[meta][PATCH 1/2] Fix some issues in the fitImage generation

Openembedded Core Discussions
 help / color / mirror / Atom feed

* [meta][PATCH 1/2] Fix some issues in the fitImage generation
@ 2017-10-25  7:32 Thomas Perrot
  2017-10-25  7:32 ` [meta][PATCH 2/2] Avoid a circular dependency between do_concat_dtb and do_assemble_fitimage Thomas Perrot
                   ` (2 more replies)
  0 siblings, 3 replies; 7+ messages in thread
From: Thomas Perrot @ 2017-10-25  7:32 UTC (permalink / raw)
  To: openembedded-core

- Bad image type replacement for aarch64
- Fix DTB path in fitimage.its
- Ignore fitImage type in do_bundle_initramfs task because the packaging is made by do_assemble_fitimage_initramfs
- the images for arm64 are not compressed the linux_comp variable must therefore be none

Signed-off-by: Thomas Perrot <thomas.perrot@tupi.fr>
---
 meta/classes/kernel-fitimage.bbclass |  7 +++++--
 meta/classes/kernel-uboot.bbclass    | 11 ++++++++---
 meta/classes/kernel.bbclass          |  4 +++-
 3 files changed, 16 insertions(+), 6 deletions(-)

diff --git a/meta/classes/kernel-fitimage.bbclass b/meta/classes/kernel-fitimage.bbclass
index 179185b6b2..0789ba611f 100644
--- a/meta/classes/kernel-fitimage.bbclass
+++ b/meta/classes/kernel-fitimage.bbclass
@@ -7,10 +7,13 @@ python __anonymous () {
         depends = "%s u-boot-mkimage-native dtc-native" % depends
         d.setVar("DEPENDS", depends)
 
-        if d.getVar("UBOOT_ARCH") == "mips":
+        uarch = d.getVar("UBOOT_ARCH")
+        if uarch == "mips":
             replacementtype = "vmlinuz.bin"
-        elif d.getVar("UBOOT_ARCH") == "x86":
+        elif uarch == "x86":
             replacementtype = "bzImage"
+        elif uarch == "arm64":
+            replacementtype = "Image"
         else:
             replacementtype = "zImage"
 
diff --git a/meta/classes/kernel-uboot.bbclass b/meta/classes/kernel-uboot.bbclass
index 87f02654fa..06d293039a 100644
--- a/meta/classes/kernel-uboot.bbclass
+++ b/meta/classes/kernel-uboot.bbclass
@@ -1,5 +1,5 @@
 uboot_prep_kimage() {
-	if [ -e arch/${ARCH}/boot/compressed/vmlinux ]; then
+	if [ -e arch/${ARCH}/boot/compressed/vmlinux ] ; then
 		vmlinux_path="arch/${ARCH}/boot/compressed/vmlinux"
 		linux_suffix=""
 		linux_comp="none"
@@ -11,8 +11,13 @@ uboot_prep_kimage() {
 		linux_comp="none"
 	else
 		vmlinux_path="vmlinux"
-		linux_suffix=".gz"
-		linux_comp="gzip"
+		if [ "${ARCH}" = "arm64" ] ; then
+			linux_suffix=""
+			linux_comp="none"
+		else
+			linux_suffix=".gz"
+			linux_comp="gzip"
+		fi
 	fi
 
 	[ -n "${vmlinux_path}" ] && ${OBJCOPY} -O binary -R .note -R .comment -S "${vmlinux_path}" linux.bin
diff --git a/meta/classes/kernel.bbclass b/meta/classes/kernel.bbclass
index 756707a3c2..7f8c8985f3 100644
--- a/meta/classes/kernel.bbclass
+++ b/meta/classes/kernel.bbclass
@@ -208,7 +208,9 @@ do_bundle_initramfs () {
 		# Backing up kernel image relies on its type(regular file or symbolic link)
 		tmp_path=""
 		for type in ${KERNEL_IMAGETYPES} ; do
-			if [ -h ${KERNEL_OUTPUT_DIR}/$type ] ; then
+			if [ "$type" = "fitImage" ] ; then
+				continue
+			elif [ -h ${KERNEL_OUTPUT_DIR}/$type ] ; then
 				linkpath=`readlink -n ${KERNEL_OUTPUT_DIR}/$type`
 				realpath=`readlink -fn ${KERNEL_OUTPUT_DIR}/$type`
 				mv -f $realpath $realpath.bak
-- 
2.13.6



^ permalink raw reply related	[flat|nested] 7+ messages in thread

* [meta][PATCH 2/2] Avoid a circular dependency between do_concat_dtb and do_assemble_fitimage
  2017-10-25  7:32 [meta][PATCH 1/2] Fix some issues in the fitImage generation Thomas Perrot
@ 2017-10-25  7:32 ` Thomas Perrot
  2017-10-25  8:05 ` ✗ patchtest: failure for "[meta] Fix some issues in the ..." and 1 more Patchwork
  2017-10-25 10:13 ` [meta][PATCH 1/2] Fix some issues in the fitImage generation André Draszik
  2 siblings, 0 replies; 7+ messages in thread
From: Thomas Perrot @ 2017-10-25  7:32 UTC (permalink / raw)
  To: openembedded-core

When UBOOT_SIGN_ENABLE is active:

Dependency loop #1 found:
  do_concat_dtb (dependent Tasks ['linux-yocto_4.10.bb:do_assemble_fitimage'])
  do_install (dependent Tasks ['pseudo_1.8.2.bb:do_populate_sysroot', 'u-boot_2017.01.bb:do_compile', 'u-boot_2017.01.bb:do_concat_dtb'])
  do_deploy (dependent Tasks ['u-boot_2017.01.bb:do_deploy_dtb', 'u-boot_2017.01.bb:do_install'])
  do_assemble_fitimage (dependent Tasks ['linux-yocto_4.10.bb:do_compile', 'u-boot_2017.01.bb:do_deploy'])

Signed-off-by: Thomas Perrot <thomas.perrot@tupi.fr>
---
 meta/classes/kernel-fitimage.bbclass |  2 +-
 meta/classes/uboot-sign.bbclass      | 21 +++++++++------------
 2 files changed, 10 insertions(+), 13 deletions(-)

diff --git a/meta/classes/kernel-fitimage.bbclass b/meta/classes/kernel-fitimage.bbclass
index 0789ba611f..13d61a2298 100644
--- a/meta/classes/kernel-fitimage.bbclass
+++ b/meta/classes/kernel-fitimage.bbclass
@@ -33,7 +33,7 @@ python __anonymous () {
         # the fitImage:
         if d.getVar('UBOOT_SIGN_ENABLE') == "1":
             uboot_pn = d.getVar('PREFERRED_PROVIDER_u-boot') or 'u-boot'
-            d.appendVarFlag('do_assemble_fitimage', 'depends', ' %s:do_deploy' % uboot_pn)
+            d.appendVarFlag('do_assemble_fitimage', 'depends', ' %s:do_deploy_dtb' % uboot_pn)
 }
 
 # Options for the device tree compiler passed to mkimage '-D' feature:
diff --git a/meta/classes/uboot-sign.bbclass b/meta/classes/uboot-sign.bbclass
index 8ee904e7df..2f616bf2a9 100644
--- a/meta/classes/uboot-sign.bbclass
+++ b/meta/classes/uboot-sign.bbclass
@@ -47,16 +47,14 @@ do_deploy_dtb () {
 	cd ${DEPLOYDIR}
 
 	if [ -f ${B}/${UBOOT_DTB_BINARY} ]; then
-		install ${B}/${UBOOT_DTB_BINARY} ${DEPLOYDIR}/${UBOOT_DTB_IMAGE}
-		rm -f ${UBOOT_DTB_BINARY} ${UBOOT_DTB_SYMLINK}
-		ln -sf ${UBOOT_DTB_IMAGE} ${UBOOT_DTB_SYMLINK}
-		ln -sf ${UBOOT_DTB_IMAGE} ${UBOOT_DTB_BINARY}
+		install ${B}/${UBOOT_DTB_BINARY} ${DEPLOY_DIR_IMAGE}/${UBOOT_DTB_IMAGE}
+		ln -sf ${DEPLOY_DIR_IMAGE}/${UBOOT_DTB_IMAGE} ${DEPLOY_DIR_IMAGE}/${UBOOT_DTB_SYMLINK}
+		ln -sf ${DEPLOY_DIR_IMAGE}/${UBOOT_DTB_IMAGE} ${DEPLOY_DIR_IMAGE}/${UBOOT_DTB_BINARY}
 	fi
 	if [ -f ${B}/${UBOOT_NODTB_BINARY} ]; then
-		install ${B}/${UBOOT_NODTB_BINARY} ${DEPLOYDIR}/${UBOOT_NODTB_IMAGE}
-		rm -f ${UBOOT_NODTB_BINARY} ${UBOOT_NODTB_SYMLINK}
-		ln -sf ${UBOOT_NODTB_IMAGE} ${UBOOT_NODTB_SYMLINK}
-		ln -sf ${UBOOT_NODTB_IMAGE} ${UBOOT_NODTB_BINARY}
+		install ${B}/${UBOOT_NODTB_BINARY} ${DEPLOY_DIR_IMAGE}/${UBOOT_NODTB_IMAGE}
+		ln -sf ${DEPLOY_DIR_IMAGE}/${UBOOT_NODTB_IMAGE} ${DEPLOY_DIR_IMAGE}/${UBOOT_NODTB_SYMLINK}
+		ln -sf ${DEPLOY_DIR_IMAGE}/${UBOOT_NODTB_IMAGE} ${DEPLOY_DIR_IMAGE}/${UBOOT_NODTB_BINARY}
 	fi
 }
 
@@ -70,11 +68,10 @@ do_concat_dtb () {
 			oe_runmake EXT_DTB=${DEPLOYDIR}/${UBOOT_DTB_IMAGE}
 			install ${B}/${UBOOT_BINARY} ${DEPLOYDIR}/${UBOOT_IMAGE}
 			install ${B}/${UBOOT_BINARY} ${DEPLOY_DIR_IMAGE}/${UBOOT_IMAGE}
-		elif [ -e "${DEPLOYDIR}/${UBOOT_NODTB_IMAGE}" -a -e "${DEPLOYDIR}/${UBOOT_DTB_IMAGE}" ]; then
-			cd ${DEPLOYDIR}
-			cat ${UBOOT_NODTB_IMAGE} ${UBOOT_DTB_IMAGE} | tee ${B}/${UBOOT_BINARY} > ${UBOOT_IMAGE}
+		elif [ -e "${DEPLOY_DIR_IMAGE}/${UBOOT_NODTB_IMAGE}" -a -e "${DEPLOY_DIR_IMAGE}/${UBOOT_DTB_IMAGE}" ]; then
+			cat ${DEPLOY_DIR_IMAGE}/${UBOOT_NODTB_IMAGE} ${DEPLOY_DIR_IMAGE}/${UBOOT_DTB_IMAGE} | tee ${B}/${UBOOT_BINARY} > ${UBOOT_IMAGE}
 		else
-			bbwarn "Failure while adding public key to u-boot binary. Verified boot won't be available."
+			bbfatal "Failure while adding public key to u-boot binary. Verified boot won't be available."
 		fi
 	fi
 }
-- 
2.13.6



^ permalink raw reply related	[flat|nested] 7+ messages in thread

* ✗ patchtest: failure for "[meta] Fix some issues in the ..." and 1 more
  2017-10-25  7:32 [meta][PATCH 1/2] Fix some issues in the fitImage generation Thomas Perrot
  2017-10-25  7:32 ` [meta][PATCH 2/2] Avoid a circular dependency between do_concat_dtb and do_assemble_fitimage Thomas Perrot
@ 2017-10-25  8:05 ` Patchwork
  2017-10-25 10:13 ` [meta][PATCH 1/2] Fix some issues in the fitImage generation André Draszik
  2 siblings, 0 replies; 7+ messages in thread
From: Patchwork @ 2017-10-25  8:05 UTC (permalink / raw)
  To: Thomas PERROT; +Cc: openembedded-core

== Series Details ==

Series: "[meta] Fix some issues in the ..." and 1 more
Revision: 1
URL   : https://patchwork.openembedded.org/series/9470/
State : failure

== Summary ==

Thank you for submitting this patch series to OpenEmbedded Core. This is
an automated response. Several tests have been executed on the proposed
series by patchtest resulting in the following failures:

* Patch            [meta,1/2] Fix some issues in the fitImage generation
 Issue             Shortlog does not follow expected format [test_shortlog_format] 
  Suggested fix    Commit shortlog (first line of commit message) should follow the format "<target>: <summary>"

If you believe any of these test results are incorrect, please reply to the
mailing list (openembedded-core@lists.openembedded.org) raising your concerns.
Otherwise we would appreciate you correcting the issues and submitting a new
version of the patchset if applicable. Please ensure you add/increment the
version number when sending the new version (i.e. [PATCH] -> [PATCH v2] ->
[PATCH v3] -> ...).

---
Guidelines:     https://www.openembedded.org/wiki/Commit_Patch_Message_Guidelines
Test framework: http://git.yoctoproject.org/cgit/cgit.cgi/patchtest
Test suite:     http://git.yoctoproject.org/cgit/cgit.cgi/patchtest-oe

^ permalink raw reply	[flat|nested] 7+ messages in thread

* Re: [meta][PATCH 1/2] Fix some issues in the fitImage generation
  2017-10-25  7:32 [meta][PATCH 1/2] Fix some issues in the fitImage generation Thomas Perrot
  2017-10-25  7:32 ` [meta][PATCH 2/2] Avoid a circular dependency between do_concat_dtb and do_assemble_fitimage Thomas Perrot
  2017-10-25  8:05 ` ✗ patchtest: failure for "[meta] Fix some issues in the ..." and 1 more Patchwork
@ 2017-10-25 10:13 ` André Draszik
  2017-10-25 11:06   ` André Draszik
  2 siblings, 1 reply; 7+ messages in thread
From: André Draszik @ 2017-10-25 10:13 UTC (permalink / raw)
  To: openembedded-core

On Wed, 2017-10-25 at 09:32 +0200, Thomas Perrot wrote:
> - Bad image type replacement for aarch64
> - Fix DTB path in fitimage.its

It doesn't look like the DTB bit is still relevant in this commit.

> - Ignore fitImage type in do_bundle_initramfs task because the packaging
> is made by do_assemble_fitimage_initramfs
> - the images for arm64 are not compressed the linux_comp variable must
> therefore be none

I think it would also make sense to split this into multiple commits, as
your aarch64 fixes and fitimage fixes aren't directly related.

> Signed-off-by: Thomas Perrot <thomas.perrot@tupi.fr>
> ---
>  meta/classes/kernel-fitimage.bbclass |  7 +++++--
>  meta/classes/kernel-uboot.bbclass    | 11 ++++++++---
>  meta/classes/kernel.bbclass          |  4 +++-
>  3 files changed, 16 insertions(+), 6 deletions(-)
> 
> diff --git a/meta/classes/kernel-fitimage.bbclass b/meta/classes/kernel-
> fitimage.bbclass
> index 179185b6b2..0789ba611f 100644
> --- a/meta/classes/kernel-fitimage.bbclass
> +++ b/meta/classes/kernel-fitimage.bbclass
> @@ -7,10 +7,13 @@ python __anonymous () {
>          depends = "%s u-boot-mkimage-native dtc-native" % depends
>          d.setVar("DEPENDS", depends)
>  
> -        if d.getVar("UBOOT_ARCH") == "mips":
> +        uarch = d.getVar("UBOOT_ARCH")
> +        if uarch == "mips":
>              replacementtype = "vmlinuz.bin"
> -        elif d.getVar("UBOOT_ARCH") == "x86":
> +        elif uarch == "x86":
>              replacementtype = "bzImage"
> +        elif uarch == "arm64":
> +            replacementtype = "Image"

Can you keep these tests in alphabetical order please?

>          else:
>              replacementtype = "zImage"
>  
> diff --git a/meta/classes/kernel-uboot.bbclass b/meta/classes/kernel-
> uboot.bbclass
> index 87f02654fa..06d293039a 100644
> --- a/meta/classes/kernel-uboot.bbclass
> +++ b/meta/classes/kernel-uboot.bbclass
> @@ -1,5 +1,5 @@
>  uboot_prep_kimage() {
> -	if [ -e arch/${ARCH}/boot/compressed/vmlinux ]; then
> +	if [ -e arch/${ARCH}/boot/compressed/vmlinux ] ; then
>  		vmlinux_path="arch/${ARCH}/boot/compressed/vmlinux"
>  		linux_suffix=""
>  		linux_comp="none"
> @@ -11,8 +11,13 @@ uboot_prep_kimage() {
>  		linux_comp="none"
>  	else
>  		vmlinux_path="vmlinux"
> -		linux_suffix=".gz"
> -		linux_comp="gzip"
> +		if [ "${ARCH}" = "arm64" ] ; then
> +			linux_suffix=""
> +			linux_comp="none"
> +		else
> +			linux_suffix=".gz"
> +			linux_comp="gzip"
> +		fi
>  	fi

I believe U-Boot supports decompression on any architecture independent of
kernel decompression support. It also supports multiple decompressors, not
just gzip. It would make sense to rather be able to configure the
compression method for U-Boot in a more generic way (including turning it
off completely) for any supported target.


Cheers,
Andre'



^ permalink raw reply	[flat|nested] 7+ messages in thread

* Re: [meta][PATCH 1/2] Fix some issues in the fitImage generation
  2017-10-25 10:13 ` [meta][PATCH 1/2] Fix some issues in the fitImage generation André Draszik
@ 2017-10-25 11:06   ` André Draszik
  2017-10-26  4:18     ` Thomas PERROT
  0 siblings, 1 reply; 7+ messages in thread
From: André Draszik @ 2017-10-25 11:06 UTC (permalink / raw)
  To: openembedded-core

On Wed, 2017-10-25 at 11:13 +0100, André Draszik wrote:
> On Wed, 2017-10-25 at 09:32 +0200, Thomas Perrot wrote:
> > - Bad image type replacement for aarch64
> > - Fix DTB path in fitimage.its
> 
> It doesn't look like the DTB bit is still relevant in this commit.
> 
> > - Ignore fitImage type in do_bundle_initramfs task because the packaging
> > is made by do_assemble_fitimage_initramfs
> > - the images for arm64 are not compressed the linux_comp variable must
> > therefore be none
> 
> I think it would also make sense to split this into multiple commits, as
> your aarch64 fixes and fitimage fixes aren't directly related.

To be more clear, your fix to kernel.bbclass is independent of everything
else and doesn't apply to aarch64 only, so it should be a separate commit.


Cheers,
Andre'



^ permalink raw reply	[flat|nested] 7+ messages in thread

* Re: [meta][PATCH 1/2] Fix some issues in the fitImage generation
  2017-10-25 11:06   ` André Draszik
@ 2017-10-26  4:18     ` Thomas PERROT
       [not found]       ` <BL2PR02MB21633D3A250997F82D3496E0B9570@BL2PR02MB2163.namprd02.prod.outlook.com>
  0 siblings, 1 reply; 7+ messages in thread
From: Thomas PERROT @ 2017-10-26  4:18 UTC (permalink / raw)
  To: André Draszik; +Cc: OE-core

[-- Attachment #1: Type: text/plain, Size: 1263 bytes --]

Hi,

I agree, so I cleaned up, split into several commits, then superseded this
one.

Best regards,
Thomas P.

On Wed, Oct 25, 2017 at 1:06 PM, André Draszik <git@andred.net> wrote:

> On Wed, 2017-10-25 at 11:13 +0100, André Draszik wrote:
> > On Wed, 2017-10-25 at 09:32 +0200, Thomas Perrot wrote:
> > > - Bad image type replacement for aarch64
> > > - Fix DTB path in fitimage.its
> >
> > It doesn't look like the DTB bit is still relevant in this commit.
> >
> > > - Ignore fitImage type in do_bundle_initramfs task because the
> packaging
> > > is made by do_assemble_fitimage_initramfs
> > > - the images for arm64 are not compressed the linux_comp variable must
> > > therefore be none
> >
> > I think it would also make sense to split this into multiple commits, as
> > your aarch64 fixes and fitimage fixes aren't directly related.
>
> To be more clear, your fix to kernel.bbclass is independent of everything
> else and doesn't apply to aarch64 only, so it should be a separate commit.
>
>
> Cheers,
> Andre'
>
> --
> _______________________________________________
> Openembedded-core mailing list
> Openembedded-core@lists.openembedded.org
> http://lists.openembedded.org/mailman/listinfo/openembedded-core
>

[-- Attachment #2: Type: text/html, Size: 2278 bytes --]

^ permalink raw reply	[flat|nested] 7+ messages in thread

[parent not found: <BL2PR02MB21633D3A250997F82D3496E0B9570@BL2PR02MB2163.namprd02.prod.outlook.com>]

* Re: [meta][PATCH 1/2] Fix some issues in the fitImage generation
       [not found]       ` <BL2PR02MB21633D3A250997F82D3496E0B9570@BL2PR02MB2163.namprd02.prod.outlook.com>
@ 2017-11-10 10:30         ` André Draszik
  0 siblings, 0 replies; 7+ messages in thread
From: André Draszik @ 2017-11-10 10:30 UTC (permalink / raw)
  To: Manjukumar Harthikote Matha; +Cc: OE-core

Hi Manju,

This one isn't the latest version, but even the latest versions had issues
reported by patchtest. So I guess no.

Cheers,
Andr'

On Thu, 2017-11-09 at 17:28 +0000, Manjukumar Harthikote Matha wrote:
> Hi All,
> 
> Is this accepted in master? We need similar patch in Rocko and I have an
> additional do_deploy failure fix patch on top of this patch.
> 
> Thanks,
> Manju
> 
> From: openembedded-core-bounces@lists.openembedded.org [mailto:openembedde
> d-core-bounces@lists.openembedded.org] On Behalf Of Thomas PERROT
> Sent: Wednesday, October 25, 2017 9:18 PM
> To: André Draszik <git@andred.net>
> Cc: OE-core <openembedded-core@lists.openembedded.org>
> Subject: Re: [OE-core] [meta][PATCH 1/2] Fix some issues in the fitImage
> generation
> 
> Hi,
> 
> I agree, so I cleaned up, split into several commits, then superseded this
> one.
> Best regards,
> Thomas P.
> 
> On Wed, Oct 25, 2017 at 1:06 PM, André Draszik <git@andred.net<mailto:git@
> andred.net>> wrote:
> On Wed, 2017-10-25 at 11:13 +0100, André Draszik wrote:
> > On Wed, 2017-10-25 at 09:32 +0200, Thomas Perrot wrote:
> > > - Bad image type replacement for aarch64
> > > - Fix DTB path in fitimage.its
> > 
> > It doesn't look like the DTB bit is still relevant in this commit.
> > 
> > > - Ignore fitImage type in do_bundle_initramfs task because the
> > > packaging
> > > is made by do_assemble_fitimage_initramfs
> > > - the images for arm64 are not compressed the linux_comp variable must
> > > therefore be none
> > 
> > I think it would also make sense to split this into multiple commits, as
> > your aarch64 fixes and fitimage fixes aren't directly related.
> 
> To be more clear, your fix to kernel.bbclass is independent of everything
> else and doesn't apply to aarch64 only, so it should be a separate commit.
> 
> 
> Cheers,
> Andre'
> 
> --
> _______________________________________________
> Openembedded-core mailing list
> Openembedded-core@lists.openembedded.org<mailto:Openembedded-core@lists.op
> enembedded.org>
> http://lists.openembedded.org/mailman/listinfo/openembedded-core
> 


^ permalink raw reply	[flat|nested] 7+ messages in thread

end of thread, other threads:[~2017-11-10 10:30 UTC | newest]

Thread overview: 7+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2017-10-25  7:32 [meta][PATCH 1/2] Fix some issues in the fitImage generation Thomas Perrot
2017-10-25  7:32 ` [meta][PATCH 2/2] Avoid a circular dependency between do_concat_dtb and do_assemble_fitimage Thomas Perrot
2017-10-25  8:05 ` ✗ patchtest: failure for "[meta] Fix some issues in the ..." and 1 more Patchwork
2017-10-25 10:13 ` [meta][PATCH 1/2] Fix some issues in the fitImage generation André Draszik
2017-10-25 11:06   ` André Draszik
2017-10-26  4:18     ` Thomas PERROT
     [not found]       ` <BL2PR02MB21633D3A250997F82D3496E0B9570@BL2PR02MB2163.namprd02.prod.outlook.com>
2017-11-10 10:30         ` André Draszik

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