Defect: Invalid RMA address range accessing coarray element with allocatable component on remote image

[everythingfunctional]

• I am reporting a bug others will be able to reproduce and not asking a question or requesting a new feature.

System information including:

• OpenCoarrays Version: 2.10.1
• Fortran Compiler: gfortran 12.2.0
• C compiler used for building lib: gcc 12.2.0
• Installation method: ./install.sh
• All flags & options passed to the installer: -M /usr
• Output of uname -a: Linux stray 6.0.11-arch1-1 #1 SMP PREEMPT_DYNAMIC Fri, 02 Dec 2022 17:25:31 +0000 x86_64 GNU/Linux
• MPI library being used: openmpi 4.1.4-4
• Machine architecture and number of physical cores: Intel x86, 4 core
• Version of CMake: 3.25.1 installed via package manager

To help us debug your issue please explain:

What you were trying to do (and why)

Use derived type with allocatable component in coarray, Example

module payload_m
    implicit none
    private
    public :: payload_t, empty_payload

    type :: payload_t
        !! A raw buffer to facilitate data transfer between  images
        !!
        !! Facilitates view of the data as either a string or raw bytes.
        !! Typical usage will be either to
        !! * produce a string representation of the data, and then parse that string to recover the original data
        !! * use the `transfer` function to copy the raw bytes of the data
        private
        integer, allocatable, public :: payload_(:)
    contains
        private
        procedure, public :: raw_payload
        procedure, public :: string_payload
    end type

    interface payload_t
        pure module function from_raw(payload) result(new_payload)
            implicit none
            integer, intent(in) :: payload(:)
            type(payload_t) :: new_payload
        end function

        pure module function from_string(payload) result(new_payload)
            implicit none
            character(len=*), intent(in) :: payload
            type(payload_t) :: new_payload
        end function

        module procedure empty_payload
    end interface

    interface
        pure module function empty_payload()
            implicit none
            type(payload_t) :: empty_payload
        end function

        pure module function raw_payload(self)
            implicit none
            class(payload_t), intent(in) :: self
            integer, allocatable :: raw_payload(:)
        end function

        pure module function string_payload(self)
            implicit none
            class(payload_t), intent(in) :: self
            character(len=:), allocatable :: string_payload
        end function
    end interface

end module

submodule(payload_m) payload_s
    implicit none
contains
    module procedure from_raw
        new_payload%payload_ = payload
    end procedure

    module procedure from_string
        new_payload = payload_t([len(payload), transfer(payload,[integer::])])
    end procedure

    module procedure empty_payload
        empty_payload%payload_  = [integer::]
    end procedure

    module procedure raw_payload
        if (allocated(self%payload_)) then
            raw_payload = self%payload_
        else
            raw_payload = [integer::]
        end if
    end procedure

    module procedure string_payload
        if (allocated(self%payload_)) then
            if (size(self%payload_) > 0) then
                allocate(character(len=self%payload_(1)) :: string_payload)
                if (len(string_payload) > 0) &
                    string_payload = transfer(self%payload_(2:),string_payload)
            else
                allocate(character(len=0) :: string_payload)
            end if
        else
            allocate(character(len=0) :: string_payload)
        end if
    end procedure

end submodule

program example
    use payload_m, only: payload_t

    character(len=*), parameter :: MESSAGE = "Hello, World!"
    type(payload_t) :: mailbox[*]

    if (this_image() == 1) then
        mailbox = payload_t(MESSAGE)
    end if
    sync all
    if (this_image() /= 1) then
        mailbox = mailbox[1]
    end if

    print *, mailbox%string_payload(), " from image: ", this_image()
end program

What happened (include command output, screenshots, logs, etc.)

$ caf main.f90 
$ cafrun -n 2 ./a.out
 Hello, World! from image:            1

Program received signal SIGSEGV: Segmentation fault - invalid memory reference.

Backtrace for this error:
#0  0x7fb8c68519ff in ???
#1  0x55dfaf068c12 in ???
#2  0x55dfaf06a3b7 in ???
#3  0x55dfaf06a4ad in ???
#4  0x7fb8c683c28f in ???
#5  0x7fb8c683c349 in ???
#6  0x55dfaf068564 in _start
	at ../sysdeps/x86_64/start.S:115
#7  0xffffffffffffffff in ???
--------------------------------------------------------------------------
Primary job  terminated normally, but 1 process returned
a non-zero exit code. Per user-direction, the job has been aborted.
--------------------------------------------------------------------------
--------------------------------------------------------------------------
mpiexec noticed that process rank 1 with PID 0 on node stray exited on signal 11 (Segmentation fault).
--------------------------------------------------------------------------
Error: Command:
   `/usr/bin/mpiexec -n 2 ./a.out`
failed to run.

with a more complex example (specifically the FEATS quadratic_solver example which can be found here in the examples folder) the error message is

[stray:40367] *** An error occurred in MPI_Get
[stray:40367] *** reported by process [2800353281,1]
[stray:40367] *** on win rdma window 5
[stray:40367] *** MPI_ERR_RMA_RANGE: invalid RMA address range
[stray:40367] *** MPI_ERRORS_ARE_FATAL (processes in this win will now abort,
[stray:40367] ***    and potentially your MPI job)
Error: Command:
   `/usr/bin/mpiexec -n 2 build/caf_3E83561A255DFE60/example/quadratic_solver`
failed to run.

What you expected to happen

The above example should print "Hello, World! from image: 1" for each image.

Step-by-step reproduction instructions to reproduce the error/bug

Simply compile the provided example as shown above. This was originally reported as follow up in #739, but was worthy of its own issue.

[vehre]

Note: Above patch needs a gfortran patched with https://gcc.gnu.org/pipermail/fortran/2024-December/061439.html

[vehre]

@zbeekman Why did you reopen this issue? The core meaning of #792 was to fix this issue.

[vehre]

Oh, I see. In the PR #794 I errorneously referenced this issue. I'll fix that.