[OpenWrt-Devel] [PATCH] ath79: Add SUPPORTED_DEVICES for Archer C7 v1/v2

[David Bauer]

Hello Christian,

On 21.04.19 14:17, Christian Lamparter wrote:
> Hello David,
> 
> On Sunday, April 21, 2019 11:42:52 AM CEST David Bauer wrote:
>> On 20.04.19 20:59, Christian Lamparter wrote:
>>> On Wednesday, April 17, 2019 3:45:52 PM CEST Adrian Schmutzler wrote:
>>>> The identifier for both devices is "archer-c7" on ar71xx, set here:
>>>> https://github.com/openwrt/openwrt/blob/master/target/linux/ar71xx/base-files/lib/ar71xx.sh#L348
>>>> https://github.com/openwrt/openwrt/blob/master/target/linux/ar71xx/base-files/lib/ar71xx.sh#L511
>>>>
>>>> Signed-off-by: Adrian Schmutzler <freifunk at adrianschmutzler.de>
>>>> ---
>>>>   target/linux/ath79/image/generic-tp-link.mk | 2 ++
>>>>   1 file changed, 2 insertions(+)
>>>>
>>>> diff --git a/target/linux/ath79/image/generic-tp-link.mk b/target/linux/ath79/image/generic-tp-link.mk
>>>> index 6853f12341..db1eabd420 100644
>>>> --- a/target/linux/ath79/image/generic-tp-link.mk
>>>> +++ b/target/linux/ath79/image/generic-tp-link.mk
>>>> @@ -70,6 +70,7 @@ define Device/tplink_archer-c7-v1
>>>>     DEVICE_TITLE := TP-Link Archer C7 v1
>>>>     DEVICE_PACKAGES := kmod-usb-core kmod-usb2 kmod-usb-ledtrig-usbport kmod-ath10k-ct ath10k-firmware-qca988x-ct
>>>>     TPLINK_HWID := 0x75000001
>>>> +  SUPPORTED_DEVICES += archer-c7
>>>>   endef
>>>>   TARGET_DEVICES += tplink_archer-c7-v1
>>>>   
>>>> @@ -79,6 +80,7 @@ define Device/tplink_archer-c7-v2
>>>>     DEVICE_TITLE := TP-Link Archer C7 v2
>>>>     DEVICE_PACKAGES := kmod-usb-core kmod-usb2 kmod-usb-ledtrig-usbport kmod-ath10k-ct ath10k-firmware-qca988x-ct
>>>>     TPLINK_HWID := 0xc7000002
>>>> +  SUPPORTED_DEVICES += archer-c7
>>> In case of the v2, I think there's still the problem that a straight up upgrade
>>> from ar71xx to ath79 will affect the 5GHz ath10k wireless because it now has a
>>> new device path and hence a new default configuration (where the card is
>>> disabled) is created.
>>
>> I recall upgrading my OCEDO Koala (which uses the same 9558/9880 combo)
>> from ar71xx to ath79 and the PCIe path being consistent on both platforms.
>>
>> This however might have changed in the meantime, so someone should
>> probably confirm this with a real C7.
> 
> On my C7 v1 with a QCA9880v2 the ar71xx installation back in
> 2018-08-17 looked like this:
> 
> config wifi-device 'radio0'
>          option type             'mac80211'
>          option country          'DE'
>          option channel          'auto'
>          option hwmode           '11g'
>          option path             'platform/qca955x_wmac'
>          option htmode           'HT20'
>          option disabled         '0'
>          option txpower          '10'
> 
> config wifi-device 'radio1'
>          option type             'mac80211'
>          option channel          '52'
>          option country          'DE'
>          option hwmode           '11a'
>          option path             'pci0000:01/0000:01:00.0'
>          option htmode           'VHT80'
>          option disabled         '0'
>          option txpower          '14'
> 
> vs ath79 (today):
> 
> config wifi-device 'radio0'
>          option type             'mac80211'
>          option country          'DE'
>          option channel          'auto'
>          option hwmode           '11g'
>          option path             'platform/ahb/ahb:apb/18100000.wmac'
>          option htmode           'HT20'
>          option disabled         '0'
>          option txpower          '10'
> 
> config wifi-device 'radio1'
>          option type             'mac80211'
>          option channel          '52'
>          option country          'DE'
>          option hwmode           '11a'
>          option path             'pci0000:00/0000:00:00.0'
>          option htmode           'VHT80'
>          option disabled         '0'
>          option txpower          '14'
> 
> so the path changed from "pci0000:01/0000:01:00.0" to
> "pci0000:00/0000:00:00.0". But again this is on a C7 v1.
> 
> Based on the bootlog on the wiki for 18.06.1 :
> https://openwrt.org/toh/tp-link/archer-c7-1750#boot_logs
> The ar71xx image enabling both pcie Root Complexes of the QCA955x.
> But unfortunately the pcie slot of the C7 is wired to the second RC,
> so the ath10k card gets pci0000:01/0000:01:00.0. Does anybody want to
> test what happens if the ath79 C7 v2 DTS enables "pcie0" too? It might
> work, but it might not (depending on whenever it might end up in a
> different pci domain like pci0001:00.).

Damn, you are right. The Koala has it's only PCIe endpoint wired to the 
first bus, so this works fine.

Regarding enabling the first bus: Personally, I would prefer a migration 
script over enabling a non-wired interface. There is already a migration 
script for exactly this case in the mpc85xx target, so most of this work 
is probably straight up copy-paste ;)

Best wishes
David

_______________________________________________
openwrt-devel mailing list
openwrt-devel at lists.openwrt.org
https://lists.openwrt.org/mailman/listinfo/openwrt-devel