[poky] [PATCH 1/1] scripts/bitbake: Remove sanity_info after run "bitbake -e"

Richard Purdie richard.purdie at linuxfoundation.org
Tue Apr 17 04:19:12 PDT 2012


On Tue, 2012-04-17 at 15:46 +0800, Dongxiao Xu wrote:
> With latest master, we found certain sanity check doesn't work
> (e.x., the network check). This is because it created the sanity_info
> file at the first time when running "bitbake -e" command in
> scripts/bitbake, therefore while running the real bitbake command,
> it finds that the sanity_check file does exist so it skip the network
> check.
> 
> The fix is to remove the sanity_info file after running "bitbake -e"
> command.
> 
> Signed-off-by: Dongxiao Xu <dongxiao.xu at intel.com>
> ---
>  scripts/bitbake |    1 +
>  1 files changed, 1 insertions(+), 0 deletions(-)
> 
> diff --git a/scripts/bitbake b/scripts/bitbake
> index 3772d82..a6fb2e5 100755
> --- a/scripts/bitbake
> +++ b/scripts/bitbake
> @@ -73,6 +73,7 @@ if [ $needpseudo = "1" ]; then
>          PSEUDOBINDIR=`cat $BUILDDIR/pseudodone`
>      else
>          PSEUDOBINDIR=`bitbake -e | grep STAGING_BINDIR_NATIVE=\" | cut -d '=' -f2 | cut -d '"' -f2`
> +        rm -rf conf/sanity_info
>      fi
>      if [ -e "$PSEUDOBINDIR/pseudo" ]; then
>          buildpseudo="0"

This forces the sanity tests to run on every bitbake run which is not
acceptable. We need to find a better solution.

This also sets off alarm bells in my mind. If the sanity tests are
failing, shouldn't they fail equally on a subsquent run of bitbake?

See my follow up patch which should fix this problem in a better way.

Cheers,

Richard




More information about the poky mailing list