-
-
Notifications
You must be signed in to change notification settings - Fork 2k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Memory Mirror commit regression #4566
Comments
Heavy Rain regressed like month ago but nvm |
There is also regression in D3 BLUS31437, fps decreases on 2nd boot, this does not happen before Memory Mirror commit even when spu cache was implemented. SPUASMJIT Recompiler |
Can someone confirm this is an issue in latest master and provide games that actually regressed on that PR and not on previous ones? (Apart from D3 report above) |
The issue is still exist on the latest master, first two games went ingame for me on the commit before 5d15d64, and i did some more tests for these two games, i tested them on older versions of Memory Mirror, just in case if they may work, and they don't work exactly the same way |
Added a fix for the performance issues with #4876 which has now been merged. Should restore GT HD performance. |
Confirm, the performance part fixed |
Fight Night Round 4 recently regressed with a "MEM Access Violation" warning as well though I'm not sure on what build it happened and if it's related to this problem. The game used to be fully playable (apart from broken audio and occasional stutter) in #4424. Force Cpu Blit does bypass the memory access violation but the player models are gone now (it might be some other regression or side effect of Force Cpu Blit). |
I can confirm that the issue on D3 has already been fix in master. |
All cases i've reported fixed now, closing this. |
Heavy Rain still stucked before menu. I dont know how about u. Tho its not showing access violation for a long time. |
Log? |
It need patch, so half fixed. |
RPCS3.log.gz |
Of course you can't, you're using an older build... |
Sh*t, I was sure I updated. Guess I'll test it again |
And there it boots, I feel so dumb |
Found a few regressions
in 5d15d64
F {PPU[0x1000009] Thread (Read Chunk Thread) [0x0056dc9c]} MEM: Access violation writing location 0xd00c08f0
F {PPU[0x10000c4] Thread (CInfoManagerPS3) [0x00705074]} MEM: Access violation writing location 0xd0330c50
3)Gran Turismo HD Concept [NPEA90002] Speed regression
before
after
did some extra debugging work for this on, since in the RP there were many versions of "Memory Mirror" i found out that 374dd7c has fine speed and b87d632 regressed, the reason in VirtualMemory.cpp changes between two versions
There're maybe other regressions, but other my games seems to work fine
The text was updated successfully, but these errors were encountered: