[Lguest] [PATCH 3/16] read/write_crX, clts and wbinvd for 64-bit paravirt

Glauber de Oliveira Costa gcosta at redhat.com
Fri Nov 2 03:13:55 EST 2007


-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Keir Fraser escreveu:
> On 1/11/07 15:30, "Jeremy Fitzhardinge" <jeremy at goop.org> wrote:
> 
>> Glauber de Oliveira Costa wrote:
>>> I in fact have seen bugs with mixed reads and writes to the same cr,
>>> (cr4), but adding the volatile
>>> flag to the read function seemed to fix it.
>> Well, volatile will make a read be repeated rather than caching the
>> previous value, but it has no effect on ordering.
> 
> volatile prevents the asm from being 'moved significantly', according to the
> gcc manual. I take that to mean that reordering is not allowed.
> 
According to a gcc developer to whom I asked this question, volatile
prevents the code
to be removed, but does not prevent it to be moved (pun indented). In
practice, it should force
a re-read, but not influence the ordering decisions from the compiler.
Besides , 'significantly'
sounds like a significantly unprecise word, whose specific meaning may
be implementation dependant.

So I agree that adding a memory location reference is probably the best
alternative.

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.7 (GNU/Linux)
Comment: Using GnuPG with Remi - http://enigmail.mozdev.org

iD8DBQFHKftDjYI8LaFUWXMRAiLTAKDqf/M8umNYw6u7r9ONozTEUVy8SwCgygma
jWNKQmxmLpyPxr00KbQy9Vg=
=JM4K
-----END PGP SIGNATURE-----



More information about the Lguest mailing list