Jachete@... writes:
> Thanks for the reply. My real question is how reliable is the 64bit
> version of sbcl? And if it is not yet stable, when will it be ?
It's reliable for me.
> I am just trying to evaluate the usefulness of continuing on a 64
> bit OS (UBUNTU Linux - Debian based), if the applications don't
> work...
>
> On the CL front:
> sbcl: behaves oddly, to say the least,
It would be good if you could be rather less unspecific here. I
understand that this won't help you, but it's been running for me
without major faults for the last couple of months.
> I believe that my code is ok, it is very simple and there is no
> reason that it should be able to load sometimes and not others, nor
> should execution of the same function+args produce different results
> since the functions do not maintain state...
There are plenty of reasons why this can happen: the first reason I
would suspect is that you have bad RAM or a bad motherboard. For what
it's worth, your code ran fine multiple times on my amd64. That said,
I'm certainly less likely to spend a long time analysing code that
breaks with (speed 3) (safety 0) (debug 0): by doing that, you've
removed all the safety nets that people have spent about 25 years
building under you, and all the ways that you can diagnose the problem
once you do fall.
> So let me know if there is anything I can do to help!
Perhaps before submitting another report you could read
<http://www.chiark.greenend.org.uk/~sgtatham/bugs.html>, to help you
focus your report. As I have said, the problem with getting 11 nested
errors after a segfault is likely to be fixed "soon" (in a week or
so); the cause of the initial segfault is by no means clear, but I
wouldn't treat your hardware as above suspicion.
Cheers,
Christophe
|