Lahey on Fedora

I'm getting mixed signals from Lahey sales as to whether or not their
v6.2 LF95 compiler will run on Fedora (their web site says it will run
on RHEL v3 WS).  Anyone have any experience running the v6.2 Lahey
compiler on Fedora?

TIA
Jim

Jim S wrote:
> I'm getting mixed signals from Lahey sales as to whether or not their
> v6.2 LF95 compiler will run on Fedora (their web site says it will run
> on RHEL v3 WS).  Anyone have any experience running the v6.2 Lahey
> compiler on Fedora?

Why not download a trial version and see? That should let you know if it wor
ks for your
set-up.

cheers,

paulv

p.s. FWIW, I've got v6.2 on RHEL v3 WS and it works just fine. Of the f95 co
mpilers I have
for linux, the Lahey one has given me the least problems (actually, I can't 
recall any).

On Mon, 18 Oct 2004, Jim S wrote:

>
>
> I'm getting mixed signals from Lahey sales as to whether or not their
> v6.2 LF95 compiler will run on Fedora (their web site says it will run
> on RHEL v3 WS).  Anyone have any experience running the v6.2 Lahey
> compiler on Fedora?
>
> TIA
> Jim
>

I use Lahey 6.2 with Gentoo linux running glibc 2.3.3 and NPTL (Native
Posix Threads Library) and it works just fine so I suspect you wouldn't
have any trouble with Fedora.

Cheers,
Jason

Jim S wrote:
> I'm getting mixed signals from Lahey sales as to whether or not their
> v6.2 LF95 compiler will run on Fedora (their web site says it will run
> on RHEL v3 WS).  Anyone have any experience running the v6.2 Lahey
> compiler on Fedora?

I use Lahey v6.2 on x86 Fedora Core 2, but I found that it started dying
after my last kernel upgrade (from 2.6.7-1.494.2.2 to 2.6.8-1.521).
So I kept the old kernel around for when I want to use Lahey.  (I also
use Intel 8.1 and 7.1.)

Regards,
--
Bil Kleb, Hampton, Virginia

Bil Kleb wrote:
>
>
> I use Lahey v6.2 on x86 Fedora Core 2, but I found that it started dying
> after my last kernel upgrade (from 2.6.7-1.494.2.2 to 2.6.8-1.521).
> So I kept the old kernel around for when I want to use Lahey.  (I also
> use Intel 8.1 and 7.1.)
>
Hello,
Since I encountered a strange problem with this combination recently, I
would like to learn more about your case:
Here it helped to increase the user stack limit (on some page related to
the error message "Insufficient system swap space" it was suggested to
increase the data size limit.)
Best wishes
Alois

Alois Steindl wrote:
>
> Since I encountered a strange problem with this combination recently, I
> would like to learn more about your case:

$ uname -r
2.6.8-1.521

$ lf95 hello.f90
f95: /usr/local/pkgs/Lahey-6.2/bin/../bin/jwd_fort: SIGSEGV signal received.

$ ulimit
unlimited

$ lf95 hello.f90
f95: /usr/local/pkgs/Lahey-6.2/bin/../bin/jwd_fort: SIGSEGV signal received.

Does that help?

Regards,
--
Bil Kleb, Hampton, Virginia

Bil Kleb wrote:
> Alois Steindl wrote:
> 
>
>
> $ uname -r
> 2.6.8-1.521
>
> $ lf95 hello.f90
> f95: /usr/local/pkgs/Lahey-6.2/bin/../bin/jwd_fort: SIGSEGV signal
> received.
>
> $ ulimit
> unlimited
>
> $ lf95 hello.f90
> f95: /usr/local/pkgs/Lahey-6.2/bin/../bin/jwd_fort: SIGSEGV signal
> received.
>
> Does that help?

What does a "ulimit -a" produce for output?

cheers,

paulv

Paul Van Delst wrote:
>
> What does a "ulimit -a" produce for output?

$ ulimit -a
core file size        (blocks, -c) 0
data seg size         (kbytes, -d) unlimited
file size             (blocks, -f) unlimited
max locked memory     (kbytes, -l) 32
max memory size       (kbytes, -m) unlimited
open files                    (-n) 1024
pipe size          (512 bytes, -p) 8
stack size            (kbytes, -s) 10240
cpu time             (seconds, -t) unlimited
max user processes            (-u) 16378
virtual memory        (kbytes, -v) unlimited

--
Bil Kleb, Hampton, Virginia

Bil Kleb <Bil.Kleb@NASA.Gov> writes:

> Paul Van Delst wrote: 
...
> stack size            (kbytes, -s) 10240

That one looks suspicious to me.  I don't know the particular
case, but I'd try removing or majorly increasing that limit
(using "ulimit -s unlimited" or "ulimit -s bigger-number").

--
Richard Maine                       |  Good judgment comes from experience;
email: my first.last at org.domain  |  experience comes from bad judgment.
org: nasa, domain: gov              |        -- Mark Twain

Bil Kleb wrote:
> Paul Van Delst wrote:
> 
>
>
> $ ulimit -a
> core file size        (blocks, -c) 0
> data seg size         (kbytes, -d) unlimited
> file size             (blocks, -f) unlimited
> max locked memory     (kbytes, -l) 32
> max memory size       (kbytes, -m) unlimited
> open files                    (-n) 1024
> pipe size          (512 bytes, -p) 8
> stack size            (kbytes, -s) 10240
> cpu time             (seconds, -t) unlimited
> max user processes            (-u) 16378
> virtual memory        (kbytes, -v) unlimited

Aha! The important number here is
stack size            (kbytes, -s) 10240

Can you increase the stack limit by doing something like:
ulimit -s unlimited

and then recompile with Lahey?

cheers,

paulv

p.s. I just found out that I'm restricted somehow in resetting limits once I
've changed
them in a terminal! Weird. In a single session I can only *decrease* the val
ue. Is that
normal? e.g.

lnx:paulv : ulimit -s 10240
lnx:paulv : ulimit -s unlimited
bash: ulimit: stack size: cannot modify limit: Operation not permitted
lnx:paulv : ulimit -s 5120
lnx:paulv : ulimit -s 10240
bash: ulimit: stack size: cannot modify limit: Operation not permitted

But if I open another window/terminal, I can push the listed ulimit stack va
lue back to
unlimited:

lnx:paulv : ulimit -s unlimited
lnx:paulv : ulimit -a
core file size        (blocks, -c) 0
data seg size         (kbytes, -d) unlimited
file size             (blocks, -f) unlimited
max locked memory     (kbytes, -l) 4
max memory size       (kbytes, -m) unlimited
open files                    (-n) 1024
pipe size          (512 bytes, -p) 8
stack size            (kbytes, -s) unlimited
cpu time             (seconds, -t) unlimited
max user processes            (-u) 7168
virtual memory        (kbytes, -v) unlimited
??
>