Re: [PATCH wireless-next v2] wifi: brcmfmac: Add support for BCM4361 PCIe chipset

On 05/08/2026 15:56, Markuss Broks wrote:
> Hi Arend,
> 
> On 8/5/26 8:46 AM, Arend van Spriel wrote:
>> On 04/08/2026 16:07, Markuss Broks via B4 Relay wrote:
>>> From: Markuss Broks <[email protected]>
>>>
>>> Those are the required changes for the BCM4361 PCIe chipset support.
>>
>> Hi Markuss,
>>
>> That is a brief commit message. Maybe you can provide a bit of context 
>> for this patch answering these questions: Are you using this device in 
>> some project? What platform are you using (if any)? What works and 
>> what does not?
> I was just looking at similar device additions in brcmfmac, and noticed 
> they didn't have much of a description, so I decided not to add much 
> either. Should I add some general info about the BCM4361 into the commit 
> message?
> 
> BCM4361 is used on the Samsung Galaxy S9 device (samsung-starlte), based 
> on Exynos9810. There it's connected via PCIe link. Not sure if perhaps 
> there's SDIO or USB or other variants of this chipset. Everything 
> "basic" seems to work, seems to connect to both 2.4GHz and 5GHz 
> networks, download speed looks good. Haven't tried station mode, but I'm 
> not sure what's the proper way to set it up anyway.

Mentioning the platform you tested these patches on is sufficient. So 
when you say "connect to both 2.4GHz and 5GHz networks" I somehow 
suspect you already covered station mode.
  >>
>>> Signed-off-by: Markuss Broks <[email protected]>
>>> ---
>>> Changes in v2:
>>> - EDITME: describe what is new in this series revision.
>>> - EDITME: use bulletpoints and terse descriptions.
>>
>> This really does not look very useful. Those EDITME: are for you! As the
>> actual patch did not change you can probably say:
>>   - fix subject format indicating wifi subsystem and wireless-next repo.
> 
> Oh, sorry, I did not notice that b4 puts the changelog there if it's not 
> in the cover letter. It's been a while since I last submitted 
> something :) I will add the changelog there for v2 and v3 once I submit 
> a v3.

Sure. No problem.

>>
>>> - Link to v1: https://patch.msgid.link/20260803-broadcom-staging- 
>>> [email protected]
>>> ---
>>> drivers/net/wireless/broadcom/brcm80211/brcmfmac/chip.c       | 3 +++
>>> drivers/net/wireless/broadcom/brcm80211/brcmfmac/pcie.c       | 3 +++
>>> drivers/net/wireless/broadcom/brcm80211/include/brcm_hw_ids.h | 2 ++
>>> 3 files changed, 8 insertions(+)
>>
>> [...]
>>
>>> diff --git a/drivers/net/wireless/broadcom/brcm80211/brcmfmac/pcie.c 
>>> b/drivers/net/wireless/broadcom/brcm80211/brcmfmac/pcie.c
>>> index 55f4d7b970f2..914ae72a7b74 100644
>>> --- a/drivers/net/wireless/broadcom/brcm80211/brcmfmac/pcie.c
>>> +++ b/drivers/net/wireless/broadcom/brcm80211/brcmfmac/pcie.c
>>> @@ -59,6 +59,7 @@ BRCMF_FW_CLM_DEF(43570, "brcmfmac43570-pcie");
>>> BRCMF_FW_DEF(4358, "brcmfmac4358-pcie");
>>> BRCMF_FW_DEF(4359, "brcmfmac4359-pcie");
>>> BRCMF_FW_DEF(4359C, "brcmfmac4359c-pcie");
>>> +BRCMF_FW_CLM_DEF(4361, "brcmfmac4361-pcie");
>>
>> Are you sure the CLM blob is required? Can you test without it being 
>> present in /lib/firmware?
> I have tried without the CLM blob, it says "no clm blob available, 
> device might have limited channels available" but seems to at least 
> somewhat work, I haven't noticed a difference. It's still a good idea to 
> load it probably, right?

Right. I was just curious whether it has a working CLM blob compiled in 
firmware or not.

>>
>>> BRCMF_FW_CLM_DEF(4364B2, "brcmfmac4364b2-pcie");
>>> BRCMF_FW_CLM_DEF(4364B3, "brcmfmac4364b3-pcie");
>>> BRCMF_FW_DEF(4365B, "brcmfmac4365b-pcie");
>>
>> [...]
>>
>>> diff --git a/drivers/net/wireless/broadcom/brcm80211/include/ 
>>> brcm_hw_ids.h b/drivers/net/wireless/broadcom/brcm80211/include/ 
>>> brcm_hw_ids.h
>>> index df3b67ba4db2..b800cbfdece1 100644
>>> --- a/drivers/net/wireless/broadcom/brcm80211/include/brcm_hw_ids.h
>>> +++ b/drivers/net/wireless/broadcom/brcm80211/include/brcm_hw_ids.h
>>
>> [...]
>>
>>> @@ -86,6 +87,7 @@
>>> #define BRCM_PCIE_43602_2G_DEVICE_ID 0x43bb
>>> #define BRCM_PCIE_43602_5G_DEVICE_ID 0x43bc
>>> #define BRCM_PCIE_43602_RAW_DEVICE_ID 43602
>>> +#define BRCM_PCIE_4361_DEVICE_ID 0x441f
>>
>> Please add:
>> +#define BRCM_PCIE_4361_2G_DEVICE_ID 0x4420
>> +#define BRCM_PCIE_4361_5G_DEVICE_ID 0x4421
> 
> Are those some special variants of the chipset?

Not really. The difference is in the radio frontend. 0x441f is a 
dual-band. The other two have only a single band so the board gets a bit 
cheaper. The device id is programmed in on-chip OTP so firmware knows 
what type of board it is running on.

Regards,
Arend