pcp
[Top] [All Lists]

Re: pmcd dumping core - multiple issues

To: "Frank Ch. Eigler" <fche@xxxxxxxxxx>
Subject: Re: pmcd dumping core - multiple issues
From: Ken McDonell <kenj@xxxxxxxxxxxxxxxx>
Date: Mon, 29 Jul 2013 11:27:11 +1000
Cc: PCP Mailing List <pcp@xxxxxxxxxxx>
Delivered-to: pcp@xxxxxxxxxxx
In-reply-to: <20130729012330.GG4664@xxxxxxxxxx>
References: <51F44856.3090308@xxxxxxxxxxxxxxxx> <y0md2q2nbv1.fsf@xxxxxxxx> <51F5AC32.4040001@xxxxxxxxxxxxxxxx> <20130729012330.GG4664@xxxxxxxxxx>
User-agent: Mozilla/5.0 (X11; Linux i686; rv:17.0) Gecko/20130623 Thunderbird/17.0.7
On 29/07/13 11:23, Frank Ch. Eigler wrote:
Hi -

Thanks for that Frank.  I'll make sure to hide the facility so it is
only enabled by explicit developer action for those of us outside the
Fedora/RHEL ecosphere.

By the way, the glibc backtrace() function, as used in
src/libpcp/src/lock.c, is a linuxy substitute for the
trace_back_stack() in src/pm{cd,ie}/src/pm{cd,ie}.c.  Perchance move
that stuff to some utility module in the library, and conditionalize
sigBAD forwarding to it based upon a command line option?

(as of last night) I already have the changes to use backtrace() ... a command line option (as opposed to compile-time I was considering) is a good idea.

Thanks, Frank.

<Prev in Thread] Current Thread [Next in Thread>