Openembedded Core Discussions
 help / color / mirror / Atom feed
* [PATCH] kernel.bbclass: Fix do_sizecheck behaviour
@ 2017-12-08 13:43 Mike Crowe
  0 siblings, 0 replies; only message in thread
From: Mike Crowe @ 2017-12-08 13:43 UTC (permalink / raw)
  To: openembedded-core; +Cc: Mike Crowe

During the introduction of support for multiple kernel image types in
849b67b2e4820564b5e5c9bd4bb293c44351c5f3, do_sizecheck was changed to only
warn if any kernel was bigger than ${KERNEL_IMAGE_MAXSIZE}. (Well, it tried
to warn - it turns out that there's no function called "warn", it should be
"bbwarn".) The previous behaviour had been to fail the build if the single
kernel image did not fit.

It seems possible that people might be generating both compressed and
uncompressed kernels and only really care whether the compressed one fits.
This means that we shouldn't just always fail if any of the images are too
large. So, let's warn (correctly this time) on every image that is too
large, but only ultimately fail if no image will fit. The build will also
fail if ${KERNEL_IMAGETYPES} is empty, but I hope that no-one needs to do
that.

While we're here correct a typo in the KERNEL_IMAGE_MAXSIZE validity check.

Signed-off-by: Mike Crowe <mac@mcrowe.com>
---
 meta/classes/kernel.bbclass | 14 +++++++++++---
 1 file changed, 11 insertions(+), 3 deletions(-)

Discussion: http://lists.openembedded.org/pipermail/openembedded-core/2017-December/145376.html

diff --git a/meta/classes/kernel.bbclass b/meta/classes/kernel.bbclass
index 756707a3c2..14f41e9b17 100644
--- a/meta/classes/kernel.bbclass
+++ b/meta/classes/kernel.bbclass
@@ -595,19 +595,27 @@ do_strip[dirs] = "${B}"
 addtask strip before do_sizecheck after do_kernel_link_images
 
 # Support checking the kernel size since some kernels need to reside in partitions
-# with a fixed length or there is a limit in transferring the kernel to memory
+# with a fixed length or there is a limit in transferring the kernel to memory.
+# If more than one image type is enabled, warn on any that don't fit but only fail
+# if none fit.
 do_sizecheck() {
 	if [ ! -z "${KERNEL_IMAGE_MAXSIZE}" ]; then
 		invalid=`echo ${KERNEL_IMAGE_MAXSIZE} | sed 's/[0-9]//g'`
 		if [ -n "$invalid" ]; then
-			die "Invalid KERNEL_IMAGE_MAXSIZE: ${KERNEL_IMAGE_MAXSIZE}, should be an integerx (The unit is Kbytes)"
+			die "Invalid KERNEL_IMAGE_MAXSIZE: ${KERNEL_IMAGE_MAXSIZE}, should be an integer (The unit is Kbytes)"
 		fi
+		at_least_one_fits=
 		for type in ${KERNEL_IMAGETYPES} ; do
 			size=`du -ks ${B}/${KERNEL_OUTPUT_DIR}/$type | awk '{print $1}'`
 			if [ $size -ge ${KERNEL_IMAGE_MAXSIZE} ]; then
-				warn "This kernel $type (size=$size(K) > ${KERNEL_IMAGE_MAXSIZE}(K)) is too big for your device. Please reduce the size of the kernel by making more of it modular."
+				bbwarn "This kernel $type (size=$size(K) > ${KERNEL_IMAGE_MAXSIZE}(K)) is too big for your device."
+			else
+				at_least_one_fits=y
 			fi
 		done
+		if [ -z "$at_least_one_fits" ]; then
+			die "All kernel images are too big for your device. Please reduce the size of the kernel by making more of it modular."
+		fi
 	fi
 }
 do_sizecheck[dirs] = "${B}"
-- 
2.11.0




^ permalink raw reply related	[flat|nested] only message in thread

only message in thread, other threads:[~2017-12-08 13:43 UTC | newest]

Thread overview: (only message) (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2017-12-08 13:43 [PATCH] kernel.bbclass: Fix do_sizecheck behaviour Mike Crowe

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox