- Status Unconfirmed
- Percent Complete
- Task Type Bugs
- Category Codecs
- Assigned To No-one
- Operating System SW-codec
- Severity Low
- Priority Very Low
- Reported Version Release 3.10
- Due in Version Undecided
-
Due Date
Undecided
- Votes
- Private
FS#12485 - Seeking in .ape file doesn't work
The .ape file I join make rockbox produce a horrible noise (like random audio) when seeking inside it. The original file was very large, so I created this one with:
dd if=original.ape of=test.ape bs=512 count=1000
So the produced file may not be correct, but it plays correctly in rockbox, and reproduces the bug, so I join it…
test.ape
(500 KiB)
Loading...
Available keyboard shortcuts
- Alt + ⇧ Shift + l Login Dialog / Logout
- Alt + ⇧ Shift + a Add new task
- Alt + ⇧ Shift + m My searches
- Alt + ⇧ Shift + t focus taskid search
Tasklist
- o open selected task
- j move cursor down
- k move cursor up
Task Details
- n Next task
- p Previous task
- Alt + ⇧ Shift + e ↵ Enter Edit this task
- Alt + ⇧ Shift + w watch task
- Alt + ⇧ Shift + y Close Task
Task Editing
- Alt + ⇧ Shift + s save task
I see that ape.c does not check for end of file. It does not check the return value of ci→seek_buffer() or ci→request_buffer(), and it doesn’t check bytesleft from ci→request_buffer(). This explains what happens when seeking past end of file in test.ape. It can even lead to crashes, eg.
at apps/codecs/demac/libdemac/entropy.c:176
176 rc.buffer = (rc.buffer « 8) | read_byte();
I also see that even seeks to valid positions within test.ape lead to noise. These lines in codec_run() change the return value of ape_calc_seekpos() and cause the problem:
/* APE’s bytestream is weird… */
firstbyte = 3 - (newfilepos & 3);
newfilepos &= ~3;
For example, frame 0 can be decoded with firstbyte = 3 (as set before the start of playback) and newfilepos = 2570 (that equals ci→curpos at start of playback and ape_ctx→seektable[0]). However, seeking works on other APE files, so more investigation is needed.