Re: Comment simplehash/dynahash trade-offs

On Mon, Jul 20, 2020 at 1:29 AM Thomas Munro <thomas.munro@gmail.com> wrote:
>
> On Fri, May 1, 2020 at 1:53 PM James Coleman <jtc331@gmail.com> wrote:
> > In another thread [1] I'd mused that "there might be some value in a
> > README or comments
> > addition that would be a guide to what the various hash
> > implementations are useful for...so that we have something to make the
> > code base a bit more discoverable."
>
> +1
>
> > I'd solicited feedback from Andres (as the author of the simplehash
> > implementation) and gotten further explanation from Tomas (both cc'd
> > here) and have tried to condense that into the comment changes in this
> > patch series.
> >
> > v1-0001-Summarize-trade-offs-between-simplehash-and-dynah.patch
> > Contains the summaries mentioned above.
>
> + * - It supports partitioning, which is useful for shared memory access using
>
> I wonder if we should say a bit more about the shared memory mode.
> Shared memory dynahash tables are allocated in a fixed size area at
> startup, and are discoverable by name in other other processes that
> need to get access to them, while simplehash assumes that it can get
> memory from a MemoryContext or an allocator with a malloc/free-style
> interface, which isn't very well suited for use in shared memory.
> (I'm sure you can convince it to work in shared memory with some
> work.)

Added.

> > v1-0002-Improve-simplehash-usage-notes.patch
>
> + *    For convenience the hash table create functions accept a void pointer
> + *    will be stored in the hash table type's member private_data.
>
> *that* will be stored?

Fixed.

> > v1-0003-Show-sample-simplehash-method-signatures.patch
> > I find it hard to read the macro code "templating" particularly for
> > seeing what the available API is and so added sample method signatures
> > in comments to the macro generated method signature defines.
>
> I didn't double-check all the expansions of the macros but +1 for this
> idea, it's very useful.

James