Red Hat Bugzilla – Bug 141589
post-increments is wreaking havoc on tests
Last modified: 2015-01-07 19:09:01 EST
Description of problem:
Definitely not sure what's happening here, but the post-increment actions in the
tests appear to be causing all sorts of chaos. Post-incrementing appears to
work just fine outside of the rhr2 scripts, so not sure if that really tells us
anything or not. Specifically the USB tests has lots of issues. Running
rhr2-rhel4-0.9-14.9e.dummyrun with "USB 2" in the hardware.conf file, I'm seeing
the following (full output.log will be attached):
+ local 'usb_device_map=004/001 "UHCI Host Controller" hub
003/001 "UHCI Host Controller" hub
002/001 "UHCI Host Controller" hub
001/001 "EHCI Host Controller" hub'
+ '[' '' = legacyfree ']'
+ '[' 0 -ge 2 ']'
+ let usb_port++
So, I get to the post-increment and then things just exit . . . very weird.
Changing these from "let <blah>++" to "let <blah>=<blah>+1" fixes the issue and
the window actually pops prompting to plug and unplug the USB devices.
Version-Release number of selected component (if applicable):
Steps to Reproduce:
Created attachment 107758 [details]
Turns out that while post increment is working it exits "1" instead of "0" I
submitted this as a bug only to be informed that it's expected behaviour (I
disagree, but I digress); I've got an updated which replaces let bob++ with let
++bob which appears to make it work (and seems backwards based on the man page,
but I digress (again)); though it's having the side effect of not stopping @ 1
port as defined in hardware.conf.
addressed in rhr2-0.9-14.9f; eg. fixed in CVS
Closing per rlandrys last comment.