On 3/10/2016 12:29 PM, Panu Matilainen wrote: > On 03/10/2016 02:22 PM, Ferruh Yigit wrote: >> On 3/10/2016 10:53 AM, Panu Matilainen wrote: >>> When doing multiple runs of validate-abi.sh, the git status check >>> will more often than not unnecessarily fail with "Working directory not >>> clean" error because of the compat_result and compile target directories >>> from the previous run. Filter out the self-generated directories >>> when checking. >>> >>> Signed-off-by: Panu Matilainen <pmatilai at redhat.com> >>> --- >>> scripts/validate-abi.sh | 3 +-- >>> 1 file changed, 1 insertion(+), 2 deletions(-) >>> >>> diff --git a/scripts/validate-abi.sh b/scripts/validate-abi.sh >>> index ea60639..a21f883 100755 >>> --- a/scripts/validate-abi.sh >>> +++ b/scripts/validate-abi.sh >>> @@ -163,8 +163,7 @@ log "INFO" "against DPDK DSOs built from version $TAG2." >>> log "INFO" "" >>> >>> # Check to make sure we have a clean tree >>> -git status | grep -q clean >>> -if [ $? -ne 0 ] >>> +if [ $(git status --porcelain | grep -vE "($TARGET|compat_reports)" | wc >>> -l) -ne 0 ] >>> then >>> log "WARN" "Working directory not clean, aborting" >>> cleanup_and_exit 1 >>> >> Hi Panu, >> >> This check catches untracked files too, does it makes sense to limit >> error only to modified files (local or staged)? > > I did ponder about that, untracked files seem mostly harmless in this > picture but erred on the side of caution. > This is something prevents me running script from working tree, and forces to create a new clone.
>> >> This also prevents specific "compat_reports" folder check. >> >> And of course mentioned change requires "git clean -fd" removed, or >> replaced with "make clean" > > Sorry, I dont understand you mean by these two comments. > If untracked files accepted by script, "compat_reports" exclusion is no more required, and "git clean -fd" needs to removed from script. Regards, ferruh
