[OE-core] [PATCH 1/1] sanity.bbclass: Don't fatal() when trying to dump environment.

Richard Purdie richard.purdie at linuxfoundation.org
Wed Nov 20 10:19:04 UTC 2013


On Tue, 2013-11-19 at 18:31 -0600, Peter Seebach wrote:
> It is distressing when bitbake -e can't run due to sanity check
> failures, because the solution often involves figuring out what
> configuration file are being read and what values they are setting.
> Solution: Don't do that. If the datasmart object reports that
> tracking is enabled, emit an error() but not a fatal(), and also
> warn the user that things may be badly broken.
> 
> This depends on a change to bitbake (adding the isTracking() method).
> 
> Signed-off-by: Peter Seebach <peter.seebach at windriver.com>
> ---
>  meta/classes/sanity.bbclass |    8 +++++++-
>  1 files changed, 7 insertions(+), 1 deletions(-)
> 
> diff --git a/meta/classes/sanity.bbclass b/meta/classes/sanity.bbclass
> index 6807a23..1ef811d 100644
> --- a/meta/classes/sanity.bbclass
> +++ b/meta/classes/sanity.bbclass
> @@ -82,11 +82,17 @@ def raise_sanity_error(msg, d, network_error=False):
>              bb.event.fire(bb.event.SanityCheckFailed(msg), d)
>          return
>  
> -    bb.fatal(""" OE-core's config sanity checker detected a potential misconfiguration.
> +    msg = (""" OE-core's config sanity checker detected a potential misconfiguration.
>      Either fix the cause of this error or at your own risk disable the checker (see sanity.conf).
>      Following is the list of potential problems / advisories:
>      
>      %s""" % msg)
> +    if d.isTracking():
> +	bb.warn("Environment tracking enabled: NOT exiting on sanity check failures.")
> +	bb.warn("Results may be inconsistent or meaningless; proceed at your own risk.")
> +        bb.error(msg)
> +    else:
> +        bb.fatal(msg)

This is creative but you just disabled the sanity tests in the graphical
UIs like hob.

isTracking() does not mean -e is being used, we use the tracking in
other places now too.

Cheers,

Richard




More information about the Openembedded-core mailing list