On Fri, Oct 19, 2018 at 12:32:47PM +0800, Haojian Zhuang wrote:
On Fri, 19 Oct 2018 at 12:30, Leif Lindholm leif.lindholm@linaro.org wrote:
On Fri, Oct 19, 2018 at 12:15:49PM +0800, Haojian Zhuang wrote:
commit 1a0f11213a539a5d0c5d05bfaa68ccc75cb340aa Author: Soby Mathew soby.mathew@arm.com Date: Mon Oct 1 16:16:34 2018 +0100
Update the version to 2.0 Change-Id: Icbc556d47a58d0870577b1bf1cd27cc5827fd56d Signed-off-by: Soby Mathew <soby.mathew@arm.com>Since ARM Trusted Firmware version is updated to 2.0, the calculation of ATF is wrong. It could only calculate version for 1.x.
Whoops.
Update the calculation to support 2.x or n.x.
Signed-off-by: Haojian Zhuang haojian.zhuang@linaro.org
atf-build.sh | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-)
diff --git a/atf-build.sh b/atf-build.sh index fb80ad0..c6ec3e3 100755 --- a/atf-build.sh +++ b/atf-build.sh @@ -25,9 +25,12 @@ function check_atf_buildver MAJOR=`grep "^VERSION_MAJOR" Makefile | sed 's/.*:= *([0-9]*).*/\1/'` [ $? -ne 0 ] && return 1 MINOR=`grep "^VERSION_MINOR" Makefile | sed 's/.*:= *([0-9]*).*/\1/'`
[ $? -ne 0 ] && return 1Why is this test deleted? This checks the command success, not the version number.
When the version is 2.0, the MINOR variable is 0. If we keep checking at here, we'll get exit and ATF_BUILDVER keeps in 1. So we should only check MAJOR variable.
I understand that. But why are you deleting a test for whether the grep command was successful?
Regards,
Leif