Nouvelle version de Mame.<\/p>\n
Source Changes Removed machine from the parameters of some of the device_get_info Changed device interfaces to pass the device_config * rather A number of changes to remove references to screens by index and * changed vblank_state_changed_func to now have a device_config Changed the MC6845 to take and pass device_config's only. The Removed video_data from running_machine. Private state screen Fixed the crash in Galaga caused by the slightly altered video timing. Converted address maps to tokens. Changed the address_map structure Fixed several instances of porttagtohandler*() in the address maps. ADDRESS_MAP_EXTERN() now is required to specify the number of Removed ADDRESS_MAP_FLAGS() grossness. There are now three new macros Changed internal memory mapping behavior to keep only a single Changed MC6845 to take a screen tag instead of a screen number. Voodoo now takes a device_config for the screen. [Zsolt Vasvari]<\/p>\n Changed TMS34010 to take a screen tag instead of screen number. Replaced the following macros (SMH == static memory handler)<\/p>\n MRA*_BANK*\/MRA*_BANK* -> SMH_BANK* This removes the silly need for a bunch of redundant constants Moved some memory system constants into memory.c. [Aaron Giles]<\/p>\n Converted atarigen and tms34061 to be handle based. [Zsolt Vasvari]<\/p>\n Changed how the PowerPC MMU is enabled at compile time, to cut down Some additional memory system cleanups: [Aaron Giles] Adds some new 6845 varieties: MC6845-1, H46505, HD6845 and SY6545-1. Moved crosshair code from video.c into its own module. Simplified Saturate MultiPCM on clip and fix envelope discontinuity. [ElSemi]<\/p>\n Fixed loading mixer settings when the driver default value is Removed cpu_scalebyfcount(). Changed input ports to register a frame Fixed VBLANK end time computation and video_screen_get_vblank() -- Fixed 68000 prefetching operation. Previous behavior was incorrect. Changed slapstic management to always install an opbase handler to superqix.c changes: [Mike Haaland] Added some clock frequency changes which were documented in u4 but Added layout to crop corners in Electric Yo-Yo. [Mr. Do!]<\/p>\n Added support for Sega Virtua Processor, to run Genesis\/MegaDrive Phase 2 of the Deal 'Em input changes (these work well with the Simplified inputs, improved DIP switches, and added DIP locations In asuka.c, created MACHINE_START( asuka ) to configure cpu2 banks Fixed M68000 build logic to only include it if enabled. [Fabio Priuli]<\/p>\n New games added or promoted from NOT_WORKING status New clones addedMAMETesters Bugs Fixed
\n----------------------
\n01216 [Tomasz Slanina]
\n00898 [couriersud]
\n01436 [Aaron Giles]
\n01480 [robiza]
\n01482 [Zsolt Vasvari]
\n01429 [RansAckeR]
\n00593 [RansAckeR]
\n01478 [RansAckeR]<\/p>\n
\n--------------
\nFixed flickering in Electric Yo-Yo. Changed qix driver to use the
\nMC6845's VSYNC notification for interrupts. [Zsolt Vasvari]<\/p>\n
\nand device_set_info calls because that information is stored
\nwith the device now. [Aaron Giles]<\/p>\n
\nthan tokens and individual bits to the device callbacks. Updated
\nall existing devices accordingly. [Aaron Giles]<\/p>\n
\nmake the video system based strictly upon the screen devices:
\n[Zsolt Vasvari]<\/p>\n
\nargument
\n* removed runnning_machine as the first parameter of a VBLANK
\ncallback
\n* removed video_data from running_machine
\n* private state screen information now hangs off the public one,
\ninstead of the other way around
\n* combined video_global and video_private into a single structure
\n* moved populating screen_state from mame.c to video.c
\n* removed defstate; the default screen parameters live in
\nscreen_config directly
\n* moved 'refresh' from public to internal video state and renamed
\nit to 'frame_period'
\n* changed all drivers that previously accessed 'refresh' to use
\nvideo_screen_get_frame_period()
\n* changed VIDEO_UPDATE signature to pass the screen device
\n* added const device_config *primary_screen to running_machine
\n* changed first argument of video_screen_* functions to take a
\ndevice_config; all callers that used 0 for the screen number now
\nuse machine->primary_screen
\n* changed game info screen and overlay UI to display the screen tag
\ninstead of screen number
\n* added video_screen_get_visible_area()
\n* changed render.c to use screen devices
\n* removed direct driver access to machine->screen[] array
\n* added video_screen_auto_bitmap_alloc(screen)
\n* the Dynax\/Don Den Lover games now do their updating in VIDEO_UPDATE
\ninstead of VIDEO_EOF. Seems to have fixed the palette problems
\n* hooked up swapped videoram in Karnov properly -- same idea as
\nBurger Time
\n* removed the public screen_state object -- all access is via
\nfunctions
\n* removed screen[] object from running_machine
\n* removed MAX_SCREENS constant -- there is no longer a hardcoded
\nupper bound<\/p>\n
\nfollowing shook out from this that are applicable to other devices
\nas well: [Zsolt Vasvari]
\n- running_machine is no longer taken, passed or stored
\n- mc6845_t is now private
\n- since passing device_config's is not device type safe, I added a
\nget_safe_token() to mc6845.c to check the device passed in for
\nthe valid type.<\/p>\n
\ninformation now hangs off the public one, instead of the other way
\naround. Combined video_global and video_private into a single
\nstructure. [Zsolt Vasvari]<\/p>\n
\n[Zsolt Vasvari]<\/p>\n
\nto house global map-wide information and hung a list of entries off
\nof it corresponding to each address range. Introduced new functions
\naddress_map_alloc() and address_map_free() to build\/destroy these
\nstructures. Updated all code as necessary. [Aaron Giles]<\/p>\n
\nDrivers should use AM_READ_PORT() macros instead. [Aaron Giles]<\/p>\n
\ndatabits, just like ADDRESS_MAP_START. [Aaron Giles]<\/p>\n
\nwhich replace its former usage. ADDRESS_MAP_GLOBAL_MASK(mask)
\nspecifies a global address-space-wide mask on all addresses. Useful
\nfor cases where one or more address lines simply are not used at
\nall. And ADDRESS_MAP_UNMAP_LOW\/HIGH specifies the behavior of
\nunmapped reads (do they come back as 0 or ~0). [Aaron Giles]<\/p>\n
\naddress map and store the byte-adjusted values next in the address
\nmap entries rather than maintaining two separate maps. Many other
\nsmall internal changes\/cleanups. [Aaron Giles]<\/p>\n
\n[Zsolt Vasvari]<\/p>\n
\n[Zsolt Vasvari]<\/p>\n
\nMRA*_RAM\/MRA*_ROM -> SMH_RAM
\nMRA*_ROM\/MWA*_ROM -> SMH_ROM
\nMRA*_NOP\/MWA*_NOP -> SMH_NOP
\nMRA*_UNMAP\/MWA*_UNMAP -> SMH_UNMAP<\/p>\n
\nwith faux type definitions that didn't buy anything. [Aaron Giles]<\/p>\n
\non the amount of unused functions in MAME. [Nathan Woods]<\/p>\n
\n* Made address versus byte offsets explicit throughout.
\n* Removed some unused parameters.
\n* Consolidated initialization process.
\n* Removed maximum memory block count.
\n* Centralized bank management.
\n* Added masked handlers and change_pc to the accessors.
\n* Added memory_get_accessors() to return a pointer to the accessors
\nfor a given address space\/databus width\/endian configuration.<\/p>\n
\n[Zsolt Vasvari]<\/p>\n
\ncrosshair logic and made it screen device based. [Zsolt Vasvari]<\/p>\n
\ncalculated. [Nicola Salmoria]<\/p>\n
\ncallback, which is called immediately after throttling and updating.
\nThis is the proper \"sync point\" between emulated time and real time.
\nMoved all analog and digital port processing into a central place
\nhere. Added tracking of time since the previous frame update and use
\nthat as an estimate for the time of the current frame. This is used
\nto scale analog ports without the use of cpu_scalebyfcount(). This
\nis not perfect in the case where frame rates are dynamic (vector
\ngames), but works well for other cases. [Aaron Giles]<\/p>\n
\nthey weren't taking into account games that use
\nMDRV_SCREEN_VBLANK_TIME. [Zsolt Vasvari]<\/p>\n
\n[Aaron Giles]<\/p>\n
\nmore aggressively catch code executing in the slapstic region.
\nUpdated all drivers to separate the slapstic region of ROM into a
\ndifferent ROM section from the fixed ROM. [Aaron Giles]<\/p>\n
\n- Changed vblank interrupts to 3 to fix screen drawing issues.
\n- Used memory_configure_bank() to configure memory.
\n- Added save states.<\/p>\n
\nwhich did not apply due to operator error. [Corrado Tomaselli]<\/p>\n
\nversion of Virtua Racing, intended to be used by MESS. It consists
\nof a CPU core SSP1601, and updates in megadriv.c: [notaz]
\n* SSP1601 replaces SSP1610, as it has been confirmed that SVP
\nactually contains SSP1601. The current SSP1610 is only a
\nplaceholder (nearly completely unimplemented) anyway.
\n* Changed megadriv.c add a new driver for Genesis\/MegaDrive+SVP
\ncombination, also add SVP memory controller logic and memory map.<\/p>\n
\ndefault cabinet type, I need to do some conditional input work
\nto fix them fully), and also a conversion to the resnet code.
\nUnfortunately, this still doesn't fix the apparently needed
\ntransparency. [James Wallace]<\/p>\n
\nto punchout.c, bankp.c, and baraduke.c. [RansAckeR]<\/p>\n
\nwith memory_configure_bank(). All drivers support save states, with
\nthe exception of cadash and clones, the background does not load
\nproperly. [Mike Haaland]<\/p>\n
\n---------------------------------------------------
\nAmuse (Version 50.08 IBA) [Pierpaolo Prazzoli]
\nJoker Poker (Version 16.03B) [Pierpaolo Prazzoli]
\nSuper Bowl (Version 16.03B) [Pierpaolo Prazzoli]
\nPit Boss II [Mariusz Wojcieszek]
\nSuper Pit Boss [Mariusz Wojcieszek]
\nPit Boss Megastar [Mariusz Wojcieszek]
\nMegatouch IV [Mariusz Wojcieszek]
\nMegatouch IV Tournament Edition [Mariusz Wojcieszek]
\nMegatouch 6 [Mariusz Wojcieszek]<\/p>\n
\n----------------
\nRaiden Fighters (Austria) [Corrado Tomaselli]
\n[\/CODE]
\n<\/p>\n