1177 replies to this topic

[jpuigs]

O.M.G.

 

I heard from a good source that it would available on April 1st.   

[dax]

I tried to fix solo2 problem. But first try was not successful

Don't give up

[betacentauri]

No, but I have not much time and a complete rebuild takes an hour or 1,5 hours(don't know exactly). And also master branch has be tested...

[betacentauri]

This seems to fix the solo2 build problem. Can somebody confirm? Please build from scratch!
 

diff --git a/conf/machine/include/vuxxo2.inc b/conf/machine/include/vuxxo2.inc
index e00b3dc..0b538fe 100644
--- a/conf/machine/include/vuxxo2.inc
+++ b/conf/machine/include/vuxxo2.inc
@@ -54,7 +54,7 @@ IMAGE_CMD_ubi_prepend = " \
 IMAGE_CMD_ubi_append = " \
        mkdir -p ${IMGDEPLOYDIR}/vuplus/${UPDATE_DIR}; \
        cp -f ${IMGDEPLOYDIR}/${IMAGE_NAME}.rootfs.ubi ${IMGDEPLOYDIR}/vuplus/${UPDATE_DIR}/root_cfe_auto.bin; \
-       cp -f ${DEPLOY_DIR_IMAGE}/${IMAGE_NAME}.vmlinux.gz ${IMGDEPLOYDIR}/vuplus/${UPDATE_DIR}/kernel_cfe_auto.bin; \
+       cp -f ${IMGDEPLOYDIR}/${IMAGE_NAME}.vmlinux.gz ${IMGDEPLOYDIR}/vuplus/${UPDATE_DIR}/kernel_cfe_auto.bin; \
        cp -f ${IMGDEPLOYDIR}/${IMAGE_NAME}.initrd_cfe_auto.bin ${IMGDEPLOYDIR}/vuplus/${UPDATE_DIR}/initrd_cfe_auto.bin; \
        touch ${IMGDEPLOYDIR}/vuplus/${UPDATE_DIR}/${FORCE_REBOOT_OPTION}.update; \
        cd ${IMGDEPLOYDIR}; \

Edited by betacentauri, 14 December 2016 - 23:41.

[Beeker]

This seems to fix the solo2 build problem. Can somebody confirm? Please build from scratch!
 

diff --git a/conf/machine/include/vuxxo2.inc b/conf/machine/include/vuxxo2.inc
index e00b3dc..0b538fe 100644
--- a/conf/machine/include/vuxxo2.inc
+++ b/conf/machine/include/vuxxo2.inc
@@ -54,7 +54,7 @@ IMAGE_CMD_ubi_prepend = " \
 IMAGE_CMD_ubi_append = " \
        mkdir -p ${IMGDEPLOYDIR}/vuplus/${UPDATE_DIR}; \
        cp -f ${IMGDEPLOYDIR}/${IMAGE_NAME}.rootfs.ubi ${IMGDEPLOYDIR}/vuplus/${UPDATE_DIR}/root_cfe_auto.bin; \
-       cp -f ${DEPLOY_DIR_IMAGE}/${IMAGE_NAME}.vmlinux.gz ${IMGDEPLOYDIR}/vuplus/${UPDATE_DIR}/kernel_cfe_auto.bin; \
+       cp -f ${IMGDEPLOYDIR}/${IMAGE_NAME}.vmlinux.gz ${IMGDEPLOYDIR}/vuplus/${UPDATE_DIR}/kernel_cfe_auto.bin; \
        cp -f ${IMGDEPLOYDIR}/${IMAGE_NAME}.initrd_cfe_auto.bin ${IMGDEPLOYDIR}/vuplus/${UPDATE_DIR}/initrd_cfe_auto.bin; \
        touch ${IMGDEPLOYDIR}/vuplus/${UPDATE_DIR}/${FORCE_REBOOT_OPTION}.update; \
        cd ${IMGDEPLOYDIR}; \

 

Build running now. Answer about a hour. Only master-next

 

If it don't succeed, i try patch of @Mirakels.

Edited by Beeker, 15 December 2016 - 09:32.

[Beeker]

Build from scratch ok solo2.

 

For clarify, how i done it.

-Fresh git clone -b master-next https://github.com/O...pli-oe-core.git

-Start build, MACHINE="vusolo2" make image.

-Stop the build when bsp layer vuduo was present.

-Adapt the file according to patch.

-Delete entire 'build' directory.

-Start again.

 

More boxes ?

Edited by Beeker, 15 December 2016 - 10:49.

[Beeker]

https://github.com/v...a8b3a5c76e9585f

 

Line 57 vuxxo2.inc  

[betacentauri]

You can try to build duo2, too.

I have tested master (only solo2).

[Beeker]

duo2 also ok.

 

Not build from scratch, but i don't think that's really necesarry(due to no errors).

Edited by Beeker, 15 December 2016 - 11:18.

[betacentauri]

Ok, thanks!

Everybody feel free to create pull request. I most likely have no time today. I will do it on Friday if nobody made it until then.

[Beeker]

PR done.

 

Solo4K ok.

 

But mut@nt hd51 fails. I updated the build-environment a few hours ago. So i have the latest bsp layer of december 14th.

NOTE: Executing SetScene Tasks
NOTE: Executing RunQueue Tasks
ERROR: openpli-enigma2-image-1.0-r0 do_image_hd_emmc: Function failed: do_image_hd_emmc (log file is located at /home/hains/test/openpli-oe-core/build/tmp/work/hd51-oe-linux-gnueabi/openpli-enigma2-image/1.0-r0/temp/log.do_image_hd_emmc.17965)
ERROR: Logfile of failure stored in: /home/hains/test/openpli-oe-core/build/tmp/work/hd51-oe-linux-gnueabi/openpli-enigma2-image/1.0-r0/temp/log.do_image_hd_emmc.17965
Log data follows:
| DEBUG: Executing python function set_image_size
| DEBUG: Python function set_image_size finished
| DEBUG: Executing shell function do_image_hd_emmc
| 0+0 records in
| 0+0 records out
| 0 bytes copied, 7.8183e-05 s, 0.0 kB/s
| 3072+0 records in
| 3072+0 records out
| 3145728 bytes (3.1 MB, 3.0 MiB) copied, 0.00439367 s, 716 MB/s
| mkfs.fat 4.0 (2016-05-06)
| Total number of sectors (6144) not a multiple of sectors per track (63)!
| Add mtools_skip_check=1 to your .mtoolsrc file to skip this test
| Copying STARTUP
| Total number of sectors (6144) not a multiple of sectors per track (63)!
| Add mtools_skip_check=1 to your .mtoolsrc file to skip this test
| Copying STARTUP_1
| Total number of sectors (6144) not a multiple of sectors per track (63)!
| Add mtools_skip_check=1 to your .mtoolsrc file to skip this test
| Copying STARTUP_2
| Total number of sectors (6144) not a multiple of sectors per track (63)!
| Add mtools_skip_check=1 to your .mtoolsrc file to skip this test
| Copying STARTUP_3
| Total number of sectors (6144) not a multiple of sectors per track (63)!
| Add mtools_skip_check=1 to your .mtoolsrc file to skip this test
| Copying STARTUP_4
| 3072+0 records in
| 3072+0 records out
| 3145728 bytes (3.1 MB, 3.0 MiB) copied, 0.00306105 s, 1.0 GB/s
| 4771+1 records in
| 4771+1 records out
| 4886136 bytes (4.9 MB, 4.7 MiB) copied, 0.00547864 s, 892 MB/s
| resize2fs 1.43 (17-May-2016)
| open: No such file or directory while opening /home/hains/test/openpli-oe-core/build/tmp/deploy/images/hd51/openpli-enigma2-5-hd51.rootfs.ext4
| WARNING: exit code 1 from a shell command.
| ERROR: Function failed: do_image_hd_emmc (log file is located at /home/hains/test/openpli-oe-core/build/tmp/work/hd51-oe-linux-gnueabi/openpli-enigma2-image/1.0-r0/temp/log.do_image_hd_emmc.17965)
ERROR: Task (/home/hains/test/openpli-oe-core/meta-openpli/recipes-openpli/images/openpli-enigma2-image.bb:do_image_hd_emmc) failed with exit code '1'
NOTE: Tasks Summary: Attempted 4643 tasks of which 4642 didn't need to be rerun and 1 failed.

Summary: 1 task failed:
  /home/hains/test/openpli-oe-core/meta-openpli/recipes-openpli/images/openpli-enigma2-image.bb:do_image_hd_emmc
Summary: There was 1 ERROR message shown, returning a non-zero exit code.
Makefile:89: recipe for target 'image' failed
make: *** [image] Error 1

Edited by Beeker, 15 December 2016 - 13:10.

[Beeker]

Image hd51 only succusfull image if i copy 2 files manually

openpli-oe-core/build/tmp/work/hd51-oe-linux-gnueabi/openpli-enigma2-image/1.0-r0/deploy-openpli-enigma2-image-image-complete/openpli-enigma2-5-hd51.rootfs.ext4

to:

openpli-oe-core/build/tmp/deploy/images/hd51/openpli-enigma2-5-hd51.rootfs.ext4

and

openpli-oe-core/build/tmp/deploy/images/hd51/openpli-enigma2-5-hd51.emmc.img

to:

openpli-oe-core/build/tmp/work/hd51-oe-linux-gnueabi/openpli-enigma2-image/1.0-r0/deploy-openpli-enigma2-image-image-complete/openpli-enigma2-5-hd51.emmc.img


I try this e.g.(any many other options) but i never succeed. I don't no why

cp ${IMGDEPLOYDIR}/${IMAGE_NAME}.rootfs.ext4 ${DEPLOY_DIR_IMAGE}/${IMAGE_NAME}.rootfs.ext4 \

I deleted

IMGDEPLOYDIR ?= "${DEPLOY_DIR_IMAGE}"

in meta-hd/conf/machine/include/hd.inc

[Erik Slagter]

I have noticed VU+ that some PR's are coming because of this issue. They may not notice it immediately though.

 

Thx Beeker.

[Beeker]

Thanks..Merged BTW.

 

Problem hd51 solved.

 

meta-hd/classes/hd-emmc_image_multi.bbclass

@@ -33,7 +33,7 @@ FOURTH_KERNEL_PARTITION_OFFSET = "$(expr ${THRID_ROOTFS_PARTITION_OFFSET} \+ ${R
 
 FOURTH_ROOTFS_PARTITION_OFFSET = "$(expr ${FOURTH_KERNEL_PARTITION_OFFSET} \+ ${KERNEL_PARTITION_SIZE})"
 
-EMMC_IMAGE = "${DEPLOY_DIR_IMAGE}/${IMAGE_NAME}.emmc.img"
+EMMC_IMAGE = "${IMGDEPLOYDIR}/${IMAGE_NAME}.emmc.img"
 EMMC_IMAGE_SIZE = "3817472"
 
 IMAGE_CMD_hd-emmc () {
@@ -62,7 +62,7 @@ IMAGE_CMD_hd-emmc () {
     mcopy -i ${WORKDIR}/boot.img -v ${WORKDIR}/STARTUP_4 ::
     dd conv=notrunc if=${WORKDIR}/boot.img of=${EMMC_IMAGE} bs=1024 seek=${BOOT_PARTITION_OFFSET}
     dd conv=notrunc if=${DEPLOY_DIR_IMAGE}/zImage of=${EMMC_IMAGE} bs=1024 seek=${KERNEL_PARTITION_OFFSET}
-    resize2fs ${DEPLOY_DIR_IMAGE}/${IMAGE_NAME}.rootfs.ext4 ${ROOTFS_PARTITION_SIZE}k
+    resize2fs ${IMGDEPLOYDIR}/${IMAGE_NAME}.rootfs.ext4 ${ROOTFS_PARTITION_SIZE}k
     # Truncate on purpose
-    dd if=${DEPLOY_DIR_IMAGE}/${IMAGE_NAME}.rootfs.ext4 of=${EMMC_IMAGE} bs=1024 seek=${ROOTFS_PARTITION_OFFSET} count=${IMAGE_ROOTFS_SIZE}
-}
\ No newline at end of file
+    dd if=${IMGDEPLOYDIR}/${IMAGE_NAME}.rootfs.ext4 of=${EMMC_IMAGE} bs=1024 seek=${ROOTFS_PARTITION_OFFSET} count=${IMAGE_ROOTFS_SIZE}
+}

And meta-hd/class/hd-emmc_image_single.bbclass

@@ -23,7 +23,7 @@ ROOTFS_PARTITION_SIZE = "1048576"
 
 STORAGE_PARTITION_OFFSET = "$(expr ${ROOTFS_PARTITION_OFFSET} \+ ${ROOTFS_PARTITION_SIZE})"
 
-EMMC_IMAGE = "${DEPLOY_DIR_IMAGE}/${IMAGE_NAME}.emmc.img"
+EMMC_IMAGE = "${IMGDEPLOYDIR}/${IMAGE_NAME}.emmc.img"
 EMMC_IMAGE_SIZE = "3817472"
 
 IMAGE_CMD_hd-emmc () {
@@ -41,9 +41,9 @@ IMAGE_CMD_hd-emmc () {
     mcopy -i ${WORKDIR}/boot.img -v ${WORKDIR}/STARTUP_1 ::
     dd conv=notrunc if=${WORKDIR}/boot.img of=${EMMC_IMAGE} bs=1024 seek=${BOOT_PARTITION_OFFSET}
     dd conv=notrunc if=${DEPLOY_DIR_IMAGE}/zImage of=${EMMC_IMAGE} bs=1024 seek=${KERNEL_PARTITION_OFFSET}
-    resize2fs ${DEPLOY_DIR_IMAGE}/${IMAGE_NAME}.rootfs.ext4 ${ROOTFS_PARTITION_SIZE}k
+    resize2fs ${IMGDEPLOYDIR}/${IMAGE_NAME}.rootfs.ext4 ${ROOTFS_PARTITION_SIZE}k
     # Truncate on purpose
-    dd if=${DEPLOY_DIR_IMAGE}/${IMAGE_NAME}.rootfs.ext4 of=${EMMC_IMAGE} bs=1024 seek=${ROOTFS_PARTITION_OFFSET} count=${IMAGE_ROOTFS_SIZE}
+    dd if=${IMGDEPLOYDIR}/${IMAGE_NAME}.rootfs.ext4 of=${EMMC_IMAGE} bs=1024 seek=${ROOTFS_PARTITION_OFFSET} count=${IMAGE_ROOTFS_SIZE}
 }
 
 kernel_do_install_append() {
@@ -53,4 +53,4 @@ kernel_do_install_append() {
 
 pkg_postinst_kernel-image () {
 	true
-}
\ No newline at end of file
+}

Only

 dd conv=notrunc if=${DEPLOY_DIR_IMAGE}/zImage of=${EMMC_IMAGE} bs=1024 seek=${KERNEL_PARTITION_OFFSET}

In hd-emmc_image_single.bbclass won't work if set to IMGDEPLOYDIR..

[Erik Slagter]

Yes, VU+ is very proactive at the moment, cheers!

 

Beeker (= hains?), thanks for your effort. Please come with something that works in both master and master-next, that's a requirement for a BSP module.

[Beeker]

Yep.

 

I gonna test master-next from scratch for hd51

 

And master-for-gcc5 is  ok instead of master?  I use Ubuntu 16.10

[Erik Slagter]

Yes master-for-gcc5 is fine, I will update it now, hold on and pull.

[Beeker]

Small issue:

 

elfutils(OE-core) in master-for-gcc5 is too old for Ubuntu 16.10. I replaced it with elfutils from master-next.

 

Seems to be fine for now....

 

Issue for TODO list.

[Beeker]

I had to upgrade glb-2.0 also..But a lot of missing bbclasses en configure failures.

 

So i can test master-next.  For master/master-for-gcc5 it's better  to downgrade Ubuntu.

[Beeker]

The entire meta-hd. to test for master / master-for-gcc5.

 

I start build from scratch for master-next

Attached Files

•  meta-hd.tar.gz   204.9KB   5 downloads