The f2fs filesystem is a newer flash filesystem that is available in
meta-filesystems. This needs a minimum sized blank space to operate
in correctly.

The f2fs has overprovision functionality and when building smaller
filesystems, you actually need more space, therefore under 500M (or so)
double the amount of space needed in order to handle the overprovision
functions in f2fs.

Signed-off-by: Saul Wold <s...@linux.intel.com>
---
 meta/classes/image_types.bbclass | 20 ++++++++++++++++++++
 1 file changed, 20 insertions(+)

diff --git a/meta/classes/image_types.bbclass b/meta/classes/image_types.bbclass
index c736f92e58c..b17c435274e 100644
--- a/meta/classes/image_types.bbclass
+++ b/meta/classes/image_types.bbclass
@@ -208,6 +208,23 @@ IMAGE_CMD_ubi () {
 
 IMAGE_CMD_ubifs = "mkfs.ubifs -r ${IMAGE_ROOTFS} -o 
${IMGDEPLOYDIR}/${IMAGE_NAME}${IMAGE_NAME_SUFFIX}.ubifs ${MKUBIFS_ARGS}"
 
+MIN_F2FS_SIZE ?= "524288"
+IMAGE_CMD_f2fs () {
+        # We need to add additional smarts here form devices smaller than 1.5G
+        # Need to scale appropriately between 40M -> 1.5G as the "overprovision
+        # ratio" goes down as the device gets bigger (70% -> 4.5%), below about
+        # 500M the standard IMAGE_OVERHEAD_FACTOR does not work, so add 
additional
+        # space here when under 500M
+       size=${ROOTFS_SIZE}
+       if [ ${size} -lt ${MIN_F2FS_SIZE} ] ; then
+               size=${MIN_F2FS_SIZE}
+               bbwarn "Rootfs size is too small for F2FS. Filesystem will be 
extended to ${size}K"
+       fi
+       dd if=/dev/zero 
of=${IMGDEPLOYDIR}/${IMAGE_NAME}${IMAGE_NAME_SUFFIX}.f2fs seek=${size} count=0 
bs=1024
+       mkfs.f2fs ${EXTRA_IMAGECMD} 
${IMGDEPLOYDIR}/${IMAGE_NAME}${IMAGE_NAME_SUFFIX}.f2fs
+       sload.f2fs -f ${IMAGE_ROOTFS} 
${IMGDEPLOYDIR}/${IMAGE_NAME}${IMAGE_NAME_SUFFIX}.f2fs
+}
+
 EXTRA_IMAGECMD = ""
 
 inherit siteinfo kernel-arch
@@ -220,6 +237,7 @@ EXTRA_IMAGECMD_ext2 ?= "-i 4096"
 EXTRA_IMAGECMD_ext3 ?= "-i 4096"
 EXTRA_IMAGECMD_ext4 ?= "-i 4096"
 EXTRA_IMAGECMD_btrfs ?= "-n 4096"
+EXTRA_IMAGECMD_f2fs ?= ""
 
 do_image_jffs2[depends] += "mtd-utils-native:do_populate_sysroot"
 do_image_cramfs[depends] += "util-linux-native:do_populate_sysroot"
@@ -234,6 +252,7 @@ do_image_squashfs_lz4[depends] += 
"squashfs-tools-native:do_populate_sysroot"
 do_image_ubi[depends] += "mtd-utils-native:do_populate_sysroot"
 do_image_ubifs[depends] += "mtd-utils-native:do_populate_sysroot"
 do_image_multiubi[depends] += "mtd-utils-native:do_populate_sysroot"
+do_image_f2fs[depends] += "f2fs-tools-native:do_populate_sysroot"
 
 # This variable is available to request which values are suitable for 
IMAGE_FSTYPES
 IMAGE_TYPES = " \
@@ -251,6 +270,7 @@ IMAGE_TYPES = " \
     cpio cpio.gz cpio.xz cpio.lzma cpio.lz4 \
     wic wic.gz wic.bz2 wic.lzma \
     container \
+    f2fs \
 "
 
 # Compression is a special case of conversion. The old variable
-- 
2.14.3

-- 
_______________________________________________
Openembedded-core mailing list
Openembedded-core@lists.openembedded.org
http://lists.openembedded.org/mailman/listinfo/openembedded-core

Reply via email to