[Bug Report] College Hoops 2k7 [NTSC-U]
#1
[Bug Report] College Hoops 2k7 [NTSC-U]
# PCSX2 version: SVN 1.1.0. 5724
# CPU options: EE/IOP defaults, VUs defaults
# Plugins used: ZZ Ogl PG CG (Debug) 0.3.0 [libzzogl-cg-0.3.0]
# Description: I can get into the menu, but after chose team and got start to the game, emulator asserted!
Code:
plugins/zzogl-pg-cg/opengl/HostMemory.cpp:212: void TransferHostLocal(const void*, u32): Assertion `start < gs_imageEnd' failed.
Reply

Sponsored links

#2
Post full emu log and try f9 .ripped your iso again and the last one try latest svn or stable one its help you in this case and see the bug still exist or not .
Core i3 9100f 3.6Ghz
RAM=8GB
nvidia GT 1030
pcsx2 version-1.3.1  
Reply
#3
(08-27-2013, 03:25 PM)Brock lessner Wrote: Post full emu log and try f9 .ripped your iso again and the last one try latest svn or stable one its help you in this case and see the bug still exist or not .

Oh boy, another person who saw posts of others trying to help and now suddenly feels like he can aid everyone with generic knowledge.

That error is irrelevant to the integrity of the .iso or rendering type. He's also on the latest SVN. And instead of apologizing, you could do better by reading better and holding your horses when you are not sure whether you can be of any help, seeing as how you're still quite clueless about the emulator.
Reply
#4
that said, i'd like to see if the problem occurs in software mode to see if the problem comes from ZZogl or from the emucore itself
CPU : AMD Ryzen 7 3800X
Mobo : Asus PRIME B450-PLUS
GPU : NVIDIA GeForce RTX 3070
RAM : 16 Go
Reply
#5
I think the problem comes from ZZogl.
According to the assertion, the error was occured in TransferHostLocal() (zzogl-pg-cg/opengl/HostMemory.cpp), actually the error data was generated by GetRectMemAddress(...), ok, I'll reproduce the situation here:

before the assertion, we call GetRectMemAddress(...) with parameters:
Code:
start, end, psm:20(PSMT4), x:0, y:0, w:1, h:1, bp:13770, bw:128

those parameters was configured by app through configuring GS registers(BITBLTBUF/TRXPOS/TRXREG/TRXDIR), after that, we got:
Code:
start:0x00296a00, end:0x00296a00, that generate the assertion!

so, the point is what was doing in GetRectMemAddress(...).
The core implementation is like that:
Code:
bits = PSMT_BITS_NUM(psm);
start = getPixelFun[psm](x, y, bp, bw);
end = getPixelFun[psm](x + w - 1, y + h - 1, bp, bw) + 1;

if (bits > 0)
{
    start *= bits;
    end *= bits;        
}
else
{
    start /= 2;
    end /= 2;
}
cuz PSMT_BTS_NUM macro will return 0 when psm is PSMT4, when 'start' and 'end' meets: end == start + 1, it'll branch to 'else', that produces the 'start' and 'end' in same.

PS: I'm not familiar with those codes yet, but I think there is something wrong here. and BTW, at first I added "if (start == end) end += 1" after "end /= 2" to "if (start == end), I got into the game, but unfortunately, few minitus later, the game throw a bad_alloc exception.
Reply
#6
(08-27-2013, 10:28 PM)Coornio Wrote: Oh boy, another person who saw posts of others trying to help and now suddenly feels like he can aid everyone with generic knowledge.

That error is irrelevant to the integrity of the .iso or rendering type. He's also on the latest SVN. And instead of apologizing, you could do better by reading better and holding your horses when you are not sure whether you can be of any help, seeing as how you're still quite clueless about the emulator.

ohh brother extremly sorry about that , :-P i think its better to kept my hourse in the farm .. :-).
Core i3 9100f 3.6Ghz
RAM=8GB
nvidia GT 1030
pcsx2 version-1.3.1  
Reply




Users browsing this thread: 1 Guest(s)