GroovyArcade live-CD 2020

[zebidia]

Instructions for linux:

Code: [Select]

cd path/to/mame/sources
patch --binary -p0 -E -i path/to/hi_155_2.txt
patch --binary -p0 -E -i path/to/0155_groovymame_015c.diff
make

Great, thanks. I think I'll put this on the back burner and wait for a new ISO.

[Ansa89]

Communication for linux users: here you can find the 15KHz patches for linux 3.18.

[akumajo]

hi

what's new in GroovyArcade-Arch2014.12.25-x86_64.iso ?

[MonkeyJug]

I'd like to know too...  Ves?  Anyone else shed some light on this?

[intealls]

Alright, here's an annoying bug that not only affects the latest GA (Dec 2014), but GM running under Ubuntu 12.04, 14.04 and 14.10 as well.

The gist of it is, start GM (I've only tried 0.156 and 0.157, any game seems to do, wrally, jackal and msword does it for sure), let it run for approximately 2 hours and 25 minutes, and watch X go from about 10% CPU utilization to 100%, which causes the game to run at about 10% speed. The cards that have been tested are Radeon HD 6310 and Radeon HD 3200. Verified on two different setups. I'm not sure it's GM's fault, since it's X that eating all the cycles.

Has anyone else seen this? I'm currently testing it on GA Dec 2014 with a 9600 NVIDIA card, to see if the problem is related to the open source Radeon driver. Edit: And it's not. It does this with the nouveau driver as well.

Edit: Going into the video settings and rotating the view fixes it for (probably) another 2 hours and 25 minutes. Could it be some integer overflow or something? Segfaults after a short while.

Edit: Only 64-bit has been tested.

Edit again: The problem occurs also when non-throttled, tested with GA 2014-04-25 in VMWare, which doesn't seem to support sync to vblank?

Edit again: Also affects GA 2013-02-24 i686

[Calamity]

Hi intealls,

Sorry for my late response and thanks a lot for looking into this.

This is an old known issue: http://forum.arcadecontrols.com/index.php/topic,130667.msg1348167.html#msg1348167

I'm totally out ideas regarding a possible cause. At least thanks to your tests now we know it's not caused by ATI drivers, vsync, etc. It'd be interesting to check if vanilla MAME has this problem too.

BTW no virtual machines I've tested support vsync.

[intealls]

Hi Calamity and thanks.

I'm testing older versions, to see if they are affected. I built 0.149 on Ubuntu 14.04.1, and ran it on GA Dec 14 (installed to hard drive). Jackal ran for 3 hours and 15 minutes, without slowdown. This is strange, since the GA 2013 dist I tested uses 0.148u1, and that did slow down after about 2 and a half hours. I need to test more, to make sure this build is unaffected.

[intealls]

I'm testing older versions, to see if they are affected. I built 0.149 on Ubuntu 14.04.1, and ran it on GA Dec 14 (installed to hard drive). Jackal ran for 3 hours and 15 minutes, without slowdown. This is strange, since the GA 2013 dist I tested uses 0.148u1, and that did slow down after about 2 and a half hours. I need to test more, to make sure this build is unaffected.

Some progress has been made. I installed GA Dec 14, and built GM 0.155 using the GA installation. It does not seem to be affected by the slowdowns. When building and testing GM 0.156 on the same install, it exhibits the slowdown problem.

Edit: When testing on Ubuntu 14.04, no slowdowns are evident with 0.155. However, the screen is completely off-center, but I suspect the blame is to be placed on SDL.

[Ansa89]

Communication for linux users: here you can find the 15KHz patches for linux 3.19.

[ves]

Hello guys, forgive my absence, I am very busy, new job etc.....

News

-Fixed Update system.
-Fixed internal system.
-Fixed Joystick Sdl2.
-Fixed broken sound to pause games.
-New support Sdl2.
-Openbox new windows manager.
-New system MegaCd, 32X from Ume.
-New Attractmode config.
-Last system version.
-New method for config roms/snap, now make cache search, select you system to add the path, no need to search for each system.


Hi Ansa, I'll send you the latest kernel diff, we have added new resolution.


Thank.

[Ansa89]

Hi Ansa, I'll send you the latest kernel diff, we have added new resolution.

Nice!

[Calamity]

Some progress has been made. I installed GA Dec 14, and built GM 0.155 using the GA installation. It does not seem to be affected by the slowdowns. When building and testing GM 0.156 on the same install, it exhibits the slowdown problem.

Edit: When testing on Ubuntu 14.04, no slowdowns are evident with 0.155. However, the screen is completely off-center, but I suspect the blame is to be placed on SDL.

This is really odd. I can't find a reason why 0.155 is affected and 0.156 is not. Have you by any chance tested this with/without multithreading?

[intealls]

Have you by any chance tested this with/without multithreading?

0.156 slows down also with -nomt 

[Calamity]

0.156 slows down also with -nomt 

Just wanted to figure out what factors may play a role in this. This will probably be a stupid suggestion, but just in case, may you check in the source trees you have built from which of them were using XINPUT (check in sdl.mak, there's a define named NO_USE_XINPUT).

[intealls]

0.156 slows down also with -nomt 

Just wanted to figure out what factors may play a role in this. This will probably be a stupid suggestion, but just in case, may you check in the source trees you have built from which of them were using XINPUT (check in sdl.mak, there's a define named NO_USE_XINPUT).

Hi Calamity and thanks, and I'm sorry for the late reply, been busy with work.

When building and testing 0.156 with NO_USE_XINPUT = 1, the slowdowns do NOT occur in GA Dec 14 with VMware. Also, when building and testing 0.158 in Ubuntu 14.04 with NO_USE_XINPUT = 1, no slowdowns either (also in VMware)!

You're the man!

Edit: Clarifications, going to test on hw now.

Edit: After leaving Jackal running for 18 hours on actual hw, no slowdown. So it's probably very safe to say that setting NO_USE_XINPUT fixes this issue.

[Doozer]

To my investigation, only 0.155 and 0.156 should be affect by default. Those two versions have NO_USE_XINPUT disabled by default.

[intealls]

To my investigation, only 0.155 and 0.156 should be affect by default. Those two versions have NO_USE_XINPUT disabled by default.

0.158
-NO_USE_XINPUT = 1
+#NO_USE_XINPUT = 1

0.157
-NO_USE_XINPUT = 1
+#NO_USE_XINPUT = 1

0.156
-NO_USE_XINPUT = 1
+#NO_USE_XINPUT = 1

0.155 leaves it set. 0155_groovymame_015b.diff leaves it set, 0155_groovymame_015c.diff does not. Sorry for the confusion.

as per https://googledrive.com/host/0B5iMjDor3P__aEFpcVNkVW5jbEE/

[Doozer]

@intealls, oops you're right. I have intentionally disabled it in my 0.156 version. I should have focussed directly on Calamity's patch and not my source code.

[Calamity]

Thanks for your tests guys. Just to clarify, all builds labelled as "wiimote" have xinput enabled, regardeless of what is in the corresponding patch. This is true for all the binaries  compiled by Ves in the GA distros. It's only since recently that xinput was addes to the main groovy patch for consistency.

[harrypancakes]

I found the most recent iso (Feb 2015) for 64-bit from the main GroovyArcade page.

I downloaded it to a USB stick, and got it running on my 15.7khz rig with J-PAC. This project is amazing. It's nice when something just works they way it is described on the tin.

I also have a 32-bit machine, but I couldn't find an iso for the 2015 release for 32bit. Does it exist?

[lacanian]

You may have to use the most current version 32 bit ISO and upgrade.

[harrypancakes]

Under Extras > Update GroovyMAME?
Should it be there? The menu is empty for me.

Or do I update from the command line?

[Elbaid]

does linux support 320x200 mode x in proper 4:3 aspect? Windows will only display this in a letterbox. thanks

[cools]

You will only get pixel perfect 320x200 4:3 if you're using a CRT and you resize it vertically.

[Elbaid]

That's what I was afraid of. I increased the vertical size of my Sony PVM, but the scanlines increased in size also, which were too thick to begin with. I should have save myself the hassle and bought a consumer grade CRT

[ih8lag]

First things first... Thanks for the awesome work on GroovyArcade and GroovyMAME... I do have a few issues with the Feb 2015 release either I am overlooking something simple or there is something wrong but here is what is going on...

1. Attract-Mode when it goes into screensaver mode the screen goes black i still hear the video playing in the background.
2. WiFi does not seem to reconnect on a reboot I have the go in through the gasetup and bring the interface back up.
3. I am using the xarcade for controls and it seems like settings are not staying... I have even dumped the xarcade.cfg file for controls into the default.cfg file in /home/arcade/cfg and it gets over written. Really the only thing thats not working is inserting coins with the L/R PB buttons... i go in to general settings map the buttons and they still do not work.

Any insight you may have would be greatly appreciated.

[ih8lag]

Good news... I think I solved #3 I was mapping it in the wrong spot too may beers that night I think.

Still working on 1 & 2 as well i seem to have discovered something else. My system keyboard is set to us and it is fine in the terminal when in Attract-Mode it is not correct when typing a / i get a - this is very minor as i can configure it mostly through config files over ssh.

[Toki]

Hi,

Anyone know how to compile xf86-video-ati (7.5-02) ?

Cools AUR package don't work anymore (https://aur.archlinux.org/packages/xf86-video-ati-groovy/) -> problem with glamor-egl dependancy.
I try to apply patch (https://code.google.com/p/groovyarcade/source/browse?repo=diff#git%2Fxf86-video-ati-7.2.0%253Fstate%253Dclosed) but the version does not match

What's xf86-video-ati'version used in latest GroovyArcade live-CD ?

[cools]

If anyone else wants to take over the AUR stuff feel free. I've lost all motivation for this, Windows just works.

[lacanian]

If anyone else wants to take over the AUR stuff feel free. I've lost all motivation for this, Windows just works.

Hi @Cools, I'm having more success with GroovyArcade on my 15kHz monitor than Windows. If I stick to using GroovyArcade I'll definitely help out as much as I can. Whats involved in looking after the package? I'm not shy to give it a go, but I won't take the lead just yet.

[Doozer]

I saw 3 new files on the google shared drive. Does someone known what they aim for?

Code: [Select]

xf86-video-ati-1_7.5.0-2-x86_64Alledi2dpm-3.pkg.tar.xz
xf86-video-ati-1_7.5.0-2-x86_64Allso-1.pkg.tar.xz
xf86-video-ati-1_7.5.0-2-x86_64ED2All-2.pkg.tar.xz


[cools]

If anyone else wants to take over the AUR stuff feel free. I've lost all motivation for this, Windows just works.

Hi @Cools, I'm having more success with GroovyArcade on my 15kHz monitor than Windows. If I stick to using GroovyArcade I'll definitely help out as much as I can. Whats involved in looking after the package? I'm not shy to give it a go, but I won't take the lead just yet.

Updating the PKGBUILD with new patches when they are released against a software version, and uploading the package to AUR.

[lacanian]

First time for everything. Here is my logic.

The 3 new files on the GroovyArcade are the patches to enable low res for the xf86-video-ati 1:7.5.0-2 package.

Download xf86-video-ati 1:7.5.0-2 and the 3 patches. Patch xf86-video-ati 1:7.5.0-2 and then use PKGBUILD to create the package and upload it to AUR.

Question time.

1) The xf86-video-ati-groovy 1:7.3.0-2 package has the following sources: (allow_low_res.diff, dpm.diff,dri2.diff)
 Are those sources included in the patch on the GroovyArcade site, or do I have to find these as well?

2) Which package do I apply the patch to? The AUR packaage https://www.archlinux.org/packages/extra/x86_64/xf86-video-ati/ or xf86-video-ati-7.5.0.tar.bz2 from http://xorg.freedesktop.org//releases/individual/driver/?

3) Is there an actual patch command to be executed or are you just overwriting the files?

On another note, when we do anything GroovyArcade related lets update the wiki.

[cools]

The patches in https://aur.archlinux.org/packages/xf86-video-ati-groovy/ are older versions than the version of this package, they are fairly generic. It's a matter of applying them to the latest version of https://www.archlinux.org/packages/extra/x86_64/xf86-video-ati/ and seeing if they cleanly apply, compile, and work.

If you check the PKGBUILD that will give you the command line to patch the files (a diff is a file of differences, not a complete file by itself).

[lacanian]

New howto guide on the GA wiki for FightCade.

https://code.google.com/p/groovyarcade/wiki/FightCade_Install_HowTo

[anemul]

I tried several times to use the LiveCD , but it never ran . I am not very fluent at all in Linux . I have a Kubuntu LiveCD , that I stuck in the Cd-Rom and it just worked . This one no . I guess I will just move on and come up with another plan . I definitely wanted to try this before I went through the trouble of installing .

[fireb1adex]

So just a quick question - I set my front end to the linux desktop - not clear on how to set it back to any of the other front ends?! - ok, as simple as logging out!

Also, I am using a vertical monitor, I read back in this thread to find help (Mame is upside down and the attract mode is horizontal) - I found this much:

"fbcon=rotate:1
To the end of the kernel boot string in /boot/syslinux/syslinux.cfg rotates the console and the setup screens, which is saving my neck. "

"In mame.ini, select either "orientation rotate_r" or "orientation rotate_l", depending on how your monitor is physically rotated."

Is there anything else I should be aware of when setting vertical? Ideally this should set itself when choosing the video setting vertical option I guess?

[skycurser_phil]

Hi Calamity,

Some game developer buddies and I have some questions about the Groovy video drivers. Would you be willing to answer some quick questions via PM?

Essentially, we're 18 months into building a new arcade game and need advice. We're considering how we can use GroovyArcade as the base OS.

Thanks and would love to chat.

Peace,

Phil and the SKYCURSER dev team.

[Calamity]

Some game developer buddies and I have some questions about the Groovy video drivers. Would you be willing to answer some quick questions via PM?

No problem. Sounds interesting.

[Doozer]

Hi Ansa, I'll send you the latest kernel diff, we have added new resolution.

Nice!

Hi Ansa, Hi Ves,

What is the status with respect to the kernel patch? I would like to give the 4.x branch a try.

Cheers

Edit: I am starting with latest 3.19 diff as it seems to apply to 4.1.4.