| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Completed Research
Forum: Help Cure Muscular Dystrophy - Phase 2 Forum
Thread: Continual error 193 |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 42
|
|
| Author |
|
|
mjeffrey
Cruncher Joined: Jan 17, 2005 Post Count: 10 Status: Offline Project Badges: 
|
I am also seeing this error on my oldest CPU systems
|
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I was neither able to successfully crunch HCMD2 on a P3 (OS is Debian Lenny) I joined WCG a couple of years ago because I wanted to contribute to HCMD ... but could not do so because it didn't work under linux. Rather ironic that now HCMD2 is out and even HCMD scientists can't crunch it under linux. |
||
|
|
JmBoullier
Former Community Advisor Normandy - France Joined: Jan 26, 2007 Post Count: 3715 Status: Offline Project Badges: 
|
I was neither able to successfully crunch HCMD2 on a P3 (OS is Debian Lenny) I joined WCG a couple of years ago because I wanted to contribute to HCMD ... but could not do so because it didn't work under linux. Rather ironic that now HCMD2 is out and even HCMD scientists can't crunch it under linux. The point raised by Michaël is not Linux but the older computer (P3 in his case). Thousands of HCMD2 WUs have been run and validated under Linux. Cheers. Jean. |
||
|
|
nasher
Veteran Cruncher USA Joined: Dec 2, 2005 Post Count: 1422 Status: Offline Project Badges: 
|
checked my computer logs and i dont see it on any of my xp or vista(1) machines
----------------------------------------i hope these errors do disapear for those who get them though  |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Don't think it would be any consolation, but I have PIIs and PIIIs on Win2000 and Win2003 (respectively) that are crunching fine - but slow. They are the only two systems I have left with this environment (one is a log collector). I do have a few WinXP machines, but they are either P4s or Athlons. Vast majority of what I run is P4/Xeon or AMD high-end on Linux; latest 2.6.2x kernel and DID get about a dozen WUs with errors. That was early on, but otherwise the Betas are ironing it out.
Cheers |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Oh, and 26 pages of PV jail
 |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
That was early on, but otherwise the Betas are ironing it out. Unfortunately, nothing's happened about this problem yet. As well as the hundreds of HCMD2 WUs that have bombed, I've got over 40 HCMD2 betas with error status. That's almost as many as all the betas I've ever successfully completed.  |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I joined WCG a couple of years ago because I wanted to contribute to HCMD ... but could not do so because it didn't work under linux. At the time of HCMD1, WCG used the United Device framework, which was only compatible with Windows. There is a thread about the switch to BOINC. Rather ironic that now HCMD2 is out and even HCMD scientists can't crunch it under linux. Please don't mix up processors and operating systems.... MAXDo, the docking application behind HCMDx, run very well on a wide variety of platforms. Even on AIX machines. HCMD2 work units run well on all my machines but one very old P3 laptop I brought back to life in the sole goal of crunching some more data. When compiling a program, you must always find a balance between optimization (i.e. speed) and compatibility. To simplify, newer processors have newer instructions set (e.g. MMX, SS1, SS2, SSE3 etc.) that can dramatically improve the speed of some kind of operations. So it's easy to understand that some optimizations used at compile time may produce an "illegal instruction" on some older processors. [Edit 1 times, last edit by Former Member at May 17, 2009 3:38:17 PM] |
||
|
|
TXR13
Cruncher Canada Joined: Dec 5, 2005 Post Count: 36 Status: Offline Project Badges:
|
Since most of my crunching machines are P3/Athlon, it is immensely frustrating to see most of the CMD2 / CMD2 Beta units go right out the window with zero runtime and illegal instruction errors.
----------------------------------------It's also a bit frustrating that CMD1 was Windows / UD Agent only, while CMD2 appears to be P4+ only. At least for the time being.  And yes, I'm being partially tongue-in-cheek here. I'm sure the illegal instruction error can eventually be cured, either by recompilation or by outright blocking of the incompatible clients. I'm not trying to point fingers or assign blame at anybody. It simply is as it is. Having looked both here and in the Beta forums, does anybody know if there is a timeline or proposed solution (if any) for fixing the illegal instruction errors on less advanced clients?  [Edit 1 times, last edit by TXR13 at May 17, 2009 6:20:30 PM] |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
At the time of HCMD1, WCG used the United Device framework, which was only compatible with Windows. WCG didn't only use UD. At the time of HCMD1, WCG was running both UD and Boinc. Boinc had been running for over a year on WCG when HCMD1 was launched. The HCMD scientists chose to use the Windows-only platform. When compiling a program, you must always find a balance between optimization (i.e. speed) and compatibility. To simplify, newer processors have newer instructions set (e.g. MMX, SS1, SS2, SSE3 etc.) that can dramatically improve the speed of some kind of operations. So it's easy to understand that some optimizations used at compile time may produce an "illegal instruction" on some older processors. It's not an issue of balance. Compiling for P4-only code is clearly a mistake in a heterogeneous environment of thousands of users. Is it "understandable"? Yes, but it should have been caught during beta tests and clearly nobody should change compile options between beta testing and going live. [Edit 1 times, last edit by Former Member at May 18, 2009 3:56:11 AM] |
||
|
|
|