Bug 14235

can you reproduce if you update you mageia ?

Keywords: (none) => NEEDINFO

Thanks for attention,

Yes,today the bug occurs yet. 

The clock is wrong. I synchronised with ntp.

The bug stays, but I have to say that it's an old computer. I just keep it to test a software installation : Noethys https://bugs.mageia.org/show_bug.cgi?id=14236

Indeed I'm not sure that mageia is updated, because I don't know how to update without this applet. I'll try with konsole.

Here the outing with konsole
[root@localhost expert]# drakrpm-update
Subroutine Gtk3::main redefined at /usr/lib/perl5/vendor_perl/5.18.1/Gtk3.pm line 296.

(drakrpm-update:5677): Gtk-WARNING **: GtkImage 0xa534978 adjusted size vertical min 47 natural 47 must not decrease below min 48 natural 48

(drakrpm-update:5677): Gtk-WARNING **: GtkImage 0xa534978 attempted to adjust its size allocation from -12,1 646x54 to 0,0 634x55. adjust_size_allocation must keep allocation inside original bounds
getting lock on urpmi
using mirror ftp://distrib-coffee.ipsl.jussieu.fr/pub/linux/Mageia/distrib/4/i586
$MIRRORLIST media/core/updates media_info/MD5SUM récupéré
comparaison de /var/cache/urpmi/partial/MD5SUM et de /var/lib/urpmi/Core Updates (distrib3)/MD5SUM
le média « Core Updates (distrib3) » est à jour
$MIRRORLIST media/nonfree/updates media_info/MD5SUM récupéré
comparaison de /var/cache/urpmi/partial/MD5SUM et de /var/lib/urpmi/Nonfree Updates (distrib13)/MD5SUM
le média « Nonfree Updates (distrib13) » est à jour
$MIRRORLIST media/tainted/updates media_info/MD5SUM récupéré
comparaison de /var/cache/urpmi/partial/MD5SUM et de /var/lib/urpmi/Tainted Updates (distrib23)/MD5SUM
le média « Tainted Updates (distrib23) » est à jour
unlocking urpmi database
getting lock on urpmi
using mirror ftp://distrib-coffee.ipsl.jussieu.fr/pub/linux/Mageia/distrib/4/i586
examen de la liste de synthèse [/var/lib/urpmi/Core Release (distrib1)/synthesis.hdlist.cz]
examen de la liste de synthèse [/var/lib/urpmi/Core Updates (distrib3)/synthesis.hdlist.cz]
invalid line <@provides@kernel[== 3.14.18-3.mga4]@kernel[== 3.14]
[...] 
about 20 pages 
[...]
@kmod(zr36060)@kmod(zr3>
problème de lecture du fichier de synthèse du média « Core Updates (distrib3) »
examen de la liste de synthèse [/var/lib/urpmi/Nonfree Release (distrib11)/synthesis.hdlist.cz]
examen de la liste de synthèse [/var/lib/urpmi/Nonfree Updates (distrib13)/synthesis.hdlist.cz]
examen de la liste de synthèse [/var/lib/urpmi/Tainted Release (distrib21)/synthesis.hdlist.cz]
examen de la liste de synthèse [/var/lib/urpmi/Tainted Updates (distrib23)/synthesis.hdlist.cz]
BDB2053 Freeing read locks for locker 0x7fe: 5619/3073300224
BDB2053 Freeing read locks for locker 0x7ff: 5619/3073300224
BDB2053 Freeing read locks for locker 0x800: 5619/3073300224
BDB2053 Freeing read locks for locker 0x801: 5619/3073300224
BDB2053 Freeing read locks for locker 0x802: 5619/3073300224
BDB2053 Freeing read locks for locker 0x803: 5619/3073300224
BDB2053 Freeing read locks for locker 0x804: 5619/3073300224
Subroutine Gtk3::main redefined at /usr/lib/perl5/vendor_perl/5.18.1/Gtk3.pm line 296.
Error creating proxy: La connexion est fermée (g-io-error-quark, 18)

*** Bug 14247 has been marked as a duplicate of this bug. ***

CC: (none) => jkowalzik

*** Bug 14289 has been marked as a duplicate of this bug. ***

CC: (none) => ftstarr

*** Bug 14317 has been marked as a duplicate of this bug. ***

CC: (none) => allan_vaughn

CC'ing Thierry who would be most able to read the perl backtrace.

Severity: normal => major
Priority: Normal => High
CC: (none) => remi, thierry.vignaud
Hardware: i586 => All

Hello,

I think I've got the same type of problem on a fresh install of mageia 4 (with live DVD) when I want to update packages with drakrpm-update: it krashes

  Can't call method "flag_skip" on an undefined value at /usr/lib/perl5/vendor_perl/5.18.1/i386-linux-thread-multi/URPM/Resolve.pm line 582.
  Perl's trace:
  standalone::bug_handler() called from /usr/lib/perl5/vendor_perl/5.18.1/i386-linux-thread-multi/URPM/Resolve.pm:582
  URPM::find_candidate_packages() called from /usr/lib/perl5/vendor_perl/5.18.1/i386-linux-thread-multi/URPM/Resolve.pm:1503
  URPM::__ANON__() called from /usr/lib/perl5/vendor_perl/5.18.1/i386-linux-thread-multi/URPM/Resolve.pm:737
  URPM::__ANON__() called from /usr/lib/perl5/vendor_perl/5.18.1/i386-linux-thread-multi/URPM/Resolve.pm:701
  URPM::__ANON__() called from /usr/lib/perl5/vendor_perl/5.18.1/i386-linux-thread-multi/URPM/Resolve.pm:703
  URPM::with_db_unsatisfied_requires() called from /usr/lib/perl5/vendor_perl/5.18.1/i386-linux-thread-multi/URPM/Resolve.pm:737
  URPM::with_any_unsatisfied_requires() called from /usr/lib/perl5/vendor_perl/5.18.1/i386-linux-thread-multi/URPM/Resolve.pm:1549
  URPM::_handle_diff_provides() called from /usr/lib/perl5/vendor_perl/5.18.1/i386-linux-thread-multi/URPM/Resolve.pm:1283
  URPM::resolve_requested__no_suggests_() called from /usr/lib/perl5/vendor_perl/5.18.1/i386-linux-thread-multi/URPM/Resolve.pm:1174
  URPM::resolve_requested__no_suggests() called from /usr/lib/perl5/vendor_perl/5.18.1/i386-linux-thread-multi/URPM/Resolve.pm:1106
  URPM::resolve_requested() called from /usr/lib/perl5/vendor_perl/5.18.1/Rpmdrake/pkg.pm:408
  Rpmdrake::pkg::get_updates_list() called from /usr/lib/perl5/vendor_perl/5.18.1/Rpmdrake/pkg.pm:509
  Rpmdrake::pkg::get_pkgs() called from /usr/lib/perl5/vendor_perl/5.18.1/Rpmdrake/gui.pm:642
  Rpmdrake::gui::pkgs_provider() called from /usr/libexec/drakrpm-update:266
  main::run_treeview_dialog() called from /usr/libexec/drakrpm-update:293


FYI :
# lspcidrake -v 
ath9k           : Qualcomm Atheros|AR9285 Wireless Network Adapter (PCI-Express) [NETWORK_OTHER] (vendor:168c device:002b subv:1a3b subd:1089) (rev: 01)
atl1c           : Qualcomm Atheros|AR8132 Fast Ethernet [NETWORK_ETHERNET] (vendor:1969 device:1062 subv:1043 subd:838a) (rev: c0)
unknown         : Intel Corporation|NM10/ICH7 Family SATA Controller [AHCI mode] [STORAGE_SATA] (vendor:8086 device:27c1 subv:1043 subd:83ad) (rev: 02)
lpc_ich         : Intel Corporation|NM10 Family LPC Controller [BRIDGE_ISA] (vendor:8086 device:27bc subv:1043 subd:83ad) (rev: 02)
unknown         : Intel Corporation|82801 Mobile PCI Bridge [BRIDGE_PCI] (vendor:8086 device:2448) (rev: e2)
ehci_pci        : Intel Corporation|NM10/ICH7 Family USB2 EHCI Controller [SERIAL_USB] (vendor:8086 device:27cc subv:1043 subd:83ad) (rev: 02)
uhci_hcd        : Intel Corporation|NM10/ICH7 Family USB UHCI Controller #4 [SERIAL_USB] (vendor:8086 device:27cb subv:1043 subd:83ad) (rev: 02)
uhci_hcd        : Intel Corporation|NM10/ICH7 Family USB UHCI Controller #3 [SERIAL_USB] (vendor:8086 device:27ca subv:1043 subd:83ad) (rev: 02)
uhci_hcd        : Intel Corporation|NM10/ICH7 Family USB UHCI Controller #2 [SERIAL_USB] (vendor:8086 device:27c9 subv:1043 subd:83ad) (rev: 02)
uhci_hcd        : Intel Corporation|NM10/ICH7 Family USB UHCI Controller #1 [SERIAL_USB] (vendor:8086 device:27c8 subv:1043 subd:83ad) (rev: 02)
shpchp          : Intel Corporation|NM10/ICH7 Family PCI Express Port 4 [BRIDGE_PCI] (vendor:8086 device:27d6) (rev: 02)
shpchp          : Intel Corporation|NM10/ICH7 Family PCI Express Port 2 [BRIDGE_PCI] (vendor:8086 device:27d2) (rev: 02)
shpchp          : Intel Corporation|NM10/ICH7 Family PCI Express Port 1 [BRIDGE_PCI] (vendor:8086 device:27d0) (rev: 02)
snd_hda_intel   : Intel Corporation|NM10/ICH7 Family High Definition Audio Controller [MULTIMEDIA_AUDIO_DEV] (vendor:8086 device:27d8 subv:1043 subd:83ce) (rev: 02)
unknown         : Intel Corporation|Atom Processor D4xx/D5xx/N4xx/N5xx Integrated Graphics Controller [DISPLAY_OTHER] (vendor:8086 device:a012 subv:1043 subd:83ac)
Card:Intel 810 and later: Intel Corporation|Atom Processor D4xx/D5xx/N4xx/N5xx Integrated Graphics Controller [DISPLAY_VGA] (vendor:8086 device:a011 subv:1043 subd:83ac)
unknown         : Intel Corporation|Atom Processor D4xx/D5xx/N4xx/N5xx DMI Bridge [BRIDGE_HOST] (vendor:8086 device:a010 subv:1043 subd:83ac)
hub             : Linux 3.12.8-desktop586-2.mga4 uhci_hcd|UHCI Host Controller [Hub|Unused|Full speed (or root) hub] (vendor:1d6b device:0001)
hub             : Linux 3.12.8-desktop586-2.mga4 ehci_hcd|EHCI Host Controller [Hub|Unused|Full speed (or root) hub] (vendor:1d6b device:0002)
hub             : Linux 3.12.8-desktop586-2.mga4 uhci_hcd|UHCI Host Controller [Hub|Unused|Full speed (or root) hub] (vendor:1d6b device:0001)
hub             : Linux 3.12.8-desktop586-2.mga4 uhci_hcd|UHCI Host Controller [Hub|Unused|Full speed (or root) hub] (vendor:1d6b device:0001)
usbhid          : eGalax Inc.|USB TouchController [Human Interface Device|No Subclass|None] (vendor:0eef device:480d)
hub             : Linux 3.12.8-desktop586-2.mga4 uhci_hcd|UHCI Host Controller [Hub|Unused|Full speed (or root) hub] (vendor:1d6b device:0001)
usbhid          : Logitech|USB Receiver [Human Interface Device|Boot Interface Subclass|Keyboard] (vendor:046d device:c52b)
hid_multitouch  : eGalax Inc. USB TouchController
hid_logitech_dj : Logitech USB Receiver
hid_logitech_dj : Logitech USB Receiver
hid_logitech_dj : Logitech USB Receiver
hid_logitech_dj : Logitech Unifying Device. Wireless PID:402d

CC: (none) => scarabeille

@Scara: You told me there was also a similar issue when running "urpmi --auto-select" IIRC. Could you post the output of this too?
As far as I understand, it would point towards an issue in urpmi (or a specific mirror?) instead of mgaapplet.

Here is (I cut the main part of the message which is always like the beginning but whith other package names)

How Can I know the mirror I'm using? (because I couldn't choose it when I added sourcs)

urpmi --auto-select
invalid line <@provides@kernel[== 3.14.18-3.mga4]@kernel[== 3.14]@alsa[== 1.0.26]@drbd-api[== 88]@kernel-server@should-restart[== system]@kernel-server-3.14.18-3.mga4[== 1-1.mga4]@kernel-server-3.14.18-3.mga4(x86-32)[== 1-1.mga4]@kmod(3c509)@kmod(3c515)[== 0.99t][...]
problème de lecture du fichier de synthèse du média « Core Updates (distrib3) »
Can't call method "flag_skip" on an undefined value at /usr/lib/perl5/vendor_perl/5.18.1/i386-linux-thread-multi/URPM/Resolve.pm line 582.

@Scara: you can use "urpmq --list-url" to see the URLs of your mirror(s).

CC'ing tmb since the error speaks about the kernel, though I'm not sure it's related to it specifically.

CC: (none) => tmb

I was on distrib-coffee and I changed the mirror to http://fr2.rpmfind.net/

the problem is always the same

here is the pastbin with the complete error : http://pastebin.com/assvgfqT

If it can help: 

(note: I have the same bug on 2 differents computers for which I did a fresh install this week-end)

# lspcidrake -v
ath9k           : Qualcomm Atheros|AR9285 Wireless Network Adapter (PCI-Express) [NETWORK_OTHER] (vendor:168c device:002b subv:1a3b subd:1089) (rev: 01)
atl1c           : Qualcomm Atheros|AR8132 Fast Ethernet [NETWORK_ETHERNET] (vendor:1969 device:1062 subv:1043 subd:838a) (rev: c0)
unknown         : Intel Corporation|NM10/ICH7 Family SATA Controller [AHCI mode] [STORAGE_SATA] (vendor:8086 device:27c1 subv:1043 subd:83ad) (rev: 02)
lpc_ich         : Intel Corporation|NM10 Family LPC Controller [BRIDGE_ISA] (vendor:8086 device:27bc subv:1043 subd:83ad) (rev: 02)
unknown         : Intel Corporation|82801 Mobile PCI Bridge [BRIDGE_PCI] (vendor:8086 device:2448) (rev: e2)
ehci_pci        : Intel Corporation|NM10/ICH7 Family USB2 EHCI Controller [SERIAL_USB] (vendor:8086 device:27cc subv:1043 subd:83ad) (rev: 02)
uhci_hcd        : Intel Corporation|NM10/ICH7 Family USB UHCI Controller #4 [SERIAL_USB] (vendor:8086 device:27cb subv:1043 subd:83ad) (rev: 02)
uhci_hcd        : Intel Corporation|NM10/ICH7 Family USB UHCI Controller #3 [SERIAL_USB] (vendor:8086 device:27ca subv:1043 subd:83ad) (rev: 02)
uhci_hcd        : Intel Corporation|NM10/ICH7 Family USB UHCI Controller #2 [SERIAL_USB] (vendor:8086 device:27c9 subv:1043 subd:83ad) (rev: 02)
uhci_hcd        : Intel Corporation|NM10/ICH7 Family USB UHCI Controller #1 [SERIAL_USB] (vendor:8086 device:27c8 subv:1043 subd:83ad) (rev: 02)
shpchp          : Intel Corporation|NM10/ICH7 Family PCI Express Port 4 [BRIDGE_PCI] (vendor:8086 device:27d6) (rev: 02)
shpchp          : Intel Corporation|NM10/ICH7 Family PCI Express Port 2 [BRIDGE_PCI] (vendor:8086 device:27d2) (rev: 02)
shpchp          : Intel Corporation|NM10/ICH7 Family PCI Express Port 1 [BRIDGE_PCI] (vendor:8086 device:27d0) (rev: 02)
snd_hda_intel   : Intel Corporation|NM10/ICH7 Family High Definition Audio Controller [MULTIMEDIA_AUDIO_DEV] (vendor:8086 device:27d8 subv:1043 subd:83ce) (rev: 02)
unknown         : Intel Corporation|Atom Processor D4xx/D5xx/N4xx/N5xx Integrated Graphics Controller [DISPLAY_OTHER] (vendor:8086 device:a012 subv:1043 subd:83ac)
Card:Intel 810 and later: Intel Corporation|Atom Processor D4xx/D5xx/N4xx/N5xx Integrated Graphics Controller [DISPLAY_VGA] (vendor:8086 device:a011 subv:1043 subd:83ac)
unknown         : Intel Corporation|Atom Processor D4xx/D5xx/N4xx/N5xx DMI Bridge [BRIDGE_HOST] (vendor:8086 device:a010 subv:1043 subd:83ac)
hub             : Linux 3.12.8-desktop586-2.mga4 ehci_hcd|EHCI Host Controller [Hub|Unused|Full speed (or root) hub] (vendor:1d6b device:0002)
hub             : Linux 3.12.8-desktop586-2.mga4 uhci_hcd|UHCI Host Controller [Hub|Unused|Full speed (or root) hub] (vendor:1d6b device:0001)
hub             : Linux 3.12.8-desktop586-2.mga4 uhci_hcd|UHCI Host Controller [Hub|Unused|Full speed (or root) hub] (vendor:1d6b device:0001)
hub             : Linux 3.12.8-desktop586-2.mga4 uhci_hcd|UHCI Host Controller [Hub|Unused|Full speed (or root) hub] (vendor:1d6b device:0001)
usbhid          : Logitech|USB Receiver [Human Interface Device|Boot Interface Subclass|Keyboard] (vendor:046d device:c52b)
usbhid          : eGalax Inc.|USB TouchController [Human Interface Device|No Subclass|None] (vendor:0eef device:480d)
hub             : Linux 3.12.8-desktop586-2.mga4 uhci_hcd|UHCI Host Controller [Hub|Unused|Full speed (or root) hub] (vendor:1d6b device:0001)
hid_multitouch  : eGalax Inc. USB TouchController
hid_logitech_dj : Logitech USB Receiver
hid_logitech_dj : Logitech USB Receiver
hid_logitech_dj : Logitech USB Receiver
hid_logitech_dj : Logitech Unifying Device. Wireless PID:402d

For the "problème de lecture du fichier de synthèse synthèse du média « Core Updates (distrib3)" or can't read the file...

I 'think' you could try removing corresponding file in /var/lib/urpmi (cache + maybe other)

For some other people, updating with urpmi --auto-update was enough to remove the "Can't call method "flag_skip" on an undefined value" issue

Thierry, another view ?

I removed all files and folders in /var/lib/urpmi and configured the packages sources again... and I have the same problem for updating the distro :(

@Scara: Thanks to a hint from spurtle on IRC, I think you are affected by bug 13370.

It would be the case if your perl-URPM package is older than perl-URPM-4.30-1.mga4, can you check? Did you install Mageia 4 instead of Mageia 4.1?

I installed with the live DVD that I downloaded when mageia 4 was out and the version of perl-URPM was 4.29-1

I downloaded perl-URPM-4.30-1.mga4.i586.rpm package and the problem is solved 

thanks :))

Assigning to Thomas: Please see the pastebin on comment 13 and my comment 17. From what I could gather users installing Mageia 4 (and not 4.1) can not update because kernel 3.14.x triggers bug 13370.

Maybe we should remove some provides for the Mageia 4 kernel?

Assignee: bugsquad => tmb

One solution could be to drop the versions from the kmod provides (like we do in mga5)

Yeah, I think we can try to remove the version from the kmod provides...

for now its an issue that can be solved on infra_* level, but if it works we need to issue an update for atleast mga4

The only kmod versions we cant drop are the dkms specific ones especially for gpus

(In reply to Thomas Backlund from comment #21)
> Yeah, I think we can try to remove the version from the kmod provides...
> 
> for now its an issue that can be solved on infra_* level,

Actually scratch that... after getting some coffee and waking the brain...
the kmod provides is generated in chroots at kernel build time, so we cant
really fix it with anyting other than an errata note about manually downloading the perl-URPM referenced in http://advisories.mageia.org/MGAA-2014-0117.html


the 3.14 kernel rpms with "too big provides list" are already on public mirrors so we cant really "fix" them anymore

Can't we provide a "bugfix" update for the kernel? Or will the already validated kernels still be an issue even if they're not the ones urpmi will try to install (when there are newer kernels in core/updates I mean)?

Technically one could maybe hack the genhdlist2 on infra_* to drop the versions from the kmod provides at hdlist generation time, but I think it would be too error prone...

(In reply to Rémi Verschelde from comment #24)
> Can't we provide a "bugfix" update for the kernel? Or will the already
> validated kernels still be an issue even if they're not the ones urpmi will
> try to install (when there are newer kernels in core/updates I mean)?

Thats exactly the issue...  the already validated 3.14 kernel has the "too big provides list" already...

Hmm... that gives one other idea... not really nice one, but maybe acceptable in this case...

the only 3.14 kernel in updates is the core kernel-3.14.18-3.mga4, the other ones are only in updates_testing.

So we could "break" the chain with pushing an updated rpm-mageia-setup-build with dropped kmod version provides, rebuild all kernels in updates_testing with it, and when we release the new kernel-3.14.22+ we drop the 3.14.18 from updates tree

or I'd have to rebuild the 3.14.18-3 and let it replace the one already validated one

*** Bug 14414 has been marked as a duplicate of this bug. ***

CC: (none) => junknospam

I decided to do it a little combo to satisfy all possibilities...

1.push rpm-mageia-setup-1.197-1.1.mga4 that dont add versions to kernel kmod provides
2. rebuild already validated kernel-3.14.18-3 with the fixed rpm-mageia-setup-build and move it to updates replacing the current one (as toolchain has not changed, the only change is the "lighter" kmod provides list (it's even lighter than the kernel-3.12 series), so the replacement is safe)
3. regenerate hdlists to pick up the lighter provides.

So this bug should now be resolved, and all kernels currently in updates_testing are updated to 3.14.23 and rebuilt with the fixed rpm-mageia-setup-build.

Resolution: (none) => FIXED
Status: NEW => RESOLVED

*** Bug 14568 has been marked as a duplicate of this bug. ***

CC: (none) => linuxgal

I get today the same bug.
The "core updates" list can't be read.
I just done a fresh install from DVD-i586 4.1, configuring the mirror. I get the error asking the update, for example with "urpmi --auto-select".
I tried with $MIRRORLIST and then with two other specific mirrors.

CC: (none) => yves.brungard_mageia

*** Bug 14225 has been marked as a duplicate of this bug. ***

CC: (none) => contact

*** Bug 14641 has been marked as a duplicate of this bug. ***

CC: (none) => nmrk.n

*** Bug 14338 has been marked as a duplicate of this bug. ***

CC: (none) => anita_de_jongh

*** Bug 13331 has been marked as a duplicate of this bug. ***

CC: (none) => magnus.mud

*** Bug 14382 has been marked as a duplicate of this bug. ***

CC: (none) => pietroamadio.deltoso

*** Bug 14316 has been marked as a duplicate of this bug. ***

*** Bug 14292 has been marked as a duplicate of this bug. ***

CC: (none) => rv.lesmond

*** Bug 14291 has been marked as a duplicate of this bug. ***

CC: (none) => nextwavecs

*** Bug 14290 has been marked as a duplicate of this bug. ***

*** Bug 14278 has been marked as a duplicate of this bug. ***

CC: (none) => andybiker

*** Bug 14507 has been marked as a duplicate of this bug. ***

CC: (none) => n1kio

*** Bug 14388 has been marked as a duplicate of this bug. ***

CC: (none) => mageia

*** Bug 14564 has been marked as a duplicate of this bug. ***

CC: (none) => vohra64

From Thierry Vignaud in https://bugs.mageia.org/show_bug.cgi?id=13331#c37
> This is already fixed.
> 
> If you still have the bug, you need to manually run (as root):
> urpmi.update -a
> urpmi perl-URPM

Still, this shows that there is still an issue with the current Core Updates synthesis file. The perl-URPM update was meant to fix upgrades from Mageia 4 to Mageia 5, but we keep getting reports of users facing the issue when updating a brand new Mageia 4.1.

@tmb: Is there still an issue with the kernel provides?

Thomas forced the regeneration of the hdlists, which fixes the issue.

People might still be faced with this issue if they can't update their local copy of the hdlists. A mention has been added for this in the Errata: https://wiki.mageia.org/en/Mageia_4_Errata#Package_installation_or_update_fails_with_.22problem_reading_synthesis_file.22