Nyaaori/rockbox
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity
17449 commits 17 branches 142 tags 291 MiB
Commit graph

87 commits

Author SHA1 Message Date
Nicolas Pennequin
f7e0e6bb1c Lock the linked list mutex when accessing the memory handles. This should fix the invalid values appearing for short times in the buffering debug screen. 
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@17497 a1c6a512-1295-4272-9138-f99709370657
 2008-05-13 20:51:06 +00:00
Steve Bavin
e02d031f4c Remove unused #define. 
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@17294 a1c6a512-1295-4272-9138-f99709370657
 2008-04-30 13:04:24 +00:00
Bertrik Sikken
c97e503a10 Cleaned up playblack.h header file and usage of it. 
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@17276 a1c6a512-1295-4272-9138-f99709370657
 2008-04-28 14:13:13 +00:00
Nicolas Pennequin
de026dcede Fix FS#8902 (NSF files get skipped). The cause was too strict metadata checking (the NSF metadata parser doesn't fill the 'length' field). Hopefully there won't be any problems with that anymore. 
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@17142 a1c6a512-1295-4272-9138-f99709370657
 2008-04-16 16:18:05 +00:00
Nicolas Pennequin
4e2de44b44 A rather big change to how tracks are loaded: there are now two parts to the process and metadata loading is done by the buffering thread (except for the first unbuffered track). The audio thread now calls audio_load_track, and once the metadata is loaded, the buffering thread sends an event which will make the audio thread call audio_finish_load_track. This one then takes care of the rest of the loading. 
This method makes skipping noticeably faster for unbuffered tracks, and especially backwards.


git-svn-id: svn://svn.rockbox.org/rockbox/trunk@17109 a1c6a512-1295-4272-9138-f99709370657
 2008-04-14 16:17:47 +00:00
Jens Arnold
2bf4178018 Make mpegplayer sleep the disk after buffering to save battery. * Add a simulator stub for ata_sleep(), and un-ifdef most calls to it. 
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@17096 a1c6a512-1295-4272-9138-f99709370657
 2008-04-13 12:24:47 +00:00
Nicolas Pennequin
33f522de8b Migrate the buffering code to the new events system. 
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@16950 a1c6a512-1295-4272-9138-f99709370657
 2008-04-03 17:51:53 +00:00
Michael Sevakis
8a6fd3ff4e Just change PRIORITY_BUFFERING to 15. Seems a teeny tiny bit helpful to mpegplayer as well. 
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@16884 a1c6a512-1295-4272-9138-f99709370657
 2008-03-29 23:21:19 +00:00
Michael Sevakis
9120c85698 Change the default buffer filechunk back to 32KB and bump the buffering thread priority up by one. 
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@16883 a1c6a512-1295-4272-9138-f99709370657
 2008-03-29 20:52:56 +00:00
Nicolas Pennequin
12b6c847c5 Fix sim warnings. 
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@16881 a1c6a512-1295-4272-9138-f99709370657
 2008-03-29 17:40:04 +00:00
Nicolas Pennequin
732df386af Boost the audio thread while it's filling the buffer. In the audio and the buffering threads, only cancel cpu boost when not filling. 
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@16880 a1c6a512-1295-4272-9138-f99709370657
 2008-03-29 17:28:30 +00:00
Nicolas Pennequin
4fd277481a Correct some windows line endings back to unix. 
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@16877 a1c6a512-1295-4272-9138-f99709370657
 2008-03-29 14:09:14 +00:00
Steve Bavin
135cc757bd Revert my earlier const madness, we'll keep the parameter lists simple. 
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@16863 a1c6a512-1295-4272-9138-f99709370657
 2008-03-28 12:51:33 +00:00
Steve Bavin
c9df8fd87b The const police raid playback.c, should be no change to behaviour. 
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@16860 a1c6a512-1295-4272-9138-f99709370657
 2008-03-28 11:24:24 +00:00
Steve Bavin
73f9863b1b More consts that require no functional change. 
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@16814 a1c6a512-1295-4272-9138-f99709370657
 2008-03-26 08:57:25 +00:00
Michael Sevakis
27cf677339 Add a complete priority inheritance implementation to the scheduler (all mutex ownership and queue_send calls are inheritable). Priorities are differential so that dispatch depends on the runnable range of priorities. Codec priority can therefore be raised in small steps (pcmbuf updated to enable). Simplify the kernel functions to ease implementation and use the same kernel.c for both sim and target (I'm tired of maintaining two ;_). 1) Not sure if a minor audio break at first buffering issue will exist on large-sector disks (the main mutex speed issue was genuinely resolved earlier). At this point it's best dealt with at the buffering level. It seems a larger filechunk could be used again. 2) Perhaps 64-bit sims will have some minor issues (finicky) but a backroll of the code of concern there is a 5-minute job. All kernel objects become incompatible so a full rebuild and update is needed. 
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@16791 a1c6a512-1295-4272-9138-f99709370657
 2008-03-25 02:34:12 +00:00
Nicolas Pennequin
483c40285b Notify the buffering thread when a handle is added, so it can go into filling mode. This is some sort of a replacement for the ATA idle callback. It will interrupt an ongoing buffering process (buffer_handle), but not for long enough to be a problem. 
Should fix a problem reported by Dave Hooper where inserting tracks into the playlist would flush the buffer and not refill it, sometimes causing the inserted tracks to be skipped.


git-svn-id: svn://svn.rockbox.org/rockbox/trunk@16295 a1c6a512-1295-4272-9138-f99709370657
 2008-02-12 23:15:59 +00:00
Nicolas Pennequin
659fe5a35b Fix FS#6215 (player crashes after trying to resume deleted music file). As mentioned in the lastest comment, the problem was in trying to resume at an offset larger than the size of the file following the one that was deleted. This revealed a crash in the buffering code, which gets a fix, but we also need to intervene earlier in the track loading so that the track that gets played will be from its start. 
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@16028 a1c6a512-1295-4272-9138-f99709370657
 2008-01-08 23:48:51 +00:00
Nicolas Pennequin
ecec94043d Check that the pointer isn't NULL before dereferencing it, just to be safe and to allow using bufgetdata to know how much data is available. Also remove a few trailing spaces. 
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@15940 a1c6a512-1295-4272-9138-f99709370657
 2007-12-16 01:38:56 +00:00
Brandon Low
e959c5aa95 Possibly save a few bytes when stripping the tags of the last allocated handle 
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@15844 a1c6a512-1295-4272-9138-f99709370657
 2007-11-28 16:39:58 +00:00
Brandon Low
3386dd7be9 Fix FS8069, because Nico_P made it easy 
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@15840 a1c6a512-1295-4272-9138-f99709370657
 2007-11-28 04:58:16 +00:00
Nicolas Pennequin
774fd19c37 Checking for total free buffer space isn't correct. What we need to check is the space available for the particular handle we want to rebuffer. 
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@15821 a1c6a512-1295-4272-9138-f99709370657
 2007-11-26 23:51:36 +00:00
Nicolas Pennequin
e24454f8b3 Buffering callbacks rework. There is now one callback for all the events that can occur. Callbacks are now registred only once instead of being removed after having been called. 
Fix FS#8092 by flushing the audio when a rebuffer is needed.
Also add some comments here and there.

git-svn-id: svn://svn.rockbox.org/rockbox/trunk@15816 a1c6a512-1295-4272-9138-f99709370657
 2007-11-26 21:13:08 +00:00
Michael Sevakis
398d9fd8b4 Hopefully the last word on getting rid of yield_codecs loops (cut read chunk to 16kB). Sansa ata driver also didn't yield enough so buffering would starve other threads. Bump priority of audio thread to keep UI more responsive. 
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@15724 a1c6a512-1295-4272-9138-f99709370657
 2007-11-20 22:45:46 +00:00
Michael Sevakis
fadbf0a6f7 Make threads responsible for explicit cancellation of their own boosted status. Sleeping and timeouts will no longer cancel it. 
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@15709 a1c6a512-1295-4272-9138-f99709370657
 2007-11-20 03:44:25 +00:00
Nicolas Pennequin
51923703fd Make the sim sleep between each file chunk, so that its buffering behaves more like that of a real target. 
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@15694 a1c6a512-1295-4272-9138-f99709370657
 2007-11-19 16:37:55 +00:00
Nicolas Pennequin
87e5b1193c The error checking for bitmap handling in bufopen was serioulsy broken, as loading a huge bitmap showed. Fix it and reorganise the code slightly. 
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@15597 a1c6a512-1295-4272-9138-f99709370657
 2007-11-12 15:16:41 +00:00
Nicolas Pennequin
5e0e2391ff Really fix red this time. 
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@15576 a1c6a512-1295-4272-9138-f99709370657
 2007-11-11 13:28:24 +00:00
Nicolas Pennequin
a384fb6d8a Fix red on the iFP. 
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@15575 a1c6a512-1295-4272-9138-f99709370657
 2007-11-11 13:15:36 +00:00
Nicolas Pennequin
9d4bed7ff0 Album art support. Based on FS#3045, but heavily modified to adapt to MoB and for cleanness. 
The cover pictures are loaded from external bitmaps. JPEG and embedded art are not supported. The pictures will only be drawn on the main display. There is no resizing but it is possible to specify the WPS bitmap size in the bitmap names (e.g. cover.100x100.bmp).
The bitmaps are stored in the main buffer and read directly from there. Currently, duplicate bitmaps will simply be present several times in the buffer, but this will be improved.
To enable for a target, #define HAVE_ALBUMART in its config file.
For more information, see the wiki page: http://www.rockbox.org/wiki/AlbumArt.


git-svn-id: svn://svn.rockbox.org/rockbox/trunk@15572 a1c6a512-1295-4272-9138-f99709370657
 2007-11-11 12:29:37 +00:00
Brandon Low
dc0f497a4f No functional change 
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@15542 a1c6a512-1295-4272-9138-f99709370657
 2007-11-08 18:59:22 +00:00
Nicolas Pennequin
b50473ffd8 Fix a bug introduced by r15533 where end-of-file wouldn't be detected correctly. 
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@15541 a1c6a512-1295-4272-9138-f99709370657
 2007-11-08 18:27:19 +00:00
Nicolas Pennequin
ca4771b40c Thanks to the changes in prep_bufdata, bufgetdata can be simplified a bit. 
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@15537 a1c6a512-1295-4272-9138-f99709370657
 2007-11-08 16:12:28 +00:00
Brandon Low
551db40abb Clean up a (nearly impossible) null check. Add a comment. 
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@15536 a1c6a512-1295-4272-9138-f99709370657
 2007-11-08 16:06:24 +00:00
Brandon Low
1db4243379 Change in function should be nearly zero, but change to proper naming and define use and allow a more liberal limit on bufgetdata to only ensure safety of wrapping, not the strict filechunk limit that I tried to impose earlier 
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@15535 a1c6a512-1295-4272-9138-f99709370657
 2007-11-08 15:52:10 +00:00
Brandon Low
6e8ee408bf Fix some bad where a handle is held across a sleep. We should probably audit this in other places as well. 
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@15533 a1c6a512-1295-4272-9138-f99709370657
 2007-11-08 15:34:23 +00:00
Nicolas Pennequin
f56cb7e09f Fix the bug where MPC files would be skipped by disabling the check of the requested size against the default filechunk size in prep_bufdata when the caller is bufread. 
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@15528 a1c6a512-1295-4272-9138-f99709370657
 2007-11-08 13:28:13 +00:00
Nicolas Pennequin
71b40994e0 Fix a flaw in prep_bufdata() that would lead to all kinds of problems with codecs that used bufread (MPC is one). 
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@15501 a1c6a512-1295-4272-9138-f99709370657
 2007-11-07 00:39:08 +00:00
Brandon Low
3eed6fc025 Fix thinko in last patch, shouldn't effect current buffering clients. 
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@15496 a1c6a512-1295-4272-9138-f99709370657
 2007-11-06 16:57:07 +00:00
Brandon Low
9821cce484 Pull the size range checking and disk request into a separate function that both bufread and bufgetdata can use. Do better range checking to ensure that we don't get 'stuck' in the calls and warn with a logf if requests are made for which we cannot guarantee satisfaction. 
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@15495 a1c6a512-1295-4272-9138-f99709370657
 2007-11-06 16:49:30 +00:00
Nicolas Pennequin
86b7c2b408 This was (probably) committed by accident, remove it. 
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@15491 a1c6a512-1295-4272-9138-f99709370657
 2007-11-05 23:46:45 +00:00
Brandon Low
7ca056dc13 Don't yield twice in a row 
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@15490 a1c6a512-1295-4272-9138-f99709370657
 2007-11-05 21:48:15 +00:00
Nicolas Pennequin
eac794de77 Types other than packet audio still need to yield while buffering. 
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@15489 a1c6a512-1295-4272-9138-f99709370657
 2007-11-05 21:46:17 +00:00
Nicolas Pennequin
0c7b26d3a7 Readd yield_codec, making it check for useful data through buffer_is_low. Fixes the PCM buffer underruns. 
Also move update_data_counters and make it static.

git-svn-id: svn://svn.rockbox.org/rockbox/trunk@15487 a1c6a512-1295-4272-9138-f99709370657
 2007-11-05 21:11:54 +00:00
Brandon Low
287f14b8ca This was a bad change on my part earlier. Do not immediately start buffering codecs when they are bufopened 
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@15486 a1c6a512-1295-4272-9138-f99709370657
 2007-11-05 18:58:50 +00:00
Nicolas Pennequin
6631e36fa7 Fix some logf warnings. 
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@15483 a1c6a512-1295-4272-9138-f99709370657
 2007-11-05 18:19:38 +00:00
Brandon Low
86830b6566 Create two separate ways for a handle buffer request to be sent. One is when the buffer has gotten low, in this case, we want to start a buffer fill. The other is when the buffer itself requests a buffer as part of a running fill. Obviously this is not a new fill start 
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@15480 a1c6a512-1295-4272-9138-f99709370657
 2007-11-05 17:51:55 +00:00
Brandon Low
76f9bfa945 When seeking foward off of the buffer, if it is a short seek don't rebuffer the whole track, just read enough to satisfy 
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@15479 a1c6a512-1295-4272-9138-f99709370657
 2007-11-05 17:50:51 +00:00
Brandon Low
3379440a4b Remove conf_filechunk, it should never have been a setting and its implementation doesn't do what it claims any way 
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@15478 a1c6a512-1295-4272-9138-f99709370657
 2007-11-05 17:48:21 +00:00
Brandon Low
1df1e7b0c7 Remove unused conf_preseek from buffering.c and all of its accountraments. This is _not_ a setting. This is a guessing tool used by either playback or buffering to serve its clients better. Use the REBUFFER_GUESS size in resume to help obviate pondlife's bug. This will also need to be used when FS8092 gets fixed correctly with a complete rebuffer for backward movements. It may also belong in buffering not playback, haven't decided for sure 
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@15475 a1c6a512-1295-4272-9138-f99709370657
 2007-11-05 16:02:12 +00:00