Re: [Libcdio-devel] libcdio-0.83git on Solaris snv134 sees no drives

[Rocky Bernstein]

Good to hear you've got this compiled and haven't given up on libcdio yet.

Right now I'm in Krakow at EuRuKo 2010 (http://www.euruko.org) and started
applying patches but due to problems in my hardware and scheduling this is
too painful. (Internet speed in Polish hotels though is faster than what I
get at work so it's not their fault, but something on my setup)

Some of the suggestions have  been applied. But some may have to remain for
when I get back to the U.S. I have a final paper due for DYLA 2010 by June
6th and am very much behind in that as well.

As for the Sun questions, have you considered one the Usenet group for
Solaris? When I get a chance I will also try to contact some of the Solaris
experts that have been helpful in the past.

Thanks again

Some other comments inline below...

On Sat, May 29, 2010 at 9:29 AM, Thomas Schmitt <address@hidden> wrote:

> Hi,
>
> i got OpenSolaris snv 134 running with gcc and
> gdb.
> Base image:
>   solaris-textinstall-134-x86.iso
> Development stuff fetched via internet by:
>   pkg install developer/gcc/address@hidden
>   pkg install developer/debug/address@hidden
>
> This suffices for building my own stuff and
> obviously for libcdio.
>
> ------------------------------------------------
> Obstacles with libcdio-0.83git of 8 May 2010:
> ------------------------------------------------
>
> make aborts with
>  ../../include/cdio++/iso9660.hpp:151: error:
>  expected identifier before numeric constant
>
> The code piece is
>
>  class FS : public CdioDevice // ISO 9660 Filesystem on a CD or CD-image
>
> I added at the beginning of the iso9660.hpp
>
>  #ifdef FS
>  #undef FS
>  #endif
>

Ok. But since this change is inside a header, the impact of this change
could be great. I wonder whether this might cause problems in C++ code that
needs assumes FS is a macro.

>
> ------------------------------------------------
>
> As with FreeBSD 8-RELEASE i got
>  /usr/bin/perl ./pod2c.pl  >usage.h
>  Expecting exactly one argument, a filename at ./pod2c.pl line 6.
>  *** Error code 255
> and copied src/cd-paranoia/usage.txt from Linux.
>

I thought we went over this. This is because you were probably not using GNU
make but the one Solaris uses, which doesn't set $<.

I've removed the occurrences of $< from Makefile.am now.


>
> ------------------------------------------------
> Warnings with libcdio-0.83git of 8 May 2010
> ------------------------------------------------
>
> Several source files do not comply to the demand
> of Solaris man lfcompile to define eventual
> _FILE_OFFSET_BITS before including any system
> headers.
> I see this macro in autools generated header
> config.h.
>
> This code
>
>  #include <sys/types.h>
>  #include <cdio/cdio.h>
>  #include <cdio/udf.h>
>
>  #include <stdio.h>
>
>  #ifdef HAVE_CONFIG_H
>  # include "config.h"
>  #endif
>
> produces this warning
>
>  In file included from udffile.c:38:
>  ../config.h:332:1: warning: "_FILE_OFFSET_BITS" redefined
>  In file included from /usr/include/sys/types.h:33,
>                   from udffile.c:31:
>
> Here is a list of affected source files.
>

I started making changes as suggested. However this is where things got too
painful so I'll have to postpone finishing for now. I've gotten up to
report.c. If you take a look at how the changes were made and do the rest
like that and send a patch for the rest I'll apply that patch.




> (header.h) gives the #include which happens too
> late.
> E.g. udf_private.h includes config.h and thus
> should be the first #include in lib/udf/udf.c
>
>  lib/udf/udf.c (udf_private.h)
>  lib/udf/udf_file.c (udf_private.h)
>  lib/cdda_interface/common_interface.c (common_interface.h)
>  src/cd-paranoia/report.c (config.h)
>  example/cdchange.c (config.h)
>  example/isofile.c (config.h)
>  example/isofile2.c (portable.h)
>  example/isofuzzy.c (portable.h)
>  example/paranoia.c (config.h)
>  example/udffile.c (config.h)
>
>
> ------------------------------------------------
> But it shows no drives
> ------------------------------------------------
>
>  $ example/drives
>  cdio 3 message: open (/vol/dev/aliases/cdrom0): No such file or directory
>  ...
>
> There is no directory /vol in the root of my
> Solaris.
>
> Two drives are attached. Command mount tells that
> the IDE DVD-ROM gets mounted via
>  /dev/dsk/c3t0d0s2
> which is a link to
>  /devices/address@hidden,0/address@hidden,1/address@hidden/address@hidden,0
> cdrecord calls it
>  3,0,0
>
> The SATA DVD burner is
>  /dev/dsk/c2t4d0s2
>  /devices/address@hidden,0/pci1458,address@hidden/address@hidden,0:c
>  2,4,0
>
>
> Are there any proposals what i should try in
>  lib/driver/solaris.c
> function
>  cdio_get_devices_solaris()
>
> ------------------------------------------------
>
> Have a nice day :)
>
> Thomas
>
>
>