On Mon, Oct 16, 2000 at 03:02:35PM +0200, Gisle S{lensminde wrote:
>
> When compiling the tests in the linux/crypto/testing directory, I
> got a compile error for the tests. The symbol 'schedule' was missing
> for several of the ciphers (maybe all), so I got a linker error.
> The kernel itself on the other hand, linked fine.
>
Argh.. it is the
if (current->need_resched)
schedule();
in gen-cbc.h and gen-ecb.h. Added in 2.2.17.8. We need this to avoid
the loopback filesystem to end up on top of the latency-tests. As a
start we need a stub for schedule().
The bad news is "current". It is a magic variable that lies at the
start of the stack-pointer, so on i386 it is derived from %esp. So
current->need_resched is peeking at places it shouldn't :-(.
The good news is that accessing "current" won't cause a segfault when
the stack starts on a page boundary. And since our stub for
schedule() is a nop, it will run correctly no matter what
current->need_resched turns out to be :-)
So the following patch should fix the problem:
--- i17xx/crypto/testing/stubs.c Mon Oct 16 15:20:25 2000
+++ i17x.nomodule/crypto/testing/stubs.c Mon Oct 16 15:17:04 2000
@@ -36,3 +36,9 @@
return (int) name;
}
+asmlinkage void schedule(void)
+{
+}
+
+
+
astor
--
Alexander Kjeldaas Mail: astor@xxxxxxx
finger astor@xxxxxxxxxxxxxxxxx for OpenPGP key.
Linux-crypto: cryptography in and on the Linux system
Archive: http://mail.nl.linux.org/linux-crypto/
