Hi Ted, On Sat, Mar 09, 2024 at 06:38:08PM +0100, Helmut Grohne wrote: > I have uploaded e2fsprogs enabling dh-sequence-movetousr to DELAYED/5 in > agreement with DevRef. Please let me know if I should delay this any > longer and find the actual patch attached.
Christian Hofstaedler reminded me that this was bad idea to the time64 + /usr-move interactions. While I checked that dh-sequence-movetousr would just work for e2fsprogs earlier, the time of check predated time64 and has become invalid. As such, I cancelled the upload. e2fsprogs has three libraries two of which are unaffected by time64. In order to avoid adding mitigations for libss2 and libcom-err2, I revert these unnecessary transitions. For libext2fs2t64, I add the relevant DEP17 mitigation. I've uploaded this to experimental directly to clear NEW and to get a dumat report quickly. Barring your opposition, I'll NMU this to unstable as well. Since you tend to backport e2fsprogs, I want to remind you that your bookworm backport must revert both the time64 transition and the /usr-move transition. In particular, the maintainer scripts I add for libext2fs2t64 must not be included in bookworm. Helmut
