centralize VM_BATCHQUEUE_SIZE definition & increase size on arm64

[gallatin]

Summary

Remove the platform-specific definitions of VM_BATCHQUEUE_SIZE
for amd64 and powerpc64, and instead treat all 64-bit platforms
identically. This has the effect of increasing the arm64
VM_BATCHQUEUE_SIZE to match that of other platforms.

Note that I was about to create a review to increase the batchqueue size on arm64, and then I realized that it all the 64-bit platforms would be the same, so it would be easier to simply define everything in one place, rather than scattered in multple vmparam.h files.

[jhb]

sys/vm/vm_pagequeue.h

77–79

Do we need to keep the #ifndef now? It seems like we can probably remove it?

[markj]

sys/vm/vm_pagequeue.h

78

I think it would be preferable to invert this and test for ILP32 instead? As far as I know CHERI doesn't define LP64 but we'd want to treat it the same as arm64.

[jhb]

sys/vm/vm_pagequeue.h

78

In CheriBSD we tend to check __SIZEOF_LONG__ == 8 rather than __LP64__.

[gallatin]

Updated patch to use SIZEOF_LONG, based on review feedback from @jhb and @markj

[gallatin]

Remove now unneeded ifndef VM_BATCHQUEUE_SIZE as suggested by @jhb

[markj]

sys/vm/vm_pagequeue.h

80

Just a comment: now that we drain per-CPU queues whenever we successfully obtain the queue lock, I think this comment about having "lots of memory" is a bit misleading. (And really it ought to say "lots of memory per CPU" if anything.) I think you could just delete the comment.