libsamplerate_resample() assumed that src_process() would
always consume the whole input buffer. That was an invalid
assumption leading to crashes.
This patch adds a leftover memchunk for storing any
non-consumed input. When pa_resampler_run() is called next
time, the leftover is prepended to the new input.
Changes in v3:
- Make the calculations in pa_resampler_result() and
pa_resampler_max_block_size() more readable and more
correct.
- Rework the leftover storing: instead of using a dedicated
buffer for it, store it in the beginning of remap_buf.
This can avoid some memory copying. (The idea was
suggested by Wang Xingchao.)
- Use a generic save_leftover() function instead of doing
the leftover copying in the resampler implementation.
- Use the leftover logic also with the speex and ffmpeg
resamplers.
[ed: dropped the speex bit since the API guarantees that
it will consume everything -- Arun]
Changes in v2:
- If add_leftover() is called with zero-length input while
the leftover length is non-zero, we don't try to acquire
the input memblock.
- Instead of taking a reference to the original input in
libsamplerate_resample(), we copy the leftover data to a
new memblock. This is done, because otherwise, if the
input is one of the internal buffers, the data can get
overwritten before reading it in add_leftover().
- Store add_leftover_buf size in bytes instead of samples
(more convenient, but less consistent with other code).
BugLink: https://bugs.freedesktop.org/show_bug.cgi?id=47156
This causes problems with 24kHz audio (results in echoing)
when upscaling to 44.1kHz or 48kHz.
It can be reapplied when the optimisation works for all cases.
This reverts commit 8539fe9765.
make speex library dependency optional, this affects the resampler
and the echo canceller module
this patch supersedes an earlier patch proposal and addresses the following
comments:
* fix order of pa_echo_canceller_method_t enum and ec_table (Frederic)
* the default resampler is speex if available as before, otherwise ffmpeg (Arun)
* does not touch the Adrian EC implementation (see separate patch) (Arun)
The algorithm had been implemented the same way as the trivial resampler. But
an important difference between the two is that the trivial resampler can write
an output as soon as the first corresponding input sample is seen, whereas the
peaks resampler must have read all input samples before writing an output
sample.
With this rework, the peaks resampler now outputs samples correctly when the
input data is spanning multiple memblocks.
The pa_xmalloc calls oom() in case of NULL pointer returned by malloc() on one
hand and dereferencing of pointer is happen early than actual check on other
hand. Thus, just remove useless checks.
Move the channel remapping bits into a separate structure. We'll make this
structure global so that optimized versions can use it to perform the channel
remapping.
Factor out the channel remap matrix code into a separate function.
Keep a pointer to the channel remapping function so we can install custom
functions.
Catch the common mono->stereo remapping case and install a custom, more
optimized function.
Get rid of the liboil dependency and reimplement the liboil functions with an
equivalent C implementation. Note that most of these functions are deprecated in
liboil and that none of them had any optimisations. We can further specialize
our handrolled versions for some extra speedups.
If there's a center channel on input that is not available on output
make sure we mix front-center only into front-left/right and rear-center
into rear-left/right.
Closes#400
