From: Sergey Vojtovich
Date: October 20 2010 8:02am
Subject: Re: bzr commit into mysql-next-mr-bugfixing branch
 (nirbhay.choubey:3326) Bug#44771
List-Archive: http://lists.mysql.com/commits/121257
Message-Id: <20101020080226.GB21709@june>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii

Hi Nirbhay,

On Tue, Oct 19, 2010 at 12:19:19PM +0000, Nirbhay Choubey wrote:
> === modified file 'storage/heap/hp_rkey.c'
> --- a/storage/heap/hp_rkey.c	2009-12-22 09:35:56 +0000
> +++ b/storage/heap/hp_rkey.c	2010-10-19 12:19:16 +0000
> @@ -63,7 +63,12 @@ int heap_rkey(HP_INFO *info, uchar *reco
>        info->update= 0;
>        DBUG_RETURN(my_errno);
>      }
> -    if (!(keyinfo->flag & HA_NOSAME))
> +    /*
> +      If key is unique and can accept NULL values, we still
> +      need to copy it to info->lastkey, which in turn is used
> +      to search subsequent records.
> +    */
> +    if (!(keyinfo->flag & HA_NOSAME) || (keyinfo->flag & HA_NULL_PART_KEY))
>        memcpy(info->lastkey, key, (size_t) keyinfo->length);
>    }
>    memcpy(record, pos, (size_t) share->reclength);
> 
Hmm... it confused me. I believe the more valuable information here
would answer a question like: "What do we do here at all?". Something
like:
/* Update lastkey if more index reads are expected. */

Regards,
Sergey
-- 
Sergey Vojtovich <svoj@stripped>
MySQL AB, Software Engineer
Izhevsk, Russia, www.mysql.com