Increment v_pdpages in the laundry queue scan.

[markj]

Summary

Also fix some inconsistencies between the different scan loops:

• Increment pdpages for every page encountered in the scan, including markers. Markers should not be common enough that this makes a difference.
• Leave newlines around the assertions to make the code a bit easier to see.
• Remove some unhelpful comments.

[alc]

I've never found pdpages to be that useful in understanding system behavior. Could I talk you into creating distinct counters for the inactive and laundry queues? And, in regards to the active queue, I'd like to be able to distinguish between idle scanning and shortage-driven scanning.

P.S. This counter also gets updated by the pseudo-scan in vm_swapout.c.

[markj]

In D14666#308086, @alc wrote:

I've never found pdpages to be that useful in understanding system behavior.

I do occasionally find it handy to watch pdpages in "systat -vm" or "vmstat 1" output.

Could I talk you into creating distinct counters for the inactive and laundry queues? And, in regards to the active queue, I'd like to be able to distinguish between idle scanning and shortage-driven scanning.

Sure. I guess we effectively want one counter per (per-domain) queue?

[alc]

In D14666#308107, @markj wrote:

In D14666#308086, @alc wrote:

Could I talk you into creating distinct counters for the inactive and laundry queues? And, in regards to the active queue, I'd like to be able to distinguish between idle scanning and shortage-driven scanning.

Sure. I guess we effectively want one counter per (per-domain) queue?

Yes.

[markj]

Add a per-queue pdpages counter.

Per-queue sysctls are added under vm.domain; the vm.stats.vm.v_pdpages
sysctl is preserved.

[alc]

I would suggest excluding markers from the count. Otherwise, this looks great.

[alc]

I would love to see this committed before we branch for 12.0.

[markj]

In D14666#349564, @alc wrote:

I would suggest excluding markers from the count. Otherwise, this looks great.

This change was partially subsumed by the batch queue patch: now we count pdpages for all queue scans. (We do currently include markers in the count, though.)

In D14666#359454, @alc wrote:

I would love to see this committed before we branch for 12.0.

As I write this, I see that I never submitted my last comment, which is part of the reason I've been dawdling. I'll get it in in the next few hours though.