On Thu, Oct 25, 2012 at 1:51 PM, Seiji Aguchi <seiji.aguchi@xxxxxxx> wrote: >> So doesn't this break erasing of existing dumps in pstore-efivars? >> Unless efi_pstore_read still understands the older variable name formats, users will be stranded with dumps consuming space in >> efivars that aren't exported via pstore anymore. > > Patch 2/5, 3/5 and 4/5 don't change an existing format of a variable name and don't break anything. > The existing format consists of type, id and ctime. And it is not moditied by these patches. > Here is a variable name (and guid) of efi_pstore in curret linus-tree. > > ls /sys/firmware/efi/vars/ > dump-type0-1-1351113059-cfc8fc79-be2e-4ddc-97f0-9f98bfe298a0 > dump-type0-2-1351113059-cfc8fc79-be2e-4ddc-97f0-9f98bfe298a0 > dump-type0-3-1351113059-cfc8fc79-be2e-4ddc-97f0-9f98bfe298a0 > > Variable Name: dump-type0-1-1351113059 > GUID: cfc8fc79-be2e-4ddc-97f0-9f98bfe298a0 Okay, this alleviates most of my concerns for ctime. > > On the other hand, patch 5/5 changes the format by adding sequence counter. > But efi_pstore_read is modied to work correctly in it. > > dump-type0-1-1-1351113059-cfc8fc79-be2e-4ddc-97f0-9f98bfe298a0 > > Variable Name: dump-type0-1-1-1351113059 > GUID: cfc8fc79-be2e-4ddc-97f0-9f98bfe298a0 > > If I need to elaborate more, please feel free to ask me:) > I'm not sure if I understand your question completely. In this case, I think efi_pstore_read in patch 5/5 should probably fall-back to trying to do a sscanf(...) == 3 if the sscanf(...) == 4 test fails so that dumps for older dumps from previous kernels are still visible to users, no? They can perhaps get a default count of 0? efi_pstore_erase would have to be updated to understand this as well. -- To unsubscribe from this list: send the line "unsubscribe linux-efi" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
