Re: [Angstrom-devel] U-Boot in meta-ti is broken
I have experience with beaglebone only, but even with it I had to bitbake u-boot-denx instead of simple u-boot. If I recall it correctly pure u-boot does not crash though, but really wrong recipe gets run. -Matti 2014/1/17 Philip Balister phi...@balister.org On 01/17/2014 02:00 PM, Ulf Samuelsson wrote: It looks like the some of the u-boot recipes in meta-ti are broken. They do not include COMPATIBLE_MACHINE statements, so when building u-boot for non-ti machines, this recipe is choosen instead. This includes the u-boot 13.01 recipe Should this email be o nthe meta-ti list? Philip BR ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel
Re: [Angstrom-devel] what file kernel configuration really uses?
2013/10/27 Koen Kooi k...@dominion.thruhere.net On Wed, 2013-10-23 at 14:44 +0300, matti kaasinen wrote: 2013/10/23 Khem Raj raj.k...@gmail.com Hi Ulf, Yes, linux.inc seems doing the job as you told - this clears a lot. I had been patching wrong file:${S}/defconfig instead of ${WORKDIR}/defconfig. It seems that I'm not alone with this mistake. ${S}/defconfig seems to be created by two patches: 0002-add-defconfig-file-to-use-as-.config.patch makes skeleton and 0073-defconfig-Update-bone-default-config.patch makes some modefications. What I mean above is that beaglebone folks have made those patches for some reason that is not quite clear tome now considering how ${S}/defconfig is produced in linux.inc. ${S}/defconfig is neither used nor produced by OE. I was wrong in that there are two patches that create ${S}/defconfig. Instead there are three of them: 0002-add-defconfig-file-to-use-as-.config.patch 0044-am33xx-Add-default-config.patch 0073-defconfig-Update-bone-default-config.patch Quoted from oe_manual The patch will be applied from the unpacked source directory, ${S}. Above patches create and modify defconfig file. 0002-add created it and next two tweak it slightly. I double checked this by first deleting: ${WORKDIR}/defconfig and ${S}/defconfig and then running: bitbake -f -c unpack linux-mainline At this point there is ${WORKDIR}/defconfig that my layer provides. Then I run: bitbake -f -c patch linux-mainline Now there is also ${S}/defconfig. Then I executed patches with those three patch files in quite a different place - and that produced there defconfig file that was identical with ${S}/defconfig. BTW Koen, please check who has signed off: https://github.com/beagleboard/meta-beagleboard/blob/master/common-bsp/recipes-kernel/linux/linux-mainline-3.8/not-capebus/0002-add-defconfig-file-to-use-as-.config.patch He might be someone you know :-) So, I would say OE produces ${S}/defconfig. ${WORKDIR}/defconfig (important one) is most likely coming from ./linux/linux-mainline-3.8/beaglebone/defconfig as there is only one difference that could have come from configuration process. It seems that configuration fragments do not work in regular Angstrom - I suppose they are just Yocto stuff. yes. Providing defconfig directly did not work - most likely it was written over by the patching the seems creating the ${WORKDIR}/defconfig what do you mean ? defconfig is provided as any other file and then munged over in WORKDIR to make a .config This is outdated information - wild quess - before I noticed how that ${S}/defconfig was really generated by those patches I explained above. As I said above, ${S}/defconfig is not used in the build. Good to know Thanks ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel
Re: [Angstrom-devel] what file kernel configuration really uses?
There is another question, I think I mentioned it earlier also in this chain. I have also asked it in other time as own subject. No-one seems to have opinion about that. Koen, you might know, if it is a feature misbehaviour or problem in my set-up. For instance when I ran those tests I did first fetch. Then I ran unpack. Unpacking ran again fetch and then unpack. Then I ran patch. Again, at least unpack was run again and then patch. Similar behaviour seems to happen with configure so that patch will be run again even though it had been ran manually before. This feature makes paching quite awkward as if I edit some file that has been unpacked and applied with other patches, that will be written over when running configure or compile to check how edits worked. I always get two warnings when I run bitbake: WARNING: No recipes available for: /home/sw/cpr3/oe/sources/meta-handheld/recipes-core/udev/udev_164.bbappend /home/sw/cpr3/oe/sources/meta-intel/meta-fri2/recipes-core/tiny-init/tiny-init.bbappend But I suppose they are nothing to do with this issue. So, Is this something that should happen or should I try to find set-up problem of some kind? Thanks, Matti 2013/10/28 matti kaasinen matti.kaasi...@gmail.com 2013/10/27 Koen Kooi k...@dominion.thruhere.net On Wed, 2013-10-23 at 14:44 +0300, matti kaasinen wrote: 2013/10/23 Khem Raj raj.k...@gmail.com Hi Ulf, Yes, linux.inc seems doing the job as you told - this clears a lot. I had been patching wrong file:${S}/defconfig instead of ${WORKDIR}/defconfig. It seems that I'm not alone with this mistake. ${S}/defconfig seems to be created by two patches: 0002-add-defconfig-file-to-use-as-.config.patch makes skeleton and 0073-defconfig-Update-bone-default-config.patch makes some modefications. What I mean above is that beaglebone folks have made those patches for some reason that is not quite clear tome now considering how ${S}/defconfig is produced in linux.inc. ${S}/defconfig is neither used nor produced by OE. I was wrong in that there are two patches that create ${S}/defconfig. Instead there are three of them: 0002-add-defconfig-file-to-use-as-.config.patch 0044-am33xx-Add-default-config.patch 0073-defconfig-Update-bone-default-config.patch Quoted from oe_manual The patch will be applied from the unpacked source directory, ${S}. Above patches create and modify defconfig file. 0002-add created it and next two tweak it slightly. I double checked this by first deleting: ${WORKDIR}/defconfig and ${S}/defconfig and then running: bitbake -f -c unpack linux-mainline At this point there is ${WORKDIR}/defconfig that my layer provides. Then I run: bitbake -f -c patch linux-mainline Now there is also ${S}/defconfig. Then I executed patches with those three patch files in quite a different place - and that produced there defconfig file that was identical with ${S}/defconfig. BTW Koen, please check who has signed off: https://github.com/beagleboard/meta-beagleboard/blob/master/common-bsp/recipes-kernel/linux/linux-mainline-3.8/not-capebus/0002-add-defconfig-file-to-use-as-.config.patch He might be someone you know :-) So, I would say OE produces ${S}/defconfig. ${WORKDIR}/defconfig (important one) is most likely coming from ./linux/linux-mainline-3.8/beaglebone/defconfig as there is only one difference that could have come from configuration process. It seems that configuration fragments do not work in regular Angstrom - I suppose they are just Yocto stuff. yes. Providing defconfig directly did not work - most likely it was written over by the patching the seems creating the ${WORKDIR}/defconfig what do you mean ? defconfig is provided as any other file and then munged over in WORKDIR to make a .config This is outdated information - wild quess - before I noticed how that ${S}/defconfig was really generated by those patches I explained above. As I said above, ${S}/defconfig is not used in the build. Good to know Thanks ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel
Re: [Angstrom-devel] ubifs operation in u-boot and Linux with am335x device
Now I tried also with erasing: I ran ubiformat and ubiattch with -O 2048 Then in uboot ubi part rootfs gave UBI: attaching mtd1 to ubi0 UBI: physical eraseblock size: 131072 bytes (128 KiB) UBI: logical eraseblock size:129024 bytes UBI: smallest flash I/O unit:2048 UBI: sub-page size: 512 UBI: VID header offset: 512 (aligned 512) UBI: data offset:2048 ECC: uncorrectable. ECC: uncorrectable. UBI error: ubi_io_read: error -74 while reading 64 bytes from PEB 0:0, read 64 bytes UBI error: validate_ec_hdr: bad VID header offset 2048, expected 512 UBI error: validate_ec_hdr: bad EC header UBI error: ubi_io_read_ec_hdr: validation failed for PEB 0 UBI error: ubi_init: cannot attach mtd1 UBI error: ubi_init: UBI error: cannot initialize UBI, error -22 UBI init error 22 Then when I ran ubi part rootfs 2048 I will get UBI: attaching mtd1 to ubi0 UBI: physical eraseblock size: 131072 bytes (128 KiB) UBI: logical eraseblock size:126976 bytes UBI: smallest flash I/O unit:2048 UBI: sub-page size: 512 UBI: VID header offset: 2048 (aligned 2048) UBI: data offset:4096 ECC: uncorrectable. ECC: uncorrectable. UBI error: ubi_io_read: error -74 while reading 64 bytes from PEB 0:0, read 64 bytes ECC: uncorrectable. ECC: uncorrectable. UBI error: ubi_io_read: error -74 while reading 64 bytes from PEB 1:0, read 64 bytes ECC: uncorrectable. . etc UBI error: ubi_io_read: error -74 while reading 64 bytes from PEB 1984:0, read 64 bytes ECC: uncorrectable. ECC: uncorrectable. UBI error: ubi_io_read: error -74 while reading 64 bytes from PEB 1985:0, read 64 bytes ECC: uncorrectable. ECC: uncorrectable. UBI error: ubi_io_read: error -74 while reading 64 bytes from PEB 1986:0, read 64 bytes ECC: uncorrectable. ECC: uncorrectable. UBI error: ubi_io_read: error -74 while reading 64 bytes from PEB 1987:0, read 64 bytes ECC: uncorrectable. ECC: uncorrectable. ECC: uncorrectable. ECC: uncorrectable. ECC: uncorrectable. etc ... ECC: uncorrectable. ECC: uncorrectable. ECC: uncorrectable. ECC: uncorrectable. ECC: uncorrectable. ECC: uncorrectable. ECC: uncorrectable. UBI error: ubi_io_read: error -74 while reading 126976 bytes from PEB 516:4096, read 126976 bytes UBI warning: ubi_eba_copy_leb: error -74 while reading data from PEB 516 UBI error: wear_leveling_worker: error -74 while moving PEB 516 to PEB 617 UBI warning: ubi_ro_mode: switch to read-only mode UBI error: do_work: work failed with error code -74 UBI: attached mtd1 to ubi0 UBI: MTD device name:mtd=7 UBI: MTD device size:248 MiB UBI: number of good PEBs:1980 UBI: number of bad PEBs: 8 UBI: max. allowed volumes: 128 UBI: wear-leveling threshold:4096 UBI: number of internal volumes: 1 UBI: number of user volumes: 1 UBI: available PEBs: 15 UBI: total number of reserved PEBs: 1965 UBI: number of PEBs reserved for bad PEB handling: 19 UBI: max/mean erase counter: 1/0 So, looks pretty much the same as before without defining -O switch. -Matti 2013/10/25 matti kaasinen matti.kaasi...@gmail.com 2013/10/25 Gary Thomas g...@mlbassoc.com Unfortunately fixing that to correct value 52 did not help either. So, I'm running out out ideas. I'll enclose steps I executed both on Linux and u-boot side and their outcome. These are from the initial state before trying to fix that omap_gpmc.h. Can you try this again (from scratch by first erasing the partition) and add the -O option? This looks exactly like the problem I described. I tried it, but I did not do erasing in between. I'll try it with erasing on Monday. Thanks a lot, Matti ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel
[Angstrom-devel] ubifs operation in u-boot and Linux with am335x device
Hi! Do anyone happen to know how u-boot should be set to access Linux created ubifs partition correctly? I'm working on beaglebone variant board with nand flash addition. Linux side (3.8.13) seems working fine provided that I don't mount it in u-boot side. However, if i do, it gets corrupted. For instance, if I attach Linux created ubifs partition (ubi part partition), i will get tons of ECC: uncorrectable errors. After re-boot also Linux side will get these errors. I'm using Angstrom distribution with following details. BB_VERSION= 1.17.0 TARGET_ARCH = arm TARGET_OS = linux-gnueabi MACHINE = beaglebone DISTRO= angstrom DISTRO_VERSION= v2012.12 TUNE_FEATURES = armv7a vfp neon cortexa8 TARGET_FPU= vfp-neon Any hints warmly appreciated. -Matti ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel
Re: [Angstrom-devel] ubifs operation in u-boot and Linux with am335x device
Hi Gary, I think I have found something that could explain this behaviour. They happen to be on u-boot side. In fact I just wrote to u-boot mailin list about this. I'll copy it below: Linux 3.8.13 seems to use 52 ecc bytes mapping them to positions 12..64. Default u-boot set-up found from include/asm/arch-am33x/omap_ gpmc.h define: #define GPMC_NAND_HW_BCH8_ECC_LAYOUT {\ .eccbytes = 56,\ .eccpos = {2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15,\ 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27,\ 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39,\ 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51,\ 52, 53, 54, 55, 56, 57},\ .oobfree = {\ {.offset = 58,\ .length = 6 } } \ } that seems using 56 ecc bytes, mapping them to positions 2..57. This 56 bytes seems fighting with the fact found from u-boot README.omap3/BCH8 set-up that points, that * CONFIG_SYS_NAND_ECCBYTES must be 13 for this BCH8 set-up On the other hand, am335x_evm.h use CONFIG_SYS_NAND_ECCBYTES to calculate CONFIG_SYS_NAND_ECCTOTAL that is same as .eccbytes in GPMC_NAND_HW_BCH8_ECC_LAYOUT definition. Value 13 gives 52 bytes that is compatible with Linux whereas 14 gives that 56 found from GPMC_NAND_HW_BCH8_ECC_LAYOUT definition above. Also text in README.omap3 states: The NAND OOB layout is the same as in linux kernel, if the linux kernel BCH8 implementation for OMAP3 works for you so the u-boot version should also. When you require the SPL to read with BCH8 there are two more configs to change: * CONFIG_SYS_NAND_ECCPOS (must be the same as .eccpos in GPMC_NAND_HW_BCH8_ECC_LAYOUT defined in arch/arm/include/asm/arch- omap3/omap_gpmc.h) * CONFIG_SYS_NAND_ECCSIZE must be 512 * CONFIG_SYS_NAND_ECCBYTES must be 13 for this BCH8 setup I would interpret this so that no modifications are needed GPMC_NAND_HW_BCH8_ECC_LAYOUT structure. That can't be true even if the mapping was the same as .eccbytes field have wrong value considering what the last line states for BCH8 set-up. If above assumption is correct, isn't this really a bug in omap_gpmc.h, not just a Linux version related problem? Regards, Matti 2013/10/25 Gary Thomas g...@mlbassoc.com On 2013-10-25 00:01, matti kaasinen wrote: Hi! Do anyone happen to know how u-boot should be set to access Linux created ubifs partition correctly? I'm working on beaglebone variant board with nand flash addition. Linux side (3.8.13) seems working fine provided that I don't mount it in u-boot side. However, if i do, it gets corrupted. For instance, if I attach Linux created ubifs partition (ubi part partition), i will get tons of ECC: uncorrectable errors. After re-boot also Linux side will get these errors. I'm using Angstrom distribution with following details. BB_VERSION= 1.17.0 TARGET_ARCH = arm TARGET_OS = linux-gnueabi MACHINE = beaglebone DISTRO= angstrom DISTRO_VERSION= v2012.12 TUNE_FEATURES = armv7a vfp neon cortexa8 TARGET_FPU= vfp-neon Any hints warmly appreciated. Details about your NAND device will probably be necessary as well as the actual errors and operations you tried. I suspect that the problem may have to do with NAND sub-page writes. I've experienced this on other platforms and was able to work around it by not letting UBI use them. For example on my 2K Micron NAND device, I use this command to attach: # ubiattach /dev/ubi_ctrl -m 3 -O 2048 The -O option forces the UBI system to only use 2048 byte writes, as opposed to the 512 byte sub-pages that it would like to use. If you want to boot with this as your file system, use these kernel command line options: root=ubi0:root ubi.mtd=3,2048 rootfstype=ubifs See if this helps. If not, send as many details as you can. -- --**-- Gary Thomas | Consulting for the MLB Associates |Embedded world --**-- __**_ Angstrom-distro-devel mailing list Angstrom-distro-devel@**linuxtogo.orgAngstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/**cgi-bin/mailman/listinfo/** angstrom-distro-develhttp://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel
Re: [Angstrom-devel] ubifs operation in u-boot and Linux with am335x device
] (ubi_attach+0x169/0x244) from [c0229373] (ubi_attach_mtd_dev+0x4ff/0x980) [ 184.429503] [c0229373] (ubi_attach_mtd_dev+0x4ff/0x980) from [c0229997] (ctrl_cdev_ioctl+0x7f/0x104) [ 184.439565] [c0229997] (ctrl_cdev_ioctl+0x7f/0x104) from [c009d6b1] (vfs_ioctl+0x15/0x20) [ 184.448616] [c009d6b1] (vfs_ioctl+0x15/0x20) from [c009de57] (do_vfs_ioctl+0x3ad/0x3e2) [ 184.457487] [c009de57] (do_vfs_ioctl+0x3ad/0x3e2) from [c009debd] (sys_ioctl+0x31/0x4c) [ 184.466358] [c009debd] (sys_ioctl+0x31/0x4c) from [c000c641] (ret_fast_syscall+0x1/0x46) [ 184.494290] uncorrectable error : [ Best Regards, Matti 2013/10/25 matti kaasinen matti.kaasi...@gmail.com Hi Gary, I think I have found something that could explain this behaviour. They happen to be on u-boot side. In fact I just wrote to u-boot mailin list about this. I'll copy it below: Linux 3.8.13 seems to use 52 ecc bytes mapping them to positions 12..64. Default u-boot set-up found from include/asm/arch-am33x/omap_ gpmc.h define: #define GPMC_NAND_HW_BCH8_ECC_LAYOUT {\ .eccbytes = 56,\ .eccpos = {2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15,\ 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27,\ 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39,\ 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51,\ 52, 53, 54, 55, 56, 57},\ .oobfree = {\ {.offset = 58,\ .length = 6 } } \ } that seems using 56 ecc bytes, mapping them to positions 2..57. This 56 bytes seems fighting with the fact found from u-boot README.omap3/BCH8 set-up that points, that * CONFIG_SYS_NAND_ECCBYTES must be 13 for this BCH8 set-up On the other hand, am335x_evm.h use CONFIG_SYS_NAND_ECCBYTES to calculate CONFIG_SYS_NAND_ECCTOTAL that is same as .eccbytes in GPMC_NAND_HW_BCH8_ECC_LAYOUT definition. Value 13 gives 52 bytes that is compatible with Linux whereas 14 gives that 56 found from GPMC_NAND_HW_BCH8_ECC_LAYOUT definition above. Also text in README.omap3 states: The NAND OOB layout is the same as in linux kernel, if the linux kernel BCH8 implementation for OMAP3 works for you so the u-boot version should also. When you require the SPL to read with BCH8 there are two more configs to change: * CONFIG_SYS_NAND_ECCPOS (must be the same as .eccpos in GPMC_NAND_HW_BCH8_ECC_LAYOUT defined in arch/arm/include/asm/arch- omap3/omap_gpmc.h) * CONFIG_SYS_NAND_ECCSIZE must be 512 * CONFIG_SYS_NAND_ECCBYTES must be 13 for this BCH8 setup I would interpret this so that no modifications are needed GPMC_NAND_HW_BCH8_ECC_LAYOUT structure. That can't be true even if the mapping was the same as .eccbytes field have wrong value considering what the last line states for BCH8 set-up. If above assumption is correct, isn't this really a bug in omap_gpmc.h, not just a Linux version related problem? Regards, Matti 2013/10/25 Gary Thomas g...@mlbassoc.com On 2013-10-25 00:01, matti kaasinen wrote: Hi! Do anyone happen to know how u-boot should be set to access Linux created ubifs partition correctly? I'm working on beaglebone variant board with nand flash addition. Linux side (3.8.13) seems working fine provided that I don't mount it in u-boot side. However, if i do, it gets corrupted. For instance, if I attach Linux created ubifs partition (ubi part partition), i will get tons of ECC: uncorrectable errors. After re-boot also Linux side will get these errors. I'm using Angstrom distribution with following details. BB_VERSION= 1.17.0 TARGET_ARCH = arm TARGET_OS = linux-gnueabi MACHINE = beaglebone DISTRO= angstrom DISTRO_VERSION= v2012.12 TUNE_FEATURES = armv7a vfp neon cortexa8 TARGET_FPU= vfp-neon Any hints warmly appreciated. Details about your NAND device will probably be necessary as well as the actual errors and operations you tried. I suspect that the problem may have to do with NAND sub-page writes. I've experienced this on other platforms and was able to work around it by not letting UBI use them. For example on my 2K Micron NAND device, I use this command to attach: # ubiattach /dev/ubi_ctrl -m 3 -O 2048 The -O option forces the UBI system to only use 2048 byte writes, as opposed to the 512 byte sub-pages that it would like to use. If you want to boot with this as your file system, use these kernel command line options: root=ubi0:root ubi.mtd=3,2048 rootfstype=ubifs See if this helps. If not, send as many details as you can. -- --**-- Gary Thomas | Consulting for the MLB Associates |Embedded world --**-- __**_ Angstrom-distro-devel mailing list Angstrom-distro-devel@**linuxtogo.orgAngstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/**cgi-bin
Re: [Angstrom-devel] ubifs operation in u-boot and Linux with am335x device
2013/10/25 Gary Thomas g...@mlbassoc.com Unfortunately fixing that to correct value 52 did not help either. So, I'm running out out ideas. I'll enclose steps I executed both on Linux and u-boot side and their outcome. These are from the initial state before trying to fix that omap_gpmc.h. Can you try this again (from scratch by first erasing the partition) and add the -O option? This looks exactly like the problem I described. I tried it, but I did not do erasing in between. I'll try it with erasing on Monday. Thanks a lot, Matti ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel
Re: [Angstrom-devel] what file kernel configuration really uses?
Hi Ulf, Yes, linux.inc seems doing the job as you told - this clears a lot. I had been patching wrong file:${S}/defconfig instead of ${WORKDIR}/defconfig. It seems that I'm not alone with this mistake. ${S}/defconfig seems to be created by two patches: 0002-add-defconfig-file-to-use-as-.config.patch makes skeleton and 0073-defconfig-Update-bone-default-config.patch makes some modefications. ${WORKDIR}/defconfig (important one) is most likely coming from ./linux/linux-mainline-3.8/beaglebone/defconfig as there is only one difference that could have come from configuration process. It seems that configuration fragments do not work in regular Angstrom - I suppose they are just Yocto stuff. Providing defconfig directly did not work - most likely it was written over by the patching the seems creating the ${WORKDIR}/defconfig Downside is that my beaglebone version of defconfig seems to get used instead of mine even though my layer should have higher priority. I hope this is the last thing I should cleared. Thanks, Matti 2013/10/22 Ulf Samuelsson angstrom-...@emagii.com On 2013-10-22 17:20, matti kaasinen wrote: Thanks Ulf, It seems to work in that way. However, I'm a bit surprised that it works so as as I mentioned above all the procedures - patching defconfig in the kernel build directory, providing defconfig in metadata and providing configuration fragments as described in the Yocto Kernel development manual - give the same outcome in the defconfig at the kernel build directory. What is happening is dependent on the kernel recipe. Typically, you find that linux.inc does the job, and in do_configure, which is run when you do: bitbake -c configure virtual/kernel ${WORKDIR}/defconfig is altered to ensure it makes sense. A lot of options are simply deleted. ${S}/.config is created as an empty file and then the deleted options are added with a proper value. At the end, defconfig is appended to the ${S}/.config so when you run bitbake -c configure virtual/kernel both ${WORKDIR}/defconfig and ${S}/.config are changed. /Ulf What command do you use when you are using .config directly? My experience is that when I for instance run: bitbake -f -c configure virtual/kernel after bitbake -f -c patch virtual/kernel bitbake executes again do_patch, that at least rides over defconfig if I edited that. In fact it seems that bitbake -c config runs always do_patch even if previous command was patch and no modifications were in between. BR, Matti 2013/10/22 Ulf Samuelsson angstrom-...@emagii.com The defconfig file is present in the meta-layers and copied to the kernel build directory. It is used to create the .config file in the kernel source directory. If you modify the .config file, you will see changes in the kernel file. if you modify the defconfig file in the build directory, nothing happens. I typically change the .config and copy the result to the defconfig in the meta-layer. Then I rebuild from scratch. bitbake -c cleansstate virtual/kernel bitbake virtual/kernel Best Regards Ulf Samuelsson u...@emagii.com +46 (722) 427 437 22 okt 2013 kl. 14:04 skrev matti kaasinen matti.kaasi...@gmail.com: Hi! What configuration kernel build really uses - .config or defconfig? It seems, that menuconfig (bitbake -c menuconfig ) use always .config file. I have problem that changes in defconfig are not seen in kernel features. Instead they seem the same that are in .config file I have tried configuration fragments, patches and providing defconfig directly. They all seem to give proper defconfig. However, menuconfig never provide the changed configurations. Also, for instance when I try to configure HW EEC operation for NAND flash using CONFIG_MTD_NAND_OMAP_BCH. omap2.c reports that CONFIG_MTD_NAND_OMAP_BCH is not enabled. I've been workin on beaglebone variant - layer over beaglebone. Build Configuration: BB_VERSION= 1.17.0 TARGET_ARCH = arm TARGET_OS = linux-gnueabi MACHINE = beaglebone DISTRO= angstrom DISTRO_VERSION= v2012.12 TUNE_FEATURES = armv7a vfp neon cortexa8 TARGET_FPU= vfp-neon oe_sitecno oe_emergence = unknown:unknown meta-angstrom = angstrom-v2012.12-yocto1.3:**b7f8207b94d9a0ece73ad212a193cb** 2c95bd17ee These setting give kernel 3.8.11. Is there something I have missed? Thanks in advance, Matti __**_ Angstrom-distro-devel mailing list Angstrom-distro-devel@**linuxtogo.orgAngstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/**cgi-bin/mailman/listinfo/** angstrom-distro-develhttp://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel __**_ Angstrom-distro-devel mailing list Angstrom-distro-devel@**linuxtogo.orgAngstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/**cgi-bin/mailman/listinfo
Re: [Angstrom-devel] what file kernel configuration really uses?
2013/10/23 Khem Raj raj.k...@gmail.com Hi Ulf, Yes, linux.inc seems doing the job as you told - this clears a lot. I had been patching wrong file:${S}/defconfig instead of ${WORKDIR}/defconfig. It seems that I'm not alone with this mistake. ${S}/defconfig seems to be created by two patches: 0002-add-defconfig-file-to-use-as-.config.patch makes skeleton and 0073-defconfig-Update-bone-default-config.patch makes some modefications. What I mean above is that beaglebone folks have made those patches for some reason that is not quite clear tome now considering how ${S}/defconfig is produced in linux.inc. ${WORKDIR}/defconfig (important one) is most likely coming from ./linux/linux-mainline-3.8/beaglebone/defconfig as there is only one difference that could have come from configuration process. It seems that configuration fragments do not work in regular Angstrom - I suppose they are just Yocto stuff. yes. Providing defconfig directly did not work - most likely it was written over by the patching the seems creating the ${WORKDIR}/defconfig what do you mean ? defconfig is provided as any other file and then munged over in WORKDIR to make a .config This is outdated information - wild quess - before I noticed how that ${S}/defconfig was really generated by those patches I explained above. usually you would keep the complete defconfig in your layer and use it. You would start with the given reference defconfig and tweak it to your interest and then do make savedefconfig which should generate a defconfig like arch/arm/configs which then you can save as a defconfig file in your layer and use it to replace the defconfig file that meta-beagleboard is providing Downside is that my beaglebone version of defconfig seems to get used instead of mine even though my layer should have higher priority. I hope this is the last thing I should cleared. for conf and include files it will use the BBPATH and not priority which means your layer should appear before meta-beagleboard in BBPATH order. Thanks, I'll check this Matti ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel
Re: [Angstrom-devel] what file kernel configuration really uses?
2013/10/23 matti kaasinen matti.kaasi...@gmail.com for conf and include files it will use the BBPATH and not priority which means your layer should appear before meta-beagleboard in BBPATH order. It did not help chnging BBPATH in layer.conf It used to be BBPATH .= :${LAYERDIR} and I changed it to: BBPATH =. ${LAYERDIR}: It still fetches beaglebone's defconfig. Should I change my bbappend file instad not using FILESEXTRAPATHS_prepend but using FILESPATH_prepend instead? Do FILESPATH have precedence over FILESEXTRAPATHS? Cheers, Matti ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel
Re: [Angstrom-devel] what file kernel configuration really uses?
Ulf, I'm not quite sure what you mean. I've understood that layers (read layer.conf files) are scanned through using locations and order found in BBLAYERS variable set in bblayers.conf file. BBPATH is initialized in bblayers.conf as ${TOPDIR}. Thereafter (my wild guess) BBPATH is appended (or prepended) with values given in the the layer.conf files. I switched to prepending and it did not help this issue. I have set my layer as the first instance in BBLAYERS variable whereas meta-beagleboard/common-bsp is somewhere in the middle. My layer is using prepending when assigning in BBPATH (should be the first instance also in BBPATH) whereas meta-beagleboard/common-bsp uses appending. Therefore, my layer should be handled before meta-beagleboard/common-bsp, should it not? I may have understood this all wrong, though. Regards, Matti 2013/10/23 Ulf Samuelsson angstrom-...@emagii.com You are looking in the wrong file. In Angstrom, you need to change BBPATH in setup-scripts/conf/bblayers.conf Not in the layer.conf in your own layer. Best Regards Ulf Samuelsson u...@emagii.com 23 okt 2013 kl. 14:24 skrev matti kaasinen matti.kaasi...@gmail.com: 2013/10/23 matti kaasinen matti.kaasi...@gmail.com for conf and include files it will use the BBPATH and not priority which means your layer should appear before meta-beagleboard in BBPATH order. It did not help chnging BBPATH in layer.conf It used to be BBPATH .= :${LAYERDIR} and I changed it to: BBPATH =. ${LAYERDIR}: It still fetches beaglebone's defconfig. Should I change my bbappend file instad not using FILESEXTRAPATHS_prepend but using FILESPATH_prepend instead? Do FILESPATH have precedence over FILESEXTRAPATHS? Cheers, Matti ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel
[Angstrom-devel] what file kernel configuration really uses?
Hi! What configuration kernel build really uses - .config or defconfig? It seems, that menuconfig (bitbake -c menuconfig ) use always .config file. I have problem that changes in defconfig are not seen in kernel features. Instead they seem the same that are in .config file I have tried configuration fragments, patches and providing defconfig directly. They all seem to give proper defconfig. However, menuconfig never provide the changed configurations. Also, for instance when I try to configure HW EEC operation for NAND flash using CONFIG_MTD_NAND_OMAP_BCH. omap2.c reports that CONFIG_MTD_NAND_OMAP_BCH is not enabled. I've been workin on beaglebone variant - layer over beaglebone. Build Configuration: BB_VERSION= 1.17.0 TARGET_ARCH = arm TARGET_OS = linux-gnueabi MACHINE = beaglebone DISTRO= angstrom DISTRO_VERSION= v2012.12 TUNE_FEATURES = armv7a vfp neon cortexa8 TARGET_FPU= vfp-neon oe_sitecno oe_emergence = unknown:unknown meta-angstrom = angstrom-v2012.12-yocto1.3:b7f8207b94d9a0ece73ad212a193cb2c95bd17ee These setting give kernel 3.8.11. Is there something I have missed? Thanks in advance, Matti ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel
Re: [Angstrom-devel] what file kernel configuration really uses?
Thanks Ulf, It seems to work in that way. However, I'm a bit surprised that it works so as as I mentioned above all the procedures - patching defconfig in the kernel build directory, providing defconfig in metadata and providing configuration fragments as described in the Yocto Kernel development manual - give the same outcome in the defconfig at the kernel build directory. What command do you use when you are using .config directly? My experience is that when I for instance run: bitbake -f -c configure virtual/kernel after bitbake -f -c patch virtual/kernel bitbake executes again do_patch, that at least rides over defconfig if I edited that. In fact it seems that bitbake -c config runs always do_patch even if previous command was patch and no modifications were in between. BR, Matti 2013/10/22 Ulf Samuelsson angstrom-...@emagii.com The defconfig file is present in the meta-layers and copied to the kernel build directory. It is used to create the .config file in the kernel source directory. If you modify the .config file, you will see changes in the kernel file. if you modify the defconfig file in the build directory, nothing happens. I typically change the .config and copy the result to the defconfig in the meta-layer. Then I rebuild from scratch. bitbake -c cleansstate virtual/kernel bitbake virtual/kernel Best Regards Ulf Samuelsson u...@emagii.com +46 (722) 427 437 22 okt 2013 kl. 14:04 skrev matti kaasinen matti.kaasi...@gmail.com: Hi! What configuration kernel build really uses - .config or defconfig? It seems, that menuconfig (bitbake -c menuconfig ) use always .config file. I have problem that changes in defconfig are not seen in kernel features. Instead they seem the same that are in .config file I have tried configuration fragments, patches and providing defconfig directly. They all seem to give proper defconfig. However, menuconfig never provide the changed configurations. Also, for instance when I try to configure HW EEC operation for NAND flash using CONFIG_MTD_NAND_OMAP_BCH. omap2.c reports that CONFIG_MTD_NAND_OMAP_BCH is not enabled. I've been workin on beaglebone variant - layer over beaglebone. Build Configuration: BB_VERSION= 1.17.0 TARGET_ARCH = arm TARGET_OS = linux-gnueabi MACHINE = beaglebone DISTRO= angstrom DISTRO_VERSION= v2012.12 TUNE_FEATURES = armv7a vfp neon cortexa8 TARGET_FPU= vfp-neon oe_sitecno oe_emergence = unknown:unknown meta-angstrom = angstrom-v2012.12-yocto1.3:b7f8207b94d9a0ece73ad212a193cb2c95bd17ee These setting give kernel 3.8.11. Is there something I have missed? Thanks in advance, Matti ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel
[Angstrom-devel] patch - edit - compile
Possibly this is not quite correct forum for my question. However, most likely you know the correct answer. I try to patch u-boot-denx for beaglebone with procedures below: 1) bitbake -f -c patch u-boot-denx 2) source editions in workdir 3) bitbake -f -c compile u-boot-denx What happens is that command at 3) executes: do_patch do_configure and do_compile. As a consequence editions are lost. I also tested just: bitbake -f -c patch u-boot-denx bitbake -f -c configure u-boot-denx resulting: do_patch do_patch do_configure Running commands below: bitbake -f -c patch u-boot-denx bitbake -f -c configure u-boot-denx bitbake -f -c configure u-boot-denx Gave tasks below: do_patch do_patch do_configure do_configure Now the question: Is there any trick to make those edition after patching so that patching is not run again by following do_configure? best regards, Matti ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel
Re: [Angstrom-devel] BUG in bash recipe / angstrom / 2011.03
Ok, they are yocto branches. Board I'm playing with is at91sam9260ek derivative (ld), so, yocto is out of question. As Atmel is not yocto participant there is nothing in yocto for this board, so it's far too long a road to walk. In fact, even the OE-classic stuff for 9260ek is very outdated. I have updated some - like kernel - for my own boards layer, if you can call that such in OE-classic. Anyhow, thanks for you information. BTW I suppose there is no point providing any new stuff to OE-classic. I wold have some, like usb_modeswitch update, chrony update, architecture independent version of isl2026 kenel module, sakis3g 3g connection script. Somehow, I feel it as waste of my and others time not sharing them. Let's take this isl2026. It really was hard to find the original version I started to work from (Buglabs version). That was very tightly dependent to their board's architecture while mine is not. On the other hand, it's not necessarily that usual that RTC is needed as it was in our case when there was a bug in at91sam9260 RTC. -Matti 2012/12/4 Koen Kooi k...@dominion.thruhere.net Op 3 dec. 2012, om 20:20 heeft matti kaasinen matti.kaasi...@gmail.com het volgende geschreven: Hi Koen, How are these versions related with OE-classic? They aren't related. Where are they sitting? They are branches of the setupscripts repo: http://www.angstrom-distribution.org/building-angstrom regards, Koen -Matti 2012/12/3 Koen Kooi k...@dominion.thruhere.net Op 17 nov. 2012, om 15:33 heeft matti kaasinen matti.kaasi...@gmail.com het volgende geschreven: Hi developers, It seems that bash-3.2-r12.bb does not make /bin/sh linking to /bin/bash. Actually, I suppose, the problem is in bash.inc file. It makes this link in pkg_postinst_bash-sh() function. However, as there is no RPROVIDES statement in the recipe, there is no way using bash-sh package that should provide this link (I think?). On the other hand, I don't think the way recipe tries to create those post inst procedures is quite by the book. Would in not be better using update-alternatives structures? That was my theory - I don't know if that is correct. However, the fact is that there is /bin/bash in the filesystem, but not the link pointing from /bin/sh to /bin/bash. This has been fixed in v2012.05 (and consequently in v2012.12 as well). regrds, Koen ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel
Re: [Angstrom-devel] BUG in bash recipe / angstrom / 2011.03
Hi Koen, How are these versions related with OE-classic? Where are they sitting? -Matti 2012/12/3 Koen Kooi k...@dominion.thruhere.net Op 17 nov. 2012, om 15:33 heeft matti kaasinen matti.kaasi...@gmail.com het volgende geschreven: Hi developers, It seems that bash-3.2-r12.bb does not make /bin/sh linking to /bin/bash. Actually, I suppose, the problem is in bash.inc file. It makes this link in pkg_postinst_bash-sh() function. However, as there is no RPROVIDES statement in the recipe, there is no way using bash-sh package that should provide this link (I think?). On the other hand, I don't think the way recipe tries to create those post inst procedures is quite by the book. Would in not be better using update-alternatives structures? That was my theory - I don't know if that is correct. However, the fact is that there is /bin/bash in the filesystem, but not the link pointing from /bin/sh to /bin/bash. This has been fixed in v2012.05 (and consequently in v2012.12 as well). regrds, Koen ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel
Re: [Angstrom-devel] BUG in bash recipe / angstrom / 2011.03
Hi developers! Consider using following patch for bash/bash.inc file to get bash-sh package available (i.e. /bin/sh - /bin/bash linking executed) in case 2011.03 maintenance is still maintained. It uses update-alternatives call instead of direct ln -s statements by the original recipe Therefore, results can now be easily accessed with update-alternatives call. I don't know how much bash is needed as it seems that also dash shell gets installed/is available with conventional console image with quite satisfactory functionality. Somehow it's installed with lower priority than busybox sh. Anyhow please find my patch below. Take care, Matti --- openembedded/recipes/bash/bash.inc2011-07-26 11:45:18.226525143 +0300 +++ oe_sitecno/overlay/bash/bash_3.2/bash.inc2012-12-01 10:33:08.915944456 +0200 @@ -37,12 +37,14 @@ PACKAGES += bash-sh RDEPENDS_bash-sh = bash +RPROVIDES_${PN} = bash bash-sh + ALLOW_EMPTY_bash-sh = 1 pkg_postinst_bash-sh() { -cd $D/bin ln -sf bash sh +update-alternatives --install ${bindir}/sh sh ${bindir}/bash 100 } pkg_postrm_bash-sh() { -ln -sf /bin/busybox /bin/sh +update-alternatives --remove sh ${bindir}/bash } 2012/11/17 matti kaasinen matti.kaasi...@gmail.com Hi developers, It seems that bash-3.2-r12.bb does not make /bin/sh linking to /bin/bash. Actually, I suppose, the problem is in bash.inc file. It makes this link in pkg_postinst_bash-sh() function. However, as there is no RPROVIDES statement in the recipe, there is no way using bash-sh package that should provide this link (I think?). On the other hand, I don't think the way recipe tries to create those post inst procedures is quite by the book. Would in not be better using update-alternatives structures? That was my theory - I don't know if that is correct. However, the fact is that there is /bin/bash in the filesystem, but not the link pointing from /bin/sh to /bin/bash. cheers, Matti ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel
[Angstrom-devel] BUG in bash recipe / angstrom / 2011.03
Hi developers, It seems that bash-3.2-r12.bb does not make /bin/sh linking to /bin/bash. Actually, I suppose, the problem is in bash.inc file. It makes this link in pkg_postinst_bash-sh() function. However, as there is no RPROVIDES statement in the recipe, there is no way using bash-sh package that should provide this link (I think?). On the other hand, I don't think the way recipe tries to create those post inst procedures is quite by the book. Would in not be better using update-alternatives structures? That was my theory - I don't know if that is correct. However, the fact is that there is /bin/bash in the filesystem, but not the link pointing from /bin/sh to /bin/bash. cheers, Matti ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel
[Angstrom-devel] bash-3.2-r12.bb does not make /bin/sh link
Hi developers, It seems that bash-3.2-r12.bb does not make /bin/sh linking to /bin/sh. Actually I suppose the problem is in bash.inc file. It makes this link in pkg_postinst_bash-sh() function. However, there is no RPROVEDES statement in recipe and therefore it does not provde this bash-sh. There is no way using bash-sh package that provides this link (I think?). -Matti ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel
Re: [Angstrom-devel] data file clash between udev and bluez4
Hi, Some more information: bluez4-4.84 does not install hid2hci, whereas bluez4-4.95 does. So, angstrom-2008.1 should restrict bluez4 version to 4.84 or perhaps advance udev version. -Matti 2012/6/15 matti kaasinen matti.kaasi...@gmail.com: Hi, Something to add: DISTRO=angstrom-2008.1 angstrom-2008-preferred-versions.inc seems to define udev version as 151 whereas angstrom-2010-preferred-versions.inc defines it as 165. Do anyone know, if this version would be compatible with other stuff in angstrom-2008? Or would it be better defining bluez4 with preferred version? On the other hand I could just write overlay recipe just deleting this hdi2hci from udev, but that would not be too elegant, would it? -Matti 2012/6/15 matti kaasinen matti.kaasi...@gmail.com: Hi! I get data file clash error between bluez4 and udev when I try to bitbake somewhat extended console image. Both try to install /lib/udev/hid2hci. Build Configuration: BB_VERSION = 1.12.0 METADATA_BRANCH = 2011.03-maintenance METADATA_REVISION = 75c7b2f TARGET_ARCH = arm TARGET_OS = linux-gnueabi MACHINE = at91sam9260ek DISTRO = angstrom DISTRO_VERSION = 2011.03 TARGET_FPU = soft Thanks, Matti ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel
[Angstrom-devel] data file clash between udev and bluez4
Hi! I get data file clash error between bluez4 and udev when I try to bitbake somewhat extended console image. Both try to install /lib/udev/hid2hci. Build Configuration: BB_VERSION= 1.12.0 METADATA_BRANCH = 2011.03-maintenance METADATA_REVISION = 75c7b2f TARGET_ARCH = arm TARGET_OS = linux-gnueabi MACHINE = at91sam9260ek DISTRO= angstrom DISTRO_VERSION= 2011.03 TARGET_FPU= soft Thanks, Matti ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel
[Angstrom-devel] DOCUMENTATION of R* variables like RDEPENDS
Hi developers! Now, just after learning lesson about currently required syntax of R* variables - http://lists.linuxtogo.org/pipermail/angstrom-distro-devel/2012-April/005928.html - I would like to add my wish to get this fact documented to OE official documentation. By my brief check I found both variations of R* variable syntax used with RDEPENDS variables in the example codes of OE official documentation (Official Manual - OpenEmbedded User Manual). I did not find any notice that not using RXXX_${PN} syntax will break image building (image class?). Problems I had was with Classical OE, so it should not be documented just on Migrating metadata to OE-core page. Thanks, Matti ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel
Re: [Angstrom-devel] RDEPENDS does not pull package to image
Hi Frank, In fact I had followed both conventions. I'm not quite sure what was with this problematic package. Anyhow, I just stripped these RDEPENDS_${PN} versions on the dependency branch - with no visible change. What I noticed from from image build with -D -D switches was that all the runtime dependencies seemed to be set quite nicely, packages were written etc. Only doubt regarding this package I found was: board-switch_1.0-r0.6_at91sam9260ek.ipk in Packages, but mtime differs - re-reading And naturally, that package and all the packages it was depending on were missing. I'll try systematic RDEPENDS_${PN}. Hope that helps. Thanks, Matti 4. huhtikuuta 2012 16.54 Frank Agius ftag...@yahoo.com kirjoitti: On 4/3/2012 4:48 PM, matti kaasinen wrote: Hi! I have faced with a strange feature with Angstrom 2011.03-maintenance. There is one package that does not get pulled to image with RDEPENDS variable and all the packages it RDEPENDS on. Do you have it as RDEPENDS or RDEPENDS_${PN}? According to the migration guide (http://www.openembedded.org/wiki/Migrating_metadata_to_OE-Core) the R* variables should always have a package name override, ie RDEPENDS_${PN}. When I bitbake it with -c clean and run then bitbake for the targeted image, this package seems to built all right ipk file is written, but it is not included in the image. If I insert it to IMAGE_INSTALL list, it (and all the packages it depends on) will be included in the image. Do anyone have hints how to debug this? -Matti ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel
Re: [Angstrom-devel] RDEPENDS does not pull package to image
Yes, Frank, you were right. _${PN} worked! Thanks, Matti 4. huhtikuuta 2012 22.05 matti kaasinen matti.kaasi...@gmail.com kirjoitti: Hi Frank, In fact I had followed both conventions. I'm not quite sure what was with this problematic package. Anyhow, I just stripped these RDEPENDS_${PN} versions on the dependency branch - with no visible change. What I noticed from from image build with -D -D switches was that all the runtime dependencies seemed to be set quite nicely, packages were written etc. Only doubt regarding this package I found was: board-switch_1.0-r0.6_at91sam9260ek.ipk in Packages, but mtime differs - re-reading And naturally, that package and all the packages it was depending on were missing. I'll try systematic RDEPENDS_${PN}. Hope that helps. Thanks, Matti 4. huhtikuuta 2012 16.54 Frank Agius ftag...@yahoo.com kirjoitti: On 4/3/2012 4:48 PM, matti kaasinen wrote: Hi! I have faced with a strange feature with Angstrom 2011.03-maintenance. There is one package that does not get pulled to image with RDEPENDS variable and all the packages it RDEPENDS on. Do you have it as RDEPENDS or RDEPENDS_${PN}? According to the migration guide (http://www.openembedded.org/wiki/Migrating_metadata_to_OE-Core) the R* variables should always have a package name override, ie RDEPENDS_${PN}. When I bitbake it with -c clean and run then bitbake for the targeted image, this package seems to built all right ipk file is written, but it is not included in the image. If I insert it to IMAGE_INSTALL list, it (and all the packages it depends on) will be included in the image. Do anyone have hints how to debug this? -Matti ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel
[Angstrom-devel] RDEPENDS does not pull package to image
Hi! I have faced with a strange feature with Angstrom 2011.03-maintenance. There is one package that does not get pulled to image with RDEPENDS variable and all the packages it RDEPENDS on. When I bitbake it with -c clean and run then bitbake for the targeted image, this package seems to built all right ipk file is written, but it is not included in the image. If I insert it to IMAGE_INSTALL list, it (and all the packages it depends on) will be included in the image. Do anyone have hints how to debug this? -Matti ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel
[Angstrom-devel] 2011.03-maintenance/tzdata_2011b.bb
Hi! I just wonder what tzdata/RCONFLICTS_${PN} = timezones timezone-africa timezone-america fields are referring to. I can't find any other reference from openembedded/recipes for instance for timezone-africa than this instance. Should they not be: tzdata-africa tzdata-america ? -Matti ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel
Re: [Angstrom-devel] 2011.03-maintenance/fakeroot_1.14.5.bb broken
Now it seems to be fixed. Before pulling update, recipe fetched far too small archive compared to my backup (below): -rw-r--r-- 1 matti matti 14143 2012-01-14 10:42 fakeroot_1.14.5.orig.tar.bz2 -rw-r--r-- 1 matti matti 280640 2010-12-13 11:51 fakeroot_1.14.5.orig.tar.bz2-backup This corrupted archive had also wrong md5 sum (naturally). After pulling bitbake update this fixed and backup size matched with fetched size. thanks, Matti 13. tammikuuta 2012 20.46 Khem Raj raj.k...@gmail.com kirjoitti: On Fri, Jan 13, 2012 at 10:23 AM, matti kaasinen matti.kaasi...@gmail.com wrote: I noticed that I sent my message below to wrong address. It happened about month ago, so I don't know if it's still valid. Have you tried latest maintenance branch ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel
Re: [Angstrom-devel] busybox-1.13.2 does not install hwclock
Hi! Just to inform that hwclock appeared after inserting couple of rtc configuration options to kernel. Those options include: CONFIG_RTC_CLASS=y CONFIG_RTC_HCTOSYS=y CONFIG_RTC_HCTOSYS_DEVICE=rtc0 CONFIG_RTC_INTF_SYSFS=y CONFIG_RTC_INTF_PROC=y CONFIG_RTC_INTF_DEV=y I don't know whether they all are needed. -Matti 11. tammikuuta 2012 21.17 matti kaasinen matti.kaasi...@gmail.com kirjoitti: Hi! Do anyone know how to configure busybox (busybox-1.13.2) for installing hwclock? Now, with default set-up of at91sam9260ek busybox-1.13.2/defconfig seems to have CONFIG_HWCLOCK=y CONFIG_FEATURE_HWCLOCK_LONG_OPTIONS=y CONFIG_FEATURE_HWCLOCK_ADJTIME_FHS=y It seems to produce binaries to work/armv5te-angstrom-linux-gnueabi/busybox-1.13.2-r42.4/busybox-1.13.2/util-linux. However, related to hwclock things, it seems to install only: hwclock.sh to /etc/init.d and hwclock-default to /etc/default/hwclock Is there some set-up I'm missing or could there be some kind of installation bug in busybox-1.13.2 recipe or 2011.03-maintenance branch? -Matti ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel
[Angstrom-devel] 2011.03-maintenance/fakeroot_1.14.5.bb broken
Hi! I noticed that I sent my message below to wrong address. It happened about month ago, so I don't know if it's still valid. It appears that fakeroot_1.14.5.bb used as fakeroot recipe for 2011.03-maintenance branch (or any branch) is broken. This has already been discussed in: http://permalink.gmane.org/gmane.comp.handhelds.openembedded/46830 As told in above link, http://snapshot.debian.net/archive/pool/f/fakeroot/fakeroot_1.14.5.orig.tar.bz2 is not there anymore. Alternative source from http://snapshot.debian.org/ that will be fetched by fakeroot_1.14.5.bb is corrupted. Checksums are wrong and in fact, archive does not open. Size of archive is far lower than the original one that I luckily do have in hands. So, I can manage with this problem as I can replace the corrupted source with original one. However, as this branch is called maintainance branch, are there any plans to fix this problem somehow, for instance adding fakeroot_1.15.1.bb recipe to 2011.03-maintenance? thanks, Matti ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel
Re: [Angstrom-devel] 2011.03-maintenance/fakeroot_1.14.5.bb broken
No, I have not ever since as I had that uncorrupted stuff available. Now it takes some time to get at the point to try as I have been playing with kernel. At this point there is one module that does not compile. I need to disable that before trying fakeroot. 13. tammikuuta 2012 20.46 Khem Raj raj.k...@gmail.com kirjoitti: On Fri, Jan 13, 2012 at 10:23 AM, matti kaasinen matti.kaasi...@gmail.com wrote: I noticed that I sent my message below to wrong address. It happened about month ago, so I don't know if it's still valid. Have you tried latest maintenance branch ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel
[Angstrom-devel] (no subject)
Hi! Do anyone know how to configure busybox (busybox-1.13.2) for installing hwclock? Now, with default set-up of at91sam9260ek busybox-1.13.2/defconfig seems to have CONFIG_HWCLOCK=y CONFIG_FEATURE_HWCLOCK_LONG_OPTIONS=y CONFIG_FEATURE_HWCLOCK_ADJTIME_FHS=y It seems to produce binaries to work/armv5te-angstrom-linux-gnueabi/busybox-1.13.2-r42.4/busybox-1.13.2/util-linux. However, related to hwclock things, it seems to install only: hwclock.sh to /etc/init.d and hwclock-default to /etc/default/hwclock Is there some set-up I'm missing or could there be some kind of installation bug in busybox-1.13.2 recipe or 2011.03-maintenance branch? -Matti ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel
[Angstrom-devel] busybox-1.13.2 does not install hwclock
Hi! Do anyone know how to configure busybox (busybox-1.13.2) for installing hwclock? Now, with default set-up of at91sam9260ek busybox-1.13.2/defconfig seems to have CONFIG_HWCLOCK=y CONFIG_FEATURE_HWCLOCK_LONG_OPTIONS=y CONFIG_FEATURE_HWCLOCK_ADJTIME_FHS=y It seems to produce binaries to work/armv5te-angstrom-linux-gnueabi/busybox-1.13.2-r42.4/busybox-1.13.2/util-linux. However, related to hwclock things, it seems to install only: hwclock.sh to /etc/init.d and hwclock-default to /etc/default/hwclock Is there some set-up I'm missing or could there be some kind of installation bug in busybox-1.13.2 recipe or 2011.03-maintenance branch? -Matti ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel
[Angstrom-devel] image packing bug?
Hi! I'm not quite sure if this is proper forum for my question, but as it somewhat smells a bug, I try first here. I was possibly faced with a bug related to image packing when bitbaking with following configuration: Build Configuration: BB_VERSION= 1.12.0 METADATA_BRANCH = 2011.03-maintenance METADATA_REVISION = f9c0bce TARGET_ARCH = arm TARGET_OS = linux-gnueabi MACHINE = at91sam9260ek DISTRO= angstrom DISTRO_VERSION= 2011.03 TARGET_FPU= soft One recipe - called recipe-aa_1.0.bb was picked from it's dev packate (recipe-aa-dev_1.0-r0.6_at91sam9260ek.ipk). It was tied in recipe-bb_1.0.bb with RDEPENDS_${PN} = recipe-aa udev Then recipe-bb_1.0.bb was included to image in initramfs-image.bbappend with statement: IMAGE_INSTALL_append = recipe-bb Anyone followed? Anyhow, everything else goes all right, but recipe-aa is always taken from wrong package (-dev). Could the reason for this be RDEPENDS_${PN} = recipe-aa udev statement as there is this 'udev'? Could bitbake use udev string to relate recipe-aa with recipe-aa-dev package? Is there any parameter I should use to force using recipe-aa(no -dev) version? Thanks, Matti ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel
Re: [Angstrom-devel] what is wrong with module_do_compile()?
Hi everyone who are interested how this proceeds! I think, I discovered the troublemaker. OK, there were many reasons, like DISTRO_VERSION= 2011.03 did not build properly set kernel. BUT, the reason why method 1 and 2 gave different results was that when do_compile of method passes list of variables with one variable, this process strips quotes away resulting: CC = arm-angstrom-linux-gnueabi-gcc -mno-thumb-interwork -mno-thumb when with method 2 CC = arm-angstrom-linux-gnueabi-gcc -mno-thumb-interwork -mno-thumb These printouts from makefile take place three times (method 2) when everything goes fine and nine times with method 1 (no compilation results at all and still compiler does not report errors). CC is always the same with method 2, but options are stripped with method 1 after the first printout. Makefile had following rule/command that assigns the CC value $(MODULES:%=%.ko): *.c *.h $(MAKE) -C $(KERNEL_SRC) ARCH=$(ARCH) SUBDIRS=$(CURDIR) CC=$(CC) modules Operation fixed with method 1 if I inserted quotes around CC parameter as follows: $(MODULES:%=%.ko): *.c *.h $(MAKE) -C $(KERNEL_SRC) ARCH=$(ARCH) SUBDIRS=$(CURDIR) CC=$(CC) modules To make this somewhat more confusing also following statement works with both methods: $(MODULES:%=%.ko): *.c *.h $(MAKE) -C $(KERNEL_SRC) ARCH=$(ARCH) SUBDIRS=$(CURDIR) modules Printouts work otherwise as explained before but now there are only three printouts with method 1 - 1: GCC having parameters, 23: GCC having no parameters (as described before). Still compilation is running fine! So, it seems that now as kernel is correctly (?) configured for this machine, compilation settings will be set automatically. GCC parameter just disturbed compilation. I hope someone could find this useful! -Matti 2011/8/18 matti kaasinen matti.kaasi...@gmail.com: Hi! I still have no proper solution! Some more information, though: KERNEL_SOURCE parameter name should be replaced with KERNEL_SRCIn order to get kmodule-test.tar.gz Makefile compatible with module_do_compile(). Methods for compilation have been written differently in module.bbclass and kmodule-test_1.0.bb. They seem to lead different CC, LD and AR values seen in Makefile. --- 1) module.bbclass convention: VAR = list of parameter definitions module_do_compile(){ unset list oe_runmake ${VAR} } 2) kmodule-test_1.0.bb convention: module_do_compile(){ unset list oe_runmake list of parameter definitions } -- 1) produces CC: arm-angstrom-linux-gnueabi-gcc -mno-thumb-interwork -mno-thumb LD: arm-angstrom-linux-gnueabi-ld --sysroot=/home/matti/projects/angstrom/setup-scripts/build/tmp-angstrom_2008_1/sysroots/armv5te-angstrom-linux-gnueabi AR: arm-angstrom-linux-gnueabi-ar 2) produces CC: arm-angstrom-linux-gnueabi-gcc LD: arm-angstrom-linux-gnueabi-ld -EL AR: arm-angstrom-linux-gnueabi-ar I also tried both of these conventions in kmodule-test_1.0.bb recipe and they acted also there like 1) or 2) respectively. With 1) compilation products are present in WORKDIR but compilation environment is not properly set (for instance library definitions are missing). For instance #include asm/io.h With 2) compilation seems going fine but compilation results are completely missing. Could anyone explain this or even better, give advices how I could get both environment set properly and results present in WORKDIR? Thanks, Matti 2011/8/14 matti kaasinen matti.kaasi...@gmail.com: Hi! It seems that module_do_compile() method loose compilation products (openembedded/2011.03-maintenance). Let's take http://www.kunen.org/uC/beagle/kmodule-test.tar.gz sources from http://www.kunen.org/uC/beagle/oe_kmod.html demo. Uncompressed sources work fine. If I comment do_compile out from kmodule-test_1.0.bb, module_do_compile should take care of compilation. It seems to do so. bitbake kmodule-test -c compile does not produce complaints However, plain: bitbake kmodule-test crashes, because kmodule_test.ko is not found. bitbake -D kmodule-test -c compile produces impressive log that would suggest that there should be kmodule_test.ko file in ${WORKDIR}, but it really does not exist there. Do I possible miss some setting? -Matti ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel
Re: [Angstrom-devel] what is wrong with module_do_compile()?
Hi Paul, Thanks for your reply. I try to stick in openembedded recipes also with kernel because I am moving from ad-hoc = everything manually environment to formal openembedded environment. True is that they need modification if platform changes. However, all the dependencies should be taken care of by the recipes. Also all the image building tasks should go smoothly - that is what I think... I tried to run the kernel module makefile manually as you suggested. Original makefile (for much older kernel) did not work only with updating variables. Rules did not work. After modifications to more close match to makefile presented in http://www.kunen.org/uC/beagle/kmodule-test.tar.gz started make procedure. However, still the compilation did not work correctly. I quess, this toolchain is somehow corrupted now. I'll try re-building it. thanks, Matti 2011/8/21 Paul Parsons lost.dista...@yahoo.com: Hi Matti, Not sure how useful this will be, but anyway... When I started to build custom kernels for my Angstrom hx4700 distribution I very quickly decided that it would be easier to build them by hand, directly from the command line. So why not do the same? All you need are the following: 1. The kernel source. 2. The kernel config file. 3. The ARM toolchain, which needs to be in your shell's search path. Then just build the kernel/drivers as you would for any Linux system, but with a few extra arguments to cross-compile. So for example, instead of typing make oldconfig, you would type something like: make ARCH=arm CROSS_COMPILE=arm-angstrom-linux-gnueabi- oldconfig No more fighting the Angstrom build scripts. Heresy perhaps, but it works for me. Regards, Paul ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel
Re: [Angstrom-devel] what is wrong with module_do_compile()?
Hi! I still have no proper solution! Some more information, though: KERNEL_SOURCE parameter name should be replaced with KERNEL_SRCIn order to get kmodule-test.tar.gz Makefile compatible with module_do_compile(). Methods for compilation have been written differently in module.bbclass and kmodule-test_1.0.bb. They seem to lead different CC, LD and AR values seen in Makefile. --- 1) module.bbclass convention: VAR = list of parameter definitions module_do_compile(){ unset list oe_runmake ${VAR} } 2) kmodule-test_1.0.bb convention: module_do_compile(){ unset list oe_runmake list of parameter definitions } -- 1) produces CC: arm-angstrom-linux-gnueabi-gcc -mno-thumb-interwork -mno-thumb LD: arm-angstrom-linux-gnueabi-ld --sysroot=/home/matti/projects/angstrom/setup-scripts/build/tmp-angstrom_2008_1/sysroots/armv5te-angstrom-linux-gnueabi AR: arm-angstrom-linux-gnueabi-ar 2) produces CC: arm-angstrom-linux-gnueabi-gcc LD: arm-angstrom-linux-gnueabi-ld -EL AR: arm-angstrom-linux-gnueabi-ar I also tried both of these conventions in kmodule-test_1.0.bb recipe and they acted also there like 1) or 2) respectively. With 1) compilation products are present in WORKDIR but compilation environment is not properly set (for instance library definitions are missing). For instance #include asm/io.h With 2) compilation seems going fine but compilation results are completely missing. Could anyone explain this or even better, give advices how I could get both environment set properly and results present in WORKDIR? Thanks, Matti 2011/8/14 matti kaasinen matti.kaasi...@gmail.com: Hi! It seems that module_do_compile() method loose compilation products (openembedded/2011.03-maintenance). Let's take http://www.kunen.org/uC/beagle/kmodule-test.tar.gz sources from http://www.kunen.org/uC/beagle/oe_kmod.html demo. Uncompressed sources work fine. If I comment do_compile out from kmodule-test_1.0.bb, module_do_compile should take care of compilation. It seems to do so. bitbake kmodule-test -c compile does not produce complaints However, plain: bitbake kmodule-test crashes, because kmodule_test.ko is not found. bitbake -D kmodule-test -c compile produces impressive log that would suggest that there should be kmodule_test.ko file in ${WORKDIR}, but it really does not exist there. Do I possible miss some setting? -Matti ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel
[Angstrom-devel] what is wrong with module_do_compile()?
Hi! It seems that module_do_compile() method loose compilation products (openembedded/2011.03-maintenance). Let's take http://www.kunen.org/uC/beagle/kmodule-test.tar.gz sources from http://www.kunen.org/uC/beagle/oe_kmod.html demo. Uncompressed sources work fine. If I comment do_compile out from kmodule-test_1.0.bb, module_do_compile should take care of compilation. It seems to do so. bitbake kmodule-test -c compile does not produce complaints However, plain: bitbake kmodule-test crashes, because kmodule_test.ko is not found. bitbake -D kmodule-test -c compile produces impressive log that would suggest that there should be kmodule_test.ko file in ${WORKDIR}, but it really does not exist there. Do I possible miss some setting? -Matti ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel
[Angstrom-devel] do anyone know how bblayers should work?
Hi OE developers! I like the idea I've learned bblayers to work. I tried it first with openembedded:2011.03-maintenance - bitbake:1.12 combination the setup-script provides. Otherwise this combination works, but it seems to me that it ignores totally bblayers.conf. Is this true or do I possibly have some kind of set-up problem? I have also tried openembedded:master - bitbake master combination. It seems to follow bblayers.conf, but unfortunately toolchain building crashes while tying to make linux native. I also tried openembedded:2011.03-maintenance - bitbake:master combination, but it crashes in very early stages. I suppose these two are not compatible or could they be made working with some configuraiton tricks? regards, Matti ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel
Re: [Angstrom-devel] do anyone know how bblayers should work?
Hi Maksym, and thanks a lot. I'll try it in that way. Funny that instructions told to use ~/setup-scripts as building location. -Matti 2011/8/11 Maksym Parkachov lazy.gop...@gmail.com: Hi Matti, if you used the setup-scripts, then bblayers.conf will be used only if you start you build from the build directory. So, you should be in ~/setup-scripts/build and your bblayers.conf should be in conf/ folder ~/setup-scripts/build$ ls conf bblayers.conf local.conf Then, bitbake will pick it up and it works without problem. I tested it on openembedded:2011.03-maintenance - bitbake:1.12 combination. Regards, Maksym. On Thu, Aug 11, 2011 at 11:13 AM, matti kaasinen matti.kaasi...@gmail.com wrote: Hi OE developers! I like the idea I've learned bblayers to work. I tried it first with openembedded:2011.03-maintenance - bitbake:1.12 combination the setup-script provides. Otherwise this combination works, but it seems to me that it ignores totally bblayers.conf. Is this true or do I possibly have some kind of set-up problem? I have also tried openembedded:master - bitbake master combination. It seems to follow bblayers.conf, but unfortunately toolchain building crashes while tying to make linux native. I also tried openembedded:2011.03-maintenance - bitbake:master combination, but it crashes in very early stages. I suppose these two are not compatible or could they be made working with some configuraiton tricks? regards, Matti ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel
Re: [Angstrom-devel] build directory
Hi Maksym, I did some more investigations. I would say bitbake (branch 1.2) that is checked out by setup-script (anyway two weeks ago) does not care about bblayers.conf and layer.conf. You can have anything at all in these files and you won't see any messages logged regarding these files. Also this version of bitbake requires you to point bitbake.conf with BBPATH in the beginning. Otherwice bitbake will complain that bitbake.conf can't be found from directories listed in BBPATH. However, this is different with bitbake Master branch. You need only point the directory where conf directory for bblayers.conf file lies. By using that bitbake can parse layer.conf files and can find that bitbake.conf file that lies in oe layer. Also, if I make errors in layer.conf files pointed by bblayers.conf file, these errors will be logged. I can't say this is quite sure as the set-up that I have for investigating this is broken - bitbake master - openembedded master combination seems to crash while creating linux native (gettext-native-0.18-r6). No-one seems tho care about that, though. I could still try bitbake/master - openembedded/2011.03-maintenance combination as 2011.03-maintenance does not crash with this gettext (actually it produces -r5) but I'm now playing over gsm modem connection, so I'll postpone that until I have proper connection. It might be good if someone more developer grade person could also comment these observations. -Matti 2011/7/26 Maksym Parkachov lazy.gop...@gmail.com: On Mon, Jul 25, 2011 at 8:02 PM, matti kaasinen matti.kaasi...@gmail.com wrote: Hii Maksym, I have to draw my answer back. I tried dirrent combinations of DISTRIBUTIONs and bitbake and angstrom revisions. Now I could not find any references from logs regarding to layers - e.g. to bblayers.conf file (bblayers.conf found or resembling). This was clearly reported when I had bblayers.conf with corrupted characters. -Matti Hi Matti, thanks for the answer. As I said in the e-mail, it does work, but only when compiling from build directory. I indeed used the same instructions and had to fix quotation as well. I'll double check it once more. It's not like it's a big issue, build works, but it took me some time to realize it (as you could imagine, it was quite sporadic, sometime works, sometime not). And I still feel, that I just don't have some variable set correctly (need to look at bitbake source code). Cheers, Maksym. ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel
Re: [Angstrom-devel] package gettext-native-0.18-r6: task do_configure: - libtoolize missing
Hi everyone, I tried this again after pulling both bitbake and openembedded master branches. It still crashes. As it happens when building native part of toolchain, I suppose this should happen with every MACHINE, not just this at91sma9260ek. Doesn't any other hit this problem? Settings after pulling for bitbake console-image are: Build Configuration: BB_VERSION= 1.13.3 METADATA_BRANCH = master METADATA_REVISION = 6303444 TARGET_ARCH = arm TARGET_OS = linux-gnueabi MACHINE = at91sam9260ek DISTRO= angstrom DISTRO_VERSION= v2011.07 TARGET_FPU= soft -Matti 2011/7/25 matti kaasinen matti.kaasi...@gmail.com: Hi everyone! I tried to send this before. I suppose, it must have before I got the registration mail back. So, please excuse, if this message is repeated. -Matti 2011/7/23 matti kaasinen matti.kaasi...@gmail.com: Hi I tried to run console-image for at91sam9260ek with following configuration: Build Configuration: BB_VERSION = 1.13.2 METADATA_BRANCH = master METADATA_REVISION = e07647d TARGET_ARCH = arm TARGET_OS = linux-gnueabi MACHINE = at91sam9260ek DISTRO = angstrom DISTRO_VERSION = v2011.07 TARGET_FPU = soft So, both the bitbake and openembedded are on master branch. Build crashes while executing: virtual:native:/home/matti/projects/angstrom/setup-scripts/sources/openembedded/recipes/gettext/gettext_0.18.bb following error log was available: | autoreconf: configure.ac: adding subdirectory libasprintf to autoreconf | autoreconf: Entering directory `libasprintf' | autoreconf: running: aclocal -I /home/matti/projects/angstrom/setup-scripts/build/tmp-angstrom_2010_x/work/i686-linux/gettext-native-0.18-r6/gettext-0.18/gnulib-local/m4/ -I /home/matti/projects/angstrom/setup-scripts/build/tmp-angstrom_2010_x/work/i686-linux/gettext-native-0.18-r6/gettext-0.18/gettext-runtime/m4 -I /home/matti/projects/angstrom/setup-scripts/build/tmp-angstrom_2010_x/work/i686-linux/gettext-native-0.18-r6/gettext-0.18/gettext-tools/m4 -I /home/matti/projects/angstrom/setup-scripts/build/tmp-angstrom_2010_x/sysroots/i686-linux/usr/share/aclocal-1.11 -I /home/matti/projects/angstrom/setup-scripts/build/tmp-angstrom_2010_x/sysroots/i686-linux/usr/share/aclocal -I /home/matti/projects/angstrom/setup-scripts/build/tmp-angstrom_2010_x/work/i686-linux/gettext-native-0.18-r6/gettext-0.18/gnulib-local/m4/ -I /home/matti/projects/angstrom/setup-scripts/build/tmp-angstrom_2010_x/work/i686-linux/gettext-native-0.18-r6/gettext-0.18/gettext-runtime/m4 -I /home/matti/projects/angstrom/setup-scripts/build/tmp-angstrom_2010_x/work/i686-linux/gettext-native-0.18-r6/gettext-0.18/gettext-tools/m4 -I /home/matti/projects/angstrom/setup-scripts/build/tmp-angstrom_2010_x/sysroots/i686-linux/usr/share/aclocal-1.11 -I /home/matti/projects/angstrom/setup-scripts/build/tmp-angstrom_2010_x/sysroots/i686-linux/usr/share/aclocal --force -I ../../m4 -I ../m4 -I gnulib-m4 | autoreconf: running: libtoolize --copy --force | autoreconf: failed to run libtoolize: No such file or directory | autoreconf: libtoolize is needed because this package uses Libtool | + oefatal 'autoreconf execution failed.' | + echo FATAL: 'autoreconf execution failed.' How to proceed? -Matti ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel
Re: [Angstrom-devel] package gettext-native-0.18-r6: task do_configure: - libtoolize missing
Hi everyone! I tried to send this before. I suppose, it must have before I got the registration mail back. So, please excuse, if this message is repeated. -Matti 2011/7/23 matti kaasinen matti.kaasi...@gmail.com: Hi I tried to run console-image for at91sam9260ek with following configuration: Build Configuration: BB_VERSION = 1.13.2 METADATA_BRANCH = master METADATA_REVISION = e07647d TARGET_ARCH = arm TARGET_OS = linux-gnueabi MACHINE = at91sam9260ek DISTRO = angstrom DISTRO_VERSION = v2011.07 TARGET_FPU = soft So, both the bitbake and openembedded are on master branch. Build crashes while executing: virtual:native:/home/matti/projects/angstrom/setup-scripts/sources/openembedded/recipes/gettext/gettext_0.18.bb following error log was available: | autoreconf: configure.ac: adding subdirectory libasprintf to autoreconf | autoreconf: Entering directory `libasprintf' | autoreconf: running: aclocal -I /home/matti/projects/angstrom/setup-scripts/build/tmp-angstrom_2010_x/work/i686-linux/gettext-native-0.18-r6/gettext-0.18/gnulib-local/m4/ -I /home/matti/projects/angstrom/setup-scripts/build/tmp-angstrom_2010_x/work/i686-linux/gettext-native-0.18-r6/gettext-0.18/gettext-runtime/m4 -I /home/matti/projects/angstrom/setup-scripts/build/tmp-angstrom_2010_x/work/i686-linux/gettext-native-0.18-r6/gettext-0.18/gettext-tools/m4 -I /home/matti/projects/angstrom/setup-scripts/build/tmp-angstrom_2010_x/sysroots/i686-linux/usr/share/aclocal-1.11 -I /home/matti/projects/angstrom/setup-scripts/build/tmp-angstrom_2010_x/sysroots/i686-linux/usr/share/aclocal -I /home/matti/projects/angstrom/setup-scripts/build/tmp-angstrom_2010_x/work/i686-linux/gettext-native-0.18-r6/gettext-0.18/gnulib-local/m4/ -I /home/matti/projects/angstrom/setup-scripts/build/tmp-angstrom_2010_x/work/i686-linux/gettext-native-0.18-r6/gettext-0.18/gettext-runtime/m4 -I /home/matti/projects/angstrom/setup-scripts/build/tmp-angstrom_2010_x/work/i686-linux/gettext-native-0.18-r6/gettext-0.18/gettext-tools/m4 -I /home/matti/projects/angstrom/setup-scripts/build/tmp-angstrom_2010_x/sysroots/i686-linux/usr/share/aclocal-1.11 -I /home/matti/projects/angstrom/setup-scripts/build/tmp-angstrom_2010_x/sysroots/i686-linux/usr/share/aclocal --force -I ../../m4 -I ../m4 -I gnulib-m4 | autoreconf: running: libtoolize --copy --force | autoreconf: failed to run libtoolize: No such file or directory | autoreconf: libtoolize is needed because this package uses Libtool | + oefatal 'autoreconf execution failed.' | + echo FATAL: 'autoreconf execution failed.' How to proceed? -Matti ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel
Re: [Angstrom-devel] build directory
Hi Maksym, I have experienced the same by using this ~/.oe/environment-2008. It sets DISTRO = angstrom_2008.1. I assume 2008.1 distro ignoress bblayers.conf. It seems that angstrom_2010.x grants it. However, if you have read/used http://sakrah.homelinux.org/blog/2010/11/bblayers-bbappend/, be careful if you copy those bblayers.conf and layers.conf settings for you usage, spaces hold unvisible characters that corrupt your settings. Also that TOPDIR settigs has wrong quotes around FILE. -Matti 2011/7/25 Maksym Parkachov lazy.gop...@gmail.com: Hi developers, I'm using your setup-scripts and providing my own layer. I'm building my image with bitbake custom-image. Everything works fine, when building the image from setup-scripts/build directory or using oebb.sh bitbake custom-image (which just does cd build before building). But trying to build from some other directory gives me an error, which basically means, that layer is not found. Of course, I'm sourcing ~/.oe/environment-2008 before. Is this how it's intended to be ? Cheers, Maksym. ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel
Re: [Angstrom-devel] build directory
Hii Maksym, I have to draw my answer back. I tried dirrent combinations of DISTRIBUTIONs and bitbake and angstrom revisions. Now I could not find any references from logs regarding to layers - e.g. to bblayers.conf file (bblayers.conf found or resembling). This was clearly reported when I had bblayers.conf with corrupted characters. -Matti 2011/7/25 matti kaasinen matti.kaasi...@gmail.com: Hi Maksym, I have experienced the same by using this ~/.oe/environment-2008. It sets DISTRO = angstrom_2008.1. I assume 2008.1 distro ignoress bblayers.conf. It seems that angstrom_2010.x grants it. However, if you have read/used http://sakrah.homelinux.org/blog/2010/11/bblayers-bbappend/, be careful if you copy those bblayers.conf and layers.conf settings for you usage, spaces hold unvisible characters that corrupt your settings. Also that TOPDIR settigs has wrong quotes around FILE. -Matti 2011/7/25 Maksym Parkachov lazy.gop...@gmail.com: Hi developers, I'm using your setup-scripts and providing my own layer. I'm building my image with bitbake custom-image. Everything works fine, when building the image from setup-scripts/build directory or using oebb.sh bitbake custom-image (which just does cd build before building). But trying to build from some other directory gives me an error, which basically means, that layer is not found. Of course, I'm sourcing ~/.oe/environment-2008 before. Is this how it's intended to be ? Cheers, Maksym. ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel ___ Angstrom-distro-devel mailing list Angstrom-distro-devel@linuxtogo.org http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/angstrom-distro-devel