gnrc_sixlowpan_frag: enable gnrc_netif_pktq if netdev_new_api is used

[benpicco]

Contribution description

6lowpan fragmentation is broken on at86rf215 since #20672. The driver no longer tries to block when a transfer is ongoing, but reports -EBUSY to the upper layer when a send is triggered while the previous frame is still being sent.

For this to work (and not drop the 2nd fragment), we need gnrc_netif_pktq.

Testing procedure

6lo fragmented packets are working again

2024-09-30 12:00:50,928 # > ping -s 512 fe80::6419:e83f:8e6b:4ce7%7
2024-09-30 12:00:51,042 # 520 bytes from fe80::6419:e83f:8e6b:4ce7%7: icmp_seq=0 ttl=64 rssi=-40 dBm time=108.982 ms
2024-09-30 12:00:52,050 # 520 bytes from fe80::6419:e83f:8e6b:4ce7%7: icmp_seq=1 ttl=64 rssi=-40 dBm time=108.986 ms
2024-09-30 12:00:53,044 # 520 bytes from fe80::6419:e83f:8e6b:4ce7%7: icmp_seq=2 ttl=64 rssi=-39 dBm time=108.983 ms
2024-09-30 12:00:53,045 # 
2024-09-30 12:00:53,058 # --- fe80::6419:e83f:8e6b:4ce7%7 PING statistics ---
2024-09-30 12:00:53,061 # 3 packets transmitted, 3 packets received, 0% packet loss
2024-09-30 12:00:53,063 # round-trip min/avg/max = 108.982/108.983/108.986 ms

Issues/PRs references

[riot-ci]

Murdock results

✔️ PASSED

0b99d4f gnrc_sixlowpan_frag: enable gnrc_netif_pktq if netdev_new_api is used

Success	Failures	Total	Runtime
10197	0	10197	13m:50s

Artifacts

• Documentation preview

[miri64]

6lowpan fragmentation is broken on at86rf215 since #20672.

How does link to gnrc_lorawan? The device is an IEEE 802.15.4 one and with LoRaWAN SCHC is recommended instead of 6LoWPAN.

[benpicco]

Oops, copy & paste error on the commit title

[maribu]

Thx for spotting this!