To reproduce, add resampler-method = ffmpeg in daemon.conf
then start PA, and load module-loopback
Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 0xb2f1db40 (LWP 23047)]
0x00000000 in ?? ()
(gdb) bt
0 0x00000000 in ?? ()
1 0xb7c463cb in pa_resampler_set_input_rate (r=0x80e9438, rate=44011) at pulsecore/resampler.c:365
2 0xb7c6321d in pa_sink_input_process_msg (o=0x80e87a0, code=3, userdata=0xabeb, offset=0, chunk=0x0)
at pulsecore/sink-input.c:1833
3 0xb7e9840b in sink_input_process_msg_cb (obj=0x80e87a0, code=3, data=0xabeb, offset=0, chunk=0x0)
at modules/module-loopback.c:538
4 0xb7c2709b in pa_asyncmsgq_dispatch (object=0x80e87a0, code=3, userdata=0xabeb, offset=0, memchunk=0xb2f1d17c)
at pulsecore/asyncmsgq.c:322
5 0xb7c4c6e3 in asyncmsgq_read_work (i=0x80dd580) at pulsecore/rtpoll.c:564
6 0xb7c4b34a in pa_rtpoll_run (p=0x80fb7e0, wait_op=true) at pulsecore/rtpoll.c:238
7 0xb7dd90af in thread_func (userdata=0x80afe88) at modules/alsa/alsa-sink.c:1785
8 0xb7bf3291 in internal_thread_func (userdata=0x8095d08) at pulsecore/thread-posix.c:83
9 0xb7ab9d4c in start_thread (arg=0xb2f1db40) at pthread_create.c:308
10 0xb79f3ace in clone () at ../sysdeps/unix/sysv/linux/i386/clone.S:130
A rewind may erase data that sink_input counted in playing_for or
underrun_for earlier. Add code adjusting those values after a rewind.
One visible symptom of this bug was problems recovering from an
underrun. When a client calls pa_stream_write() with a large block of
memory, the function can split that into smaller pieces before sending
it to the server. When receiving new data for a stream that had
silence queued due to underrun, the server would do a rewind to
replace the queued-but-not-played silence with the new data. Because
of the bug, this rewind itself would not change underrun_for. It's
possible for multiple rewinds to be done without filling the sink
buffer in between (which is what would eventually reset underrun_for).
In this case, the server rapidly processing the split packets would
rewind the stream for _each_ of them (as underrun_for would stay set),
erasing valid audio as a result.
As Peter Meerwald <p.meerwald@bct-electronic.com> discovered, our ARM
svolume code performance is quite terrible when the incoming samples are
not word-aligned. This can very easily be the case, since the
architecture only requires that the samples be 16-bit aligned, and we
might end up running the innermost loop after processing modulo-4
samples. The performance degradation was ~50x on a Cortex A9
(Pandaboard).
This reworks the svolume logic to first consume enough samples to make
sure the rest is word aligned, and reordering the processing to work
with 4 samples at a time first, and then finally deal with the
remainder.
With this, performance is comparable for arbitrary alignments (~3x
faster than the C code).
This fixes at least one crash that has been observed. The
multiplication in trivial_resample() overflowed when
resampling from 96 kHz to 48 kHz, causing an assertion
error:
Assertion 'o_index * fz < pa_memblock_get_length(output->memblock)' failed at pulsecore/resampler.c:1521, function trivial_resample(). Aborting.
Without the assertion, the memcpy() after the assertion
would have overwritten some random heap memory.
When compiling without HAVE_SYMLINK the runtime dir is a real directory,
which is attempted to be created. In the case it already exists we shouldn't
error out. The HAVE_SYMLINK-enabled code already does this.
Currently, Windows versions of pacat and friends fail because the current
poll emulation is not sufficient (it only works for socket fds).
Luckily Gnulib has a much better emulation that seems to work good enough.
The implementation has been largely copied (except a few bug fix
regarding timeout handling, to be pushed upstream) and works on pipes
and files as well. The copy has been obtained through their gnulib-tool utility,
which gives a LGPLv2.1+ licensed file.
This fixes the "Assertion (!e->dead) failed" error coming and lets pacat
and friends stream happily to/from a server (I didn't actually test parec).
Rounding with 0.5 causes us to always round up for any value of the form
x.5. IEEE754 specifies round-to-nearest-even as the behaviour in this
case. This might not always be possible with NEON code, but this change
gets us much closer to it.
final:
* includes some minor style fixes and build-time changes to allow
building a single binary for neon and non-neon systems
v4:
* fix for sample length < 4
v3:
* convert from intrinsics to inline assembly
v2:
* load and store data with vld1/vld1q and vst1/vst1q, resp., to work
around alignment issues of compiler-generated vldmia instruction
* remove redundant check for NEON flags
Ubuntu/Linaro gcc 4.6.3
arm-linux-gnueabi-gcc -O2 -mcpu=cortex-a8 -mfloat-abi=softfp -mfpu=neon
runtime on beagle-xm:
D: [pulseaudio] sconv_neon.c: checking NEON sconv_s16le_from_float
I: [pulseaudio] sconv_neon.c: NEON: 3754 usec.
I: [pulseaudio] sconv_neon.c: ref: 58594 usec.
D: [pulseaudio] sconv_neon.c: checking NEON sconv_s16le_to_float
I: [pulseaudio] sconv_neon.c: NEON: 1831 usec.
I: [pulseaudio] sconv_neon.c: ref: 10528 usec.
I: [pulseaudio] sconv_neon.c: Initialising ARM NEON optimized conversions.
conversion may be off by one for some samples due to rounding issues
CC libpulsecore_2.98_la-svolume_arm.lo
pulsecore/svolume_arm.c: In function 'pa_volume_s16ne_arm':
pulsecore/svolume_arm.c:50:8: warning: assignment discards 'const' qualifier from pointer target type [enabled by default]
Signed-off-by: Peter Meerwald <p.meerwald@bct-electronic.com>
On Windows, strerror can actually return "Unknown Error"
(e.g. for large errnums). The code assumes the return value to be helpful.
Make it slightly more helpful by catching the message and appending the
errnum.
On Windows, fdsem.c:flush() fails because sockets are set to non-blocking
mode, since pa_read() returns -1 (and errno == EWOULDBLOCK). I guess pa_read()
is expected to block in this case so make it actually block by calling poll().
These days we don't set core->default_sink/source as soon as somebody
asks for it. To retain consistent behaviour (i e the asterisk),
we need to call pa_namereg_get_default_sink/source.
Reported-by: Daniel Manrique <daniel.manrique@canonical.com>
Reported-by: Brendan Donegan <brendan.donegan@canonical.com>
Signed-off-by: David Henningsson <david.henningsson@canonical.com>
Nowadays, we are using more hashmaps and other things, than we did
before. Therefore, I often get the "flist is full (don't worry)"
message. This change should avoid that message. I was unable to find
any significance in increase of memory footprint from this change.
Signed-off-by: David Henningsson <david.henningsson@canonical.com>
pa_sink_input_seek() calculates output lenth (slength) and
corresponding input length (ilength). During an underrun, the function
generates slength bytes of silence and adds ilength to the
underrun_for value. However, the ilength value may be shortened to
match resampler limits, and there's no corresponding adjustment to
slength. Thus, the length of the generated silence is longer than
resampler output would have been, and underrun_for should be increased
by more than the limited ilength. This error makes the user-visible
since_underrun field in struct pa_timing_info too small. Fix by using
the original value calculated before limiting in this case.
The problem that the comment mentions doesn't actually
exist, because when the sink latency is changed to a smaller
value, the sink implementor will request the required
rewind.
On ARM, pa_object has less strict alignment requirements
than e.g. pa_sink and pa_source, so when pa_object is cast
to pa_sink, the compiler thinks that it's unsafe. In this
case, however, the pointer given to pa_sink_ref() was a
pa_sink pointer to begin with, so casting it first to
pa_object and then back to pa_sink is entirely safe.
This particular source of warnings is extremely annoying,
because this message is printed for any compilation unit
that includes sink.h, source.h or any other header that
defines a class, and the message tends to get printed
multiple times for one compilation unit:
In file included from ./pulsecore/source-output.h:37:0,
from ./pulsecore/source.h:49,
from ./pulsecore/sink.h:40,
from ./pulsecore/core.h:50,
from daemon/daemon-conf.h:31,
from daemon/cmdline.h:25,
from daemon/cmdline.c:38:
./pulsecore/sink-input.h: In function 'pa_sink_input_ref':
./pulsecore/sink-input.h:245:1: warning: cast increases required alignment of target type [-Wcast-align]
This makes updating of permissions on existing directories optional with
pa_make_secure_dir() and pa_make_secure_parent_dir(). This makes sure
that the recursive directory creation doesn't end up modifying existing
directories, and also fixes a problem where creating an auth cookie
(specifically ~/.esd_auth) would end up modifying permissions on ~.
Thanks to Frédéric Danis for reporting this.
This includes updating the native protocol and the client API.
A new command was added to allow setting the latency offset.
Also the card list command now shows the latency offset if there
are ports available.
Update protocol to 27.
Set the state variable immediately to zero so if we fail to open the
configuration file we don't check an uninitialized pointer and free an
nonexistent proplist.
In practice there is always at least one profile, and I
don't think there will ever be cards without profiles.
Therefore, I added assertions to pa_card_new() stating that
the card new data must always contain at least one profile.
Now a lot of code can be simplified, because it's guaranteed
that the profiles hashmap and the active_profile field are
always non-NULL.
In my opinion, pa_card_set_profile() should assert that name
is not NULL, and it would be the job of the client interface
to filter out NULLs from the client input, but this is done
this way also when setting sink and source ports, so for
consistency I'll do this this way for now.
