Align to the new heap memory map and allocator, smaller SYSTEM and
RUNTIME zones are used on apollolake now.
Signed-off-by: Keyon Jie <yang.jie@linux.intel.com>
Some DAI devices need a delay between their PRE_START and START
trigger commands, and similarly between PRE_RELEASE and RELEASE.
Add a DAI driver operation to get that delay time and use it between
the two commands.
Signed-off-by: Guennadi Liakhovetski <guennadi.liakhovetski@linux.intel.com>
We need to split the START trigger command into two commands because
some components have a long delay inside their START handling. This
patch introduces two new trigger commands: PRE_START and
PRE_RELEASE and a new state PRE_ACTIVE to prepare for that split.
For simmetry POST_STOP and POST_PAUSE are also added, however they
aren't used yet.
Signed-off-by: Guennadi Liakhovetski <guennadi.liakhovetski@linux.intel.com>
From deep down trace.c:va_tracelog() up to the _log_message() level.
Also rename va_tracelog() to the more specific dma_tracelog()
Preparation to support the DMA trace in Zephyr.
The only functional change in this commit is that DMA messages copied to
the shared memory are not de-duplicated any more (a.k.a "adaptive rate
limiting" or CONFIG_TRACE_FILTERING_ADAPTIVE). These are generally
supposed to be high level hence rare enough; otherwise there is probably
a "bigger problem".
Signed-off-by: Marc Herbert <marc.herbert@intel.com>
It is failing and skipped only when compiled on the host, runs fine with
xt-run.
This is temporarily needed to add host-based unit tests to CI now and
catch any regression in any other test now.
Signed-off-by: Marc Herbert <marc.herbert@intel.com>
Fixes 'no previous prototype for trace_log_filtered' -Werr and
redefinition of ‘trace_flush_dma_to_mbox’ error.
Signed-off-by: Marc Herbert <marc.herbert@intel.com>
This change avoids build error:
cc1: warnings being treated as errors
sof/test/cmocka/src/common_mocks.c: In function ‘__panic’:
sof/test/cmocka/src/common_mocks.c💯 warning: ‘noreturn’ function does return
Signed-off-by: Seppo Ingalsuo <seppo.ingalsuo@linux.intel.com>
This enables support for running the allocator mocks with valgrind by
building the allocator for the host library target using a similar
heap map to Intel CAVS targets (memory.c is almost identical copy of CAVS
version).
Signed-off-by: Liam Girdwood <liam.r.girdwood@linux.intel.com>
Pipeline components will be freed by the host before the
pipeline widget itself is freed. No need to walk the pipeline
and free anything during pipeline_free().
Fix the unit tests to remove the tests that are not relevant anymore.
Signed-off-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Pure rename, zero functional change.
The very poorly named trace_flush() function sounds like earlier
scheduling of something that would the same later anyway but that's
absolutely not what it does. Instead it copies pending DMA traces to the
shared mailbox.
As an example, in June 2020, PR #3195 commit 6c14e76c0d ("trace: Log
FW ABI and hash numbers") added a tr_info() "banner" immediately after
to make sure tracing works. That included a likely misunderstood
trace_flush() call immediately after the tr_info().
Signed-off-by: Marc Herbert <marc.herbert@intel.com>
Prepare for supporting new IPC major versions by partitioning
IPC code into directories.
This is a code move only, no code changes except Makefiles.
Signed-off-by: Liam Girdwood <liam.r.girdwood@linux.intel.com>
Several code paths use the existing clock_ms_to_ticks() function in
some not ideal ways, potentially leading to computation overhead or
to precision loss. The function is often called to calculate ticks
for 1 millisecond, which then is recalculated to the required time
interval. It's better to let the function calculate the number of
ticks for the required time interval directly. E.g. instead of
clock_ms_to_ticks(PLATFORM_DEFAULT_CLOCK, 1) * milliseconds
it's better to call
clock_ms_to_ticks(PLATFORM_DEFAULT_CLOCK, milliseconds)
directly. For microseconds however replacing
clock_ms_to_ticks(PLATFORM_DEFAULT_CLOCK, 1) * microseconds / 1000
with
clock_ms_to_ticks(PLATFORM_DEFAULT_CLOCK, microseconds / 1000)
can lead to a loss of precision. To avoid that a new function
clock_us_to_ticks() is added.
Signed-off-by: Guennadi Liakhovetski <guennadi.liakhovetski@linux.intel.com>
Cordic sin cos input value range is [-2*pi to 2*pi]
and output range is [-1 to +1]
This is common function to calculate trignometric sine
and cosine using separate lookup table size for speeds
and accuracy calculation.
For 32bit sine and cosine
Error (max = 0.000000011175871), THD+N = -170.152933
For 16bit sine and cosine
Error (max = 0.000061), THD+N = -91.518584
Signed-off-by: ShriramShastry <malladi.sastry@intel.com>
Make sure component creation is not tightly coupled to a particular IPC
version. This is mostly a mechanical change of structures being passed
to the comp creation APIs away from IPC specific to general structures.
Highlevel changes
1) Pass a common component data object and a component
specific data object during create().
2) Mark the component IPC derived data as "ipc_config" within the
component device to help developers track the data source origin.
3) Pass component specific data during creation so that componets
can allocate and copy to thier private data.
4) Comp_dev no longer has component specific data appended to it.
Instead we can store all in the comp private data (and hence use the
compiler to access it rather than by developer access methods).
Signed-off-by: Liam Girdwood <liam.r.girdwood@linux.intel.com>
Move the code that sets the scheduling comp for
a pipeline to allow setting it in ipc_pipeline_complete().
This removes the restriction that the scheduling comp must
be set up before the pipeline widget and provides the
flexibility in the kernel to set up the widgets in any order
while parsing topology.
Signed-off-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Do not reset component status in the case of xrun.
An irrecoverable xrun will lead to the pipeline being
stopped by the host and resetting the status will result
in an error stating invalid current state during the STOP
trigger.
Also update the unit test to retain the same state during an xrun.
Signed-off-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Allows the mocks to be built only for xtensa GCC. i.e a build test only.
Running on qemu requires more work.
Signed-off-by: Liam Girdwood <liam.r.girdwood@linux.intel.com>
Allows the mocks to be quickly run on the host with full access to host
debug and development tooling.
Signed-off-by: Liam Girdwood <liam.r.girdwood@linux.intel.com>
Reduce the number of functions we mock and use the real functions where
we can. This brings in more runtime code so will increase the test
coverage.
Signed-off-by: Liam Girdwood <liam.r.girdwood@linux.intel.com>
Decouple the IPC ABI major version from pipeline creation. This moves
struct sof_ipc_pipe_new from the pipeline and replaces it with it's
members (saving 8 bytes as no header is needed).
Add new feature specific pipeline APIs to configure pipeline at creation.
Additionally align testbench and UTs to API change.
Signed-off-by: Liam Girdwood <liam.r.girdwood@linux.intel.com>
Add cmocka unit test for FFT library, including fft with size 1024/256,
fft + ifft with SNR test, fft for 2 channels.
Implement helpers to accept samples put in comp_buffer for both real and
complex FFT, the helper fft_real_2() is implemented to speed up the FFT
for dual channels inputs up to two times. All theses helpers should be
used for unit test only.
Suggested-by: Amery Song <chao.song@intel.com>
Signed-off-by: Keyon Jie <yang.jie@linux.intel.com>
Split IPC APIs out by feature so that the IPC layer to help future
IPC infrastructure changes support more than one IPC ABI MAJOR version.
No code changes here, only code partitioning and Doxygen comments.
Signed-off-by: Liam Girdwood <liam.r.girdwood@linux.intel.com>
Several files had multiple blank lines even before the removal of
platform_shared_commit(). Fix them with "cat -s"
Signed-off-by: Liam Girdwood <liam.r.girdwood@linux.intel.com>
Split the pipeline logic up into stream, params, graph, scheduling and xrun
so that it easier to follow and understand. This is to help on the TSC work
to integrate the codec adaptor into core logic.
This PR is all mechanical code moves. There are NO functional changes. Some
function that were static are now public.
Followup patches will add more documentation updates alongside.
Signed-off-by: Liam Girdwood <liam.r.girdwood@linux.intel.com>
Array streams represents streams on "many" side i.e. input
for MUX and output for DEMUX.
For DEMUX each stream has masks array - 1 mask per output
channel. Each mask shows, from which input channel data
should be taken.
This commit reverts "demux" part of commit:
"b1b31e7154a5c159d81459634eabd8013b434181"
Signed-off-by: Bartosz Kokoszko <bartoszx.kokoszko@linux.intel.com>
Input and output channel settings from matrix row and column
are in reverse order, so they should be swapped.
Signed-off-by: Jaska Uimonen <jaska.uimonen@intel.com>
Also create trace_log_unfiltered() function, to be used internally
by tracing subsystem to emit information about suppressed trace messages.
The *unfiltered function could be also used as replacement
in _log_message() macro, to disable log filtering for debugging purposes.
Signed-off-by: Artur Kloniecki <arturx.kloniecki@linux.intel.com>
Assertions don't show proper location for functions implemented
in header file.
Before changes, example location of this assert was:
Location: src/audio/pipeline.c:303
Signed-off-by: Karol Trzcinski <karolx.trzcinski@linux.intel.com>
Shared memory shouldn't be connected with any cpu core,
because in general such an resource may be shared between
any number of arbitrary chosen cores.
Connecting it with with core 0 heap, leads to unbalanced
heap usage between core 0 and secondary cores.
Moreover operations like alloc and free may be called from
two different cores, what's not possible in current implementation.
Such an situation is especially possibly during handling resources
describing hardware components, eg. DAI or DMA, from two different
cores.
Moving shared memory management to dedicated heap, allows to easily
use another memory region, what may be needed to fulfill platform
specific demands.
Signed-off-by: Karol Trzcinski <karolx.trzcinski@linux.intel.com>
Currently pcm converter unit tests use mocked-up audio_stream
implementation which should be replaced with real FW implementation.
Signed-off-by: Artur Kloniecki <arturx.kloniecki@linux.intel.com>
UUID objects are represented by struct sof_uuid_entry instances in SOF.
Instead of casting pointers to them to integers for passing around,
carry them as pointers until they have to be cast to integers for
packing into data structures. This also fixes printing UUID, using a
"%pU" format.
Signed-off-by: Guennadi Liakhovetski <guennadi.liakhovetski@linux.intel.com>
Currently selector unit tests mock audio_stream by hand, which
adds maintenance overhead in case of changes in audio_stream
implementation, which are under way.
Signed-off-by: Artur Kloniecki <arturx.kloniecki@linux.intel.com>
Currently volume unit tests mock buffer components by hand, which
adds maintenance overhead in case of changes in buffer implementation,
which are under way.
Signed-off-by: Artur Kloniecki <arturx.kloniecki@linux.intel.com>
Currently buffer is filled by memcpy_s and verified by memcmp, which
would require more UT maintenance during upcoming buffer changes.
With this change, usage of the buffer better reflects real-life usage.
Signed-off-by: Artur Kloniecki <arturx.kloniecki@linux.intel.com>
All audio_stream clients should use provided audio_stream_get_avail*/free*
API, instead of directly accessing .avail and .free member fields.
Signed-off-by: Artur Kloniecki <arturx.kloniecki@linux.intel.com>
This kconfig collides with one of the same name in Zephyr. Rename it
for clarity ("MULTICORE" was picked for symmetry with the existing
"CORE_COUNT", though it's admittedly a little long...).
Signed-off-by: Andy Ross <andrew.j.ross@intel.com>
This commit simplifies mux processing functions and
does not allow mux to mix channels.
Signed-off-by: Bartosz Kokoszko <bartoszx.kokoszko@linux.intel.com>
Mechanism of comparison pipe_id from ipc_comp_dev will be used in
trace filtering, so this part should be moved to separate function
and allow code reusability.
Moreover function usage is compact and descriptive method to define
what given block of code should do, so after such a refactor code
readability is improved.
Include from pipeline.h is removed to omit circular reference.
Signed-off-by: Karol Trzcinski <karolx.trzcinski@linux.intel.com>
This commit rewrites functions responsible for large config configuration
from detect_test component and puts them in component.c file as
generic ones - as a result they can be used by other components.
Signed-off-by: Bartosz Kokoszko <bartoszx.kokoszko@linux.intel.com>
Ability to change log level per UUID component improve user
experience during debugging firmware. Threshold trace level
defined in log message with value from trace context, where
trece context is related with component instance (local or
global instance).
Signed-off-by: Karol Trzcinski <karolx.trzcinski@linux.intel.com>
UUID key is too detailed information for logging function to allow
easy functionality extension in future. Trace context structure is
designed to work directly with logging function, so there is no
reason to unpack values from this struct in trace macro-functions
and pass them as separate arguments to trace_log().
Signed-off-by: Karol Trzcinski <karolx.trzcinski@linux.intel.com>
Each trace function should refer to trace context, where are
stored information like uuid and runtime editable logs level,
to unify trace API and allow trace filtering in future.
Remove buf_cl_* functions, as no longer supported.
Signed-off-by: Karol Trzcinski <karolx.trzcinski@linux.intel.com>
Used method need floating point unit in DSP core.
Vectorized conversion functions are inaccessible for used cores.
Signed-off-by: Karol Trzcinski <karolx.trzcinski@linux.intel.com>
Respect fixed point data format as Q1.X instead of QX.
After change values in float will be in range <-1;1)
what will correspnd to fixed point data MAX/MIN value
in Q1.X format.
Signed-off-by: Karol Trzcinski <karolx.trzcinski@linux.intel.com>
All trace entries are identified by uuids only now.
Previous TRACE_CLASS_ identification removed completely.
UUIDs are passed to the tracing subsystem inside a trace
context. Each trace context defines run-time log level.
The level is initialized to LOG_LEVEL_INFO (this may be
change per instance is needed) and may be re-configured
at run-time.
Signed-off-by: Marcin Maka <marcin.maka@linux.intel.com>
Actions perfomed by the default handler are required to keep
the component device state consistent (e.g. dev->frames are
initialized). While the current comp_params() silently skips
the component if params are not defined.
New simple components will not be required to define a handler
if no extra steps are needed.
It is also an opportunity to reduce the code of existing components
Note that most of the extra steps are moved to prepare() in many
implementations already (like volume), others may follow and remove
their params() handlers once reduced to a single call to the default
handler (like mux recently did).
Signed-off-by: Marcin Maka <marcin.maka@linux.intel.com>
Functions written with metaprogramming are hard to read and debug,
so newly introduced version should be easier to understand for
new developers. Moreover introduced version is more generic and
compact because of usage single logging function implementation
regardless of given parameters number and usage of atomic or
mbox version.
There is no longer a need to heavy test logging function under
UT with different number of arguments, because of it singular
implementation.
Signed-off-by: Karol Trzcinski <karolx.trzcinski@linux.intel.com>
rrealloc function is no longer in use, so it is better
to remove it. Unused code is no longer tested so its
quality is unknown.
The current implementation is simple but expensive since
extra calls to malloc/free requires time and risky since
at some point requires double allocation of potentially
large object. Please consider re-implementation instead
of restoring.
Signed-off-by: Marcin Maka <marcin.maka@linux.intel.com>
Changes schedule operations to return error in case
chosen scheduler is not found or scheduler doesn't
support mandatory operation.
Signed-off-by: Tomasz Lauda <tomasz.lauda@linux.intel.com>
Previous memcpy of bytes is unsafe if new object is larger
then old one.
Typically realloc retrieves the old size internally and does
not need this information passed as argument but sof heap
implementation does not store the exact size of the allocated
memory buffer.
Signed-off-by: Marcin Maka <marcin.maka@linux.intel.com>
Previous memcpy of bytes is unsafe if new object is larger
then old one (which is probably most popular realloc use).
Typically realloc retrieves the old size internally and does
not need this information passed as argument but sof heap
implementation does not store the exact size of the allocated
memory buffer.
Signed-off-by: Marcin Maka <marcin.maka@linux.intel.com>
Few artifacts of the past were left after dependency removal. This patch
is there to address that.
Signed-off-by: Artur Kloniecki <arturx.kloniecki@linux.intel.com>
In the past, when parameters were held by components, it was impossible
to manage input/output streams with different channels' count.
Now, after params were moved to buffers, we can properly handle such
scenario, which in turn allows for more flexible approach in terms of
runtime configurability.
Signed-off-by: Artur Kloniecki <arturx.kloniecki@linux.intel.com>
(de)mux_get_processing_function() call depended on config->frame_format
being properly configured, which was obsolete from the very beginning.
This patch removed that dependency in favor of reading frame_fmt from
downstream buffer during the call, rendering that field completely
obsolete.
Signed-off-by: Artur Kloniecki <arturx.kloniecki@linux.intel.com>
Fixes:
error: 'err' may be used uninitialized in this function
[-Werror=maybe-uninitialized]
if (err)
^
As I found nothing in the entire code base checking the return value of
comp_register() (the real function seems to always succeed), I also
simplify the main path and pass through the value returned by
memcpy_s().
Signed-off-by: Marc Herbert <marc.herbert@intel.com>
Once the debug version of allocation functions are removed,
macros may be replaced with functions declared without leading
_ which makes the documentation (and the header) more readable,
the defines section does not need to contain doubled documentation
of alloc api.
bzero macro is not esasily removable due to include deps which
bring earlier declaration of non-inline prototype.
Signed-off-by: Marcin Maka <marcin.maka@linux.intel.com>
All clients that just wants to allocate and free memory
does not need to know all the heap management api
and internals.
Signed-off-by: Marcin Maka <marcin.maka@linux.intel.com>
Previously used by kpb tests only and defined locally.
New pieces of the code, like perf macros need them
globally.
Signed-off-by: Marcin Maka <marcin.maka@linux.intel.com>
New is c++ keyword which causes errors while generating
the documentation for component api with sphinx and breathe.
Signed-off-by: Marcin Maka <marcin.maka@linux.intel.com>
As reported by gcc.
Fixes: 0b279e06bc. This is a partial, one-line revert of that commit
that decreased the size of this array for no obvious reason.
Signed-off-by: Marc Herbert <marc.herbert@intel.com>
Add "UPDATE_DISCONNECTED True" property in ExternalProject_Add() so:
- Unit tests can be rebuilt without a network connection
- The code can't change from one build to the next without the developer
being aware.
Found by chance thanks to some network issue.
Signed-off-by: Marc Herbert <marc.herbert@intel.com>
The current content of component.h is a mix of basic api, common
basic helpers for every component developers as well as advanced
functions and macros used by infrastructure and very specialized
components like host, dai, kpb etc.
This patch moves the advanced part to component_ext.h and keeps
only basic part in component.h to avoid overloading effects
component developers with declarations and code they do not use.
Signed-off-by: Marcin Maka <marcin.maka@linux.intel.com>
New trace entry parameter is defined: uuid of the source.
Technically, this is an address of uuid entry linked
inside a special section in elf.
Has_ids is removed from the static log entry since eventually
all entries will use uuids and -1 detection is good enough
detection for those who still do not provide ids.
Note: both changes are done in a single patch to avoid two
major dbg-abi version bumps.
Signed-off-by: Marcin Maka <marcin.maka@linux.intel.com>
Unique identification in form of Guid is much more
scalable for fw parts (components, dais, ...) then
existing ids and trace class ids.
Signed-off-by: Marcin Maka <marcin.maka@linux.intel.com>
Changes implementation of notifications allocation.
Instead of using common empty list of messages let's
switch to allocating message per source of notification.
This way we won't have problem with memory and lack of
empty messages and also we will be able to optimize memory.
Signed-off-by: Tomasz Lauda <tomasz.lauda@linux.intel.com>
This commit moves fetching mux/demux processing functions from
mux_prepare() to mux_set_values(). It insures us against
situation when we fetch proc_func in mux_prepare() and
after that comes mux_set_values() with new mux config.
Signed-off-by: Bartosz Kokoszko <bartoszx.kokoszko@linux.intel.com>
Implements remote operations for params, prepare, trigger
and reset calls. If the target component is running on
different core, we should send IDC in order to execute
appropriate operation.
Signed-off-by: Tomasz Lauda <tomasz.lauda@linux.intel.com>
Implements locking and cache synchronization for buffers,
which are connecting components running on different cores.
We need to explicitly use cache here (if platform supports it),
so the shared API cannot be used in this particular case.
Shared is going through uncached area on cAVS platforms,
but it's too slow in the case of frequent buffer accesses.
Also loop in pipeline_set_params has been changed to avoid
unnecessary cache accesses after unlocking the buffer.
Signed-off-by: Tomasz Lauda <tomasz.lauda@linux.intel.com>
Adds separate task to components, which are supposed to be scheduled
on different core than the rest of the pipeline. These are so called
shared components. This way components requiring huge processing power
can be offloaded to slave cores.
Signed-off-by: Tomasz Lauda <tomasz.lauda@linux.intel.com>
Implements component reallocation in shared memory. This way
cores can share components, which are part of one core's pipeline,
but runs on other core.
Signed-off-by: Tomasz Lauda <tomasz.lauda@linux.intel.com>
Back pointer must be initialized before any instance
level logging is performed by the new() operation since
the instance identification requires access to the uuid
that will be implemented on the component driver level.
Signed-off-by: Marcin Maka <marcin.maka@linux.intel.com>
Static inline getters provides better control of use
component properties as L- and R-values.
They also make changes to the common component data defs
easier and transparent for clients (no code changes, just
recompilation required).
Signed-off-by: Marcin Maka <marcin.maka@linux.intel.com>
Implements comp_verify_params() function, which can be used in order
to set and update some of sink (playback) or source (capture) buffer
parameters with pcm parameters. There is a possibility to specify
which of parameters won't be overwritten (e.g. SRC component should
not overwrite rate parameter, because it is able to change it).
Commit also adds macros to buffer.h, which can be used in order
to specify which parameter of source/sink buffer should not be
updated.
Signed-off-by: Bartosz Kokoszko <bartoszx.kokoszko@linux.intel.com>
This patch refactor KPBs macros:
KPB_SAMPLING_WIDTH to KPB_SAMPLES_PER_MS as this is
exactly what this macro stands for - the amount of
samples we take each milisecond.
KPB_NR_OF_CHANNELS to KPB_NUM_OF_CHANNELS as the later
is more meaningful
Signed-off-by: Marcin Rajwa <marcin.rajwa@linux.intel.com>
When new component will need to process audio data in two steps,
then temporary circular buffer will be needed. Also extra information
about data stream may be needed to choose right algorithm or assert data
compatibility.
Currently only 'comp_buffer' has such a possibility but it is component
defined in topology - with much wider capabilities - and shouldn't be
created inside component just to hold temporary data stream.
Use introduced structure in processing functions, move api responsible
for data processing from buffer to audio_stream module.
Signed-off-by: Karol Trzcinski <karolx.trzcinski@linux.intel.com>
Link common_mock to each test by default.
Removed duplicated mocks in component specific folders.
Keep old mocks where implementation differs from
implementation in common_moks.c
Signed-off-by: Karol Trzcinski <karolx.trzcinski@linux.intel.com>
In unit tests there are a lot of mocks duplication for mocks related
with memory management and panic routine.
As long as implementation doesn't have any module specific routines
and may be reused, it should be mocked in common mock space.
Marked each mock as WEAK to make it possible to overwrite this
implementation in specific test.
To use assert_false in __panic function, added dependence of
universal_mock from cmock.
Signed-off-by: Karol Trzcinski <karolx.trzcinski@linux.intel.com>
SHARED_DATA macros doesn't put shared data into separate section
when unit tests are built and uncached memory region is not used.
Signed-off-by: Tomasz Lauda <tomasz.lauda@linux.intel.com>
Changes component driver registration to use comp_driver_info
as an element to be put into the drivers list. This way
comp_driver can be constant and comp_driver_info can be
marked as shared in the future.
Signed-off-by: Tomasz Lauda <tomasz.lauda@linux.intel.com>
Keyon Jie