Re: pre-2.3.1-1 fails compile (smbfs)

Linus Torvalds (torvalds@transmeta.com)
Tue, 11 May 1999 20:15:46 -0700 (PDT)


On Tue, 11 May 1999, Pete Clements wrote:
>
> gcc -D__KERNEL__ -I/hdb3/usr/src/linux-2.3.1/include -Wall -Wstrict-prototypes -O2 -fomit-frame-pointer -pipe -fno-strength-reduce -m486 -malign-loops=2 -malign-jumps=2 -malign-functions=2 -DCPU=586 -c -o inode.o inode.c
> inode.c:71: `MUTEX' undeclared here (not in a function)

A lot of the wait-queue and semaphore initializations have to be changed.
It's not a big piece of work, but it's "details". Most of them are caught
by the compiler, but not all (for example, it used to be ok to initialize
the wait-queue heads to 0 - so there are places that just depend on the
normal BSS initialization or do "memset(xx, 0, size)").

The things that aren't caught statically by the compiler are usually
caught reasonably gracefully at run-time (ie if you fix smbfs to compile
without warnings it may still not work, but it will probably give you a
well-behaved oops telling you where it is using an uninitialized wait-
queue).

And yes, this is why this was done in 2.3.x instead of just doing the
cleanup (because cleanup it is) in 2.2.x...

I'll happily accept patches - you can look at the pre-patch to see what
kinds of changes are necessary,

Linus

-
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.tux.org/lkml/