Content-type: multipart/alternative; Boundary="1__=08BBE564DFC7C4B78f9e8a93df938690918c08BBE564DFC7C4B7" --1__=08BBE564DFC7C4B78f9e8a93df938690918c08BBE564DFC7C4B7 Content-type: text/plain; charset=US-ASCII Content-transfer-encoding: quoted-printable Oprofile users: I have been looking through the Oprofile code to see what is going on w= ith the mmcra_has_sihv flag. In the SLES 2.6.5-7.97 kernel the kernel sets the mmcra_has_sihv bit to= a 1 if the hardware is specifically not power4 or power4plus. Or in other words, the mmcra_has_sihv flag gets set on 970 and power 5. It should = not get set. Hence you need to comment out the line that sets the mmcra_has_sihv flag for the 970 machine. On the SLES 2.6.5-7.139 kernel the test was changed to check if the hardware supports the SIHV feature. The code is if (cur_cpu_spec->cpu_features & CPU_FTR_MMCRA_SIHV) mmcra_has_sihv =3D 1;= Presumably this will only come out true on power 5 and beyond. Hence i= t will not get set on 970. I checked on RH, the RH 2.6.9-5.EL kernel has the hardware support chec= k not the power 4 proc type check. In summary, the correct check for 970 is now in place to fix the proble= m that Joerg was seeing. Carl Love = Joerg Maier = = To Sent by: oprofile-list@lists.sourceforge.= net oprofile-list-adm = cc in@lists.sourcefo = rge.net Subj= ect Re: separating kernel code <-> u= ser code = 05/01/2005 09:45 = AM = = = = = That did it! thanks a lot! Cant we check if the cpu is ppc970 as well and then dont set mmcra_has_sihv? > > More recent kernels have the correct check. You can simply comment = out > > the code in op_model_power4.c that sets mmcra_has_sihv to 1. This > > feature works as expected in POWER5, so we enable it there in mainl= ine. regards, joerg ------------------------------------------------------- This SF.Net email is sponsored by: NEC IT Guy Games. Get your fingers limbered up and give it your best shot. 4 great events= , 4 opportunities to win big! Highest score wins.NEC IT Guy Games. Play to win an NEC 61 plasma display. Visit http://www.necitguy.com/?r=3D20 _______________________________________________ oprofile-list mailing list oprofile-list@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/oprofile-list = --1__=08BBE564DFC7C4B78f9e8a93df938690918c08BBE564DFC7C4B7 Content-type: text/html; charset=US-ASCII Content-Disposition: inline Content-transfer-encoding: quoted-printable

Oprofile users:

I have been looking through the Oprofile code to see what is going on w= ith the mmcra_has_sihv flag.

In the SLES 2.6.5-7.97 kernel the kernel sets the mmcra_has_sihv bit to= a 1 if the hardware is specifically not power4 or power4plus. Or in o= ther words, the mmcra_has_sihv flag gets set on 970 and power 5. It sh= ould not get set. Hence you need to comment out the line that sets the= mmcra_has_sihv flag for the 970 machine.

On the SLES 2.6.5-7.139 kernel the test was changed to check if the har= dware supports the SIHV feature. The code is if (cur_cpu_spec->cpu_= features & CPU_FTR_MMCRA_SIHV) mmcra_has_sihv =3D 1; Presumably th= is will only come out true on power 5 and beyond. Hence it will not ge= t set on 970.

I checked on RH, the RH 2.6.9-5.EL kernel has the hardware support chec= k not the power 4 proc type check.

In summary, the correct check for 970 is now in place to fix the proble= m that Joerg was seeing.

Carl Love

3D"InactiveJoerg Maier <joerg.maier@rz.uni-man= nheim.de>


=
          Joerg Maier <joerg.maier@rz.uni-mannheim.de&= gt;
          Sent by: oprofile-list-admin@lists.sourceforge.net

          05/01/2005 09:45 AM

=
3D""
To
3D""
oprofile-list@lists.sourceforge.net
3D""
cc
3D""
3D""
Subject
3D""
Re: separating kernel code <-> user code<= /td>
=3D""3D""<= /td>

That did it! thanks a lot! Cant we check if the cpu is ppc970 as we= ll
and then dont set mmcra_has_sihv?

> > More recent kernels have the correct check. You can simply co= mment out
> > the code in op_model_power4.c that sets mmcra_has_sihv to 1. = This
> > feature works as expected in POWER5, so we enable it there in= mainline.

regards,
joerg


-------------------------------------------------------
This SF.Net email is sponsored by: NEC IT Guy Games.
Get your fingers limbered up and give it your best shot. 4 great events= , 4
opportunities to win big! Highest score wins.NEC IT Guy Games. Play to<= br> win an NEC 61 plasma display. Visit
http://www.necitguy.com/?r=3D20
_______________________________________________
oprofile-list mailing list
oprofile-list@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/oprofile-list=

= --1__=08BBE564DFC7C4B78f9e8a93df938690918c08BBE564DFC7C4B7--