Re: [PATCH 0/2] zswap pool per-CPU acomp_ctx simplifications

[Yosry Ahmed]

> > > > Actually looking at zswap_cpu_comp_dead(), is the IS_ERR_OR_NULL()
> > > > check on acomp_ctx also misleading? Should that also just be a NULL
> > > > check?
> > >
> > > Even a NULL check would be redundant in this case, per my
> > > understanding, because if the alloc_percpu() call in
> > > zswap_pool_create() had failed, pool creation would have failed.
> > >
> > > I think a NULL check on the acomp_ctx would still be a good idea, just
> > > in case, since this is all part of CPU hotplug. I agree, we don't need
> > > an IS_ERR() check on acomp_ctx.
> >
> > So I think we do one patch to convert both IS_ERR_OR_NULL() to NULL
> > checks, and then the current patch 1, right?
>
> Yes.

Sounds good, thanks for bearing with me :)