Re: [PATCH] [SECURITY] suid procs exec'd with bad 0,1,2 fds

Theodore Y. Ts'o (tytso@MIT.EDU)
Tue, 4 Aug 1998 17:35:51 -0400


From: Perry Harrington <pedward@sun4.apsoft.com>
Date: Tue, 4 Aug 1998 13:42:56 -0700 (PDT)

The trampoline thing has already been discussed. The other use,
interpreters, is odd. You're telling me that people have written
interpreters that place code to actually execute on the stack? I've
read a bit about stack machines, and wrote one, the implementations
I've seen would never do such a machine dependent and unorthodox
thing.

You'll often see this sort of thing done for speed reasons; for example,
a instruction-level emulator (say to emulate a 68k for a Mac-emulator)
might very well use just-in-time compilation tricks to translate 68k
code into i386, and then execute the code on the stack. This can make a
very big difference for tight CPU loops (like string copy, etc.).

It would also be very reasonable for a Java JIT compiler/interpreter to
do the same thing. Yes, it's machine dependent, but so is a C
compiler!

- Ted

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.altern.org/andrebalsa/doc/lkml-faq.html