From: Peter Ujfalusi <peter.ujfalusi@linux.intel.com>
To: lgirdwood@gmail.com, broonie@kernel.org
Cc: alsa-devel@alsa-project.org,
pierre-louis.bossart@linux.intel.com,
ranjani.sridharan@linux.intel.com, kai.vehmanen@linux.intel.com,
rander.wang@intel.com, yung-chuan.liao@linux.intel.com
Subject: [PATCH 1/3] ASoC: SOF: Intel: hda: Do not re-enable L1 if disabled before suspend
Date: Mon, 20 Feb 2023 09:58:02 +0200 [thread overview]
Message-ID: <20230220075804.4829-2-peter.ujfalusi@linux.intel.com> (raw)
In-Reply-To: <20230220075804.4829-1-peter.ujfalusi@linux.intel.com>
From: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
We have a workaround in place to address a known issue with host DMA
running into xruns when capture streams are running. But when resuming
from Sx, we unconditionally re-enable DMI L1 without taking the
workaround into account and this could lead to xruns when a suspended
capture stream is restarted.
To fix this rename the flag l1_support_enabled to l1_disabled in struct
sof_intel_hda_dev to save the L1 disabled status which can be
set/cleared when we get/put a stream and use the flag to determine if DMI
L1 should enabled or not during the post_fw_run op.
Signed-off-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Bard Liao <yung-chuan.liao@linux.intel.com>
Reviewed-by: Rander Wang <rander.wang@intel.com>
Signed-off-by: Peter Ujfalusi <peter.ujfalusi@linux.intel.com>
---
sound/soc/sof/intel/hda-ctrl.c | 8 +++++---
sound/soc/sof/intel/hda-dsp.c | 10 ++++------
sound/soc/sof/intel/hda-stream.c | 10 ++++++++--
sound/soc/sof/intel/hda.h | 2 +-
4 files changed, 18 insertions(+), 12 deletions(-)
diff --git a/sound/soc/sof/intel/hda-ctrl.c b/sound/soc/sof/intel/hda-ctrl.c
index 3aea36c077c9..a1037512da1f 100644
--- a/sound/soc/sof/intel/hda-ctrl.c
+++ b/sound/soc/sof/intel/hda-ctrl.c
@@ -158,16 +158,18 @@ void hda_dsp_ctrl_misc_clock_gating(struct snd_sof_dev *sdev, bool enable)
*/
int hda_dsp_ctrl_clock_power_gating(struct snd_sof_dev *sdev, bool enable)
{
+ struct sof_intel_hda_dev *hda = sdev->pdata->hw_pdata;
u32 val;
/* enable/disable audio dsp clock gating */
val = enable ? PCI_CGCTL_ADSPDCGE : 0;
snd_sof_pci_update_bits(sdev, PCI_CGCTL, PCI_CGCTL_ADSPDCGE, val);
- /* enable/disable DMI Link L1 support */
+ /* disable the DMI link when requested. But enable only if it wasn't disabled previously */
val = enable ? HDA_VS_INTEL_EM2_L1SEN : 0;
- snd_sof_dsp_update_bits(sdev, HDA_DSP_HDA_BAR, HDA_VS_INTEL_EM2,
- HDA_VS_INTEL_EM2_L1SEN, val);
+ if (!enable || !hda->l1_disabled)
+ snd_sof_dsp_update_bits(sdev, HDA_DSP_HDA_BAR, HDA_VS_INTEL_EM2,
+ HDA_VS_INTEL_EM2_L1SEN, val);
/* enable/disable audio dsp power gating */
val = enable ? 0 : PCI_PGCTL_ADSPPGD;
diff --git a/sound/soc/sof/intel/hda-dsp.c b/sound/soc/sof/intel/hda-dsp.c
index 68eb06f13a1f..e3b69dbc1308 100644
--- a/sound/soc/sof/intel/hda-dsp.c
+++ b/sound/soc/sof/intel/hda-dsp.c
@@ -776,7 +776,7 @@ int hda_dsp_resume(struct snd_sof_dev *sdev)
}
/* restore L1SEN bit */
- if (hda->l1_support_changed)
+ if (hda->l1_disabled)
snd_sof_dsp_update_bits(sdev, HDA_DSP_HDA_BAR,
HDA_VS_INTEL_EM2,
HDA_VS_INTEL_EM2_L1SEN, 0);
@@ -868,11 +868,9 @@ int hda_dsp_suspend(struct snd_sof_dev *sdev, u32 target_state)
}
/* enable L1SEN to make sure the system can enter S0Ix */
- hda->l1_support_changed =
- snd_sof_dsp_update_bits(sdev, HDA_DSP_HDA_BAR,
- HDA_VS_INTEL_EM2,
- HDA_VS_INTEL_EM2_L1SEN,
- HDA_VS_INTEL_EM2_L1SEN);
+ if (hda->l1_disabled)
+ snd_sof_dsp_update_bits(sdev, HDA_DSP_HDA_BAR, HDA_VS_INTEL_EM2,
+ HDA_VS_INTEL_EM2_L1SEN, HDA_VS_INTEL_EM2_L1SEN);
/* stop the CORB/RIRB DMA if it is On */
hda_codec_suspend_cmd_io(sdev);
diff --git a/sound/soc/sof/intel/hda-stream.c b/sound/soc/sof/intel/hda-stream.c
index 7f0fd05a96e6..d96d9cd9e62f 100644
--- a/sound/soc/sof/intel/hda-stream.c
+++ b/sound/soc/sof/intel/hda-stream.c
@@ -182,6 +182,7 @@ int hda_dsp_stream_spib_config(struct snd_sof_dev *sdev,
struct hdac_ext_stream *
hda_dsp_stream_get(struct snd_sof_dev *sdev, int direction, u32 flags)
{
+ struct sof_intel_hda_dev *hda = sdev->pdata->hw_pdata;
struct hdac_bus *bus = sof_to_bus(sdev);
struct sof_intel_hda_stream *hda_stream;
struct hdac_ext_stream *hext_stream = NULL;
@@ -222,10 +223,12 @@ hda_dsp_stream_get(struct snd_sof_dev *sdev, int direction, u32 flags)
* Workaround to address a known issue with host DMA that results
* in xruns during pause/release in capture scenarios.
*/
- if (!(flags & SOF_HDA_STREAM_DMI_L1_COMPATIBLE))
+ if (!(flags & SOF_HDA_STREAM_DMI_L1_COMPATIBLE)) {
snd_sof_dsp_update_bits(sdev, HDA_DSP_HDA_BAR,
HDA_VS_INTEL_EM2,
HDA_VS_INTEL_EM2_L1SEN, 0);
+ hda->l1_disabled = true;
+ }
return hext_stream;
}
@@ -233,6 +236,7 @@ hda_dsp_stream_get(struct snd_sof_dev *sdev, int direction, u32 flags)
/* free a stream */
int hda_dsp_stream_put(struct snd_sof_dev *sdev, int direction, int stream_tag)
{
+ struct sof_intel_hda_dev *hda = sdev->pdata->hw_pdata;
struct hdac_bus *bus = sof_to_bus(sdev);
struct sof_intel_hda_stream *hda_stream;
struct hdac_ext_stream *hext_stream;
@@ -264,9 +268,11 @@ int hda_dsp_stream_put(struct snd_sof_dev *sdev, int direction, int stream_tag)
spin_unlock_irq(&bus->reg_lock);
/* Enable DMI L1 if permitted */
- if (dmi_l1_enable)
+ if (dmi_l1_enable) {
snd_sof_dsp_update_bits(sdev, HDA_DSP_HDA_BAR, HDA_VS_INTEL_EM2,
HDA_VS_INTEL_EM2_L1SEN, HDA_VS_INTEL_EM2_L1SEN);
+ hda->l1_disabled = false;
+ }
if (!found) {
dev_err(sdev->dev, "%s: stream_tag %d not opened!\n",
diff --git a/sound/soc/sof/intel/hda.h b/sound/soc/sof/intel/hda.h
index 45f9d4248f14..0679bebe4ad7 100644
--- a/sound/soc/sof/intel/hda.h
+++ b/sound/soc/sof/intel/hda.h
@@ -502,7 +502,7 @@ struct sof_intel_hda_dev {
u32 stream_max;
/* PM related */
- bool l1_support_changed;/* during suspend, is L1SEN changed or not */
+ bool l1_disabled;/* is DMI link L1 disabled? */
/* DMIC device */
struct platform_device *dmic_dev;
--
2.39.2
next prev parent reply other threads:[~2023-02-20 8:03 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-02-20 7:58 [PATCH 0/3] SoC: SOF: Intel: hda/mtl: Improve and enable DMI L1 haldling Peter Ujfalusi
2023-02-20 7:58 ` Peter Ujfalusi [this message]
2023-02-20 7:58 ` [PATCH 2/3] ASoC: SOF: Intel: hda: Restrict DMI L1 disable workaround Peter Ujfalusi
2023-02-20 7:58 ` [PATCH 3/3] ASoC: SOF: Intel: MTL: Enable DMI L1 Peter Ujfalusi
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=20230220075804.4829-2-peter.ujfalusi@linux.intel.com \
--to=peter.ujfalusi@linux.intel.com \
--cc=alsa-devel@alsa-project.org \
--cc=broonie@kernel.org \
--cc=kai.vehmanen@linux.intel.com \
--cc=lgirdwood@gmail.com \
--cc=pierre-louis.bossart@linux.intel.com \
--cc=rander.wang@intel.com \
--cc=ranjani.sridharan@linux.intel.com \
--cc=yung-chuan.liao@linux.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