From: Rui Miguel Silva <rui.silva@xxxxxxxxxx> [ Upstream commit f757f9291f920e1da4c6cfd4064c6bf59639983e ] The loops to setup the memory pool were skipping some blocks, that was not visible on the ISP1763 because it has fewer blocks than the ISP1761. But won testing on that IP from the family that would be an issue. Reported-by: Dietmar Eggemann <dietmar.eggemann@xxxxxxx> Tested-by: Dietmar Eggemann <dietmar.eggemann@xxxxxxx> Signed-off-by: Rui Miguel Silva <rui.silva@xxxxxxxxxx> Link: https://lore.kernel.org/r/20210827131154.4151862-2-rui.silva@xxxxxxxxxx Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx> Signed-off-by: Sasha Levin <sashal@xxxxxxxxxx> --- drivers/usb/isp1760/isp1760-hcd.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/usb/isp1760/isp1760-hcd.c b/drivers/usb/isp1760/isp1760-hcd.c index 27168b4a4ef2..ffb3a0c8c909 100644 --- a/drivers/usb/isp1760/isp1760-hcd.c +++ b/drivers/usb/isp1760/isp1760-hcd.c @@ -588,8 +588,8 @@ static void init_memory(struct isp1760_hcd *priv) payload_addr = PAYLOAD_OFFSET; - for (i = 0, curr = 0; i < ARRAY_SIZE(mem->blocks); i++) { - for (j = 0; j < mem->blocks[i]; j++, curr++) { + for (i = 0, curr = 0; i < ARRAY_SIZE(mem->blocks); i++, curr += j) { + for (j = 0; j < mem->blocks[i]; j++) { priv->memory_pool[curr + j].start = payload_addr; priv->memory_pool[curr + j].size = mem->blocks_size[i]; priv->memory_pool[curr + j].free = 1; -- 2.30.2
