case-lib: hijack.sh use trap to hijack exit command #229
Conversation
|
@Bin-QA Once we use this version of exit, we need to clean up old exit calls in test case, right? |
|
@aiChaoSONG current without more old exit |
|
@Bin-QA , Yes, you are right. when use trap, the func_hijack_exit will be automatically called. Now got the idea |
case-lib/hijack.sh
Outdated
| ;; | ||
| esac | ||
|
|
||
| trap '' EXIT |
There was a problem hiding this comment.
I don't think you need this.
There was a problem hiding this comment.
I missed all the processing done on the exit code, sorry. I don't think you don't have access to the exit code in trap EXIT, so I don't think this can work. How did you test it?
There was a problem hiding this comment.
yes, you are right, I already test it without this line, it can be worked without recursive call the trap function
There was a problem hiding this comment.
@Bin-QA I don't understand your last comment, I don't understand how this can work and I would like to know with what failure and non-zero exit you tested this change because I really don't understand how this can report failures. Try this:
trap 'func_exit_handler' EXIT
func_exit_handler()
{
printf '$1 = |%s|\n' "$1"
local exit_status=${1:-0}
printf '(wrong) exit status = |%s|\n' "${exit_status}"
}
exit 5This prints:
$1 = ||
(wrong) exit status = |0|
Cc: @xiulipan
There was a problem hiding this comment.
Sorry for miss your point
use 'trap' command hijack exit command instead of function 'exit' which perpare for 'set -e' Signed-off-by: Wu, BinX <binx.wu@intel.com>
xiulipan
left a comment
xiulipan
left a comment
There was a problem hiding this comment.
LGTM, let's track the test status to see if this will bring any regression.
|
I still don't understand sorry. Which failure and non-zero exit_status was #229 tested with? |
|
There are still a number of |
In case test verify and simple step, it is not effect in current script, |
4 participants
use 'trap' command hijack exit command instead of function
'exit' which perpare for 'set -e'
Signed-off-by: Wu, BinX binx.wu@intel.com