When data is not in packed-16-bit mode, flushing the fifo
may result in swapping left and right channels if there
happens to be an odd number of entries in the FIFO.
This is especially likely when switching sample frequencies
for some reason.
When stopping PCM DMA, disable DMA and Underrun Interrupts
and then wait for FIFO to be empty before stopping AIC's playback.
Change-Id: I45b6b022c9e3889627842663cd9b7d2e0affb7c6
opx shortcuts allow an easy way to add parameters to plugins
you should be able to set them as shortcuts now
the plugin hd a use after free bug on the dat file file descriptor
Change-Id: I5fe3b0628e7da003c03b9b54f5788b0d0dbfea74
save tsr plugin path for later
resume tsr plugin when user stops the interrupting plugin
expand return of tsr_exit function to allow
continue, suspend, terminate
tsr plugins check parameter at start to determine if
the plugin is being resumed
Change-Id: I6fc70de664c7771e7dbc9a1af7a831e7b50b1d15
Control cache entries cost 24 bytes per command, but cacheable
commands are always less than that when written out to the file.
We can actually cache *more* data by writing commands directly
to the fd (native Rockbox has a 512-byte cache per fd) and it's
much simpler.
Change-Id: Ibb1b9ffa56ef17431b281419a04082e14d0cbd85
The intent behind the check is to only do this if the filename
comes from an on-disk file. With the RAM buffer is gone this is
now "always" the case.
Change-Id: I4c26e32c482dde176b69a6071e9562e9955d7171
This checks for an out of memory condition when the filenames
were buffered in RAM. The playlist no longer buffers anything
in RAM so the code is unreachable.
Change-Id: I4c2d0b2ac4ba3b7e7c94fe554a0cf7cb1a798e09
Use the playlist control file for directory playback instead of
storing filenames in RAM. The implementation of that feature is
very iffy and probably responsible for random crashes that may
occur when skipping through directories rapidly.
Change-Id: I3863940cd4542410d8046a3ca47508b5d97309a1
GCC 4.9 always emits assembly with divided syntax. Setting unified
syntax in inline assembly causes the assembler to complain about
GCC's generated code, because the directive extends past the scope
of the inline asm. Fix this by setting divided mode at the end of
the inline assembly block.
The assembler directives are hidden behind macros because later
versions of GCC won't need this workaround: they can be told to
use the unified syntax with -masm-syntax-unified.
Change-Id: Ic09e729e5bbb6fd44d08dac348daf6f55c75d7d8
Any modifications to the playlist (insert, delete, shuffle, etc)
will cause the modified flag to be set. The flag is cleared when
the playlist is saved. Code that generates playlists can manually
clear the modified flag if appropriate; there is now a proper API
for this so the tagcache and pictureflow don't need to resort to
hacks.
Change-Id: I8d3c723265a41db07a13de3f1d2abb0444528d57
Replace the minimum version bound with a check on the size of
the API struct. The version only needs to be incremented for
ABI breaking changes. Additions to the API won't need to touch
the version number, resulting in fewer merge conflicts.
Change-Id: I916a04a7bf5890dcf5d615ce30087643165f8e1f
Make 4.3 and newer doesn't interpret comments in a shell call,
whereas Make 4.2 and older do. Escaping the comment directly
works on old makes, but on new makes the backslash is passed
as well -- which we want to avoid.
The safe way to pass a literal "#" character to the shell on
both versions is by embedding it in a variable and expanding
that. It's ugly, but it works...
Change-Id: I1a217c42d747fd5aa83f9990c234e06966ac1a00
GNU grep 3.8 warns about "\#" when compiling. The "#" has to be
escaped for Makefiles and shell, but it isn't special to grep so
grep thinks it's an invalid escape sequence.
Change-Id: Ieb2607d42b7daa1939ca0101b915ec0c7afd4298
Enabling locked actions for all softlock targets accidentally
broke keylock on touchscreens because the generic touchscreen
keymap was missed. Trying to lookup CONTEXT_WPS|CONTEXT_LOCKED
returned the mapping for CONTEXT_STD because the locked version
wasn't explicitly handled.
But on almost all cases, a context's keymap does not change when
the screen is locked. It makes more sense to mask out the locked
flag and only check for it where needed.
Change-Id: I65cda2de82950d272d4394fd772286699e7c3779
rockbox_browse() overrides the desired dirfilter with
global_settings.dirfilter if you aren't using one of the
SHOW_ modes > NUM_FILTER_MODES (which come with their own sideeffects)
add flag BROWSE_DIRFILTER to
override global_settings.dirfilter with browse_context.dirfilter
add ability to set tc->browse to NULL to exit dirbrowse immediately
Change-Id: I2f40d394f9dc0864b2041293eda219f7436a7bf0
old_chunk has already been unpinned therefore the address
could be incorrect
NULL will ensure buflib_shrink uses the same start address
Change-Id: I9371e4f8263c6e5bf769108a8558dc938d4c8b87
find_font_index can use font_filename_matches_loaded_id()
font_path_to_glyph_path doesn't need strcat
Change-Id: I8d1d36a68abbc700078d651eed930035641b6240
Clip+ displays the logo with the LCD split inbetween. Calculate the y
offset to be directly after the split if possible. Thanks go to William
Wilgus for fixing a similar issue and pointing me to macros I was
looking for.
Change-Id: I2183aec1149ce74ffb840716910d2051e4a284ed
stupid bug on my part font_filename was pointing to a movable
alloc therefore that data could be moved out from under the returned pointer
instead pin the internal calls to core_get_data and
remove font_filename in favor of a function that checks if a
filename matches the loaded font as this was all it was used for anyway
Change-Id: Iebb47ffe3f81aa9d5e3968975f26d64283633ffc
store tag length to shortcut strcmp
this is nearly as fast as making a hash table using gperf
its not the hottest path but its
even slightly faster to shortcut
based on string length
if (tagstr_len > match->len)
continue;
else if (tagstr_len < match->len)
break;
but I found no measurable difference
I don't think its worth the extra constraint of keeping tags sorted
Change-Id: I4bb47cc6c5b8266d5f13c4ac78ae11d55ecb2d67
Plain import of the library parts first. Adaptions to Rockbox will
follow. A *lot* of kudos go to Mihaly Horvath for creating this library
from his already lightweight cSID-light, mainly for Rockbox. Besides a
lot of other things, he made his algorithms integer-only and
significantly improved the C64 emulation, so finally RSIDs could be
played as well as PSIDs. TinySID was nice for what it is, but this is a
quantum leap in SID playback quality for Rockbox. Check for example:
https://hvsc.csdb.dk/MUSICIANS/P/Page_Jason/Eighth.sidhttps://hvsc.csdb.dk/MUSICIANS/J/Jeff/Blowing.sid
Change-Id: I353e12fbfd7cd8696b834616e55743e7b844a73e
add some faster options to the track menu
add a shortcut for basename formatting
sort basename results
filename - by sorted filename
title - just title no formatting untagged items are unsorted as well
title mm:ss - title and duration
Change-Id: I90aea95051f6231580ef7f1fc08cd2d0d69601fb
This reverts commit 8379c6eb07.
Reason for revert:
tcs.result_len is not the same as strlen(dptr->name) + 1
since it gets filled with filename
Change-Id: I1264afc0a43ffe75adf1de00d27ab98e3adbca19
the tagcache counts the null terminator strlen however, does not
no matching entries makes for a terrible amount of disk searching
and a hang on devices without dircache to save them
Change-Id: Id3460037199ee9853de0f11e4763a7ff11130e45
only alloc enough space for the path instead of a fixed buffer
round up the path buffer size to slightly more thna the longest
common path length to reduce reallocations
Change-Id: I51b8b17584d6a905ea3a66a7c38b7b1b9da2e786
Distinguish read locks from write locks to aid converting the
mutex to an MRSW lock at some point in the future. Everything
is a write lock for now and a mutex is still used, so there's
no change in behavior.
Change-Id: I25aeed580580bdb0c3048fc7e1eaad1c32c30159
This was insanely slow for large playlists. Implement it as a
constant-time operation to avoid the O(N!) blowup, and add a
new "clear" command to the control file to avoid flooding it
with individual delete commands.
Change-Id: Ied82d1cba015591edd583f2a2ff6772108f4e176
This is all but guaranteed to be broken since pdbox doesn't use
SHAREDBSS_ATTR or SHAREDDATA_ATTR, needed to share memory safely
with COP threads.
Judging by the codec performance comparison on the wiki, if the
iRiver H100 and H300 can run PDbox on one core, PP targets should
have no problem doing it - they are 2-4x faster even on codecs
that don't have multi-core optimizations.
Change-Id: I8f556013146cb670f69cc42e80b1c768ba0f74cd
move loading file reference to a separate function
some optimization of the disk search function check tag length before
reading the actual entry & reduce number of lseek calls
this affects startup resume as well before tagcache is done loading
Change-Id: I2bd2fc53c4870416ecd1a4034b34cb6984d8ad51
Apparently a response is coming out of nowhere and tripping this
check. I can't be bothered to look into it; it would be better to
just update the ARC USB driver to the new control request API...
Change-Id: Ic5062443e060534f170d3afe17c00d3c25d1d3bd
Fix deadlocks in the dircache scan thread caused by incorrect
lock ordering. Mutating operations need to stop the thread to
prevent it from accessing invalid data; this must always be
done before taking the playlist lock to avoid deadlocking the
scan thread.
Change-Id: If719a8b28ed0b0b3eac068073581e606c4a5f58a
