[Intel-gfx] [PATCH i-g-t] docs/chamelium: Explain that the Chamelium should only target one DUT
Lyude Paul
lyude at redhat.com
Wed Aug 30 18:40:43 UTC 2017
R-B'd and pushed, thanks!
On Tue, 2017-08-29 at 18:11 +0300, Paul Kocialkowski wrote:
> This adds an explanation about why the Chamelium should only be
> connected to one target device at once to the in-tree documentation.
>
> Signed-off-by: Paul Kocialkowski <paul.kocialkowski at linux.intel.com>
> ---
> docs/chamelium.txt | 6 ++++++
> 1 file changed, 6 insertions(+)
>
> diff --git a/docs/chamelium.txt b/docs/chamelium.txt
> index 77594284..ae7ac34a 100644
> --- a/docs/chamelium.txt
> +++ b/docs/chamelium.txt
> @@ -54,6 +54,12 @@ CI system with a shared testlist) to remove the
> Chamelium configuration from the
> hosts that shouldn't connect to the Chamelium so that they can be
> skipped, which
> is faster than a network timeout.
>
> +It should also be noted that each Chamelium platform should only be
> used for
> +testing a single target device at a time. This is because the reset
> call issued
> +by the IGT tests is common to all connectors and thus one machine
> running a test
> +on a given connector may reset the chamelium while another machine
> is running
> +a test on another connector.
> +
> An example fully-featured configuration follows:
> [Common]
> FrameDumpPath=/root/
--
Cheers,
Lyude
More information about the Intel-gfx
mailing list