On Fri, Dec 15, 2023 at 02:59:35PM +0100, Francesco Dolcini wrote:
On Fri, Dec 15, 2023 at 02:51:09PM +0100, Johan Hovold wrote:
On Thu, Dec 14, 2023 at 06:01:46PM +0100, Francesco Dolcini wrote:
From: Francesco Dolcini francesco.dolcini@toradex.com
receive_buf() is called from ttyport_receive_buf() that expects values ">= 0" from serdev_controller_receive_buf(), change its return type from ssize_t to size_t.
drivers/bluetooth/btmtkuart.c | 4 ++-- drivers/bluetooth/btnxpuart.c | 4 ++--
-static ssize_t btnxpuart_receive_buf(struct serdev_device *serdev,
const u8 *data, size_t count)+static size_t btnxpuart_receive_buf(struct serdev_device *serdev,
const u8 *data, size_t count){ struct btnxpuart_dev *nxpdev = serdev_device_get_drvdata(serdev);
A quick check of just the first two functions here shows that they can return negative values.
This is already fixed. Patches are in next.
There were 3 buggy user of this API.
- 1 patch was merged a few days ago in mainline
- 2 patches are in next, the maintainer decided to wait for the next merge window
commit 687d2de93b11 ("Bluetooth: btmtkuart: fix recv_buf() return value") commit 7954bbcdd7ea ("Bluetooth: btnxpuart: fix recv_buf() return value") commit c8820c92caf0 ("platform/surface: aggregator: fix recv_buf() return value")
Then why was that not mentioned in the patch (e.g. below the --- line)?
You are certainly not making it easy for reviewers, but good to see that you thought about this, and I see now when reviewing the mail archives that those bugs were the reason for this patch in the first place.
Johan