If the i2c bus driver ignores the I2C_M_RECV_LEN flag (as some of
them do), it is possible for an I2C_SMBUS_BLOCK_DATA read issued
on some random device to return an arbitrary value in the first
byte (and nothing else). When this happens, i2c_smbus_xfer_emulated()
will happily write past the end of the supplied data buffer, thus
causing Bad Things to happen. To prevent this, check the size
before copying the data block and return an error if it is too large.
Fixes: 209d27c3b167 ("i2c: Emulate SMBus block read over I2C")
Signed-off-by: Mans Rullgard <[email protected]>
---
drivers/i2c/i2c-core-smbus.c | 7 +++++++
1 file changed, 7 insertions(+)
diff --git a/drivers/i2c/i2c-core-smbus.c b/drivers/i2c/i2c-core-smbus.c
index 3ac426a8ab5a..a719c26b98ac 100644
--- a/drivers/i2c/i2c-core-smbus.c
+++ b/drivers/i2c/i2c-core-smbus.c
@@ -495,6 +495,13 @@ static s32 i2c_smbus_xfer_emulated(struct i2c_adapter *adapter, u16 addr,
break;
case I2C_SMBUS_BLOCK_DATA:
case I2C_SMBUS_BLOCK_PROC_CALL:
+ if (msg[1].buf[0] > I2C_SMBUS_BLOCK_MAX) {
+ dev_err(&adapter->dev,
+ "Invalid block size returned: %d\n",
+ msg[1].buf[0]);
+ status = -EINVAL;
+ goto cleanup;
+ }
for (i = 0; i < msg[1].buf[0] + 1; i++)
data->block[i] = msg[1].buf[i];
break;
--
2.27.0
On Sat, Jun 13, 2020 at 11:41:09AM +0100, Mans Rullgard wrote:
> If the i2c bus driver ignores the I2C_M_RECV_LEN flag (as some of
> them do), it is possible for an I2C_SMBUS_BLOCK_DATA read issued
Out of interest, which driver did you use?
> on some random device to return an arbitrary value in the first
> byte (and nothing else). When this happens, i2c_smbus_xfer_emulated()
> will happily write past the end of the supplied data buffer, thus
> causing Bad Things to happen. To prevent this, check the size
> before copying the data block and return an error if it is too large.
Good catch, we were relying on the drivers too much here. I think the
same fix is needed for the non-emulated case as well. Will have a look.
> + if (msg[1].buf[0] > I2C_SMBUS_BLOCK_MAX) {
> + dev_err(&adapter->dev,
> + "Invalid block size returned: %d\n",
> + msg[1].buf[0]);
> + status = -EINVAL;
I changed this to -EPROTO as described in
Documentation/i2c/fault-codes.rst.
Applied to for-current, thanks!
Wolfram Sang <[email protected]> writes:
> On Sat, Jun 13, 2020 at 11:41:09AM +0100, Mans Rullgard wrote:
>> If the i2c bus driver ignores the I2C_M_RECV_LEN flag (as some of
>> them do), it is possible for an I2C_SMBUS_BLOCK_DATA read issued
>
> Out of interest, which driver did you use?
I saw it with the mv64xxx (Allwinner) and omap (Beaglebone) drivers.
From a quick look, it seems like quite a few others have the same
problem.
>> on some random device to return an arbitrary value in the first
>> byte (and nothing else). When this happens, i2c_smbus_xfer_emulated()
>> will happily write past the end of the supplied data buffer, thus
>> causing Bad Things to happen. To prevent this, check the size
>> before copying the data block and return an error if it is too large.
>
> Good catch, we were relying on the drivers too much here. I think the
> same fix is needed for the non-emulated case as well. Will have a look.
>
>> + if (msg[1].buf[0] > I2C_SMBUS_BLOCK_MAX) {
>> + dev_err(&adapter->dev,
>> + "Invalid block size returned: %d\n",
>> + msg[1].buf[0]);
>> + status = -EINVAL;
>
> I changed this to -EPROTO as described in
> Documentation/i2c/fault-codes.rst.
>
> Applied to for-current, thanks!
>
--
M?ns Rullg?rd
> I saw it with the mv64xxx (Allwinner) and omap (Beaglebone) drivers.
> From a quick look, it seems like quite a few others have the same
> problem.
Okay, well, to be fair, both drivers don't advertise
I2C_SMBUS_BLOCK_DATA. The client driver should check for that. Anyhow,
it makes sense to have your additional check in the core as fallback
safety.