Commit Graph

  • 0bd7d8dd67 aica: 44100 samples a second, exactly LibretroAdmin 2026-10-10 22:24:47 +02:00
  • f8fe442bd1 aica: the DMA that clears registers stops at its length; a verify() the DSP could reach LibretroAdmin 2026-10-10 22:24:47 +02:00
  • 57bdf2a8dd aica: the pitch keeps all its bits LibretroAdmin 2026-10-10 22:24:47 +02:00
  • 06f09dafae aica: the amplitude LFO is half as deep; its noise wave has 8 bits LibretroAdmin 2026-10-10 22:24:47 +02:00
  • 8bb566d971 sh4/dmac: channel 2's length of 0, where it leaves the address, flags not set by writing 1 LibretroAdmin 2026-10-10 22:24:46 +02:00
  • 3e66b13eaf sh4/intc: interrupts of the same level in the processor's order LibretroAdmin 2026-10-10 22:24:46 +02:00
  • 35c5a6fa96 sh4/tmu: a timer's period is TCOR + 1; exact in 64 bits; flags not set by writing 1 LibretroAdmin 2026-10-10 22:24:46 +02:00
  • 48ffdd2c53 tools/threads: a state of another machine is refused LibretroAdmin 2026-10-10 19:30:38 +02:00
  • 226921c2f1 aica: the envelope's low two bits LibretroAdmin 2026-10-10 19:30:38 +02:00
  • 186d346e80 aica: channels are added up at 20 bits, and level and pan are one multiplication LibretroAdmin 2026-10-10 19:30:38 +02:00
  • 43c450e452 rend: dumped textures are widened as drawn ones are LibretroAdmin 2026-10-10 19:30:38 +02:00
  • 31c2b79f8b rend: palette colours read at their centres, punch-through at the reference, precision LibretroAdmin 2026-10-10 19:30:37 +02:00
  • b7bbd421e1 serialize: what a state says about the machine is checked, not taken LibretroAdmin 2026-10-10 19:30:37 +02:00
  • 42aa69f75f arm7: the pc has no low two bits LibretroAdmin 2026-10-10 19:30:37 +02:00
  • 434738c965 aica: a DMA that starts itself; the high bytes of the interrupt and timer registers LibretroAdmin 2026-10-10 19:30:37 +02:00
  • 87f7957c44 naomi: the board serial's index, shift counts; no LAN thread with nobody to talk to LibretroAdmin 2026-10-10 19:30:37 +02:00
  • 75ee3253ba naomi: cartridge reads past the end, a missing disc image, values from a state LibretroAdmin 2026-10-10 19:30:37 +02:00
  • b465fb1a2e libretro: the Dreamcast mouse is read once per poll LibretroAdmin 2026-10-10 19:30:37 +02:00
  • ca211ee386 maple: frames are as long as their headers say; JVS and firmware bounds LibretroAdmin 2026-10-10 19:30:36 +02:00
  • 94f11231bb network: no SIGPIPE; a close passed on after the data before it; old listening sockets LibretroAdmin 2026-10-10 19:30:36 +02:00
  • 7f0b597fdc modem: the DSP memory's last address; a soft reset LibretroAdmin 2026-10-10 19:30:36 +02:00
  • 1a626c8496 picotcp: PPP packets are as long as they say; the receive buffer; a frame never freed LibretroAdmin 2026-10-10 19:30:36 +02:00
  • 7917281c05 rend: render-to-texture format 7 is a warning, not the end LibretroAdmin 2026-10-10 19:30:36 +02:00
  • 70c3d76f46 vulkan: a framebuffer drawn to for the first time starts as the border colour LibretroAdmin 2026-10-10 19:30:36 +02:00
  • 234e0a7065 pvr: Sort-DMA and the YUV converter stay inside memory LibretroAdmin 2026-10-10 19:30:35 +02:00
  • 6493ba7c6b pvr: a list that overruns hands out its elements from the restarted list LibretroAdmin 2026-10-10 19:30:35 +02:00
  • e0ae6f16a0 reios: the HLE BIOS's disc and flash calls keep to the memory they are given LibretroAdmin 2026-10-10 19:30:35 +02:00
  • 51604bcca9 imgread: no disc, a file shorter than its track, a track with no length LibretroAdmin 2026-10-10 19:30:35 +02:00
  • 216d6d8093 gdrom: unknown play and seek parameters, a DMA length not a multiple of 32, PIO of one byte LibretroAdmin 2026-10-10 19:30:35 +02:00
  • 3473c54849 sh4_mem: a block copy stops at the end of the memory it is in LibretroAdmin 2026-10-10 19:30:34 +02:00
  • 1c045c8858 sh4: a DMA count of 0 is 16 M; timer prescaler modes 5 to 7; BL set at an exception LibretroAdmin 2026-10-10 19:30:34 +02:00
  • 4ddd046a30 sh4: RTE's delay slot runs with the new SR; pending interrupts are looked at once LibretroAdmin 2026-10-10 19:30:34 +02:00
  • 04d93f31cd sh4_sched: an event registered again keeps its place; resets drop their own events LibretroAdmin 2026-10-10 19:30:34 +02:00
  • 990da23822 tools/threads: MAC with S, the MMU turned on once, the TLB's counter LibretroAdmin 2026-10-10 17:04:20 +02:00
  • d5340b6f7e picoppp: what a socket does not take is kept; a UDP socket on the game's port LibretroAdmin 2026-10-10 17:04:20 +02:00
  • a8752f22c0 modem: writes that stopped the emulator; the line drops at reset LibretroAdmin 2026-10-10 17:04:19 +02:00
  • c389b5bc4f picotcp: PPP's timer, a late Configure-Ack, the VJ option; segments and timers LibretroAdmin 2026-10-10 17:04:19 +02:00
  • b394653bfa naomi: the comm board and the Atomiswave's devices on a reset LibretroAdmin 2026-10-10 17:04:19 +02:00
  • d4611bd4b4 naomi: regions for 21 sets, and the sizes of two LibretroAdmin 2026-10-10 17:04:19 +02:00
  • 9e81799b19 naomi: cartridge reads stay inside the cartridge; a DMA goes to its end LibretroAdmin 2026-10-10 17:04:18 +02:00
  • e60e95ae6a libretro: an arcade machine's ports once a frame; trackball motion; a mixed path LibretroAdmin 2026-10-10 17:04:18 +02:00
  • ba5d45f55a maple: JVS replies, the EEPROM kept in memory, players after a state load LibretroAdmin 2026-10-10 17:04:18 +02:00
  • f95b3ab721 cheats: three widescreen entries LibretroAdmin 2026-10-10 17:04:17 +02:00
  • c2603e17d9 reios: the drive's music commands, the DMA interrupt, SYS_MISC LibretroAdmin 2026-10-10 17:04:17 +02:00
  • f8a9f86b6b imgread: 99 tracks at most, sector sizes the reader knows, broken CDI files LibretroAdmin 2026-10-10 17:04:17 +02:00
  • 6beadac619 gdrom: a save state's drive buffers are checked LibretroAdmin 2026-10-10 17:04:17 +02:00
  • f6d8923331 rend: a custom texture is not palette indices; the cache cleared while one loads LibretroAdmin 2026-10-10 17:04:17 +02:00
  • 4388d158b0 pvr: list ends, short register writes, strips joined, YUV per macroblock LibretroAdmin 2026-10-10 17:04:16 +02:00
  • 257efa584e pvr: a reset leaves nothing of the last run; the light gun's line LibretroAdmin 2026-10-10 17:04:16 +02:00
  • 5968fe2a70 aica: a channel that reaches past the end of sound memory LibretroAdmin 2026-10-10 17:04:16 +02:00
  • 9a0ef6f6ec aica: ends of arrays, VOFF by the byte, and code buffers on Windows LibretroAdmin 2026-10-10 17:04:15 +02:00
  • 219ade341c savestate: room for the state of the next frame LibretroAdmin 2026-10-10 17:04:15 +02:00
  • 365642ffbf vmem: a descriptor, a flag and an address left over LibretroAdmin 2026-10-10 17:04:15 +02:00
  • dc600e4d83 dynarec: a linked block is entered past its address test (x86-64) LibretroAdmin 2026-10-10 17:04:14 +02:00
  • 2d938f6f5f dynarec: XTRCT and FSCA with a constant on ARM; what a write handler is given LibretroAdmin 2026-10-10 17:04:14 +02:00
  • 590879e504 dynarec: an odd address is an address error, not code LibretroAdmin 2026-10-10 17:04:14 +02:00
  • 28166f5bf1 dynarec: a block discarded on its own, and then the cache emptied LibretroAdmin 2026-10-10 17:04:14 +02:00
  • 58001b762a dynarec: the block that turns the MMU on is run once (x86-64) LibretroAdmin 2026-10-10 17:04:14 +02:00
  • fdc8b2d89b sh4/mmu: the TLB's replacement counter moves on; store queues on reset LibretroAdmin 2026-10-10 17:04:13 +02:00
  • cfdc1b3499 sh4: channel 2 DMA goes by the address the DMAC puts on the bus LibretroAdmin 2026-10-10 17:04:13 +02:00
  • baeb291f64 sh4: the game's rounding mode stays off the frontend's thread LibretroAdmin 2026-10-10 17:04:13 +02:00
  • 70a2fcb946 sh4: the scheduler's clock across a reset LibretroAdmin 2026-10-10 17:04:13 +02:00
  • 5f3740e1ea sh4: on-chip registers keep only the bits they have LibretroAdmin 2026-10-10 17:04:12 +02:00
  • d81ed78044 sh4: RTE's delay slot, a jump to an odd address, Stop() in the interpreter LibretroAdmin 2026-10-10 17:04:12 +02:00
  • c797705b08 sh4: MAC.W and MAC.L with the S flag set LibretroAdmin 2026-10-10 17:04:12 +02:00
  • 12ba338a53 tools/threads: S flag, shifts by 32, code at two addresses, late TLB LibretroAdmin 2026-10-10 14:35:29 +02:00
  • b79b16d1a8 picotcp: an option of no length ends the options LibretroAdmin 2026-10-10 14:35:28 +02:00
  • 2bf8bb470a naomi: the cartridge's DMA offset moves on, a game's settings are its own LibretroAdmin 2026-10-10 14:35:28 +02:00
  • 99566a0625 maple: all arcade players without threaded rendering; mouse; coins LibretroAdmin 2026-10-10 14:35:28 +02:00
  • 0901816b0e imgread: the TOC as a drive gives it; bounds, and broken image files LibretroAdmin 2026-10-10 14:35:28 +02:00
  • e3ce1f292e gdrom: unknown commands end; replies no longer than there is of them LibretroAdmin 2026-10-10 14:35:28 +02:00
  • 32d526c62f aica: the LFO across a state load, the DSP's output registers, two crashes LibretroAdmin 2026-10-10 14:35:28 +02:00
  • 5448b336d0 savestate: a state that does not fit is refused, and an old state's list LibretroAdmin 2026-10-10 14:35:28 +02:00
  • dd3d84fa75 pvr: a reset value, and four ways a game could stop the emulator LibretroAdmin 2026-10-10 14:35:27 +02:00
  • 688de3bf33 vmem: the memory file closed, no gap while it is mapped, and a build fix LibretroAdmin 2026-10-10 14:35:27 +02:00
  • fe65999c78 dynarec: with the MMU on, every block asks whether it is at its address LibretroAdmin 2026-10-10 14:35:27 +02:00
  • 9140befd81 dynarec: room for the largest block, and a block that is already gone LibretroAdmin 2026-10-10 14:35:27 +02:00
  • b462778b68 dynarec: two wrong results from the optimiser LibretroAdmin 2026-10-10 14:35:27 +02:00
  • 73cd7d063b sh4: a timer reloaded with next to nothing; channel 2 DMA from anywhere LibretroAdmin 2026-10-10 14:35:27 +02:00
  • e31ba85b7d sh4/mmu: a TLB of the program's own, loaded after translation is turned on LibretroAdmin 2026-10-10 14:35:27 +02:00
  • a9b045e415 dynarec: a block that is running when the cache is emptied twice LibretroAdmin 2026-10-10 14:35:26 +02:00
  • 603ff108e9 sh4/mmu: MMUCR's fields where they are; a TLB multiple hit is an exception LibretroAdmin 2026-10-10 14:35:26 +02:00
  • 8e12f7a1dd sh4: FPSCR's cause field and reset value; FIPR, FNEG, FABS with PR set LibretroAdmin 2026-10-10 14:35:26 +02:00
  • 151eaf1826 sh4: CLRS and SETS change the S flag in recompiled code LibretroAdmin 2026-10-10 14:35:26 +02:00
  • 9200931e30 dynarec/arm64: constant addresses from x26 LibretroAdmin 2026-10-10 09:00:47 +02:00
  • 7f44247072 dynarec/arm64: the cycle counter stays in its register with the MMU on LibretroAdmin 2026-10-09 21:08:25 +02:00
  • 2ac67a2341 dynarec/arm64: a load or store in one instruction, on the block's registers LibretroAdmin 2026-10-09 21:08:25 +02:00
  • b483a8deff dynarec/x64: the cycle counter in r14 LibretroAdmin 2026-10-09 21:08:25 +02:00
  • 73e53685b8 dynarec/x64: constant addresses from r15, the rest of the context in one go LibretroAdmin 2026-10-09 21:08:25 +02:00
  • a90834b0b0 dynarec/x64: a load or store is one instruction on Windows too LibretroAdmin 2026-10-09 21:08:24 +02:00
  • 1b37a4b2dd vmem/windows: the SH4's 4 GB laid out on 64-bit, and all of it given back LibretroAdmin 2026-10-09 21:08:24 +02:00
  • 68dddce1ee tools/threads: NEGC twice on one register LibretroAdmin 2026-10-09 19:11:52 +02:00
  • 9fcb1a7570 dynarec/x64: a load or store with the registers the block has LibretroAdmin 2026-10-09 19:11:52 +02:00
  • f7c81a063e dynarec/x64: a value to and from the context in one instruction LibretroAdmin 2026-10-09 19:11:19 +02:00
  • ed35b46289 dynarec/x64: NEGC leaves 32 bits in its register LibretroAdmin 2026-10-09 19:11:19 +02:00
  • 71aed156a2 dynarec/arm64: no nop after a load or store LibretroAdmin 2026-10-09 18:22:30 +02:00
  • d959f22ba1 dynarec/x64: a load or store is one instruction LibretroAdmin 2026-10-09 18:22:30 +02:00
  • f23af78e11 rom_luts: no column for switching vmem32 off LibretroAdmin 2026-10-09 17:55:21 +02:00
  • bbc9d0ae3c vmem32 is gone LibretroAdmin 2026-10-09 17:55:21 +02:00
  • 3fcb4547e1 tools/threads: 64 bits in two pages, and every core through the table LibretroAdmin 2026-10-09 16:32:46 +02:00