Re: Trying to track down problem with packed.com from testsuite

[john o goyo]

Greetings, Gaius.

On 2024-09-02 09:00, Gaius Mulley wrote (in part):
> john o goyo <jog37@riddermarkfarm.ca> writes (in part):
>
> > [...]
> > Here is what I tried:
> >
> > [~/opt/gm2/testsuite]=> gm2-14 packed.mod -fiso -O2 -wrapper gdb,--args
> > GNU gdb (GDB) 15.1
> > [...]
> > This GDB was configured as "sparcv9-sun-solaris2.11".
> > [...]
> > Reading symbols from
> > /home/build/gcc/14.2.0/libexec/gcc/sparc-sun-solaris2.11/14.2.0/cc1gm2...
> > [...]
> > (gdb) break gimplify_function_tree(tree_node*)
> > Breakpoint 1 at 0x8a6f40: file
> > /home/build/opt/gcc/src/gcc-14.2.0/gcc/gimplify.cc, line 19333.
> > (gdb) r
> > Starting program:
> > /home/build/gcc/14.2.0/libexec/gcc/sparc-sun-solaris2.11/14.2.0/cc1gm2
> > -quiet -dumpdir a- -dumpbase packed.mod -dumpbase-ext .mod -mcpu=v9
> > -O2 -fiso -fm2-pathname=- -fm2-pathnameI. -fgen-module-list=-
> > -fscaffold-dynamic -fscaffold-main -flibs=m2iso,m2cor,m2pim,m2log
> > -fm2-pathname=- -fm2-pathnameI. packed.mod -o /var/tmp//ccb3aGgQ.s
> > [Thread debugging using libthread_db enabled]
> > [New Thread 1 (LWP 1)]
> > [Switching to Thread 1 (LWP 1)]
> >
> > Thread 2 hit Breakpoint 1, gimplify_function_tree (fndecl=0xfb474700)
> >      at /home/build/opt/gcc/src/gcc-14.2.0/gcc/gimplify.cc:19333
> > 19333   {
> > (gdb) l
> > 19328      Return the sequence of GIMPLE statements corresponding to
> > the body
> > 19329      of FNDECL.  */
> > 19330
> > 19331   void
> > 19332   gimplify_function_tree (tree fndecl)
> > 19333   {
> > 19334     gimple_seq seq;
> > 19335     gbind *bind;
> > 19336
> > 19337     gcc_assert (!gimple_body (fndecl));
> > [...]
> I wonder what the output from the tree dump would be at this point?
>
> (gdb) print debug_tree (fndecl)
(gdb) print debug_tree (fndecl)
 <function_decl fb474700 _M2_packed_init
    type <function_type fb432ae0
        type <void_type fb9109c0 PROC VOID
            align:8 warn_if_not_align:0 symtab:0 alias-set -1 
canonical-type fb9109c0
            pointer_to_this <pointer_type fb910a20 ADDRESS>>
        SI
        size <integer_cst fb40abd0 constant 32>
        unit-size <integer_cst fb40abe8 constant 4>
        align:32 warn_if_not_align:0 symtab:0 alias-set -1 
canonical-type fb432ae0
        arg-types <tree_list fb43dde8 value <integer_type fb9103c0 INTEGER>
            chain <tree_list fb43ddd0 value <pointer_type fb910a20 ADDRESS>
                chain <tree_list fb43ddb8 value <pointer_type fb910a20 
ADDRESS>
                    chain <tree_list fb40b050 value <void_type fb9109c0 
PROC>>>>>
        pointer_to_this <pointer_type fb433f20>>
    addressable used public static SI (null):0:0 align:32 
warn_if_not_align:0 context <translation_unit_decl fb4180d8 D.80> 
initial <block fb4240e0>
    result <result_decl fb4762d0 D.1631 type <void_type fb9109c0 PROC>
        VOID (null):0:0
        align:8 warn_if_not_align:0 context <function_decl fb474700 
_M2_packed_init>>
    arguments <parm_decl fb473540 argc
        type <integer_type fb9103c0 INTEGER public SI size <integer_cst 
fb40abd0 32> unit-size <integer_cst fb40abe8 4>
            align:32 warn_if_not_align:0 symtab:0 alias-set -1 
canonical-type fb9103c0 precision:32 min <integer_cst fb40ae10 
-2147483648> max <integer_cst fb40ae28 2147483647>
            pointer_to_this <pointer_type fb910fc0> reference_to_this 
<reference_type fb452960>>
        SI packed.mod:57:1 size <integer_cst fb40abd0 32> unit-size 
<integer_cst fb40abe8 4>
        align:32 warn_if_not_align:0 context <function_decl fb474700 
_M2_packed_init> arg-type <integer_type fb9103c0 INTEGER>
        chain <parm_decl fb4734f0 argv type <pointer_type fb910a20 ADDRESS>
            unsigned SI packed.mod:57:1 size <integer_cst fb40abd0 32> 
unit-size <integer_cst fb40abe8 4>
            align:32 warn_if_not_align:0 context <function_decl 
fb474700 _M2_packed_init> arg-type <pointer_type fb910a20 ADDRESS> chain 
<parm_decl fb4734a0 envp>>>
    struct-function fb484000 chain <function_decl fb474680 _M2_packed_dep>>
$1 = void


[I cannot say that I am enlightened, though. #6-)]

> or perhaps 'pf' might be able to get a m2 description of the tree:
>
> (gdb) print pf (fndecl)

(gdb) print pf (fndecl)

Thread 2 received signal SIGSEGV, Segmentation fault.
0xf888bcf8 in putc () from /lib/libc.so.1
The program being debugged was signaled while in a function called from GDB.
GDB remains in the frame where the signal was received.
To change this behavior use "set unwind-on-signal on".
Evaluation of the expression containing the function
(modula2::pf(tree)) will be abandoned.
When the function is done executing, GDB will silently stop.

[I am in uncharted territory here, Gaius.]

Sincerely,
john

P.S. Configure did not honour the flags specified so I used variables 
CC="/home/build/gcc/14.2.0/bin/gcc -g -O0" and so on.  But this was 
still ignored building the libraries.