aboutsummaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* song: don't make the struct packedMax Kellermann2008-10-112-2/+2
| | | | | | The "packed" attribute may have negative side effects on performance. Remove the "packed" attribute, and increase the size of "song.url" to a multiple of the machine word size.
* directory: fix return value in removeDeletedFromDirectoryEric Wong2008-10-071-1/+1
| | | | oops :x
* Merge branch 'ew/update-thrsafe'Eric Wong2008-10-077-114/+167
|\ | | | | | | | | | | | | | | | | | | | | | | | | | | * ew/update-thrsafe: directory: serialize song deletes from playlist during update directory: use songvec_for_each for iterators dbUtils: more cleanups song: Add song_print_url_x dbUtils/directory: traverseAllIn forEachSong returns -1 on error songvec: lock traversals for thread-safe updates/reads song: add print_song_info_x for iterators tha pass void * songvec: add songvec_for_each iterator song: replace printSong* with song_print_* Assert if we don't have song or song->url set
| * directory: serialize song deletes from playlist during updateEric Wong2008-10-073-11/+27
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This makes the update code thread-safe and doesn't penalize the playlist code by complicating it with complicated and error-prone locks (and the associated overhead, not everybody has a thread-implementation as good as NPTL). The update task blocks during the delete; but the update task is a slow task anyways so we can block w/o people caring too much. This was also our only freeSong call site, so remove that function. Note that deleting entire directories is not fully thread-safe, yet; as their traversals are not yet locked.
| * directory: use songvec_for_each for iteratorsEric Wong2008-10-073-63/+56
| | | | | | | | Get rid of songvec_write so we can enforce proper locking
| * dbUtils: more cleanupsEric Wong2008-10-071-11/+3
| | | | | | | | | | | | | | Remove unnecessary wrapper function now that we have song_print_url_x and also return the results directly since we'll know the song_print_* functions will all return -1 on error.
| * song: Add song_print_url_xEric Wong2008-10-072-0/+8
| | | | | | | | | | It'll be handy for passing throug songvec_for_each like song_print_info_x.
| * dbUtils/directory: traverseAllIn forEachSong returns -1 on errorEric Wong2008-10-072-22/+14
| | | | | | | | Being consistent with most UNIX functions...
| * songvec: lock traversals for thread-safe updates/readsEric Wong2008-10-071-10/+34
| | | | | | | | | | | | | | | | | | | | Only one lock is used for all songvec traversals since they're rarely changed. Also, minimize lock time and release it before calling iterator functions since they may block (updateSongInfo => stat/open/seek/read). This lock only protects songvecs (and all of them) during traversals; not the individual song structures themselves.
| * song: add print_song_info_x for iterators tha pass void *Eric Wong2008-10-062-0/+8
| | | | | | | | | | traverseAllIn will be modified to take < 0 as errors instead of non-zero...
| * songvec: add songvec_for_each iteratorEric Wong2008-10-052-0/+15
| | | | | | | | | | This is so we can more consistently deal with locking needed for thread-safety in iterator functions.
| * song: replace printSong* with song_print_*Eric Wong2008-10-055-24/+23
| | | | | | | | | | | | This make argument order more consistent for iterators. Additionally, these now return ssize_t results for error checking.
| * Assert if we don't have song or song->url setEric Wong2008-10-052-4/+10
| | | | | | | | song objects cannot exist without a path or URL
* | wavpack: remove C99 initializersEric Wong2008-10-051-12/+16
| | | | | | | | | | More power to those who can't afford newer computers capable of installing/building/running newer toolchains.
* | dbUtils: remove C99 initializerEric Wong2008-10-051-3/+2
|/ | | | | | | Upgrading a toolchain is still a nasty operation and they're never easy to build, so don't force people to upgrade from their old compilers (especially since some people are still stuck with Linux 2.4 and gcc 2.95).
* autotools: tidy up make distEric Wong2008-10-053-145/+3
| | | | | | | | * Add missing headers in Makefile.am * remove mp4ff.dsp (Win32 crap) * Add scripts, m4, bs, autogen.sh to allow for hotfixes by the SCM-challenged. (downloading the source via git is NOT a lightweight operation for everybody).
* autotools: add pthreads linker supportEric Wong2008-10-052-0/+279
| | | | | This is needed for people that don't use any of the following: JACK, ALSA, libmikmod, Shout.
* Add .gitignoreEric Wong2008-10-051-0/+35
| | | | | These patterns have been sitting in my private .git/info/exclude files since 2005 because we used to use SVN.
* song: stop storing song_typeEric Wong2008-10-055-29/+22
| | | | | | | We already know if a song is a URL or not based on whether it has parentDir defined or not. Hopefully one day in the future we can drop HTTP support from MPD entirely when an HTTP filesystem comes along and we can access streams via open(2).
* song: use flex arrays to store song->urlEric Wong2008-10-053-19/+24
| | | | | | Reduce the number of allocations we make, so there's less pressure on the allocator and less overhead to keep track of the allocations in.
* song: converted "type" to enumMax Kellermann2008-10-052-5/+8
| | | | | | Having an enum type is much nicer than an anonymous integer plus CPP macros. Note that the old code didn't save any space by declaring the variable 8 bit, due to padding.
* song: get rid of newNullSong()Eric Wong2008-10-051-15/+4
| | | | | It didn't save us any lines of code nor did it do anything useful since we would overwrite everything anyways.
* song: call freeJustSong if newSong failsEric Wong2008-10-051-1/+1
| | | | | There's no reason to scan the playlist for a song we just allocated.
* playlist.h: fix build for folks that don't -DMPD_PATH_MAX=...Eric Wong2008-10-051-0/+1
| | | | Thanks (again) to unK for the report :x
* decode: fix build for folks that don't -DMPD_PATH_MAX=...Eric Wong2008-10-051-0/+1
| | | | Thanks to unK for the report
* directory: fix writeDirectoryDBEric Wong2008-10-051-2/+4
| | | | Wow, I must have been halfway asleep when I did that...
* command: fix return statusEric Wong2008-10-051-16/+6
| | | | | | This got broken when listHandlerFunc was removed. Since we no longer need it and it's confusing, remove processCommandInternal and just use process_command.
* Merge commit '1533279fb18b2b0f53345f0e0727d2f0f968c00d'Eric Wong2008-10-0515-393/+283
|\ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | * commit '1533279fb18b2b0f53345f0e0727d2f0f968c00d': directory: reuse existing directory if found on update song: better handling of existing songs when rereading DB tag: merge clearMpdTag into tag_free song: start avoiding race in updateSongInfo command: get rid of specialized list handlers directory: simplify list update handling logic main_notify: define main_task so we can use it for assertions directory: streamline deletes song: start avoiding race in updateSongInfo playlist: small lines of code reduction Revert "Start using song pointers in core data structures" Conflicts: src/playlist.c
| * Merge commit 'box/song-locks' into ew/song-locksEric Wong2008-10-043-20/+24
| |\ | | | | | | | | | | | | | | | | | | | | | * commit 'box/song-locks-wip': directory: reuse existing directory if found on update song: better handling of existing songs when rereading DB tag: merge clearMpdTag into tag_free song: start avoiding race in updateSongInfo
| | * directory: reuse existing directory if found on updateEric Wong2008-10-041-4/+9
| | | | | | | | | | | | | | | | | | | | | | | | | | | Instead of allocating a new one, just reuse an existing one if one is found when rereading the DB. This is a small makes the previous commit work on subdirectories of the root music directory. [1] "song: better handling of existing songs when rereading DB"
| | * song: better handling of existing songs when rereading DBEric Wong2008-10-041-4/+13
| | | | | | | | | | | | | | | | | | Don't reallocate existing tags if they haven't changed. This isn't used as often anymore, but we still take HUP signals to reread the DB if other processes changed it.
| | * tag: merge clearMpdTag into tag_freeEric Wong2008-10-031-12/+2
| | | | | | | | | | | | | | | avoid some redundant clearing logic as well, since the tag is getting freed.
| | * song: start avoiding race in updateSongInfoEric Wong2008-10-031-9/+12
| | | | | | | | | | | | | | | This is still not SMP-safe yet, as it needs at least a barrier before calling tag_free(old_tag).
| * | Merge commit 'box/directory' into ew/song-locksEric Wong2008-10-045-225/+159
| |\ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | * commit 'box/directory': command: get rid of specialized list handlers directory: simplify list update handling logic main_notify: define main_task so we can use it for assertions directory: streamline deletes Conflicts: src/command.c
| | * | command: get rid of specialized list handlersEric Wong2008-10-031-106/+67
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | commands should really not behave differently if they're issued inside a command list or not; so stop having special handler functions to deal with them. "update" was the only command that used this functionality and I changed that in the last commit to serialize access.
| | * | directory: simplify list update handling logicEric Wong2008-10-033-100/+71
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Now the "update" command can be issued multiple times regardless of whether the client is in list mode or not. We serialize the update tasks to prevent updates from trampling over each other and will spawn another update task once the current one is finished updating and reaped. Right now we cap the queue size to 32 which is probably enough (I bet most people usually run update with no argument anyways); but we can make it grow/shrink dynamically if needed. There'll still be a hard-coded limit to prevent DoS attacks, though.
| | * | main_notify: define main_task so we can use it for assertionsEric Wong2008-10-032-0/+7
| | | | | | | | | | | | | | | | | | | | It'll be easier to keep track of what code runs in what task/thread this way.
| | * | directory: streamline deletesEric Wong2008-10-031-17/+15
| | | | | | | | | | | | | | | | | | | | Instead of relying on the shortname, just pass the song pointer to prevent redundant lookups during deletes.
| * | | song: start avoiding race in updateSongInfoEric Wong2008-10-031-9/+12
| | |/ | |/| | | | | | | | | | | | | | | | | | | | | | This is still not SMP-safe yet, as it needs at least a barrier before calling tag_free(old_tag). Locks may be simpler to implement and the potential performance gain of avoiding locks may not be worth it on infrequently modified data structures...
| * | playlist: small lines of code reductionEric Wong2008-10-031-96/+47
| | | | | | | | | | | | Hopefully this makes the code feel less claustrophobic...
| * | Revert "Start using song pointers in core data structures"Eric Wong2008-10-028-49/+44
| | | | | | | | | | | | | | | | | | | | | This actually opened us up to making lock dependencies more difficult than they needed to be now that we have threaded updates. We would always use the memory anyways, just in the stack instead of bss.
* | | playlist: fix "currentsong" at end-of-playlistEric Wong2008-10-041-4/+11
|/ / | | | | | | | | | | | | The current song information could get cleared if we got to the end of the playlist and mpd is not in repeat mode. This also prevents "currentsong" from returning information if mpd is not playing.
* | Switch to C99 types (retaining compat with old compilers)Eric Wong2008-09-2931-132/+150
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Seeing the "mpd_" prefix _everywhere_ is mind-numbing as the mind needs to retrain itself to skip over the first 4 tokens of a type to get to its meaning. So avoid having extra characters on my terminal to make it easier to follow code at 2:30 am in the morning. Please report any new issues you may come across on Free toolchains. I realize how difficult it can be to build/maintain cross-compiling toolchains and I have no intention of forcing people to upgrade their toolchains to build mpd. Tested with gcc 2.95.4 and and gcc 4.3.1 on x86-32.
* | allow searching for albums with an empty tagEric Wong2008-09-292-1/+25
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | tfing wrote: > I have quite some files with an empty album tag as they do not come > from a particular album. > > If I want to look for those files and browse them, this happens: > :: nc localhost 6600 > OK MPD 0.12.0 > find album "" > ACK [2@0] {find} too few arguments for "find" > > I'd like to be able to browse those files in a client like gmpc. > So these 2 items would have to be developed: > - list album should report that some files have an empty tag > - it should be possible to search for an empty tag with the find command Patch-by: Marc Pavot ref: http://musicpd.org/mantis/view.php?id=464
* | Merge branch 'ew/directory'Eric Wong2008-09-2912-375/+341
|\| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | * ew/directory: directory: remove redundant sanitizePathDup update: move path sanitation up the stack to avoid extra copies clean up updateInit calling and error handling directory: isRootDirectory() is a one-liner directory: writeDirectoryInfo propagates errors directory: make it clear that DIRECTORY_MTIME is deprecated directory: remove "Mp3" references playlist: deleteASongFromPlaylist takes a const Song * songvec: songvec_delete takes a const Song pointer directory: remove shortname arguments everywhere path: add mpd_basename() function directory.h: remove directory_sigChldHandler decl directory: replace DirectoryList with dirvec directory: remove unused CPP defines songvec_free => songvec_destroy directory.c: kill unnecessary includes
| * directory: remove redundant sanitizePathDupEric Wong2008-09-291-15/+7
| | | | | | | | | | | | | | We already sanitize and duplicated our paths before calling updateInit() to get pre-pthread_create() error-checking along with heap allocation reduction because we don't have to redupe because our parent stack went away.
| * update: move path sanitation up the stack to avoid extra copiesEric Wong2008-09-293-30/+58
| | | | | | | | | | | | | | | | | | | | | | | | | | | | Remove yet another use of our old malloc-happy linked list implementation and replace it with a simple array of strings. This also implements more eager error handling of invalid paths (still centralized in updateInit) so we can filter out bad paths before we spawn a thread. This also does its part to fix the "update" command inside list mode which lost its static variable in ada24f9a921ff95d874195acf253b5a9dd12213d (although it was broken and requires the fix in 769939b62f7557f8e7c483223d68a8b39af43e37, too).
| * clean up updateInit calling and error handlingEric Wong2008-09-283-13/+22
| | | | | | | | | | Move error reporting to command.c so directory.c does not deal with client error handling any more.
| * directory: isRootDirectory() is a one-linerEric Wong2008-09-281-4/+1
| | | | | | | | Improving the signal to noise ratio...
| * directory: writeDirectoryInfo propagates errorsEric Wong2008-09-281-28/+19
| | | | | | | | | | If a write failed, it's a good sign subsequent writes will fail, too, so propgate errors all the way up the stack.