Arch Linux

Please read this before reporting a bug:
https://wiki.archlinux.org/title/Bug_reporting_guidelines

Do NOT report bugs when a package is just outdated, or it is in the AUR. Use the 'flag out of date' link on the package page, or the Mailing List.

REPEAT: Do NOT report bugs for outdated packages!
Tasklist

FS#60001 - gfortran runtime segfault at formated screen output by write(*,fmt) statement

Attached to Project: Arch Linux
Opened by Johannes Rieke (jrieke) - Sunday, 09 September 2018, 21:19 GMT
Last edited by Evangelos Foutras (foutrelis) - Monday, 10 September 2018, 04:42 GMT
Task Type Bug Report
Category Packages: Core
Status Closed
Assigned To No-one
Architecture x86_64
Severity Medium
Priority Normal
Reported Version
Due in Version Undecided
Due Date Undecided
Percent Complete 100%
Votes 0
Private No

Details

Dear Arch-GCC team,

I stumbled over an issue at current gfortran:

Description:
Simple Fortran code fails after building without errors with segfault at runtime in line with formatted write statement:

* minimum working example:

program test_gfortran8_2_write_format_error
implicit none
real(8) :: real_dummy
real_dummy = 42.0_8
write(*,*) real_dummy
write(*,'("This is the answer to all questions:",f4.0)') real_dummy ! this line causes runtime error
end program test_gfortran8_2_write_format_error


* runtime error:

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

Backtrace for this error:
#0 0x7f0d8e13cdff in ???
#1 0x7f0d8e19ec84 in ???
#2 0x7f0d8e18e4e1 in ???
#3 0x7f0d8eae64c7 in data_transfer_init
at ../../../gcc-8-20171217/libgfortran/io/transfer.c:2791
#4 0x55ecc044926e in ???
#5 0x55ecc04492e8 in ???
#6 0x7f0d8e129222 in ???
#7 0x55ecc04490bd in ???
#8 0xffffffffffffffff in ???
Segmentation fault (core dumped)




Additional info:
* package version(s):
pacman -Si gcc
Repository : core
Name : gcc
Version : 8.2.1+20180831-1
Description : The GNU Compiler Collection - C and C++ frontends
Architecture : x86_64
URL : http://gcc.gnu.org
Licenses : GPL LGPL FDL custom
Groups : base-devel
Provides : gcc-multilib
Depends On : gcc-libs=8.2.1+20180831-1 binutils>=2.28 libmpc
Optional Deps : lib32-gcc-libs: for generating code for 32-bit ABI
Conflicts With : None
Replaces : gcc-multilib
Download Size : 32,89 MiB
Installed Size : 132,45 MiB
Packager : Evangelos Foutras <evangelos@foutrelis.com>
Build Date : Fr 31 Aug 2018 14:44:58 CEST
Validated By : MD5 Sum SHA-256 Sum Signature

* system:
Kernel: 4.17.19-1-MANJARO x86_64 bits: 64 Desktop: Xfce 4.12.4 Distro: Manjaro Linux


Steps to reproduce:
1) Compile test_gfortran8_2_write_format_error.f90 without additional options --> gfortran test_gfortran8_2_write_format_error.f90
2) run ./a.out

I had a similar error with gcc 7.2 at the beginning of this year: https://forum.manjaro.org/t/latest-update-to-gcc-7-2-1-20171128-causes-error-in-fortran-code-file-io-writing-lenghty-strings-in-formatted-form/36746
This old one was filed already. Custom builds with 7.2 from nightly had not this issue for the old bug. I found no information on this bug and I have not tried to compile GCC on my own this time. gfortran 7.2.0 runs fine with this code.

Best regards,
Johannes
This task depends upon

Closed by  Evangelos Foutras (foutrelis)
Monday, 10 September 2018, 04:42 GMT
Reason for closing:  Works for me
Additional comments about closing:  Wrong libgfortran linked/used.
Comment by Evangelos Foutras (foutrelis) - Monday, 10 September 2018, 04:37 GMT
Works for me. Your backtrace also indicates that you're using the wrong libgfortran:

> at ../../../gcc-8-20171217/libgfortran/io/transfer.c:2791

Loading...