ Fi.CI.CHECKPATCH: warning for " Patchwork
2025-02-03 17:15 ` ✗ Fi.CI.SPARSE: " Patchwork
2025-02-03 17:31 ` ✓ i915.CI.BAT: success " Patchwork
2025-02-03 20:06 ` ✗ i915.CI.Full: failure " Patchwork
2025-02-12 11:49 ` [PATCH v2] drm/i915/dmc: Create debugfs entry " Mohammed Thasleem
2025-02-19 1:33 ` [v2] " Almahallawy, Khaled
2025-02-21 17:53 ` Rodrigo Vivi
2025-02-21 17:49 ` [PATCH v2] " Rodrigo Vivi
2025-02-21 18:35 ` Imre Deak
2025-02-21 18:44 ` Imre Deak
2025-02-12 14:17 ` ✗ Fi.CI.CHECKPATCH: warning for drm/i915/dmc: Add debugfs for dc6 counter (rev2) Patchwork
2025-02-12 14:17 ` ✗ Fi.CI.SPARSE: " Patchwork
2025-02-12 14:51 ` ✓ i915.CI.BAT: success " Patchwork
2025-02-12 21:47 ` ✗ i915.CI.Full: failure " Patchwork
2025-03-03 19:23 ` [PATCH v3] drm/i915/dmc: Create debugfs entry for dc6 counter Mohammed Thasleem
2025-03-04 8:32 ` Jani Nikula
2025-03-04 8:33 ` Jani Nikula
2025-03-04 11:00 ` Imre Deak
2025-03-04 12:16 ` Jani Nikula
2025-03-04 12:22 ` Imre Deak
2025-03-09 8:10 ` [PATCH v4] " Mohammed Thasleem
2025-03-10 15:04 ` Imre Deak
2025-03-10 15:12 ` Imre Deak
2025-03-03 21:23 ` ✗ Fi.CI.CHECKPATCH: warning for drm/i915/dmc: Add debugfs for dc6 counter (rev3) Patchwork
2025-03-03 21:23 ` ✗ Fi.CI.SPARSE: " Patchwork
2025-03-03 21:42 ` ✗ i915.CI.BAT: failure " Patchwork
2025-03-09 9:28 ` ✗ i915.CI.BAT: failure for drm/i915/dmc: Add debugfs for dc6 counter (rev4) Patchwork
2025-03-12 14:43 ` [PATCH v5] drm/i915/dmc: Create debugfs entry for dc6 counter Mohammed Thasleem
2025-03-12 15:08 ` Imre Deak
2025-03-12 18:14 ` Naladala, Ramanaidu [this message]
2025-03-12 18:49 ` Imre Deak
2025-03-12 19:32 ` Naladala, Ramanaidu
2025-03-12 20:06 ` Imre Deak
2025-03-12 16:48 ` ✗ i915.CI.BAT: failure for drm/i915/dmc: Add debugfs for dc6 counter (rev5) Patchwork
2025-03-12 20:30 ` [PATCH v6] drm/i915/dmc: Create debugfs entry for dc6 counter Mohammed Thasleem
2025-03-12 20:54 ` [PATCH v7] " Mohammed Thasleem
2025-03-13 10:51 ` Jani Nikula
2025-03-12 21:17 ` ✗ i915.CI.BAT: failure for drm/i915/dmc: Add debugfs for dc6 counter (rev6) Patchwork
2025-03-12 21:52 ` ✗ i915.CI.BAT: failure for drm/i915/dmc: Add debugfs for dc6 counter (rev7) Patchwork
2025-03-14 14:56 ` ✓ i915.CI.BAT: success " Patchwork
2025-03-14 15:24 ` Patchwork
2025-03-19 11:48 ` ✓ i915.CI.Full: " Patchwork
2025-03-19 15:14 ` 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=29cd2d2d-5c10-4188-82e7-be83590426a1@intel.com \
--to=ramanaidu.naladala@intel.com \
--cc=intel-gfx@lists.freedesktop.org \
/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