[dpdk-dev] Minutes of Technical Board Meeting, 2021-Nov-03

[Maxime Coquelin <maxime.coquelin@redhat.com>]

Minutes of Technical Board Meeting, 2021-Nov-03

Members Attending
-----------------
-Aaron
-Ferruh
-Hemant
-Honnappa
-Jerin
-Kevin
-Konstantin
-Maxime (Chair)
-Olivier
-Stephen
-Thomas


NOTE: The technical board meetings every second Wednesday at 
https://meet.jit.si/DPDK at 3 pm UTC.
Meetings are public, and DPDK community members are welcome to attend.

NOTE: Next meeting will be on Wednesday 2021-Nov-17 @3pm UTC, and will 
be chaired by Olivier.

# ENETFEC driver
- TB discussed whether depending on an out-of-tree Kernel module is 
acceptable
-- TB voted to accept that ENETFEC PMD relies on out-of-tree kernel module
-- TB recommends avoiding out-of-tree Kernel modules, but the kernel 
module required by the ENETFEC PMD is in the process of being upstreamed
- TB discussed whether having this drivers as a VDEV is acceptable or if 
a bus driver is required, knowing that only this device would use it
-- TB voted to accept this driver as a VDEV

# IP frag ABI change in v21.11 [0]
- This ABI change was not announced so TB approval was required
-- TB voted to accept this ABI change

# Communication plan around v21.11 release
- Thomas highlighted that a lot of changes are being introduced in 
v21.11 release.
- In addition to the usual release blog post, blog posts about specific 
new features would be welcomed
-- TB calls for ideas to maintainers and contributors

# Feedback from Governing Board on proposal for technical board process 
updates
- Honnappa proposes a new spreadsheet to improve the communication 
between the technical and governing boards

# L3 forward mode in testpmd [1]
- Honnappa presented the reasons of this new forwarding mode
-- L3FWD is a standard benchmark for DPDK
-- L3FWD example misses debugging features present in testpmd
- Concerns raised about code duplication and bloating of testpmd
- Suggestions that adding more statistics and interactive mode to L3 FWD 
would be preferable
-- But concerns that it would make this application too much complex, 
defeating the initial purpose of this example
- As no consensus has been reached, Honnappa proposed to reject/defer it 
for now

# DMARC configuration
- Ali monitored the DMARC configuration changes done on the user and web 
mailing lists
-- Better results have been observed
- TB voted to apply the new policy to the other mailing lists
- Ali will apply the new policy by the end of next week

# Patch from AMD to raise the maximum number of lcores
- Ran out of time, adding this item to the next meeting

[0]: 
https://patches.dpdk.org/project/dpdk/patch/20211102190309.5795-1-konstantin.ananyev@intel.com/
[1]: 
https://patchwork.dpdk.org/project/dpdk/patch/20210430213747.41530-2-kathleen.capella@arm.com/