[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Bug#994542: Bug#994543: Workaround glibc bug BZ#26401 (regression in lchmod and fchmodat when /proc is not mounted)



Control: retitle -1 Workaround glibc bug BZ#26401 (regression in lchmod and fchmodat when /proc is not mounted)

Hi

On Fri, Sep 17, 2021 at 03:29:26PM +0200, Salvatore Bonaccorso wrote:
> Source: glibc
> Version: 2.32-3
> Severity: important
> Tags: upstream
> Forwarded: https://sourceware.org/bugzilla/show_bug.cgi?id=26401
> X-Debbugs-Cc: carnil@debian.org
> Control: affects -1 rsync
> Control: clone -1 -2
> Control: reassing -2 src:rsync 3.2.3-6
> Control: retitle -2 rsync: Workaround glibc bug BZ#26401 (regression in lchmod and fchmodat when /proc is not mountet)
> Control: forwarded -2 https://github.com/WayneD/rsync/issues/109
> 
> Hi
> 
> There is a regression in lchmod and fchmodat when /proc is not
> mounted, the upstream report about it is at [1] and affects as well
> e.g. rsync[2]. It relates to #951191.
> 
>  [1] https://sourceware.org/bugzilla/show_bug.cgi?id=26401
>  [2] https://github.com/WayneD/rsync/issues/109
> 
> For rsync the following where applied:
> 
> https://github.com/WayneD/rsync/commit/85b8dc8abaca96fc3ea7421e09101b6ac41b6718
> https://github.com/WayneD/rsync/commit/9dd62525f3b98d692e031f22c02be8f775966503

FWIW, the autopkgtest regresssons are maybe related to this. I
encountered it while perfoming a linux build in a chroot where /proc
was not mounted.

Tested cherry-picking the workaround on top of 3.2.3-6.

Regards,
Salvatore


Reply to: