AIX Open Source

 View Only
  • 1.  dnf update "successfull" but "Segmentation fault(coredump)"

    Posted Mon August 22, 2022 09:46 AM
      |   view attached
    Hello,
    Please have a look to the enclosed file.
    The "dnf update" is successfull but we have a return code 139 and a core dump.
    Do we have to be worried about this ?
    We ran "dnf check" two times.
    We ran "dnf update" a  second time but no update was available. Message "Nothing to do".
    We thank you very much four help,
    Christophe

    ps : Maybe it comes from our AIX Operating System LPAR.

    ------------------------------
    christophe derouet
    ------------------------------

    Attachment(s)



  • 2.  RE: dnf update "successfull" but "Segmentation fault(coredump)"

    Posted Thu September 01, 2022 01:07 PM
    Hi Christophe,

    Sorry for delay in response. With the attached output it is hard to tell the core dump was from which application.
    As dnf internally calls many other libraries and programs. 
    As you dnf check and update running later are fine so I don't think you have any issue.
    But it is good to understand what caused the first core dump.
    If you can attach the core file and provide more information about your system.
    1. What level of AIX
    2. Output of "rpm -qa"
    3. output of "lslpp -l rpm.rte"


    ------------------------------
    SANKET RATHI
    ------------------------------



  • 3.  RE: dnf update "successfull" but "Segmentation fault(coredump)"

    Posted Tue September 06, 2022 05:38 AM
    Hello Sanket,

    I work with Christophe,

    Our test were made on an lpar who was upadted last days : (7200-05-04-2220).

    I had the core dump too on an 7200-05-03-2136.

    I reinstall DNF today after migrate from 7200-05-03-2136 to 7200-05-04-2220 and i dont reproduce the core dump.

    So i send you the core i had last month in 7200-05-03-2136 (Christophe had the same core in same OS aix version).

    Thank you very much 

    regards

    Nicolas

    ------------------------------
    Nicolas KAPLIN
    ------------------------------



  • 4.  RE: dnf update "successfull" but "Segmentation fault(coredump)"

    Posted Tue September 06, 2022 05:41 AM
      |   view attached
    The core file...

    ------------------------------
    Nicolas KAPLIN
    ------------------------------

    Attachment(s)

    rar
    core.rar   5.15 MB 1 version


  • 5.  RE: dnf update "successfull" but "Segmentation fault(coredump)"

    Posted Mon September 12, 2022 07:45 AM
    Hi Nicolas,

    Thanks for the core file.

    It seems to be a memory allocation issue for one of the glib2 object at first glance.
    It also doesn't seem to be consistent and we haven't seen this kind of core dump in our environment.


    Core was generated by `python3_32'.
    Program terminated with signal SIGSEGV, Segmentation fault.
    #0 0xd7a09e40 in allocator_slab_stack_push (allocator=0xd7a09fc8 <allocator_add_slab+104>, ix=4053775540,
    sinfo=0xb080ee4c) at gslice.c:1252
    1252 gslice.c: No such file or directory.
    [Current thread is 1 (process 1)]
    (gdb) bt
    #0 0xd7a09e40 in allocator_slab_stack_push (allocator=0xd7a09fc8 <allocator_add_slab+104>, ix=4053775540,
    sinfo=0xb080ee4c) at gslice.c:1252
    #1 0xd7a09fc8 in allocator_add_slab (allocator=0xd7a0a254 <slab_allocator_alloc_chunk+132>, ix=811461976,
    chunk_size=804391072) at gslice.c:1285
    #2 0xd7a0a254 in slab_allocator_alloc_chunk (chunk_size=3617623428) at gslice.c:1330
    #3 0xd7a08d84 in magazine_cache_pop_magazine (ix=3617624504, countp=0x1) at gslice.c:726
    #4 0xd7a091b8 in thread_memory_magazine1_reload (tmem=0xd7a096fc <g_slice_alloc+168>, ix=804391416)
    at gslice.c:801
    #5 0xd7a096fc in g_slice_alloc (mem_size=3617626080) at gslice.c:1014
    #6 0xd7a097e0 in g_slice_alloc0 (mem_size=3618268168) at gslice.c:1051
    #7 0xd7aa6408 in g_type_create_instance (type=3618197140) at gtype.c:1848
    #8 0xd7a94e94 in g_object_new_internal (class=0xd7a955b8 <g_object_new_with_properties+716>,
    params=0x30504858, n_params=1) at gobject.c:1799
    ........

    (gdb) p prev
    $6 = (SlabInfo *) 0x0
    (gdb)

    ....
    0xd7a09e3c <+140>: l r10,96(r31)
    => 0xd7a09e40 <+144>: st r10,8(r9)
    0xd7a09e44 <+148>: l r9,96(r31)
    .......

    (gdb) p $r9
    $3 = 0

    Do you have ulimit values set to unlimited ?

    I will try to find some more info from the core file.


    ------------------------------
    SANGAMESH
    ------------------------------



  • 6.  RE: dnf update "successfull" but "Segmentation fault(coredump)"

    Posted Tue September 20, 2022 10:48 AM
    Hi Sangamesh,

    Maybe it can help you.

    I face this issue too on most of my AIX servers on which I'm installing dnf.
    I have the same message after the first "dnf update" at the end of the execution.

      Verifying        : python3-dnf-plugin-migrate-4.0.16-32_50.ppc        116/123 
      Verifying        : python3-dnf-plugin-migrate-4.0.16-32_1.noarch      117/123 
      Verifying        : python3-dnf-plugins-core-4.0.16-32_50.ppc          118/123 
      Verifying        : python3-dnf-plugins-core-4.0.16-32_1.noarch        119/123 
      Verifying        : python3-hawkey-0.39.1-32_50.ppc                    120/123 
      Verifying        : python3-hawkey-0.39.1-32_2.ppc                     121/123 
      Verifying        : python3-libdnf-0.39.1-32_50.ppc                    122/123 
      Verifying        : python3-libdnf-0.39.1-32_2.ppc                     123/123 
    Memory fault(coredump)
    
    warning: rpmdb: Freeing read locks for locker 0x1d1: 15204740/1
    warning: rpmdb: Freeing read locks for locker 0x1d3: 15204740/1
    warning: rpmdb: Freeing read locks for locker 0x1d4: 15204740/1
    warning: rpmdb: Freeing read locks for locker 0x1d5: 15204740/1
    warning: rpmdb: Freeing read locks for locker 0x1d6: 15204740/1
    warning: rpmdb: Freeing read locks for locker 0x1d7: 15204740/1
    warning: rpmdb: Freeing read locks for locker 0x1d8: 15204740/1
    warning: rpmdb: Freeing read locks for locker 0x1d9: 15204740/1
    warning: rpmdb: Freeing read locks for locker 0x1da: 15204740/1
    warning: rpmdb: Freeing read locks for locker 0x1db: 15204740/1
    warning: rpmdb: Freeing read locks for locker 0x1dc: 15204740/1
    warning: rpmdb: Freeing read locks for locker 0x1dd: 15204740/1
    warning: rpmdb: Freeing read locks for locker 0x1de: 15204740/1
    warning: rpmdb: Freeing read locks for locker 0x1df: 15204740/1
    warning: rpmdb: Freeing read locks for locker 0x1e0: 15204740/1
    warning: rpmdb: Freeing read locks for locker 0x1e1: 15204740/1
    warning: rpmdb: Freeing read locks for locker 0x1e2: 15204740/1
    warning: rpmdb: Freeing read locks for locker 0x1e3: 15204740/1
    warning: rpmdb: Freeing read locks for locker 0x1e4: 15204740/1​

    But everything seems to be OK.
    I just noted that there is a full memory consumption during the execution of the "dnf update".

    Our AIX version is 7200-05-02-2114.
    rpm.rte is 4.15.1.3

    These are the list of rpms installed after "dnf update" (most from IBM, other are customs with no dependencies) :
    xxxx-nrpe-2.15.0-9.noarch
    hmsu-1.4.1-0.ppc
    lpar2rrd-agent-6.15-2.ppc
    rubrik-agent-5.3.2.p3.19174.HPD179046-1.0_aix7.1.ppc
    rpm-python3-4.15.1-32_1.ppc
    python3-six-1.13.0-1.noarch
    AIX-rpm-7.2.5.1-16.ppc
    rsync-3.2.3-1.ppc
    libgcc-8-1.ppc
    libffi-3.4.2-1.ppc
    libzstd-1.4.4-100.ppc
    ncurses-6.2-3.ppc
    expat-2.4.6-1.ppc
    libtasn1-4.16.0-1.ppc
    check-0.13.0-1.ppc
    db-5.3.28-1.ppc
    libgomp-8-1.ppc
    libxml2-2.9.14-1.ppc
    libunistring-0.9.10-1.ppc
    gettext-0.21-1.ppc
    info-6.7-1.ppc
    libassuan-2.5.5-1.ppc
    sqlite-3.37.2-1.ppc
    libgcrypt-1.9.4-1.ppc
    p11-kit-tools-0.23.22-1.ppc
    libksba-1.4.0-1.ppc
    nettle-3.7.3-1.ppc
    gdbm-1.23-1.ppc
    krb5-libs-1.18.4-1.ppc
    python3-libcomps-0.1.11-100.ppc
    openldap-2.4.58-3.ppc
    curl-7.79.1-1.ppc
    libsolv-0.7.9-32_51.ppc
    libmodulemd-1.5.2-100.ppc
    gnupg2-2.2.35-1.ppc
    librepo-1.11.0-100.ppc
    python3-libdnf-0.39.1-32_50.ppc
    python3-gpg-1.13.1-100.ppc
    dnf-data-4.2.17-32_50.ppc
    dnf-4.2.17-32_50.ppc
    python3-dnf-plugin-migrate-4.0.16-32_50.ppc
    jq-1.6-1.ppc
    zchunk-1.1.4-101.ppc
    python3-devel-3.7.12-1.ppc
    python3-test-3.7.12-1.ppc
    lz4-1.9.3-1.ppc
    sudo-1.6.9p23-2noldap.ppc
    MFEcma-4.8.0-1995.ppc
    puppet-agent-6.17.0-1.ppc
    python3-dateutil-2.8.0-1.noarch
    libgcc8-8.3.0-6.ppc
    zlib-1.2.12-1.ppc
    libstdc++8-8.3.0-6.ppc
    libstdc++-8-1.ppc
    bzip2-1.0.8-2.ppc
    json-c-0.15-2.ppc
    xz-libs-5.2.5-1.ppc
    gmp-6.2.1-1.ppc
    libgomp8-8.3.0-6.ppc
    libtextstyle-0.21-1.ppc
    glib2-2.56.1-3.ppc
    libiconv-1.17-1.ppc
    bash-5.1.4-2.ppc
    libgpg-error-1.42-1.ppc
    readline-8.1-1.ppc
    libsmartcols-2.34-100.ppc
    p11-kit-0.23.22-1.ppc
    ca-certificates-2021.2.52-3.ppc
    pinentry-1.0.0-1.ppc
    gnutls-3.7.1-1.ppc
    python3-3.7.12-1.ppc
    libcomps-0.1.11-100.ppc
    cyrus-sasl-2.1.28-1.ppc
    libssh2-1.10.0-1.ppc
    zchunk-libs-1.1.4-101.ppc
    libyaml-0.2.5-1.ppc
    npth-1.5-1.ppc
    gpgme-1.13.1-100.ppc
    libdnf-0.39.1-32_50.ppc
    python3-hawkey-0.39.1-32_50.ppc
    oniguruma-6.9.6-1.ppc
    python3-dnf-4.2.17-32_50.ppc
    python3-dnf-plugins-core-4.0.16-32_50.ppc
    dnf-automatic-4.2.17-32_50.ppc
    python3-librepo-1.11.0-100.ppc
    zchunk-devel-1.1.4-101.ppc
    python3-docs-3.7.12-1.ppc
    python3-tools-3.7.12-1.ppc

    regards.




    ------------------------------
    Damien LEMAGNEN
    AIX System Engineer/Architect
    NETIXWARE
    ------------------------------



  • 7.  RE: dnf update "successfull" but "Segmentation fault(coredump)"

    Posted Tue September 20, 2022 03:02 PM
    Hi Damien,

    Please let me know your AIX server configuration.

    ------------------------------
    SANGAMESH
    ------------------------------