1# OpenBMC kernel development 2 3The OpenBMC project maintains a kernel tree for use by the project. The tree's general development policy is that code must be upstream first. This is a strong requirement, not a hard requirement, and exceptions will be made on a case-by-case basis. If in doubt, start a discussion on the mailing list. 4 5The OpenBMC kernel tree is hosted at https://github.com/openbmc/linux and contains the set of patches that we carry. Ideally there would be no patches carried, as everything should be upstream. 6 7Your code will make it into the OpenBMC tree in these ways, from most to least desirable: 8 91. When the OpenBMC kernel moves to a new upstream release 102. By backporting upstream commits from a newer kernel version to the OpenBMC kernel 113. Patches included in the OpenBMC tree temporarily 12 13## TL;DR 14 15If you require a patch added to the tree, follow these steps: 16 171. Submit your patch upstream. It doesn't need to be upstream, but it should be on it's way 182. Use ```git format-patch --subject-prefix="PATCH linux dev-4.7" --to=openbmc@lists.ozlabs.org --to=joel@jms.id.au``` to create a formatted patch 19 20## Developing a new driver 21 22When developing a new driver, your goal is to have the code accepted upstream. The first step should be to check that there is no existing driver for the hardware you wish to support. Check the OpenBMC `-dev` tree, check upstream, and if you do not find support there ask on the mailing list. 23 24Once you are sure a driver needs to be written, you should develop and test the driver, before sending it upstream to the relevant maintainers. You should feel welcome to cc the OpenBMC list when sending upstream, so other kernel developers can provide input where appropriate. Be sure to follow the (upstream development process)[https://www.kernel.org/doc/Documentation/process/submitting-patches.rst]. 25 26In the past patches underwent 'pre-review' on the OpenBMC mailing list. While this is useful for developers who are not familiar with writing kernel code, it has lead to confusion about the upstreaming process, so now we do all of our development in the community. 27 28Once the driver has been accepted upstream, send the good news to the OpenBMC list with a reference to the upstream tree. This may be Linus' tree, or it might be one of the subsystem maintainers. From there the OpenBMC kernel team can decide how best to include your code in the OpenBMC tree. 29 30### Exceptions 31 32There are cases where waiting for upstream acceptance will delay the bring-up of a new system. This should be avoided through careful planning and early development of the features upstream, but where this has not happened we can chose to carry the patches in the OpenBMC tree while the upstream development is ongoing. 33 34Another exception to the upstream first rule is where patches are modifying files that are not upstream. This currently includes the aspeed board file `arch/arm/mach-aspeed/aspeed.c`, and the device tree source files `dts`. The board file should go away when we get drivers written for all of the functionality; for now it contains some hacks relating to LPC and early init. 35 36If you find yourself adding to `arch/arm/mach-aspeed/aspeed.c`, first send an email to the OpenBMC list to get the opinion of the kernel developers. Patches to `aspeed.c` will be treated with some prejudice as the file will be removed once we have drivers for all of the Aspeed peripherals. 37 38## Getting existing code in the tree 39 40The OpenBMC kernel is currently based on the 4.7 series. If there is upstream code you would like backported, send it to the list. Be sure to include the upstream commit SHA in the commit message. 41 42## Testing 43 44When modifying the tree we currently test on the following platforms: 45 46 - Palmetto, an OpenPower Power8 box containing an ast2400 with NCSI networking 47 - ast2500-evb, the Aspeed dev board with two PHYs 48 - Witherspoon, an OpenPower Power9 box containing an ast2500 with NCSI networking 49 - qemu-plametto and qemu-romulus 50 51Before submitting patches it is recommended you boot test on at least the Qemu platforms, and whatever hardware you have available. 52 53# Tips and Tricks 54 55Some general hints for kernel development 56 57## Out-of-tree builds 58 59You can build a kernel out of the yocto environment, by using the initramfs 60(from a pre-existing yocto build) directly: 61 62``` 63make ARCH=arm \ 64 O=obj \ 65 CROSS_COMPILE=arm-linux-gnueabihf- \ 66 CONFIG_INITRAMFS_SOURCE=/path/tp/obmc-phosphor-image-palmetto.cpio.gz 67``` 68 69(adjust `O` and `CROSS_COMPILE` parameters as appropriate). 70 71You'll need to use `aspeed_g4_defconfig` or `aspeed_g5_defconfig` as your base 72kernel configuration. 73 74The cpio can be found in the following yocto output directory: 75 76``` 77 build/tmp/deploy/images/palmetto/ 78``` 79 80## Building a uImage 81 82To build a uImage (for example, to netboot): 83 84``` 85# build a zImage using the obmc rootfs 86make ARCH=arm \ 87 O=obj \ 88 CROSS_COMPILE=arm-linux-gnueabihf- \ 89 CONFIG_INITRAMFS_SOURCE=/path/tp/obmc-phosphor-image-palmetto.cpio.gz 90 91# create a combined zImage + DTB image 92cat obj/arch/arm/boot/zImage \ 93 obj/arch/arm/boot/dts/aspeed-bmc-opp-palmetto.dtb \ 94 > obj/aspeed-zimage 95 96# create a uImage 97./scripts/mkuboot.sh -A arm -O linux -C none -T kernel \ 98 -a 0x40008000 -e 0x40008000 -n $USER-`date +%Y%m%d%H%M` \ 99 -d obj/aspeed-zimage obj/uImage 100``` 101 102Note that some systems may have upgraded to a FIT-based u-boot, where the old 103uImage format is no longer accepted. 104