Hi, my last week was almost exclusively spent on
- fwget(8) updates, which were then requested to be broken into flavors (D44918). While I was reluctant to do it now as there's too much on the TODO and it dodn't seem as important as continueing first to move firmware out of base; in hindsight it was the right request given now it is done and we don't have to start all over again. - With further updates it was suggested that these updates could be generated automatically. I set down and went through each of the drivers again enhancing the mostly single-line helpers I had into full scripts doing extra checks and dealing with "special" parts of each. These scripts are "not beautiful" but "do the job" as the commit message indicates. They are currently on github: https://github.com/freebsd/wireless/commit/aa03599103d7f7843ce1f0d0b38eadb4fbd29384 The https://github.com/freebsd/wireless/tree/bz_wifi_drv_update branch also includes the updats of fwget(8): https://github.com/freebsd/wireless/commit/04d43be2cfb3b8d9ed9b3f3a568a1a8fd40be224 - Updates to the wifi-firmware-$name-kmod ports and adding iwlwifi. Flavors as requested are now supported as well and needed the extra round-trip to get right (I am not a ports person yet). The entire distfiles/flavor sections are now also generated from the scripts mentioned above so updates in the future should be like drivers--doable in a short time. The update is in D44945 and jrm has been helpful reviewing. - For iwlwifi the script also generate the iwlwififw(4) table (and the one for the wiki page https://wiki.freebsd.org/WiFi/Iwlwifi/Chipsets which got updated) - I also started to look at bsdinstall and the release building process in order to get fwget(8) support into it and also on how to get the firmware packages installed into an installation media (likely memstick or a varian thereof). - Anyone wanting to test fwget installing the correct flavor of firmware for their card will be more than welcome. At least please keep an eye out for when these changes land (likely along the driver update to main). Testing with more than just 1 chipset per driver is time-consuming so any feedback will help. - There was very little feedback on any of the driver updates on github. Thanks for the people testing rtw8x. I had one request for a stable/14 based branched. Is the conclusion that things need to be in the tree for people to have a look? - The usual bits on some reviews for LinuxKPI etc. and some emails which did not hit the list. I'll be less available/responsive for the next week given Eastern Easter is this weekend. Lots of health and joy, Bjoern -- Bjoern A. Zeeb r15:7
