[PATCH RFC v12 5/20] initramfs|security: Add security hook to initramfs unpack

[Fan Wu]

On 2/3/2024 2:25 PM, Paul Moore wrote:
> On Jan 30, 2024 Fan Wu <wufan at linux.microsoft.com> wrote:
>>
>> This patch introduces a new hook to notify security system that the
>> content of initramfs has been unpacked into the rootfs.
>>
>> Upon receiving this notification, the security system can activate
>> a policy to allow only files that originated from the initramfs to
>> execute or load into kernel during the early stages of booting.
>>
>> This approach is crucial for minimizing the attack surface by
>> ensuring that only trusted files from the initramfs are operational
>> in the critical boot phase.
>>
>> Signed-off-by: Fan Wu <wufan at linux.microsoft.com>
>> ---
>> v1-v11:
>>    + Not present
>>
>> v12:
>>    + Introduced
>> ---
>>   include/linux/lsm_hook_defs.h |  4 ++++
>>   include/linux/security.h      | 10 ++++++++++
>>   init/initramfs.c              |  3 +++
>>   security/security.c           | 12 ++++++++++++
>>   4 files changed, 29 insertions(+)
>>
>> diff --git a/include/linux/lsm_hook_defs.h b/include/linux/lsm_hook_defs.h
>> index 185924c56378..b247388786a9 100644
>> --- a/include/linux/lsm_hook_defs.h
>> +++ b/include/linux/lsm_hook_defs.h
>> @@ -425,3 +425,7 @@ LSM_HOOK(int, 0, uring_override_creds, const struct cred *new)
>>   LSM_HOOK(int, 0, uring_sqpoll, void)
>>   LSM_HOOK(int, 0, uring_cmd, struct io_uring_cmd *ioucmd)
>>   #endif /* CONFIG_IO_URING */
>> +
>> +#ifdef CONFIG_BLK_DEV_INITRD
>> +LSM_HOOK(void, LSM_RET_VOID, unpack_initramfs_security, void)
>> +#endif /* CONFIG_BLK_DEV_INITRD */
> 
> Let's just call it "unpack_initramfs", the "_security" part is somewhat
> implied since we are talking about a LSM hook ;)
> 
>> diff --git a/init/initramfs.c b/init/initramfs.c
>> index 76deb48c38cb..075a5794cde5 100644
>> --- a/init/initramfs.c
>> +++ b/init/initramfs.c
>> @@ -18,6 +18,7 @@
>>   #include <linux/init_syscalls.h>
>>   #include <linux/task_work.h>
>>   #include <linux/umh.h>
>> +#include <linux/security.h>
>>   
>>   static __initdata bool csum_present;
>>   static __initdata u32 io_csum;
>> @@ -720,6 +721,8 @@ static void __init do_populate_rootfs(void *unused, async_cookie_t cookie)
>>   #endif
>>   	}
>>   
>> +	security_unpack_initramfs();
> 
> Given the caller, what do you think of changing the hook name to
> "security_initramfs_populated()"?  I think this not only matches up
> better with the caller, "do_populate_rootfs()", but since in using the
> past tense we help indicate that this hook happens *after* the rootfs
> is populated with the initramfs data.
> 

Yeah, I agree this sounds better. I will update this part.

-Fan
>>   done:
>>   	/*
>>   	 * If the initrd region is overlapped with crashkernel reserved region,
>> diff --git a/security/security.c b/security/security.c
>> index ddf2e69cf8f2..2a527d4c69bc 100644
>> --- a/security/security.c
>> +++ b/security/security.c
>> @@ -5581,3 +5581,15 @@ int security_uring_cmd(struct io_uring_cmd *ioucmd)
>>   	return call_int_hook(uring_cmd, 0, ioucmd);
>>   }
>>   #endif /* CONFIG_IO_URING */
>> +
>> +#ifdef CONFIG_BLK_DEV_INITRD
>> +/**
>> + * security_unpack_initramfs() - Notify LSM that initramfs has been loaded
>> + *
>> + * Tells the LSM the initramfs has been unpacked into the rootfs.
>> + */
>> +void security_unpack_initramfs(void)
>> +{
>> +	call_void_hook(unpack_initramfs_security);
>> +}
>> +#endif /* CONFIG_BLK_DEV_INITRD */
>> -- 
>> 2.43.0
> 
> --
> paul-moore.com