| Age | Commit message (Collapse) | Author |
|---|
|
ffmpeg/libav - addition to #13096
|
|
e064049 - fixes #13096
|
|
|
|
|
|
bump libnfs to 1.3
|
|
at reasonable speed is feasible
|
|
pointer will be deleted, not trap'
|
|
Some multi-volume RAR files have their unpacked size set as 0 in all
volumes except the first one.
Use the previous unpacked size instead of 0 in such cases in order to
support such files properly.
|
|
Certain seek patterns on a file inside a non-compressed rar file can
cause CmdExtract::UnstoreFile() to think that the destination buffer has
been filled (as DestUnpSize counter, originally set to the file size,
reaches zero).
However, counting written bytes using DestUnpSize doesn't make sense for
the UnpackToMemory codepath used for non-compressed rar files, as there
can be seeks which can eventually cause more data to be read than what
the actual file size was. The actual output buffer is internally handled
by ComprDataIO.
The check in UnstoreFile() will result in not all data being written to
the destination buffer, causing CRarFile::Read() to return old stale
data.
Fix that by dropping the unnecessary DestUnpSize handling in
UnpackToMemory codepath of CmdExtract::UnstoreFile().
|
|
on using 'gas-preprocessor.pl ' as that fails with clang, refactor the configure of ffmpeg to allow passing -as= when expands to multiple items
|
|
texturepacker
|
|
x86_64 in CXXFLAGS as a hint to build it 32 or 64 bit
|
|
|
|
|
|
FileTransfer buffer.
|
|
add FreeBSD support
|
|
|
|
|
|
do not use DWORD or LPDWORD to store a thread id anymore
use LONG or LPLONG instead
switch back dword to int
|
|
TODO: understand that error
dcr.c:148:8: error: member reference base type 'dcr_stream_obj' (aka 'void') is not a structure or union
if (fileno(dcr.obj_) > 2) (*dcr.ops_->close_)(dcr.obj_);
^~~~~~~~~~~~~~~~
In file included from dcr.c:43:
/usr/include/stdio.h:483:36: note: instantiated from:
define fileno(p) (!__isthreaded ? __sfileno(p) : (fileno)(p))
^
/usr/include/stdio.h:474:28: note: instantiated from:
define __sfileno(p) ((p)->_file)
~~~ ^
|
|
Support for MythTV 0.25 in libcmyth
|
|
|
|
from our mirrors
|
|
|
|
|
|
Gentoo seems to have already updated to the trunk version of MythTV after the 0.25 release.
No changes necessary as libcmyth doesn't currently have support for RESCHEDULE_RECORDINGS.
|
|
No changes necessary as new QUERY_ACTIVE_BACKENDS method is not used yet.
|
|
It's not entirely clear what the protocol changes were as it looks like the changes only affected the scheduler. libcmyth isn't using ASK_RECORDER or GET_FREE_INPUTS so assume no changes are necessary.
|
|
No changes necessary as the new QUERY_FILETRANSFER REOPEN method is not yet used.
|
|
No changes necessary as QUERY FILE HASH is not used by libcmyth.
|
|
No changes required as the VIDEO_LIST_CHANGE event is not used in libcmyth.
|
|
Added new proginfo_season, proginfo_episode, proginfo_inetref properties and altered all of the places where the proginfo object is passed or received over the myth protocol.
|
|
All 64 bit integers are now passed and received as a single 64 bit number rather than two 32 bit hi and lo integers.
|
|
Final step in the threading refactor: CThread
|
|
|
|
Also:
[fix] threads: fix race condition when stopping thread - cherry pick merge
Also, from jimfcarroll:
[fixed] AirPlayServer now complies with the new CThread constructor requirements (that is, you must supply a thread name).
replace custom thread local storage with the XbmcThreads::ThreadLocal template.
[cosmetic] removed incorrect hungarian notation.
[cleanup] Removed XSyncUtils. Moved the GlobalMemoryStatus call to XMemUtil. Cleaned up code that used the #define INFINITE from XSyncUtils.h
[cleanup] Some miscellaneous cleanup. Fixed initialization warning in LCD. Removed some unused functions and members.
[cleanup] PAPlayer warning due to constructor initialization list order.
|
|
Add support for new ffmpeg-10/11 api.
|
|
from a previous build
|
|
|
|
This is very hackish, but it's the old solution we had for dvd
still frames. The problem is that the demuxer asks for probing
of the codec in the mpeg stream. This causes lavf to read the
whole menu structure into internal buffers. After which, it
won't read from input stream anymore and no events triggers.
The only proper fix to avoid this is to allow this situation
in our navigator. This needs further work so that we can
process HOP/ACTIVATION events in libdvdnav without advancing
stream.
|
|
|
|
The QUERY_RECORDINGS command needed to be changed to support the new sort API.
|
|
|
|
|
|
|
|
Code will fall back to longest item if there is not first play
hdmv item. However it will still try menu's if there is only
some unnsupported BD-J titles. This could possible fail to
play if it switches into BD-J mode later in menu's.
Starting index.bdmv will still only play longest title,
to attempt playback with menu's, start the MovieObject.bdmv
|
|
xbmc will abort configure if an old libbluray is found
and configure is auto or yes. To skip libbluray support
in that case explicitly disable it.
|
|
decrypt
|
|
|
|
|
