headers
Marc Alff | 1 Sep 2006 05:17
Picon
Favicon

Re: SHOW PROFILE and PATCH/5.1: Change thd->proc_info accesses to a macro


Hi All

Elliot Murphy wrote:
> This reminds me - I asked the runtime team about it,
> and Kostja thought the change was OK, but preferred
> to see it as an inline function rather than a macro.
>
> Jeremy, I'm sorry I neglected to forward that feedback on.
>
> JimW also commented on-list suggesting that a method should be
> added to the THD class instead.
>
> Jeremy, what do you think of of these review comments?
> It should be easy to add this once we agree on the changes.
>

Tentatively answering for Jeremy.

THD_PROC_INFO probably needs to be a macro,
since __FILE__ and __LINE__ will be captured with later changes,
to display the source information.

See the result of "show profile source for query 8;" in the example given :

http://jcole.us/files/show_profile2.txt

Looks very cool.

Cheers,
(Continue reading)

Permalink | Reply | 1 comment |
headers
Jeremy Cole | 1 Sep 2006 05:41

Re: SHOW PROFILE and PATCH/5.1: Change thd->proc_info accesses to a macro

Hi Marc, All,

> Tentatively answering for Jeremy.
> 
> THD_PROC_INFO probably needs to be a macro,
> since __FILE__ and __LINE__ will be captured with later changes,
> to display the source information.
> 
> See the result of "show profile source for query 8;" in the example given :
> 
> http://jcole.us/files/show_profile2.txt
> 
> Looks very cool.

Yes, you're quite right that in order to get the file/line source 
information, it does need to be a macro unfortunately.

Regards,

Jeremy

-- 
high performance mysql consulting
www.provenscaling.com

--

-- 
MySQL Internals Mailing List
For list archives: http://lists.mysql.com/internals
To unsubscribe:    http://lists.mysql.com/internals?unsub=gcdmd-internals <at> m.gmane.org
(Continue reading)

Permalink | Reply | 0 comments |
headers
Green, Paul | 1 Sep 2006 13:56
Picon

RE: Patches to correct build issues on Stratus VOS

Lenz Grimmer [mailto:lenz <at> mysql.com] wrote:
> Hi Paul,
> 
> On Tue, 30 May 2006, Green, Paul wrote:
> 
> > >That you had built and run MySQL, and all tests pass, and no new 
> > >patches are necessary.
> > >For example, we may want to wait till we have complete set 
> > >of patches that are required for a working binary.
> > 
> > That makes sense.  I will make a point to send you another set of 
> > patches once all the tests work.  That way, there will be a 
> > record of 
> > all the patches needed to make it work on VOS.  If you 
> > would prefer to 
> > wait and review the patches at that time, that's fine by me.
> 
> Just catching up on some old stuff here - is there any update 
> regarding these patches to get MySQL working on VOS?

I have been unable to get finish this work.  I'm trying to find a VOS
customer to sponsor the work so that I can devote my full attention to
it.  I don't have a specific timeframe for completing the port, but I'm
hopeful that it will finish sometime in 2007.

Thanks
PG

--

-- 
MySQL Internals Mailing List
(Continue reading)

Permalink | Reply | 501 comments |
headers
Lenz Grimmer | 1 Sep 2006 14:00
Picon
Favicon

RE: Patches to correct build issues on Stratus VOS


Hi Paul,

On Fri, 1 Sep 2006, Green, Paul wrote:

> I have been unable to get finish this work.  I'm trying to find a VOS
> customer to sponsor the work so that I can devote my full attention to
> it.  I don't have a specific timeframe for completing the port, but I'm
> hopeful that it will finish sometime in 2007.

Thanks for the update, I appreciate it! Let us know if there is anything
we can do to help you.

Bye,
	LenZ
--

-- 
 Lenz Grimmer <lenz <at> mysql.com>
 Community Relations Manager, EMEA
 MySQL GmbH, http://www.mysql.de/, Hamburg, Germany
 Visit the MySQL Forge at http://forge.mysql.com/
Permalink | Reply | 501 comments |
headers
Marc Alff | 5 Sep 2006 20:10
Picon
Favicon

Re: SHOW PROFILE and PATCH/5.1: Change thd->proc_info accesses to a macro


Hi Jeremy

>
> On Mon, 31 Jul 2006, Jeremy Cole wrote:
>
> >> Some of you have already heard about an upcoming feature for
> tracing queries
> >> through the MySQL server, tentatively called SHOW PROFILE. If you'd
> like to
> >> get excited about it, here's an example:
> >>
> >> http://jcole.us/files/show_profile2.txt
> >>
> >> I am contributing and doing the work on this feature in a few stages:
> >>
> >> 0. Change thd->proc_info access to a macro
> >> 1. Add code for SHOW PROFILE command itself
> >> 2. Clean up nonsensical or useless proc_info changes
> >> 3. Add more useful proc_info changes where appropriate
> >>
The code will be changed for step [0] to a macro, as you proposed.

Some minor changes :

- the macro name is THD_SET_PROC_INFO
- it will expand to thd->set_proc_info(proc_info, __FILE__, __LINE__)

THD_SET_PROC_INFO should provide the entry point needed.
(Continue reading)

Permalink | Reply | 1 comment |
headers
Jeremy Cole | 5 Sep 2006 19:34

Re: SHOW PROFILE and PATCH/5.1: Change thd->proc_info accesses to a macro

Hi Marc,

First, it's good to see you with a mysql.com email address. :)  Two 
months ago with your contributions I said "he will work for MySQL soon" :)

> The code will be changed for step [0] to a macro, as you proposed.
> 
> Some minor changes :
> 
> - the macro name is THD_SET_PROC_INFO
> - it will expand to thd->set_proc_info(proc_info, __FILE__, __LINE__)
> 
> THD_SET_PROC_INFO should provide the entry point needed.
> 
> The THD::set_proc_info() method is only tentative for now,
> feel free to change it with later patches if needed.
> 
> I assume these minor changes are ok, please let us know otherwise.

Yes, no problem at all.

Is there a patch available yet?  (Or, has it been pushed to BK yet?)

Regards,

Jeremy

--

-- 
high performance mysql consulting
www.provenscaling.com
(Continue reading)

Permalink | Reply | 0 comments |
headers
Arnold Daniels | 6 Sep 2006 13:11
Picon

Re: UDF XML functions

Hi,

First of all the first version of the UDF library to create XML is 
finished. It can be found at http://libmysqlbnxml.sourceforge.net/. I 
could appreciate it if people could have a look and give any remarks 
about the code, especially security flaws.

Next, I'm not able to get a tree result within a decent time, because 
I'm unable to get MySQL to do what I want it to do. The problem breaks 
down to this:

SELECT id, bedrijfsnaam, achternamen, adressen FROM relatie
 LEFT JOIN (SELECT relatie_id, GROUP_CONCAT(achternaam) AS achternamen 
FROM contactpersoon GROUP BY relatie_id) AS contactpersoon ON relatie.id 
= contactpersoon.relatie_id
 LEFT JOIN (SELECT relatie_id, GROUP_CONCAT(adres) AS adressen FROM 
relatie_adres GROUP BY relatie_id) AS relatie_adres ON relatie.id = 
relatie_adres.relatie_id
LIMIT 50

Executing a query like this is extremely slow. All subqueries are 
executed fully and the joins are done by table scans. Does anyone have 
any suggestion in how to speed up getting aggregated data from the 
database without have to split up query and doing all the work at the 
client.

Thanks,
Arnold

Brian Aker schreef:
(Continue reading)

Permalink | Reply | 0 comments |
headers
Wei Li | 6 Sep 2006 21:35
Picon
Favicon

how to enable DBUG_EXECUTE_IF()

Hi,

In the code, we see there some DBUG_EXECUTE_IF() macros probably for
testing, like the following:
  DBUG_EXECUTE_IF("crash_commit_after_log", abort(););

I wonder how we can enable the test?

Thanks,
Wei

--

-- 
MySQL Internals Mailing List
For list archives: http://lists.mysql.com/internals
To unsubscribe:    http://lists.mysql.com/internals?unsub=gcdmd-internals <at> m.gmane.org
Permalink | Reply | 1 comment |
headers
Bastiaan Zapf | 6 Sep 2006 23:35
Picon
Favicon

Modifying the Fulltext index


I'm trying to search fulltext-wise over vast amounts of data, some of
which includes real dashes (that aren't intended as markup or something)
- ISBN numbers and the like. Now I'd like to make the Fulltext-index
work with these "words".

I'd have to change the find-subroutines for sure. Would I have to change
the indexing subroutines too?

How much work (which order of magnitude? Hours? Days? Weeks?) would that
be for a practiced programmer?

Greetings,
   Basti

--

-- 
MySQL Internals Mailing List
For list archives: http://lists.mysql.com/internals
To unsubscribe:    http://lists.mysql.com/internals?unsub=gcdmd-internals <at> m.gmane.org
Permalink | Reply | 1 comment |
headers
xiao.feng | 7 Sep 2006 05:42
Picon

How to detect SQL needed Columns in Storage Engine,not all columns?

hi,
  How to detect SQL needed Columns in Storage Engine,not all columns?
Question:
  table1:  column1, column2, column3.    
    column1, column2 withn small data length, may be int or long or datetime.
    column3 with very long data length, may be 64k or 2M high.

  SQL: select column1, column2 from table1 where column3 ='value'
  In My storage engine, how to detect SQL level only wants column1 and column2's value, no need column3'
value? Because column3's data very long, may cause  system slower.

 Federated engine select all fields from remote MySQL, I think it not the best solution.

 Thanks you for any hints!

--

-- 
MySQL Internals Mailing List
For list archives: http://lists.mysql.com/internals
To unsubscribe:    http://lists.mysql.com/internals?unsub=gcdmd-internals <at> m.gmane.org
Permalink | Reply | 7 comments |

Gmane