summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* Merge remote-tracking branch 'origin/master'wm42012-05-2020-1770/+760
|\
| * font_load_ft.c: fix possible crash with recent FontconfigUoti Urpala2012-05-171-1/+1
| | | | | | | | | | | | | | | | | | | | | | Calling FcFontMatch with last argument NULL triggers an assertion failure in latest libfontconfig (2.9.0). Change the code to use a dummy address, like another nearby call to FcFontMatch already did. Whether the code triggering the assertion failure was run depended on font settings and/or available system fonts (the code looks like you'd have to get a non-scalable font first). Thus it didn't usually happen even with problematic libfontconfig version.
| * configure: OSX: fix check for compiler-specific cflagsStefano Pigozzi2012-05-171-3/+4
| | | | | | | | | | | | | | | | | | | | | | | | | | The configure script adds some compiler-specific flags for GCC on OSX. This was done under a check for compiler binary name not being clang. Move the test to after $cc_vendor has been determined and check against that instead. The previous test worked for clang if you explicitly specified --cc=clang, but not if the default system "cc" command was used and mapped to clang. Recent versions of Xcode changed the default compiler (cc) to clang. This caused a lot of spam from clang, which complained about the unknown flags when compiling with no explicit --cc option.
| * configure: don't enable GL/X11 without X11Stefano Pigozzi2012-05-171-9/+11
| | | | | | | | | | | | | | | | | | | | | | | | | | The OpenGL autodetection checked for all backends regardless of which features had been enabled previously. Stop checking for X11 backend if X11 support is disabled, and stop checking for w32 backend if we are not on Windows. This makes the changes in commit 3862d469ae ("configure: OSX: check for X11 header conflict with corevideo") work as intended. That commit disabled X11 when a header conflict was detected, but the GL X11 backend could still be autodetected despite that and trigger a compilation failure.
| * core: fix EOF handling with untimed audio outputsUoti Urpala2012-05-141-14/+12
| | | | | | | | | | | | | | | | | | | | | | When using an audio output without a native playback rate (such as ao_pcm), the code plays audio further when the current write position is behind video. After support for continuing audio after the end of video was added, this could cause a deadlock: audio was not played further, but neither was EOF triggered. Fix the code to properly handle playback of remaining audio after video ends in the untimed audio case (audio-only case was not affected, only the case where a video stream exists but ends before the audio stream).
| * options: simplify option parsing/setting machineryUoti Urpala2012-05-086-500/+160
| | | | | | | | | | | | | | | | | | | | | | | | Each option type had three separate operations to copy option values between memory locations: copy between general memory locations ("copy"), copy from general memory to active configuration of the program ("set"), and in the other direction ("save"). No normal option depends on this distinction any more. Change everything to define and use a single "copy" operation only. Change the special options "include" and "profile", which depended on hacky option types, to be special-cased directly in option parsing instead. Remove the now unused option types m_option_type_func and m_option_type_func_param.
| * options: change -v parsingUoti Urpala2012-05-074-10/+12
| | | | | | | | | | | | | | | | | | | | | | | | | | Handle -v flags as a special case in command line preparsing stage, and change the option entry into a dummy one. Specifying "v" in config file no longer works (and the dummy entry shows an error in this case); "msglevel" can still be used for that purpose. Because the flag is now interpreted at an earlier parsing stage, it now affects the printing of some early messages that were only affected by the MPLAYER_VERBOSE environment variable before. The main motivation for this change is to get rid of the last CONF_TYPE_FUNC option.
| * options: change --input=keylist, cmdlist implementationUoti Urpala2012-05-071-8/+8
| | | | | | | | | | | | | | | | Change the --input=keylist and --input=cmdlist suboptions to use the "print function" option type. This changes their semantics somewhat, and now some other output can appear after the printed lists (before, they called "exit(0)" directly). I'm not aware of any program parsing the output which could be affected.
| * ad_mpg123: update libmpg123 API useThomas Orgis2012-05-072-169/+84
| | | | | | | | | | | | Improve ad_mpg123, including use of the more efficient framewise decoding with mpg123 version 1.14 or later (older versions are still supported).
| * configure: use pkg-config for vdpau, libmpg123, libblurayUoti Urpala2012-05-061-5/+3
| |
| * ao_arts, ao_esd: remove these AOsUoti Urpala2012-05-065-691/+0
| | | | | | | | Delete ao_arts and ao_esd. Both have been deprecated upstream.
| * build: remove IRIX supportUoti Urpala2012-05-064-361/+2
| |
| * ao_portaudio: add new PortAudio audio output driverwm42012-05-064-0/+464
| | | | | | | | | | | | | | | | | | | | This AO has potential to be useful on platforms other than Linux. On Windows in particular, PortAudio can make use of newer/better audio APIs like WASAPI, instead of DirectSound. As an implementation choice, the PortAudio callback API was used. The blocking API might be a better match for mplayer's requirements, but caused severe problems on Linux/ALSA (possibly PortAudio bugs).
| * ao_pulse: fix specifying host/sink after 4fed8ad197Uoti Urpala2012-05-031-1/+1
| | | | | | | | | | | | | | Commit 4fed8ad197 ("ao_pulse: convert to new AO API") failed to change the variable name used on one line in suboption handling. This caused a crash due to NULL dereference if you tried to specify any suboptions for the AO (as in "--ao=pulse:foo"). Fix.
* | screenshot: fix minor memory leakwm42012-05-151-0/+1
| | | | | | | | Introduced in commit 5fe14d5bf "screenshot: improve error messages".
* | cocoa: return the real CGLPixelFormatObjStefano Pigozzi2012-05-151-1/+1
| | | | | | | | | | vo_cocoa_cgl_pixel_format is returning a cached CGLPixelFormatObj. Return the current one by querying the OpenGL context.
* | cocoa: don't request a depth buffer sizeStefano Pigozzi2012-05-151-2/+0
| |
* | cocoa: gl3: make window creation fail on <10.7Stefano Pigozzi2012-05-151-0/+5
| | | | | | | | | | Versions of OSX prior to 10.7 do not support OpenGL 3. Fail the window creation when that is the case.
* | cocoa: gl3: support querying of colors bit depthStefano Pigozzi2012-05-153-0/+21
| | | | | | | | | | Add support for querying the bit depth of the colors from the OpenGL context. This allows to perform dithering correctly.
* | cocoa: handle failure on pixel format creationStefano Pigozzi2012-05-141-0/+6
| | | | | | | | | | | | | | | | This didn't make any difference on with OpenGL 2.1, but with the introduction of OpenGL3.2 it's possible for the pixel format creation to fail (if OpenGL3.2 is not supported). This code handles the failure case accordingly.
* | screenshot: improve error messageswm42012-04-291-6/+6
| | | | | | | | | | | | | | Include filename in the error message, if the screenshot destination file already exists. Consistently quote filenames with 'quotes' in messages.
* | screenshot, vo_png: fix dependency on sizeof(AVFrame)wm42012-04-292-11/+22
| | | | | | | | | | | | | | | | | | In order to stay binary compatible with libavcodec, applications should not dependent on sizeof(AVFrame). This means allocating AVFrame on the stack is not allowed, and the function avcodec_alloc_frame() must be used to allocate an AVFrame instead. Partially based on a patch by uau.
* | Merge remote-tracking branch 'origin/master'wm42012-04-2921-1256/+1204
|\| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Conflicts: bstr.c bstr.h libvo/cocoa_common.m libvo/gl_common.c libvo/video_out.c mplayer.c screenshot.c sub/subassconvert.c Merge of cocoa_common.m done by pigoz. Picking my version of screenshot.c. The fix in commit aadf1002f8a will be redone in a follow-up commit, as the original commit causes too many conflicts with the work done locally in this branch, and other work in progress.
| * win32: get_path(): fix undefined behaviorMartin Herkt2012-04-281-3/+5
| | | | | | | | | | | | | | MSWindows-specific code in get_path() declared a stack array (exedir[]) in an inner scope, then kept a reference to the array beyond the end of the that scope. Fix. This caused visible breakage with GCC 4.7.
| * stream_cdda: print CDTEXT if availablewm42012-04-281-17/+51
| | | | | | | | | | | | | | | | | | | | The per-CD info will be printed on playback start, per-track info when a track is played. (This is not a technical restriction, and just goes along with the existing code.) The following fields are not included in output, because these are supposedly binary: CDTEXT_DISCID, CDTEXT_GENRE, CDTEXT_SIZE_INFO, CDTEXT_TOC_INFO, CDTEXT_TOC_INFO2.
| * configure, stream_cdda: remove libcdparanoia supportwm42012-04-282-104/+20
| | | | | | | | | | libcdparanoia is barely developed anymore, while libcdio is still quite active.
| * cosmetics: stream_cdda.c: reformatwm42012-04-281-359/+368
| |
| * stream_cdda: various fixesreimar2012-04-281-10/+28
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Fix cdda speed default value, range and use more robust condition. Based on patch by Ingo Brückl [ib wupperonline de]. git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@34458 b3059339-0415-0410-9bf9-f77b7e298cf2 Do not call paranoia_overlapset with 0, it actually causes cdparanoia to just hang. Instead use it to set/unset PARANOIA_MODE_OVERLAP. git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@34459 b3059339-0415-0410-9bf9-f77b7e298cf2 Fail if trying to seek beyond the last chapter, not just if it is beyond the end of the disc. git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@34460 b3059339-0415-0410-9bf9-f77b7e298cf2 cdda: set position to an actual EOF position when we set EOF. This avoids some inconsistency like the stream indicating EOF but a read still returning more data. git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@34462 b3059339-0415-0410-9bf9-f77b7e298cf2 Allow PARANOIA_MODE_FULL with skipping. git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@34467 b3059339-0415-0410-9bf9-f77b7e298cf2 Don't call paranoia_modeset() for PARANOIA_MODE_DISABLE. cdparanoia destroys start sector information after such a call. Since it is pointless without setting a mode anyway, don't do it. git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@34468 b3059339-0415-0410-9bf9-f77b7e298cf2 Add comment to a condition that is just a hack around a cdparanoia bug. git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@34472 b3059339-0415-0410-9bf9-f77b7e298cf2 Add checks for errors in stream_cdda's get_track_by_sector(). git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@34495 b3059339-0415-0410-9bf9-f77b7e298cf2 Fix seeking beyond EOF in stream_cdda to work with cache. git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@34577 b3059339-0415-0410-9bf9-f77b7e298cf2
| * cocoa_common: fix problems with alt-tab window changesStefano Pigozzi2012-04-261-4/+5
| | | | | | | | | | | | | | | | | | | | Fix alt tabbing to another window in the same workspace. The player window stayed on top because of a missing call to orderBack:. Fix alt tabbing to the player window from a different workspace. The window didn't get activated. Turns out that you must call makeKeyAndOrderFront: before setLevel: or setPresentationOptions: or the window will not properly ask for focus.
| * configure: OSX: check for X11 header conflict with corevideoStefano Pigozzi2012-04-261-78/+95
| | | | | | | | | | | | | | | | | | Check that headers from ApplicationServices and X11 do not conflict before enabling X11 support on OSX. Both headers would be included in vo_corevideo.m (through QuartzCore/QuartzCore.h and gl_common.h). The conflict exists on versions of Mac OSX prior to 10.7, where ApplicationServices includes the deprecated QuickDraw framework, resulting in a clash on the Cursor type definition.
| * Makefile: fix OSX compilation with --disable-corevideoStefano Pigozzi2012-04-261-3/+4
| | | | | | | | | | cocoa_common.m (enabled under COCOA) depends on osx_common.c, but the latter was erroneously only enabled under COREVIDEO.
| * cocoa_common, gl_common: add OSX specific getProcAddressStefano Pigozzi2012-04-264-2/+18
| | | | | | | | | | | | | | | | Run dlopen on the OpenGL dynamic library instead of on the binary. This should prevent crashes due to function conflicts when X11/lGL is linked. Remove mutual exclusion of the X11 and Cocoa backends.
| * OSX, input: implement wakeup in response to Cocoa eventsStefano Pigozzi2012-04-267-1/+204
| | | | | | | | | | | | | | | | | | Add code to wake up the select() call in input.c when an OSX event is available and a Cocoa OpenGL backend is initialized. Fixes the slow response to input or other events in Cocoa-based VOs during long select() sleeps (e.g., when mplayer2 is paused) introduced by commit 7040968.
| * vo_sharedbuffer: add this video outputStefano Pigozzi2012-04-265-0/+345
| | | | | | | | | | | | | | | | | | | | This OSX video output is replaces the previous shared_buffer mode of vo_corevideo. It manages a shared buffer and a Cocoa distributed object to communicate with GUIs. Splitting this code into a separate VO allows to get rid of harmful code coupling, performance inefficiencies (useless image memory copies) and ugly code (big if-else conditionals).
| * vo_corevideo: restructure this video outputStefano Pigozzi2012-04-265-352/+349
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Restructure this video output to be similar to vo_gl, even if simpler and less feature complete (for example it's still missing EOSD support). Ideally, it should act as a decent fallback in the case where something breaks in the OSX support of vo_gl. Here's a summary of what changed: * Remove the shared buffer code since it wasn't using any function from the CoreVideo API. Moreover, its presence in vo_corevideo was forcing the non-GUI related code to perform more image copies than necessary. Equivalent shared-buffer functionality will be added in a separate new VO in the next commit (this means OSX GUIs will need to specify a different VO). * Clean up the code to conform a bit more to the mplayer2 conventions. Enforce 80 column wrapping, use a private struct for file variables, use the new libvo api. * Add OSD rendering using OpenGL instead of writing directly on the video image data. * Simplify the logic for the rendering function when dealing with panscan. * Add VOCTRL_REDRAW_FRAME support. * Add colormatrix support by using the built-in API provided by CoreVideo.
| * vo_corevideo: use cocoa_common to display the windowStefano Pigozzi2012-04-265-863/+185
| | | | | | | | | | | | | | Change vo_corevideo to use cocoa_common to create and manage the window. This doesn't affect external OSX GUIs, since they don't use vo_corevideo window management, but only read the image data from the shared buffer.
| * vo_corevideo: use soft tabs (4 spaces)Stefano Pigozzi2012-04-261-684/+684
| |
| * win32: core: wake up more often to poll for inputUoti Urpala2012-04-261-2/+13
| | | | | | | | | | | | | | | | | | MSWindows does not have properly working support for detecting events on file descriptors. As a result the current mplayer2 code does not support waking up when new input events occur. Make the central playloop wake up more often to poll for events; otherwise response would be a lot laggier than on better operating systems during pause or other cases where the process would not otherwise wake up.
| * core: change initial sync with --delay, video stream switchUoti Urpala2012-04-231-1/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | Make A/V sync at the start of playback with nonzero --delay behave the same way as it does when seeking to the beginning later, meaning video plays from the start and audio is truncated or padded with silence to match timing. This was already the default behavior in case the streams in the file started at different times, but not if the mismatch was due to --delay. Trigger similar audio synchronization when switching to a new video stream. Previously, switching a video stream on after playing for some time in audio-only mode was buggy and caused initial desync equal to the duration of prior audio-only playback.
| * core: uninitialize VO and AO when no track playsUoti Urpala2012-04-231-10/+9
| | | | | | | | | | | | | | | | | | | | | | | | Uninitialize video and audio outputs when switching to a file without a corresponding track (audio-only file / file with no sound), or when entering --idle mode. Switching track choice to "off" during playback already did this. It could be useful to have a mode where the video window stays open even when no video plays, but implementing that properly would require more than just leaving the window on screen like the code did before this commit.
| * configure: --enable-debug: don't change other compiler flagsUoti Urpala2012-04-191-13/+8
| | | | | | | | | | | | | | | | The --enable-debug and --enable-profile options set their own compiler flags, completely different from normal flag selection. These flags sucked; especially '-W' (an obsolete alias for '-Wextra') generated a huge number of irrelevant warnings. Change configure to only add "-g" or similar to the flags that would be used otherwise.
| * ad_ffmpeg: switch to avcodec_decode_audio4()Uoti Urpala2012-04-191-69/+112
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Switch libavcodec audio decoding from avcodec_decode_audio3() to avcodec_decode_audio4(). Instead of decoding directly to the output buffer, the data is now copied from the libavcodec output packet, adding an extra memory copy (optimizing this would require some interface changes). After libavcodec added avcodec_decode_audio4() earlier, it dropped support for splitting large audio packets into output chunks of size AVCODEC_MAX_AUDIO_FRAME_SIZE or less. This caused a regression with the previous API: audio files with huge packets could fail to decode, as libavcodec refused to write into the AVCODEC_MAX_AUDIO_FRAME_SIZE buffer provided by mplayer2. This occurrend mainly with some lossless au