When the system is suspended while a PCM is paused, it doesn't receive
the SUSPEND trigger. So, the SOF driver has to ensure that the PCM and
the widgets associated with the paused PCM are freed in the firmware
during suspend. This is handled in the
sof_tear_down_left_over_pipelines() call. But since the state of this
PCM is SUSPENDED, we end up clearing the prepared flag for the PCM
before freeing it. This results in IPC errors while freeing the widgets.
But because the widget use_counts are reset to 0 even though the IPC
fails, releasing the paused stream after resuming from suspend proceeds
normally.
Fix the IPC errors by removing the clearing of the prepared flag in
sof_set_hw_params_upon_resume(). In fact, we can remove the
sof_set_hw_params_upon_resume() and call
snd_sof_dsp_hw_params_upon_resume() directly. This will ensure that the
PCM is freed in the firmware before the IPC's for freeing the widgets
are sent.
BugLink: https://github.com/thesofproject/linux/issues/3543
Reviewed-by: Bard Liao <yung-chuan.liao@linux.intel.com>
Reviewed-by: Péter Ujfalusi <peter.ujfalusi@linux.intel.com>
Signed-off-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/20220421203201.1550328-2-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Hi,
The aim of this series is to clean up, make it easier to interpret and less
'chatty' prints aimed for debugging errors.
For example currently the DSP/IPC dump is printed every time we have an IPC
timeout and it is posible to lost the first and more indicative dump to find the
rootcause.
Regards,
Peter
---
Peter Ujfalusi (18):
ASoC: SOF: debug: Swap the dsp_dump and ipc_dump sequence for
fw_exception
ASoC: SOF: ipc and dsp dump: Add markers for better visibility
ASoC: SOF: Print the dbg_dump and ipc_dump once to reduce kernel log
noise
ASoC: SOF: loader: Print the DSP dump if boot fails
ASoC: SOF: intel: atom: No need to do a DSP dump in atom_run()
ASoC: SOF: debug/ops: Move the IPC and DSP dump functions out from the
header
ASoC: SOF: debug: Add SOF_DBG_DUMP_OPTIONAL flag for DSP dumping
ASoC: SOF: intel: hda-loader: Use snd_sof_dsp_dbg_dump() for DSP dump
ASoC: SOF: Drop SOF_DBG_DUMP_FORCE_ERR_LEVEL and sof_dev_dbg_or_err
ASoC: SOF: debug: Print out the fw_state along with the DSP dump
ASoC: SOF: ipc: Re-enable dumps after successful IPC tx
ASoC: SOF: ops: Force DSP panic dumps to be printed
ASoC: SOF: Introduce macro to set the firmware state
ASoC: SOF: intel: hda: Drop 'error' prefix from error dump functions
ASoC: SOF: core: Clean up snd_sof_get_status() prints
ASoC: SOF: loader: Drop SOF_DBG_DUMP_REGS flag when firmware start
fails
ASoC: SOF: Intel: hda-loader: Drop SOF_DBG_DUMP_REGS flag from
dbg_dump calls
ASoC: SOF: Intel: hda: Dump registers and stack when SOF_DBG_DUMP_REGS
is set
Pierre-Louis Bossart (1):
ASoC: SOF: core: debug: force all processing on primary core
sound/soc/sof/core.c | 24 ++++++-------
sound/soc/sof/debug.c | 61 ++++++++++++++++++++++++++++++--
sound/soc/sof/intel/atom.c | 5 +--
sound/soc/sof/intel/hda-loader.c | 11 +++---
sound/soc/sof/intel/hda.c | 16 +++------
sound/soc/sof/ipc.c | 10 ++++--
sound/soc/sof/loader.c | 11 ++++--
sound/soc/sof/ops.c | 3 ++
sound/soc/sof/ops.h | 12 +------
sound/soc/sof/pm.c | 6 ++--
sound/soc/sof/sof-priv.h | 31 ++++++++++------
sound/soc/sof/topology.c | 6 ++++
12 files changed, 131 insertions(+), 65 deletions(-)
--
2.33.0
The DSP device substates such as D0I0/D0I3
are platform-specific. Therefore, the d0_substate
field of struct snd_sof_dev is replaced
with the dsp_power_state field which represents the current
state of the DSP. This field holds both the device state
and the platform-specific substate values.
With the DSP device substates being platform-specific,
the DSP power state transitions need to be performed in
the platform-specific suspend/resume ops as well.
In order to achieve this, the ops signature has to be
modified to pass the target device state as an
argument. The target substate will be determined by
the platform-specific ops before performing the transition.
For example, in the case of the system suspending to S0IX,
the top-level SOF device suspend callback needs to
only determine if the DSP will be entering
D3 or remain in D0. The target substate in case the device
needs to remain in D0 (D0I0 or D0I3) will be determined
by the platform-specific suspend op.
With the addition of the extended set of power states for the DSP,
the set_power_state op for HDA platforms has to be extended
to handle only the appropriate state transitions. So, the
implementation for the Intel HDA platforms is also modified.
Signed-off-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/20200129220726.31792-6-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Add a new enum sof_dsp_power_states for all the possible
the DSP device states. The SOF driver currently handles
only the D0 and D3 states and support for other states
will be added later as needed.
Also, add a helper to determine the target DSP power state
based on the system suspend target.
The snd_sof_dsp_d0i3_on_suspend() function is renamed to
snd_sof_stream_suspend_ignored() to be more indicative
of what it does and it used to determine the target
DSP state during system suspend.
Signed-off-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/20200129220726.31792-5-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Add a state machine for FW boot to track the
different stages of FW boot and replace the boot_complete
field with fw_state field in struct snd_sof_dev.
This will be used to determine the actions to be performed
during system suspend.
One of the main motivations for adding this change is the
fact that errors during the top-level SOF device probe cannot
be propagated and therefore suspending the SOF device normally
during system suspend could potentially run into errors.
For example, with the current flow, if the FW boot failed
for some reason and the system suspends, the SOF device
suspend could fail because the CTX_SAVE IPC would be attempted
even though the FW never really booted successfully causing it
to time out. Another scenario that the state machine fixes
is when the runtime suspend for the SOF device fails and
the DSP is powered down nevertheless, the CTX_SAVE IPC during
system suspend would timeout because the DSP is already
powered down.
Reviewed-by: Curtis Malainey <cujomalainey@chromium.org>
Reviewed-by: Daniel Baluta <daniel.baluta@nxp.com>
Signed-off-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/20191218002616.7652-2-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Move all the audio-specific code in the core,
audio-specific logic in the top-level PM callbacks
and the core header files into a separate file
(sof-audio.*) in preparation for adding an
audio client device.
In the process of moving all structure definitions
for widget, routes, pcm's etc, the snd_sof_dev
member in all these structs is replaced with
the snd_soc_component member. Also, use the component
device instead of the snd_sof_dev device wherever
possible in the PCM component driver,
control IO functions and the topology parser as the
component device will be moved over to the client
device later on.
Signed-off-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/20191204211556.12671-9-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
When application goes through SUSPEND/STOP->PREPARE->START
cycle, we should always reprogram the SOF device to start
DMA from a known state so that hw_ptr/appl_ptrs remain valid.
This is expected by ALSA core as it resets the buffer
state as part of prepare (see snd_pcm_do_prepare()).
Fix the issue by forcing reconfiguration of the FW with
STREAM_PCM_PARAMS in prepare(). Use combined logic to handle
prepare and the existing flow to reprogram hw-params after
system suspend.
Without the fix, first call to pcm pointer() will return
an invalid hw_ptr and application may immediately observe XRUN
status, unless "start_threshold" SW parameter is set to maximum
value by the application.
Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/20190722141402.7194-3-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Paused streams do not get suspended when the system enters S3.
So, clear and release link DMA channel for such streams in the
hda_dsp_set_hw_params_upon_resume() callback. Also, invalidate
the link DMA channel in the DAI config before restoring the
dai config upon resume. Also, modify the signature for the
set_hw_params_upon_resume() op to return an int.
Signed-off-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
As part of the suspend flow, a context save IPC message is
sent to the firmware before powering down the DSP. If errors
are met, the suspend flow is aborted with current code.
Change the behaviour such that if firmware returns -EBUSY or
-EAGAIN, return the error codes to PM core as before. The device
is left in active state in this case.
If other errors are reported, print a warning but do not block the
suspend flow. As per interface specification, no valid error can be
returned in this scenario. If the hardware has hit a fatal error and
is not able to respond successfully, best recovery method is to
proceed with suspend and power off the DSP.
Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
The prepare() ioctl for BE dai link gets called both
when the stream is started and when it is resumed from
suspend. SOF uses this ioctl to set the hw params
again only if the stream has been suspended.
When the stream is started, the hw_params ioctl gets called
before prepare() and hw_params is set for the BE dai link.
So the prepare call does not need to do anything further.
When the stream resumes after system suspend, SOF requires
that the hw_params be set again for the BE dai. In order
to determine which streams should set the hw params
during prepare(), an internal flag called "hw_params_upon_resume"
is introduced in struct sof_intel_hda_stream. The flag is set
for hda streams when the sof device suspends and is
cleared after hw_params is set.
Signed-off-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Add support for saving and restoring DSP context in D3 to host DDR.
The suspend callback includes: suspend all pcm's stream that are running,
send CTX_SAVE ipc, drop all ipc's, release trace dma and then
power off the DSP.
And the resume callback performs the following steps: load FW, run FW,
re-initialize trace, restore pipeline, restore the kcontrol values
and finally send the ctx restore ipc to the dsp.
The streams that are suspended are resumed by the ALSA resume trigger.
If the streams are paused during system suspend, they are marked
explicitly so they can be restored during PAUSE_RELEASE.
Signed-off-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Signed-off-by: Liam Girdwood <liam.r.girdwood@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Mark Brown <broonie@kernel.org>
