[vox-tech] core files

[Peter Jay Salzman]

On Sun 15 Feb 04,  7:02 PM, Michael Wenk <wenk@praxis.homedns.org> said:
> On Sunday 15 February 2004 05:57 pm, Peter Jay Salzman wrote:
> > On Sun 15 Feb 04,  5:53 PM, mrp <mrp@sonic.net> said:
> > > On Sun, Feb 15, 2004 at 05:38:47PM -0800, Peter Jay Salzman wrote:
> > > > I've noticed some core files have a number appended to them, like:
> > > >
> > > >    core.2342
> > > >
> > > > The number is most likely a pid, but I'm curious how the core file got
> > > > named this way.  Is there a libc function that sets the name of a
> > > > possible future core file?
> > >
> > > I think it's probably a function of the default signal handler for
> > > the SIGABORT signal.
> >
> > i've used abort() tonight in one of my programs and it just gave a file
> > named "core".
> >
> > > I don't know why it would use "core" sometimes and
> > > "core.<pid>" other times.. maybe if there is already a core file it uses
> > > the second form instead to avoid overwriting an existing core file.
> >
> > can't be this.  i've had MANY core files written in my life.  ;-)  core
> > files are just overwritten, not renamed (permission granting).
> >
> > i've noticed that whenever opera dumps core, the corefile always has the
> > number in its name.  that's kind of why i figured it might be a libc
> > thing that you can control.
> >
> > > You might also want to look at the setrlimit(2) man page.
> >
> > i'll look into it.  thanks!
> >
> > pete
> 
> Did a bit of googling, cuz I noticed that as well and found the following 
> site: 
> http://www.sinz.org/Michael.Sinz/Linux/core.html
> 
> M

nice!  that got bookmarked.

talk about noticing, i've also noticed that somewhere along the way, the
'file' utility stopped reporting what signal prompted the core file to
be written.  at least on my system (debian testing).

core: ELF 32-bit LSB core file Intel 80386, version 1 (SYSV),
SVR4-style, SVR4-style, from 'hello'

not terribly important, but i liked the old way better.

pete

-- 
Make everything as simple as possible, but no simpler.  -- Albert Einstein
GPG Instructions: http://www.dirac.org/linux/gpg
GPG Fingerprint: B9F1 6CF3 47C4 7CD8 D33E 70A9 A3B9 1945 67EA 951D