Re: [PATCHv4 5/5] io_uring: cache nodes and mapped buffers

[Pavel Begunkov <asml.silence@gmail.com>]

On 2/24/25 21:04, Keith Busch wrote:
> On Thu, Feb 20, 2025 at 04:06:21PM +0000, Pavel Begunkov wrote:
>> On 2/20/25 15:24, Keith Busch wrote:
>>>>> +		node = io_cache_alloc(&ctx->buf_table.node_cache, GFP_KERNEL);
>>>>
>>>> That's why node allocators shouldn't be a part of the buffer table.
>>>
>>> Are you saying you want file nodes to also subscribe to the cache? The
>>
>> Yes, but it might be easier for you to focus on finalising the essential
>> parts, and then we can improve later.
>>
>>> two tables can be resized independently of each other, we don't know how
>>> many elements the cache needs to hold.
>>
>> I wouldn't try to correlate table sizes with desired cache sizes,
>> users can have quite different patterns like allocating a barely used
>> huge table. And you care about the speed of node change, which at
>> extremes is rather limited by CPU and performance and not spatiality
>> of the table. And you can also reallocate it as well.
> 
> Having the cache size and lifetime match a table that it's providing

You don't need to align it with the table size, I actually
overlooked it, will comment on v5.

> seems as simple as I can make this. This is still an optimization at the
> end of the day, so it's not strictly necessary to take the last two
> patches from this series to make zero copy work if you don't want to
> include it from the beginning.

Ok. As I mentioned above, I do think it's wrong for nodes, but
we can merge the patch and rework on top.

-- 
Pavel Begunkov