aboutsummaryrefslogtreecommitdiffstats
path: root/src (unfollow)
Commit message (Collapse)AuthorFilesLines
2008-04-12added output_buffer_expand()Max Kellermann1-13/+22
output_buffer_expand() moves the cb->end to the new position (only its current successor is allowed) and wakes up the player if is waiting for the decoder. This simplifies flushOutputBuffer(). git-svn-id: https://svn.musicpd.org/mpd/trunk@7338 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12initialize all elements in initOutputBuffer()Max Kellermann1-0/+2
The current OutputBuffer object is allocated statically, i.e. it is zeroed. To be safe for other cases in the future, also initialize the other elements. git-svn-id: https://svn.musicpd.org/mpd/trunk@7337 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12wake up player on demandMax Kellermann1-1/+9
The decoder should not wake up the player when it did not produce a flushed chunk. Move the decoder_wakeup_player() call to flushOutputBuffer() and invoke it only if the buffer was previously empty. git-svn-id: https://svn.musicpd.org/mpd/trunk@7336 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12unlink the socket before bind()Max Kellermann1-0/+2
If a unix domain socket already exists, bind() fails. Unlink the socket file. git-svn-id: https://svn.musicpd.org/mpd/trunk@7335 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12replaced assertion with checkMax Kellermann1-2/+1
During my tests, it happened that data->position>newPosition. I have not yet fully understood why this can happen; for now, replace this with a run-time check. git-svn-id: https://svn.musicpd.org/mpd/trunk@7334 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12fix several errors/warnings which only appear with -O3Max Kellermann2-1/+3
git-svn-id: https://svn.musicpd.org/mpd/trunk@7333 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12multiply num_samples with bytes_per_channelMax Kellermann1-1/+1
The patch "convert blocks until the buffer is full" did not update data->chunk_length correctly: it added the number of samples, not the number of bytes. Multiply that with bytes_per_channel git-svn-id: https://svn.musicpd.org/mpd/trunk@7332 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12missing num_channels check in previous patchMax Kellermann1-1/+1
In the patch "special optimized case for 16bit stereo", the check for "num_channels==2" was missing. git-svn-id: https://svn.musicpd.org/mpd/trunk@7331 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12special optimized case for 16bit stereoMax Kellermann1-3/+20
Not having to loop for every sample byte (depending on a variable unknown at compile time) saves a lot of CPU cycles. We could consider reimplementing this function with liboil... git-svn-id: https://svn.musicpd.org/mpd/trunk@7330 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12read num_channels onceMax Kellermann1-3/+4
Read frame->header.channels once, and pass only this integer to flac_convert(). git-svn-id: https://svn.musicpd.org/mpd/trunk@7329 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12assume the buffer is empty in flacWrite()Max Kellermann1-4/+3
flacWrite() is the only function which sets data->chunk_length. If we flush the buffer before we return, we can assume that it is always empty upon entering flacWrite(). git-svn-id: https://svn.musicpd.org/mpd/trunk@7328 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12convert blocks until the buffer is fullMax Kellermann1-23/+43
Move the inner loop which converts samples to flac_convert(). There it is isolated and easier to optimize. This function does not have to worry about buffer boundaries; the caller (i.e. flacWrite()) calculates how much is left and is responsible for flushing. That saves a lot of superfluous range checks within the loop. git-svn-id: https://svn.musicpd.org/mpd/trunk@7327 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12calculate bytes_per_channel, check for buffer flush onceMax Kellermann1-11/+14
Check for flushing the chunk buffer only once per sample, before iterating over channels and bytes. This saves another 5% CPU cycles. git-svn-id: https://svn.musicpd.org/mpd/trunk@7326 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12don't calculate bytes per sample within the loopMax Kellermann1-1/+2
AudioFormat.bits is volatile, and to read it, 3 pointers had to be deferenced. Calculate this value once. This speeds up this function by 5%. git-svn-id: https://svn.musicpd.org/mpd/trunk@7325 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12use unsigned integers and size_t in the flac pluginMax Kellermann2-4/+9
git-svn-id: https://svn.musicpd.org/mpd/trunk@7324 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12pass dc to decoder_task()Max Kellermann3-6/+6
git-svn-id: https://svn.musicpd.org/mpd/trunk@7323 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12pass pc to set_current_song()Max Kellermann1-6/+4
git-svn-id: https://svn.musicpd.org/mpd/trunk@7322 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12pass pc to player_task()Max Kellermann3-6/+6
Another global variable cleanup patch. git-svn-id: https://svn.musicpd.org/mpd/trunk@7321 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12whitespace cleanupMax Kellermann1-14/+14
git-svn-id: https://svn.musicpd.org/mpd/trunk@7320 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12clean up CPP includesMax Kellermann55-110/+33
Try to only include headers which are really needed. We should particularly check all "headers including other headers". The long-term goal is to have a manageable, small API for plugins (decoders, output) without so many mpd internals cluttering the namespace. git-svn-id: https://svn.musicpd.org/mpd/trunk@7319 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12removed "volatile" from OutputBufferChunk propertiesMax Kellermann1-3/+3
There is no danger that gcc will optimize access to OutputBufferChunk properties, since decoder and player work in different chunk objects. It is safe to remove "volatile" here. git-svn-id: https://svn.musicpd.org/mpd/trunk@7318 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12pass pc to player notify functionsMax Kellermann3-30/+29
Same as the previous patch: less global variables. git-svn-id: https://svn.musicpd.org/mpd/trunk@7317 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12pass DecoderControl object to decoder_sleep()Max Kellermann3-17/+14
Less global variables: at any invocation of decoder_sleep(), we have a reference to the DecoderControl anyway, so we should pass it. This costs less than having to call getPlayerData() in every tiny function. Maybe some day we will be able to have multiple decoders at the same time... git-svn-id: https://svn.musicpd.org/mpd/trunk@7316 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12added output_buffer_free()Max Kellermann3-0/+9
To do proper cleanup before exiting, we have to provide a destructor for OutputBuffer. One day, valgrind will not complain about memory leaks! git-svn-id: https://svn.musicpd.org/mpd/trunk@7315 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12remove the local variable "end"Max Kellermann1-5/+0
"end" is not being used anymore, since we moved most OutputBuffer struct accesses into methods. git-svn-id: https://svn.musicpd.org/mpd/trunk@7314 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12don't use short chunk numbersMax Kellermann3-7/+7
Don't be mean with integer sizes. Although we will probably never have more than 32k buffered chunks, we should use 32 bit integers for addressing them. We do not save very much (some of the saved space is eaten by alignment anyway), but we save at least one assembler instruction for converting short to int. This change requires some more explicit casts, because gcc was less picky when comparing short with a full int. git-svn-id: https://svn.musicpd.org/mpd/trunk@7313 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12added output_buffer_skip()Max Kellermann3-8/+9
First patch without camelCase ;) output_buffer_skip() lets us eliminate advanceOutputBufferTo(), and removes yet another external OutputBuffer struct access. git-svn-id: https://svn.musicpd.org/mpd/trunk@7312 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12remove buffered_chunks, use cb->sizeMax Kellermann3-3/+3
Second patch to make OutputBuffer self-contained: since OutputBuffer now knows its own size, we do not need the global variable "buffered_chunks" anymore. git-svn-id: https://svn.musicpd.org/mpd/trunk@7311 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12pass buffered_chunks to initOutputBuffer()Max Kellermann3-16/+21
Try to make OutputBuffer self-contained, without depending on a global variable. git-svn-id: https://svn.musicpd.org/mpd/trunk@7310 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12make playerData_pd staticMax Kellermann1-39/+32
git-svn-id: https://svn.musicpd.org/mpd/trunk@7309 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12allocate playerData_pd from heap instead of shmMax Kellermann2-12/+2
git-svn-id: https://svn.musicpd.org/mpd/trunk@7308 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12fixed another shadow warningMax Kellermann1-11/+11
git-svn-id: https://svn.musicpd.org/mpd/trunk@7307 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12don't allocate pd.auddioDeviceStates from shmMax Kellermann1-5/+4
git-svn-id: https://svn.musicpd.org/mpd/trunk@7306 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12fix typo in commentMax Kellermann1-3/+3
git-svn-id: https://svn.musicpd.org/mpd/trunk@7305 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12let initOutputBuffer() allocate memoryMax Kellermann3-10/+6
This is the first patch in a series which removes the shared memory, and moves all the playerData objects into the normal libc heap. git-svn-id: https://svn.musicpd.org/mpd/trunk@7304 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12fix deadlock at end of songMax Kellermann1-0/+1
Now that we do proper locking and signalling instead of continuous polling for IPC, a deadlock was found: at the end of a song, the player thread waits until the main thread sets pc->queueState from PLAYER_QUEUE_DECODE to PLAYER_QUEUE_PLAY. He is never woken up, since syncPlaylistWithQueue() does not activate the notification. I added wakeup_player_nb() to setQueueState(), since the player must be signalled whenever pc->queueState changes. git-svn-id: https://svn.musicpd.org/mpd/trunk@7303 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12decode: fix some uninitialized variable warningsEric Wong2-3/+3
They're probably not needed, but less noise => faster debugging git-svn-id: https://svn.musicpd.org/mpd/trunk@7302 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12decode: fix unsigned comparision and add some paranoid assertionsEric Wong1-8/+9
git-svn-id: https://svn.musicpd.org/mpd/trunk@7301 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12Fix a few more warnings from -WshadowEric Wong7-36/+38
git-svn-id: https://svn.musicpd.org/mpd/trunk@7300 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12missing const in disabled code pathMax Kellermann1-1/+1
git-svn-id: https://svn.musicpd.org/mpd/trunk@7298 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12some assertions to check the unsigned changesMax Kellermann1-0/+3
git-svn-id: https://svn.musicpd.org/mpd/trunk@7297 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12use size_t in audio.cMax Kellermann2-5/+5
Buffer sizes should be size_t. This is safe here, at least not unsafer than without the patch. I have no idea why audioBufferSize and audioBufferPos were explicitly declared as signed integer. git-svn-id: https://svn.musicpd.org/mpd/trunk@7296 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12use free()/malloc() instead of realloc()Max Kellermann1-2/+3
When growing the audioOutput->convBuffer, we can use free()+malloc() instead of realloc(), which saves a memcpy(). git-svn-id: https://svn.musicpd.org/mpd/trunk@7295 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12use size_t in audioOutput.c internalsMax Kellermann2-2/+2
git-svn-id: https://svn.musicpd.org/mpd/trunk@7294 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12use size_t and constant pointer in ao pluginsMax Kellermann12-36/+66
The audio output plugins should get a constant pointer, because they must not modify the buffer. Since the size is a non-negative buffer size in bytes, we should change its type to size_t. git-svn-id: https://svn.musicpd.org/mpd/trunk@7293 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12use size_t for pcm_sizeOfConvBuffer() resultMax Kellermann1-4/+4
In my previous patch set, I forgot to change the pcm_sizeOfConvBuffer() invocation in convertAudioFormat() to also use size_t. git-svn-id: https://svn.musicpd.org/mpd/trunk@7292 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12yet more unsigned integersMax Kellermann1-4/+4
calculateCrossFadeChunks() still returns int, although the caller uses it as an unsigned value. Since the function body checks for negative values, it is safe to cast to unsigned. crossFade() takes signed parameters, although it callers pass unsigned integers. Change declaration to unsigned. git-svn-id: https://svn.musicpd.org/mpd/trunk@7291 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12constant pointersMax Kellermann3-9/+9
There were some const pointers missing in the previous const-cleanup patch. git-svn-id: https://svn.musicpd.org/mpd/trunk@7290 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12fix parameter types in the faad callsMax Kellermann2-3/+3
libfaad wants uint32_t pointers. Passing a long pointer is bugged on amd64. git-svn-id: https://svn.musicpd.org/mpd/trunk@7289 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12fix compilation error in wavpack pluginMax Kellermann1-3/+6
The patch "Start using song pointers in core data structures" removed dc->utf8url, and the adaption for wavpack_plugin.c was missing. git-svn-id: https://svn.musicpd.org/mpd/trunk@7288 09075e82-0dd4-0310-85a5-a0d7c8717e4f
generated by cgit v1.2.3 (git 2.25.1) at 2024-07-27 20:17:25 +0000