Description of problem: I was working on script for fixing broken non-utf8 file names (qt/kde can't handle them well), but found out that bash have some problems with iso-8859-2 strings when locale is set to utf-8 Version-Release number of selected component (if applicable): bash-4.2.7-1.fc15.x86_64 How reproducible: always Steps to Reproduce: 1. echo foo | tr [$'\xFD'$'\xBA'] [$'\xFD'$'\xBE'] Actual results: bash loops with 100 % CPU Expected results: foo Additional info: Tested with ksh and mksh shells and it works there. When you set LC_ALL=en_US.iso-8859-1 then it works in bash too: LC_ALL=en_US.iso-8859-1; echo foo | tr [$'\xFD'$'\xBA'] [$'\xFD'$'\xBE'] prints "foo" as expected.
Thanks for the report. The bug is in the bash globbing. Easy reproducer: [$'\xFD'$'\xBA'] Reported upstream.
bash-4.2.7-2.fc15 has been submitted as an update for Fedora 15. https://admin.fedoraproject.org/updates/bash-4.2.7-2.fc15
bash-4.2.8-2.fc15 has been submitted as an update for Fedora 15. https://admin.fedoraproject.org/updates/bash-4.2.8-2.fc15
bash-4.2.8-2.fc15 reverts the previous change. It introduced another bugs (bug #689868). This bug will be fixed in future version, as announced upstream.