New core starting to come together!

News from administrators.
User avatar
chriztr
Posts: 38
Joined: Sun Oct 06, 2013 4:15 pm

Re: New core starting to come together!

Post by chriztr » Fri Jan 02, 2015 4:41 pm

Nintendo Maniac 64 wrote: Is that with the latest 2015-01-02 04:14 build? If so, how did you get Link to have a sword? At least with the SSE41 build I start with no sword and tons of ocarinas.
Kiosk / preview rom dosen't have the sudden death / fuck load of ocarinas problem.
Image Image

User avatar
Nintendo Maniac 64
Posts: 185
Joined: Fri Oct 04, 2013 11:37 pm

Re: New core starting to come together!

Post by Nintendo Maniac 64 » Fri Jan 02, 2015 6:18 pm

chriztr wrote:Kiosk / preview rom dosen't have the sudden death / fuck load of ocarinas problem.
Ahhh... well technically the issue doesn't seem to actually get in the way of continuing the game since the gaurd(s) still allow you to leave Clock Town and you can still attack via Deku Scrub.

----------------------------------------------------------------

Anyway, here's a weird glitch that I've come accross a couple times now - somehow in Smash Bros. controller 2 was able to pause - it ended up contantly pausing and unpausing and even if I managed to pause between P2's pauses and quit the match to start a new one, the issue would still occur. Interestly, it occured the moment after I (player 1) paused, but I never hit the enter key - it said player 1 paused after I hit the C key (B button).

I also noticed a second issue; it would seem that the image being output isn't truely 320x240 with 200% nearest neighbor. I've added bright pink lines in overscan area that shows the rows of pixels that are 3 pixels high rather than the correct 2 - they repeat in a pattern of occuring after 78 pixels, 76 pixels, 78, 76, 78.

Here's a screenshot showing both the controller 2 pause issue and the not 200% nearest neighbor pixel issue:

Image
CEN64 Forum's resident straight-male kuutsundere
(just "tsundere" makes people think of "Shana clones" *shivers*)

CPU+iGPU: Pentium G3258 @ 4.6GHz/1.281v
dGPU: Radeon HD5870 1GB
RAM: Vengeance 1600 4x4GB
OS: Windows 7

User avatar
darkszero
Posts: 4
Joined: Fri Oct 25, 2013 12:52 pm

Re: New core starting to come together!

Post by darkszero » Fri Jan 02, 2015 8:44 pm

Removing the ifdefs at the .s files "fixed" the build here. Performance was roughly 25% or less than the provided binary. Enabling the VR4300_BUSY_WAIT_DETECTION flag improved it to 50%.
I've got a hunch that I'm not building a 'Release' build. When I set CMAKE_BUILD_TYPE to Release, I get an error telling that LTO is unavailable. I guess I should get another MinGW distro or build it myself.

Maybe I'll try using the clang binary I've built myself for other reasons, but not now.

User avatar
MarathonMan
Site Admin
Posts: 692
Joined: Fri Oct 04, 2013 4:49 pm

Re: New core starting to come together!

Post by MarathonMan » Fri Jan 02, 2015 8:49 pm

darkszero wrote:I've got a hunch that I'm not building a 'Release' build. When I set CMAKE_BUILD_TYPE to Release, I get an error telling that LTO is unavailable. I guess I should get another MinGW distro or build it myself.
Yeah... if you don't have LTO, it's usually game over as far as performance goes. :/

EDIT: Lighting issues mostly gone now. Just need VMACU and LTV/STV. Fortunately, VMACU should be easy now that I just did VMACF.

User avatar
Snowstorm64
Posts: 303
Joined: Sun Oct 20, 2013 8:22 pm

Re: New core starting to come together!

Post by Snowstorm64 » Fri Jan 02, 2015 9:22 pm

I don't know what happened here, but with latest commits the graphics are less stable in SM64. It is worse with the debug build, like this:
sm64bug.png
sm64bug.png (98.26 KiB) Viewed 33602 times
EDIT: Looks like this commit has caused this regression.
OS: Debian GNU/Linux Jessie (8.0)
CPU: Intel i7 4770K @ 3.5 GHz
Build: AVX (compiled from git)

Presence
Posts: 51
Joined: Fri Oct 18, 2013 9:27 am

Re: New core starting to come together!

Post by Presence » Fri Jan 02, 2015 9:45 pm

MarathonMan wrote:I broke SSE2 builds in getting the new core slapped together (sorry). That function only gets compiled when SSE2 is used as a build type, so double check your CMake configuration.
Thanks. Wasn't aware I had to manually specify the architecture within CMakeLists.txt.

For anyone interested, I've commented out the controller/save parameters from CEN64-Qt so it can launch the new core. I've built binaries for Linux and Windows and uploaded them to the release page on Github (here).

User avatar
MarathonMan
Site Admin
Posts: 692
Joined: Fri Oct 04, 2013 4:49 pm

Re: New core starting to come together!

Post by MarathonMan » Fri Jan 02, 2015 9:49 pm

Drat. Yeah, so the whole method of how I've been using register caching is incredibly hacky right now. Perhaps it's best to just disable it for the time being until I can come up with a more stable solution.

On the bright side, VMACU is finished. :D

User avatar
Narann
Posts: 154
Joined: Mon Jun 16, 2014 4:25 pm
Contact:

Re: New core starting to come together!

Post by Narann » Fri Jan 02, 2015 10:52 pm

MarathonMan wrote:Yeah, so the whole method of how I've been using register caching is incredibly hacky right now. Perhaps it's best to just disable it for the time being until I can come up with a more stable solution.
You know the rule :D
MarathonMan wrote:On the bright side, VMACU is finished. :D
Cool! So only LTV/STV is missing to fix lighting troubles?

User avatar
iwasaperson
Posts: 49
Joined: Tue Apr 22, 2014 12:50 am

Re: New core starting to come together!

Post by iwasaperson » Fri Jan 02, 2015 11:21 pm

OK. WTH?

Majora's Mask works fine on the Windows build (with WINE), but not on the Linux build (my own build, but I've also tried the SSE4.1 build).

https://i.imgur.com/MCvXEGK.png (WINE on left, native on right)

User avatar
Ambient_Malice
Posts: 21
Joined: Mon Aug 25, 2014 9:04 am

Re: New core starting to come together!

Post by Ambient_Malice » Sat Jan 03, 2015 2:01 am

I used a portable linux distro to test the linux build. Worked like a charm. (Which should make creating a "portable N64" fairly simple in the future.)

Performance? Robot on Wheels into averaged about 25VIs on Windows and 30VIs on Linux. So that's a 5fps/20% boost on Linux. (SSE4.1 build)

I've been testing homebrew.

ScummVM - no boot.
Retroactive's 2011 MGC demo - no boot.
Wolfenstein3D N64 homebrew port - seems to run fine.

User avatar
MarathonMan
Site Admin
Posts: 692
Joined: Fri Oct 04, 2013 4:49 pm

Re: New core starting to come together!

Post by MarathonMan » Sat Jan 03, 2015 2:13 am

Narann wrote:You know the rule :D
But look how pretty it is!
https://www.youtube.com/watch?v=6QPN3rOhuu0
MarathonMan wrote:On the bright side, VMACU is finished. :D
Cool! So only LTV/STV is missing to fix lighting troubles?[/quote]

Yep! Lighting looks pretty good for the most part now, too (even w/o it).
iwasaperson wrote:Majora's Mask works fine on the Windows build (with WINE), but not on the Linux build (my own build, but I've also tried the SSE4.1 build).
Try again -- I disabled the optimization that should be responsible for that. The Windows x64 calling convention allocates fewer volatile registers so this is less likely to happen with that optimization enabled.

User avatar
iwasaperson
Posts: 49
Joined: Tue Apr 22, 2014 12:50 am

Re: New core starting to come together!

Post by iwasaperson » Sat Jan 03, 2015 2:27 am

MarathonMan wrote:Try again -- I disabled the optimization that should be responsible for that. The Windows x64 calling convention allocates fewer volatile registers so this is less likely to happen with that optimization enabled.
Nope. Still happening.

User avatar
MarathonMan
Site Admin
Posts: 692
Joined: Fri Oct 04, 2013 4:49 pm

Re: New core starting to come together!

Post by MarathonMan » Sat Jan 03, 2015 2:29 am

iwasaperson wrote:
MarathonMan wrote:Try again -- I disabled the optimization that should be responsible for that. The Windows x64 calling convention allocates fewer volatile registers so this is less likely to happen with that optimization enabled.
Nope. Still happening.
Ok that's really weird then. I'll look into that tomorrow or so. That's just the vanilla kart or the kiosk?

User avatar
iwasaperson
Posts: 49
Joined: Tue Apr 22, 2014 12:50 am

Re: New core starting to come together!

Post by iwasaperson » Sat Jan 03, 2015 2:41 am

MarathonMan wrote:
iwasaperson wrote:
MarathonMan wrote:Try again -- I disabled the optimization that should be responsible for that. The Windows x64 calling convention allocates fewer volatile registers so this is less likely to happen with that optimization enabled.
Nope. Still happening.
Ok that's really weird then. I'll look into that tomorrow or so. That's just the vanilla kart or the kiosk?
Vanilla (US).

User avatar
The Extremist
Posts: 29
Joined: Sun Nov 03, 2013 6:11 pm
Location: Canadian Prairie

Re: New core starting to come together!

Post by The Extremist » Sat Jan 03, 2015 5:06 am

MarathonMan wrote:
But look how pretty it is!
https://www.youtube.com/watch?v=6QPN3rOhuu0
Swell! :)

User avatar
chriztr
Posts: 38
Joined: Sun Oct 06, 2013 4:15 pm

Re: New core starting to come together!

Post by chriztr » Sat Jan 03, 2015 6:32 am

Why is DK64 all flickery?

Image

Video:
http://webmshare.com/34PX3

On the other hand, OOT MM is working flawless:
zelda_oot_mq_01.png
zelda_oot_mq_01.png (98.18 KiB) Viewed 33697 times

User avatar
Snowstorm64
Posts: 303
Joined: Sun Oct 20, 2013 8:22 pm

Re: New core starting to come together!

Post by Snowstorm64 » Sat Jan 03, 2015 8:20 am

After LTV/STV, I guess it will be the time to fix things? I mean, look at some of these games that crash or cause some weird issue after a while, no matter if with new or old RSP, even with old core:
  • Mario Party 2: If you do the tutorial, in the tunnel at a certain point some weird issues happen like the camera out of place.
  • Iggy's Reckin' Balls: At the end of a level (not the tutorial) the game will always crash.
  • Earthworm Jim 3D: bad graphical corruption at the boot, will crash after a while.
  • Paper Mario: It will always crash at main intro. Maybe it is related to the audio.
  • Top Gear Rally: At some point the car will fall out of the track.
OS: Debian GNU/Linux Jessie (8.0)
CPU: Intel i7 4770K @ 3.5 GHz
Build: AVX (compiled from git)

User avatar
MarathonMan
Site Admin
Posts: 692
Joined: Fri Oct 04, 2013 4:49 pm

Re: New core starting to come together!

Post by MarathonMan » Sat Jan 03, 2015 11:09 am

chriztr wrote:Why is DK64 all flickery?
This game has always been difficult. Speaking of which, how did you get it to boot? It won't boot for me?
Snowstorm64 wrote:After LTV/STV, I guess it will be the time to fix things? I mean, look at some of these games that crash or cause some weird issue after a while, no matter if with new or old RSP, even with old core:
  • Mario Party 2: If you do the tutorial, in the tunnel at a certain point some weird issues happen like the camera out of place.
  • Iggy's Reckin' Balls: At the end of a level (not the tutorial) the game will always crash.
  • Earthworm Jim 3D: bad graphical corruption at the boot, will crash after a while.
  • Paper Mario: It will always crash at main intro. Maybe it is related to the audio.
  • Top Gear Rally: At some point the car will fall out of the track.
Ah excellent! I knew about the ROMs that wouldn't boot, but not these. They all suggest that I'm misunderstanding or overlooking some detail somewhere in the pipeline that needs serious fixing! I might have to implement a debugger to actually solve these... otherwise I'm going to lose my mind. :p

User avatar
Snowstorm64
Posts: 303
Joined: Sun Oct 20, 2013 8:22 pm

Re: New core starting to come together!

Post by Snowstorm64 » Sat Jan 03, 2015 11:32 am

MarathonMan wrote:
This game has always been difficult. Speaking of which, how did you get it to boot? It won't boot for me?
I guess there's some issues with Linux build, because neither I can boot DK64, and Majora's Mask runs very bad.
MarathonMan wrote: Ah excellent! I knew about the ROMs that wouldn't boot, but not these. They all suggest that I'm misunderstanding or overlooking some detail somewhere in the pipeline that needs serious fixing! I might have to implement a debugger to actually solve these... otherwise I'm going to lose my mind. :p
Yeah, a debugger is a good idea.
Also, I forgot that Worms Armageddon will crash if you try to start a match.
OS: Debian GNU/Linux Jessie (8.0)
CPU: Intel i7 4770K @ 3.5 GHz
Build: AVX (compiled from git)

User avatar
Ambient_Malice
Posts: 21
Joined: Mon Aug 25, 2014 9:04 am

Re: New core starting to come together!

Post by Ambient_Malice » Sat Jan 03, 2015 11:33 am

It just dawned on me that Wonder Project J2's menu failing to initialise correctly is very much likely due to missing save support.

Duke Nukem: Zero Hour freezes after starting a new game. The World is Not Enough does the same thing. They're both Eurocom games. I would test 40 Winks, but 40 Winks is PAL. I wonder whether they're affected by missing save support, or whether some other bug/unhandled behavior is responsible.

User avatar
MarathonMan
Site Admin
Posts: 692
Joined: Fri Oct 04, 2013 4:49 pm

Re: New core starting to come together!

Post by MarathonMan » Sat Jan 03, 2015 2:10 pm

iwasaperson wrote:OK. WTH?

Majora's Mask works fine on the Windows build (with WINE), but not on the Linux build (my own build, but I've also tried the SSE4.1 build).
Fixed.

angrylion added a check to clamp the RDRAM range to 4MiB/8MiB using a Windows interface that checks the validity of the virtual address space surrounding the RDRAM. If you're on Linux, well, then it always clamped to 4MiB.

So it was breaking all games that required the expansion pack on Linux. :D

Nice find.

User avatar
Nacho
Posts: 66
Joined: Thu Nov 07, 2013 9:25 am

Re: New core starting to come together!

Post by Nacho » Sat Jan 03, 2015 2:26 pm

Confirmed that. DK64 now boots on Linux. Flickering issues included.

Regarding Expansion Pack, it may be a nice feature a 4/8MB switch. Some games would behave differently without expansion pack (say Perfect Dark)
Testing CEN64 on: Intel Core i5 520M 2.4 GHz. SSE2 SSE3 SSE4.1 SSE4.2 SSSE3, but no AVX. Ubuntu Linux

User avatar
iwasaperson
Posts: 49
Joined: Tue Apr 22, 2014 12:50 am

Re: New core starting to come together!

Post by iwasaperson » Sat Jan 03, 2015 2:31 pm

MarathonMan wrote:
iwasaperson wrote:OK. WTH?

Majora's Mask works fine on the Windows build (with WINE), but not on the Linux build (my own build, but I've also tried the SSE4.1 build).
Fixed.

angrylion added a check to clamp the RDRAM range to 4MiB/8MiB using a Windows interface that checks the validity of the virtual address space surrounding the RDRAM. If you're on Linux, well, then it always clamped to 4MiB.

So it was breaking all games that required the expansion pack on Linux. :D

Nice find.
Can confirm working :D .
Seems like angrylion uses a lot of Windows-only code.

User avatar
MarathonMan
Site Admin
Posts: 692
Joined: Fri Oct 04, 2013 4:49 pm

Re: New core starting to come together!

Post by MarathonMan » Sat Jan 03, 2015 4:16 pm

Nacho wrote:Regarding Expansion Pack, it may be a nice feature a 4/8MB switch. Some games would behave differently without expansion pack (say Perfect Dark)
Certainly, in due time. :D

User avatar
chriztr
Posts: 38
Joined: Sun Oct 06, 2013 4:15 pm

Re: New core starting to come together!

Post by chriztr » Sat Jan 03, 2015 6:44 pm

Some of the coloring isn't quite right yet:

Image Image Image

But on the other hand, here's some images from an i5 4670k @ 4.2GHz with windows

Banjo Tooie (showing white-screen)
Image

Diddy kong racing:
Image Image

Mario Party:
Image Image Image

Paper Mario (where it's crashing / hanging)
Image

Zelda MM:
Image Image Image

User avatar
Snowstorm64
Posts: 303
Joined: Sun Oct 20, 2013 8:22 pm

Re: New core starting to come together!

Post by Snowstorm64 » Sat Jan 03, 2015 7:40 pm

CTC2 is required by Mortal Kombat 4 to work properly:
Attachments
mortalkombat4.png
mortalkombat4.png (48.76 KiB) Viewed 34208 times
OS: Debian GNU/Linux Jessie (8.0)
CPU: Intel i7 4770K @ 3.5 GHz
Build: AVX (compiled from git)

User avatar
Ambient_Malice
Posts: 21
Joined: Mon Aug 25, 2014 9:04 am

Re: New core starting to come together!

Post by Ambient_Malice » Sat Jan 03, 2015 7:42 pm

Nacho wrote:Confirmed that. DK64 now boots on Linux. Flickering issues included.

Regarding Expansion Pack, it may be a nice feature a 4/8MB switch. Some games would behave differently without expansion pack (say Perfect Dark)
Hybrid Heaven, Resident Evil 2, and Indiana Jones and the Infernal Machine raise resolution automatically with the Expansion Pack enabled.

Space Station Silicon Valley (U) won't run with the EP inserted.

User avatar
MarathonMan
Site Admin
Posts: 692
Joined: Fri Oct 04, 2013 4:49 pm

Re: New core starting to come together!

Post by MarathonMan » Sun Jan 04, 2015 1:29 pm

Fixed another couple bugs in the VR4300/RSP, half-frame rendering issues, and implemented WatchLo/Hi support.

The latter should help with GS hacks/cheats:
http://forums.cen64.com/jungerman.png

User avatar
Snowstorm64
Posts: 303
Joined: Sun Oct 20, 2013 8:22 pm

Re: New core starting to come together!

Post by Snowstorm64 » Sun Jan 04, 2015 3:07 pm

Nice addition!

I may have found a VR4300 bug that doesn't happen with old core:
PaperBoy.png
PaperBoy.png (19.01 KiB) Viewed 33983 times
(Paperboy, at the boot)
With old core, Paperboy's intro is running fine.

Also, checking the debug output of F-Zero X, I found some interesting things:
bus_read_word: Failed to access: 0x05000508
bus_write_word: Failed to access: 0x00800000
bus_write_word: Failed to access: 0x008FFFFC
bus_read_word: Failed to access: 0x00800000
bus_read_word: Failed to access: 0x05000508
bus_read_word: Failed to access: 0x05000508
bus_write_word: Failed to access: 0x05000520
bus_read_word: Failed to access: 0x05000508
bus_read_word: Failed to access: 0x05000540
bus_read_word: Failed to access: 0x05000508
bus_write_word: Failed to access: 0x05000500
bus_write_word: Failed to access: 0x05000508
The red lines are attempts to check 64DD status but they fail, the orange lines are 64DD-related. Maybe this is what prevents F-Zero X from booting on CEN64. A stub that permanently sets the 64DD status on OFF (or something like that) may help, who knows. Blue lines are attempts to read the Controller Pak.

Speaking of which, another stub may be useful to turn off those annoying messages that ask if we want to repair the Controller Pak, or warn us that the Controller Pak is damaged, this may 'unlock' some other games that don't let us continue because of it.

MarathonMan, do you think it's worth to give it a try? Those stubs may be useful to protect the game by unpredictable behaviors.

EDIT: Looks like F-Zero X USA will fail to boot on a real N64 if there is a 64DD attached.
OS: Debian GNU/Linux Jessie (8.0)
CPU: Intel i7 4770K @ 3.5 GHz
Build: AVX (compiled from git)

User avatar
MarathonMan
Site Admin
Posts: 692
Joined: Fri Oct 04, 2013 4:49 pm

Re: New core starting to come together!

Post by MarathonMan » Sun Jan 04, 2015 9:02 pm

Snowstorm64 wrote:The red lines are attempts to check 64DD status but they fail, the orange lines are 64DD-related. Maybe this is what prevents F-Zero X from booting on CEN64. A stub that permanently sets the 64DD status on OFF (or something like that) may help, who knows. Blue lines are attempts to read the Controller Pak.

Speaking of which, another stub may be useful to turn off those annoying messages that ask if we want to repair the Controller Pak, or warn us that the Controller Pak is damaged, this may 'unlock' some other games that don't let us continue because of it.

MarathonMan, do you think it's worth to give it a try? Those stubs may be useful to protect the game by unpredictable behaviors.

EDIT: Looks like F-Zero X USA will fail to boot on a real N64 if there is a 64DD attached.
I don't really want to spend time tracking down issues with a single game at this point in time. Especially when there are so many crucial pieces of the system that are unimplemented.

Same goes for the stub/controller pack hack, to an extent: time would be better spent actually implementing the features instead of implementing a stub that's going to be removed anyways.

Currently focusing on implementing the remaining portions of the VR4300 and hunting bugs down there... after that, I'll probably start (finally) working on a debugger/finishing up the rest of the RSP.

User avatar
Snowstorm64
Posts: 303
Joined: Sun Oct 20, 2013 8:22 pm

Re: New core starting to come together!

Post by Snowstorm64 » Sun Jan 04, 2015 9:29 pm

I thought it would be easy to set those bytes on zero (and I'd still want those annoying messages to go away :P), but I guess I'm wrong. Probably you're right, with a little more patience it's better to implement completely those features rather than just add some stubs, after all.
OS: Debian GNU/Linux Jessie (8.0)
CPU: Intel i7 4770K @ 3.5 GHz
Build: AVX (compiled from git)

User avatar
Nintendo Maniac 64
Posts: 185
Joined: Fri Oct 04, 2013 11:37 pm

Re: New core starting to come together!

Post by Nintendo Maniac 64 » Mon Jan 05, 2015 12:05 am

Mario Tennis boots now but crashes before showing the intro, and unlike Mario Golf you can't mash START / A button to make it to the title screen.

The intro for Star Wars: Episode 1 - Racer no longer has the bottom half of the screen cut off, but controller input is still broken.

And interestingly enough, San Francisco Rush 2049 now crashes before reaching the title screen... (it was able to reach the title screen before)
CEN64 Forum's resident straight-male kuutsundere
(just "tsundere" makes people think of "Shana clones" *shivers*)

CPU+iGPU: Pentium G3258 @ 4.6GHz/1.281v
dGPU: Radeon HD5870 1GB
RAM: Vengeance 1600 4x4GB
OS: Windows 7

User avatar
Ambient_Malice
Posts: 21
Joined: Mon Aug 25, 2014 9:04 am

Re: New core starting to come together!

Post by Ambient_Malice » Mon Jan 05, 2015 2:52 am

ImageImage
Last Legion UX runs, but has strange flickering colours on surfaces.

beannaich
Posts: 149
Joined: Mon Oct 21, 2013 2:43 pm

Re: New core starting to come together!

Post by beannaich » Tue Jan 06, 2015 1:19 am

Ambient_Malice wrote:Another thing - I keep the Cen64 dowload link handy, but will that be added to the front page anytime soon?
Done.

User avatar
Ambient_Malice
Posts: 21
Joined: Mon Aug 25, 2014 9:04 am

Re: New core starting to come together!

Post by Ambient_Malice » Tue Jan 06, 2015 1:22 am

beannaich wrote:
Ambient_Malice wrote:Another thing - I keep the Cen64 dowload link handy, but will that be added to the front page anytime soon?
Done.
Bless your heart.

IMO, Cen64-QT should be the official front-end and be linked on the front page, but that's a whole other kettle of fish. You're all doing a wonderful job.

User avatar
Nintendo Maniac 64
Posts: 185
Joined: Fri Oct 04, 2013 11:37 pm

Re: New core starting to come together!

Post by Nintendo Maniac 64 » Tue Jan 06, 2015 4:53 pm

So MarathonMan, I saw that image you posted before you deleted it. ;) I even had a chance to screenshot your post because, without closing the tab, I ended up reloading the page in a new tab which allowed me to see that you had deleted the post. Instant email notification of replies FTW.

Currently the screenshot of your post is just sitting on my computer, and at least for now I won't repost it in this thread out of respect...


To everyone else, outside of PMing the image, here's a hint - the songs titled "Regeneration" and "Rollercoaster" are currently available only via lossy MP3 encodings from analog recording - said songs were never made available on any kind of soundtrack. If MarathonMan is able to fully get working what the screenshot was implying, then we should finally be able to get nice mostly-lossless audio rips in the future.

(is my hint too obscure? :P)

EDIT:If you figure out my hint, note that the screenshot did not actually show what my hint is, well, hinting at, but something else that would take advantage of what the screenshot was implying that MarathonMan is working on.
Last edited by Nintendo Maniac 64 on Tue Jan 06, 2015 5:10 pm, edited 3 times in total.
CEN64 Forum's resident straight-male kuutsundere
(just "tsundere" makes people think of "Shana clones" *shivers*)

CPU+iGPU: Pentium G3258 @ 4.6GHz/1.281v
dGPU: Radeon HD5870 1GB
RAM: Vengeance 1600 4x4GB
OS: Windows 7

User avatar
iwasaperson
Posts: 49
Joined: Tue Apr 22, 2014 12:50 am

Re: New core starting to come together!

Post by iwasaperson » Tue Jan 06, 2015 5:06 pm

Nintendo Maniac 64 wrote:So MarathonMan, I saw that image you posted before you deleted it. ;) I even had a chance to screenshot your post because, without closing the tab, I ended up reloading the page in a new tab which allowed me to see that you had deleted the post. Instant email notification of replies FTW.

Currently the screenshot of your post is just sitting on my computer, and at least for now I won't repost it in this thread out of respect...


To everyone else, outside of PMing the image, here's a hint - the songs titled "Regeneration" and "Rollercoaster" are currently available only via lossy MP3 encodings from analog recording - said songs were never made available on any kind of soundtrack. If MarathonMan is able to fully get working what the screenshot was implying, then we should finally be able to get nice mostly-lossless audio rips in the future.

(is my hint too obscure? :P)
He's working on perfect N64 audio?!?!?!

User avatar
Nintendo Maniac 64
Posts: 185
Joined: Fri Oct 04, 2013 11:37 pm

Re: New core starting to come together!

Post by Nintendo Maniac 64 » Tue Jan 06, 2015 5:07 pm

iwasaperson wrote:He's working on perfect N64 audio?!?!?!
Perhaps, but no, that's not what the screenshot implied. Maybe my hint is too obsure, but I figured it'd be better to be obscure than to be blantantly-obvious since MarathonMan obviously deleted his post for a reason.

Also note my edit.

EDIT:
Speaking of edits, any guesses after this one I will not say whether it is correct or not - I will leave it up to you guys to speculate. The main thing was that I wanted to point out that the screenshot did not imply at all that he was working on audio.
Last edited by Nintendo Maniac 64 on Tue Jan 06, 2015 5:12 pm, edited 1 time in total.
CEN64 Forum's resident straight-male kuutsundere
(just "tsundere" makes people think of "Shana clones" *shivers*)

CPU+iGPU: Pentium G3258 @ 4.6GHz/1.281v
dGPU: Radeon HD5870 1GB
RAM: Vengeance 1600 4x4GB
OS: Windows 7

User avatar
MarathonMan
Site Admin
Posts: 692
Joined: Fri Oct 04, 2013 4:49 pm

Re: New core starting to come together!

Post by MarathonMan » Tue Jan 06, 2015 5:10 pm

Nintendo Maniac 64 wrote:...
I created a new thread as this one has grown a little wrong; screenshot is still there. :p

Thanks for having the courtesy, though.

User avatar
Nintendo Maniac 64
Posts: 185
Joined: Fri Oct 04, 2013 11:37 pm

Re: New core starting to come together!

Post by Nintendo Maniac 64 » Tue Jan 06, 2015 5:13 pm

MarathonMan wrote:I created a new thread as this one has grown a little wrong; screenshot is still there. :p

Thanks for having the courtesy, though.
:( I was having fun with that guessing game!
CEN64 Forum's resident straight-male kuutsundere
(just "tsundere" makes people think of "Shana clones" *shivers*)

CPU+iGPU: Pentium G3258 @ 4.6GHz/1.281v
dGPU: Radeon HD5870 1GB
RAM: Vengeance 1600 4x4GB
OS: Windows 7

User avatar
iwasaperson
Posts: 49
Joined: Tue Apr 22, 2014 12:50 am

Re: New core starting to come together!

Post by iwasaperson » Tue Jan 06, 2015 10:42 pm

...Well that didn't work: https://i.imgur.com/yOm4hC6.png (mgc_2011)

Guess it will be a while until CEN64 is truly indistinguishable from a real N64.

User avatar
MarathonMan
Site Admin
Posts: 692
Joined: Fri Oct 04, 2013 4:49 pm

Re: New core starting to come together!

Post by MarathonMan » Fri Jan 09, 2015 1:22 pm

krom gave me a ROM that pointed out a bug in the PI controller. TIL 64MiB ROMs didn't work.

The Zelda:OoT debug ROM boots now, which is kinda cool. :)

EDIT: Almost makes me wonder how many other ROMs don't boot because of poorly implemented or non-existant PI/SI/AI...

User avatar
Snowstorm64
Posts: 303
Joined: Sun Oct 20, 2013 8:22 pm

Re: New core starting to come together!

Post by Snowstorm64 » Fri Jan 09, 2015 1:30 pm

Pokemonstadium2.png
Pokemonstadium2.png (243 KiB) Viewed 34143 times
:)
OS: Debian GNU/Linux Jessie (8.0)
CPU: Intel i7 4770K @ 3.5 GHz
Build: AVX (compiled from git)

User avatar
MarathonMan
Site Admin
Posts: 692
Joined: Fri Oct 04, 2013 4:49 pm

Re: New core starting to come together!

Post by MarathonMan » Fri Jan 09, 2015 1:32 pm

Background is probably the JPEG decoder ucode, which needs LTV/STV (same as in OoT when Link is sleeping in his room in the intro scene).

User avatar
Snowstorm64
Posts: 303
Joined: Sun Oct 20, 2013 8:22 pm

Re: New core starting to come together!

Post by Snowstorm64 » Fri Jan 09, 2015 1:45 pm

MarathonMan wrote:Background is probably the JPEG decoder ucode, which needs LTV/STV (same as in OoT when Link is sleeping in his room in the intro scene).
Don't forget CTC2 which is required by Mortal Kombat 4 and Majora's Mask (?).
Also:
OgreBattle64.png
OgreBattle64.png (105.71 KiB) Viewed 34138 times
Other 40/64MB games like Conker's Bad Fur Day and Resident Evil 4 still won't boot. Paper Mario doesn't work anymore with that VR4300 pipeline clean-up some days ago. F-Zero X (which is 16 MB) causes segfault now, that's weird...
OS: Debian GNU/Linux Jessie (8.0)
CPU: Intel i7 4770K @ 3.5 GHz
Build: AVX (compiled from git)

User avatar
MarathonMan
Site Admin
Posts: 692
Joined: Fri Oct 04, 2013 4:49 pm

Re: New core starting to come together!

Post by MarathonMan » Fri Jan 09, 2015 2:01 pm

Snowstorm64 wrote:F-Zero X (which is 16 MB) causes segfault now, that's weird...
Carts that use SRAM save states might be botched now. I'll give the PI a solid hammering when I get the chance.

The Paper Mario thing I'm aware of -- it was using CACHE instructions with mapped virtual addresses. I wasn't doing the address translation for the CACHE instructions (oddly enough, the game still worked). When I implemented the address translation for CACHE instructions, the game broke. Not sure why yet.

User avatar
Nintendo Maniac 64
Posts: 185
Joined: Fri Oct 04, 2013 11:37 pm

Re: New core starting to come together!

Post by Nintendo Maniac 64 » Sat Jan 10, 2015 11:36 pm

So um, it would seem that Majora's Mask no longer boots...
Snowstorm64 wrote:[Screenshot of Pokemon Stadium 2]

:)
How the heck? For me it crashes right at launch just like Majora's Mask does now.
CEN64 Forum's resident straight-male kuutsundere
(just "tsundere" makes people think of "Shana clones" *shivers*)

CPU+iGPU: Pentium G3258 @ 4.6GHz/1.281v
dGPU: Radeon HD5870 1GB
RAM: Vengeance 1600 4x4GB
OS: Windows 7

User avatar
Snowstorm64
Posts: 303
Joined: Sun Oct 20, 2013 8:22 pm

Re: New core starting to come together!

Post by Snowstorm64 » Sun Jan 11, 2015 8:12 am

Majora's Mask still works fine for me.
OS: Debian GNU/Linux Jessie (8.0)
CPU: Intel i7 4770K @ 3.5 GHz
Build: AVX (compiled from git)

User avatar
teres
Posts: 19
Joined: Fri Apr 11, 2014 9:44 am

Re: New core starting to come together!

Post by teres » Sun Jan 11, 2015 9:58 am

FWIW, MM is working for me, but PS2 just seems to stay on the N64 logo screen indefinitely.

User avatar
Nacho
Posts: 66
Joined: Thu Nov 07, 2013 9:25 am

Re: New core starting to come together!

Post by Nacho » Sun Jan 11, 2015 11:37 am

Majora's Mask and Pokemon Stadium 2 are booting for me. PS2 does not hang and is fully playable. CEN64 up-to-date.

md5sum:

1561c75d11cedf356a8ddb1a4a5f9d5d *../Pokemon Stadium 2 (U) [!].z64
2a0a8acb61538235bc1094d297fb6556 *../Legend of Zelda, The - Majora's Mask (U) [!].z64

EDIT: md5sums corrected
Last edited by Nacho on Sun Jan 11, 2015 12:39 pm, edited 1 time in total.
Testing CEN64 on: Intel Core i5 520M 2.4 GHz. SSE2 SSE3 SSE4.1 SSE4.2 SSSE3, but no AVX. Ubuntu Linux

Post Reply

Who is online

Users browsing this forum: No registered users and 1 guest