[PATCH 0/4] image_types.bbclass: fix btrfs, ubifs and elf

[PATCH 0/4] image_types.bbclass: fix btrfs, ubifs and elf

[Robert Yang]

The following changes since commit 5ebea1114d54120d5ff5d8a6cd148110d0fda23f:

  build-appliance-image: Update to master head revision (2016-04-06 23:54:57 +0100)

are available in the git repository at:

  git://git.openembedded.org/openembedded-core-contrib rbt/img
  http://cgit.openembedded.org/cgit.cgi/openembedded-core-contrib/log/?h=rbt/img

Robert Yang (4):
  image_types.bbclass: set nodesize for btrfs
  image_types.bbclass: fix ubi ubifs and multiubi
  image_types.bbclass: fix elf
  image-vm.bbclass/image_types.bbclass: IMAGE_NAME -> IMAGE_LINK_NAME

 meta/classes/image-vm.bbclass    |  2 +-
 meta/classes/image_types.bbclass | 10 +++++++---
 2 files changed, 8 insertions(+), 4 deletions(-)

-- 
2.7.4

[PATCH 1/4] image_types.bbclass: set nodesize for btrfs

[Robert Yang]

The default value is 16K which is too big to create image for
core-image-minimal:
rootfs.btrfs is too small to make a usable filesystem
Minimum size for each btrfs device is 41943040.

Use 4K as ext2/3/4 to fix the problem.

Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
---
 meta/classes/image_types.bbclass | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/meta/classes/image_types.bbclass b/meta/classes/image_types.bbclass
index 3e98024..3f97ddd 100644
--- a/meta/classes/image_types.bbclass
+++ b/meta/classes/image_types.bbclass
@@ -218,7 +218,7 @@ EXTRA_IMAGECMD_jffs2 ?= "--pad ${JFFS2_ENDIANNESS} --eraseblock=${JFFS2_ERASEBLO
 EXTRA_IMAGECMD_ext2 ?= "-i 4096"
 EXTRA_IMAGECMD_ext3 ?= "-i 4096"
 EXTRA_IMAGECMD_ext4 ?= "-i 4096"
-EXTRA_IMAGECMD_btrfs ?= ""
+EXTRA_IMAGECMD_btrfs ?= "-n 4096"
 EXTRA_IMAGECMD_elf ?= ""
 
 IMAGE_DEPENDS = ""
-- 
2.7.4

[PATCH 2/4] image_types.bbclass: fix ubi ubifs and multiubi

[Robert Yang]

Fixed when do_image_ubi:
| DEBUG: Executing shell function do_image_ubifs
| Error: min. I/O unit was not specified (use -h for help)
| WARNING: exit code 255 from a shell command.

Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
---
 meta/classes/image_types.bbclass | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/meta/classes/image_types.bbclass b/meta/classes/image_types.bbclass
index 3f97ddd..caa4dbb 100644
--- a/meta/classes/image_types.bbclass
+++ b/meta/classes/image_types.bbclass
@@ -120,6 +120,9 @@ IMAGE_CMD_elf () {
 }
 IMAGE_TYPEDEP_elf = "cpio.gz"
 
+MKUBIFS_ARGS ?= "-m 1KiB -e 128KiB -c $ROOTFS_SIZE"
+UBINIZE_ARGS ?= "-m 1KiB -p 128KiB"
+
 UBI_VOLNAME ?= "${MACHINE}-rootfs"
 
 multiubi_mkfs() {
-- 
2.7.4

Re: [PATCH 2/4] image_types.bbclass: fix ubi ubifs and multiubi

[Stefano Babic]

Hi Robert,

On 08/04/2016 09:15, Robert Yang wrote:
> Fixed when do_image_ubi:
> | DEBUG: Executing shell function do_image_ubifs
> | Error: min. I/O unit was not specified (use -h for help)
> | WARNING: exit code 255 from a shell command.
> 
> Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
> ---
>  meta/classes/image_types.bbclass | 3 +++
>  1 file changed, 3 insertions(+)
> 
> diff --git a/meta/classes/image_types.bbclass b/meta/classes/image_types.bbclass
> index 3f97ddd..caa4dbb 100644
> --- a/meta/classes/image_types.bbclass
> +++ b/meta/classes/image_types.bbclass
> @@ -120,6 +120,9 @@ IMAGE_CMD_elf () {
>  }
>  IMAGE_TYPEDEP_elf = "cpio.gz"
>  
> +MKUBIFS_ARGS ?= "-m 1KiB -e 128KiB -c $ROOTFS_SIZE"
> +UBINIZE_ARGS ?= "-m 1KiB -p 128KiB"
> +

These are machine specific and are generally set in the machine
configuration file. Default values you set here could be wrong in most
cases, depending on the NAND chip.

Is it not better to get the error ? Then I know that I forget to set the
two variables and I can easy fix it. If no error is raised, we get a
ubifs that never run on the machine and as rootfs is not mounted - and
finding the reason is much more difficult.

Best regards,
Stefano

-- 
=====================================================================
DENX Software Engineering GmbH,      Managing Director: Wolfgang Denk
HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany
Phone: +49-8142-66989-53 Fax: +49-8142-66989-80 Email: sbabic@denx.de
=====================================================================

Re: [PATCH 2/4] image_types.bbclass: fix ubi ubifs and multiubi

[Anders Darander]

* Stefano Babic <sbabic@denx.de> [160408 11:19]:
> On 08/04/2016 09:15, Robert Yang wrote:
> > Fixed when do_image_ubi:
> > | DEBUG: Executing shell function do_image_ubifs
> > | Error: min. I/O unit was not specified (use -h for help)
> > | WARNING: exit code 255 from a shell command.

NAK

> > +MKUBIFS_ARGS ?= "-m 1KiB -e 128KiB -c $ROOTFS_SIZE"
> > +UBINIZE_ARGS ?= "-m 1KiB -p 128KiB"
> > +

> These are machine specific and are generally set in the machine
> configuration file. Default values you set here could be wrong in most
> cases, depending on the NAND chip.

> Is it not better to get the error ? Then I know that I forget to set the
> two variables and I can easy fix it. If no error is raised, we get a
> ubifs that never run on the machine and as rootfs is not mounted - and
> finding the reason is much more difficult.

I agree with Stefano here; I'd also prefer a build failure instead of a
non-working image.

Cheers,
Anders

-- 
Anders Darander, Senior System Architect
ChargeStorm AB / eStorm AB

Re: [PATCH 2/4] image_types.bbclass: fix ubi ubifs and multiubi

[Robert Yang]

On 04/08/2016 05:54 PM, Anders Darander wrote:
> * Stefano Babic <sbabic@denx.de> [160408 11:19]:
>> On 08/04/2016 09:15, Robert Yang wrote:
>>> Fixed when do_image_ubi:
>>> | DEBUG: Executing shell function do_image_ubifs
>>> | Error: min. I/O unit was not specified (use -h for help)
>>> | WARNING: exit code 255 from a shell command.
>
> NAK
>
>>> +MKUBIFS_ARGS ?= "-m 1KiB -e 128KiB -c $ROOTFS_SIZE"
>>> +UBINIZE_ARGS ?= "-m 1KiB -p 128KiB"
>>> +
>
>> These are machine specific and are generally set in the machine
>> configuration file. Default values you set here could be wrong in most
>> cases, depending on the NAND chip.
>
>> Is it not better to get the error ? Then I know that I forget to set the
>> two variables and I can easy fix it. If no error is raised, we get a
>> ubifs that never run on the machine and as rootfs is not mounted - and
>> finding the reason is much more difficult.
>
> I agree with Stefano here; I'd also prefer a build failure instead of a
> non-working image.

My concerns was that if they can't be built out by default, then we have no
idea about whether they work or not. But as you said ubifs is a little
special, so I removed the patch from the repo. In YP 2.2, maybe we can
make it work a with a qemu machine ?

// Robert

>
> Cheers,
> Anders
>

[PATCH 3/4] image_types.bbclass: fix elf

[Robert Yang]

Fixed:
| DEBUG: Executing shell function do_image_elf
| Cannot open `/path/to/tmp/sysroots/qemux86-64/usr/src/kernel/bzImage': (null)
| WARNING: exit code 1 from a shell command.

The bzImage is in DEPLOY_DIR_IMAGE

Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
---
 meta/classes/image_types.bbclass | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/meta/classes/image_types.bbclass b/meta/classes/image_types.bbclass
index caa4dbb..0683c47 100644
--- a/meta/classes/image_types.bbclass
+++ b/meta/classes/image_types.bbclass
@@ -111,13 +111,14 @@ IMAGE_CMD_cpio () {
 	fi
 }
 
-ELF_KERNEL ?= "${STAGING_DIR_HOST}/usr/src/kernel/${KERNEL_IMAGETYPE}"
+ELF_KERNEL ?= "${DEPLOY_DIR_IMAGE}/${KERNEL_IMAGETYPE}"
 ELF_APPEND ?= "ramdisk_size=32768 root=/dev/ram0 rw console="
 
 IMAGE_CMD_elf () {
 	test -f ${DEPLOY_DIR_IMAGE}/${IMAGE_NAME}${IMAGE_NAME_SUFFIX}.elf && rm -f ${DEPLOY_DIR_IMAGE}/${IMAGE_NAME}${IMAGE_NAME_SUFFIX}.elf
 	mkelfImage --kernel=${ELF_KERNEL} --initrd=${DEPLOY_DIR_IMAGE}/${IMAGE_NAME}${IMAGE_NAME_SUFFIX}.cpio.gz --output=${DEPLOY_DIR_IMAGE}/${IMAGE_NAME}${IMAGE_NAME_SUFFIX}.elf --append='${ELF_APPEND}' ${EXTRA_IMAGECMD}
 }
+
 IMAGE_TYPEDEP_elf = "cpio.gz"
 
 MKUBIFS_ARGS ?= "-m 1KiB -e 128KiB -c $ROOTFS_SIZE"
-- 
2.7.4

[PATCH 4/4] image-vm.bbclass/image_types.bbclass: IMAGE_NAME -> IMAGE_LINK_NAME

[Robert Yang]

If we create hdddirect in the first time, and run bitbake to create
vmimg in the second time, then the previouse created
${IMAGE_LINK_NAME}.hdddirect may can not be found since it contains the
data string which are different. Use IMAGE_LINK_NAME to fix the problem.

Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
---
 meta/classes/image-vm.bbclass    | 2 +-
 meta/classes/image_types.bbclass | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/meta/classes/image-vm.bbclass b/meta/classes/image-vm.bbclass
index 2cd2640..47f7326 100644
--- a/meta/classes/image-vm.bbclass
+++ b/meta/classes/image-vm.bbclass
@@ -146,7 +146,7 @@ DISK_SIGNATURE_GENERATED := "${@generate_disk_signature()}"
 
 run_qemu_img (){
     type="$1"
-    qemu-img convert -O $type ${DEPLOY_DIR_IMAGE}/${IMAGE_NAME}.hdddirect ${DEPLOY_DIR_IMAGE}/${IMAGE_NAME}.$type
+    qemu-img convert -O $type ${DEPLOY_DIR_IMAGE}/${IMAGE_LINK_NAME}.hdddirect ${DEPLOY_DIR_IMAGE}/${IMAGE_NAME}.$type
     ln -sf ${IMAGE_NAME}.$type ${DEPLOY_DIR_IMAGE}/${IMAGE_LINK_NAME}.$type
 }
 create_vmdk_image () {
diff --git a/meta/classes/image_types.bbclass b/meta/classes/image_types.bbclass
index 0683c47..b702bc1 100644
--- a/meta/classes/image_types.bbclass
+++ b/meta/classes/image_types.bbclass
@@ -116,7 +116,7 @@ ELF_APPEND ?= "ramdisk_size=32768 root=/dev/ram0 rw console="
 
 IMAGE_CMD_elf () {
 	test -f ${DEPLOY_DIR_IMAGE}/${IMAGE_NAME}${IMAGE_NAME_SUFFIX}.elf && rm -f ${DEPLOY_DIR_IMAGE}/${IMAGE_NAME}${IMAGE_NAME_SUFFIX}.elf
-	mkelfImage --kernel=${ELF_KERNEL} --initrd=${DEPLOY_DIR_IMAGE}/${IMAGE_NAME}${IMAGE_NAME_SUFFIX}.cpio.gz --output=${DEPLOY_DIR_IMAGE}/${IMAGE_NAME}${IMAGE_NAME_SUFFIX}.elf --append='${ELF_APPEND}' ${EXTRA_IMAGECMD}
+	mkelfImage --kernel=${ELF_KERNEL} --initrd=${DEPLOY_DIR_IMAGE}/${IMAGE_LINK_NAME}.cpio.gz --output=${DEPLOY_DIR_IMAGE}/${IMAGE_NAME}${IMAGE_NAME_SUFFIX}.elf --append='${ELF_APPEND}' ${EXTRA_IMAGECMD}
 }
 
 IMAGE_TYPEDEP_elf = "cpio.gz"
-- 
2.7.4

Re: [PATCH 0/4] image_types.bbclass: fix btrfs, ubifs and elf

[Robert Yang]

Removed patch 2/4 and updated in the repo:

git://git.openembedded.org/openembedded-core-contrib rbt/img


Robert Yang (3):
   image_types.bbclass: set nodesize for btrfs
   image_types.bbclass: fix elf
   image-vm.bbclass/image_types.bbclass: IMAGE_NAME -> IMAGE_LINK_NAME

So there are only 3 patches in the repo.

// Robert

On 04/08/2016 03:15 PM, Robert Yang wrote:
> The following changes since commit 5ebea1114d54120d5ff5d8a6cd148110d0fda23f:
>
>    build-appliance-image: Update to master head revision (2016-04-06 23:54:57 +0100)
>
> are available in the git repository at:
>
>    git://git.openembedded.org/openembedded-core-contrib rbt/img
>    http://cgit.openembedded.org/cgit.cgi/openembedded-core-contrib/log/?h=rbt/img
>
> Robert Yang (4):
>    image_types.bbclass: set nodesize for btrfs
>    image_types.bbclass: fix ubi ubifs and multiubi
>    image_types.bbclass: fix elf
>    image-vm.bbclass/image_types.bbclass: IMAGE_NAME -> IMAGE_LINK_NAME
>
>   meta/classes/image-vm.bbclass    |  2 +-
>   meta/classes/image_types.bbclass | 10 +++++++---
>   2 files changed, 8 insertions(+), 4 deletions(-)
>