|
Gyruss slowdown in 148u1 possible bug.
#305264 - 03/04/13 09:34 PM
|
|
|
Hi, speed in gyruss is dropping from 100% to 90% and back in 148u1. Unthrottled about 400%. Is anyone else seeing this?
Ive tried mametesters but the site is down at the moment. Thanks guys.
|
|
|
Tafoid |
|
I keep on testing.. testing.. testing... into the future!
|
|
|
|
|
Reged: 04/19/06
|
|
Posts: 3135
|
|
Loc: USA
|
|
|
|
Send PM
|
|
|
Re: Gyruss slowdown in 148u1 possible bug.
[Re: russ h.]
#305265 - 03/04/13 09:54 PM
|
|
|
> Hi, speed in gyruss is dropping from 100% to 90% and back in 148u1. Unthrottled about
> 400%. Is anyone else seeing this?
>
> Ive tried mametesters but the site is down at the moment. Thanks guys.
Try shutting off multithreading (-nomt) and see if that helps.
Other than that there are few things that can cause fluxs from 100% speed like that.
|
|
|
|
Re: Gyruss slowdown in 148u1 possible bug.
[Re: Tafoid]
#305272 - 03/05/13 02:02 AM
|
|
|
I've had gyruss slow down on my pc as well. Turning off multithreading didn't solve it for me, I had to lower the numprocessors settings in mame.ini.
|
|
|
|
Re: Gyruss slowdown in 148u1 possible bug.
[Re: hap]
#305273 - 03/05/13 02:30 AM
|
|
|
Also happening on my end.
|
|
|
| Sune |
|
Connected
|
|
|
|
|
Reged: 09/21/03
|
|
Posts: 5648
|
|
Loc: Lagoa Santa, Brasil
|
|
|
|
Send PM
|
|
|
It'll get worse
[Re: joey35car]
#305275 - 03/05/13 02:59 AM
|
|
|
|
|
|
Re: It'll get worse
[Re: Sune]
#305286 - 03/05/13 08:49 AM
|
|
|
I compiled a MAME with this revision and 90% of Konami CPU games are broken. They have very bad gfx, reset themselves at startup, don't boot at all, crashes with some unhandled OPCODE exception and so on..... Good luck the the submitter in fixing all the timings.
|
|
|
|
Re: Gyruss slowdown in 148u1 possible bug.
[Re: joey35car]
#305290 - 03/05/13 01:07 PM
|
|
|
Does setting mame.ini numprocessors to a lower value (4 or 2) circumvent the slowdowns for you?
|
|
|
|
Re: Gyruss slowdown in 148u1 possible bug. (edited)
[Re: hap]
#305293 - 03/05/13 03:44 PM
|
|
|
Yes, Iv just changed from auto to 4 and its now working normally. nice one.
I'm using win7 64bit with i7 cpu.
Also Iv just tried ridge racer with it on 4 and that's not slowing down where it used to.
not tried much else yet, but looks like the issue is with numprocessors.
I think I'll leave mine on 4 for now. Thanks for taking a look.
Edited by russ h. (03/05/13 04:23 PM)
|
|
|
|
Re: Gyruss slowdown in 148u1 possible bug. (edited)
[Re: russ h.]
#305296 - 03/05/13 04:54 PM
|
|
|
Good =) I'm kinda glad I'm not the only one that has(had) this slowdown issue.
i7 and win7 64bit as well here, MAME autodetects numprocessors as 8 (4 physical cores + hyperthreading).
|
|
|
| B2K24 |
|
MAME @ 15 kHz Sony Trinitron CRT user
|
|
|
|
|
Reged: 10/25/10
|
|
Posts: 2663
|
|
|
|
|
|
Send PM
|
|
|
Re: It'll get worse
[Re: Mamesick]
#305351 - 03/07/13 12:31 AM
|
|
|
> I compiled a MAME with this revision and 90% of Konami CPU games are broken. They
> have very bad gfx, reset themselves at startup, don't boot at all, crashes with some
> unhandled OPCODE exception and so on..... Good luck the the submitter in fixing all
> the timings.
After reading this comment by David Haywood, it made me think of this current Konami stuff that's happening now.
|
|
|
| R. Belmont |
|
Cuckoo for IGAvania
|
|
|
|
|
|
Reged: 09/21/03
|
|
Posts: 9716
|
|
Loc: ECV-197 The Orville
|
|
|
|
Send PM
|
|
|
Re: It'll get worse
[Re: B2K24]
#305395 - 03/07/13 07:46 PM
|
|
|
> After reading this comment by David Haywood, it made me think of this current Konami
> stuff that's happening now.
If I understand you correctly, that's a tragic misinterpretation of what's going on. Nathan is replacing the old MAME-spec "close enough for government work" 6809 family CPU core with a new cycle-by-cycle accurate one, similar to what OG recently did with the 6502 family. There are obviously going to be teething pains, but it's certainly not indicative of a lack of quality focus; it's actually the exact opposite and will enable *far* more accurate emulation of the games in the future.
|
|
|
| B2K24 |
|
MAME @ 15 kHz Sony Trinitron CRT user
|
|
|
|
|
|
Reged: 10/25/10
|
|
Posts: 2663
|
|
|
|
|
|
Send PM
|
|
|
Re: It'll get worse
[Re: R. Belmont]
#305400 - 03/07/13 08:33 PM
|
|
|
I would never dare to criticize past or current/future work and I apologize if it came across that way.
What I meant to say was when looking at some points made in the comment and more specifically
"when the very same code actually falls apart when presented with real challenges to test its accuracy"
we seem to have been shown an example or validation of that with current progress.
|
|
|
| R. Belmont |
|
Cuckoo for IGAvania
|
|
|
|
|
|
Reged: 09/21/03
|
|
Posts: 9716
|
|
Loc: ECV-197 The Orville
|
|
|
|
Send PM
|
|
|
Re: It'll get worse
[Re: B2K24]
#305404 - 03/07/13 09:30 PM
|
|
|
> "when the very same code actually falls apart when presented with real challenges to
> test its accuracy"
>
> we seem to have been shown an example or validation of that with current progress.
Yes, indeed. There's some timing-related tricks on some MESS systems that weren't possible with the current code.
|
|
|
|
Re: It'll get worse
[Re: R. Belmont]
#305475 - 03/09/13 10:22 AM
|
|
|
Latest SVN I compiled is really fine. All things in KONAMI CPU games are come back as before if not better. Sadly the infamous KONAMI+YM2151 sound bug is still there. Though I'm really happy about this cycle-exact rewrites. I'm writing for my personal taste a CBM64 emulator based on MAME/MESS source and the M6502/M6510 cycle-exact core helped me a lot. So, as always, thanks to all people who make MAME/MESS always more accurate.
|
|
|
|
Re: It'll get worse
[Re: B2K24]
#305476 - 03/09/13 12:01 PM
|
|
|
> "when the very same code actually falls apart when presented with real challenges to
> test its accuracy"
>
> we seem to have been shown an example or validation of that with current progress.
We've been telling Haze that ever since he started his daft merging rant, I guess it finally sunk in.
|
|
|
Index 
Threaded 
Gyruss slowdown in 148u1 possible bug.
Reply
[Re: 
