Re: builtin man page for wait omits information from SIGNALS

[AA]

Greg:  I included a link to the github. There are only 3 PRs, one of 
which is mine. That's why I was suspicious it's the wrong place.

Here is a link to the actual patch: 
https://github.com/gitGNU/gnu_bash/pull/5/files

Robert: Yes, I meant the section of "man bash" that discusses the 
builtins, not man 2 wait, or some unrelated man page.

I am unaware of the man page politics of gnu bash. Therefore, if you 
could educate/direct me to the right place to ensure the patch at the 
link above is considered, I'd be happy to follow up. Seems to me at 
least an xref to the SIGNALS section seems warranted. One otherwise has 
to have read the entire man page to happen across this "early exit" 
behavior of wait. I suppose that's why wizards have long beards.


What follows touches on aspects of wait and bash that lead to the 
foregoing doc PR.

when implementing a wrapper for an external executable running a 
kubernetes pod, I catch SIGTERM, forward it to the child process, and 
then I wish to wait for the child process to actually terminate. I do 
this by:

_sigterm() {   kill -TERM 6; 2>/dev/null; }; trap _sigterm SIGTERM;  
python foo.py  2>&1 | rotatelogs -v -e -L foo.log 
foo.logfile.%Y-%m-%d-%H_%M_%S 1G 2>> rotatelogs.log & wait 6; wait 6; }

(1) The wait 6 is a reference to the PID of python. It's always the same 
because the pod startup is deterministic. Getting at that PID 
symbolically seems quite painful. It wasn't clear from the manpage if %- 
would give me that PID, nor did I find any good way to refer to the job 
IDs of members of the most recent pipe. If it doesn't exist, this would 
seem to be a useful addition to the bash jobs reference variable syntax. 
If %- works then perhaps the syntax could be %-N to refer to the N'th 
back PID in the most recent pipe built by bash?

Building a named pipe so I can execute the components separately in 
order to access $! seems absurdly roundabout. I refuse! :)

(2) The double wait 6 is so that the parent bash doesn't immediately die 
after delivering the SIGTERM to the child python. It wasn't clear from 
the man page whether one can or should wait inside a signal handler 
function. Would that have been the better place to put the second wait 
6? Perhaps it's exactly equivalent. Does the answer warrant an 
additional doc PR? Happy to write it.

All the best,
A.



On 6/15/22 5:52 AM, Robert Elz wrote:
>      Date:        Wed, 15 Jun 2022 07:37:07 -0400
>      From:        Greg Wooledge<greg@wooledge.org>
>      Message-ID:<YqnEYyQUBSeT7WEB@wooledge.org>
>
>    | Bash does not ship a man page named "wait".  What are you actually
>    | reading?
>
> I assumed the section of "man bash" which describes the wait
> builtin command, describes the various exit status possibilities
> from wait, but doesn't mention the case where wait is interrupted.
>
> Whether it is worth mentioning there, or providing a "for what
> happens if a signal occurs during the wait see..." xref isn't
> my call, but it is a notable omission.
>
> kre