Subject: Re: firefox-bin-java anyone?
To: Todd Vierling <firstname.lastname@example.org>
From: Steven M. Bellovin <email@example.com>
Date: 09/29/2005 12:59:35
In message <Pine.NEB.firstname.lastname@example.org>, Todd Vierling wri
>On Thu, 29 Sep 2005, Christos Zoulas wrote:
>> >Speaking of which, this should become the default in any kernel with
>> >COMPAT_LINUX enabled (or just turn on the compat bits always!). It's long
>> >past time to stop requiring a special mount option just to get 15-20% of an
>> >emulated platform's binaries to work.
>> You are suggesting that we add:
>> char buf;
>> if (sysctlbyname("emul.linux.kern.ostype",
>> NULL, 0, buf, sizeof(buf)) != -1)
>> flags |= PROCFSMNT_LINUXCOMPAT;
>> to mount_procfs?
>Yes, something like that. In fact, I like that quite a lot as a first step.
>(I'd personally prefer to see the check added to the kernel directly and rip
>out the mount_procfs bits altogether. Just using COMPAT_LINUX as a switch
>there isn't very useful, since it leaves the compat_linux.o LKM in the
What is the disadvantage of that flag always being enabled? Does it
need to be an option?
--Steven M. Bellovin, http://www.cs.columbia.edu/~smb