On 9/19/25 10:28 AM, anlauf at gcc dot gnu.org wrote:
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=122002
--- Comment #5 from anlauf at gcc dot gnu.org ---
(In reply to Jerry DeLisle from comment #4)
Now I wonder why I dont see it here.
Can you run:
valgrind --leak-check=full --track-origins=yes /path/to/f951 -quiet -pedantic
pdt_13.f03
I get something similar to Sam:
==17800== Conditional jump or move depends on uninitialised value(s)
==17800== at 0x590F46A: __printf_buffer (in /lib64/libc.so.6)
==17800== by 0x592A1F8: __vsprintf_internal (in /lib64/libc.so.6)
==17800== by 0x590B180: sprintf (in /lib64/libc.so.6)
==17800== by 0xA24B24: gfc_get_pdt_instance(gfc_actual_arglist*,
gfc_symbol**, gfc_actual_arglist**) (decl.cc:4065)
==17800== by 0xAC5F35: gfc_match_rvalue(gfc_expr**) (primary.cc:4091)
==17800== by 0xA80D12: match_primary (matchexp.cc:217)
==17800== by 0xA80D12: match_level_1 (matchexp.cc:267)
==17800== by 0xA80D12: match_mult_operand(gfc_expr**) (matchexp.cc:323)
==17800== by 0xA80FDC: match_add_operand(gfc_expr**) (matchexp.cc:412)
==17800== by 0xA812D9: match_level_2(gfc_expr**) (matchexp.cc:536)
==17800== by 0xA814BA: match_level_3(gfc_expr**) (matchexp.cc:607)
==17800== by 0xA81602: match_level_4 (matchexp.cc:655)
==17800== by 0xA81602: match_and_operand(gfc_expr**) (matchexp.cc:749)
==17800== by 0xA817DA: match_or_operand(gfc_expr**) (matchexp.cc:778)
==17800== by 0xA818EA: match_equiv_operand(gfc_expr**) (matchexp.cc:821)
==17800== Uninitialised value was created by a stack allocation
==17800== at 0xA2498F: gfc_get_pdt_instance(gfc_actual_arglist*,
gfc_symbol**, gfc_actual_arglist**) (decl.cc:3878)
==17800==
==17800== Use of uninitialised value of size 8
==17800== at 0x590436B: _itoa_word (in /lib64/libc.so.6)
==17800== by 0x590E913: __printf_buffer (in /lib64/libc.so.6)
==17800== by 0x592A1F8: __vsprintf_internal (in /lib64/libc.so.6)
==17800== by 0x590B180: sprintf (in /lib64/libc.so.6)
==17800== by 0xA24B24: gfc_get_pdt_instance(gfc_actual_arglist*,
gfc_symbol**, gfc_actual_arglist**) (decl.cc:4065)
...
I have to head out for a bit. I did regression test the patch that was
submitted. I am guessing something got mixed up on the commit which I am just
now building. I will get back to this when I return.
Jerry
