summaryrefslogtreecommitdiffstatshomepage
path: root/src/devices/machine
Commit message (Collapse)AuthorAgeFilesLines
* cking_master: use clock device for irq signal hap2021-09-101-1/+1
|
* clock: add setters for period, duty cycle, pulse width hap2021-09-102-26/+150
|
* apple2: support for the AppleIISD card [R. Belmont, Florian Reitz] arbee2021-09-072-0/+328
|
* -Reworked device type definition macros a little and added more Doxygen. Vas Crabb2021-09-072-2/+1
| | | | | | | | | | | | | | | * Reworked device type definition macros to eliminate one level of indirection when using device types by name. * Fixed a potential initialisation order issue that could affect device parent ROMs. * Eliminated the need for DEFINE_DEVICE_TYPE_NS - just use DEFINE_DEVICE_TYPE or DEFINE_DEVICE_TYPE_PRIVATE with fully-qualified names. * Changed device type aliases to static auto references in the headers. * Added Doxygen comments for system/device definition macros and system flags. -Added ROM parents for m68705u3 and a2diskiing.
* machine/fdc37c665gt.cpp: Fixed address usage for serial port registers. (#8536) 9871238791132021-09-051-6/+6
|
* i8257: fix it so tc can work with multiple devices cracyc2021-09-041-2/+4
|
* 8042kbdc.cpp: mouse reset answers at least one byte if mouse disabled. yz70s2021-08-301-3/+11
| | | | Makes nforcepc bios happy.
* gts1: various fixes. Still not working though. Robbbert2021-08-271-8/+11
|
* fdc37c93x.cpp: update to support ps/2 mouse yz70s2021-08-251-0/+2
|
* 8042kbdc.cpp: improve management of ps/2 mouse yz70s2021-08-252-26/+105
| | | | | | | | - add mouse interrupt - support more mouse commands - use command byte to enable/disable mouse and keyboard interrupts Now the ps/2 mouse works with windows 98
* Major state refactoring of pc98 based HW (#8475) Angelo Salese2021-08-242-0/+199
| | | | | | | | | | | | | | | | | | | | | | | | | | | First major step in splitting up machine classes in NEC PC98 family tree: * Splits up derivative HWs into own state machine and files (pc98ha.cpp, pc9821.cpp and pc9801_epson.cpp); * Adds a preliminary uPD4991a parallel RTC, used by pc98ha; * Fix -26, -86, -118 C-bus sound board dips or jumper settings; * Fix default sound card for pc9821 (-86) and pc9821ce2 / pc9821cx3 (-118); * Adds preliminary MAD Factory Otomichan-kai C-bus sound board; * Adds boilerplate code for C-bus installing board I/Os, avoiding the possible inconvenience of board(s) getting silently unmapped by other installed boards; * Major refactoring of HW dip switches readouts, using required_ioports instead of scattering things around in PPI hooks; * Extensive QA rundown, including research on missing features and undumped machines. pc9801.cpp: Don't passthrough mouse irq frequency when cycle setting is setup too, fixes jastrike mouse input on options menu. pc9801.cpp: Fix kanji RAM window LR readback, makes telenetm to properly display 8x16 chars on RS and derivative machines. pc8801.cpp: fixed OPNA RAM readback, allowing SWs to playback ADPCMs properly. New machines marked as NOT_WORKING ---------------------------------- PC-98LT [anonymous] PC-98HA "Handy98" [anonymous] PC-9821Nr15 (98NOTE Lavie) [flyingharuka] PC-9821Nr166 (98NOTE Lavie) [flyingharuka] PC-9821Nw150 (98NOTE Lavie) [flyingharuka] PC-9821Ra266 (98MATE R) [flyingharuka] PC-9821Cx3 (98MULTi CanBe) [flyingharuka] PC-9801VX [qazmko1029] PC-9801US [CoolMod]
* formats, osd, util: Started refactoring file I/O stuff. (#8456) Vas Crabb2021-08-225-45/+63
| | | | | | | | | Added more modern generic I/O interfaces with implementation backed by stdio, osd_file and core_file, replacing io_generic. Also replaced core_file's build-in zlib compression with a filter. unzip.cpp, un7z.cpp: Added option to supply abstract I/O interface rather than filename. Converted osd_file, core_file, archive_file, chd_file and device_image_interface to use std::error_condition rather than their own error enums. Allow mounting TI-99 RPK from inside archives.
* nscsi: Added initialization of sense buffer on reset. Michael Zapf2021-08-201-0/+1
|
* swim1, swim2: Implement callback for DAT1BYTE output AJR2021-08-134-2/+55
|
* ch376.cpp: fix mount status and entry formatting [Peter Ferrie] Peter Ferrie2021-08-121-2/+4
| | | | | - BOOTi can now find .DO and .PO files - BOOTi can now boot HDVs
* ap2000: hook up reset to Centronics init signal (#8363) goldnchild2021-08-092-1/+20
|
* fdc37c93x.cpp: actually read logical device registers yz70s2021-08-072-21/+10
|
* z8536: Avoid magic numbers Dirk Best2021-08-052-8/+17
|
* z8536: Better daisy chain support, fix port assignment when writing data Dirk Best2021-08-042-5/+30
|
* fdc37c93x.cpp: use ns16550 instead of ns16450 for rs232 as per datasheet yz70s2021-08-032-4/+4
|
* smc91c96 updates (a2osx w/ lancegs) Kelvin Sherlock2021-08-022-8/+16
| | | | | | | 1. m_loopback_result needs to be a signed into so negative status codes remain negative. 2. FDSE bit was masked out 3. when full duplex switched internet is active, deferrals and collisions not used 4. implement AUTO_RELEASE
* swim1: Disable read side effects to help debugging AJR2021-07-302-10/+21
|
* applepic: Add timer and remaining GPIO AJR2021-07-302-16/+58
|
* applepic: Add device files missing from 918da8411db59ec376d800ae2a424344a6736a25 AJR2021-07-292-0/+511
|
* maciifx: Add preliminary PIC (Peripheral Interface Controller) devices and ↵ AJR2021-07-292-0/+0
| | | | | | OSS interrupt control * machine/applefdintf.*: Correct permissions
* i82371sb.cpp: various changes, windows 98 recognizes the ide busmaster ↵ yz70s2021-07-292-17/+94
| | | | | | | | | controller in pcipc driver - corrections to pci configuration space maps - remove useless redefinitions of command_w/r in ide device - placeholders for nmi, stpclk and extsmi lines of isa bridge - correctly support 8 bit writes to idetim registers
* i82439hx.cpp: corrections to pci configuration space map and decoding to ↵ yz70s2021-07-292-2/+86
| | | | debug log of pam register writes
* fdc37c93x.cpp: add placeholder routine for mouse irq yz70s2021-07-292-0/+8
|
* ns32082: implement program flow trace support Patrick Mackinlay2021-07-282-3/+35
|
* ns32202: fix syntax errors Patrick Mackinlay2021-07-281-3/+3
|
* ns32202: implement counter current values Patrick Mackinlay2021-07-282-6/+67
|
* hp_taco: commands to read HP9825 tapes added fulivi2021-07-251-15/+28
|
* srcclean for 0.234 branch Vas Crabb2021-07-251-2/+2
|
* am9517a: setting mask bits also sets status Patrick Mackinlay2021-07-231-0/+3
|
* ns32082: remove obsolete comment Patrick Mackinlay2021-07-231-5/+0
|
* ncr5380n: reclaim ncr5380 device name Patrick Mackinlay2021-07-202-58/+58
|
* API cleanups and miscellaneous fixes. Vas Crabb2021-07-152-13/+12
| | | | | | | | | | | | | | | | emu/ioport.cpp: Allow controller files to override input sequences for inputs that don't use defaults, and to override the toggle setting for digital inputs. emu/config.cpp: Expose configuration level (mostly matters for controller files), improved verbose diagnostic messages, and moved a few things out of the global and preprocessor namespaces. docs: Added documentation for some controller configuration file features. The device mapping feature documentation will be merged in at some point. util/unicode.cpp, emu/input.cpp: API cleanups.
* fidel_sc6: add note hap2021-07-141-1/+2
|
* New working machines hap2021-07-132-0/+147
| | | | | -------------------- TI-1680 [hap, Sean Riddle]
* -bus/a2bus: Added //SHH SYSTEME LANceGS Card. (#8264) ksherlock2021-07-092-28/+44
| | | | | | | | | | | -machine/smc91c9x.cpp adjustments: * Reset should disable promiscuous mode. * RCR - handle promiscuous changes, adjust soft reset handling. * EPH_STATUS should be LINK_OK by default (previous code wouldn't set LINK_OK unless RX_EN is set but LANceGS won't set RX_EN unless LINK_OK is set). * B0_BANK bits 4-8 are unspecified in the documentation but LANceGS expects a 3 (used in Card detection logic). * MIR values depend on device type and resets when the mmu resets. * Reading/writing the data register needs to be aware of 8-bit I/O. * Calculate FCS. * Removed WMS OUI filtering hack.
* video/voodoo.cpp: Major rewrite: (#8267) Aaron Giles2021-07-094-3/+13
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | * Changed to use modern poly.h instead of polylgcy.h. * Moved helper classes into separate voodoo namespace. * Derived device classes from video_device_interface. * Split classes so that later versions derive from earlier versions. * Created device maps to be directly included. * Redesigned register mapping to use helper classes and delegates. * Rewrote rasterizers to use C++ templates instead of macros. * Added logic to compute equations for color/texture combine units. * Added special generic identity-texel rasterizer cases. * Removed pipeline stalls on texture and palette changes. * Removed pipeline stalls on most all parameter changes. * Generally re-thought and cleaned up logic throughout. * Parameterized cycle stealing on status reads; updated all existing voodoo consumers to configure it as it was before. -vidoe/poly.h: Various improvements: * Exposed poly_array class for broader use. * Changed poly_array to intelligently determine maximum size. * Added logic to track multiple "last" instances in poly_array. * Extended logic to support up to 16m work items. * Removed MaxPolys parameter from poly_manager template. * Added Flags parameter to poly_manager template. * Added POLY_FLAG_NO_CLIPPING flag to remove clipping code when not needed. * poly_manager now supports a MaxParams value of 0. * Made paramcount a template parameter for render_* functions. * Added reset_after_wait() method to be overridden by derived classes. * Switched to using std:: helpers instead of internal methods. * Removed useless dependency on screen_device. * TRACK_POLY_WAITS now produces more complete statistics. -video/polylgcy.cpp: Removed legacy implementation of polygon renderer. -machine/gt64xxx.cpp: Prevent lockups by disallowing 0-duration timers. -machine/pci.cpp: Added support for adding subdevice maps directly. -emu/video/rgbsse.h: Improved min/max for SSE4.1+ and scale+clamp operations for all. -emu/vidoe/rgbutil.h: Made palette expansion constexpr and added argbexpand function. -osd/osdcore.cpp: Changed osd_ticks to use QueryPerformanceCounter on Windows since the mingw std::chrono::high_resolution_clock is anything but.
* Remove some tag lookups by name (#8252) Miodrag Milanović2021-07-056-22/+26
|
* heromem, sbmjb: Map I/O expanders (and some RAM for sbmjb) AJR2021-07-032-0/+19
|
* k056230: clear network interrupt properly (fixes screen flashing in zr107) Ville Linde2021-07-012-24/+12
|
* srcclean in preparation for release Vas Crabb2021-06-271-3/+3
|
* ds6417: fix mask cracyc2021-06-261-1/+1
|
* hd63450: unhalt cpu if a dma bus error occurs cracyc2021-06-251-0/+2
|
* vis: make memcard writing work cracyc2021-06-251-1/+1
|
* machine/latch8.cpp: Clean up a little. Vas Crabb2021-06-242-76/+72
| | | | | | | | | Making the bit read functions honour the XOR mask is easy in theory, but dkong.cpp is doing scary things with the XOR mask and bit read handlers on one of its latches, so it could end up breaking something. Honouring the read callbacks would add two tests to each of the bit read functions, which could hurt performance. The whole design of this device seems somewhat incoherent.
* vis: fix card comms and crc cracyc2021-06-232-11/+28
|