Re: [i-g-t PATCH v2 1/4] lib/igt_core: Add igt_exec helpers

From: Chris Wilson <chris@chris-wilson.co.uk>
To: Abdiel Janulgue <abdiel.janulgue@linux.intel.com>
Cc: intel-gfx@lists.freedesktop.org
Subject: Re: [i-g-t PATCH v2 1/4] lib/igt_core: Add igt_exec helpers
Date: Tue, 30 May 2017 11:55:08 +0100	[thread overview]
Message-ID: <20170530105508.GG23936@nuc-i3427.alporthouse.com> (raw)
In-Reply-To: <47bb4176-0895-4813-d563-21e7b1c955de@linux.intel.com>

On Tue, May 30, 2017 at 01:31:46PM +0300, Abdiel Janulgue wrote:
> 
> 
> On 29.05.2017 16:01, Chris Wilson wrote:
> > On Mon, May 29, 2017 at 03:08:07PM +0300, Abdiel Janulgue wrote:
> >> Support executing external processes with the goal of capturing its
> >> standard streams to the igt logging infrastructure in addition to its
> >> exit status.
> > 
> > This is not an exec wrapper. This is a system() replacement. That it
> > invokes shell to evaluate the cmd should be front and centre.
> > 
> 
> The problem with system() is that it waits until the process is
> completed. If the process' output is more than PIPE_BUF, it'll block. Do
> you have any suggestions?

I'm just saying the name is misleading. igt_system() and explain the
improvements (concurrency and accessible piped output).
-Chris

-- 
Chris Wilson, Intel Open Source Technology Centre
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx