On Fri, Aug 09, 2013 at 09:12:37AM +1000, Michael Neuling wrote: > Anton Blanchard <an...@samba.org> wrote: > > > > > Hi, > > > > > The CONFIG_VIRTUALIZATION disabling should be done in the Kconfig not > > > here. > > > > > > I'm not that keen on another defconfig. benh is already talking about > > > having a powernv defconfig. I'm worried we are going to fragment the > > > defconfigs. If you want something special like LE, then change the > > > default one. > > > > I agree we don't want machine specific defconfigs, but I think it makes > > sense to have ones that cover the key options that conflict. I'm > > thinking 32bit, 64bit, 64bit BookE, 64bit LE etc. > > > > One bonus is if we have a smaller set of defconfigs we might actually > > get better testing. I have no idea which 32bit defconfigs I should test > > for example, and I'm not going to test them all! > > FWIW, I test with these configs and it seems to catch most of 32/64 bit, > SMP/UP, differnt MMU issues: > > pseries_defconfig > ppc64_defconfig > ppc64e_defconfig > pmac32_defconfig > ppc44x_defconfig > ppc6xx_defconfig > mpc85xx_smp_defconfig > mpc85xx_defconfig > chroma_defconfig
> ps3_defconfig > celleb_defconfig I think cell_defconfig pretty much covers these too, and also gets you coverage of the IBM blades. And add randconfig too :) cheers _______________________________________________ Linuxppc-dev mailing list Linuxppc-dev@lists.ozlabs.org https://lists.ozlabs.org/listinfo/linuxppc-dev
