From: Bill O'Donnell <bodonnel@redhat.com>
To: cem@kernel.org
Cc: linux-xfs@vger.kernel.org, ebiggers@google.com, wbx@openadk.org
Subject: Re: [PATCH] xfs_io: Fix fscrypt macros ordering
Date: Fri, 16 Aug 2024 15:20:34 -0500 [thread overview]
Message-ID: <Zr-0kuyHlKBW25Yd@redhat.com> (raw)
In-Reply-To: <20240816193957.42626-1-cem@kernel.org>
On Fri, Aug 16, 2024 at 09:39:38PM +0200, cem@kernel.org wrote:
> From: Carlos Maiolino <cem@kernel.org>
>
> We've been reported a failure to build xfsprogs within buildroot's
> environment when they tried to upgrade xfsprogs from 6.4 to 6.9:
>
> encrypt.c:53:36: error: 'FSCRYPT_KEY_IDENTIFIER_SIZE' undeclared
> here (not in a function)
> 53 | __u8
> master_key_identifier[FSCRYPT_KEY_IDENTIFIER_SIZE];
> |
> ^~~~~~~~~~~~~~~~~~~~~~~~~~~
> encrypt.c:61:42: error: field 'v1' has incomplete type
> 61 | struct fscrypt_policy_v1 v1;
> | ^~
>
> They were using a kernel version without FS_IOC_GET_ENCRYPTION_POLICY_EX
> set and OVERRIDE_SYSTEM_FSCRYPT_POLICY_V2 was unset.
> This combination caused xfsprogs to attempt to define fscrypt_policy_v2
> locally, which uses:
> __u8 master_key_identifier[FSCRYPT_KEY_IDENTIFIER_SIZE];
>
> The problem is FSCRYPT_KEY_IDENTIFIER_SIZE is only after this block of
> code, so we need to define it earlier.
>
> This also attempts to use fscrypt_policy_v1, which is defined only
> later.
>
> To fix this, just reorder both ifdef blocks, but we need to move the
> definition of FS_IOC_GET_ENCRYPTION_POLICY_EX to the later, otherwise,
> the later definitions won't be enabled causing havoc.
>
> Fixes: e97caf714697a ("xfs_io/encrypt: support specifying crypto data unit size")
> Reported-by: Waldemar Brodkorb <wbx@openadk.org>
> Signed-off-by: Carlos Maiolino <cmaiolino@redhat.com>
Reviewed-by: Bill O'Donnell <bodonnel@redhat.com>
> ---
> io/encrypt.c | 64 ++++++++++++++++++++++++++--------------------------
> 1 file changed, 32 insertions(+), 32 deletions(-)
>
> diff --git a/io/encrypt.c b/io/encrypt.c
> index 79061b07c..97abb964e 100644
> --- a/io/encrypt.c
> +++ b/io/encrypt.c
> @@ -35,35 +35,6 @@
> #define FS_IOC_GET_ENCRYPTION_POLICY _IOW('f', 21, struct fscrypt_policy)
> #endif
>
> -/*
> - * Since the log2_data_unit_size field was added later than fscrypt_policy_v2
> - * itself, we may need to override the system definition to get that field.
> - * And also fscrypt_get_policy_ex_arg since it contains fscrypt_policy_v2.
> - */
> -#if !defined(FS_IOC_GET_ENCRYPTION_POLICY_EX) || \
> - defined(OVERRIDE_SYSTEM_FSCRYPT_POLICY_V2)
> -#undef fscrypt_policy_v2
> -struct fscrypt_policy_v2 {
> - __u8 version;
> - __u8 contents_encryption_mode;
> - __u8 filenames_encryption_mode;
> - __u8 flags;
> - __u8 log2_data_unit_size;
> - __u8 __reserved[3];
> - __u8 master_key_identifier[FSCRYPT_KEY_IDENTIFIER_SIZE];
> -};
> -
> -#undef fscrypt_get_policy_ex_arg
> -struct fscrypt_get_policy_ex_arg {
> - __u64 policy_size; /* input/output */
> - union {
> - __u8 version;
> - struct fscrypt_policy_v1 v1;
> - struct fscrypt_policy_v2 v2;
> - } policy; /* output */
> -};
> -#endif
> -
> /*
> * Second batch of ioctls (Linux headers v5.4+), plus some renamings from FS_ to
> * FSCRYPT_. We don't bother defining the old names here.
> @@ -106,9 +77,6 @@ struct fscrypt_policy_v1 {
>
> #define FSCRYPT_MAX_KEY_SIZE 64
>
> -#define FS_IOC_GET_ENCRYPTION_POLICY_EX _IOWR('f', 22, __u8[9]) /* size + version */
> -/* struct fscrypt_get_policy_ex_arg was defined earlier */
> -
> #define FSCRYPT_KEY_SPEC_TYPE_DESCRIPTOR 1
> #define FSCRYPT_KEY_SPEC_TYPE_IDENTIFIER 2
> struct fscrypt_key_specifier {
> @@ -152,6 +120,38 @@ struct fscrypt_get_key_status_arg {
>
> #endif /* !FS_IOC_GET_ENCRYPTION_POLICY_EX */
>
> +/*
> + * Since the log2_data_unit_size field was added later than fscrypt_policy_v2
> + * itself, we may need to override the system definition to get that field.
> + * And also fscrypt_get_policy_ex_arg since it contains fscrypt_policy_v2.
> + */
> +#if !defined(FS_IOC_GET_ENCRYPTION_POLICY_EX) || \
> + defined(OVERRIDE_SYSTEM_FSCRYPT_POLICY_V2)
> +#undef fscrypt_policy_v2
> +struct fscrypt_policy_v2 {
> + __u8 version;
> + __u8 contents_encryption_mode;
> + __u8 filenames_encryption_mode;
> + __u8 flags;
> + __u8 log2_data_unit_size;
> + __u8 __reserved[3];
> + __u8 master_key_identifier[FSCRYPT_KEY_IDENTIFIER_SIZE];
> +};
> +
> +#undef fscrypt_get_policy_ex_arg
> +struct fscrypt_get_policy_ex_arg {
> + __u64 policy_size; /* input/output */
> + union {
> + __u8 version;
> + struct fscrypt_policy_v1 v1;
> + struct fscrypt_policy_v2 v2;
> + } policy; /* output */
> +};
> +
> +#define FS_IOC_GET_ENCRYPTION_POLICY_EX _IOWR('f', 22, __u8[9]) /* size + version */
> +
> +#endif
> +
> /*
> * Since the key_id field was added later than struct fscrypt_add_key_arg
> * itself, we may need to override the system definition to get that field.
> --
> 2.46.0
>
>
prev parent reply other threads:[~2024-08-16 20:20 UTC|newest]
Thread overview: 3+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-08-16 19:39 [PATCH] xfs_io: Fix fscrypt macros ordering cem
2024-08-16 20:05 ` Eric Biggers
2024-08-16 20:20 ` Bill O'Donnell [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=Zr-0kuyHlKBW25Yd@redhat.com \
--to=bodonnel@redhat.com \
--cc=cem@kernel.org \
--cc=ebiggers@google.com \
--cc=linux-xfs@vger.kernel.org \
--cc=wbx@openadk.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox