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

[john o goyo]

Greetings, Gaius.

On 2024-08-30 06:26, Gaius Mulley wrote (in part):
[...]
> Hi John,
>
> Does the following work on Solaris with gdb, it will invoke cc1gm2
> with all default options and the source file hello.mod.
>
>
> $ gm2 hello.mod -wrapper gdb,--args
>
> GNU gdb (Debian 13.1-3) 13.1
> Copyright (C) 2023 Free Software Foundation, Inc.
> License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
> This is free software: you are free to change and redistribute it.
> There is NO WARRANTY, to the extent permitted by law.
> Type "show copying" and "show warranty" for details.
> This GDB was configured as "x86_64-linux-gnu".
> Type "show configuration" for configuration details.
> For bug reporting instructions, please see:
> <https://www.gnu.org/software/gdb/bugs/>.
> Find the GDB manual and other documentation resources online at:
>      <http://www.gnu.org/software/gdb/documentation/>.
>
> For help, type "help".
> Type "apropos word" to search for commands related to "word"...
> Reading symbols from 
> /home/gaius/opt/libexec/gcc/x86_64-pc-linux-gnu/15.0.0/cc1gm2...
> (gdb) break gimplify_function_tree(tree_node*)
> Breakpoint 1 at 0xe52080: file ../../gcc/tree.h, line 3632.
> (gdb) run
> Starting program: /home/gaius/opt/libexec/gcc/x86_64-pc-linux-gnu/15.0.0/cc1gm2 
> -imultiarch x86_64-linux-gnu -quiet -dumpdir a- -dumpbase hello.mod 
> -dumpbase-ext .mod -mtune=generic -march=x86-64 -g -fm2-pathname=- 
> -fm2-pathnameI. -fgen-module-list=- -fscaffold-dynamic -fscaffold-main 
> -flibs=m2cor,m2log,m2pim,m2iso -fm2-pathname=- -fm2-pathnameI. hello.mod -o 
> /tmp/ccQdbK7K.s
> [Thread debugging using libthread_db enabled]
> Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".
>
> Breakpoint 1, gimplify_function_tree (fndecl=0x7ffff76cc300) at 
> ../../gcc/tree.h:3632
> 3632      if (TREE_CODE (__t) != __c)
> (gdb)

Yes, the above does indeed work but I am still in the dark.  Please 
remember that I am woefully ignorant of the inner workings of gcc.

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));
[...]
(gdb) del 1
(gdb) c
Continuing.

Thread 2 received signal SIGSEGV, Segmentation fault.
build_call_expr (fndecl=0x0, n=1) at 
/home/build/opt/gcc/src/gcc-14.2.0/gcc/tree.cc:10913
10913     return build_call_expr_loc_array (UNKNOWN_LOCATION, fndecl, n, 
argarray);

[set breakpoint at l10913 and reran]
Thread 2 hit Breakpoint 2, build_call_expr_loc_array (loc=0, fndecl=0x0, 
n=1, argarray=0xffbfe850)
    at /home/build/opt/gcc/src/gcc-14.2.0/gcc/tree.cc:10863
10863     tree fntype = TREE_TYPE (fndecl);
(gdb) bt
#0  build_call_expr_loc_array (loc=0, fndecl=0x0, n=1, argarray=0xffbfe850)
    at /home/build/opt/gcc/src/gcc-14.2.0/gcc/tree.cc:10863
#1  build_call_expr (fndecl=0x0, n=1) at 
/home/build/opt/gcc/src/gcc-14.2.0/gcc/tree.cc:10913
#2  0x00d2ece8 in build_cltz_expr (src=0xfb481680, leading=<optimized 
out>, define_at_zero=<optimized out>)
    at /home/build/opt/gcc/src/gcc-14.2.0/gcc/tree-ssa-loop-niter.cc:2290
#3  0x00d3c5fc in number_of_iterations_cltz_complement (exit=<optimized 
out>, loop=0xfb412f80,
    code=<optimized out>, niter=0xffbfeb88) at 
/home/build/opt/gcc/src/gcc-14.2.0/gcc/tree-ssa-loop-niter.cc:2542
#4  number_of_iterations_bitcount (loop=0xfb412f80, exit=<optimized 
out>, code=<optimized out>, niter=0xffbfeb88)
    at /home/build/opt/gcc/src/gcc-14.2.0/gcc/tree-ssa-loop-niter.cc:2628
[...]
(gdb) l
10858      ARGARRAY.  */
10859
10860   tree
10861   build_call_expr_loc_array (location_t loc, tree fndecl, int n, 
tree *argarray)
10862   {
10863     tree fntype = TREE_TYPE (fndecl);
10864     tree fn = build1 (ADDR_EXPR, build_pointer_type (fntype), fndecl);
10865
10866     return fold_build_call_array_loc (loc, TREE_TYPE (fntype), fn, 
n, argarray);
10867   }
(gdb) n

Thread 2 received signal SIGSEGV, Segmentation fault.
build_call_expr (fndecl=0x0, n=1) at 
/home/build/opt/gcc/src/gcc-14.2.0/gcc/tree.cc:10913
10913     return build_call_expr_loc_array (UNKNOWN_LOCATION, fndecl, n, 
argarray);

Does this offer you any clues on the problem or how to proceed?

Sincerely,
john