Exactly as stated - Reicast crashes when loading games using Dynarec. But if I use interpreter, they work but they're unfortunately slow and laggy. The emulator just seems to be broken and useless. These games did used to work, but no longer do.
I'm using the latest Reicast from Google Play
Amlogic Q-Box
Android 4.4.2
Games neither work as CHD nor GDI
This could be fixed by #1710. Are you using betas or the public version?
Nice. Im using the public version, the public version, from the Google play store.
r19.07.3-3-g29f95ee2
Where can we find the beta versions?
Beta r19.07.3-52-gfe2ba8cc should be out within 1h, please test and report back
http://enroll.reicast.com to get betas via google play, or http://build.reicast.com to download them manually
Ah damn. Looks like the issue is still there. Still crashes.
Can you join http://chat.reicast.com and poke me (skmp) there? I can't repro here, so I need your help to figure out what's wrong
Yeah sure. I can do that. I'll just sign up there.
After a long remote debugging session w/ @Cee123, this seems to be the libart bug I've worked around before in older versions of the shield. It's a bug that got fixed around the android6/7 timeline (2-3 years ago iirc), so mostly people with android 5.xx are affected.
While this is strictly not a reicast bug, I'll try to make a workaround for people with the buggy version
Beta r19.07.3-67-g4b0fce26 (building right now) should fix those issues
Err didn't meant to close. @Cee123 @DitzRoom @spielvan @Emulatorer please test when possible, http://enroll.reicast.com is the link to activate betas for your account
@skmp the one I reported is android Xiaomi Redmi 6a Helio P22 version oreo but in my phone mali-450 android 5.1 Reicast is ok in the latest build.
Last night it kept crashing afterwards and started doing it with interpreter as well.
Still crashes after the bios screen man. Must be something else wrong there.
v19.07.4-1-g6a462750
I think it's this crappy version of Android 5.1.1. Wish there was a way of updating it.
Sigh 😞
On Wed, 31 Jul 2019, 04:44 Cee123, notifications@github.com wrote:
Still crashes after the bios screen man. Must be something else wrong
there.—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
https://github.com/reicast/reicast-emulator/issues/1709?email_source=notifications&email_token=AADAAMRPSV4G7HXTJYG72RLQCD4APA5CNFSM4IHX2CY2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD3F4OII#issuecomment-516671265,
or mute the thread
https://github.com/notifications/unsubscribe-auth/AADAAMW4PEKOIVBIF6PHREDQCD4APANCNFSM4IHX2CYQ
.
Well, if we do a couple of debugging sessions I'm sure we'll figure it out. These things just take time.
PPSSPP also is having a problem with android box 🤔
See~https://github.com/hrydgard/ppsspp/issues/10625
Because my android phone version is lollipop but I haven't encounter crashing the latest build of reicast so maybe this is just a platform specific to "android box" ?
•sorry for my english 🙂
PPSSPP seems to work on mine. Which is strange because it used to be really slow.
FYI @skmp
Found the reason why the app is crashing using dynarec Android (TV).
The issue is the 4189cd432f6ae36d90d8cab62283eaa42713cf3d commit.
git revert 4189cd432f6ae36d90d8cab62283eaa42713cf3d works as a temporary solution.
Looks like bm_RamPageHasData function is not working correctly and returns false when it should not to, the behaviour should be verified and fixed.
Error from debugging (program counter points to a memory area which probably is not an instruction memory but a data memory, or maybe some stack overflow?):
A/libc: Fatal signal 4 (SIGILL), code 1, fault addr 0xd2a3d3ae in tid 21075 (Thread-5), pid 20972 (eicast.emulator)
Awesome, so we know that the issue is now should be possible to fix that somehow.
@Cee123 should be fixed in r20.02b, can you re-test and confirm?
@skmp Yeah I'll test it out see if it works now. Thanks so much. That's excellent!
@skmp Oh my god! Thank you. Excellent work!! I can confirm that it works now. The crashing issue is gone.
:)
Most helpful comment
@Cee123 should be fixed in r20.02b, can you re-test and confirm?