------- Comment on attachment From <email address hidden> 2019-04-12 05:43 EDT-------
Hi Manoj, thank you. I attached a tarball with patches - in addition to that one it looked best to add these in:
8cf4c05712f0 powerpc/lib/code-patching: refactor patch_instruction()
8183d99f4a22 powerpc/lib/feature-fixups: use raw_patch_instruction()
51c3c62b58b3 powerpc: Avoid code patching freed init sections
b45ba4a51cde powerpc/lib: fix book3s/32 boot failure due to code patching
The first two are just pre-reqs to keep it cleaner, but since 37bc3e5fd764 is in that seems the right thing to do. The last mostly fixes an error on 32 bit ppc kernels, which aren't supported, but this keeps it closer to upstream in the event of needing some further changes and cleans it up.
------- Comment on attachment From <email address hidden> 2019-04-12 05:43 EDT-------
Hi Manoj, thank you. I attached a tarball with patches - in addition to that one it looked best to add these in: lib/code- patching: refactor patch_instruction() lib/feature- fixups: use raw_patch_ instruction( )
8cf4c05712f0 powerpc/
8183d99f4a22 powerpc/
51c3c62b58b3 powerpc: Avoid code patching freed init sections
b45ba4a51cde powerpc/lib: fix book3s/32 boot failure due to code patching
The first two are just pre-reqs to keep it cleaner, but since 37bc3e5fd764 is in that seems the right thing to do. The last mostly fixes an error on 32 bit ppc kernels, which aren't supported, but this keeps it closer to upstream in the event of needing some further changes and cleans it up.
I tested this with: wspoon5: /home/mranweil# echo 0 > /sys/kernel/ debug/powerpc/ barrier_ nospec wspoon5: /home/mranweil# dmesg |grep -i skip wspoon5: /home/mranweil#
root@ltc-
root@ltc-
[ 345.961730] Skipping init section patching addr: 0xc0000000010e2b1c
root@ltc-
In addition to the previous tests.