[PATCH] find-debuginfo.sh: Exit with real exit status in parallel jobs

[Keith Seitz]

On 8/17/24 2:20 PM, Mark Wielaard wrote:
> On Fri, Aug 16, 2024 at 11:54:20AM -0700, Keith Seitz wrote:
>> Currently, when the script is executed in parallel (-jN), the
>> resulting exit status will always be 0.
>>
>> The script execs an appropriate number of clones of itself, calling
>> run_job to run the actual workload. This then calls do_file(), saving
>> the exit status into "res.$jobid".
>>
>> In do_file(), though, if an error occurs, exit is called. This causes
>> the entire exec'd shell to exit with status 0 (since there are almost
>> always echo calls as the last executed statement). The real exit
>> status is therefor never written to the "res.$jobid" files by run_job().
>>
>> The simple solution is to use 'return' instead of 'exit'. A number
>> of minor adjustments are also made to propagate this properly so that
>> it is reported as the correct exit status.
> 
> Good find. Code looks correct. Except don't we now also need to check
> the result of do_file () in the non-parallel case? Something like:
> 
> diff --git a/scripts/find-debuginfo.in b/scripts/find-debuginfo.in
> index f40e566f86c3..5998b9d1fa6c 100755
> --- a/scripts/find-debuginfo.in
> +++ b/scripts/find-debuginfo.in
> @@ -579,6 +579,10 @@ fi
>   if [ $n_jobs -le 1 ]; then
>     while read nlinks inum f; do
>       do_file "$nlinks" "$inum" "$f"
> +    res=$?
> +    if [ "$res" != "0" ]; then
> +      exit $res
> +    fi
>     done <"$temp/primary"
>   else
>     for ((i = 1; i <= n_files; i++)); do

Yes, that does look appropriate. My apologies for not following through
with the non-parallel invocation. I guess I could have (further) 
modified my custom rpm package to specifically test this case, too,
and submit a more complete patch.

>> While at it, I've incorporated a patch for find-debuginfo/30505.
>> Using this patch and another patch to the RPM package (submitted as
>> github issue #3215), failures of gdb-add-index.sh will now properly fail
>> the build instead of being swallowed. It should be much easier for
>> developers to figure out why their builds have failed should gdb crash.
> 
> Thanks, that has been a long standing issue.  Smart to introduce a new
> _find_debuginfo_exit_on_error for rpm-config. The reason errors
> weren't reported in the past was because people would then disable
> creating debuginfo packages completely. But these days we really
> should report (and fix) any errors.
> 
> There are a couple of other places, the main debugedit invocation,
> strip_to_debug, add_minidebug, which really should report errors
> too. But that can wait.

Indeed. I'm selfish, so I am attempting to improve testing for
areas for which I am directly responsible, but perhaps I can add
that to a TODO and give it a try with my mass prebuilder some time
to verify.

In any case, a step in the right direction. Thank you for the review!

Keith