In D13431#400286, @jbeich wrote:

@johalun0_gmail.com, does rP489493 work for you?

Yeah, Sway is on hold until we can find time to get the new version that is based on wlroots ported. I might give this a shot during the holidays. Marking this as abandoned.

I created the necessary git branches to point these ports to. It would be good if we could let 11.2 use 4.11 so we can scrap 4.9 all together (and thus remove some legacy code from lkpi).

Thanks for doing this! Works great on my Dell 7490.
Suspend/resume not tested. Please also add EVDEV to this driver so we can get things like horizontal scrolling :)

Fix comments.

Changes: Disable phys_avail until KPI is in.

In D16660#376951, @imp wrote:

So why the special case for vgapci0?

In D16660#376945, @imp wrote:

I have no clue what this is trying to fix.
I am aware of a bug that we need to read all the files in the path (which part of this does), but the special case for vgapci0 seems unwarranted and dangerous.
So there's some relevance to this, but I have some concerns.

In D16660#376108, @hselasky wrote:

Ping is this patch still relevant?

Patch srat.c to use phys_avail_reserve()

physmem.h has moved from machine/ to x86/

phys_avail_reserve has been moved to a separate diff https://reviews.freebsd.org/D17431 and this diff should be applied after D17431.

In D17267#368025, @yuripv_yuripv.net wrote:

In D17267#367979, @johalun0_gmail.com wrote:

Does this require any changes to devd.conf to work? Only applying this patch and copy over devd binary to my test machine doesn't change anything for me, the regexp still starts with ^$'(.....

Weird, works for me, and regex in Testing device-name line looks correct now without any changes to devd.conf. Make sure you are running correct binary?

Does this require any changes to devd.conf to work? Only applying this patch and copy over devd binary to my test machine doesn't change anything for me, the regexp still starts with ^$'(.....

Yeah, CK work was committed soon after so_reuseport_lb was merged so this was missed. Thanks for fixing!

Thanks for cleaning this up.

LGTM and I assume there's another patch coming that will fix the loader issues?

Thanks for finding and fixing this!

Thank you <3 This will make life easier for us drm developers.

In D16719#355599, @imp wrote:

This strikes me as two different changes. One for the phys_avail_reserve() and another to use it for the pci early stuff. I'm a little uneasy with the pci early name, but the code is clearly PCI specific.
Would there ever be a need to do things like this when running the 32-bit kernel? I know we don't support graphics cards there, but when we're bringing in new interfaces, it's a question that suggests itself.
RPi has this sort of thing too, but it's dealt with in the bowels of the FDT code...

Results with I219LM (em0) TX and Macbook thunderbolt dongle (bge0) RX

I tried the first version again and it seems OK on the TX side. Not sure about RX, seeing some poll error but that might be my messed up I218LM which can't do more than 70 Kpps TX (it does 14 Mpps with FreeBSD 11) but that's another problem.
Between I219LM and a Realtek NIC I get stable 150 Kpps and I guess that's the limit for emulated netmap.

In D16300#346610, @shurd wrote:

In D16300#346609, @johalun0_gmail.com wrote:

OK, switching TX/RX between I128LM and I219LM makes no difference, still 5 Kpps. iperf between the two machine measure +900 Mbps.

Are the counters in dev.em.X.mac_stats.xoff_txd and dev.em.X.mac_stats.xon_txd incrementing? If so, you can try turning off flow control... sysctl dev.em.X.fc=0

OK, switching TX/RX between I128LM and I219LM makes no difference, still 5 Kpps. iperf between the two machine measure +900 Mbps.

In D16300#346577, @shurd wrote:

Use the same callout reset logic in the admin task as the txq timer.

@johalun0_gmail.com I wonder if the admin task was overriding the txqs
with a longer timeout in your case.

Test between two 1000baseT FD, connected with 1GB switch. I128LM is sending, I129LM is receiving.
Kernel from 2018-07-18 (how to find svn revision when building from github?)
Patched kernel only on sending side. Receiving side is same revision but unpatched.

Committed.

Rebase diff

Add license and clean up some reorder diffs

Rebase to r334720

Rebase against today's head

Some stuff has been committed so rebase against head.

Don't increment index in radix_tree_iter_delete

Update since some parts have been committed to head.
Also,

• Remove fcheck() macro, patch driver source code instead to use fget() and fput().
• Derive PAGE_MASK from PAGE_SIZE since PAGE_MASK is different on FreeBSD and Linux

Remove whitespace change.

Updated diff against today's kernel.

Revert change to MPASS test in scheduler code.

Added TASK_KILLABLE to TASK_NORMAL so that the assert don't cause panic in linux_add_to_sleepqueue().

In D15567#328787, @cem wrote:

I don't have any doc for the new Zen APUs, sorry. My TR doesn't have any 15eb devices.

Have you addressed all of @rwatson's comments?

And netmap is alive again on I218-LM :)

• Omit the inheritance part for now so that we can get this committed. Need more time and eyes to figure out a safe way to solve the locking issues. (this is anyway not a critical function but more a nice-to-have thing)
• Add leading zeros to the rest of the SO_* options to indicate that they are now 32 bit.

In D15385#324410, @hselasky wrote:

Looks good. Approved!

In D15385#324408, @hselasky wrote:

Did you check if this driver is already supported by webcamd?

@rwatson If you have time, I would appreciate if you could take a look at the updates.

Added comment - unsure about locking in one place...

Improving locking
Refactor
Fix style

Add entry in man page

Full context for all files.

@gnn @rwatson Are we good to go?

Thanks for the feedback!

Changes:

• Fix style
• Add lock assertions
• Acquire INP_HASH_RLOCK in_pcblookup_lbgroup_last()
• Acquire INP_WLOCK before calling soinherit()

Update to apply cleanly to HEAD as of 20180420.

Together with r331692 this issue is solved.

Committed.

Everything in HEAD now.

Minor style fixes

Simplify tasklet_enable function.

Fix comments and various style fixes.

Fixed some comments. Will update the diff when the last one is fixed.
@hselasky please check my comments in list.h and wait.h.

@hselasky
Thanks for the feedback! I will look at the issues. Please look at my replies to your comments.
The work was started before the split of linuxkpi so merging it back into the two different parts is somewhat messy...

Not relevant anymore.