On (24/01/15 08:47), Yosry Ahmed wrote:

quoted

At this point I NACK this patch. We're about to submit an allocator
which is clearly better that z3fold and is faster that zsmalloc in
most cases and that submission will mark z3fold as deprecated. But for
now this move is premature.

I think unless there are current users of z3fold that cannot use
zsmalloc, the introduction of a new allocator should be irrelevant to
deprecating z3fold. Do you know of such users? Can you explain why
zsmalloc is not usable for them?

I second this. I believe "do we know whether z3fold has users"
is a legit question that means no offense.