[Solved] Pacman warning in update.

judd1 · 2018-06-08 11:07:05

When updating today, I find a pacman warning:

warning: could not get information from the file usr/share/libalpm/hooks/gio-querymodules-32.hook

Obviously I do not have anything in that file, but then, just in case I regenerated the image with mkinitcpio -p linux and there was no problem.

I publish because I do not understand the warning, although I searched the forum I do not find anything.

pacman.log: https://hastebin.com/raw/ekikazewiz

Last edited by judd1 (2018-06-10 14:34:36)

eschwartz · 2018-06-08 12:52:18

judd1 wrote:

When updating today, I find a pacman warning:
warning: could not get information from the file usr/share/libalpm/hooks/gio-querymodules-32.hook
Obviously I do not have anything in that file,

Why is this obvious? Why do you have a broken hook there?

More importantly, where does this error message really come from, since the string 'get information' does not appear anywhere in the pacman codebase.

but then, just in case I regenerated the image with mkinitcpio -p linux and there was no problem.

I'm totally lost. What does "the image" have to do with anything here? What image? You mentioned mkinitcpio, but the kernel initramfs has nothing to do with either pacman or the glib2 package from which "gio-querymodules" (and its multilib variant) come from.

I publish because I do not understand the warning, although I searched the forum I do not find anything.
pacman.log: https://hastebin.com/raw/ekikazewiz

I'm more interested in the contents of

pacman -Qs glib2
pacman -Qo /usr/share/libalpm/hooks/*
cat /usr/share/libalpm/hooks/gio-querymodules-32.hook

Where did this hook file come from, what's in it, and who is complaining about it?

judd1 · 2018-06-08 13:25:25

Eschwartz wrote:

Why is this obvious? Why do you have a broken hook there?

I do not know why the hook is there, I simply relate it to mkinitcpio (bad of me)

Eschwartz wrote:

I'm more interested in the contents of

pacman -Qs glib2
pacman -Qo /usr/share/libalpm/hooks/*
cat /usr/share/libalpm/hooks/gio-querymodules-32.hook

$ pacman -Qs glib2
local/glib2 2.56.1-1
    Low level core library
local/lib32-glib2 2.56.1-1
    Low level core library (32-bit)

$ pacman -Qo /usr/share/libalpm/hooks/*
/usr/share/libalpm/hooks/60-linux.hook is contained in linux 4.16.13-2
/usr/share/libalpm/hooks/90-linux.hook is contained in linux 4.16.13-2
/usr/share/libalpm/hooks/detect-old-perl-modules.hook is contained in perl 5.26.2-1
/usr/share/libalpm/hooks/gio-querymodules-32.hook is contained in lib32-glib2 2.56.1-1
/usr/share/libalpm/hooks/gvfsd.hook is contained in gvfs 1.36.2-2
/usr/share/libalpm/hooks/systemd-binfmt.hook is contained in systemd 238.133-1
/usr/share/libalpm/hooks/systemd-catalog.hook is contained in systemd 238.133-1
/usr/share/libalpm/hooks/systemd-daemon-reload.hook is contained in systemd 238.133-1
/usr/share/libalpm/hooks/systemd-hwdb.hook is contained in systemd 238.133-1
/usr/share/libalpm/hooks/systemd-sysctl.hook is contained in systemd 238.133-1
/usr/share/libalpm/hooks/systemd-sysusers.hook is contained in systemd 238.133-1
/usr/share/libalpm/hooks/systemd-tmpfiles.hook is contained in systemd 238.133-1
/usr/share/libalpm/hooks/systemd-udev-reload.hook is contained in systemd 238.133-1
/usr/share/libalpm/hooks/systemd-update.hook is contained in systemd 238.133-1
/usr/share/libalpm/hooks/texinfo-install.hook is contained in texinfo 6.5-1
/usr/share/libalpm/hooks/texinfo-remove.hook is contained in texinfo 6.5-1

Yeah sure, I forgot to put the bar to /usr ... after $ cat

$ cat /usr/share/libalpm/hooks/gio-querymodules-32.hook
[Trigger]
Type = File
Operation = Install
Operation = Upgrade
Operation = Remove
Target = usr/lib32/gio/modules/*.so

[Action]
Description = Updating 32-bit GIO module cache...
When = PostTransaction
Exec = /usr/bin/gio-querymodules-32 /usr/lib32/gio/modules

Last edited by judd1 (2018-06-08 13:32:19)

eschwartz · 2018-06-08 13:40:26

This message should be generated by printf(_("%s is owned by %s%s %s%s%s\n"), filename, colstr->title,

As seen in https://git.archlinux.org/pacman.git/tr … v5.1.0#n83

I'm unwilling to engage with people who modify the output of commands before posting it.

I've got no idea where the error message you posted in the initial post came from, but I'm no longer confident I can figure out what it really said either. Since what you're posting is neither the LANG=C output, nor even any of the official localizations, the only alternative to your malicious modification of output is that you're using your own handcrafted locales which you built and copied into /usr/share/locale/ by hand and forgot to tell us. I *severely* doubt this is the cause.

judd1 · 2018-06-08 13:54:10

Eschwartz wrote:

I'm unwilling to engage with people who modify the output of commands before posting it.

There is no malice on my part, not at all, and I would not come here to bother, I am not one of those people and least of all at 52 years old. On the other hand, I'm not willing to be answered that way.

I just started this post, from what I understand, it can be a problem. Only that.

(I better keep listening to blues ...)

Last edited by judd1 (2018-06-08 14:06:04)

Trilby · 2018-06-08 14:23:19

There is a bit of miscommunication here, clearly. Judd1, we expect that error messages are reproduced verbatim. There are often little queues that may seem irrelevant to some, but turn out to be quite important, so we really need to see exactly what the output was (remember, all diagnosticians are from Missouri). This is best acheived by either copy and pasing from the terminal, or piping the stdout (and stderr) to a file and posting that file.

I suspect you must have typed out that the error message you posted yourself. Might you have also paraphrased or accidentally modified it in some way?

I agree raising the suggestion that the output could have been maliciously modified was unproductive. But note that Eschwartz is trying to help, and the information you are giving him just doesn't add up. If that warning came from pacman, it should be in pacman's log. But there's nothing even remotely closely resembling that message there. So it begs the question, what did generate that message. Do you use a pacman wrapper and/or aur-helper to do you updates?

Are you able to replicate this message?

judd1 · 2018-06-08 15:27:37

Trilby

First of all, I appreciate your deference for your treatment of me.

The only thing that generated this problem (and I assume it) is that when copying from console, having finished today's update, I put the mouse in usr when I should have copied it from the bar before usr, that is from / usr and not from usr, I mean? It was just a bad mouse action on my part, nothing more than that.

With respect to my "bad intentions", I have not paraphrased or modified anything in the post, nor in the answers that I have given by request of Eschwartz. I also understand and nobility obliges, that the post is not well written, I am not good at making me understand, although starting, should not be treated well to someone who writes with the best good intentions, which is my case. Besides, I do not have the knowledge that you have at all, I want to learn what I can most from those who know by experience, intelligence and ability on this IT task. (A little patience please ...)

Exactly, as his question at the end says: I updated, unintentionally and I had the error of not telling Eschwartz with yaourt -Syu
This is what then generated that warning of pacman, now I see it, late but I see it at the end.

So that everything is clearer, whether or not it works, this is my update today, with pacman ... (of course)

[judd@arch ~]$ sudo pacman -Syyu
[sudo] password for judd: 
:: Sincronizando las bases de datos de los paquetes...
 core                     129,9 KiB   500K/s 00:00 [----------------------] 100%
 extra                   1631,7 KiB   922K/s 00:02 [----------------------] 100%
 community                  4,4 MiB   400K/s 00:11 [----------------------] 100%
 multilib                 171,1 KiB   514K/s 00:00 [----------------------] 100%
:: Iniciando actualización completa del sistema...
resolviendo dependencias...
buscando conflictos entre paquetes...

Paquetes (12) boost-libs-1.67.0-4  firefox-60.0.2-1
              firefox-i18n-es-ar-60.0.2-1  flashplugin-30.0.0.113-1
              hardinfo-0.5.1-7  lib32-glib2-2.56.1-1  libcmis-0.5.1-9
              libixion-0.13.0-2  liborcus-0.13.4-2  ncmpcpp-0.8.2-2
              pepper-flash-30.0.0.113-1  pkcs11-helper-1.23-1

Tamaño total de la descarga:     57,04 MiB
Tamaño total de la instalación:  202,33 MiB
Tamaño neto tras actualizar:      0,33 MiB

:: ¿Continuar con la instalación? [S/n] 
:: Obteniendo los paquetes...
 pkcs11-helper-1.23-...    65,3 KiB   416K/s 00:00 [----------------------] 100%
 boost-libs-1.67.0-4...     2,4 MiB   839K/s 00:03 [----------------------] 100%
 firefox-60.0.2-1-x86_64   36,7 MiB   690K/s 00:54 [----------------------] 100%
 firefox-i18n-es-ar-...   465,9 KiB   759K/s 00:01 [----------------------] 100%
 flashplugin-30.0.0....     6,9 MiB   769K/s 00:09 [----------------------] 100%
 libcmis-0.5.1-9-x86_64   509,9 KiB   721K/s 00:01 [----------------------] 100%
 libixion-0.13.0-2-x...   224,4 KiB   672K/s 00:00 [----------------------] 100%
 liborcus-0.13.4-2-x...   598,5 KiB   682K/s 00:01 [----------------------] 100%
 pepper-flash-30.0.0...     7,3 MiB   672K/s 00:11 [----------------------] 100%
 hardinfo-0.5.1-7-x86_64  221,6 KiB   852K/s 00:00 [----------------------] 100%
 ncmpcpp-0.8.2-2-x86_64   676,2 KiB   562K/s 00:01 [----------------------] 100%
 lib32-glib2-2.56.1-...  1020,5 KiB   136K/s 00:08 [----------------------] 100%
(12/12) comprobando las claves del depósito        [----------------------] 100%
(12/12) verificando la integridad de los paquetes  [----------------------] 100%
(12/12) cargando los archivos de los paquetes      [----------------------] 100%
(12/12) comprobando conflictos entre archivos      [----------------------] 100%
(12/12) comprobando el espacio disponible en e...  [----------------------] 100%
advertencia: no se pudo obtener información del archivo usr/share/libalpm/hooks/gio-querymodules-32.hook
:: Procesando los cambios de los paquetes...
( 1/12) actualizando boost-libs                    [----------------------] 100%
( 2/12) actualizando firefox                       [----------------------] 100%
( 3/12) actualizando firefox-i18n-es-ar            [----------------------] 100%
( 4/12) actualizando flashplugin                   [----------------------] 100%
( 5/12) actualizando hardinfo                      [----------------------] 100%
( 6/12) actualizando lib32-glib2                   [----------------------] 100%
( 7/12) actualizando libcmis                       [----------------------] 100%
( 8/12) actualizando libixion                      [----------------------] 100%
( 9/12) actualizando liborcus                      [----------------------] 100%
(10/12) actualizando ncmpcpp                       [----------------------] 100%
(11/12) actualizando pepper-flash                  [----------------------] 100%
(12/12) actualizando pkcs11-helper                 [----------------------] 100%
:: Ejecutando los «hooks» de posinstalación...
(1/1) Arming ConditionNeedsUpdate...

I apologize to you and Eschwartz for my carelessness and outrage when it comes to addressing a question the way I have.

My respects.

Last edited by judd1 (2018-06-08 15:48:16)

apg · 2018-06-08 21:48:43

At this point I have to assume that you translated the messages into English yourself for your first couple posts. Do not do that. That is exactly the sort of thing that Trilby and Eschwartz are talking about. Run commands with LANG=C to get output in English. What is the output of `stat /usr/share/libalpm/hooks/gio-querymodules-32.hook'?

jasonwryan · 2018-06-08 22:03:23

Why is this marked as solved? Please read the Code of Conduct: https://wiki.archlinux.org/index.php/Co … way_street

judd1 · 2018-06-08 22:46:01

apg wrote:

At this point I have to assume that you translated the messages into English yourself for your first couple posts. Do not do that. That is exactly the sort of thing that Trilby and Eschwartz are talking about.

I understand, thank you

apg wrote:

Run commands with LANG=C to get output in English. What is the output of `stat /usr/share/libalpm/hooks/gio-querymodules-32.hook'?

$ LANG=C stat /usr/share/libalpm/hooks/gio-querymodules-32.hook
  File: /usr/share/libalpm/hooks/gio-querymodules-32.hook
  Size: 259       	Blocks: 8          IO Block: 4096   regular file
Device: 812h/2066d	Inode: 828402      Links: 1
Access: (0644/-rw-r--r--)  Uid: (    0/    root)   Gid: (    0/    root)
Access: 2018-06-08 07:48:08.486763439 -0300
Modify: 2018-06-07 11:14:08.000000000 -0300
Change: 2018-06-08 07:48:07.586763407 -0300
 Birth: -

Last edited by judd1 (2018-06-08 22:56:22)

judd1 · 2018-06-08 23:14:02

jasonwryan wrote:

Why is this marked as solved?

Unmarked.

jasonwryan wrote:

Please read the Code of Conduct: https://wiki.archlinux.org/index.php/Co … way_street

I have already learned it.

eschwartz · 2018-06-10 05:34:53

judd1 wrote:

With respect to my "bad intentions", I have not paraphrased or modified anything in the post, nor in the answers that I have given by request of Eschwartz. I also understand and nobility obliges, that the post is not well written, I am not good at making me understand, although starting, should not be treated well to someone who writes with the best good intentions, which is my case. Besides, I do not have the knowledge that you have at all, I want to learn what I can most from those who know by experience, intelligence and ability on this IT task. (A little patience please ...)
Exactly, as his question at the end says: I updated, unintentionally and I had the error of not telling Eschwartz with yaourt -Syu
This is what then generated that warning of pacman, now I see it, late but I see it at the end.

If you post stuff in Spanish, the worst that will happen is we'll ask you to re-run the command using LANG=C. If you post in the "other languages" subforum, you can even find people who can discuss this directly in your native language.

If you try translating it yourself, you may *think* you're being helpful, but really you just end up getting people upset when they don't understand your translations and cannot figure out where they come from. If I wanted the original English, I have two options: I can look up the translation files from pacman, or I can use Google Translate *myself*. Instead, you posted one line, without context, so I had no cues as to where it might have come from, and the one thing I knew was that that exact error message did not exist.

Round-trip translation is typically used as a form of casual entertainment for comedic purposes, and obviously has no place in a technical help forum. Most importantly, it has no place here when other people do not know they are dealing with such translations!

In short, use LANG=C whenever possible, but if you cannot repeat the error, at least give us the original version in Spanish and let *us* do the translating. At least then we will understand that we're dealing with localized messages, and have some basis for deciphering it, rather than being tricked into thinking that that is the real, original English message.

So that everything is clearer, whether or not it works, this is my update today, with pacman ... (of course)

[judd@arch ~]$ sudo pacman -Syyu
[...]
(12/12) comprobando las claves del depósito        [----------------------] 100%
(12/12) verificando la integridad de los paquetes  [----------------------] 100%
(12/12) cargando los archivos de los paquetes      [----------------------] 100%
(12/12) comprobando conflictos entre archivos      [----------------------] 100%
(12/12) comprobando el espacio disponible en e...  [----------------------] 100%
advertencia: no se pudo obtener información del archivo usr/share/libalpm/hooks/gio-querymodules-32.hook
:: Procesando los cambios de los paquetes...
( 1/12) actualizando boost-libs                    [----------------------] 100%
( 2/12) actualizando firefox                       [----------------------] 100%
( 3/12) actualizando firefox-i18n-es-ar            [----------------------] 100%
( 4/12) actualizando flashplugin                   [----------------------] 100%
[...]

I apologize to you and Eschwartz for my carelessness and outrage when it comes to addressing a question the way I have.

My respects.

So this finally tells me something.

I grepped in the translation files for pacman, to figure out what the original message was. While somewhat awkward (I prefer LANG=C obviously), it's at least something I'm physically capable of... once I see the exact untranslated version.

This message is:

(12/12) checking available disk space...
warning: could not get file information for usr/share/libalpm/hooks/gio-querymodules-32.hook

It comes from https://git.archlinux.org/pacman.git/tr … 5.1.0#n241

What it is saying is, that the file in question *was* owned by a package, lib32-glib2, which was part of the list of packages which this -Syu updated.

It is, in fact, still owned by that package. But at the time of the update, the file did not exist on disk, even though pacman thought there should be a file there.

So pacman warned you that it could not find the file in order to delete the old version of the file during the update process. But it's not an error, since by upgrading and getting the file installed any issues would, presumably, be fixed.

...

This warning happens if for some reason a file which pacman installed, was deleted by hand. The next time the package which owns the file gets updated, pacman will warn you that the file doesn't exist, then reinstall it as part of the upgrade.

I don't know why this hook was deleted in the past, but now that it exists again everything should be fine.

Last edited by eschwartz (2018-06-10 05:37:57)

judd1 · 2018-06-10 14:32:48

Eschwartz wrote:

judd1 wrote:
With respect to my "bad intentions", I have not paraphrased or modified anything in the post, nor in the answers that I have given by request of Eschwartz. I also understand and nobility obliges, that the post is not well written, I am not good at making me understand, although starting, should not be treated well to someone who writes with the best good intentions, which is my case. Besides, I do not have the knowledge that you have at all, I want to learn what I can most from those who know by experience, intelligence and ability on this IT task. (A little patience please ...)
Exactly, as his question at the end says: I updated, unintentionally and I had the error of not telling Eschwartz with yaourt -Syu
This is what then generated that warning of pacman, now I see it, late but I see it at the end.
If you post stuff in Spanish, the worst that will happen is we'll ask you to re-run the command using LANG=C. If you post in the "other languages" subforum, you can even find people who can discuss this directly in your native language.
If you try translating it yourself, you may *think* you're being helpful, but really you just end up getting people upset when they don't understand your translations and cannot figure out where they come from. If I wanted the original English, I have two options: I can look up the translation files from pacman, or I can use Google Translate *myself*. Instead, you posted one line, without context, so I had no cues as to where it might have come from, and the one thing I knew was that that exact error message did not exist.
Round-trip translation is typically used as a form of casual entertainment for comedic purposes, and obviously has no place in a technical help forum. Most importantly, it has no place here when other people do not know they are dealing with such translations!
In short, use LANG=C whenever possible, but if you cannot repeat the error, at least give us the original version in Spanish and let *us* do the translating. At least then we will understand that we're dealing with localized messages, and have some basis for deciphering it, rather than being tricked into thinking that that is the real, original English message.
So that everything is clearer, whether or not it works, this is my update today, with pacman ... (of course)
[judd@arch ~]$ sudo pacman -Syyu
[...]
(12/12) comprobando las claves del depósito        [----------------------] 100%
(12/12) verificando la integridad de los paquetes  [----------------------] 100%
(12/12) cargando los archivos de los paquetes      [----------------------] 100%
(12/12) comprobando conflictos entre archivos      [----------------------] 100%
(12/12) comprobando el espacio disponible en e...  [----------------------] 100%
advertencia: no se pudo obtener información del archivo usr/share/libalpm/hooks/gio-querymodules-32.hook
:: Procesando los cambios de los paquetes...
( 1/12) actualizando boost-libs                    [----------------------] 100%
( 2/12) actualizando firefox                       [----------------------] 100%
( 3/12) actualizando firefox-i18n-es-ar            [----------------------] 100%
( 4/12) actualizando flashplugin                   [----------------------] 100%
[...]
I apologize to you and Eschwartz for my carelessness and outrage when it comes to addressing a question the way I have.
My respects.

From this moment, I will follow your advice and teachings on how I have to proceed; not to bother opening a topic in the forum.

Eschwartz wrote:

So this finally tells me something.
I grepped in the translation files for pacman, to figure out what the original message was. While somewhat awkward (I prefer LANG=C obviously), it's at least something I'm physically capable of... once I see the exact untranslated version.
This message is:
(12/12) checking available disk space...
warning: could not get file information for usr/share/libalpm/hooks/gio-querymodules-32.hook
It comes from https://git.archlinux.org/pacman.git/tr … 5.1.0#n241
What it is saying is, that the file in question *was* owned by a package, lib32-glib2, which was part of the list of packages which this -Syu updated.
It is, in fact, still owned by that package. But at the time of the update, the file did not exist on disk, even though pacman thought there should be a file there.
So pacman warned you that it could not find the file in order to delete the old version of the file during the update process. But it's not an error, since by upgrading and getting the file installed any issues would, presumably, be fixed.
...
This warning happens if for some reason a file which pacman installed, was deleted by hand. The next time the package which owns the file gets updated, pacman will warn you that the file doesn't exist, then reinstall it as part of the upgrade.
I don't know why this hook was deleted in the past, but now that it exists again everything should be fine.

Now I understand how Pacman thought that file was on the disk and when he did not find it he decided to notice what was happening.
Impeccable your explanation.

Well, in the end and like almost everything in life "mistakes are learned". I am truly grateful for your teaching.

Eschwartz: again I apologize for the bad time that has gone through my wrongdoing.

My respects.

Arch Linux

#1 2018-06-08 11:07:05

[Solved] Pacman warning in update.

#2 2018-06-08 12:52:18

Re: [Solved] Pacman warning in update.

#3 2018-06-08 13:25:25

Re: [Solved] Pacman warning in update.

#4 2018-06-08 13:40:26

Re: [Solved] Pacman warning in update.

#5 2018-06-08 13:54:10

Re: [Solved] Pacman warning in update.

#6 2018-06-08 14:23:19

Re: [Solved] Pacman warning in update.

#7 2018-06-08 15:27:37

Re: [Solved] Pacman warning in update.

#8 2018-06-08 21:48:43

Re: [Solved] Pacman warning in update.

#9 2018-06-08 22:03:23

Re: [Solved] Pacman warning in update.

#10 2018-06-08 22:46:01

Re: [Solved] Pacman warning in update.

#11 2018-06-08 23:14:02

Re: [Solved] Pacman warning in update.

#12 2018-06-10 05:34:53

Re: [Solved] Pacman warning in update.

#13 2018-06-10 14:32:48

Re: [Solved] Pacman warning in update.

Board footer