lists.openwall.net   lists  /  announce  owl-users  owl-dev  john-users  john-dev  passwdqc-users  yescrypt  popa3d-users  /  oss-security  kernel-hardening  musl  sabotage  tlsify  passwords  /  crypt-dev  xvendor  /  Bugtraq  Full-Disclosure  linux-kernel  linux-netdev  linux-ext4  linux-hardening  linux-cve-announce  PHC 
Open Source and information security mailing list archives
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date:	Thu, 18 Feb 2016 14:16:31 +0800
From:	Ming Lei <ming.lei@...onical.com>
To:	Kent Overstreet <kent.overstreet@...il.com>
Cc:	Sagi Grimberg <sagig@....mellanox.co.il>,
	Jens Axboe <axboe@...nel.dk>,
	Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
	linux-block@...r.kernel.org, Christoph Hellwig <hch@...radead.org>,
	stable <stable@...r.kernel.org>,
	Keith Busch <keith.busch@...el.com>
Subject: Re: [PATCH 1/4] block: bio: introduce helpers to get the 1st and last bvec

Hi Kent,

Thanks for your review.

On Thu, Feb 18, 2016 at 12:24 PM, Kent Overstreet
<kent.overstreet@...il.com> wrote:
> On Mon, Feb 15, 2016 at 05:42:12PM +0800, Ming Lei wrote:
>> Cc Kent and Keith.
>>
>> Follows another version which should be more efficient.
>> Kent and Keith, I appreciate much if you may give a review on it.
>>
>> diff --git a/include/linux/bio.h b/include/linux/bio.h
>> index 56d2db8..ef45fec 100644
>> --- a/include/linux/bio.h
>> +++ b/include/linux/bio.h
>> @@ -278,11 +278,21 @@ static inline void bio_get_first_bvec(struct bio *bio, struct bio_vec *bv)
>>   */
>>  static inline void bio_get_last_bvec(struct bio *bio, struct bio_vec *bv)
>>  {
>> -     struct bvec_iter iter;
>> +     struct bvec_iter iter = bio->bi_iter;
>> +     int idx;
>> +
>> +     bio_advance_iter(bio, &iter, iter.bi_size);
>> +
>> +     WARN_ON(!iter.bi_idx && !iter.bi_bvec_done);
>> +
>> +     if (!iter.bi_bvec_done)
>> +             idx = iter.bi_idx - 1;
>> +     else    /* in the middle of bvec */
>> +             idx = iter.bi_idx;
>>
>> -     bio_for_each_segment(*bv, bio, iter)
>> -             if (bv->bv_len == iter.bi_size)
>> -                     break;
>> +     *bv = bio->bi_io_vec[idx];
>> +     if (iter.bi_bvec_done)
>> +             bv->bv_len = iter.bi_bvec_done;
>>  }
>
> It can't be done correctly without a loop.

As we discussed in gtalk, the only case this patch can't cope with
is that one single bvec doesn't use up the remained io vector,
but it can be handled by putting the following code at the
function entry:

if (!bio_multiple_segments(bio)) {
       *bv = bio_iovec(bio);
       return;
}

>
> The reason is that if the bio was split in the middle of a segment, bv->bv_len
> on the last biovec will be larger than what's actually used by the bio (it's
> being shared between the two splits!).

The last two lines in this helper should handle the situation.

>
> You have to iterate over all the biovecs so that you can see where
> bi_iter->bi_size ends.

I understand your concern is that this patch may not be much more
efficient than bio_for_each_segment().

IMO, one win of the patch is that 16bytes bvec copy is saved for all
vectors, and another 'win' is to just run bvec_iter_advance() once(
like move the outside for loop inside).

I will run some benchmark to see if there is any performance
difference between the two patches.

Thanks,
Ming

Powered by blists - more mailing lists