summaryrefslogtreecommitdiff
path: root/classes/kernel.bbclass
AgeCommit message (Collapse)AuthorFiles
2009-05-07kernel.bbclass: also stage include/trace if it's availableDmitry Eremin-Solenikov1
Signed-off-by: Dmitry Eremin-Solenikov <dbaryshkov@gmail.com>
2009-04-06kernel,module-base class, bitbake.conf: introduce MACHINE_KERNEL_PRKoen Kooi1
* bandaid till we have a better way to track ABI changes
2009-03-24kernel.bbclass: stage more filesKoen Kooi1
2009-03-12Kernel bbclass "too many open files" bugfixJader H. Silva1
tempfile.mkstemp() is used to create a temporary file. This function return a tuple with an OS file descriptor and a filename. Filename is stored in "tmpfile" but descriptor is not stored anywhere, but it is still open because it's only an integer to python so it is not closed at the end of the function. For each iteration in which this function is called, a new OS file descriptor is opened, but not closed. The solution is to store the file descriptor and close it.
2009-01-30kernel.bbclass: move initramfs stuff to configure so we can do ↵Koen Kooi1
postprocessing on it with do_configure_append
2009-01-25kernel.bbclass: Remove some repetition in last commit.Tim 'timtim' Ellis1
2009-01-25kernel.bbclass: Add missing entry for scsi headers in staging, as needed by ↵Tim 'timtim' Ellis1
iscsi-target Signed-off by Koen Kooi <koen@openembedded.org>
2009-01-12kernel.bbclass: do not use insecure tmpnam() function for generating unique ↵Marcin Juszkiewicz1
names
2008-11-23kernel.bbclass: stage more headers for the .27 header moveKoen Kooi1
2008-11-15kernel.bbclass: stage more headers for >=2.6.27 kernelsKoen Kooi1
2008-10-30kernel bbclass: start adding support for >=2.6.27 where headers movedKoen Kooi1
* still lack a symlink for mach/, dunno how to do that properly since the information is dynamically generated with makefiles
2008-09-26kernel.bbclass: prepare for 2.6.27 style external firmwareKoen Kooi1
* see http://lwn.net/Articles/284932/ * does *NOT* have dependency code, RRECOMMENDS for firmwares need to be added manually
2008-08-13kernel.bbclass: make aes-generic module provide aes Thomas Kunze1
* the kernel guys renamed the aes module wifi with wpa should work again
2008-08-12kernel.bbclass: fix typo in do_sizecheck()Thomas Kunze1
2008-08-03kernel.bbclass: use KERNEL_OUTPUT where possibleTom Rini1
2008-07-27kernel.bbclass: -change initramfs-logicThomas Kunze1
-add parameter to do_sizecheck do make it optional
2008-05-31classes/kernen.bbclass: Fix INITRAMFS logic to stop breaking builds for ↵Richard Purdie1
people not using it
2008-05-30kernel.bbclass: get initramfs stuff workingThomas Kunze1
To use this you have to INITRAMFS_IMAGE to the desired image. In addition the you must have cpio.gz in IMAGE_FSTYPES (at least for that image see packages/images/initramfs-kexec-image.bb for an example)
2008-05-25kernel bbclass: stage vmlinux if presentKoen Kooi1
2008-05-22kernel.bbclass: comment initramfs stuff till rootfs locking works properly ↵Koen Kooi1
(initramfs should be optional as well)
2008-04-30kernel bbclass: ${D} doesn't always have a trailing /, so put one in, sh is ↵Koen Kooi1
flexible enough to take care of '//' inside paths
2008-04-28classes/kernel.bbclass : fix for situation in which we generate no modules.Graeme Gregory1
2008-04-15kernel bbclass: have modules.tgz have the same version and packages and ↵Koen Kooi1
deployed images (PV) instead of pulling it out of the makefile (KERNEL_VERSION) * before: 836K modules-2.6.25-rc8-davinci1-r1-davinci-sffsdr.tgz 1.4M uImage-2.6.24+2.6.25-rc8-r1-davinci-sffsdr.bin * after: 836K modules-2.6.24+2.6.25-rc8-r1-davinci-sffsdr.tgz 1.4M uImage-2.6.24+2.6.25-rc8-r1-davinci-sffsdr.bin
2008-03-31kernel.bbclass,linux.inc: cleanup definition of ${KERNEL_IMAGE_BASE_NAME}Jeremy Laine1
* drop ${DATETIME} from ${KERNEL_IMAGE_BASE_NAME} in kernel.bbclass, see: http://lists.linuxtogo.org/pipermail/openembedded-devel/2008-March/005017.html * make ${KERNEL_IMAGE_BASE_NAME} and ${KERNEL_IMAGE_SYMLINK_NAME} ?=, see: http://lists.linuxtogo.org/pipermail/openembedded-devel/2008-March/005016.html * remove ${KERNEL_IMAGE_BASE_NAME} and ${KERNEL_IMAGE_SYMLINK_NAME} from linux.inc * use ${KERNEL_IMAGE_BASE_NAME} for uImage deploy
2008-03-28kernel.bbclass: fix generation of uImage on powerpc platformsJeremy Laine1
* add a map_uboot_arch function to kernel.bbclass which changes "powerpc" to "ppc" and export UBOOT_ARCH * pass -A ${UBOOT_ARCH} instead of -A ${ARCH} to mkimage in kernel.bbclass
2008-03-27kernel.bbclass,linux.inc: move dependency on u-boot-openmoko-native to ↵Jeremy Laine1
kernel.bbclass
2008-03-27kernel.bbclass: unbreak uImage generation.Koen Kooi1
2008-03-27kernel.bbclass: add generic do_deploy taskThomas Kunze1
2008-02-21kernel.bbclass: Fis typo in do_builtin_initramfs.Paul Sokolovsky1
2008-02-19classes/kernel.bbclass: fix staging for x86 on kernels >= 2.6.24Jeremy Laine1
2008-02-17classes/kernel.bbclass: Remove usage of non-standard variables (DEPLOY_TO) ↵Richard Purdie1
and random renaming of output.
2008-02-15kernel.bbclass: Add satndalone task "builtin_initramfs".Paul Sokolovsky1
* Builds a zImage with linked in initramfs specified by INITRAMFS_LOC param.
2008-01-24kernel: add wifi suspend fix to kernel-module-(hostap|orinoco)-cs. Closes 3664.Rolf Leggewie1
* applies only to kernel later than 2.6.17 * added base_version_less_or_equal function to base.bbclass (thanks, zecke and mickey!) to do the comparison. This does not change existing functionality. Skipped RFC.
2008-01-21kernel.bbclass: Account for sa1100 rtc renaming (#3710)Richard Purdie1
2008-01-18kernel.bbclass: added sizecheck task from linux recipe (merge from Poky)Marcin Juszkiewicz1
2008-01-18kernel.bbclass: stage also include/video directory (from Poky)Marcin Juszkiewicz1
2007-12-07kernel.bbclass: When processing module_autoload's, also try canonical name ↵Paul Sokolovsky1
(with dashes). * To remind, modprobe accepts both '-' & '_' in module names interchangeably. Actual module files use both. Still, in OE, some consistency is needed. And its natural to use dash as the separator, as package names already use that. So, if canonical naming is used, it all will "just work", refardless of what separator chars specific kernel module filename randomly contains.
2007-12-07kernel.bbclass: Add kernel-vmlinux to kernel-modules dependency blacklist.Paul Sokolovsky1
* This cuts x11-gpe-image by 3M for example.
2007-09-29kernel.bbclass: Add menuconfig task (from poky)Richard Purdie1
2007-09-29kernel.bbclass: Package vmlinux file (used by oprofile) (from poky)Richard Purdie1
2007-08-29kernel.bbclass: Allow machines to override the kernel-base runtime dependenciesRichard Purdie1
2007-07-25kernel.bbclass: Add support for module config per 2.6 conevntions.Paul Sokolovsky1
* 2.6 modprobe expects per-module configuration in /etc/modprobe.d/ . * Closes #2669.
2007-07-05kernel.bbclass: Oh ok, we need both DEPENDS and RDEPENDS.Paul Sokolovsky1
2007-07-05kernel.bbclass: DEPEND on bluez-dtl1-workaround, not just RDEPEND.Paul Sokolovsky1
2007-07-04bluez: Stop 02dtl1_cs.sh the nasty hack from breaking everyone's BT connection.Paul Sokolovsky1
* This sloppily written script just unconditionally breaks a BT connection on suspend. * So, instead, make only DTL1 victims suffer it thru: 1. Make a separate package for it. 2. Make kernel-module-dtl1-cs and nothing else RDEPENDS on it.
2007-04-09kernel.bbclass: Add kernel-base package which kernel modules depend on ↵Richard Purdie1
instead of kernel-image. This allows machines to specify that the kernel-image shouldn't be included in a default rootfs, without needing kernel-image to be empty. Set RDEPENDS_kernel-base_MACHINE = for machines that need this as per a discussion on the mailing list.
2007-02-14kernel.bbclass: Remove PARALLEL_INSTALL_MODULES. If we want to do this we ↵Richard Purdie1
should rethink and have something that works for multiple kernel versions generically not just a 2.4/2.6 split. There are certainly better ways to do it than this.
2007-01-07kernel.bbclass: Remove the KERNEL_RELEASE variable and set it to ↵Richard Purdie1
KERNEL_VERSION. This should removes a large source of kernel module problems as the variables then 'just work'
2007-01-06kernel.bbclass: Create relative symlinks from "zImage" to actual kernel ↵Paul Sokolovsky1
image in use. * Make sure update-alternatives created relative symlinks in /boot (or whatever) kernel image directory is used. This is required for tools like LAB (Linux As Bootloader) which may mount rootfs over some other temporary fs tree to get access to kernel to boot. * Closes #1569.
2006-11-20kernel.bbclass: unset MACHINE otherwise things break (from poky)Richard Purdie1