Generally speaking, lot of style issues.

When used with latest DTB we still face panic, however related to event timer:

There is no visible benefits so far from that change apart from more efficient usage of descriptors. Abandoning it for now.

@kd Feel free to commit it - I have no particular preferences on committer.

Bump date for ifconfig.8

Adjust ifconfig.8 after review.

Update according to comments. IDMAC_MAX_SIZE will be changed in another MR.

Update patch according to latest comments.

@manu: The handling of bouncing buffers itself is correct in this case. I did not observe any issues in sys/arm64/arm64/busdma_bounce.c. However this is needed, because currently it returns invalid value: instead of making reservation for 2 segments it makes reservation for 3/4 segment which in turns leads to crashes in bouncing buffers code. Comment above the line is valid - for sure we need to reserve two segments (it works for one reserved segment too, but I am not sure about all edge cases), but we need to correctly reserve 2 segments, not 3 sectors.

@mmel I think we need to do something with this review. We have at least one bug report: https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=256822 (similar to what I observe).

Works nice on my RockPro64 board. Thanks!

I performed again test with

Superseded by https://reviews.freebsd.org/D30387

Remove redundant space.

Submitted by: Konrad Sewiłło-Jopek

Provide different dbg messages for different types of errors.

Actually, we don't need to use errno_copy in first occurence of described problem.

Fix similar problems in different locations. Save original errno value from function it originally generated it.

Update patch accordingly to comments.

Oh, that means the current (in repo) code is wrong anyway. close(2) should be then called just before return.

Fixed calibration issue: we should add one tick, not substract.

Rework code to remove workarounds, remove clock definition, be more precise on calibration.

Add calibration.

your patch works perfectly. BladeRF is usable again.

Maybe I will share results from my investigation.

Regarding sysctl interface: allwinner driver has data cells hardcoded - should I follow the same way and add hardcoded values to this driver also? Or maybe there is some more elegant way to do so?

Rework patch (part 1 of 2):

This MR solved problem with eMMC cards on RK3399. Previously card slot was not initialized correctly due to some DMA mapping issues.

Fix interrupt handling: we should clear interrupt bit only if we want to tap watchdog. Minor fixes to build system, remove clock configuration in RK3399 - this should be done in separated commit.

@hselasky
I think anyone with src commit bit (at least I don't have one). dmesg output is attached at the end of my comment.
Also output of usbconfig:

root@1e66c783a77f:~ # usbconfig -d 0.2 dump_device_desc
ugen0.2: <Realtek 802.11ac WLAN Adapter> at usbus0, cfg=0 md=HOST spd=HIGH (480Mbps) pwr=ON (500mA)