aboutsummaryrefslogtreecommitdiffstats
path: root/README.md
diff options
context:
space:
mode:
authorSzabolcs Hubai <szab.hu@gmail.com>2022-09-26 08:42:54 +0200
committerHauke Mehrtens <hauke@hauke-m.de>2022-11-12 18:15:55 +0100
commite7ad68d682bdb73b7d13d6c4b8d1d65d9f050138 (patch)
tree4907cd885cda8b1b91766d984615d987a0f2ccb7 /README.md
parentffa4b5283b390099a207dcd22c6e9b0b528ac1e4 (diff)
downloadupstream-e7ad68d682bdb73b7d13d6c4b8d1d65d9f050138.tar.gz
upstream-e7ad68d682bdb73b7d13d6c4b8d1d65d9f050138.tar.bz2
upstream-e7ad68d682bdb73b7d13d6c4b8d1d65d9f050138.zip
ramips: define lzma-loader recipe for SEAMA devices
Define "Device/seama-lzma-loader" recipe for SEAMA devices to help contributors avoid doing recipe mistakes. In a forum topic [0] I was under the impression that the good old uimage-lzma-loader didn't fix the LZMA ERROR 1 for a device. It was found out, that the uimage-lzma-loader never worked because the KERNEL variable was overriden earlier (also an LZMA ERROR 1 related commit, 6fba88de1913), and the "use lzma-loader" fix (commit 09faa73c53bd) didn't catch that to include the "loader-kernel" part. I contributed an LZMA ERROR 1 fix (commit ce1957100411) for the SEAMA device D-Link DIR-860L B1, where I had to duplicate the whole uimage-lzma-loader recipe because of the special needs of the vendor bootloader. This new recipe reuse most of uimage-lzma-loader's KERNEL definiton to avoid duplication. It uses "relocate-kernel" as it needed for D-Link DIR-860L B1 to boot from flash, and it's compatible with D-Link DIR-645 too. It repacks lzma-loader with lzma for kernel (without uImage), because these weird hacked vendor bootloaders accepts only LZMA compressed kernels from flash: We have SEAMA, Image Size = 4759794 Verifying Checksum ... Uncompressing SEAMA linux.lzma ... OK It uses uImage header for initramfs kernel to be little bit verbose. 0: https://forum.openwrt.org/t/136435/10 Signed-off-by: Szabolcs Hubai <szab.hu@gmail.com>
Diffstat (limited to 'README.md')
0 files changed, 0 insertions, 0 deletions