Thanks to @ripsaw8080 for insight into CD-DA channel mapping,
@hail-to-the-ryzen for testing and flagging a position-tracking bug,
and @dreamer_ for guidance and code review.
The CD-DA volume and channel mapping loops were moved to generic mixer
calls and no longer require a pre-processing loop:
- Application-controlled CD-DA volume adjustment is now applied using
an existing mixer volume scalar that was previously unused by the
CD-DA code.
- Mapping of CD-DA left and right channels is now applied at the tail
end of the mixer's sample ingest sequence.
The following have been removed:
- The CD-DA callback chunk-wise circular buffer
- The decode buffers in the Opus and MP3 decoders
- The decode buffer and conversion buffers in SDL_Sound
These removals and API changes allow the image player's buffer
to be passed-through ultimately to the audio codec, skipping multiple
intermediate buffers.
This way there's no need to prepend every line in build job with a path
to MSYS2-installed bash, and deal with problems related to escaping
embedded shell invocations.
Apt does not have a stable CLI interface, therefore should be avoided in
scripts. Using apt-get should be fine.
Split 'apt-get update' to a separate step. It makes it easier to
check, what mirrors and repositories are being used by CI machines.
Remove SPDX identifier - it's missing from other .yml files (I would
consider these configuration files and not source code, so not covered
by copyright).
Use the same build dependencies for static analysis build as in other
Linux jobs.
Runtime improvements:
- Replaces the existing audio callback routine with an efficient chunked
circular-buffer audio reader
- Replaces assumptions that all audio tracks are 44.1 kHz & stereo.
The mixer is now fed data at the actual compressed track's rate and
channel count
- Eliminates all SDL locks in the audio code in favour of mixer state
control
- Queries the codec for track-length instead of using hundreds of
iterative decimating seeks to determine track length (loading
a 99-track CUE now takes 0.1 user-seconds versus 3+ seconds
previously)
- Seeks are performed within the already-decoded buffer (for all
codecs) instead of discarding and re-running the decode sequence
- SDL_Sound's buffer-size is now set once and never resized, which
eliminates repeated re-malloc'ing in the library
- Only one seek is performed per-playback sequence followed by
sequential decodes, similar to a physical CDROM (The baseline dosbox
performs a seek for every 2352-bytes of uncompressed audio)
- The DOSBox mixer is now only active during playback sequences and
fully dormant otherwise (baseline dosbox instead performs hundreds of
calls/second with empty data)
- When using Opus audio tracks, and if your dosbox.conf [mixer]
rate=48000, then you will (very likely) achieve sample-exact
unadulterated pass-through along your entire audio chain from the
decoder, to DOSBox, to your operating system's software mixer,
to your sound card driver, to your sound card, to your speakers,
or to your digital receiver / USB speakers/headphones / or HDMI
TV/screen. This is because almost all modern hardware DACs use
a native sample rate of 48000
Source-level maintenance improvements:
- It strips all pre-processor #ifdef branching for SDL_Sound from
the code
- Fixes all codec compiler warnings (in the modified files); builds
have been tested with GCC 4 to 10, Clang 6 to 10, and MSVC 14
- Tested on Linux, macOS (Xcode), and Windows (MinGW MSYS 1.0)
operating systems
- Tested on i386, x86_64, ARM, and PowerPC (big-endian) architectures
So far, a splash screen exists only in raster format, encoded with RLE
for easy inclusion in .cpp files (src/gui/dosbox_splash.h).
Vector format is easier to modify and can be used as basis for more
artwork, to provide splash screen in better resolutions, etc.
This vector was obtained by resizing raster image to high resolution,
using a set of Gimp filters to sharpen it, then converted in Inkscape
using Path Bitmap functionality and finally optimized by hand to e.g.
remove excessive path nodes.
To convert it to a format easily usable from C/C++: open the file in
Gimp, edit to fit your requirements and export as C header.
This null-check resolves a potential issue detected by static analysis.
It might be the case, that crash never happens due to the way this
static function is used at the moment, and because DYNFLG_CHANGED is being
cleared in dnew->genreg->Clear() few lines before, but the crash might
still happen if initial state of flags is inconsistent or surrounding
code will be changed even a little bit.
This nullcheck makes the code more robust at no performance penalty.
Macros FRAMES_TO_MSF and MSF_TO_FRAMES come from SDL_cdrom library.
SDL_cdrom was removed from SDL2 and these macros are used also in code
not directly related to SDL_cdrom library, so a replacement is needed.
Turns out output pointer type is different on Windows (int*) than it is on
Linux (unsigned char*); given choice between using void* for parameters and
a template function, I prefer template.
The commit makes the following changes:
- The package listing script now requires the user specify which package manager
they're using. This approach resolves the ambiguity if a system has more than
one package manager (ie: macports & brew)
- Adds packages for Fedora, RedHat/CentOS, Arch, and OpenSuse
- Eliminates unecessary code in the package manager script
(more can be eliminate at the expense of complexity)
- Made a couple minor fixes to the build script
- Tried to further "standardize" the workflows as follows:
- names are Compiler Version (Environment)
- Sorted them alphabetically in their respective YAMLs
- Minor spacing adjustment to align values (where it makes sense)
- Dropped quotes around some of the string values because I'd
rather our YAML be consistent and propper instead of changing our
YAML to suite the limitations of an editor (can a different plugin
or better parser be used?)
- Added macOS workflows for Homebrew and MacPorts, both ready to
go and tested, but with the build step just commented out
- Removes extraneous (and unused) includes
- Cleans up many compiler warnings
- Fixes SDL include statements (using proper `#include <file.h>` directives
given the prior `#include "file.h"` directives should only be used when
including a projects' own header files instead of -Ipath-provided system
headers)
- Eliminates extraneous codecs
Many of these were made after building SDL_Sound under various compilers
and operating systems and hitting various issues. It currently builds
clean without errors or warnings on all our workflow compilers and
versions.
This change makes a couple changes to the CI workflow:
- Adds more compiler coverage:
- gcc to MacOS (see note below)
- 32 and 64bit gcc and clang to Windows
- With more builds, this separates them into per-OS workflow YAMLs
(laying the foundation for more build environments: BSD? DOS? ... )
- Moves all functional commands from GitHub-syntax-YAML into scripts,
which (besides eliminating repeated code), now serve a dual-purpose
of being runnable outside of GitHub.
- One script takes care of listing dependent packages for the given
runtime environment
- Another script takes care of configuring and building
These scripts can be leveraged by a nightly build & asset generator in
the future.
Note: adding GCC to MacOS is now "correct" from a build perspective,
however to keep this PR focussed on the CI workflow I have not included
the coreMIDI / AppleBlocks code-fixes here (so for now, the gcc macOS
builds will fail; we will merge the coreMIDI / AppleBlocks later
depending on how upstream wants to handle it).
log-env.sh is cross-platform (works on Linux, MacOS and Windows)
log-env.ps1 is Windows-only and requirs specifying pwsh shell, but
provides some Windows-specific information, that might be useful e.g.
for MSVC builds.
