GDB 5.1 on Tru64 UNIX 4.0D, 5.0, 5.1

GDB 5.1 on Tru64 UNIX 4.0D, 5.0, 5.1

[gdb]

GDB 5.1 builds fine on Tru64 UNIX 4.0D, 5.0A, and 5.1. However, when
trying to run it:

$ gdb a.out
GNU gdb 5.1
Copyright 2001 Free Software Foundation, Inc.
GDB is free software, covered by the GNU General Public License, and you are
welcome to change it and/or distribute copies of it under certain conditions.
Type "show copying" to see the conditions.
There is absolutely no warranty for GDB.  Type "show warranty" for details.
This GDB was configured as "alpha-dec-osf4.0d"...mdebugread.c:2448:
gdb-internal-error: Section index is uninitialized

An internal GDB error was detected.  This may make further
debugging unreliable.  Continue this debugging session? (y or n) n

Create a core file containing the current state of GDB? (y or n) n

This is with a.out generated using 'cc -g' or 'gcc -g'.

-- 
albert chin (china@thewrittenword.com)

Re: GDB 5.1 on Tru64 UNIX 4.0D, 5.0, 5.1

[Eli Zaretskii]

> From: gdb@thewrittenword.com
> Date: Mon, 26 Nov 2001 10:45:05 -0600
> 
> GDB 5.1 builds fine on Tru64 UNIX 4.0D, 5.0A, and 5.1. However, when
> trying to run it:
> [...]
> An internal GDB error was detected.  This may make further
> debugging unreliable.  Continue this debugging session? (y or n) n

Search the file README for "alpha-dec-osf4", it's all there.

Re: GDB 5.1 on Tru64 UNIX 4.0D, 5.0, 5.1

[gdb]

On Mon, Nov 26, 2001 at 09:10:30PM +0200, Eli Zaretskii wrote:
> > From: gdb@thewrittenword.com
> > Date: Mon, 26 Nov 2001 10:45:05 -0600
> > 
> > GDB 5.1 builds fine on Tru64 UNIX 4.0D, 5.0A, and 5.1. However, when
> > trying to run it:
> > [...]
> > An internal GDB error was detected.  This may make further
> > debugging unreliable.  Continue this debugging session? (y or n) n
> 
> Search the file README for "alpha-dec-osf4", it's all there.

Are you sure?

$ cd gdb-5.1
$ grep -i alpha gdb/README
[no output]

-- 
albert chin (china@thewrittenword.com)

Re: GDB 5.1 on Tru64 UNIX 4.0D, 5.0, 5.1

[Eli Zaretskii]

> From: gdb@thewrittenword.com
> Date: Mon, 26 Nov 2001 13:48:26 -0600
> > 
> > Search the file README for "alpha-dec-osf4", it's all there.
> 
> Are you sure?
> 
> $ cd gdb-5.1
> $ grep -i alpha gdb/README
> [no output]

Yes, I'm sure:

$ grep -i alpha gdb/README
alpha-dec-osf5.1
alpha-dec-osf4.0e

Here's the entry in its entirety:

    alpha-dec-osf4.0e

    GDB 5.1 is known to have problems on this platform (encounters an
    internal error in the symbol table reader).

Re: GDB 5.1 on Tru64 UNIX 4.0D, 5.0, 5.1

[gdb]

On Mon, Nov 26, 2001 at 10:31:04PM +0200, Eli Zaretskii wrote:
> > From: gdb@thewrittenword.com
> > Date: Mon, 26 Nov 2001 13:48:26 -0600
> > > 
> > > Search the file README for "alpha-dec-osf4", it's all there.
> > 
> > Are you sure?
> > 
> > $ cd gdb-5.1
> > $ grep -i alpha gdb/README
> > [no output]
> 
> Yes, I'm sure:
> 
> $ grep -i alpha gdb/README
> alpha-dec-osf5.1
> alpha-dec-osf4.0e
> 
> Here's the entry in its entirety:
> 
>     alpha-dec-osf4.0e
> 
>     GDB 5.1 is known to have problems on this platform (encounters an
>     internal error in the symbol table reader).

Oh well. Doesn't work for me.

-- 
albert chin (china@thewrittenword.com)

Re: GDB 5.1 on Tru64 UNIX 4.0D, 5.0, 5.1

[gdb]

On Mon, Nov 26, 2001 at 04:36:01PM -0600, gdb@thewrittenword.com wrote:
> On Mon, Nov 26, 2001 at 10:31:04PM +0200, Eli Zaretskii wrote:
> > > From: gdb@thewrittenword.com
> > > Date: Mon, 26 Nov 2001 13:48:26 -0600
> > > > 
> > > > Search the file README for "alpha-dec-osf4", it's all there.
> > > 
> > > Are you sure?
> > > 
> > > $ cd gdb-5.1
> > > $ grep -i alpha gdb/README
> > > [no output]
> > 
> > Yes, I'm sure:
> > 
> > $ grep -i alpha gdb/README
> > alpha-dec-osf5.1
> > alpha-dec-osf4.0e
> > 
> > Here's the entry in its entirety:
> > 
> >     alpha-dec-osf4.0e
> > 
> >     GDB 5.1 is known to have problems on this platform (encounters an
> >     internal error in the symbol table reader).
> 
> Oh well. Doesn't work for me.

To clarify, CC='cc -DUSE_LDR_ROUTINES' produced the same results on
Tru64 UNIX 5.1.

-- 
albert chin (china@thewrittenword.com)

Re: GDB 5.1 on Tru64 UNIX 4.0D, 5.0, 5.1

[Scott Snyder]

>gdb-internal-error: Section index is uninitialized

These hacks seemed to work around the worst of the symbol-reading problems
for me.  There's still other problems on this platform, though.
sss


Index: gdb/mdebugread.c
===================================================================
RCS file: /cvs/src/src/gdb/mdebugread.c,v
retrieving revision 1.16
diff -u -r1.16 mdebugread.c
--- mdebugread.c	2001/10/12 23:51:28	1.16
+++ mdebugread.c	2001/11/02 20:03:52
@@ -2424,11 +2424,19 @@
 	    }
 	  else if (SC_IS_DATA (ext_in->asym.sc))
 	    {
+#if 1 /* sss */
+              if (objfile->sect_index_data == -1)
+                continue;
+#endif
 	      ms_type = mst_data;
 	      svalue += ANOFFSET (objfile->section_offsets, SECT_OFF_DATA (objfile));
 	    }
 	  else if (SC_IS_BSS (ext_in->asym.sc))
 	    {
+#if 1 /* sss */
+              if (objfile->sect_index_bss == -1)
+                continue;
+#endif
 	      ms_type = mst_bss;
 	      svalue += ANOFFSET (objfile->section_offsets, SECT_OFF_BSS (objfile));
 	    }
@@ -2444,6 +2452,10 @@
 	    }
 	  else if (SC_IS_DATA (ext_in->asym.sc))
 	    {
+#if 1 /* sss */
+              if (objfile->sect_index_data == -1)
+                continue;
+#endif
 	      ms_type = mst_file_data;
 	      svalue += ANOFFSET (objfile->section_offsets, SECT_OFF_DATA (objfile));
 	    }
Index: bfd/ecoff.c
===================================================================
RCS file: /cvs/src/src/bfd/ecoff.c,v
retrieving revision 1.13
diff -u -r1.13 ecoff.c
--- ecoff.c	2001/10/10 12:08:28	1.13
+++ ecoff.c	2001/11/27 06:06:03
@@ -565,7 +565,7 @@
   UPDATE_RAW_END (cbDnOffset, idnMax, backend->debug_swap.external_dnr_size);
   UPDATE_RAW_END (cbPdOffset, ipdMax, backend->debug_swap.external_pdr_size);
   UPDATE_RAW_END (cbSymOffset, isymMax, backend->debug_swap.external_sym_size);
-  UPDATE_RAW_END (cbOptOffset, ioptMax, backend->debug_swap.external_opt_size);
+  UPDATE_RAW_END (cbOptOffset, ioptMax, 1/*backend->debug_swap.external_opt_size*/);
   UPDATE_RAW_END (cbAuxOffset, iauxMax, sizeof (union aux_ext));
   UPDATE_RAW_END (cbSsOffset, issMax, sizeof (char));
   UPDATE_RAW_END (cbSsExtOffset, issExtMax, sizeof (char));

Re: GDB 5.1 on Tru64 UNIX 4.0D, 5.0, 5.1

[Joel Brobecker]

> These hacks seemed to work around the worst of the symbol-reading problems
> for me.  There's still other problems on this platform, though.

I actually sent a message in gdb-patches ~10 days ago about these
errors. I explained what I understand the problem is, and was asking for
advice in order to implement a clean solution.

See http://sources.redhat.com/ml/gdb-patches/2001-11/msg00296.html

-- 
Joel

GDB 5.1 on Tru64 UNIX 4.0D, 5.0, 5.1

[gdb]

GDB 5.1 builds fine on Tru64 UNIX 4.0D, 5.0A, and 5.1. However, when
trying to run it:

$ gdb a.out
GNU gdb 5.1
Copyright 2001 Free Software Foundation, Inc.
GDB is free software, covered by the GNU General Public License, and you are
welcome to change it and/or distribute copies of it under certain conditions.
Type "show copying" to see the conditions.
There is absolutely no warranty for GDB.  Type "show warranty" for details.
This GDB was configured as "alpha-dec-osf4.0d"...mdebugread.c:2448:
gdb-internal-error: Section index is uninitialized

An internal GDB error was detected.  This may make further
debugging unreliable.  Continue this debugging session? (y or n) n

Create a core file containing the current state of GDB? (y or n) n

This is with a.out generated using 'cc -g' or 'gcc -g'.

-- 
albert chin (china@thewrittenword.com)

Re: GDB 5.1 on Tru64 UNIX 4.0D, 5.0, 5.1

[Eli Zaretskii]

> From: gdb@thewrittenword.com
> Date: Mon, 26 Nov 2001 10:45:05 -0600
> 
> GDB 5.1 builds fine on Tru64 UNIX 4.0D, 5.0A, and 5.1. However, when
> trying to run it:
> [...]
> An internal GDB error was detected.  This may make further
> debugging unreliable.  Continue this debugging session? (y or n) n

Search the file README for "alpha-dec-osf4", it's all there.

Re: GDB 5.1 on Tru64 UNIX 4.0D, 5.0, 5.1

[gdb]

On Mon, Nov 26, 2001 at 09:10:30PM +0200, Eli Zaretskii wrote:
> > From: gdb@thewrittenword.com
> > Date: Mon, 26 Nov 2001 10:45:05 -0600
> > 
> > GDB 5.1 builds fine on Tru64 UNIX 4.0D, 5.0A, and 5.1. However, when
> > trying to run it:
> > [...]
> > An internal GDB error was detected.  This may make further
> > debugging unreliable.  Continue this debugging session? (y or n) n
> 
> Search the file README for "alpha-dec-osf4", it's all there.

Are you sure?

$ cd gdb-5.1
$ grep -i alpha gdb/README
[no output]

-- 
albert chin (china@thewrittenword.com)

Re: GDB 5.1 on Tru64 UNIX 4.0D, 5.0, 5.1

[Eli Zaretskii]

> From: gdb@thewrittenword.com
> Date: Mon, 26 Nov 2001 13:48:26 -0600
> > 
> > Search the file README for "alpha-dec-osf4", it's all there.
> 
> Are you sure?
> 
> $ cd gdb-5.1
> $ grep -i alpha gdb/README
> [no output]

Yes, I'm sure:

$ grep -i alpha gdb/README
alpha-dec-osf5.1
alpha-dec-osf4.0e

Here's the entry in its entirety:

    alpha-dec-osf4.0e

    GDB 5.1 is known to have problems on this platform (encounters an
    internal error in the symbol table reader).

Re: GDB 5.1 on Tru64 UNIX 4.0D, 5.0, 5.1

[gdb]

On Mon, Nov 26, 2001 at 04:36:01PM -0600, gdb@thewrittenword.com wrote:
> On Mon, Nov 26, 2001 at 10:31:04PM +0200, Eli Zaretskii wrote:
> > > From: gdb@thewrittenword.com
> > > Date: Mon, 26 Nov 2001 13:48:26 -0600
> > > > 
> > > > Search the file README for "alpha-dec-osf4", it's all there.
> > > 
> > > Are you sure?
> > > 
> > > $ cd gdb-5.1
> > > $ grep -i alpha gdb/README
> > > [no output]
> > 
> > Yes, I'm sure:
> > 
> > $ grep -i alpha gdb/README
> > alpha-dec-osf5.1
> > alpha-dec-osf4.0e
> > 
> > Here's the entry in its entirety:
> > 
> >     alpha-dec-osf4.0e
> > 
> >     GDB 5.1 is known to have problems on this platform (encounters an
> >     internal error in the symbol table reader).
> 
> Oh well. Doesn't work for me.

To clarify, CC='cc -DUSE_LDR_ROUTINES' produced the same results on
Tru64 UNIX 5.1.

-- 
albert chin (china@thewrittenword.com)

Re: GDB 5.1 on Tru64 UNIX 4.0D, 5.0, 5.1

[Scott Snyder]

>gdb-internal-error: Section index is uninitialized

These hacks seemed to work around the worst of the symbol-reading problems
for me.  There's still other problems on this platform, though.
sss


Index: gdb/mdebugread.c
===================================================================
RCS file: /cvs/src/src/gdb/mdebugread.c,v
retrieving revision 1.16
diff -u -r1.16 mdebugread.c
--- mdebugread.c	2001/10/12 23:51:28	1.16
+++ mdebugread.c	2001/11/02 20:03:52
@@ -2424,11 +2424,19 @@
 	    }
 	  else if (SC_IS_DATA (ext_in->asym.sc))
 	    {
+#if 1 /* sss */
+              if (objfile->sect_index_data == -1)
+                continue;
+#endif
 	      ms_type = mst_data;
 	      svalue += ANOFFSET (objfile->section_offsets, SECT_OFF_DATA (objfile));
 	    }
 	  else if (SC_IS_BSS (ext_in->asym.sc))
 	    {
+#if 1 /* sss */
+              if (objfile->sect_index_bss == -1)
+                continue;
+#endif
 	      ms_type = mst_bss;
 	      svalue += ANOFFSET (objfile->section_offsets, SECT_OFF_BSS (objfile));
 	    }
@@ -2444,6 +2452,10 @@
 	    }
 	  else if (SC_IS_DATA (ext_in->asym.sc))
 	    {
+#if 1 /* sss */
+              if (objfile->sect_index_data == -1)
+                continue;
+#endif
 	      ms_type = mst_file_data;
 	      svalue += ANOFFSET (objfile->section_offsets, SECT_OFF_DATA (objfile));
 	    }
Index: bfd/ecoff.c
===================================================================
RCS file: /cvs/src/src/bfd/ecoff.c,v
retrieving revision 1.13
diff -u -r1.13 ecoff.c
--- ecoff.c	2001/10/10 12:08:28	1.13
+++ ecoff.c	2001/11/27 06:06:03
@@ -565,7 +565,7 @@
   UPDATE_RAW_END (cbDnOffset, idnMax, backend->debug_swap.external_dnr_size);
   UPDATE_RAW_END (cbPdOffset, ipdMax, backend->debug_swap.external_pdr_size);
   UPDATE_RAW_END (cbSymOffset, isymMax, backend->debug_swap.external_sym_size);
-  UPDATE_RAW_END (cbOptOffset, ioptMax, backend->debug_swap.external_opt_size);
+  UPDATE_RAW_END (cbOptOffset, ioptMax, 1/*backend->debug_swap.external_opt_size*/);
   UPDATE_RAW_END (cbAuxOffset, iauxMax, sizeof (union aux_ext));
   UPDATE_RAW_END (cbSsOffset, issMax, sizeof (char));
   UPDATE_RAW_END (cbSsExtOffset, issExtMax, sizeof (char));

Re: GDB 5.1 on Tru64 UNIX 4.0D, 5.0, 5.1

[Joel Brobecker]

> These hacks seemed to work around the worst of the symbol-reading problems
> for me.  There's still other problems on this platform, though.

I actually sent a message in gdb-patches ~10 days ago about these
errors. I explained what I understand the problem is, and was asking for
advice in order to implement a clean solution.

See http://sources.redhat.com/ml/gdb-patches/2001-11/msg00296.html

-- 
Joel