Re: _mdfd_getseg can be expensive - Mailing list pgsql-hackers

From Andres Freund
Subject Re: _mdfd_getseg can be expensive
Date
Msg-id 20141101172347.GP13584@awork2.anarazel.de
Whole thread Raw
In response to Re: _mdfd_getseg can be expensive  (Tom Lane <tgl@sss.pgh.pa.us>)
Responses Re: _mdfd_getseg can be expensive  (Andres Freund <andres@anarazel.de>)
List pgsql-hackers
On 2014-11-01 12:57:40 -0400, Tom Lane wrote:
> Andres Freund <andres@2ndquadrant.com> writes:
> > On 2014-10-31 18:48:45 -0400, Tom Lane wrote:
> >> While the basic idea is sound, this particular implementation seems
> >> pretty bizarre.  What's with the "md_seg_no" stuff, and why is that
> >> array typed size_t?
> 
> > It stores the length of the array of _MdfdVec entries.
> 
> Oh.  "seg_no" seems like not a very good choice of name then.
> Perhaps "md_seg_count" or something like that would be more intelligible.

That's fine with me.

> And personally I'd have made it an int, because we are certainly not doing
> segment-number arithmetic in anything wider than int anywhere else.

Fine with me too. I picked size_t by habit, because there's projects
that don't allow anything else to be used for lengths of memory...

I've, during testing, also noticed it has accidentally introduced a
vfd/memory leak...

So I'll repost a version with those fixes.

Greetings,

Andres Freund

-- Andres Freund                       http://www.2ndQuadrant.com/PostgreSQL Development, 24x7 Support, Training &
Services



pgsql-hackers by date:

Previous
From: Tom Lane
Date:
Subject: Re: [BUGS] ltree::text not immutable?
Next
From: Andres Freund
Date:
Subject: Re: Let's drop two obsolete features which are bear-traps for novices