Re: [Libreboot-dev] Why do you carry 0042-cpu-amd-car-Initialize-entire-CAR-space-instead-of-o.patch ?

[Francis Rowe]

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

ask tpearson on freenode. I'll be rebasing those patches soon anyway,
with the newest patcheset

Op 22/12/15 om 22:19 schreef address@hidden:
> Hello, and thank you very much for your work.
> 
> I'm looking at the code and I see that libreboot is patching
> coreboot with a change that was abandoned in coreboot:
> 
> resources/libreboot/patch/kgpe-d16/0042-cpu-amd-car-Initialize-entire-
CAR-space-instead-of-o.patch
>
>  http://review.coreboot.org/#/c/11974/
> 
> I'm not good at assembler but from what I've read I think the 
> original code was right and the patch introduces a bug, because rep
> lodsl or rep stosl copy 4 bytes, so we only have to repeat this
> CacheSize/4 times (= CacheSize / 2^2 = CacheSize >> 2).
> 
> I can't test now but I'll reply if/ when I can test it. But I
> thought I'd ask because I may be missing something.
> 
> Thanks.
> 
> Xavi
> 
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.22 (GNU/Linux)

iQEcBAEBAgAGBQJWekNMAAoJEP9Ft0z50c+U0xgH/ArMf1PlonmX+M4JdkLUN39i
A6OVxmRFE2XYT3Spgw3iqbGp/kvMcn3EAialgOHiFy4/zlTZAE0lN8pNS3y2K88f
L1UiZh8u54IIZmFWEBK1hWzlDJnOXku5lbhsLQ9WbRpyRT6N4LQ2V47NA3qKByST
QVmI3heMW53GDtTtzxRsPovnea3YtM7q0Hq60uoGGR4TYLgY6mcUK71z5IHFq+Ox
MY1YAhxzTx7EfmxArWZIzHScfNSbc60l7QG/wk5RzbIuwGn2IVGfdv3PmJ9Jlid5
WIX9vBSSplBVGc6fwDfIlnqj9QZCaobGyICuYmPki9jfy+20VUX7UOq07U0o6b4=
=tV95
-----END PGP SIGNATURE-----