Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 28 Dec 2007 19:47:41 +0000
From:      Rui Paulo <rpaulo@fnop.net>
To:        Michael Butler <imb@protected-networks.net>
Cc:        Kris Kennaway <kris@FreeBSD.org>, FreeBSD current <freebsd-current@freebsd.org>, Jens Schweikhardt <schweikh@schweikhardt.net>
Subject:   Re: Garbled shutdown message
Message-ID:  <86bq8ar482.wl%rpaulo@fnop.net>
In-Reply-To: <47754D37.4050702@protected-networks.net>
References:  <20071228120152.GB1544@schweikhardt.net> <4774F09A.3050405@FreeBSD.org> <47754D37.4050702@protected-networks.net>

next in thread | previous in thread | raw e-mail | index | archive | help
At Fri, 28 Dec 2007 14:23:35 -0500,
Michael Butler wrote:
> 
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
> 
> Kris Kennaway wrote:
> > Jens Schweikhardt wrote:
> >> If you look at this closely you can find "Syncing (max 60 seconds) for
> >> system process" if you read about every other character. It's always the
> >> third of the three "Waiting" lines. Anyone else seeing this?
> > 
> > You have an SMP system and do not have the PRINTF_BUFR option or
> > whatever it is.  Without it multiple CPUs have no synchronization
> > between printfs and will interleave.
> 
> So the recommended value for PRINTF_BUFR_SIZE is ... ?
> 
> In my (very brief) browse through kern/subr_prf.c, I didn't see any
> overflow protection on the stack-allocated buffer ... which makes me
> rather nervous.
> 
> Interleaved text is *way* better than clobbered kernel stack ..

	pca.n_bufr = sizeof(bufr);
	pca.remain = sizeof(bufr);

I think this is the "overflow protection".

--
Rui Paulo



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?86bq8ar482.wl%rpaulo>