It's the title of the Window / Process, it says Dolphin [Debug] SVN R 6580M (ATM)
Ahh much better speedstep, this is gonna be awesome
Ahh much better speedstep, this is gonna be awesome
Dolphin i7 mobile problem
|
It's the title of the Window / Process, it says Dolphin [Debug] SVN R 6580M (ATM)
Ahh much better speedstep, this is gonna be awesome 12-15-2010, 10:30 AM
(This post was last modified: 12-15-2010, 10:38 AM by NaturalViolence.)
We just need to test it with some other people with different HTed cpus to see if they behave similarly before we commit it.
Edit: I would recommend you get on irc and talk to skid_au at this point. Network: irc.colosolutions.net (or any efnet) Channel: #dolphin-emu
"Normally if given a choice between doing something and nothing, I’d choose to do nothing. But I would do something if it helps someone else do nothing. I’d work all night if it meant nothing got done."
-Ron Swanson "I shall be a good politician, even if it kills me. Or if it kills anyone else for that matter. " -Mark Antony
This is building way too fast for even -O2 optimization I think, hmm.. no I guess not.
OK Well I can get a standard build out there while I am waiting to be able to get Intel's C++ Compiler working to get SSE4.2 optimization (is this how the other builds are made? How much of a performance boost will compile-time SSE4.2 actually offer anyways?) I guess I should make a 32-bit build too. Can I just make a thread and attach it, or is that too much for the forums? Pretty small 7zipped..
Talked to skid_au, after seeing a comment here: http://code.google.com/p/dolphin-emu/sou...ail?r=6579 we discovered that the number is a bitmask, and 1 and 4 would also do the trick, and it did, and he's going to go ahead and commit that... scratch that, did commit that. I will post for testing!
Bah I built it but it's too big to attach on the forum. Gotta wait for the automated builds to come up. SVN 6584. I'd like to make a topic and have some people try it and see what happens! 12-15-2010, 03:46 PM
(This post was last modified: 12-15-2010, 03:53 PM by NaturalViolence.)
Congratulations! You have officially already contributed more to the project than 99.9% of forum members and 99.9999999999999999999% of users.
This leaves 4 things on my to do list. 1. Wait for marco to fix the fifo 2. Get rodolfo to revert his pixel depth commit and redo it properly this time using the original code from 5895 and 5896. 3. Fix the stuttering (likely an issue with the jit cache or shader cache, more likely the jit cache). 4. Add fractional efb scaling and 2xSSAA. If we can do all of that the devs may be ready to release a rc for 3.0 I would love it if you could help me with some compiling/testing. Quote:I'd like to make a topic and have some people try it and see what happens! No need at this point. It's fixed/over.
"Normally if given a choice between doing something and nothing, I’d choose to do nothing. But I would do something if it helps someone else do nothing. I’d work all night if it meant nothing got done."
-Ron Swanson "I shall be a good politician, even if it kills me. Or if it kills anyone else for that matter. " -Mark Antony (12-15-2010, 03:46 PM)NaturalViolence Wrote: Congratulations! You have officially already contributed more to the project than 99.9% of forum members and 99.9999999999999999999% of users. Yay!.. Quote:I would love it if you could help me with some compiling/testing. Sure. Compiling/testing is my specialty . Are you officially on this project? Quote:Quote:I'd like to make a topic and have some people try it and see what happens! I am just curious if there anyone gets any noticeable speedups from it. Do you have any idea how much ICC (i.e. SSE4.2) would speed up the build vs VC? I'd want to build it with gcc/mingw-w64 for fun but I am sure it would just be slower alas.
I got ICC working, am trying a build with it now. Now THIS is what optimized compiling should look like, my CPU groaning
Hey maybe you can give me one final HT lesson.. how is it possible for all 8 threads to be in 100% use, 100% CPU use reported, if they are not all available at-once? Or is stuff bouncing around SO fast that it just looks that way, with god knows how many parallel compiling threads going on? I should go check out your chart again for you. 12-16-2010, 02:11 AM
Quote:Hey maybe you can give me one final HT lesson.. how is it possible for all 8 threads to be in 100% use, 100% CPU use reported, if they are not all available at-once? Or is stuff bouncing around SO fast that it just looks that way, with god knows how many parallel compiling threads going on? From what I understand they are all available at once just not used at once. You just need 8 software threads and let the OS kernel deal with it. Quote:Do you have any idea how much ICC (i.e. SSE4.2) would speed up the build vs VC? No I do not since I don't have an sse4.2 cpu. However based on responses from the forums I can imagine that the difference would be very small. Quote:Sure. Compiling/testing is my specialty . Are you officially on this project? As a dev....no. But that doesn't really limit what you can do as much as you would think. The only thing you can't do if you're not a dev is commit changes. But you can still make a patched file and give it to the devs to commit, as many people here on the forums do. Or you can find issues in your games, track down the revision that caused it and inform the devs. Basically the debugging process is exactly what you just went through.
"Normally if given a choice between doing something and nothing, I’d choose to do nothing. But I would do something if it helps someone else do nothing. I’d work all night if it meant nothing got done."
-Ron Swanson "I shall be a good politician, even if it kills me. Or if it kills anyone else for that matter. " -Mark Antony |
« Next Oldest | Next Newest »
|