Re: Kernel 2.2.11 crash...

Andrea Arcangeli (andrea@suse.de)
Tue, 5 Oct 1999 17:21:21 +0200 (CEST)


On Tue, 5 Oct 1999, Alan Cox wrote:

>Those modules allow vmware itself to do a lot of clever unsafe things
>behind the kernels back. So we don't have a good way to track what

I had to gone into such modules at once and I could mainly find code to
port kernel data structures, GFP,FP wrappers to userspace. It seems to me
that the vmware modules are there only to write kernel code into
userspace. So they avoid completly any module linking problem as you can
recompile the wrapper against the new symbols. So basically they get the
kernel data into userspace and they then manage it in userspace. With only
the kernel modules you can't do anything to change vmware or to fix any
vmware bugs or to port vmware to the new kernel code.

I even think that rotating the PG_* defines (in mm.h) and recompiling the
kernel will break vmware too. I can't try this out though as I don't have
a vmware licence.

While I like very much things like vmware, if a kernel breaks someway with
a vmware module previously or currently loaded, then the report _must_ not
be addresses to l-k. If you think it's not vmware the problem, then try to
reproduce with vmware _never_ loaded. Only if you'll be able to reproduce
then send to l-k.

Andrea

-
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/