[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <20250714154608.GP721198@horms.kernel.org>
Date: Mon, 14 Jul 2025 16:46:08 +0100
From: Simon Horman <horms@...nel.org>
To: Pauli Virtanen <pav@....fi>
Cc: linux-bluetooth@...r.kernel.org, marcel@...tmann.org,
johan.hedberg@...il.com, luiz.dentz@...il.com, davem@...emloft.net,
edumazet@...gle.com, kuba@...nel.org, pabeni@...hat.com,
netdev@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] Bluetooth: ISO: add socket option to report packet
seqnum via CMSG
On Mon, Jul 14, 2025 at 05:02:57PM +0300, Pauli Virtanen wrote:
> User applications need a way to track which ISO interval a given SDU
> belongs to, to properly detect packet loss. All controllers do not set
> timestamps, and it's not guaranteed user application receives all packet
> reports (small socket buffer, or controller doesn't send all reports
> like Intel AX210 is doing).
>
> Add socket option BT_PKT_SEQNUM that enables reporting of received
> packet ISO sequence number in BT_SCM_PKT_SEQNUM CMSG.
>
> Signed-off-by: Pauli Virtanen <pav@....fi>
Hi Pauli,
Some minor feedback from my side.
The byte order annotations around the sequence number seem inconsistent.
And my guess is that __le16 should be consistently used to hold the sequence
number.
Sparse says:
net/bluetooth/iso.c:2322:28: warning: incorrect type in assignment (different base types)
net/bluetooth/iso.c:2322:28: expected unsigned short [usertype] sn
net/bluetooth/iso.c:2322:28: got restricted __le16 [usertype] sn
net/bluetooth/iso.c:2333:28: warning: incorrect type in assignment (different base types)
net/bluetooth/iso.c:2333:28: expected unsigned short [usertype] sn
net/bluetooth/iso.c:2333:28: got restricted __le16 [usertype] sn
Powered by blists - more mailing lists