On 06.07.2023 00:51, Stefano Stabellini wrote: > On Tue, 4 Jul 2023, Jan Beulich wrote: >> On 29.06.2023 17:55, Federico Serafini wrote: >>> --- a/xen/arch/x86/cpu/mcheck/mce.h >>> +++ b/xen/arch/x86/cpu/mcheck/mce.h >>> @@ -44,7 +44,7 @@ extern uint8_t cmci_apic_vector; >>> extern bool lmce_support; >>> >>> /* Init functions */ >>> -enum mcheck_type amd_mcheck_init(struct cpuinfo_x86 *c); >>> +enum mcheck_type amd_mcheck_init(struct cpuinfo_x86 *ci); >> >> Supported even by ... >> >>> enum mcheck_type intel_mcheck_init(struct cpuinfo_x86 *c, bool bsp); >> >> ... immediate context, c is the name we use for parameters of this type. >> You want to change the definition in such cases instead of the >> declaration. >> >> I also think this patch could do with splitting, not the least to >> reduce the Cc list(s) needed. > > This is a very large patch but it is entirely mechanical (good!) > > How would you see it split? By individual files (too many in my > opinion), or maybe by directory? By directory it would be something > like: > > xen/arch/x86/cpu/mcheck > xen/arch/x86/hvm > xen/arch/x86/include/asm
As for the other patch and as mentioned in the earlier replay: By maintainership area, such that we don't need to accumulate a dozen of acks (easy to miss one) in order for the patch to go in. As much as the patch is mechanical, it ought to be largely mechanical to split the patch at maintainership boundaries. Jan
