create-diff-object: Add support for x86_64 CONFIG_CALL_PADDING #1487
+37
−0
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
For x86_64 this option uses -fpatchable-function-entry=$(CONFIG_FUNCTION_PADDING_BYTES),$(CONFIG_FUNCTION_PADDING_BYTES), so CONFIG_FUNCTION_PADDING_BYTES NOPs are placed before the function entry point (in order to store a pointer to ftrace_ops).When calculating function padding, check for the presence of the CONFIG_FUNCTION_PADDING_BYTES NOPs, and adjust the padding size by CONFIG_FUNCTION_PADDING_BYTES if they are found. Signed-off-by: lu21253 <lu.haoA@h3c.com> Reviewed-by: w28894 <wang.shijie@h3c.com> Reviewed-by: z17497 <zhang.chunA@h3c.com>
lu21253
changed the title
Contributor
|
/packit test |
Contributor
|
Hi @lu21253 : just curious, what was the failure mode that you encountered that this patch fixes? And if you could attach your complete kernel configuration. I ask as RHEL-10 sets CONFIG_CALL_PADDING=y and CONFIG_FUNCTION_PADDING_BYTES=16, yet there are no problems (at least using gcc). Thanks and btw, many of us here will be away at LPC over the next week or two, so replies may take a while. Thanks! |
Author
|
Hi, @joe-lawrence : make: Leaving directory '/home/l21253/V9/B93D001/PUBLIC/kernel' /home/l21253/V9/B93D001/kpatch-master/kpatch-build/create-diff-object /home/l21253/V9/B93D001/objs64/orig/n_tty.o /home/l21253/V9/B93D001/objs64/patched/n_tty.o vmlinux /home/l21253/V9/B93D001/objs64/vmlinux.symtab /home/l21253/V9/B93D001/TARGET/arch/x86_64/release/Module.symvers livepatch-livepatch-vmlinux-tty.ko /home/l21253/V9/B93D001/objs64/output/n_tty.o create-diff-object: ERROR: home/l21253/V9/B93D001/objs64/orig/n_tty.o: kpatch_bundle_symbols: 355: symbol echo_set_canon_col at offset 16 within section .text.echo_set_canon_col, expected 0 |
Contributor
|
@lu21253 : can you attach your kernel .config? (Or indicate which distro kernel you may be using.) Thanks! |
Author
|
@joe-lawrence kernel version is 6.6.0, Thanks! |
Contributor
Yes that's the version, but the kernel config? Perhaps it's at |
Author
|
Contributor
In #1411 we added a kpatch-build assertion against this option (IIRC there was some explanation in #1320), so you should had gotten an "ERROR: kernel option 'CONFIG_X86_KERNEL_IBT' not supported" error message right off the bat. The short of it is that with this kernel option, and a few others now, the kernel build will defer objtool object file manipulation to vmlinux.o instead of on individual object files. This breaks kpatch-build assumptions and we never found a great workaround, so to address that Josh created an upstream, in-kernel-tree approach just recently merged: https://www.phoronix.com/news/Linux-6.19-objtool-klp-build That said, maybe this kpatch-build patch is reasonable for completeness on !CONFIG_X86_KERNEL_IBT configs. How did you hit the create-diff-object error (ie, which version of kpatch-build are you using) and do you happen to have an input kpatch for repro? Thanks! |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
For x86_64 this option uses -fpatchable-function-entry=$(CONFIG_FUNCTION_PADDING_BYTES),$(CONFIG_FUNCTION_PADDING_BYTES), so CONFIG_FUNCTION_PADDING_BYTES NOPs are placed before the function entry point (in order to store a pointer to ftrace_ops).When calculating function padding, check for the presence of the CONFIG_FUNCTION_PADDING_BYTES NOPs, and adjust the padding size by CONFIG_FUNCTION_PADDING_BYTES if they are found.
Signed-off-by: l21253 lu.haoA@h3c.com
Reviewed-by: w28894 wang.shijie@h3c.com
Reviewed-by: z17497 zhang.chunA@h3c.com