The new output type `texturepp' was added, which implements
pixel-perfect scaling using SDL's hardware-accelerated texture output.
In pixel-pefect mode, each original pixel is displayed as a rectangle m
by n pixels, so that m:n yields a reasonably good approximation of the
pixel aspect ratio (PAR) of the emulated graphical mode while using as
much screen space as possible. The balance between the precision of
aspect ratio and the utilisation of screen space is specified as the
`parweight' parameter to pp_getscale() and is currently hard-coded in
sdlmain.cpp.
This implementation emulatates pixel-perfect mode as a special case of
nearest-neighbor interpolation when the horisontal and vertical scaling
factors are integers.
The hashest generated by xxHash have changed between v0.7.2 and v0.7.3,
therefore the hashes in prior-genreated lookup tables will no longer
be valid. By bumping the file's identifier, the fast-seek code will
reject these older files and regenerate new ones.
Context: The codecs implement a read-write callback function
(RWops) used to read N bytes from the underlying binary stream
into a buffer. In some cases, a codec might only return a
subset of the requested bytes and requires subsequent read()
requests to get the remaining bytes. Internally, the codec
might have to reposition or run second decode sequence to get
the bytes.
The RWops callbacks for the various codecs were inconsistently
implemented: some performed the above mentioned subsequent
re-read attempts while others simply accepted whatever we got
after the first read attempt. This commit makes them the same
by attempting to re-read ("get the requested bytes at all
costs") until the underly stream goes EOF.
Some of these RWops functions also contained a book-keeping
bug from upstream that resulted in over-reading after
under-delivering on the first read attempt. The concequence
being that too much data would be written to the buffer
(writing past its end) and leaving the underlying stream's
file position too-far-forward.
- Cleanup some types where precision is lost.
- Explicitly cast between types (fixes all effc++ warnings)
- Simplify lossy and cast-heavy floating-point coversions with ceil_divide
Adds a `--disable-opus-cdda` flag that explicitly disables support
for Ogg Opus CDDA tracks and in turn avoid the need for the Opus package
dependencies such as the opusfile, opus, and ogg libraries.
This feature does not alter the default operation of ./configure, which
is to enable Opus CDDA support and quit if the Opus dependency package,
opusfile, is not found. The user can then choose to either a) install
the package or b) explicitly disable Opus support.
This commit also includes:
- fixes for a double-free in the MP3 close routine that
was discovered during testing
- a message if a CD audio track cannot be added during CDROM
mounting (such as attempting to use Opus tracks when the binary
does not support them).
- the --disable-opus-cdda flag in our config heavy workflow
Coverity finds use after free on value pointed by sample_list, but fails
to detect, that this is a global variable, that is being updated by
Sound_FreeSample before freeing any memory.
This code is obviously written with the purpose of freeing whole list,
so there's nothing unexpected here - this is definitely a false
positive finding.
We could mark this issue as false-positive in Coverity, but I think it's
better to avoid tripping it in the first place.
The eliminated code removes the (ch == 1) branch, which is scoped
within this if condition: `if (rtype == 2 && ch != 1)`, therefore
the (ch == 1) branch will never be taken.
In the call to decode_residue:
decode_residue(f, residue_buffers, ch, n2, r, do_not_decode);
The channel count is previously intialized as zero and incremented
based on a for-loop (f->channels) plus a conditional,
if (map->chan[j].mux == i). If this doesn't happen then 'ch'
remains zero.
Once inside decode_residue(..), the code has three branches based
on channel count: stereo (ch == 2), mono (ch == 1), and then the
exception if it's neither of those (simple 'else'). It's in here
where a zero-valued 'ch' can be used as the denominator in these
calculations:
int c_inter = z % ch
p_inter = z/ch;
Obviously this 'else' branch is meant for channel counts greater
than two an not for zero channels; so this change simply makes
that branch only valid if (ch > 2).
