While raw_msg isn't a fixed size, it does have a maximum size. Adjust the
struct to represent this and avoid the following warning when building
with -Wzero-length-bounds:
drivers/net/can/usb/etas_es58x/es58x_fd.c: In function 'es58x_fd_tx_can_msg':
drivers/net/can/usb/etas_es58x/es58x_fd.c:360:35: warning: array subscript 65535 is outside the bounds of an interior zero-length array 'u8[0]' {aka 'unsigned char[]'} [-Wzero-length-bounds]
360 | tx_can_msg = (typeof(tx_can_msg))&es58x_fd_urb_cmd->raw_msg[msg_len];
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
In file included from drivers/net/can/usb/etas_es58x/es58x_core.h:22,
from drivers/net/can/usb/etas_es58x/es58x_fd.c:17:
drivers/net/can/usb/etas_es58x/es58x_fd.h:231:6: note: while referencing 'raw_msg'
231 | u8 raw_msg[0];
| ^~~~~~~
Cc: Wolfgang Grandegger <redacted>
Cc: Marc Kleine-Budde <mkl@pengutronix.de>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Jakub Kicinski <kuba@kernel.org>
Cc: Arunachalam Santhanam <redacted>
Cc: Vincent Mailhol <redacted>
Cc: linux-can@vger.kernel.org
Cc: netdev@vger.kernel.org
Signed-off-by: Kees Cook <redacted>
---
drivers/net/can/usb/etas_es58x/es581_4.h | 2 +-
drivers/net/can/usb/etas_es58x/es58x_fd.h | 2 +-
2 files changed, 2 insertions(+), 2 deletions(-)
From: Vincent MAILHOL <hidden> Date: 2021-08-18 05:14:08
On Wed. 18 Aug 2021 at 12:40, Kees Cook [off-list ref] wrote:
quoted hunk
While raw_msg isn't a fixed size, it does have a maximum size. Adjust the
struct to represent this and avoid the following warning when building
with -Wzero-length-bounds:
drivers/net/can/usb/etas_es58x/es58x_fd.c: In function 'es58x_fd_tx_can_msg':
drivers/net/can/usb/etas_es58x/es58x_fd.c:360:35: warning: array subscript 65535 is outside the bounds of an interior zero-length array 'u8[0]' {aka 'unsigned char[]'} [-Wzero-length-bounds]
360 | tx_can_msg = (typeof(tx_can_msg))&es58x_fd_urb_cmd->raw_msg[msg_len];
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
In file included from drivers/net/can/usb/etas_es58x/es58x_core.h:22,
from drivers/net/can/usb/etas_es58x/es58x_fd.c:17:
drivers/net/can/usb/etas_es58x/es58x_fd.h:231:6: note: while referencing 'raw_msg'
231 | u8 raw_msg[0];
| ^~~~~~~
Cc: Wolfgang Grandegger <redacted>
Cc: Marc Kleine-Budde <mkl@pengutronix.de>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Jakub Kicinski <kuba@kernel.org>
Cc: Arunachalam Santhanam <redacted>
Cc: Vincent Mailhol <redacted>
Cc: linux-can@vger.kernel.org
Cc: netdev@vger.kernel.org
Signed-off-by: Kees Cook <redacted>
---
drivers/net/can/usb/etas_es58x/es581_4.h | 2 +-
drivers/net/can/usb/etas_es58x/es58x_fd.h | 2 +-
2 files changed, 2 insertions(+), 2 deletions(-)
raw_msg is part of a union so its maximum size is implicitly the
biggest size of the other member of that union:
| struct es58x_fd_urb_cmd {
| __le16 SOF;
| u8 cmd_type;
| u8 cmd_id;
| u8 channel_idx;
| __le16 msg_len;
|
| union {
| struct es58x_fd_tx_conf_msg tx_conf_msg;
| u8 tx_can_msg_buf[ES58X_FD_TX_BULK_MAX * ES58X_FD_CANFD_TX_LEN];
| u8 rx_can_msg_buf[ES58X_FD_RX_BULK_MAX * ES58X_FD_CANFD_RX_LEN];
| struct es58x_fd_echo_msg echo_msg[ES58X_FD_ECHO_BULK_MAX];
| struct es58x_fd_rx_event_msg rx_event_msg;
| struct es58x_fd_tx_ack_msg tx_ack_msg;
| __le64 timestamp;
| __le32 rx_cmd_ret_le32;
| u8 raw_msg[0];
| } __packed;
|
| __le16 reserved_for_crc16_do_not_use;
| } __packed;
ram_msg can then be used to manipulate the other fields at the byte level.
I am sorry but I fail to understand why this is an issue.
Also, the proposed fix drastically increases the size of the structure.
Yours sincerely,
Vincent
On Wed, Aug 18, 2021 at 02:13:51PM +0900, Vincent MAILHOL wrote:
On Wed. 18 Aug 2021 at 12:40, Kees Cook [off-list ref] wrote:
quoted
While raw_msg isn't a fixed size, it does have a maximum size. Adjust the
struct to represent this and avoid the following warning when building
with -Wzero-length-bounds:
drivers/net/can/usb/etas_es58x/es58x_fd.c: In function 'es58x_fd_tx_can_msg':
drivers/net/can/usb/etas_es58x/es58x_fd.c:360:35: warning: array subscript 65535 is outside the bounds of an interior zero-length array 'u8[0]' {aka 'unsigned char[]'} [-Wzero-length-bounds]
360 | tx_can_msg = (typeof(tx_can_msg))&es58x_fd_urb_cmd->raw_msg[msg_len];
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
In file included from drivers/net/can/usb/etas_es58x/es58x_core.h:22,
from drivers/net/can/usb/etas_es58x/es58x_fd.c:17:
drivers/net/can/usb/etas_es58x/es58x_fd.h:231:6: note: while referencing 'raw_msg'
231 | u8 raw_msg[0];
| ^~~~~~~
Cc: Wolfgang Grandegger <redacted>
Cc: Marc Kleine-Budde <mkl@pengutronix.de>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Jakub Kicinski <kuba@kernel.org>
Cc: Arunachalam Santhanam <redacted>
Cc: Vincent Mailhol <redacted>
Cc: linux-can@vger.kernel.org
Cc: netdev@vger.kernel.org
Signed-off-by: Kees Cook <redacted>
---
drivers/net/can/usb/etas_es58x/es581_4.h | 2 +-
drivers/net/can/usb/etas_es58x/es58x_fd.h | 2 +-
2 files changed, 2 insertions(+), 2 deletions(-)
raw_msg is part of a union so its maximum size is implicitly the
biggest size of the other member of that union:
Yup, understood. See below...
| struct es58x_fd_urb_cmd {
| __le16 SOF;
| u8 cmd_type;
| u8 cmd_id;
| u8 channel_idx;
| __le16 msg_len;
|
| union {
| struct es58x_fd_tx_conf_msg tx_conf_msg;
| u8 tx_can_msg_buf[ES58X_FD_TX_BULK_MAX * ES58X_FD_CANFD_TX_LEN];
| u8 rx_can_msg_buf[ES58X_FD_RX_BULK_MAX * ES58X_FD_CANFD_RX_LEN];
| struct es58x_fd_echo_msg echo_msg[ES58X_FD_ECHO_BULK_MAX];
| struct es58x_fd_rx_event_msg rx_event_msg;
| struct es58x_fd_tx_ack_msg tx_ack_msg;
| __le64 timestamp;
| __le32 rx_cmd_ret_le32;
| u8 raw_msg[0];
| } __packed;
|
| __le16 reserved_for_crc16_do_not_use;
| } __packed;
ram_msg can then be used to manipulate the other fields at the byte level.
I am sorry but I fail to understand why this is an issue.
The issue is with using a 0-element array (these are being removed from
the kernel[1] so we can add -Warray-bounds). Normally in this situation I
would replace the 0-element array with a flexible array, but this
case is unusual in several ways:
- There is a trailing struct member (reserved_for_crc16_do_not_use),
which is never accessed (good), and documented as "please never access
this".
- struct es58x_fd_urb_cmd is statically allocated (it is written into
from the URB handler).
- The message lengths coming from the USB device are stored in a u16,
which looked like it was possible to overflow the buffer.
In taking a closer look, I see that the URB command length is checked,
and the in-data length is checked as well, so the overflow concern
appears to be addressed.
Also, the proposed fix drastically increases the size of the structure.
From: Vincent MAILHOL <hidden> Date: 2021-08-18 07:55:48
On Wed. 18 Aug 2021 at 15:48, Kees Cook [off-list ref] wrote:
On Wed, Aug 18, 2021 at 02:13:51PM +0900, Vincent MAILHOL wrote:
quoted
On Wed. 18 Aug 2021 at 12:40, Kees Cook [off-list ref] wrote:
quoted
While raw_msg isn't a fixed size, it does have a maximum size. Adjust the
struct to represent this and avoid the following warning when building
with -Wzero-length-bounds:
drivers/net/can/usb/etas_es58x/es58x_fd.c: In function 'es58x_fd_tx_can_msg':
drivers/net/can/usb/etas_es58x/es58x_fd.c:360:35: warning: array subscript 65535 is outside the bounds of an interior zero-length array 'u8[0]' {aka 'unsigned char[]'} [-Wzero-length-bounds]
360 | tx_can_msg = (typeof(tx_can_msg))&es58x_fd_urb_cmd->raw_msg[msg_len];
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
In file included from drivers/net/can/usb/etas_es58x/es58x_core.h:22,
from drivers/net/can/usb/etas_es58x/es58x_fd.c:17:
drivers/net/can/usb/etas_es58x/es58x_fd.h:231:6: note: while referencing 'raw_msg'
231 | u8 raw_msg[0];
| ^~~~~~~
Cc: Wolfgang Grandegger <redacted>
Cc: Marc Kleine-Budde <mkl@pengutronix.de>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Jakub Kicinski <kuba@kernel.org>
Cc: Arunachalam Santhanam <redacted>
Cc: Vincent Mailhol <redacted>
Cc: linux-can@vger.kernel.org
Cc: netdev@vger.kernel.org
Signed-off-by: Kees Cook <redacted>
---
drivers/net/can/usb/etas_es58x/es581_4.h | 2 +-
drivers/net/can/usb/etas_es58x/es58x_fd.h | 2 +-
2 files changed, 2 insertions(+), 2 deletions(-)
raw_msg is part of a union so its maximum size is implicitly the
biggest size of the other member of that union:
Yup, understood. See below...
quoted
| struct es58x_fd_urb_cmd {
| __le16 SOF;
| u8 cmd_type;
| u8 cmd_id;
| u8 channel_idx;
| __le16 msg_len;
|
| union {
| struct es58x_fd_tx_conf_msg tx_conf_msg;
| u8 tx_can_msg_buf[ES58X_FD_TX_BULK_MAX * ES58X_FD_CANFD_TX_LEN];
| u8 rx_can_msg_buf[ES58X_FD_RX_BULK_MAX * ES58X_FD_CANFD_RX_LEN];
| struct es58x_fd_echo_msg echo_msg[ES58X_FD_ECHO_BULK_MAX];
| struct es58x_fd_rx_event_msg rx_event_msg;
| struct es58x_fd_tx_ack_msg tx_ack_msg;
| __le64 timestamp;
| __le32 rx_cmd_ret_le32;
| u8 raw_msg[0];
| } __packed;
|
| __le16 reserved_for_crc16_do_not_use;
| } __packed;
ram_msg can then be used to manipulate the other fields at the byte level.
I am sorry but I fail to understand why this is an issue.
The issue is with using a 0-element array (these are being removed from
the kernel[1] so we can add -Warray-bounds). Normally in this situation I
would replace the 0-element array with a flexible array, but this
case is unusual in several ways:
- There is a trailing struct member (reserved_for_crc16_do_not_use),
which is never accessed (good), and documented as "please never access
this".
Yes. And FYI, this field is here so that
| sizeof(struct es58x_fd_urb_cmd)
returns the correct maximum size.
And, of course, because this structure will be sent to the
device, there is no possibility to reorder those fields.
- struct es58x_fd_urb_cmd is statically allocated (it is written into
from the URB handler).
- The message lengths coming from the USB device are stored in a u16,
which looked like it was possible to overflow the buffer.
In taking a closer look, I see that the URB command length is checked,
and the in-data length is checked as well, so the overflow concern
appears to be addressed.
quoted
Also, the proposed fix drastically increases the size of the structure.
Indeed. I will send a v2, now that I see that the overflow concern isn't
an issue.
Thanks for the explanation. That makes sense.
At the end, the only goal of raw_msg[] is to have a tag pointing
to the beginning of the union. It would be virtually identical to
something like:
| u8 raw_msg[];
| union {
| /* ... */
| } __packed ;
I had a look at your work and especially at your struct_group() macro.
Do you think it would make sense to introduce a union_group()?
Result would look like:
| union_group_attr(urb_msg, __packed, /* raw_msg renamed to urb_msg */
| struct es58x_fd_tx_conf_msg tx_conf_msg;
| u8 tx_can_msg_buf[ES58X_FD_TX_BULK_MAX * ES58X_FD_CANFD_TX_LEN];
| u8 rx_can_msg_buf[ES58X_FD_RX_BULK_MAX * ES58X_FD_CANFD_RX_LEN];
| struct es58x_fd_echo_msg echo_msg[ES58X_FD_ECHO_BULK_MAX];
| struct es58x_fd_rx_event_msg rx_event_msg;
| struct es58x_fd_tx_ack_msg tx_ack_msg;
| __le64 timestamp;
| __le32 rx_cmd_ret_le32;
| );
And I can then use urb_msg in place of the old raw_msg (might
need a bit of rework here and there but I can take care of it).
This is the most pretty way I can think of to remove this zero length array.
Keeping the raw_msg[] but with another size seems odd to me.
Or maybe I would be the only one using this feature in the full
tree? In that case, maybe it would make sense to keep the
union_group_attr() macro local to the etas_es58x driver?
Yours sincerely,
Vincent
On Wed, Aug 18, 2021 at 04:55:20PM +0900, Vincent MAILHOL wrote:
At the end, the only goal of raw_msg[] is to have a tag pointing
to the beginning of the union. It would be virtually identical to
something like:
| u8 raw_msg[];
| union {
| /* ... */
| } __packed ;
I had a look at your work and especially at your struct_group() macro.
Do you think it would make sense to introduce a union_group()?
Result would look like:
| union_group_attr(urb_msg, __packed, /* raw_msg renamed to urb_msg */
| struct es58x_fd_tx_conf_msg tx_conf_msg;
| u8 tx_can_msg_buf[ES58X_FD_TX_BULK_MAX * ES58X_FD_CANFD_TX_LEN];
| u8 rx_can_msg_buf[ES58X_FD_RX_BULK_MAX * ES58X_FD_CANFD_RX_LEN];
| struct es58x_fd_echo_msg echo_msg[ES58X_FD_ECHO_BULK_MAX];
| struct es58x_fd_rx_event_msg rx_event_msg;
| struct es58x_fd_tx_ack_msg tx_ack_msg;
| __le64 timestamp;
| __le32 rx_cmd_ret_le32;
| );
And I can then use urb_msg in place of the old raw_msg (might
need a bit of rework here and there but I can take care of it).
This is the most pretty way I can think of to remove this zero length array.
Keeping the raw_msg[] but with another size seems odd to me.
Or maybe I would be the only one using this feature in the full
tree? In that case, maybe it would make sense to keep the
union_group_attr() macro local to the etas_es58x driver?
From: Vincent MAILHOL <hidden> Date: 2021-08-18 09:35:14
On Wed. 18 Aug 2021 at 18:03, Kees Cook [off-list ref] wrote:
On Wed, Aug 18, 2021 at 04:55:20PM +0900, Vincent MAILHOL wrote:
quoted
At the end, the only goal of raw_msg[] is to have a tag pointing
to the beginning of the union. It would be virtually identical to
something like:
| u8 raw_msg[];
| union {
| /* ... */
| } __packed ;
I had a look at your work and especially at your struct_group() macro.
Do you think it would make sense to introduce a union_group()?
Result would look like:
| union_group_attr(urb_msg, __packed, /* raw_msg renamed to urb_msg */
| struct es58x_fd_tx_conf_msg tx_conf_msg;
| u8 tx_can_msg_buf[ES58X_FD_TX_BULK_MAX * ES58X_FD_CANFD_TX_LEN];
| u8 rx_can_msg_buf[ES58X_FD_RX_BULK_MAX * ES58X_FD_CANFD_RX_LEN];
| struct es58x_fd_echo_msg echo_msg[ES58X_FD_ECHO_BULK_MAX];
| struct es58x_fd_rx_event_msg rx_event_msg;
| struct es58x_fd_tx_ack_msg tx_ack_msg;
| __le64 timestamp;
| __le32 rx_cmd_ret_le32;
| );
And I can then use urb_msg in place of the old raw_msg (might
need a bit of rework here and there but I can take care of it).
This is the most pretty way I can think of to remove this zero length array.
Keeping the raw_msg[] but with another size seems odd to me.
Or maybe I would be the only one using this feature in the full
tree? In that case, maybe it would make sense to keep the
union_group_attr() macro local to the etas_es58x driver?
The __flex_array itself is a nasty hack :D but the rest is clean.
Is this compliant to the C standard? Well, I guess that as long
as both GCC and LLVM supports it, it is safe to add it to the
kernel.
I like the final result. I will do a bit more testing and give my
acknowledgement if everything goes well.
Yours sincerely,
Vincent
On Wed, Aug 18, 2021 at 06:33:39PM +0900, Vincent MAILHOL wrote:
On Wed. 18 Aug 2021 at 18:03, Kees Cook [off-list ref] wrote:
quoted
On Wed, Aug 18, 2021 at 04:55:20PM +0900, Vincent MAILHOL wrote:
quoted
At the end, the only goal of raw_msg[] is to have a tag pointing
to the beginning of the union. It would be virtually identical to
something like:
| u8 raw_msg[];
| union {
| /* ... */
| } __packed ;
I had a look at your work and especially at your struct_group() macro.
Do you think it would make sense to introduce a union_group()?
Result would look like:
| union_group_attr(urb_msg, __packed, /* raw_msg renamed to urb_msg */
| struct es58x_fd_tx_conf_msg tx_conf_msg;
| u8 tx_can_msg_buf[ES58X_FD_TX_BULK_MAX * ES58X_FD_CANFD_TX_LEN];
| u8 rx_can_msg_buf[ES58X_FD_RX_BULK_MAX * ES58X_FD_CANFD_RX_LEN];
| struct es58x_fd_echo_msg echo_msg[ES58X_FD_ECHO_BULK_MAX];
| struct es58x_fd_rx_event_msg rx_event_msg;
| struct es58x_fd_tx_ack_msg tx_ack_msg;
| __le64 timestamp;
| __le32 rx_cmd_ret_le32;
| );
And I can then use urb_msg in place of the old raw_msg (might
need a bit of rework here and there but I can take care of it).
This is the most pretty way I can think of to remove this zero length array.
Keeping the raw_msg[] but with another size seems odd to me.
Or maybe I would be the only one using this feature in the full
tree? In that case, maybe it would make sense to keep the
union_group_attr() macro local to the etas_es58x driver?
Is this compliant to the C standard? Well, I guess that as long
as both GCC and LLVM supports it, it is safe to add it to the
kernel.
The kernel already uses a bunch of compiler extensions, none of which
were legal under the C standard to begin with. :) So, really, this is
about normalizing what we're already doing and finding a single hack
that helps the code base for readability and robustness.
I like the final result. I will do a bit more testing and give my
acknowledgement if everything goes well.