On Thu, 13 Jun 2019 20:07:22 -0500
Josh Poimboeuf <jpoim...@redhat.com> wrote:
> It's possible for livepatch and ftrace to be toggling a module's text
> permissions at the same time, resulting in the following panic:
>
[..]
> The above panic occurs when loading two modules at the same time with
> ftrace enabled, where at least one of the modules is a livepatch module:
>
> CPU0 CPU1
> klp_enable_patch()
> klp_init_object_loaded()
> module_disable_ro()
> ftrace_module_enable()
> ftrace_arch_code_modify_post_process()
> set_all_modules_text_ro()
> klp_write_object_relocations()
> apply_relocate_add()
> *patches read-only code* - BOOM
>
> A similar race exists when toggling ftrace while loading a livepatch
> module.
>
> Fix it by ensuring that the livepatch and ftrace code patching
> operations -- and their respective permissions changes -- are protected
> by the text_mutex.
>
> Reported-by: Johannes Erdfelt <johan...@erdfelt.com>
> Fixes: 444d13ff10fb ("modules: add ro_after_init support")
> Signed-off-by: Josh Poimboeuf <jpoim...@redhat.com>
> Acked-by: Jessica Yu <j...@kernel.org>
> Reviewed-by: Petr Mladek <pmla...@suse.com>
> Reviewed-by: Miroslav Benes <mbe...@suse.cz>
This patch looks uncontroversial. I'm going to pull this one in and
start testing it. And if it works, I'll push to Linus.
-- Steve
