>>> Yi Sun <yi.y....@linux.intel.com> 08/04/17 5:23 AM >>> >On 17-08-03 03:37:06, Jan Beulich wrote: >> >>> On 03.08.17 at 04:10, <yi.y....@linux.intel.com> wrote: >> > @@ -1283,10 +1344,20 @@ static void psr_cpu_init(void) >> > feat = feat_l3; >> > feat_l3 = NULL; >> > >> > - if ( !cat_init_feature(®s, feat, info, FEAT_TYPE_L3_CAT) ) >> > - feat_props[FEAT_TYPE_L3_CAT] = &l3_cat_props; >> > - else >> > - feat_l3 = feat; >> > + if ( (regs.c & PSR_CAT_CDP_CAPABILITY) && (opt_psr & PSR_CDP) ) >> > + { >> > + /* If CDP init fails, try to work as L3 CAT. */ >> > + if ( !cat_init_feature(®s, feat, info, FEAT_TYPE_L3_CDP) ) >> > + feat_props[FEAT_TYPE_L3_CDP] = &l3_cdp_props; >> >> I'm disappointed to see that you didn't merge to two if()s, as >> pointed out in v15 review. >> >Sorry, I missed this. A new patch 15.2 is sent out. Please help to review.
Well, I suppose you've seen that I've committed the series (adjusting the code above). Hence from now on it'll need to be incremental changes, unless we'll find it necessary to revert due to the regression already reported to you by Andrew. Jan _______________________________________________ Xen-devel mailing list Xen-devel@lists.xen.org https://lists.xen.org/xen-devel