hpoj-devel

Re: [hpoj-devel] Re: [hpoj-announce] hpoj-0.7 is available

[<pa...@rc...>]

Joe Piolunek wrote:
> I got the official 0.7 release from sourceforge today and installed
> it. There  weren't any problems during the build or installation.
> Printing using  ptal-printd works well. It's certainly a lot easier to
> set up.  Congratulations. 

Hi, Joe.  Thanks for the installation report; I'm glad you like it.  Also
thank you for your contributions to 0.7 as well as your constant willingness
to try out bleeding-edge code and give valuable feedback.

I'm sorry I wasn't able to fix the "xojpanel no-start" problem for you in this
release.  I worked on it some last night before starting the release process
and found that the failure code path starts at line 1501 in link_receive()
in ieee12844.c.  It seems to think that the peripheral is sending a command
packet with an unknown socket ID pair as part of the command, but I don't yet
know at what point during the negotiation sequence it happens.  It's obviously
a race condition, because I discovered that it doesn't fail if I'm sitting
at the computer waiting for it to happen, but as soon as I walk away and
let the screensaver (kforest) kick in, it fails left and right, which would
explain the high failure rate I previously reported.

David

Re: [hpoj-devel] Re: [hpoj-announce] hpoj-0.7 is available

[<pa...@rc...>]

Dag Nygren wrote:
> I am the one that had problems with SMP and printing
> hanging....
> 
> Downloaded 0.7  and tried again:
> 
> Didn't check scanning yet but I still see ptal-printd
> hanging and becoming unkillable (hang in the driver
> I presume).
> The strange thing is that the first printing usually works....
> 
> Is there anything I can do to help you (or me) debug this
> problem ?

Hi, Dag.  There are "debug=" options that can be passed to the kernel-mode
drivers (see the INSTALL file).  To start out, you could try enabling
debug output for each one individually and then both together, but chances
are there will be so much output that syslog can't keep up.  But it's a
start, anyway.

Also, could you send me print samples (in PostScript format) that cause this
problem for you, or URLs of webpages I can try printing to duplicate the
problem?  I have an SMP box at work, but I wasn't able to lock it up from
printing RedHat test pages.

David