Re: Powertweak release 0.99.5

[<man...@t-...>]

On 07/09/03 16:37, Sven Dowideit wrote:
> Hi guys,
> 
> We have a new release. I'll start working on the 0.99.6 roadmap soon, and 
> would like some suggestions for network protocol. I'm currently leaning towards
> xmlrpc or SOAP, perhaps utilising the DMTF WBEM standard (the same guys that did
> the DMI/SMBIOS standard).
> 
> Please tell me if i've forgotten anything as this is my first sf release.
> 
> I'll be updating the web page tomorrow :)
> 
> Dave - can you please update the freshmeat.net page?
> 
> cheers
> Sven
> 
> ------------------------------------------------------------------------
> from the NEWS file
> 
> v0.99.5 [Release 22. -- The 'long in the tooth' release ]
[...]

Hi Sven,

nice to have a new release finally!!!

Somehow I'm missing the CPU[0] Tweaks tab completely. I've rechecked 
things like MSR support in kernel and compatibility of my PIII 
Coppermine CPU (family 6, model 8, stepping 6, brand 2).

When running gpowertweak I get
Gtk-CRITICAL **: file gtkwidget.c: line 1427 (gtk_widget_show): 
assertion `widget != NULL' failed.

Gtk-CRITICAL **: file gtkbox.c: line 331 (gtk_box_pack_start): assertion 
`child != NULL' failed.
now (instead?!?).

(But the lines above only when run without powertweakd in the background.)

In Powertweak 0.99.4 all of the above is working well. And two more 
questions regarding the PentiumIII.xml:
* Why did you remove the In-order-queue-depth tweak?
* Was this register change from 0.99.4 to 0.99.5 intended:
   from:   <MSR register="0x0000002A" bit="30">
            <WIDGETTEXT>Streaming buffer disable</WIDGETTEXT>
            <CONFIGNAME>P6_STREAMING_BUFFER_DISABLE</CONFIGNAME>
     	  ...
   to:     <MSR register="0x00000033" bit="30">
            <WIDGETTEXT>Streaming buffer disable</WIDGETTEXT>
            <CONFIGNAME>P6_STREAMING_BUFFER_DISABLE</CONFIGNAME>
           ...

Another thing, can you have a look at the attached diff that updates the 
old bdflush section of memory.xml to some kernel 2.4.x style (and fixes 
one minor thing in proc.xml, too). Maybe it's worth to be applied in 
some kind.


Thanks and best regards,

  Manuel