RCN kernel build directory patch

GoogleGroups
1

I'm guessing that there's a good reason for this, but I was debugging
an issue with the RobertCNelson/linux-dev.git kernel (am33x-v3.8
branch) and found that the issue had been fixed, but was commented
out. The patch in question was
"patches/build/0004-Make-single-.dtb-targets-also-with-DTC_FLAGS.patch".
Commenting in the patch in the patch.sh file fixes my issue, but it
makes me wonder whether something more is going on that I should be
aware of.

So, is there anything more to this patch? If not, could it be
commented in? Also, any issues with the other build directory patches?
They look interesting too.

Thanks,
Frank

2

I'm guessing that there's a good reason for this, but I was debugging
an issue with the RobertCNelson/linux-dev.git kernel (am33x-v3.8
branch) and found that the issue had been fixed, but was commented
out. The patch in question was
"patches/build/0004-Make-single-.dtb-targets-also-with-DTC_FLAGS.patch".
Commenting in the patch in the patch.sh file fixes my issue, but it
makes me wonder whether something more is going on that I should be
aware of.

So... What bug/problem does it fix?

So, is there anything more to this patch? If not, could it be
commented in? Also, any issues with the other build directory patches?
They look interesting too.

Regards,

3

I'm guessing that there's a good reason for this, but I was debugging
an issue with the RobertCNelson/linux-dev.git kernel (am33x-v3.8
branch) and found that the issue had been fixed, but was commented
out. The patch in question was
"patches/build/0004-Make-single-.dtb-targets-also-with-DTC_FLAGS.patch".
Commenting in the patch in the patch.sh file fixes my issue, but it
makes me wonder whether something more is going on that I should be
aware of.

So... What bug/problem does it fix?

For me, the eMMC cape and presumably other ones failed to "resolve
tree" at runtime. I traced this down to the am335x-boneblack.dtb being
different (around 16 KB) than the one produce by the
beagleboard/kernel (24129 bytes). This turned out to be due to some
arguments not being passed to dtc. That particular patch passed the
dtc arguments and fixed the issue.