[PATCH 08/27] kexec_file: Restrict at runtime if the kernel is locked down

Mimi Zohar zohar at linux.ibm.com
Fri Mar 1 02:05:53 UTC 2019


Hi Jiri,

Thank you for thinking about IMA.

On Thu, 2019-02-28 at 15:11 -0800, Matthew Garrett wrote:
> From: Jiri Bohac <jbohac at suse.cz>
> 
> When KEXEC_SIG is not enabled, kernel should not load images through
> kexec_file systemcall if the kernel is locked down unless IMA can be used
> to validate the image.

This patch is a partial solution, but it doesn't take into account the
architecture specific or custom policies.

Mimi

> [Modified by David Howells to fit with modifications to the previous patch
>  and to return -EPERM if the kernel is locked down for consistency with
>  other lockdowns]
> 
> Signed-off-by: Jiri Bohac <jbohac at suse.cz>
> Signed-off-by: David Howells <dhowells at redhat.com>
> Reviewed-by: Jiri Bohac <jbohac at suse.cz>
> Cc: Matthew Garrett <mjg59 at srcf.ucam.org>
> cc: Chun-Yi Lee <jlee at suse.com>
> cc: kexec at lists.infradead.org
> ---
>  include/linux/ima.h | 6 ++++++
>  kernel/kexec_file.c | 8 ++++++++
>  2 files changed, 14 insertions(+)
> 
> diff --git a/include/linux/ima.h b/include/linux/ima.h
> index b5e16b8c50b7..b35ed0725a05 100644
> --- a/include/linux/ima.h
> +++ b/include/linux/ima.h
> @@ -127,4 +127,10 @@ static inline int ima_inode_removexattr(struct dentry *dentry,
>  	return 0;
>  }
>  #endif /* CONFIG_IMA_APPRAISE */
> +
> +static inline bool is_ima_kexec_appraise_enabled(void)
> +{
> +	return IS_ENABLED(CONFIG_IMA_APPRAISE_REQUIRE_KEXEC_SIGS);
> +}
> +
>  #endif /* _LINUX_IMA_H */
> diff --git a/kernel/kexec_file.c b/kernel/kexec_file.c
> index 67f3a866eabe..b4e938dff4be 100644
> --- a/kernel/kexec_file.c
> +++ b/kernel/kexec_file.c
> @@ -239,6 +239,14 @@ kimage_file_prepare_segments(struct kimage *image, int kernel_fd, int initrd_fd,
>  		}
> 
>  		ret = 0;
> +		if (is_ima_kexec_appraise_enabled())
> +			break;
> +
> +		if (kernel_is_locked_down(reason)) {
> +			ret = -EPERM;
> +			goto out;
> +		}
> +
>  		break;
> 
>  		/* All other errors are fatal, including nomem, unparseable



More information about the Linux-security-module-archive mailing list