Braid [NPUB30133]
Started by ssshadow




46 posts in this topic
jacky400
Unregistered


 
07-02-2014, 12:43 PM -
#21
it is now with texts in intro and in-game menu.



nevermind1711
Unregistered


 
08-27-2014, 10:53 AM -
#22
I tested it on i5 2500, nvidia gtx 770, windows 7 x64. Depth and color buffer on, 1280x720 and it crashes after the loading screen Sad



Here is the log:

http://pastebin.com/0nsnpgaS
pnkie666
Unregistered


 
08-27-2014, 01:23 PM -
#23
(08-27-2014, 10:53 AM)nevermind1711 Wrote: I tested it on i5 2500, nvidia gtx 770, windows 7 x64. Depth and color buffer on, 1280x720 and it crashes after the loading screen Sad



Here is the log:

http://pastebin.com/0nsnpgaS

it's strange because i launch this game without problem, look here:
nevermind1711
Unregistered


 
08-27-2014, 01:30 PM -
#24
(08-27-2014, 01:23 PM)pnkie666 Wrote:
(08-27-2014, 10:53 AM)nevermind1711 Wrote: I tested it on i5 2500, nvidia gtx 770, windows 7 x64. Depth and color buffer on, 1280x720 and it crashes after the loading screen Sad



Here is the log:

http://pastebin.com/0nsnpgaS

it's strange because i launch this game without problem, look here:

Do you use any special setting? Have you noticed any difference in the log or can you post it here?
Pachus
Unregistered


 
10-23-2014, 12:30 AM -
#25
Hello everybody, here's my first post. The game stoped working for me in Nekotekina's Spurs commit (more or less I think), now It runs, but the game textures are bugged as hell:

[Image: GIXkxr9l.png]


I wonder if I'm the only one, check it out. See u.

Log:

Quote:https://mega.co.nz/#!4cBmiRZJ!xyONtdO-6F...dc7p0lvC6o
nevermind1711
Unregistered


 
10-23-2014, 09:01 PM -
#26
same here Wink well, at least it doesn't crash anymore for me Big Grin
DaHandy
Unregistered


 
12-16-2014, 11:09 PM -
#27
(10-23-2014, 12:30 AM)Pachus Wrote: Hello everybody, here's my first post. The game stoped working for me in Nekotekina's Spurs commit (more or less I think), now It runs, but the game textures are bugged as hell:

img

I wonder if I'm the only one, check it out. See u.

Log:

Quote:https://mega.co.nz/#!4cBmiRZJ!xyONtdO-6F...dc7p0lvC6o

Hello rpcs3 community!

I have been browsing through the rpcs3 code on GitHub and to me it seems that for example CELL_GCM_NV4097_SET_VERTEX_TEXTURE_ADDRESS is already an implemented function (in RSXTexture methodRegisters, with a different name though: NV4097_SET_TEXTURE_ADDRESS).

Why does RSXThread still throw a TODO message into the log if that is the case?

Sorry if I am just being stupid here... Angel

Edit:
I just realized that for example SetVertexTextureAddress (0x00040908) and SetTextureAddress (0x00041A08) are two different functions.

However NV4097_SET_TEXTURE_ADDRESS in RSXTexture methodRegisters seems to behave just like SetVertexTextureAddress in RSXFIFOCommands (http://www.psdevwiki.com/ps3/RSXFIFOComm...ureAddress) except with a different address. Is there a mistake in the code or am I making wrong interpretations?

If both functions behave in the same way, it should be very easy to also implement the corresponding vertex functions such as NV4097_SET_VERTEX_TEXTURE_ADDRESS based on the already existing code.

Here is also my log: http://tny.cz/8fb35a73
DaHandy
Unregistered


 
12-17-2014, 04:23 PM -
#28
(12-17-2014, 05:16 AM)AlexAltea Wrote: @DaHandy: Right, as you said NV4097_SET_VERTEX_TEXTURE_ADDRESS (0x908) and NV4097_SET_VERTEX_TEXTURE_ADDRESS (0x1A08) are different RSX methods. Just FYI, RSX methods work like this:

[Image: ecad0e5f9b1428e520254fcec53b5c61.png]
Sorry for the #define's. Gotta replace them with BitField<from, to> or something.

Up to the date, I'm not quite sure what's the deal with the RSX methods, they seem to be both a way to store information somewhere in the GPU using their offset as pointer *AND* to perform operations.

Regarding the first aspect (storing information), RPCS3 has members on RSXThread for most of the methods which are updated on the big switch case. Only a few of them (e.g. RSXTexture objects), are actually handled in a way that resembles more the PS3 (using a methodRegisters object).

I'd say all the other methods should switch to this methodRegisters approach as well.

Thanks for your answer Alex!

I'm hopefully going to look more into that during the Christmas holidays as I get bored. I guess no one except Sony knows how the RSX really operates? Its just going to be a lot of guessing along with reversing the real PS3 console.

PS: Related to this, I just noticed that jacky400 commited some kind of a vertex texture implementation 2 hours ago, today.
vlj
Unregistered


 
10-18-2015, 09:49 PM -
#29
It almost works with dx12. It looks like the geometry is wrong, but the texture are inplace.
vlj
Unregistered


 
10-25-2015, 01:14 AM -
#30
The game use primitive restart, that's why there was some rendering issue.
Doing some dirty hack I almost got the game sort-of working.


Forum Jump:


Users browsing this thread: 1 Guest(s)