[Lcdproc] LC16M imon + LCD 0.5.1

[Peter Marschall]

Hi,

On Tuesday, 31. October 2006 04:54, Matthew Hall wrote:
> gettimeofday({1162267079, 507854}, NULL) = 0
> select(1024, [3 5], NULL, NULL, {0, 0}) = 0 (Timeout)
> nanosleep({0, 26886000}, NULL)          = 0
> gettimeofday({1162267079, 539921}, NULL) = 0
> select(1024, [3 5], NULL, NULL, {0, 0}) = 0 (Timeout)
> write(4, "                    7 57 PM     ", 32) = -1 ENODEV (No such
> device)
> nanosleep({0, 31250000}, NULL)          = 0
> gettimeofday({1162267079, 575910}, NULL) = 0
> select(1024, [3 5], NULL, NULL, {0, 0}) = 0 (Timeout)
> nanosleep({0, 31250000}, NULL)          = 0
> ...
>
> While if it was working, it would = 32, not -1 ENODEV.

Explainable: write() returns the number of chars successfully 
written, or -1 in case of error.

> So what's happening here? Killing and restarting LCDd makes
> it all better. Where does my device go? It's really, truly
> there, and works the majority of the time.

Unfortunately the strace output does not help me very much.
It makes it only more mysterious:
nanosleep does neither occur in the imon driver nor in the
server core. So I guess it is part of some included library.

Can you try compiling LCDproc using the --enable-debug=yes 
configure switch and running LCDd in gdb?

Peter

-- 
Peter Marschall
peter@adpm.de