* [PATCH] Bluetooth: btbcm: Add entries for BCM4364B3 UART Bluetooth
@ 2026-11-16 21:38 Jason Perlow
2026-11-22 8:27 ` Paul Menzel
0 siblings, 1 reply; 3+ messages in thread
From: Jason Perlow @ 2026-11-16 21:38 UTC (permalink / raw)
To: Marcel Holtmann, Luiz Augusto von Dentz, linux-bluetooth; +Cc: linux-kernel
The BCM4364B3 combo chip on Apple T2 Macs, e.g. the MacBookPro16,1, has a
UART Bluetooth controller that reports LMP subversion 0x3032 (001.016.050)
on some units and 0x305b (001.016.091) on others. Neither has an entry in
bcm_uart_subver_table, so the controller is logged without a name:
Bluetooth: hci0: BCM4364B3 Trinidad Olympic GEN (MFG)
Bluetooth: hci0: BCM (001.016.091) build 0115
and the firmware lookup only tries the generic "brcm/BCM.hcd".
Add both subversions so the controller is named and its firmware is looked
up as "brcm/BCM4364B3.hcd" first. Tested on two MacBookPro16,1 with a
v7.4-rc3 based kernel carrying an equivalent change, one reporting 0x3032
and one reporting 0x305b. They now log:
Bluetooth: hci0: BCM4364B3 (001.016.050) build 0082
Bluetooth: hci0: BCM4364B3 (001.016.091) build 0115
Signed-off-by: Jason Perlow <jperlow@gmail.com>
---
drivers/bluetooth/btbcm.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/drivers/bluetooth/btbcm.c b/drivers/bluetooth/btbcm.c
index 065536d4ec..92ba7aeac3 100644
--- a/drivers/bluetooth/btbcm.c
+++ b/drivers/bluetooth/btbcm.c
@@ -519,6 +519,8 @@ static const struct bcm_subver_table
bcm_uart_subver_table[] = {
{ 0x2119, "BCM4373A0" }, /* 001.001.025 */
{ 0x2128, "BCM4384A0" }, /* 001.001.040 */
{ 0x4119, "BCM4384B0" }, /* 002.001.025 */
+ { 0x3032, "BCM4364B3" }, /* 001.016.050 */
+ { 0x305b, "BCM4364B3" }, /* 001.016.091 */
{ }
};
-- 2.56.0
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: [PATCH] Bluetooth: btbcm: Add entries for BCM4364B3 UART Bluetooth
2026-11-16 21:38 [PATCH] Bluetooth: btbcm: Add entries for BCM4364B3 UART Bluetooth Jason Perlow
@ 2026-11-22 8:27 ` Paul Menzel
2026-11-22 17:14 ` Jason Perlow
0 siblings, 1 reply; 3+ messages in thread
From: Paul Menzel @ 2026-11-22 8:27 UTC (permalink / raw)
To: Jason Perlow
Cc: Marcel Holtmann, Luiz Augusto von Dentz, linux-bluetooth, linux-kernel
Dear Jason,
Thank you for your patch.
Am 16.11.26 um 22:38 schrieb Jason Perlow:
> The BCM4364B3 combo chip on Apple T2 Macs, e.g. the MacBookPro16,1, has a
> UART Bluetooth controller that reports LMP subversion 0x3032 (001.016.050)
> on some units and 0x305b (001.016.091) on others. Neither has an entry in
> bcm_uart_subver_table, so the controller is logged without a name:
>
> Bluetooth: hci0: BCM4364B3 Trinidad Olympic GEN (MFG)
> Bluetooth: hci0: BCM (001.016.091) build 0115
>
> and the firmware lookup only tries the generic "brcm/BCM.hcd".
>
> Add both subversions so the controller is named and its firmware is looked
> up as "brcm/BCM4364B3.hcd" first. Tested on two MacBookPro16,1 with a
> v7.4-rc3 based kernel carrying an equivalent change, one reporting 0x3032
> and one reporting 0x305b. They now log:
>
> Bluetooth: hci0: BCM4364B3 (001.016.050) build 0082
> Bluetooth: hci0: BCM4364B3 (001.016.091) build 0115
Is this only about the logs, or does functionality also change, that BT
starts working, or did it work before already?
> Signed-off-by: Jason Perlow <jperlow@gmail.com>
> ---
> drivers/bluetooth/btbcm.c | 2 ++
> 1 file changed, 2 insertions(+)
>
> diff --git a/drivers/bluetooth/btbcm.c b/drivers/bluetooth/btbcm.c
> index 065536d4ec..92ba7aeac3 100644
> --- a/drivers/bluetooth/btbcm.c
> +++ b/drivers/bluetooth/btbcm.c
> @@ -519,6 +519,8 @@ static const struct bcm_subver_table
> bcm_uart_subver_table[] = {
> { 0x2119, "BCM4373A0" }, /* 001.001.025 */
> { 0x2128, "BCM4384A0" }, /* 001.001.040 */
> { 0x4119, "BCM4384B0" }, /* 002.001.025 */
> + { 0x3032, "BCM4364B3" }, /* 001.016.050 */
> + { 0x305b, "BCM4364B3" }, /* 001.016.091 */
Sort it?
> { }
> };
With my comments addressed, feel free to add:
Reviewed-by: Paul Menzel <pmenzel@molgen.mpg.de>
Kind regards,
Paul
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: [PATCH] Bluetooth: btbcm: Add entries for BCM4364B3 UART Bluetooth
2026-11-22 8:27 ` Paul Menzel
@ 2026-11-22 17:14 ` Jason Perlow
0 siblings, 0 replies; 3+ messages in thread
From: Jason Perlow @ 2026-11-22 17:14 UTC (permalink / raw)
To: Paul Menzel
Cc: Marcel Holtmann, Luiz Augusto von Dentz, linux-bluetooth, linux-kernel
Hi Paul,
Thanks for looking at it. (Resending as plain text; the first copy of
this reply bounced from the lists because of an HTML part.)
As tested it is about naming and the firmware lookup. The controller
already came up without the entries, but it was logged without a name
and the driver only tried the generic brcm/BCM.hcd. With the entries it
is named BCM4364B3 and brcm/BCM4364B3.hcd is tried first. I only checked
the log output and that the controller still initialises on both units;
I have not compared pairing or throughput before and after, and I am
not aware of a unit where Bluetooth did not work without this. The
commit message now says that explicitly so it does not read as a bug
fix.
You are right about the order. The table is ascending by subversion
here, so v2 puts 0x3032 and 0x305b between 0x2128 and 0x4119.
v2 is already out as a separate mail
(<20261122165132.14015-1-jperlow@gmail.com>) with your Reviewed-by
added, as you offered.
Thanks,
Jason
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2026-11-22 17:14 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2026-11-16 21:38 [PATCH] Bluetooth: btbcm: Add entries for BCM4364B3 UART Bluetooth Jason Perlow
2026-11-22 8:27 ` Paul Menzel
2026-11-22 17:14 ` Jason Perlow
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
all inboxes | Powered by JetHome®