invalid operand combination: xorb (Openserver 6) - SCO

This is a discussion on invalid operand combination: xorb (Openserver 6) - SCO ; Hi. I'm trying to compile findutils under Openserver 6. The following error message happens pretty much right the way: cc -DHAVE_CONFIG_H -I. -I. -I../.. -g -c argmatch.c UX:as: ERROR: /var/tmp/ctm2AAAa006a3:2939:invalid operand combination: xorb Any thoughts? Thanks. -3-...

+ Reply to Thread
Results 1 to 8 of 8

Thread: invalid operand combination: xorb (Openserver 6)

  1. invalid operand combination: xorb (Openserver 6)

    Hi.

    I'm trying to compile findutils under Openserver 6. The following error
    message happens pretty much right the way:

    cc -DHAVE_CONFIG_H -I. -I. -I../.. -g -c argmatch.c
    UX:as: ERROR: /var/tmp/ctm2AAAa006a3:2939:invalid operand combination: xorb

    Any thoughts? Thanks.

    -3-

  2. Re: invalid operand combination: xorb (Openserver 6)

    third@whasup.com wrote:
    > Hi.
    >
    > I'm trying to compile findutils under Openserver 6. The following error
    > message happens pretty much right the way:
    >
    > cc -DHAVE_CONFIG_H -I. -I. -I../.. -g -c argmatch.c
    > UX:as: ERROR: /var/tmp/ctm2AAAa006a3:2939:invalid operand combination: xorb


    This is a known problem in the OpenServer 6 compiler, related to
    the bool type. I believe it is fixed in OSR6.0.0 maint pack 1, which
    should hit the website in the next few days.

    Jonathan Schilling


  3. Re: invalid operand combination: xorb (Openserver 6)


    jlselsewhere@my-deja.com wrote:
    > I believe it is fixed in OSR6.0.0 maint pack 1, which
    > should hit the website in the next few days.


    Maint pack 1 is there now, at
    http://www.sco.com/support/update/do...ase.php?rid=77

    Jonathan Schilling


  4. Re: invalid operand combination: xorb (Openserver 6)

    As per jlselsewhere@my-deja.com (jlselsewhere@my-deja.com):

    > third@whasup.com wrote:
    > > Hi.
    > >
    > > I'm trying to compile findutils under Openserver 6. The following error
    > > message happens pretty much right the way:
    > >
    > > cc -DHAVE_CONFIG_H -I. -I. -I../.. -g -c argmatch.c
    > > UX:as: ERROR: /var/tmp/ctm2AAAa006a3:2939:invalid operand combination: xorb

    >
    > This is a known problem in the OpenServer 6 compiler, related to
    > the bool type. I believe it is fixed in OSR6.0.0 maint pack 1, which
    > should hit the website in the next few days.


    Thanks. Installed it, the error is gone. Didn't go much further though.

    cc -DHAVE_CONFIG_H -I. -I. -I../.. -g -c fnmatch.c
    UX:as: ERROR: /var/tmp/ctm2AAAa007FB:1540:invalid operand combination: movb
    UX:as: ERROR: /var/tmp/ctm2AAAa007FB:3156:invalid operand combination: movb
    UX:as: ERROR: /var/tmp/ctm2AAAa007FB:5954:invalid operand combination: movb
    UX:as: ERROR: /var/tmp/ctm2AAAa007FB:7598:invalid operand combination: movb

    -3-

  5. Re: invalid operand combination: xorb (Openserver 6)

    t...@whasup.com wrote:
    >
    > Thanks. Installed it, the error is gone. Didn't go much further though.
    >
    > cc -DHAVE_CONFIG_H -I. -I. -I../.. -g -c fnmatch.c
    > UX:as: ERROR: /var/tmp/ctm2AAAa007FB:1540:invalid operand combination: movb
    > UX:as: ERROR: /var/tmp/ctm2AAAa007FB:3156:invalid operand combination: movb
    > UX:as: ERROR: /var/tmp/ctm2AAAa007FB:5954:invalid operand combination: movb
    > UX:as: ERROR: /var/tmp/ctm2AAAa007FB:7598:invalid operand combination: movb


    Hmmm. We think we know what went wrong here, but to be sure,
    please do the same compile adding the -P option, thus producing a .i
    file,
    then post the .i file here.

    Jonathan Schilling


  6. Re: invalid operand combination: xorb (Openserver 6)

    As per jlselsewhere@my-deja.com (jlselsewhere@my-deja.com):

    > t...@whasup.com wrote:
    > >
    > > Thanks. Installed it, the error is gone. Didn't go much further though.
    > >
    > > cc -DHAVE_CONFIG_H -I. -I. -I../.. -g -c fnmatch.c
    > > UX:as: ERROR: /var/tmp/ctm2AAAa007FB:1540:invalid operand combination: movb
    > > UX:as: ERROR: /var/tmp/ctm2AAAa007FB:3156:invalid operand combination: movb
    > > UX:as: ERROR: /var/tmp/ctm2AAAa007FB:5954:invalid operand combination: movb
    > > UX:as: ERROR: /var/tmp/ctm2AAAa007FB:7598:invalid operand combination: movb

    >
    > Hmmm. We think we know what went wrong here, but to be sure,
    > please do the same compile adding the -P option, thus producing a .i
    > file,
    > then post the .i file here.


    The file is too big. You will find it here:

    http://whasup.com/openserver6/fnmatch.i

    There is a lot of apparently empty lines in it, but I left it the way it
    was generated.

    -3-

  7. Re: invalid operand combination: xorb (Openserver 6)

    third@whasup.com wrote:
    > >
    > > Hmmm. We think we know what went wrong here, but to be sure,
    > > please do the same compile adding the -P option, thus producing a .i
    > > file,
    > > then post the .i file here.

    >
    > The file is too big. You will find it here:
    >
    > http://whasup.com/openserver6/fnmatch.i


    OK, thanks. It is indeed a second but related compiler bug,
    which we have now fixed. Since Maint pack 1 is already out,
    it'll have to wait until Maint pack 2. If you need it immediately,
    you'll have to go through normal SCO customer support channels to
    get it into the escalations process.

    Jonathan Schilling


  8. Re: invalid operand combination: xorb (Openserver 6)

    As per jlselsewhere@my-deja.com (jlselsewhere@my-deja.com):

    > third@whasup.com wrote:
    > > >
    > > > Hmmm. We think we know what went wrong here, but to be sure,
    > > > please do the same compile adding the -P option, thus producing a .i
    > > > file,
    > > > then post the .i file here.

    > >
    > > The file is too big. You will find it here:
    > >
    > > http://whasup.com/openserver6/fnmatch.i

    >
    > OK, thanks. It is indeed a second but related compiler bug,
    > which we have now fixed. Since Maint pack 1 is already out,
    > it'll have to wait until Maint pack 2. If you need it immediately,
    > you'll have to go through normal SCO customer support channels to
    > get it into the escalations process.


    Thanks. I guess I'll have to wait. Maybe it's a dumb suggestion, but perhaps
    developers should try compiling a couple of gnu packages (findutils would
    be a good choice) after applying that second fix. Thanks again.

    -3-

+ Reply to Thread