HylaFAX The world's most advanced open source fax server

[Date Prev][Date Next][Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: faxq hanging - some progress



I have been having this same problem. I spent a couple of hours last
night tracing it and yes, it is definately in the page chop code
that it is getting stuck I found what could be an infinite loop in the
code, but it gets stuck inside a loop within that loop. The fact that
the process has a delete file open leads me to believe that there
may be a pointer problem in here somwhere. I'll have to spend some
more time on it. Just for referece, this is with the beta release on
a Red Hat 6.0 system.

Jay Davis


Neal Holtz wrote:

>
> There have been a few reports of faxq hanging, using 100% of CPU
> and doing nothing useful.  I encountered the same problem,
> with 4.1beta2 on a Caldera OpenLinux 1.2 system.  I compiled
> the binaries with gcc version 2.7.2.3.
>
> I traced the problem to logic that is looking for trailing whitespace
> in pages.  Attach gdb to the running faxq process and CTRL-C it a
> few times, you get the following backtrace:
>
> > 0x805a340 in G3Decoder::decodeRow (this=0xbffff280, scanline=0x0, lastx=1728) at G3Decoder.c++:228
> > 228             EXPAND1D(Nop1d);
> > (gdb) bt
> > #0  0x805a340 in G3Decoder::decodeRow (this=0xbffff280, scanline=0x0, lastx=1728) at G3Decoder.c++:228
> > #1  0x804fe7c in MemoryDecoder::scanPageForBlanks (this=0xbffff280, fillorder=2, params=@0xbffff3d8)
> >     at faxQueueApp.c++:858
> > #2  0x804ff60 in faxQueueApp::preparePageChop (this=0x80a8338, req=@0x80ae540, tif=0x80ae6a8,
> >     params=@0xbffff3d8, pagehandling=@0x80ae5f8) at faxQueueApp.c++:881
> > #3  0x804f9ed in faxQueueApp::preparePageHandling (this=0x80a8338, req=@0x80ae540, info=@0x80aead8,
> >     dci=@0x80a1834, emsg=@0x80ae5e8) at faxQueueApp.c++:703
> > #4  0x804f69f in faxQueueApp::prepareJob (this=0x80a8338, job=@0x80ae3a0, req=@0x80ae540,
> >     info=@0x80aead8, dci=@0x80a1834) at faxQueueApp.c++:578
> > ... etc
>
> Now, that EXPAND1D macro is large, and I have not looked for the real
> cause
> in there (and I may not), but it at least allowed me to configure a work
> around.  Put the following in  /var/spool/fax/etc/config:
>
> PageChop:               none
>
> Strangely - this hanging only happens when I ask for a cover page, yet
> it did
> not appear to be when processing the cover page, but I could be wrong
> about that.
>
> Hope this helps.
>
> --
> Neal Holtz                             http://www.docuweb.ca/~nholtz
> Dept. of Civil and Environmental Engineering,   Carleton University,
> Ottawa, Ontario, Canada K1S 5B6.                   nholtz@docuweb.ca




Project hosted by iFAX Solutions