- Previous message: mouss: "Re: Command output: Invalid user"
- Maybe in reply to: Duane Hill: "fatal: accept connection: Invalid argument"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] [ attachment ]
|
|
|
From: Wietse Venema (no email)
Date: Tue Feb 27 2007 - 19:10:38 EST
Victor Duchovni:
> On Wed, Feb 28, 2007 at 09:42:50AM +1100, Andrew McNamara wrote:
>
> > >>> I could try dropping to a 32 bit kernel, however, the other server I have
> > >>> is running a 32 bit kernel and gets a few of these same errors.
> > >>
> > >> Okay, that suggests it isn't a 32 bit vs 64 bit issue.
> > >>
> > >> Is the other server SMP also?
> > >
> > >Yes.
> >
> > Given the difficulty trying a non-SMP kernel on the primary machine, maybe
> > one could be tried on this secondary machine?
>
> Has anyone checked the FreeBSD source for EINVAL return codes from
> accept() on a TCP socket? If this is a feature we can stop the wild
> speculation. I vote this issue is moved to a suitable FreeBSD list until
> it becomes clear why accept(2) sets errno=EINVAL.
I have FreeBSD 6.2 installed on a test machine, and I can look at it
tomorrow. Of course if the bug is SMP related (not confirmed) then
all bets would be off as far as error returns codes.
Wietse
|
|
|