Re: [PATCH v1 1/3] Bluetooth: btnxpuart: fix recv_buf() return value

[Jiri Slaby]

On 27. 11. 23, 20:14, Francesco Dolcini wrote:
> From: Francesco Dolcini <francesco.dolcini@xxxxxxxxxxx>
>
> Serdev recv_buf() callback is supposed to return the amount of bytes
> consumed, therefore an int in between 0 and count.
>
> Do not return negative number in case of issue, just print an error and
> return count.  This fixes a WARN in ttyport_receive_buf().
>
> [    9.962266] Bluetooth: hci0: Frame reassembly failed (-84)
> [    9.972939] ------------[ cut here ]------------
> [    9.977922] serial serial0: receive_buf returns -84 (count = 6)
> [    9.994857] WARNING: CPU: 0 PID: 37 at drivers/tty/serdev/serdev-ttyport.c:37 ttyport_receive_buf+0xd8/0xf8
> [   10.004840] Modules linked in: mwifiex_sdio(+) mwifiex snd_soc_simple_card crct10dif_ce cfg80211 snd_soc_simple_card_utils k3_j72xx_bandgap rti_wdt rtc_ti_k3 btnxpuart bluetooth sa2ul ecdh_generic ecc sha256_generic tidss rfkill libsha256 drm_dma_helper snd_soc_davinci_mcasp authenc omap_mailbox snd_soc_ti_udma snd_soc_ti_edma snd_soc_ti_sdma atmel_mxt_ts ina2xx snd_soc_nau8822 ti_sn65dsi83 tc358768 ti_ads1015 tps65219_pwrbutton at24 m_can_platform industrialio_triggered_buffer drm_kms_helper m_can kfifo_buf rtc_ds1307 lm75 pwm_tiehrpwm can_dev spi_omap2_mcspi panel_lvds pwm_bl libcomposite fuse drm backlight ipv6
> [   10.059984] CPU: 0 PID: 37 Comm: kworker/u4:2 Not tainted 6.7.0-rc2-00147-gf1a09972a45a #1
> [   10.071793] Hardware name: Toradex Verdin AM62 WB on Verdin Development Board (DT)
> [   10.082898] Workqueue: events_unbound flush_to_ldisc
> [   10.091345] pstate: 60000005 (nZCv daif -PAN -UAO -TCO -DIT -SSBS BTYPE=--)
> [   10.101820] pc : ttyport_receive_buf+0xd8/0xf8
> [   10.109712] lr : ttyport_receive_buf+0xd8/0xf8

From here:

> [   10.117581] sp : ffff800082b9bd20
> [   10.124202] x29: ffff800082b9bd20 x28: ffff00000000ee05 x27: ffff0000002f21c0
> [   10.134735] x26: ffff000002931820 x25: 61c8864680b583eb x24: ffff0000002f21b8
> [   10.145209] x23: ffff00000026e740 x22: ffff0000002f21e0 x21: ffffffffffffffac
> [   10.155686] x20: ffff000000da5c00 x19: 0000000000000006 x18: 0000000000000000
> [   10.166178] x17: ffff7fffbe0e7000 x16: ffff800080000000 x15: 000039966db1c650
> [   10.176564] x14: 000000000000022c x13: 000000000000022c x12: 0000000000000000
> [   10.186979] x11: 000000000000000a x10: 0000000000000a60 x9 : ffff800082b9bb80
> [   10.197352] x8 : ffff00000026f200 x7 : ffff00003fd90080 x6 : 00000000000022e5
> [   10.207680] x5 : 00000000410fd030 x4 : 0000000000c0000e x3 : ffff7fffbe0e7000
> [   10.218051] x2 : 0000000000000002 x1 : 0000000000000000 x0 : 0000000000000000

Please trim this. No need to dump registers into the commit log. Also 
the module list is usually not so useful. You can prune everything after 
mwifiex.

> [   10.228393] Call trace:
> [   10.233989]  ttyport_receive_buf+0xd8/0xf8
> [   10.241224]  flush_to_ldisc+0xbc/0x1a4
> [   10.248117]  process_scheduled_works+0x16c/0x28c

And these are as well not interesting:

> [   10.255851]  worker_thread+0x16c/0x2e0
> [   10.262673]  kthread+0x11c/0x128
> [   10.268953]  ret_from_fork+0x10/0x20
> [   10.275460] ---[ end trace 0000000000000000 ]---

  ^^^^^^^^^^^^^^
So are not the timestamps.

> Closes: https://lore.kernel.org/all/ZWEIhcUXfutb5SY6@xxxxxxxxxxxxxxxxxxxxxxxxxxxx/
> Fixes: 689ca16e5232 ("Bluetooth: NXP: Add protocol support for NXP Bluetooth chipsets")
> Signed-off-by: Francesco Dolcini <francesco.dolcini@xxxxxxxxxxx>
> ---
>   drivers/bluetooth/btnxpuart.c | 7 +++----
>   1 file changed, 3 insertions(+), 4 deletions(-)
>
> diff --git a/drivers/bluetooth/btnxpuart.c b/drivers/bluetooth/btnxpuart.c
> index b7e66b7ac570..951fe3014a3f 100644
> --- a/drivers/bluetooth/btnxpuart.c
> +++ b/drivers/bluetooth/btnxpuart.c
> @@ -1276,11 +1276,10 @@ static int btnxpuart_receive_buf(struct serdev_device *serdev, const u8 *data,
>   	if (IS_ERR(nxpdev->rx_skb)) {
>   		int err = PTR_ERR(nxpdev->rx_skb);
>   		/* Safe to ignore out-of-sync bootloader signatures */
> -		if (is_fw_downloading(nxpdev))
> -			return count;
> -		bt_dev_err(nxpdev->hdev, "Frame reassembly failed (%d)", err);
> +		if (!is_fw_downloading(nxpdev))
> +			bt_dev_err(nxpdev->hdev, "Frame reassembly failed (%d)", err);
>   		nxpdev->rx_skb = NULL;

Is this NULLing not needed in the good case?

> -		return err;
> +		return count;

Should you return 0? I don't know, maybe not, but you should document it 
in the commit log.

thanks,
--
js
suse labs