[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <12f8070a-5358-af01-8535-e94405bea931@huaweicloud.com>
Date: Fri, 1 Aug 2025 15:44:24 +0800
From: Yu Kuai <yukuai1@...weicloud.com>
To: Yu Kuai <yukuai1@...weicloud.com>, Jan Kara <jack@...e.cz>
Cc: yukuai@...nel.org, axboe@...nel.dk, akpm@...ux-foundation.org,
yang.yang@...o.com, dlemoal@...nel.org, ming.lei@...hat.com,
linux-block@...r.kernel.org, linux-kernel@...r.kernel.org,
yi.zhang@...wei.com, yangerkun@...wei.com, johnny.chenyi@...wei.com,
Omar Sandoval <osandov@...com>, "yukuai (C)" <yukuai3@...wei.com>
Subject: Re: [PATCH v2 1/2] lib/sbitmap: convert shallow_depth from one word
to the whole sbitmap
Hi,
在 2025/08/01 8:24, Yu Kuai 写道:
> Hi,
>
> 在 2025/08/01 0:27, Jan Kara 写道:
>> On Thu 31-07-25 10:38:58, Yu Kuai wrote:
>>> Hi,
>>>
>>> 在 2025/07/31 2:24, Yu Kuai 写道:
>>>> hi, Jan!
>>>>
>>>> 在 2025/7/30 21:03, Jan Kara 写道:
>>>>> I think having two APIs will be even more confusing than the current
>>>>> state.
>>>>> But as I wrote I think you can have API to specify shallow depth in
>>>>> total
>>>>> size and in sbitmap_queue_get_shallow() do:
>>>>>
>>>>> shallow_per_word = (shallow_depth << sb->shift) / sb->depth;
>>> In order to consider the last word, I think we should use __map_depth()
>>> here.
>>
>> Right.
>>
>>>>> rounding_index = shallow_depth - shallow_per_word * sb->depth;
>>> And then it's not possible to calculate this rounding index easily. How
>>> about following, although the reminder handling is not perfect.
>>>
>>> static unsigned int __map_depth_with_shallow(const struct sbitmap *sb,
>>> int index,
>>> unsigned int
>>> shallow_depth)
>>> {
>>> unsigned int word_depth = __map_depth(sb, index);
>>> unsigned int shallow_word_depth = word_depth * shallow_depth;
>>> unsigned reminder = do_div(shallow_word_depth, sb->depth);
>>>
>>> if (reminder && !(index & 0x1))
>>
>> Well, why not:
>> if (remainder > index)
> Do you mean reminder > index * shallow_depth? This looks correct, and
> with the consideration for the last word:
>
> if (index == sb->map_nr - 1)
> shallow_word_depth = max(shallow_word_depth, 1);
> else if (reminder > index * shallow_depth)
Sorry there is a mistake, should use the word_depth here, following is
an example for 4 word sbitmap(64+64+64+32), I do the math manually, and
the results look perfect :)
| shallow_depth | word0 | word1 | word2 | word3 | total |
| ------------- | ------ | ------ | ------ | ----- | ----- |
| 224 | 64 | 64 | 64 | 32 | 224 |
| 112 | 32 | 32 | 32 | 16 | 112 |
| 113 | 32 + 1 | 32 | 32 | 16 | 113 |
| 114 | 32 + 1 | 32 + 1 | 32 | 16 | 114 |
| 115 | 32 + 1 | 32 + 1 | 32 + 1 | 16 | 115 |
| 116 | 33 | 33 | 33 | 16+1 | 116 |
> shallow_word_depth++;
>
> Thanks,
> Kuai
>
>> ?
>>
>> That should accurately distribute the remainder across the remaining
>> words,
>> shouldn't it?
>>
>>> shallow_word_depth++;
>>>
>>> return shallow_word_depth;
>>> }
>>
>> Honza
>>
>
> .
>
Powered by blists - more mailing lists