[PATCH] D14706: [lit] Implement support of per test timeout in lit.

[Daniel Dunbar via llvm-commits]

Thank you for working on this!

The implementation sounds good to me -- I am fine having an initial version
of this feature that only works with an extra dependency, if it at least
helps to get the feature in place. It would still be nice to get a
completely built in version, eventually, but I think it is more important
to have a start than have it completely finished.

On that note, it might just be a matter of ensuring we spawn processes into
their own process group and killing the full process group to get rid of
the dependency, if you are interested in investigating.

My only comment on the implementation is that I would prefer not to have
the timeout line printed following the header on each invocation. I don't
think that is important enough info to merit always being output. I'm also
not sure it should be a warning if the user overrides the timeout on the
command line, honoring that seems expected behavior.

 - Daniel

On Friday, December 4, 2015, Dan Liew <dan at su-root.co.uk> wrote:

> On 27 November 2015 at 13:18, Dan Liew <dan at su-root.co.uk> wrote:
> > On 27 November 2015 at 13:16, Dan Liew <dan at su-root.co.uk> wrote:
> >> delcypher added a reviewer: MatzeB.
> >>
> >> http://reviews.llvm.org/D14706
> >
> > @ddunbar: Ping.
>
> @ddunbar: Ping.
>
> If I don't hear anything within a week. I'll just go ahead and merge this.
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20151214/b630f227/attachment.html>