Thanks for pointing out that the workers get signalled multiple times, I missed that indeed. In that case, termination of the module under valgrind takes a little longer then I thought it did, yet the problem remains the same. 
So the upper boundary of 1000 ms for the iteration has to remain fixed ? In that case, we'll have a patch to maintain (or see if we can round up in less time). 


On Mon, Mar 23, 2015 at 11:18 PM, Maxim Dounin <> wrote:

On Mon, Mar 23, 2015 at 08:39:46PM +0100, Otto van der Schaaf wrote:

> Hi,
> For testing quick termination during high loads, while running with
> valgrind, it might be useful to be able to extend the amount of time nginx
> allows child processes to wrap up before sending SIGKILL.
> For ngx_pagespeed, the current hard-coded default of 1 second seems to be
> just short of what we need to be able to reliably test just this scenario,
> so I've made a patch so we can run with different values for with and
> without valgrind.
> Would the following patch be acceptable?

No.  Note well that current limit isn't 1 second, you are
misunderstanding the code.  It'll signal workers multiple times,
doubling wait time on each iteration, from 50ms to 1000ms

Maxim Dounin

nginx-devel mailing list