On 4/14/10 9:35 PM, Myles Watson wrote: > > > On Wed, Apr 14, 2010 at 1:22 PM, Stefan Reinauer > <[email protected] <mailto:[email protected]>> wrote: > > On 4/14/10 6:44 PM, Myles Watson wrote: >> >> Index: src/northbridge/amd/gx2/chipsetinit.c >> =================================================================== >> --- src/northbridge/amd/gx2/chipsetinit.c (revision 5425) >> +++ src/northbridge/amd/gx2/chipsetinit.c (working copy) >> @@ -275,7 +275,7 @@ >> if ((msr.lo&0xff) == 0x11) >> return; >> >> - totalmem = sizeram() << 20 - 1; >> + totalmem = (sizeram() << 20) - 1; >> >> I agree that it looks right, but it changes the answer, since >> subtraction has a higher precedence than shift. >> >> I don't have a board to test which is right. Did you check v3? > Nor do I... > > v3 never supported gx2. > > > I guess this should be the patch, then: > > - totalmem = sizeram() << 20 - 1; > + totalmem = sizeram() << (20 - 1); If we can't decide on the right fix we should leave the warning there as it keeps the question open until someone can answer it.
-- coreboot mailing list: [email protected] http://www.coreboot.org/mailman/listinfo/coreboot
