summaryrefslogtreecommitdiffstatshomepage
path: root/3rdparty
Commit message (Collapse)AuthorAgeFilesLines
* Update bgfx to latest Бранимир Караџић2021-08-107652-1088655/+173249
|
* Update bimg to latest Бранимир Караџић2021-08-1041-11473/+18285
|
* Update bx to latest Бранимир Караџић2021-08-10140-1622/+3184
|
* ymf278b: Allow PCM writes if new2flag is set, even if newflag is not. Fixes ↵ Aaron Giles2021-07-241-10/+12
| | | | MT08033.
* ymfm: Return masked status from set/reset function. Fixes MT08034. Aaron Giles2021-07-231-1/+1
|
* ymfm: Sync with upstream Aaron Giles2021-07-2311-60/+78
| | | | | * Fix LFO disable on YM2151 * Fix swapped lower nibbles for 12-bit PCM on OPL4
* Implement 68k FPU FATAN instruction (#8287) Calvin Buckley2021-07-182-0/+392
|
* Fix cross compiling benchmarks Miodrag Milanovic2021-06-224-8/+8
|
* ymfm: Play ADPCM-A samples through the end of the end block inclusive Aaron Giles2021-06-221-10/+15
|
* Fix reporting of ADPCM-B end-of-sample. Seems to fix MT08006; hopefully will ↵ Aaron Giles2021-06-212-6/+12
| | | | fix other neogeo issues as well.
* ymfm: Sync with upstream: Aaron Giles2021-06-018-214/+391
| | | | | * Fix panning calculations in OPL4 (YMF278B) * Make OPQ reverb less ridiculous
* ymfm: Sync with upstream: Aaron Giles2021-05-298-52/+50
| | | | | * Match OPN LFO frequencies to hardware measurements * Improve OPQ detune, reverb, and KSR
* ymfm: Don't mask off top bit of octave for OPQ. Aaron Giles2021-05-231-2/+2
|
* ymfm: Improve OPQ behavior for timers and register access. Add YM3533 device ↵ Aaron Giles2021-05-232-11/+28
| | | | and use that instead of YM3806.
* ymfm: Fix OPZ fine tuning and envelope shift. Now pretty much all the TX81Z ↵ Aaron Giles2021-05-232-13/+16
| | | | built-in instruments sound reasonable.
* ymfm: Some progress on OPZ. Some instruments in the TX81Z sound kind of ok now. Aaron Giles2021-05-232-68/+122
|
* ymfm: Fix OPL3 waveform 7 to better match hardware. Aaron Giles2021-05-221-1/+1
|
* ymfm: Add missing clamp in YM2608 Aaron Giles2021-05-221-1/+2
|
* ymfm: Boost ADPCM-B volume by a factor of 2 to more closely match previous ↵ Aaron Giles2021-05-221-2/+2
| | | | levels. Addresses MT #07956
* ymfm: Sync with latest, add complete YMF278B support (#8090) Aaron Giles2021-05-2217-240/+2164
| | | | | | | | | | | * Sync with upstream. I/O callbacks are now consolidated into a single read callback and a single write callback, with an access type specifier. * Initial working implementation of YM278B. Most features implemented, except vibrato. * Implement vibrato and status register flags. Fix envelope rate computation. * Rename ymfm_interface::external_type to access_class and clean up the fallout. * Formally replace the old YMF278B engine with the one from ymfm * Rotated YMF278B outputs into a more logical order. * Re-evaluted envelope calculations and 2x works better than the weird 15/8 I came up with before. Also changed the way FM resampling is computed to be more precise (and simpler). Turned off extraneous debugging. * Start of/reset to a null state with no loaded waveforms. * Fix YM2608 I/O ports.
* ymfm: Switch to internal SSG for YM2203/2608/2610. Clean up code to ↵ Aaron Giles2021-05-151-3/+6
| | | | determine the number of SSG outputs.
* ymfm: Refactor new FM engine into a 3rdparty library (#8046) Aaron Giles2021-05-1426-0/+15646
| | | | | | | | | | | | | | | ymfm: refactor the code into a separate 3rdparty library * Moved ymfm core implementation to 3rdparty/ymfm * Split out each family (OPM/OPN/OPL/etc) into its own source file * Added preliminary OPQ and OPZ support, still WIP * Put all 3rdparty code into its own namespace ymfm * Fixed various bugs reported in #8042 * Created interface class for communication between the 3rdparty engine and the emulator * Standardized MAME implementation of all Yamaha devices based on a template class * Created standard base class ym_generic that can be used when multiple YM chips are swapped in * Changed YM2203/2608/2610 to embed a YM2149 as a subdevice instead of deriving from ay8910_device * Also provided compile-time option to use a simplified built-in SSG rather than using MAME's at all (currently off) * Consolidated MAME header files from one-per-chip (ym2151.h, ym2203.h, etc) to one-per-family (ymopm.h, ymopn.h, etc)
* -Improved accuracy of System 573's digital I/O audio emulation. (#7664) 9871238791132021-01-162-190/+1240
| | | | | * Refactored mas3507d and k573dio/fpga to be more accurate to real hardware. -3rdparty/minimp3: Updated to latest master and removed local changes.
* SDL2: temp macOS compile fix for bundled SDL2 pending the 2.0.14 release [R. ↵ arbee2020-12-121-0/+4
| | | | Belmont]
* linenoise: properly handle numpad enter cracyc2020-12-051-0/+6
|
* Minimal changes to sol2 and bgfx to compile as C++17. Vas Crabb2020-11-152-2/+2
| | | | These changes should be sent upstream as pull requests.
* -Switch to building MAME as C++17. Vas Crabb2020-11-15308-63299/+31239
| | | | | | | * Updated sol2 to 3.2.2 * Updated pugixml to 1.10 * Increased minimum clang version to 6 * Cleaned up some stuff that can use new features
* -3rdparty/expat: Update to 2.2.10." Vas Crabb2020-10-15179-24358/+44285
| | | | -Fixed tiny build (missing s11c_bg_device) and unused lambda capture in emu/rendlay.cpp.
* 3rdparty/nanosvg: Re-base on latest upstream. Vas Crabb2020-10-055-107/+279
| | | | | | | Now based on upstream cc6c08d3a80f1a305021af3d6394cdf1535d02a2. Among other things, this version is supposed to be less sensitive to the global locale.
* 3rdparty/sqlite3: Updated to SQLite 3.33.0 Vas Crabb2020-10-044-30673/+78635
|
* Fix SOL2 build on GCC 10.2 by working around overload resolution problem AJR2020-07-287-18/+18
|
* asmjit: sync with upstream (nw) Patrick Mackinlay2020-06-182-2/+5
|
* asmjit: mame changes (nw) Patrick Mackinlay2020-06-122-1/+3
|
* asmjist: sync with upstream (nw) Patrick Mackinlay2020-06-12122-11020/+14214
|
* Fix build even more (nw) AJR2020-05-301-2/+2
|
* Fix build (nw) AJR2020-05-301-1/+1
|
* asmjit: new 3rdparty library Patrick Mackinlay2020-05-28128-0/+77219
|
* rv unintentional change (nw) Justin Kerk2020-05-221-0/+242
|
* New not working software list additions Justin Kerk2020-05-221-242/+0
| | | | | | | --------------------------------------- ibm5150: Wizzardz & War Lordz [LGR] (nw) cleanups, move SimLife to ibm5170 (#6728)
* Removed winpcap and cleaned up network module selection. Vas Crabb2020-03-05318-56378/+0
| | | | | | | | (nw) The pcap.h header itself has the problematic original BSD license, including the obnoxious advertising clause. Using tap/tun networking on Windows provides a much better experience, so the extra setup is worth it. This patch also allows you to enable pcap on platforms where it's disabled by default with USE_PCAP=1 if you really want to use it.
* Maintenance: Vas Crabb2020-03-043-5/+1283
| | | | | | | * Removed empty nl_examples from dist.mak * Added copyright acknowledgements and full text of licenses to binary distribution * Fixed up the list of third-party libraries * Moved WDL fft.c to 3rdparty
* Sync with GENie upstream revision ce9f3c5 (#6262) Julian Sikorski2020-02-0615-136/+177
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | * Change makerules to take variable $(PROJECT_TYPE) instead of hardcoded 'gmake' This allows to run `make projgen PROJECT_TYPE=ninja` to generate ninja build files instead. and to build GENie using ninja by running `make release PROJECT_TYPE=ninja`. Using ninja improves build times, e.g. for macOS: 12.47s with gmake goes down to 2.05s with ninja. Signed-off-by: Julian Sikorski <belegdol+github@gmail.com> * Set prefer project set as solution.startproject as default target Signed-off-by: Julian Sikorski <belegdol+github@gmail.com> * Adding `GenerateMapFiles` flag. Causes Visual Studio's linker to generate .map files for that configuration. Signed-off-by: Julian Sikorski <belegdol+github@gmail.com> * Add ninja support for 'wholearchive' libraries Signed-off-by: Julian Sikorski <belegdol+github@gmail.com> * Fixup ninja.esc to gracefully ignore nil passed as value Signed-off-by: Julian Sikorski <belegdol+github@gmail.com> * Use -Wl,-force_load for wholearchive libs when building for macosx Signed-off-by: Julian Sikorski <belegdol+github@gmail.com> * Updated README. Signed-off-by: Julian Sikorski <belegdol+github@gmail.com> * Add space after filename in 'Generating' message This makes the filename 'clickable' to open in iTerm and VSCode Signed-off-by: Julian Sikorski <belegdol+github@gmail.com> * Print generated filenames as quoted string Signed-off-by: Julian Sikorski <belegdol+github@gmail.com> * Remove '...' after filename Signed-off-by: Julian Sikorski <belegdol+github@gmail.com> * Pop cwd after pushing to run file Signed-off-by: Julian Sikorski <belegdol+github@gmail.com> * Make paths in embed.lua rely on script dir Signed-off-by: Julian Sikorski <belegdol+github@gmail.com> * Updated README. Signed-off-by: Julian Sikorski <belegdol+github@gmail.com> * Fix issues caused by make-4.3 no longer adding spaces to variables in some cases The fix was found by @asavah. Signed-off-by: Julian Sikorski <belegdol+github@gmail.com> * Fixed release script. Signed-off-by: Julian Sikorski <belegdol+github@gmail.com> * Updated README. Signed-off-by: Julian Sikorski <belegdol+github@gmail.com> * Update scripts.c Co-authored-by: Christian Helmich <kagekirin@gmail.com> Co-authored-by: Johan Sköld <johan@skold.cc> Co-authored-by: Бранимир Караџић <branimirkaradzic@gmail.com>
* Revert "Sync with GENie upstream revision ce9f3c5 (#6262)" (#6263) Robert2020-02-0615-177/+136
| | | This reverts commit a06d8e9cc68e6c91fcd58770e972ff81ffd20275.
* Sync with GENie upstream revision ce9f3c5 (#6262) Julian Sikorski2020-02-0515-136/+177
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | * Change makerules to take variable $(PROJECT_TYPE) instead of hardcoded 'gmake' This allows to run `make projgen PROJECT_TYPE=ninja` to generate ninja build files instead. and to build GENie using ninja by running `make release PROJECT_TYPE=ninja`. Using ninja improves build times, e.g. for macOS: 12.47s with gmake goes down to 2.05s with ninja. Signed-off-by: Julian Sikorski <belegdol+github@gmail.com> * Set prefer project set as solution.startproject as default target Signed-off-by: Julian Sikorski <belegdol+github@gmail.com> * Adding `GenerateMapFiles` flag. Causes Visual Studio's linker to generate .map files for that configuration. Signed-off-by: Julian Sikorski <belegdol+github@gmail.com> * Add ninja support for 'wholearchive' libraries Signed-off-by: Julian Sikorski <belegdol+github@gmail.com> * Fixup ninja.esc to gracefully ignore nil passed as value Signed-off-by: Julian Sikorski <belegdol+github@gmail.com> * Use -Wl,-force_load for wholearchive libs when building for macosx Signed-off-by: Julian Sikorski <belegdol+github@gmail.com> * Updated README. Signed-off-by: Julian Sikorski <belegdol+github@gmail.com> * Add space after filename in 'Generating' message This makes the filename 'clickable' to open in iTerm and VSCode Signed-off-by: Julian Sikorski <belegdol+github@gmail.com> * Print generated filenames as quoted string Signed-off-by: Julian Sikorski <belegdol+github@gmail.com> * Remove '...' after filename Signed-off-by: Julian Sikorski <belegdol+github@gmail.com> * Pop cwd after pushing to run file Signed-off-by: Julian Sikorski <belegdol+github@gmail.com> * Make paths in embed.lua rely on script dir Signed-off-by: Julian Sikorski <belegdol+github@gmail.com> * Updated README. Signed-off-by: Julian Sikorski <belegdol+github@gmail.com> * Fix issues caused by make-4.3 no longer adding spaces to variables in some cases The fix was found by @asavah. Signed-off-by: Julian Sikorski <belegdol+github@gmail.com> * Fixed release script. Signed-off-by: Julian Sikorski <belegdol+github@gmail.com> * Updated README. Signed-off-by: Julian Sikorski <belegdol+github@gmail.com> * Update scripts.c Co-authored-by: Christian Helmich <kagekirin@gmail.com> Co-authored-by: Johan Sköld <johan@skold.cc> Co-authored-by: Бранимир Караџић <branimirkaradzic@gmail.com>
* Sync bgfx, bx and bimg with upstream; fix the hlsl fs_chroma.sc matrix ↵ Julian Sikorski2019-11-121300-11732/+78037
| | | | | | | | | | | | | | transposition properly (#5875) * Sync bgfx with upstream revision 280420d * Sync bx with upstream revision 267727d * Sync bimg with upstream revision c1bab10 * Use bgfx helper functions to ensure correct matrix orientation * Rebuild hlsl chain
* Fix waitvsync for bgfx vulkan backend. [Couriersud] Couriersud2019-10-271-0/+39
| | | | | | | | | Currently waitvsync is broken in the bgfx vulkan backend. The option is just ignored and a mode waiting for vsync is selected. This also breaks throttle. -nothrottle doesn't work. These changes fix this. If available a non-syncing mode is used if -nowaitvsync is used.
* Merge pull request #5758 from vadosnaprimer/luaengine_ram R. Belmont2019-10-191-1/+1
|\ | | | | luaengine: reduce memory usage during compilation
| * luaengine: reduce memory usage during compilation feos2019-10-191-1/+1
| | | | | | | | | | | | | | | | | | | | Previous semantics were encouraged by sol2, and ThePhD admitted it was a bad idea for usertypes with lots of members. sol3 allows to assign members outside a constructor for regular usertypes, but mame can't move to sol3 because it requires C++17. Turns out sol2 also has a way to add members separately, and that is what I used. This helps immensely. This approach required a fix to warning C5046 copied from here https://github.com/ThePhD/sol2/commit/57990845726e17fba11a39cfcb1fc0127a7ea638 Tested current mame plugins, everything seems to work. PS: Due to come black magic, if I use simple_usertype::set() with "manager" as well, "machine" member dies (others are fine!), so I had to leave it intact.
* | Revert accidental commit of bgfx hacks. (nw) couriersud2019-10-182-6/+3
| |
* | Continuous integration improvements (#5703) Julian Sikorski2019-10-181-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | * Add workaround for imgtool and jedutil failing vs2019 debug builds with /ZI * No longer allow msvc build to fail * Enable tools build for travis to make it more useful * Switch travis to Xcode 11 in order to fix nltool linking failure * Prefer 64-bit compiler with VS 2019 too * Setting PreferredToolArchitecture to x64 is not needed, genie puts it into the project files for vs2015 or later * OPTIMIZE=1 build is faster that OPTIMIZE=0 for some reason. So fast in fact, that TOOLS=1 can be enabled without hitting the 60 minute timeout * Switch MINGW build to VS 2017 image until appveyor figure out why builds on VS 2019 are almost twice as slow * Run pacman twice to account for core system upgrades