diff options
| author | Ben Hutchings <ben.hutchings@codethink.co.uk> | 2018-06-19 11:17:32 -0700 |
|---|---|---|
| committer | Greg Kroah-Hartman <gregkh@linuxfoundation.org> | 2018-07-03 11:23:17 +0200 |
| commit | 54ae564b35423f25b763bc9e01b489d9f11c034e (patch) | |
| tree | a10a3892fa8d2d6bea0d48eb301a3478e845cb17 /drivers/phy/phy-qcom-ufs-qmp-14nm.c | |
| parent | 2a1b1234d0502237872f6a11016061328528b86d (diff) | |
Input: elan_i2c_smbus - fix more potential stack buffer overflows
commit 50fc7b61959af4b95fafce7fe5dd565199e0b61a upstream.
Commit 40f7090bb1b4 ("Input: elan_i2c_smbus - fix corrupted stack")
fixed most of the functions using i2c_smbus_read_block_data() to
allocate a buffer with the maximum block size. However three
functions were left unchanged:
* In elan_smbus_initialize(), increase the buffer size in the same
way.
* In elan_smbus_calibrate_result(), the buffer is provided by the
caller (calibrate_store()), so introduce a bounce buffer. Also
name the result buffer size.
* In elan_smbus_get_report(), the buffer is provided by the caller
but happens to be the right length. Add a compile-time assertion
to ensure this remains the case.
Cc: <stable@vger.kernel.org> # 3.19+
Signed-off-by: Ben Hutchings <ben.hutchings@codethink.co.uk>
Reviewed-by: Benjamin Tissoires <benjamin.tissoires@redhat.com>
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Diffstat (limited to 'drivers/phy/phy-qcom-ufs-qmp-14nm.c')
0 files changed, 0 insertions, 0 deletions