[Bug fortran/87359] [9 regression] pointer being freed was not allocated

[pault at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=87359

--- Comment #10 from Paul Thomas <pault at gcc dot gnu.org> ---
(In reply to Thomas Koenig from comment #7)
> Well, I can confirm this.  Output from valgrind shows as the first error:
> 
> Running test: event_transforms_1| Process library 'event_transforms_1_lib':
> initialized
> ==23957== Invalid read of size 8
> ==23957==    at 0xB22334:
> __mci_midpoint_MOD___final_mci_midpoint_Mci_midpoint_t (mci_midpoint.f90:475)
> ==23957==    by 0xA57C34: __process_mci_MOD_process_mci_entry_init
> (process_mci.f90:187)
> ==23957==    by 0xA8E78B: __process_MOD_process_setup_mci (process.f90:1150)
> ==23957==    by 0xB30828: __event_transforms_uti_MOD_event_transforms_1
> (event_transforms_uti.f90:85)
> ==23957==    by 0x422BB9: __unit_tests_MOD_test (unit_tests.f90:175)
> ==23957==    by 0xB4FF46: __event_transforms_ut_MOD_event_transforms_test
> (event_transforms_ut.f90:45)
> ==23957==    by 0xB51801: whizard_check.3816 (in
> /home/ig25/Downloads/Whiz/whizard_test)
> ==23957==    by 0xB53255: MAIN__ (in /home/ig25/Downloads/Whiz/whizard_test)
> ==23957==    by 0xB535A9: main (in /home/ig25/Downloads/Whiz/whizard_test)
> ==23957==  Address 0x61375a0 is 16 bytes inside a block of size 480 free'd
> ==23957==    at 0x4C2B28A: free (vg_replace_malloc.c:530)
> ==23957==    by 0xA57C03: __process_mci_MOD_process_mci_entry_init
> (process_mci.f90:187)
> ==23957==    by 0xA8E78B: __process_MOD_process_setup_mci (process.f90:1150)
> ==23957==    by 0xB30828: __event_transforms_uti_MOD_event_transforms_1
> (event_transforms_uti.f90:85)
> ==23957==    by 0x422BB9: __unit_tests_MOD_test (unit_tests.f90:175)
> ==23957==    by 0xB4FF46: __event_transforms_ut_MOD_event_transforms_test
> (event_transforms_ut.f90:45)
> ==23957==    by 0xB51801: whizard_check.3816 (in
> /home/ig25/Downloads/Whiz/whizard_test)
> ==23957==    by 0xB53255: MAIN__ (in /home/ig25/Downloads/Whiz/whizard_test)
> ==23957==    by 0xB535A9: main (in /home/ig25/Downloads/Whiz/whizard_test)
> ==23957==  Block was alloc'd at
> ==23957==    at 0x4C2A0DD: malloc (vg_replace_malloc.c:299)
> ==23957==    by 0xA4539B:
> __process_config_MOD_process_component_extract_mci_template
> (process_config.f90:834)
> ==23957==    by 0xA579D2: __process_mci_MOD_process_mci_entry_init
> (process_mci.f90:187)
> ==23957==    by 0xA8E78B: __process_MOD_process_setup_mci (process.f90:1150)
> ==23957==    by 0xB30828: __event_transforms_uti_MOD_event_transforms_1
> (event_transforms_uti.f90:85)
> ==23957==    by 0x422BB9: __unit_tests_MOD_test (unit_tests.f90:175)
> ==23957==    by 0xB4FF46: __event_transforms_ut_MOD_event_transforms_test
> (event_transforms_ut.f90:45)
> ==23957==    by 0xB51801: whizard_check.3816 (in
> /home/ig25/Downloads/Whiz/whizard_test)
> ==23957==    by 0xB53255: MAIN__ (in /home/ig25/Downloads/Whiz/whizard_test)
> ==23957==    by 0xB535A9: main (in /home/ig25/Downloads/Whiz/whizard_test)
> 
> So, it is a finalizer problem.  The allocate statement was at
> 
>   function process_component_extract_mci_template (component) &
>          result (mci_template)
>     class(mci_t), allocatable :: mci_template
>     class(process_component_t), intent(in) :: component
>     if (allocated (component%mci_template)) &
>        allocate (mci_template, source = component%mci_template)
>   end function process_component_extract_mci_template
> 
> Hopefully, this can narrow things down a bit.

If this a finalizer problem, try reverting r263916.

I am not in a position to work on this right now, except to revert the above
revision on trunk. Please let me know if this does indeed fix your proble.

Cheers

Paul