Some (broken) streams don't have the extended stream properties in
the header, resulting in applying a duration of zero on outgoing
buffers.
Fixes#611473
Some files have payload with timestamps smaller than the preroll duration.
Instead of blindly substracting the preroll value (and ending up with
insanely high timestamps on the outgoing buffers), we make sure we
never go below 0.
Fixes#610432
Fix off-by-one bug when constructing the avcC header chunk: we
wrote wrong profile info into the header. The first byte in the
SPS NAL we get from x264 is the nal type, not the profile_idc.
Also add some debug logging.
Fixes#610089.
We don't want C specific flags in GST_CXXFLAGS, so base it on the
GST_CFLAGS that only contains the pkg-config CFLAGS but none of
the GST_OPTION_CFLAGS. Also, we only need the local includes once.
Generate win32/common/config.h-new directly from config.h.in,
using shell variables in configure and some hard-coded information.
Change top-level makefile so that 'make win32-update' copies the
generated file to win32/common/config.h, which we keep in source
control. It's kept in source control so that the git tree is
buildable from VS.
This change is similar to the ones applied a while ago to GStreamer
core, gst-plugins-base and gst-plugins-good and prevents configure
from changing files in source control. The generated config.h
should be ok, but needs testing.
Turns out 4 + 4 + 2 + (4 * 2) is actually 18 and not 22. This avoids
a presumably unintentional padding of uninitialised bytes at the end
of the CONT tags chunk, which should be harmless but causes warnings
in valgrind (see #608533 for a test URL).
We previously only aggregated flow returns after the while(push) loop,
which meant that in some cases we would end-up not properly aggregating
the flow returns.
This is based on the same flow aggregation algorithm as oggdemux.
The correct basis for (Xing, VBRI) seek table calculations is the
byte size and duration provided by that metadata, rather than some
other (possibly even estimated) one. This also prevents an infinite
conversion loop in (unlikely) case where a TOC is provided without
such corresponding (duration) metdata.
Use the same strategy as accurate seeks to store
pending non-accurate seeks to avoid overwriting non-definite
stop times. When doing non-accurate seeks our position
reporting might drift off by some secs and the stream can
end up before it should.
Fixes#603695
lame's 'lgv' variable is only initialized when the caps
is negotiated, whenever a seek happens before that, it would
attempt to call a function on an empty pointer, causing the crash.
Fixes#603515
