On 9/7/18 9:26 AM, Konstantin Belousov wrote: > On Fri, Sep 07, 2018 at 09:00:20AM -0700, John Baldwin wrote: >> On 9/7/18 8:09 AM, Konstantin Belousov wrote: >>> Author: kib >>> Date: Fri Sep 7 15:09:56 2018 >>> New Revision: 338523 >>> URL: https://svnweb.freebsd.org/changeset/base/338523 >>> >>> Log: >>> Teach sysctl(8) about the Persistent memory type. >>> >>> Add PersistentMemory to the list of sysctl's known memory types >>> when decoding an EFI memory map. >>> >>> Submitted by: D Scott Phillips <[email protected]> >>> MFC after: 1 week >>> Approved by: re (rgrimes) >>> >>> Modified: >>> head/sbin/sysctl/sysctl.c >>> >>> Modified: head/sbin/sysctl/sysctl.c >>> ============================================================================== >>> --- head/sbin/sysctl/sysctl.c Fri Sep 7 14:37:44 2018 >>> (r338522) >>> +++ head/sbin/sysctl/sysctl.c Fri Sep 7 15:09:56 2018 >>> (r338523) >>> @@ -704,7 +704,8 @@ S_efi_map(size_t l2, void *p) >>> "ACPIMemoryNVS", >>> "MemoryMappedIO", >>> "MemoryMappedIOPortSpace", >>> - "PalCode" >>> + "PalCode", >>> + "PersistentMemory" >>> }; >>> >>> /* >>> @@ -733,7 +734,7 @@ S_efi_map(size_t l2, void *p) >>> >>> for (i = 0; i < ndesc; i++, >>> map = efi_next_descriptor(map, efihdr->descriptor_size)) { >>> - if (map->md_type <= EFI_MD_TYPE_PALCODE) >>> + if (map->md_type <= EFI_MD_TYPE_PERSISTENT) >> >> Perhaps this should use nitems(types) instead? (And I believe it's my >> fault it didn't originally.) > > For me, it was more an issue that the code assumes contiguous values for > the EFI_MD_TYPEs constants. > > What about the following:
Looks good to me. -- John Baldwin _______________________________________________ [email protected] mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "[email protected]"
