efibootmgr-0.5.4-x86_64 vs efibootmgr-0.6.0-x86_64

Olaus · 10-19-2016, 06:08 AM

Hi, the standard installation of Slackware64-14.2 includes efibootmgr-0.5.4-x86_64. There is also efibootmgr-0.6.0-x86_64 in slackware64-14.2/testing/packages/. Could anyone please tell the differences? Is the latter version somewhat experimental? According to http://linux.dell.com/efibootmgr/efi....5.4/ChangeLog, the last commit for 0.5.4 is 6e6bf6fc7665851798a6c2c92893ebb629e42aff, which is 6 commits behind the latest commit for http://linux.dell.com/efibootmgr/efi....6.0/ChangeLog. There don't seem to be that many significant differences?

zk1234 · 10-19-2016, 11:01 AM

I don't think that there are significant differences between them. Both seem to be "old-fashioned", meaning that they use the efivars kernel module (and not the efvars file system).

.

bassmadrigal · 10-19-2016, 06:02 PM

This never saw any change from 14.1. Here's the changelog entry when he reverted from 0.6.0 to 0.5.4 and his reasoning for it.

Code:

+--------------------------+
Thu Sep 19 06:48:59 UTC 2013
a/efibootmgr-0.5.4-x86_64-1.txz:  Upgraded.
  Well, reverted to an older version really.  :)  It was reported that
  efibootmgr-0.6.0 was silently failing to actually write the new variables
  to EFI, but that 0.5.4 works.  I'm currently unable to test this here as I'm
  still using DUET UEFI, and changes to the UEFI variables do not persist
  between boots.  I also added a few patches from Fedora's repo that seem
  reasonable to include.  The old (newer) version was moved to /testing in
  case it might work better for someone.
  Thanks to John Yost for the bug report.
--snip--
testing/packages/efibootmgr-0.6.0-x86_64-1.txz:  Moved here from A series.

This never saw a recompile or upgrade (or downgrade) after this entry (including during the 14.2 development), so the 14.2 package is the same one from 14.1.

Olaus · 10-25-2016, 12:32 AM

Quote:

Originally Posted by zk1234

I don't think that there are significant differences between them. Both seem to be "old-fashioned", meaning that they use the efivars kernel module (and not the efvars file system).

.

Is this "old-fashioned" way a problem? In that case, what could be used instead?

zk1234 · 10-26-2016, 12:42 AM

Quote:

Originally Posted by Olaus

Is this "old-fashioned" way a problem?

No, it is not.