spi-imx: prevent overflow when estimating transfer time

The words delay is computed by multiplying two unsigned ints
and by adding up the result to a u64 variable.

The multiplication, however, is performed with 32bit math
thus losing data when the actual result is larger than UINT32_MAX.

Fix the operation by casting the first operand to u64, thus forcing
the multiplication to be performed with 64bit math.

This fixes 1 OVERFLOW_BEFORE_WIDEN issue reported by Coverity
Report: CID 1601859:  Integer handling issues  (OVERFLOW_BEFORE_WIDEN)

Cc: Mark Brown <broonie@kernel.org>
Cc: Shawn Guo <shawnguo@kernel.org>
Cc: Sascha Hauer <s.hauer@pengutronix.de>
Cc: Pengutronix Kernel Team <kernel@pengutronix.de>
Cc: Fabio Estevam <festevam@gmail.com>
Cc: imx@lists.linux.dev
Cc: linux-arm-kernel@lists.infradead.org
Cc: linux-kernel@vger.kernel.org
Signed-off-by: Antonio Quartulli <antonio@mandelbit.com>
Link: https://patch.msgid.link/20241115220202.31086-1-antonio@mandelbit.com
Signed-off-by: Mark Brown <broonie@kernel.org>

diff --git a/drivers/spi/spi-imx.c b/drivers/spi/spi-imx.c
index 0b6b015..eeb7d08 100644 (file)
--- a/drivers/spi/spi-imx.c
+++ b/drivers/spi/spi-imx.c
@@ -1685,7 +1685,7 @@ static unsigned int spi_imx_transfer_estimate_time_us(struct spi_transfer *trans
                words = DIV_ROUND_UP(transfer->len * BITS_PER_BYTE, transfer->bits_per_word);
                word_delay_us = DIV_ROUND_CLOSEST(spi_delay_to_ns(&transfer->word_delay, transfer),
                                                  NSEC_PER_USEC);
-               result += words * word_delay_us;
+               result += (u64)words * word_delay_us;
        }
 
        return min(result, U32_MAX);