Re: Checking if processor honours the WP ... Stops the computer

Kevin M. Bealer (kmb203@psu.edu)
Wed, 02 Jul 1997 02:43:17 -0400

Martin Mares wrote:
>Hi,
>
>> I have the same problem, in 2.1.43, 2.1.42 and before. I traced it
>> down to the inline assembly block in arch/i386/mm/init.c.
>>
>> I don't know -why- it is dying, or what the WP is, but if you comment
>> out the inline assembly and force the "test" to "no" you can get it to
>> run again.
>
> (1) What CPU do you use?

486 DX, 33 MHz, in an IBM. It's an intel, even has the littl' sticker
on the box. Note that I have been compiling kernels since 1.2.13/1.3.?
and never had a problem with WP bit before. What is the WP bit?

> (2) Does the following patch help?
>

It took a while to get back to you because it takes a while to make a
kernel on this system (also I had a 70 MB download that I wanted to
finish..)

Sorry, it doesn't seem to have an effect, the system still hangs w/o
giving a yes or no. However, my removing the inline assembly and
forcing a "No" seems to work.

What exactly does your patch attempt to do?

> Have a nice day
> Martin
>
>--- init.c.mj Wed Jun 25 10:47:10 1997
>+++ init.c Wed Jun 25 10:48:18 1997
>@@ -348,7 +348,7 @@
> local_flush_tlb();
> current->mm->mmap->vm_start += PAGE_SIZE;
> __asm__ __volatile__(
>- "movb %0,%1 ; movb %1,%0"
>+ "movb %0,%1 ; movb %1,%0 ; jmp 1f; 1:"
> :"=m" (*(char *) __va(0)),
> "=q" (tmp_reg)
> :/* no inputs */

--kmb203@psu.edu----------------Debian/GNU--1.3---Linux--2.0.30---
"The C Programming Language -- a language that combines all the
elegance and power of assembly language with all the readability
and maintainability of assembly language." (version 2.0 of this
signature, thanks Pete :)