On Tue, Aug 27, 2024 at 05:59:41PM GMT, Christian Brauner wrote:
> When a kmem cache is created with SLAB_TYPESAFE_BY_RCU the free pointer
> must be located outside of the object because we don't know what part of
> the memory can safely be overwritten as it may be needed to prevent
> object recycling.
>
> That has the consequence that SLAB_TYPESAFE_BY_RCU may end up adding a
> new cacheline. This is the case for .e.g, struct file. After having it
> shrunk down by 40 bytes and having it fit in three cachelines we still
> have SLAB_TYPESAFE_BY_RCU adding a fourth cacheline because it needs to
> accomodate the free pointer and is hardware cacheline aligned.
>
> I tried to find ways to rectify this as struct file is pretty much
> everywhere and having it use less memory is a good thing. So here's a
> proposal.
>
> I was hoping to get something to this effect into v6.12.
>
> If we really want to switch to a struct to pass kmem_cache parameters I
> can do the preparatory patch to convert all kmem_cache_create() and
> kmem_cache_create_usercopy() callers to use a struct for initialization
> of course. I can do this as a preparatory work or as follow-up work to
> this series. Thoughts?
So one thing I can do is to add:
struct kmem_cache_args {
.freeptr_offset,
.useroffset,
.flags,
.name,
};
accompanied by:
int kmem_create_cache(struct kmem_cache_args *args);
and then switch both the filp cache and Jens' io_kiocb cache over to use
these two helpers. Then we can convert other callers one by one.
@Vlastimil, @Jens, @Linus what do you think?
