[Bug fortran/97039] New: -fbounds-check misses violation with slice of array but not an element

[Bug fortran/97039] New: -fbounds-check misses violation with slice of array but not an element

[anthony.debeus at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=97039

            Bug ID: 97039
           Summary: -fbounds-check misses violation with slice of array
                    but not an element
           Product: gcc
           Version: 10.2.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: fortran
          Assignee: unassigned at gcc dot gnu.org
          Reporter: anthony.debeus at gmail dot com
  Target Milestone: ---

Using built-in specs.
COLLECT_GCC=gfortran
COLLECT_LTO_WRAPPER=/usr/lib/gcc/x86_64-pc-linux-gnu/10.2.0/lto-wrapper
Target: x86_64-pc-linux-gnu
Configured with: /build/gcc/src/gcc/configure --prefix=/usr --libdir=/usr/lib
--libexecdir=/usr/lib --mandir=/usr/share/man --infodir=/usr/share/info
--with-bugurl=https://bugs.archlinux.org/
--enable-languages=c,c++,ada,fortran,go,lto,objc,obj-c++,d --with-isl
--with-linker-hash-style=gnu --with-system-zlib --enable-__cxa_atexit
--enable-cet=auto --enable-checking=release --enable-clocale=gnu
--enable-default-pie --enable-default-ssp --enable-gnu-indirect-function
--enable-gnu-unique-object --enable-install-libiberty --enable-linker-build-id
--enable-lto --enable-multilib --enable-plugin --enable-shared
--enable-threads=posix --disable-libssp --disable-libstdcxx-pch
--disable-libunwind-exceptions --disable-werror
gdc_include_dir=/usr/include/dlang/gdc
Thread model: posix
Supported LTO compression algorithms: zlib zstd
gcc version 10.2.0 (GCC)

gfortran -Wall -Wextra -fbounds-check -fno-strict-aliasing -fwrapv
-fno-aggressive-loop-optimizations -fsanitize=undefined arrays.f90 test.f90

or just

gfortran -fbounds-check arrays.f90 test.f90

test.f90

 PROGRAM test
 USE arrays

 call init_mat(MM,N,RM)

 write(*,*) RM%r(257,:)  ! this prints out the out-of-bounds slice INCORRECTLY
 write(*,*) RM%r(257,1)  ! this is caught correctly (line 7 below)
 write(*,*) size(RM%r)
 END

arrays.f90

MODULE arrays

 INTEGER, PARAMETER :: MM=180, N=22 

 TYPE RMatrix
   REAL, ALLOCATABLE :: r(:,:)
 END TYPE RMatrix

 TYPE(RMatrix) :: RM

 CONTAINS

subroutine init_mat(MM,N,RM) ! allocate arrays
  INTEGER, INTENT(IN) :: MM,N
  TYPE(RMatrix) :: RM  
  allocate (RM%r(MM,N))  
end subroutine init_mat

end module

produces
./a.out
   0.00000000       0.00000000       0.00000000       0.00000000      
0.00000000       0.00000000       0.00000000       0.00000000       0.00000000 
     0.00000000       0.00000000       0.00000000       0.00000000      
0.00000000       0.00000000       0.00000000       0.00000000       0.00000000 
     0.00000000       0.00000000       0.00000000       0.00000000    
At line 7 of file test.f90
Fortran runtime error: Index '257' of dimension 1 of array 'rm%r' above upper
bound of 180

Error termination. Backtrace:
#0  0x55edbad311c8 in ???
#1  0x55edbad31394 in ???
#2  0x7fc2b3ff7151 in ???
#3  0x55edbad2f16d in ???
#4  0xffffffffffffffff in ???

the correct result from the PGI/NVIDIA (v 20.7) compiler follows

pgf90 -C arrays.f90 test.f90

./a.out
0: Subscript out of range for array rm%r (test.f90: 6)
    subscript=257, lower bound=1, upper bound=180, dimension=1

[Bug fortran/97039] -fbounds-check misses violation with slice of array but not an element

[kargl at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=97039

kargl at gcc dot gnu.org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |kargl at gcc dot gnu.org

--- Comment #1 from kargl at gcc dot gnu.org ---
(In reply to Anthony M de Beus from comment #0)
> 
> the correct result from the PGI/NVIDIA (v 20.7) compiler follows
> 

As the Fortran code is invalid, there are no correct results.

[Bug fortran/97039] -fbounds-check misses violation with slice of array but not an element

[anthony.debeus at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=97039

--- Comment #2 from Anthony M de Beus <anthony.debeus at gmail dot com> ---
Clarification for anyone confused, "correct" results by a fortran compiler with
bounds-checking enabled would include finding/checking incorrect bounds in
(deliberately) incorrect fortran code given in the illustrative example.  

Hence in the example given, the pgf90 v.20.7 compiler correctly identifies the
incorrect fortran code involving the array slice, whereas the gfortran compiler
only catches the second error in the code, involving a specific element, but
not the first error, involving a slice.

[Bug fortran/97039] -fbounds-check misses violation with slice of array but not an element

[anlauf at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=97039

anlauf at gcc dot gnu.org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Keywords|                            |diagnostic
   Last reconfirmed|                            |2020-10-16
     Ever confirmed|0                           |1
             Status|UNCONFIRMED                 |NEW
                 CC|                            |anlauf at gcc dot gnu.org

--- Comment #3 from anlauf at gcc dot gnu.org ---
Reduced testcase:

program test
  implicit none
  integer, parameter :: m=2, n=3
  real, allocatable  :: r(:,:)
  allocate (r(m,n))  
  write(*,*) r(m+1,:)
  write(*,*) r(m+1,1)
end

Looking at the tree dump, there are array bounds checks generated for both
dimensions in the latter write, but there are suspiciously looking checks
for dimension 2 for the first case (r(m+1,:)).

[Bug fortran/97039] -fbounds-check misses violation with slice of array but not an element

[anlauf at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=97039

anlauf at gcc dot gnu.org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         Resolution|---                         |DUPLICATE
             Status|NEW                         |RESOLVED

--- Comment #4 from anlauf at gcc dot gnu.org ---
Should be fixed by r14-4039-g1cbf18978aa384 for pr30802.
Thus marking as duplicate.

*** This bug has been marked as a duplicate of bug 30802 ***