Resolved: [Bug Report] Tomb Raider: Legend [PAL]

[Maetel85]

# PCSX2 version: v1.2.1 and 1.4.0
# Plugins used: GSdx, SPU2-X, Lilypad, cddvdGigaherz,rest null plugins
# Plugin settings: mostly default, the game is being played at 4x resolution and 8x AA

PC specs:

- Win 8.1 64 bit
- 8GB DDR3 RAM
- Intel 3570k
- AMD Radeon 7950 OC (need more specs?)

-Description: in certain locations the screen gets filled with small monocolour squares which appear only in hardware mode as software mode isn't affected. I tried v.1.2.1 6 months ago and 1.4.0 now but the issues is still here. I also tried the game with a Lenovo laptop with which, although clearly with lower performance, I experienced the same issue in the same location.
The issue stops once you finish that location which could be a single corridor, an entire hall, etc.
I also tried playing with hacks, mtvu, etc. and switching from DX11 hardware to dx9 or OGL to no avail. I see no info about this issue and I don't even know the particular name, so I can't find a solution. Here is a GS dump:

[gregory]

Do you have the issue in SW mode if you disable mipmap support?

[Maetel85]

Do you have the issue in SW mode if you disable mipmap support?

No I haven't it

[Ryudo]

Same issue - http://forums.pcsx2.net/Thread-GSDX-Tomb...#pid465230

Both Underworld and Legend suffer from it, but Underworld is broken far worse, where even software mode is glitchy.

[CK1]

Code:

bool GSC_TombRaiderLegend(const GSFrameInfo& fi, int& skip)
{
    if(skip == 0)
    {
        if(fi.TME && fi.FBP == 0x01000 && fi.FPSM == fi.TPSM && fi.TPSM == PSM_PSMCT32 && (fi.TBP0 == 0x2b60 ||fi.TBP0 == 0x2b80 || fi.TBP0 == 0x2E60 ||fi.TBP0 ==0x3020 ||fi.TBP0 == 0x3200 || fi.TBP0 == 0x3320))
        {
            skip = 1;
        }
        else if(fi.TPSM == PSM_PSMCT32 && (fi.TPSM | fi.FBP)==0x2fa0 && (fi.TBP0==0x2bc0 ) && fi.FBMSK ==0)  
        {
            skip = 2;
        }
        
        
    }// ||fi.TBP0 ==0x2F00

    return true;
}

bool GSC_TombRaiderUnderWorld(const GSFrameInfo& fi, int& skip)
{
    if(skip == 0)
    {
        if(fi.TME && fi.FBP == 0x01000 && fi.FPSM == fi.TPSM && fi.TPSM == PSM_PSMCT32 && (fi.TBP0 == 0x2B60 /*|| fi.TBP0 == 0x2EFF || fi.TBP0 ==0x2F00 || fi.TBP0 == 0x3020*/ || fi.TBP0 >= 0x2C01 && fi.TBP0!=0x3029 && fi.TBP0!=0x302d))
        {
            skip = 1;
        }
        else if(fi.TPSM == PSM_PSMCT32 && (fi.TPSM | fi.FBP)==0x2c00 && (fi.TBP0 ==0x0ee0) && fi.FBMSK ==0)  
        {
            skip = 2;
        }
        /*else if(fi.TPSM == PSM_PSMCT16 && (fi.TPSM | fi.FBP)>=0x0 && (fi.TBP0 >=0x0) && fi.FBMSK ==0)  
        {
            skip = 600;
        }*/
    }

    return true;
}

@gregory - Found hacks directly referencing both games in GSState.cpp

Maybe these aren't needed anymore?

[gregory]

Yes maybe it is working on openGL. I don't know what the hack does. If you can confirm me that removing hack break rendering on Dx but improve rendering on openGL I will gladly put them on the Dx level.

[lightningterror]

Resolved in:
https://github.com/PCSX2/pcsx2/pull/11299

Thank you for your report.
This bug report has now been marked as Resolved since it has been fixed on our code base.

This thread will now be closed and moved to the resolved bug reports subforum.