Experimental Physics and
Industrial Control System

"Jeff Hill" <[email protected]> · Tue, 26 Apr 2011 10:01:19 -0600

Hi Pete,

Sorry, your original message was lost in my inbox. To get this fixed I need
a stack trace for all of the threads. If there is a SIGBUS then a core file
can be generated. If you type "ulimit -c unlimited" into bash then core file
generation will be enabled. Once you have the core file you can get a
complete dump of all threads as follows. Send that output and I will have a
look. 

gdb <executable name and path> <core file name and path>
thread apply all bt full

This, will more than likely, be quite a large output, but it should
hopefully be comprehensive enough to resolve the issue. 

You might send also the version of EPICS base that the gateway linked
against.

Thanks for reporting the bug,

Jeff
______________________________________________________
Jeffrey O. Hill           Email        [email protected]
LANL MS H820              Voice        505 665 1831
Los Alamos NM 87545 USA   FAX          505 665 5107

Message content: TSPA

With sufficient thrust, pigs fly just fine. However, this is
not necessarily a good idea. It is hard to be sure where they
are going to land, and it could be dangerous sitting under them
as they fly overhead. -- RFC 1925

> -----Original Message-----
> From: [email protected] [mailto:[email protected]]
> On Behalf Of Martin L. Smith
> Sent: Tuesday, April 26, 2011 7:07 AM
> To: [email protected]
> Cc: [email protected]
> Subject: Re: gateway aborted by SIGBUS
> 
> Hi Pete,
> 
> I have not seen this in quite a while at APS where I have about 40
> gateways
> running that are read/write; although I'm using version 2.0.3.0
> 
> If your gateways are running under Linux my understanding is that you may
> want
> to increase the stack size limit. I have a comment in my starter file:
> 
> # Limit stacksize for Linux to increase number of threads
> # 1024 is 204 threads (default)
> #  4096 is 512 threads
> ulimit -s 4096
> 
> As you can see I use 4096 although I'm not sure that this will help
> especially
> if you are not running on Linux.
> If this is reproducible the gateway documentation mentions that there
> is a command line debug option which may help but I expect that it would
> slow
> things down a bit.
> 
> -debug <value>
> Enter a value between 0-100. 50 gives lots of information, 1 gives a small
> amount. For developers.
> 
> Marty
> 
> [email protected] wrote:
> > Recently we have had three read/write gateways (version 2.0.4.0) crash
> at different times with the following error appearing in the gateway log.
> >
> > PV Gateway Aborted (SIGBUS)
> >
> > The crashes have all occurred when doing a caput thorough the gateway.
> >
> > The server option on the gateway causes an immediate restart of the
> gateway however the restart doesn't seem to cure the problem with the
> particular PV's we are writing to although read access to all PV's is
> fine. Manually stopping and restarting the gateway get this back to a
> working state so presumably the time taken to do the restart manually is
> significant.
> >
> > So a couple of questions:
> >
> > 1) Has anyone else seen this?
> > 2) What's the best way to debug this? Presumably as its SIGBUS it has
> something to do with stack size or byte alignment or similar. So
> presumably I need to get a core dump. Is there any additional debugging in
> the gateway I can enable which may help?
> >
> > Pete Leicester
> > Diamond Light Source
> >
> >
> >
> >
> >
> >

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System