Bug#935496: Fails to install: Invalid action '-p'
Hi, On Fri, Aug 30, 2019 at 12:03 PM Simon McVittie wrote: > On Fri, 23 Aug 2019 at 10:04:28 +0200, Michael Biebl wrote: > > Setting up fuse (2.9.9-1) ... > > dpkg: error processing package fuse (--configure): > > installed fuse package post-installation script subprocess returned error > > exit status 1 > ... > > # udevadm test --action -p /devices/virtual/misc/fuse > > Invalid action '-p' Ouch, it's a leftover (outdated lines) from the time when fuse installed its udev rules. Now it should be something like this: "udevadm test --action=change /devices/virtual/misc/fuse". But now as udev ships its fuse rules (in 50-udev-default.rules and 99-systemd.rules) this is not needed anymore. > Similar to the equivalent fuse3 bug #934293, this seems to be a regression > since buster: the same binary package installs OK on buster. Maybe > udevadm became more strict about its parameter parsing? As mentioned elsewhere, this seems to be the case. About the the transition, it is expected but I would like to get more information on it. > Similar to fuse3, it would be helpful if the maintainer script had less > "> /dev/null 2>&1" so that error messages would appear. Well, the expected output (on my Buster system) is 131 lines long. These are not relevant for normal users / usage. > > I'm not exactly sure what this code is supposed to achieve. > > Since fuse no longer ships its own udev rules, maybe it can be dropped > > altogether? Trigger an udev rules change after the package installed its udev rule. > Or if the postinst is still necessary, maybe fuse3 could take over the > fuse binary package name for bullseye (with a transitional package) so that > bugs like this one don't need to be fixed in both places? For the time I will fix it independently. I can't promise when the actual transition will take place. But if you can, please check the proposed package update[1]. Thanks, Laszlo/GCS [1] dget -x http://www.barcikacomp.hu/gcs/fuse_2.9.9-2.dsc
Bug#935496: Fails to install: Invalid action '-p'
On Fri, 30 Aug 2019 18:57:11 +0200 Michael Biebl wrote: > Am 30.08.19 um 11:04 schrieb Simon McVittie: > > Control: tags -1 + bullseye sid > > > > On Fri, 23 Aug 2019 at 10:04:28 +0200, Michael Biebl wrote: > >> Setting up fuse (2.9.9-1) ... > >> dpkg: error processing package fuse (--configure): > >> installed fuse package post-installation script subprocess returned error > >> exit status 1 > > ... > >> # udevadm test --action -p /devices/virtual/misc/fuse > >> Invalid action '-p' > > > > Similar to the equivalent fuse3 bug #934293, this seems to be a regression > > since buster: the same binary package installs OK on buster. Maybe > > udevadm became more strict about its parameter parsing? > > I'm currently investigating, whether this is a regression in udevadm or not Quoting upstream https://github.com/systemd/systemd/issues/13442#issuecomment-526688796 > I do not think it is an our bug, but the bug in the script used in the > install script of fuse. udevadm test does not support -p, and --action option > requires argument. Also, of course, -p is not a valid action. -- Why is it that all of the instruments seeking intelligent life in the universe are pointed away from Earth? signature.asc Description: OpenPGP digital signature
Bug#935496: Fails to install: Invalid action '-p'
Am 30.08.19 um 11:04 schrieb Simon McVittie: > Control: tags -1 + bullseye sid > > On Fri, 23 Aug 2019 at 10:04:28 +0200, Michael Biebl wrote: >> Setting up fuse (2.9.9-1) ... >> dpkg: error processing package fuse (--configure): >> installed fuse package post-installation script subprocess returned error >> exit status 1 > ... >> # udevadm test --action -p /devices/virtual/misc/fuse >> Invalid action '-p' > > Similar to the equivalent fuse3 bug #934293, this seems to be a regression > since buster: the same binary package installs OK on buster. Maybe > udevadm became more strict about its parameter parsing? I'm currently investigating, whether this is a regression in udevadm or not That said: > Similar to fuse3, it would be helpful if the maintainer script had less > "> /dev/null 2>&1" so that error messages would appear. > >> I'm not exactly sure what this code is supposed to achieve. >> Since fuse no longer ships its own udev rules, maybe it can be dropped >> altogether? > > Or if the postinst is still necessary, maybe fuse3 could take over the > fuse binary package name for bullseye (with a transitional package) so that > bugs like this one don't need to be fixed in both places? That unhelpful redirection aside, that maintainer scripts code (in fuse/fuse3) looks fishy and should be carefully reviewed if it is actually still needed and if so, add a comment or two. -- Why is it that all of the instruments seeking intelligent life in the universe are pointed away from Earth? signature.asc Description: OpenPGP digital signature
Bug#935496: Fails to install: Invalid action '-p'
Control: tags -1 + bullseye sid On Fri, 23 Aug 2019 at 10:04:28 +0200, Michael Biebl wrote: > Setting up fuse (2.9.9-1) ... > dpkg: error processing package fuse (--configure): > installed fuse package post-installation script subprocess returned error > exit status 1 ... > # udevadm test --action -p /devices/virtual/misc/fuse > Invalid action '-p' Similar to the equivalent fuse3 bug #934293, this seems to be a regression since buster: the same binary package installs OK on buster. Maybe udevadm became more strict about its parameter parsing? Similar to fuse3, it would be helpful if the maintainer script had less "> /dev/null 2>&1" so that error messages would appear. > I'm not exactly sure what this code is supposed to achieve. > Since fuse no longer ships its own udev rules, maybe it can be dropped > altogether? Or if the postinst is still necessary, maybe fuse3 could take over the fuse binary package name for bullseye (with a transitional package) so that bugs like this one don't need to be fixed in both places? smcv
Bug#935496: Fails to install: Invalid action '-p'
Package: fuse Version: 2.9.9-1 Severity: serious Trying to run apt (re)install fuse results in $ sudo apt install --reinstall fuse Reading package lists... Done Building dependency tree Reading state information... Done 0 upgraded, 0 newly installed, 1 reinstalled, 0 to remove and 0 not upgraded. Need to get 72.3 kB of archives. After this operation, 0 B of additional disk space will be used. Get:1 http://ftp.debian.org/debian sid/main amd64 fuse amd64 2.9.9-1 [72.3 kB] Fetched 72.3 kB in 0s (219 kB/s) [master ba9ff87] saving uncommitted changes in /etc prior to apt run Author: Michael Biebl 1 file changed, 5 insertions(+), 5 deletions(-) (Reading database ... 389058 files and directories currently installed.) Preparing to unpack .../fuse_2.9.9-1_amd64.deb ... Unpacking fuse (2.9.9-1) over (2.9.9-1) ... Setting up fuse (2.9.9-1) ... dpkg: error processing package fuse (--configure): installed fuse package post-installation script subprocess returned error exit status 1 Processing triggers for man-db (2.8.6.1-1) ... Errors were encountered while processing: fuse E: Sub-process /usr/bin/dpkg returned an error code (1) Looking at the postinst code, the faulty line is udevadm test --action -p $(udevadm info -q path -n /dev/fuse) > /dev/null 2>&1 Running that yields # udevadm test --action -p /devices/virtual/misc/fuse Invalid action '-p' I'm not exactly sure what this code is supposed to achieve. Since fuse no longer ships its own udev rules, maybe it can be dropped altogether? -- System Information: Debian Release: bullseye/sid APT prefers unstable APT policy: (500, 'unstable'), (200, 'experimental') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 4.19.0-5-amd64 (SMP w/4 CPU cores) Kernel taint flags: TAINT_OOT_MODULE, TAINT_UNSIGNED_MODULE Locale: LANG=de_DE.UTF-8, LC_CTYPE=de_DE.UTF-8 (charmap=UTF-8), LANGUAGE=de_DE.UTF-8 (charmap=UTF-8) Shell: /bin/sh linked to /bin/dash Init: systemd (via /run/systemd/system) LSM: AppArmor: enabled Versions of packages fuse depends on: ii adduser 3.118 ii libc6 2.28-10 ii libfuse2 2.9.9-1 ii mount 2.34-0.1 ii sed 4.7-1 fuse recommends no packages. fuse suggests no packages. -- Configuration Files: /etc/fuse.conf [Errno 13] Keine Berechtigung: '/etc/fuse.conf' -- no debconf information
