[SOLVED] makepkg -s with nested AUR deps on "userless" systems

hi..
so for the last two years i was able to solve every Arch Linux problem i had from infos around the wiki, the forum, etc. now i am stuck.
here's the scenario:
i am automating installation of a full systems for headless usage.
no user required and/or wanted.
this works perfectly until i need something special from the AUR.
because if a package from the AUR has no dependencies we can run makepkg as user nobody:
..# sudo -u nobody makepkg
but as soon as we need
..# sudo -u nobody makepkg -s
to resolve dependencies from the AUR this fails. "nobody" has no password, therefore we can not sudo.
my preferred idea would be to create a shortlived dummyuser. but how to get this dummy to behave correctly with creation and entering of the password without actual user interaction? (remember: headless automation)
i would really like to avoid the special workaround of running my own build-server with all the packages i need on different target machines, putting those made packages somewhere on the internet and then be able to download them to the designated machine and "pacman -U" them there.
i hope someone with a similar use-case can give me some hints.
cheers,
grubernd
Last edited by grubernd (2015-05-16 17:17:58)

thanks.
progandy wrote:…or if you use su, then allow password-free su execution for the wheel group (/etc/pam.d/su) and…
nice usage of su, will have to dig more into that and i should be able to not need sudo as an extra package.
but i am really trying to avoid the NOPASSWD route for security reasons.
apologies, i forgot to mention that in my original post.
Trilby wrote:
Why not just script it:
source PKGBUILD
pacman -S --asdeps $depends $makedepends
sudo -u nobody makepkg
thanks. seeing it written down differently showed me the flaw in my wishful thinking:
makepkg can of course resolve dependencies but it cannot ever resolve dependencies where one package in the AUR requires another un-made package from the AUR. just doesn't work. that was the block i had.
although not really beautiful i will just drill down the package-chain and resolve dependencies myself in the installer scripts.
(marking thread as solved)

Similar Messages

  • [SOLVED] Makepkg, install dependencies from AUR.

    Is it possible for makepkg to install the dependencies from AUR with makepkg -si?
    Last edited by lemanski6 (2012-09-20 18:05:38)

    Nope.  Many "AUR helpers" do this though.  In my limited experience yaourt seems to be one that handles this seemlessly, but I'd say it comes with some costs of hiding what is doing from the user.

  • [SOLVED] Saving PKGBUILDs with makepkg or AUR helpers

    Hi all,
    I am trying to save automatically the package PKGBUILD when I install a package with an aur helper. I use packer and aurget.
    I modified makepkg.conf and set this relevant variables:
    # PACKAGE OUTPUT
    # Default: put built package and cached source in build directory
    #-- Destination: specify a fixed directory where all packages will be placed
    #PKGDEST=/home/packages
    PKGDEST=/var/cache/pacman/pkg/aur
    #-- Source cache: specify a fixed directory where source files will be cached
    #SRCDEST=/home/sources
    SRCDEST=/var/cache/pacman/pkg/aur/src
    The PKGDEST works fine and packages are saved in the specified path.
    However SRCDEST variable doesn't seem to work, as I expected the PKBUILDs and other sources to be saved there, but there are no files in the path.
    In the end what I am trying to emulate is something similar to ABS but for installed packages from AUR.
    Last edited by adrianbs (2012-01-29 12:18:41)

    adrianbs wrote:
    jasonwryan wrote:
    Cower (as awesome as it is) doesn't open the PKGBUILD in vim, build the package and then ask if you want to keep the build directory.
    I reckon my 10 lines are pretty well spent...
    Cower doesn't but pacaur (which wraps it) does
    It has an option it the config file to keep the build dir.
    And for only an extra 872 lines of code...

  • Issues with nested for loops - saving images from a camera

    Hi all,
    I've written a vi. to capture a specific number of images ('Image No') and save these images, outputted to a folder of my choice.  Each image is identified sequentially.  However, I wish to do a number of iterations ('Run') of this capture sequence, such that the filename of each image would be 'Filename (Run)_(Image No).png', e.g. run 5, image 10 would be 'Filename 5_10.png'.  I have tried a nested for loop for this but I receive an error 'Asynchronous I/O operation in progress' (I've attached a printscreen).
    Can anyone assist me in solving this problem? I preiously posted this in machine Vision but got no response (http://forums.ni.com/t5/Machine-Vision/Capturing-image-sequences-issues-with-nested-for-loops/m-p/19...).  Please find attached my vi.
    Kindest regards and thanks,
    Miika
    Solved!
    Go to Solution.
    Attachments:
    Labview problem.jpg ‏3841 KB
    Image sequence save to file.vi ‏48 KB

    Miika,
    the problem is not the filenam, but the name of the folder (AHHHHH!). You try to create the same folder in the outer for loop over and over again.... (it is the error message above the '======', not below )
    Norbert
    CEO: What exactly is stopping us from doing this?
    Expert: Geometry
    Marketing Manager: Just ignore it.

  • [SOLVED]What to do when AUR provides source files, but no 'configure'?

    Not sure if this is better posted here in "Newbie Corner" or in the "AUR" forum.... but I suppose it's a newbie-type question.
    There's a package I've installed from AUR on another Arch system and the install process went smoothly then.  New box; everything else is working fine and dandy, but need help with a package found in AUR.
    The package  installs source files -- no binaries available -- and there's no 'configure' file.  There's a 'configure.in' file that gets parsed by 'autoconf', but there are errors, and no 'configure' gets created.
    The PDF documentation from the author implies there's a 'configure' already prepared (the author writes the standard "Enter these commands:  ./configure && make install"), and located in the src directory.
    Is this an issue for the Arch package maintainer, or for the upstream author?  And might I be pointed to a suitable resource to continue my own investigation?
    Thanks!
    Last edited by fionn (2013-10-06 19:49:25)

    skottish wrote:
    fionn wrote:
    Trilby wrote:You *might* be looking for `makepkg -o` which is what gets the actual source code for the package.
    Oh, Cr*p.  This means I need to do it all over again and burn another CD.
    I have a tinfoil-hat wearing conspiracy-theory nut of a brother-in-law who wants a new system, compiled from sources, "without Government Interference".  He wants an "air-gap" (not networked) box to prevent NSA boogeymen from monkeying around with our Precious Bodily Fluids or some damned thing like that.  Why is it always a brother-in-law who goes off the deep end?  Why can't I have one who's a lawyer or doctor or something useful?
    You know, you may gently want to look into the idea that you're taking advice from a hat wearing bird. I'm not suggesting anything, but it is something that you may want to consider.
    On the upside, it does look to be an aluminum-free chapeau...

  • [Solved] Makepkg hanging on checking runtime dependencies.

    Hey guys, as above really.
    Recently I've been having problems with makepkg and installing from AUR. It seems to fetch dependencies from the official repositories fine, but when makepkg goes to compile the PKGBUILD, it hangs at '==> checking runtime dependencies'.
    I've checked htop and I've a yaourt process running like:
    /bin/bash /usr/bin/yaourt -T gtk-perl
    the process is using around 40% CPU and even if I leave it running for a while, nothing happens.
    I've made no system changes (besides changing http to https in the yaourt config to stop the 301 error after AUR came out of maintenance https://bbs.archlinux.org/viewtopic.php?pid=1182722) since this started happening so I'm pretty stumped to what it could be.
    Any ideas?
    edit: It works if I run as root, strange, but not ideal.
    Last edited by dunks (2012-10-28 01:38:46)

    I do not use yaourt.  We have been having issues with the AUR server over the last couple days.  It is back up now on a new server, but there were issues relating to that move.  Does this thread have any bearing on the issue?
    Edit: Never mind.  I am distracted and get low scores for comprehension today
    Last edited by ewaller (2012-10-27 23:14:57)

  • [SOLVED] Cannot build openswan from AUR

    Hello everybody,
    I have an error when I try to upgrade openswan with yaourt/AUR (2.6.24-2 => 2.6.25-1) on a i686 arch install
    This is the error :
    make[3]: Entering directory `/tmp/yaourt-tmp-olivier/aur-openswan/openswan/src/openswan-2.6.25/OBJ.linux.i386/programs/addconn'
    cc -c -fno-strict-aliasing -g -Werror -DKLIPS -DSCANDIR_HAS_CONST -I/tmp/yaourt-tmp-olivier/aur-openswan/openswan/src/openswan-2.6.25/ports/linux/include -I/tmp/yaourt-tmp-olivier/aur-openswan/openswan/src/openswan-2.6.25/ports/linux/include -I/tmp/yaourt-tmp-olivier/aur-openswan/openswan/src/openswan-2.6.25/ports/linux/include -I/tmp/yaourt-tmp-olivier/aur-openswan/openswan/src/openswan-2.6.25/ports/linux/include -I/tmp/yaourt-tmp-olivier/aur-openswan/openswan/src/openswan-2.6.25/programs/pluto -I/tmp/yaourt-tmp-olivier/aur-openswan/openswan/src/openswan-2.6.25/programs/addconn/ -I/tmp/yaourt-tmp-olivier/aur-openswan/openswan/src/openswan-2.6.25 -I/tmp/yaourt-tmp-olivier/aur-openswan/openswan/src/openswan-2.6.25/linux/include -I/tmp/yaourt-tmp-olivier/aur-openswan/openswan/src/openswan-2.6.25/include -Wall -Wpointer-arith -Wcast-qual -Wstrict-prototypes -Wbad-function-cast -DKLIPS -DNAT_TRAVERSAL -DNAT_TRAVERSAL -DKERNEL_ALG -DIKE_ALG -DFINALCONFDIR=\"/etc\" -DFINALCONFDDIR=\"/etc/ipsec.d\" -DFINALCONFFILE=\"/etc/ipsec.conf\" -DFINALVARDIR=\"/var\" -Werror /tmp/yaourt-tmp-olivier/aur-openswan/openswan/src/openswan-2.6.25/programs/addconn/addconn.c
    cc1: warnings being treated as errors
    In file included from /tmp/yaourt-tmp-olivier/aur-openswan/openswan/src/openswan-2.6.25/programs/addconn/addconn.c:51:0:
    /tmp/yaourt-tmp-olivier/aur-openswan/openswan/src/openswan-2.6.25/include/ipsecconf/confread.h:37:19: error: comparison between 'enum keyword_string_config_field' and 'enum keyword_string_conn_field'
    /tmp/yaourt-tmp-olivier/aur-openswan/openswan/src/openswan-2.6.25/include/ipsecconf/confread.h:38:19: error: comparison between 'enum keyword_numeric_config_field' and 'enum keyword_numeric_conn_field'
    /tmp/yaourt-tmp-olivier/aur-openswan/openswan/src/openswan-2.6.25/include/ipsecconf/confread.h:39:34: error: comparison between 'enum keyword_string_config_field' and 'enum keyword_string_conn_field'
    /tmp/yaourt-tmp-olivier/aur-openswan/openswan/src/openswan-2.6.25/include/ipsecconf/confread.h:40:34: error: comparison between 'enum keyword_numeric_config_field' and 'enum keyword_numeric_conn_field'
    make[3]: *** [addconn.o] Error 1
    make[3]: Leaving directory `/tmp/yaourt-tmp-olivier/aur-openswan/openswan/src/openswan-2.6.25/OBJ.linux.i386/programs/addconn'
    make[2]: *** [programs] Error 1
    make[2]: Leaving directory `/tmp/yaourt-tmp-olivier/aur-openswan/openswan/src/openswan-2.6.25/OBJ.linux.i386/programs'
    make[1]: *** [programs] Error 1
    make[1]: Leaving directory `/tmp/yaourt-tmp-olivier/aur-openswan/openswan/src/openswan-2.6.25/OBJ.linux.i386'
    make: *** [programs] Error 2
    ==> ERROR: Build Failed.
    Aborting...
    Error: Makepkg was unable to build openswan package.
    Error: unable to update openswan
    Following packages have not been installed:
    openswan
    Is a bug with the AUR package or is a problème on my install ?
    Thx for the help.
    Olive
    Last edited by nolive (2010-05-05 06:58:38)

    ok, that's build when I comment this in /tmp/yaourt-tmp-olivier/aur-openswan/openswan/src/openswan-2.6.25/programs/Makefile.program
    # die if there are any warnings
    ifndef WERROR
    #WERROR:= -Werror
    endif
    And run the compile directly (not with yaourt or pkgbuild) with
    sudo make programs install
    in /tmp/yaourt-tmp-olivier/aur-openswan/openswan/src/openswan-2.6.25
    But my ipsec VPN don't work :
    sudo ipsec auto --up toto
    003 ERROR: "toto" #1: sendto on eth0 to xxx.xxx.xxx.xxx:500 failed in main_outI1. Errno 22: Invalid argument
    104 "toto" #1: STATE_MAIN_I1: initiate
    010 "toto" #1: STATE_MAIN_I1: retransmission; will wait 20s for response
    003 ERROR: "toto" #1: sendto on eth0 to xxx.xxx.xxx.xxx:500 failed in EVENT_RETRANSMIT. Errno 22: Invalid argument
    010 "toto" #1: STATE_MAIN_I1: retransmission; will wait 40s for response
    003 ERROR: "toto" #1: sendto on eth0 to xxx.xxx.xxx.xxx:500 failed in EVENT_RETRANSMIT. Errno 22: Invalid argument
    Thank's for your help. I haven't the skills for checking the code.

  • Hello  Simple problem - don,t know how to solve it.  With Premiere CC when I try to do a selection (click... drag... release the click) very often it stop way before the end of the move I'm swinging the Magic Mouse. I taught that the mouse clicking was de

    Hello
    Simple problem - don,t know how to solve it.
    With Premiere CC when I try to do a selection (click... drag... release the click) very often it stop way before the end of the move I'm swinging the Magic Mouse. I taught that the mouse clicking was defective and went to get a new Magic Mouse after lots of frustration. Today, I have an edit to do it it does the SAME thing !!
    I was like ????#$%?&*(???
    Opened all the lights and taught I've trow the new mouse to the garbage and was using the defective mouse again... no !! - ??
    Actually, the bran new mouse is doing the same thing. What I understand after investigating on the motion and watching carefully my fingers !! -  is that when I click I have to keep my finger at the EXACT same place on the mouse... drag and release and it's fine. If I click by pushing on the mouse and my finder is moving of a 1/32th of a millimeter, it will release and my selection will be to redo. You can understand my frustration ! - 75$ later... same problem, but I know that if I click with about 5 pounds of pressure and trying to pass my finger through the plastic of the mouse, it you stay steady and make it !
    The problem is that scrolling is enable while clicking and it bugs.
    How to disable it ??
    Simple question - can't find the answer !

    Helllooo !?
    sorry but the Magic Mouse is just useless with the new Adobe Premiere CC and since I'm not the only one but can't find answer this is really disappointing. This mouse is just fantastic and now I have to swap from a USB mouse to the Magic Mouse every times I do some editing. My USB mouse if hurting my hand somehow and I want to got back to the Magic Mouse asap. Please - for sure there is a simple solution !
    Thanks !!

  • I just updated my latest java but the update is causing problems with some externale devices. So i would like to uninstall this latest java update and get back the previous one. That should solve to problems with my external device

    i just updated my latest java but the update is causing problems with some external devices. So i would like to uninstall this latest java update and get back the previous one. That should solve to problems with my external device.
    Is this possible and how do i do that?
    Anyone who responds thanks for that!
    Juko
    I am running
    Hardware Overview:
      Model Name:          Mac Pro
      Model Identifier:          MacPro1,1
      Processor Name:          Dual-Core Intel Xeon
      Processor Speed:          2,66 GHz
      Number of Processors:          2
      Total Number of Cores:          4
      L2 Cache (per Processor):          4 MB
      Memory:          6 GB
      Bus Speed:          1,33 GHz
      Boot ROM Version:          MP11.005D.B00
      SMC Version (system):          1.7f10
      Serial Number (system):          CK7XXXXXXGP
      Hardware UUID:          00000000-0000-1000-8000-0017F20F82F0
    System Software Overview:
      System Version:          Mac OS X 10.7.5 (11G63)
      Kernel Version:          Darwin 11.4.2
      Boot Volume:          Macintosh HD(2)
      Boot Mode:          Normal
      Computer Name:          Mac Pro van Juko de Vries
      User Name:          Juko de Vries (jukodevries)
      Secure Virtual Memory:          Enabled
      64-bit Kernel and Extensions:          No
      Time since boot:          11 days 20:39
    Message was edited by Host

    Java 6 you can't as Apple maintains it, and Java 7 you could if you uninstall it and Oracle provides the earlier version which they likely won't his last update fixed 37 remote exploits.
    Java broken some software here and there, all you'll have to do is wait for a update from the other parties.

  • TS1567 i have uninstalled then reinstalled itunes to solve a problem with apple mobile device, but i get this message during the install. Service  Apple Mobile Device  failed to start. Verify that you have sufficent priveleges to start system services.

    I Have reinstalled itunes to solve a problem with Apple mobile support, but during install i get this message. Service 'Apple Mobile, ( Apple Mobile Device ) failed to start. Verify that you have sufficent privileges to start system services. I'd appreciate any help on this.
    Thanks bluebanana 66

    Hi bluebanana 66,
    Thanks for visiting Apple Support Communities.
    There are a few reasons why you might receive this message after installing iTunes. You may need to verify that iTunes has been completely removed before trying to install again. See this article for the steps to completely remove and reinstall iTunes:
    Removing and reinstalling iTunes, QuickTime, and other software components for Windows Vista or Windows 7
    http://support.apple.com/kb/ht1923
    If the error message persists, see the "Additional troubleshooting" section of this article for some steps that can help solve it:
    Trouble installing iTunes or QuickTime for Windows
    http://support.apple.com/kb/ht1926
    Best,
    Jeremy

  • How to set the screen size with nested Iview screen

    Can someone show me how to set the screen size with nested IView  screen.
    at the moment i only try it with trial and error untill it looks somewhat close. Is there a property in VC to set the size for Height and width.
    The same question is for the form. the form elements inside the form is easy but the actual form is still a challenge.

    Hi,
    Yes you are correct ,Still the Form and iview are not able to change thier Layout Structure using a Propery.
    it is available with VC7.1
    Govindu

  • Custom taglib with nested tag not working

    Hi everybody,
    I have a question concerning a custom taglib. I want to create a tag with nested child tags. The parent tag is some kind of iterator, the child elements shall do something with each iterated object. The parent tag extends BodyTagSupport, i have overriden the methods doInitBody and doAfterBody. doInitBody initializes the iterator and puts the first object in the pageContext to be used by the child tag. doAfterBody gets the next iterator object and puts that in the pageContext, if possible. It returns with BodyTag.EVAL_BODY_AGAIN when another object is available, otherwise BodyTag.SKIP_BODY.
    The child tag extends SimpleTagSupport and does something with the given object, if it's there.
    In the tld-file I have configured both tags with name, class and body-content (tagdependent for the parent, empty for the child).
    The parent tag is being executed as I expected. But unfortunately the nested child tag does not get executed. If I define that one outside of its parent, it works fine (without object, of course).
    Can somebody tell me what I might have missed? Do I have to do something special with a nested tag inside a custom tag?
    Any help is greatly appreciated!
    Thanks a lot in advance!
    Greetings,
    Peter

    Hi again,
    unfortunately this didn't work.
    I prepared a simple example to show what isn't working. Perhaps it's easier then to show what my problem is:
    I have the following two tag classes:
    public class TestIterator extends BodyTagSupport {
        private Iterator testIteratorChild;
        @Override
        public void doInitBody() throws JspException {
            super.doInitBody();
            System.out.println("TestIterator: doInitBody");
            List list = Arrays.asList(new String[] { "one", "two", "three" });
            testIteratorChild = list.iterator();
        @Override
        public int doAfterBody() throws JspException {
            int result = BodyTag.SKIP_BODY;
            System.out.println("TestIterator: doAfterBody");
            if (testIteratorChild.hasNext()) {
                pageContext.setAttribute("child", testIteratorChild.next());
                result = BodyTag.EVAL_BODY_AGAIN;
            return result;
    public class TestIteratorChild extends SimpleTagSupport {
        @Override
        public void doTag() throws JspException, IOException {
            super.doTag();
            System.out.println(getJspContext().getAttribute("child"));
            System.out.println("TestIteratorChild: doTag");
    }The Iterator is the parent tag, the Child shall be shown in each iteration. My taglib.tld looks like the following:
    <?xml version="1.0" encoding="UTF-8"?>
    <taglib
         xsi:schemaLocation="http://java.sun.com/xml/ns/javaee web-jsptaglibrary_2_1.xsd"
         xmlns="http://java.sun.com/xml/ns/javaee"
         xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" version="2.1">
         <tlib-version>1.0</tlib-version>
         <short-name>cms-taglib</short-name>
         <uri>http://www.pgoetz.de/taglibs/cms</uri>
         <tag>
              <name>test-iterator</name>
              <tag-class>de.pgoetz.cms.taglib.TestIterator</tag-class>
              <body-content>tagdependent</body-content>
         </tag>
         <tag>
              <name>test-iterator-child</name>
              <tag-class>de.pgoetz.cms.taglib.TestIteratorChild</tag-class>
              <body-content>empty</body-content>
         </tag>
    </taglib>And the snippet of my jsp is as follows:
         <!-- TestIterator -->
         <cms:test-iterator>
              <cms:test-iterator-child />
         </cms:test-iterator>The result is that on my console I get the following output:
    09:28:01,656 INFO [STDOUT] TestIterator: doInitBody
    09:28:01,656 INFO [STDOUT] TestIterator: doAfterBody
    09:28:01,656 INFO [STDOUT] TestIterator: doAfterBody
    09:28:01,656 INFO [STDOUT] TestIterator: doAfterBody
    09:28:01,656 INFO [STDOUT] TestIterator: doAfterBody
    So the child is never executed.
    It would be a great help if anybody could tell me what's going wrong here.
    Thanks and greetings from germany!
    Peter
    Message was edited by:
    Peter_Goetz

  • HT4623 How can i solve installation problem with iOS 6.

    How can i solve installation problem with iOS 6. I've tried to install it on my iPhone 4 but nothing turned out. Thanks in advance!!!!

    can you connect your device to Tiunes and check to see if there are any files in
    OSX: ~/Library/Logs/Crashreporter/Mobiledevice/<Devicename>/
    Windows: %appdata%\Apple Computer\Logs\Crashreporter\MobileDevice\<Devicename>
    specifically named OTA......
      Also for the failed itunes restore please attach the contents of the latest log in
    OSX:   ~/Library/Logs/<iphone or ipod or ipad> updater logs
    Windows: C:/User/ Appdata/Roaming/iTunes/<iphone or ipod or ipad> updater logs

  • UnmarshalException with nested EOFException on Naming.lookup()

    I have an RMI server object that creates its own RMI registry and binds, like this:
         Registry registry = LocateRegistry.createRegistry(1099);
         registry.bind("JournalReceiver", this);I run this from the command line to get it started. Everything looks good.
    My client is running as a webapp in Tomcat (on the same machine), and it executes this call:
            receiver = (RmiJournalReceiverInterface) Naming.lookup("JournalReceiver");and gets this error
    java.rmi.UnmarshalException: error unmarshalling return; nested exception is:
         java.io.EOFException
         at sun.rmi.registry.RegistryImpl_Stub.lookup(Unknown Source)
         at java.rmi.Naming.lookup(Naming.java:84)It is clear that I'm reaching the registry, and if I modify the code to intentionally lookup the wrong name I get a "not bound error", so I'm pretty confident that my server object is indeed bound to that name. Beyond that, I'm lost.
    Any thoughts?
    jb

    OK - the problem was an incorrect codebase.
    My server object creates its own registry with
           Registry registry = LocateRegistry.createRegistry(1099);
         registry.bind("JournalReceiver", this);and I get the UnmarshalException with nested EOFException, as shown above.
    However, if I set up an outside registry with 'rmiregistry', and use
            Naming.rebind("JournalReceiver", this)then I get the UnmarshalException with nested ClassNotFoundException, which has shown up in several threads.
    So, following the advice given in those threads, I fix the codebase like this:
    java -Djava.rmi.server.codebase=file:/var/local/jb/RmiReceiver.jar \
       org.nsdl.RmiJournalReceiverand it works as advertised, even when I go back to my original lines of binding code.
    Thanks again for the help.
    jb

  • Import tables with nested table : ORA-00600

    In Oracle 9.2
    Create object, type as table, and table with nested table (store as syms_ntab) are successfully.
    Also its export.
    In process of import on another server (also 9.2, 'fromuser=one touser=two') shows errors:
    . . importing table "SYMS_NTAB"
    IMP-00058: ORACLE error 600 encountered
    ORA-00600: internal error code, arguments: [kokeeafi1], [2], [2], [], [], [], [], []
    IMP-00075: Warning: The nested table may contain partial rows or duplicate rows
    But for all that table is created and error occur on phase inserting strings.
    What is this?
    In Oracle 8.0.5 i perform similar operation without error.

    From Oracle error messages and codes manual:
    ORA-00600 internal error code, arguments: [string], [string], [string], [string], [string], [string], [string], [string]
    Cause: This is the generic internal error number for Oracle program exceptions. It indicates that a process has encountered a low-level, unexpected condition. Causes of this message include:
    * timeouts
    * file corruption
    * failed data checks in memory
    * hardware, memory, or I/O errors
    * incorrectly restored files
    The first argument is the internal message number. Other arguments are various numbers, names, and character strings. The numbers may change meanings between different versions of Oracle.
    Action: Report this error to Oracle Support Services after gathering the following information:
    * events that led up to the error
    * the operations that were attempted that led to the error
    * the conditions of the operating system and databases at the time of the error
    * any unusual circumstances that occurred before receiving the ORA-00600 message
    * contents of any trace files generated by the error
    * the relevant portions of the Alter files
    Note: The cause of this message may manifest itself as different errors at different times. Be aware of the history of errors that occurred before this internal error.

Maybe you are looking for