KWSN Simon and group,

Great work on the Rev2 Windows science apps stuff...

Any Rev 2s for Linux? Please?...

In 64-bits?

For AMD? Athlon X2 even...?


;-)

Happy crunchin',
Martin

---

See new freedom: Mageia Linux
Take a look for yourself: Linux Format
The Future is what We all make IT (GPLv3)

KWSN Simon and group,

Great work on the Rev2 Windows science apps stuff...

Any Rev 2s for Linux? Please?...

In 64-bits?

For AMD? Athlon X2 even...?


;-)

Happy crunchin',
Martin

Nobody who has tried has been able to get the 2.x sources to build for Linux yet. They do build with DevC++/MinGW/GCC 3.4.2 for Windows, so hopefully the incompatibility could be resolved by someone thoroughly familiar with Linux. I have never ventured into *nix yet, though hope to find the time to do so.

The Windows Rev-2.2B Sources are available, of course.

                                                               Joe

KWSN Simon and group,

Great work on the Rev2 Windows science apps stuff...

Any Rev 2s for Linux? Please?...

In 64-bits?

For AMD? Athlon X2 even...?

Nobody who has tried has been able to get the 2.x sources to build for Linux yet. They do build with DevC++/MinGW/GCC 3.4.2 for Windows, so hopefully the incompatibility could be resolved by someone thoroughly familiar with Linux. I have never ventured into *nix yet, though hope to find the time to do so.

The Windows Rev-2.2B Sources are available, of course.

I thought part of the clever performance trick was "just" to compile using the Intel ICC system with static libraries and the Intel anti-AMD code dongle deactivated.

NB: High geek-factor required to overcome the ""just"" bit!


Or has gcc + flags got a fighting chance at 64-bits?

Regards,
Martin

---

See new freedom: Mageia Linux
Take a look for yourself: Linux Format
The Future is what We all make IT (GPLv3)

Martin,

in my experience, GCC does not stand a chance. Recent (4.xx) versions have lessened the performance differential, but they're still quite a bit slower than ICC.

You could try using IPP with GCC - that would require pretty large source edits, though, AFAICS.

I've managed to compile most source files of the science app under Linux using 2.0 sources, but could never get it to work totally. Other people have tried but I haven't heard of any success stories, so far.

This is something I'd very much like to change - but lacking coding skills, it may take me a while to do so. So if you'd like to step in and give us a push, you're very welcome :o)

Intel's ICC and IPP for Linux are available for free as full versions for personal, non-commercial usage (links in the How-To's on lunatics.at).

HTH,
Simon.

---

Donate to SETI@Home via PayPal!

Optimized SETI@Home apps + Information

KWSN Simon and group,

Great work on the Rev2 Windows science apps stuff...

Any Rev 2s for Linux? Please?...

In 64-bits?

For AMD? Athlon X2 even...?

Nobody who has tried has been able to get the 2.x sources to build for Linux yet. They do build with DevC++/MinGW/GCC 3.4.2 for Windows, so hopefully the incompatibility could be resolved by someone thoroughly familiar with Linux. I have never ventured into *nix yet, though hope to find the time to do so.

The Windows Rev-2.2B Sources are available, of course.

I thought part of the clever performance trick was "just" to compile using the Intel ICC system with static libraries and the Intel anti-AMD code dongle deactivated.

NB: High geek-factor required to overcome the ""just"" bit!


Or has gcc + flags got a fighting chance at 64-bits?

Regards,
Martin

The noncommercial/Linux Intel compiler uses GCC syntax, though probably a more recent dialect than GCC 3.4.2. I think the intent is to be able to compile from the same Linux code as current GCC, then add Intel specific tweaks. Simon of course would not violate the Intel license terms by distributing apps with the GenuineIntel checks removed.

As for 64-bit builds, there are probably additional difficulties. I have no way to judge how many, but I'll freely admit that because I don't know either Linux or 64-bit systems I've probably put inappropriate preprocessor conditionals into the code.

                                                                Joe

Intel's ICC and IPP for Linux are available for free as full versions for personal, non-commercial usage (links in the How-To's on lunatics.at).

In going the ICC route, is there then any turgidness about releasing the resulting binaries?...

Cheers,
Martin

---

See new freedom: Mageia Linux
Take a look for yourself: Linux Format
The Future is what We all make IT (GPLv3)

Martin,

not as long as you do not make any money out of it or derivatives.
Basically, offering apps for free is quite fine.

HTH,
Simon.

---

Donate to SETI@Home via PayPal!

Optimized SETI@Home apps + Information

Hm,

I think I found a source for a large number of compile errors I was getting - mainly, Linux doesn't seem to like the "__fastcall" compiler directive, much.

Any code containing this (there are quite a few occurrences) errors out. Actually, the compiler should choose correctly on its own, so I removed those directives, and voila, less than half the compile problems.

There's probably a cleaner way to do this by redefining __fastcall to something supported under Linux; I'll have to rely on someone more knowledgeable for that, though.

Still, quite a few to go ;o)

Regards,
Simon.

---

Donate to SETI@Home via PayPal!

Optimized SETI@Home apps + Information

Hm,

I think I found a source for a large number of compile errors I was getting - mainly, Linux doesn't seem to like the "__fastcall" compiler directive, much.

Any code containing this (there are quite a few occurrences) errors out. Actually, the compiler should choose correctly on its own, so I removed those directives, and voila, less than half the compile problems.

There's probably a cleaner way to do this by redefining __fastcall to something supported under Linux; I'll have to rely on someone more knowledgeable for that, though.

Still, quite a few to go ;o)

Regards,
Simon.

Compiling the source is not the issue, i successfully did that ;-)

If you want i can share the modified source and the script for building the app.
(and a little howto)

---

Join BOINC United now!

Hi Crunch3r,

that'd be pretty neat, actually. Right now, I've almost got it done; but still I'm getting lots of undefined index errors on linking that I can't get rid of.

So if you want to share how you got it to compile, scripts and/or a How-To, I'd appreciate it a lot :o)

Since you're registered at lunatics.at, you could to attach it to a post there or email me at simon at zadra dot org.

Netter Zug von Dir :o)

Regards,
Simon.

---

Donate to SETI@Home via PayPal!

Optimized SETI@Home apps + Information

Hi Crunch3r,

that'd be pretty neat, actually. Right now, I've almost got it done; but still I'm getting lots of undefined index errors on linking that I can't get rid of.

So if you want to share how you got it to compile, scripts and/or a How-To, I'd appreciate it a lot :o)

Since you're registered at lunatics.at, you could to attach it to a post there or email me at simon at zadra dot org.

Netter Zug von Dir :o)

Regards,
Simon.

OK.

i'll zip the files and will send you an email in about 20 min.

---

Join BOINC United now!

That really sounds like good news!!!

I got some Linux machines (32 and 64 bit, suse and ubuntu) if you need a place to test what you compiled... :-)

grüßle
mic.

---

mic.

I'm interesed also about it :) I totaly switch to linux, but app rev1.3 is so slow in comparition to app rev2.4 in windows... (~1300s slower on 30Credit)

I hold thumbs too fast as posbile realese of linux app rev >2 :D

---

Hi,

This is just for INFO.

Today i had some major progress with porting the code to linux... ATM i think i'm at 60 to 75% done porting the code.

Today i solved/debugged a major part of the code that previously was the cause for crashing the app... and i think that there's some kind of a breakthrough towards porting the code. ( It's kinda hard if you're not a C or c++ coder after all... but i guess i got some experience from pre "ENHANCED" code after all ;-) )

However... there are still some things to be done, but that depends on the support i get from Ben, Joe and Alex...

We'll see if we can solve that issue :-)

---

Join BOINC United now!

Sounds good. We've got a dual quad core Xeon running Linux that could use a boost. On a per core basis my C2D machine running Vista is kicking its you know what.

---

Hi Crunch3r,

check back at lunatics.at, I just gave you access to the development area. Thanks for getting involved!

Regards,
Simon.

---

Donate to SETI@Home via PayPal!

Optimized SETI@Home apps + Information

We'll see if we can solve that issue :-)

I'm happy to hear that this is going ok... glad someone can do this stuff, I'm just trying to stay on top of learning linux : )

Looking forward to the v2 for linux, whenever it arrives!

Many thanks,
N.

---

Crunching SETI@Home as a member of the Whirlpool BOINC Teams

Looking forward to the v2 for linux, whenever it arrives!

Hi folks,

today i finnished porting the code thanks to Joe and Ben.

The application is working and should hopefull be ready for release this week.

---

Join BOINC United now!

Hi folks,
today i finnished porting the code thanks to Joe and Ben.
The application is working and should hopefull be ready for release this week.

great news, well done Crunch3r and crew

---

Ooh, ooh!
I'm ready!!!p