Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 5 Nov 2003 12:41:53 -0500 (EST)
From:      Garrett Wollman <wollman@khavrinen.lcs.mit.edu>
To:        Kim Needham <kim.needham@rocksoft.com>
Cc:        freebsd-security@freebsd.org
Subject:   Re: Veractiy and FreeBSD
Message-ID:  <200311051741.hA5HfrBJ074765@khavrinen.lcs.mit.edu>
In-Reply-To: <5.2.0.9.0.20031105114806.00b34b98@mail.chariot.net.au>
References:  <20031104180932.GA70958@rot13.obsecurity.org> <001901c3a2de$7ab72c10$6855cd80@dhcp.hwi.buffalo.edu> <4529.141.149.242.190.1067993236.squirrel@mail.hwi.buffalo. edu> <5.2.0.9.0.20031105114806.00b34b98@mail.chariot.net.au>

index | next in thread | previous in thread | raw e-mail

<<On Wed, 05 Nov 2003 11:58:21 +1030, Kim Needham <kim.needham@rocksoft.com> said:

> This is caused by a check that determines if the /proc filesystem is on
> the same device as the file being read to prevent reading in files of 
> infinite length.

The correct way to perform this check is to call fstatfs() on the file
descriptor and examine f_fstypename[] to see what it is.  You probably
shouldn't look explicitly at the type name; use getvfsbyname() on the
type name; if (vfc_flags & VFCF_SYNTHETIC) then the contents of the
file system are generated by the kernel.

-GAWollman


help

Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200311051741.hA5HfrBJ074765>