[PATCH] PM patches: logic fix, feature, cleaning, memory downclocking

Rafał Miłecki

22 May 2009 22 May '09

09:50

Well, topic says everything :) First 3 patches are quite nice, stable changes. Not sure if you will want to include 4th patch which adds code (commented out for now) for almost untested memory downclocking. I suggest commiting this, then we can ask ppl to test this. If you prefer other way, let me know. -- Rafał Miłecki

Attachments:

0001-PM-fix-broken-engine-clock-setting-logic.patch (application/octet-stream — 2.6 KB)
0002-PM-add-RHDGetDefaultMemoryClock.patch (application/octet-stream — 1.4 KB)
0003-PM-get-rid-of-EnableForced-field.patch (application/octet-stream — 2.5 KB)
0004-PM-add-memory-downclocking.-Commented-out-for-now.patch (application/octet-stream — 3.0 KB)

Show replies by date

David C. Rankin, J.D.,P.E.

26 May 26 May

05:41

On or about Friday 22 May 2009 at approximately 04:50:43 Rafał Miłecki composed:

...

Well, topic says everything :)

First 3 patches are quite nice, stable changes. Not sure if you will want to include 4th patch which adds code (commented out for now) for almost untested memory downclocking. I suggest commiting this, then we can ask ppl to test this. If you prefer other way, let me know.

Rafał Thank you. No how do I apply the patches to test? I have all of the 4 patches saved in my radeonhd build directory under a directory named "patches". So my git clone setup is: ~/archlinux/radeonhd (dir created with git cloone) ~/archlinux/radeonhd/patches (your 4 patches) I'm ready to test. What parameters do I look at to know if it is having the desired effect, of do I just tape a thermometer to the palmrest on my laptop? -- David C. Rankin, J.D.,P.E. Rankin Law Firm, PLLC 510 Ochiltree Street Nacogdoches, Texas 75961 Telephone: (936) 715-9333 Facsimile: (936) 715-9339 www.rankinlawfirm.com -- To unsubscribe, e-mail: radeonhd+unsubscribe@opensuse.org For additional commands, e-mail: radeonhd+help@opensuse.org

Rafał Miłecki

07:26

New subject: [PATCH] PM patches: logic fix, feature, cleaning, memory downclocking

2009/5/26 David C. Rankin, J.D.,P.E. :

...

Thank you. No how do I apply the patches to test? I have all of the 4 patches saved in my radeonhd build directory under a directory named "patches". So my git clone setup is:

~/archlinux/radeonhd (dir created with git cloone) ~/archlinux/radeonhd/patches (your 4 patches)

I'm ready to test. What parameters do I look at to know if it is having the desired effect, of do I just tape a thermometer to the palmrest on my laptop?

First patch prevents printing incorrect warning "ForceLowPowerMode disabled: could not determine default engine clock" in Xorg.0.log in case you use "ForceLowPowerMode" and don't use "LowPowerModeEngineClock". It's easy patch, nothing to regress. Second patch adds function RHDGetDefaultMemoryClock, but doesn't make any use of it yet. Third patch is just little simplifing code. Fourth patch finally adds something you may test. It tries downclocking your Memory by 50%. But ith's disabled by default. You need to apply this patch, then manually edit rhd_pm.c. If you want to test this, find "#if 0" in rhd_pm.c and change it to "#if 1" (it occures in two places in code). After restarting X just use "grep Memory /var/log/Xorg.0.log". How to apply patches? I suggest creating separated local branch: cd ~/archlinux/radeonhd/ git checkout -t -b downclocking origin/master git am ~/archlinux/radeonhd/patches/*patch [now edit rhd_pm.c] make? sudo make install? -- Rafał Miłecki -- To unsubscribe, e-mail: radeonhd+unsubscribe@opensuse.org For additional commands, e-mail: radeonhd+help@opensuse.org

David C. Rankin, J.D.,P.E.

29 May 29 May

04:52

New subject: [PATCH] PM patches: logic fix, feature, cleaning, memory downclocking

On or about Tuesday 26 May 2009 at approximately 02:26:59 Rafał Miłecki composed:

...

2009/5/26 David C. Rankin, J.D.,P.E. :

...
Thank you. No how do I apply the patches to test? I have all of the 4 patches saved in my radeonhd build directory under a directory named "patches". So my git clone setup is:

~/archlinux/radeonhd (dir created with git cloone) ~/archlinux/radeonhd/patches (your 4 patches)

I'm ready to test. What parameters do I look at to know if it is having the desired effect, of do I just tape a thermometer to the palmrest on my laptop?

First patch prevents printing incorrect warning "ForceLowPowerMode disabled: could not determine default engine clock" in Xorg.0.log in case you use "ForceLowPowerMode" and don't use "LowPowerModeEngineClock". It's easy patch, nothing to regress.

Second patch adds function RHDGetDefaultMemoryClock, but doesn't make any use of it yet.

Third patch is just little simplifing code.

Fourth patch finally adds something you may test. It tries downclocking your Memory by 50%. But ith's disabled by default. You need to apply this patch, then manually edit rhd_pm.c. If you want to test this, find "#if 0" in rhd_pm.c and change it to "#if 1" (it occures in two places in code).

After restarting X just use "grep Memory /var/log/Xorg.0.log".

How to apply patches?

I suggest creating separated local branch: cd ~/archlinux/radeonhd/ git checkout -t -b downclocking origin/master git am ~/archlinux/radeonhd/patches/*patch [now edit rhd_pm.c] make? sudo make install?

-- Rafał Miłecki

Rafał, That's what I needed to know. I'm am not familiar with 'git' and the git checkout -t -b downclocking origin/master git am ~/archlinux/radeonhd/patches/*patch parts were the parts I knew nothing about. I'll pull the patches, apply and report back. Something really has to give here. The heat is unbearable. To get the actual temp data, I have a digital multimeter with a great surface temperature probe on it ( Craftsman Model No. 82400 ) and I took the temperature readings on the left palm rest and left fan discharge of my laptop with the radeonhd driver active. The results: Left Palm Rest Temp: 96 Deg. F Left Fan Discharge: 147 Deg. F The inside of this box is cooking.... That's with: Option "ForceLowPowerMode" Option "LowPowerModeEngineClock" "140000" I'll keep my fingers crossed and report back. -- David C. Rankin, J.D.,P.E. Rankin Law Firm, PLLC 510 Ochiltree Street Nacogdoches, Texas 75961 Telephone: (936) 715-9333 Facsimile: (936) 715-9339 www.rankinlawfirm.com -- To unsubscribe, e-mail: radeonhd+unsubscribe@opensuse.org For additional commands, e-mail: radeonhd+help@opensuse.org

Rafał Miłecki

05:01

New subject: [PATCH] PM patches: logic fix, feature, cleaning, memory downclocking

2009/5/29 David C. Rankin, J.D.,P.E. :

...

Rafał,

That's what I needed to know. I'm am not familiar with 'git' and the

git checkout -t -b downclocking origin/master git am ~/archlinux/radeonhd/patches/*patch

parts were the parts I knew nothing about. I'll pull the patches, apply and report back. Something really has to give here. The heat is unbearable. To get the actual temp data, I have a digital multimeter with a great surface temperature probe on it ( Craftsman Model No. 82400 ) and I took the temperature readings on the left palm rest and left fan discharge of my laptop with the radeonhd driver active. The results:

Left Palm Rest Temp: 96 Deg. F

Left Fan Discharge: 147 Deg. F

The inside of this box is cooking.... That's with:

Option "ForceLowPowerMode" Option "LowPowerModeEngineClock" "140000"

I'll keep my fingers crossed and report back.

Nice, tests with your multimeter would be absolutely great. You may also try to reduce EngineClock later. For example change it from 140000 to 120000. If it doesn't cause any corruptions you can still try decreasing that. Don't worry about values you may use 139999, 139998, 139997, end so on.AtomBIOS will round it to something acceptable. -- Rafał Miłecki -- To unsubscribe, e-mail: radeonhd+unsubscribe@opensuse.org For additional commands, e-mail: radeonhd+help@opensuse.org

David C. Rankin, J.D.,P.E.

05:26

New subject: [PATCH] PM patches: logic fix, feature, cleaning, memory downclocking

On or about Friday 29 May 2009 at approximately 00:01:26 Rafał Miłecki composed:

...

2009/5/29 David C. Rankin, J.D.,P.E. :

...
Rafał,

That's what I needed to know. I'm am not familiar with 'git' and the

git checkout -t -b downclocking origin/master git am ~/archlinux/radeonhd/patches/*patch

parts were the parts I knew nothing about. I'll pull the patches, apply and report back. Something really has to give here. The heat is unbearable. To get the actual temp data, I have a digital multimeter with a great surface temperature probe on it ( Craftsman Model No. 82400 ) and I took the temperature readings on the left palm rest and left fan discharge of my laptop with the radeonhd driver active. The results:

Left Palm Rest Temp: 96 Deg. F

Left Fan Discharge: 147 Deg. F

The inside of this box is cooking.... That's with:

Option "ForceLowPowerMode" Option "LowPowerModeEngineClock" "140000"

I'll keep my fingers crossed and report back.

Nice, tests with your multimeter would be absolutely great. You may also try to reduce EngineClock later. For example change it from 140000 to 120000. If it doesn't cause any corruptions you can still try decreasing that. Don't worry about values you may use 139999, 139998, 139997, end so on.AtomBIOS will round it to something acceptable.

-- Rafał Miłecki

Rafał, (also cc'ed to archlinux list for those interested) Here is a quick Summary of what I did: 23:54 alchemy:~/archlinux/apps/radeonhd> git checkout -t -b downclocking origin/master Branch downclocking set up to track remote branch master from origin. Switched to a new branch 'downclocking' 23:55 alchemy:~/archlinux/apps/radeonhd> git am patches/*patch Applying: PM: fix broken engine clock setting logic /home/david/archlinux/apps/radeonhd/.git/rebase-apply/patch:30: trailing whitespace. "calculated engine clock at %ldHz\n", warning: 1 line adds whitespace errors. Applying: PM: add RHDGetDefaultMemoryClock Applying: PM: get rid of EnableForced field Applying: PM: add memory downclocking. Commented out for now. /home/david/archlinux/apps/radeonhd/.git/rebase-apply/patch:28: trailing whitespace. "calculated memory clock at %ldHz\n", warning: 1 line adds whitespace errors. 23:56 alchemy:~/archlinux/apps/radeonhd> find . -name 'rhd_pm.c' ./src/rhd_pm.c 23:57 alchemy:~/archlinux/apps/radeonhd> vi src/rhd_pm.c #if 1 unsigned long defaultMemory = RHDGetDefaultMemoryClock(rhdPtr); if (defaultMemory) { Pm->ForcedMemoryClock = defaultMemory / 2; xf86DrvMsg(rhdPtr->scrnIndex, X_INFO, "ForceLowPowerMode: " "calculated memory clock at %ldHz\n", Pm->ForcedEngineClock); } <snip> Reboot & check /var/log/Xorg.0.log <snip> (II) RADEONHD: version 1.2.5, built from git branch downclocking, commit 0de8b484 + changes <snip> The current log file is at: http://www.3111skyline.com/download/linux/radeonhd/downclock/Xorg.0.log.2900... I haven't pulled any additional temperatures for comparison because there wouldn't be that much difference, it's still HOT. Looking through the log file, it is quite a coincidence that the downclocking chose 140000 as the memory clock downclock because that was exactly what I had been setting with the: Option "LowPowerModeEngineClock" "140000" At least we know that the patch applies OK and that it has the desired effect on the memory clock. Now if we can just figure out what else we need to power down to get the temps under control, we will be home free! Thank you for your patch. As I said before, I have the perfect box for testing, so let me know when you have the next one ready and I'll be happy to test it. -- David C. Rankin, J.D.,P.E. Rankin Law Firm, PLLC 510 Ochiltree Street Nacogdoches, Texas 75961 Telephone: (936) 715-9333 Facsimile: (936) 715-9339 www.rankinlawfirm.com -- To unsubscribe, e-mail: radeonhd+unsubscribe@opensuse.org For additional commands, e-mail: radeonhd+help@opensuse.org

David C. Rankin, J.D.,P.E.

05:41

New subject: [PATCH] PM patches: logic fix, feature, cleaning, memory downclocking

On or about Friday 29 May 2009 at approximately 00:01:26 Rafał Miłecki composed:

...

2009/5/29 David C. Rankin, J.D.,P.E. :

...
Rafał,

That's what I needed to know. I'm am not familiar with 'git' and the

git checkout -t -b downclocking origin/master git am ~/archlinux/radeonhd/patches/*patch

parts were the parts I knew nothing about. I'll pull the patches, apply and report back. Something really has to give here. The heat is unbearable. To get the actual temp data, I have a digital multimeter with a great surface temperature probe on it ( Craftsman Model No. 82400 ) and I took the temperature readings on the left palm rest and left fan discharge of my laptop with the radeonhd driver active. The results:

Left Palm Rest Temp: 96 Deg. F

Left Fan Discharge: 147 Deg. F

The inside of this box is cooking.... That's with:

Option "ForceLowPowerMode" Option "LowPowerModeEngineClock" "140000"

I'll keep my fingers crossed and report back.

Nice, tests with your multimeter would be absolutely great. You may also try to reduce EngineClock later. For example change it from 140000 to 120000. If it doesn't cause any corruptions you can still try decreasing that. Don't worry about values you may use 139999, 139998, 139997, end so on.AtomBIOS will round it to something acceptable.

-- Rafał Miłecki

Uh oh, I have some good news and I have some bad news. I did some additional testing with compiz and on cube rotate, I can't view the cube-caps any longer (the cube keeps jumping around) and when trying to rotate the cube, it will not rotate all the way around. It gets roughly 1/2 - 3/4 way around and then stops and you are left on desktop 3 or 4. Strange, I haven't seen that with the radeonhd driver before. The good news is I went ahead an used the multimeter again to confirm the temps. Still 96 Deg. F on the left palm rest and 147 Deg. F on the left fan discharge. I'll drop the EngineClock to 120000 and test again. -- David C. Rankin, J.D.,P.E. Rankin Law Firm, PLLC 510 Ochiltree Street Nacogdoches, Texas 75961 Telephone: (936) 715-9333 Facsimile: (936) 715-9339 www.rankinlawfirm.com -- To unsubscribe, e-mail: radeonhd+unsubscribe@opensuse.org For additional commands, e-mail: radeonhd+help@opensuse.org

David C. Rankin, J.D.,P.E.

05:57

New subject: [PATCH] PM patches: logic fix, feature, cleaning, memory downclocking

On or about Friday 29 May 2009 at approximately 00:41:01 David C. Rankin, J.D.,P.E. composed:

...

On or about Friday 29 May 2009 at approximately 00:01:26 Rafał Miłecki

composed:

...
2009/5/29 David C. Rankin, J.D.,P.E. :

...
Rafał,

That's what I needed to know. I'm am not familiar with 'git' and the

git checkout -t -b downclocking origin/master git am ~/archlinux/radeonhd/patches/*patch

parts were the parts I knew nothing about. I'll pull the patches, apply and report back. Something really has to give here. The heat is unbearable. To get the actual temp data, I have a digital multimeter with a great surface temperature probe on it ( Craftsman Model No. 82400 ) and I took the temperature readings on the left palm rest and left fan discharge of my laptop with the radeonhd driver active. The results:

Left Palm Rest Temp: 96 Deg. F

Left Fan Discharge: 147 Deg. F

The inside of this box is cooking.... That's with:

Option "ForceLowPowerMode" Option "LowPowerModeEngineClock" "140000"

I'll keep my fingers crossed and report back.

Nice, tests with your multimeter would be absolutely great. You may also try to reduce EngineClock later. For example change it from 140000 to 120000. If it doesn't cause any corruptions you can still try decreasing that. Don't worry about values you may use 139999, 139998, 139997, end so on.AtomBIOS will round it to something acceptable.

-- Rafał Miłecki

Uh oh,

I have some good news and I have some bad news. I did some additional testing with compiz and on cube rotate, I can't view the cube-caps any longer (the cube keeps jumping around) and when trying to rotate the cube, it will not rotate all the way around. It gets roughly 1/2 - 3/4 way around and then stops and you are left on desktop 3 or 4. Strange, I haven't seen that with the radeonhd driver before.

For testing purposes, I backed out the downclocking by setting: #if 0 in src/rhd_pm.c and compiz returned to normal. I don't know why it makes a difference, but it makes a real difference to the cube rotate function. Just one more challenge for the driver team ;-) Thanks again Rafał. -- David C. Rankin, J.D.,P.E. Rankin Law Firm, PLLC 510 Ochiltree Street Nacogdoches, Texas 75961 Telephone: (936) 715-9333 Facsimile: (936) 715-9339 www.rankinlawfirm.com -- To unsubscribe, e-mail: radeonhd+unsubscribe@opensuse.org For additional commands, e-mail: radeonhd+help@opensuse.org

Alex Deucher

13:40

New subject: [PATCH] PM patches: logic fix, feature, cleaning, memory downclocking

On Fri, May 29, 2009 at 1:57 AM, David C. Rankin, J.D.,P.E. wrote:

...

On or about Friday 29 May 2009 at approximately 00:41:01 David C. Rankin, J.D.,P.E. composed:

...
On or about Friday 29 May 2009 at approximately 00:01:26 Rafał Miłecki

composed:

...
2009/5/29 David C. Rankin, J.D.,P.E. :

...
Rafał,

That's what I needed to know. I'm am not familiar with 'git' and the

git checkout -t -b downclocking origin/master git am ~/archlinux/radeonhd/patches/*patch

parts were the parts I knew nothing about. I'll pull the patches, apply and report back. Something really has to give here. The heat is unbearable. To get the actual temp data, I have a digital multimeter with a great surface temperature probe on it ( Craftsman Model No. 82400 ) and I took the temperature readings on the left palm rest and left fan discharge of my laptop with the radeonhd driver active. The results:

Left Palm Rest Temp: 96 Deg. F

Left Fan Discharge: 147 Deg. F

The inside of this box is cooking.... That's with:

Option "ForceLowPowerMode" Option "LowPowerModeEngineClock" "140000"

I'll keep my fingers crossed and report back.

Nice, tests with your multimeter would be absolutely great. You may also try to reduce EngineClock later. For example change it from 140000 to 120000. If it doesn't cause any corruptions you can still try decreasing that. Don't worry about values you may use 139999, 139998, 139997, end so on.AtomBIOS will round it to something acceptable.

-- Rafał Miłecki

Uh oh,

I have some good news and I have some bad news. I did some additional testing with compiz and on cube rotate, I can't view the cube-caps any longer (the cube keeps jumping around) and when trying to rotate the cube, it will not rotate all the way around. It gets roughly 1/2 - 3/4 way around and then stops and you are left on desktop 3 or 4. Strange, I haven't seen that with the radeonhd driver before.

For testing purposes, I backed out the downclocking by setting:

#if 0

in src/rhd_pm.c and compiz returned to normal. I don't know why it makes a difference, but it makes a real difference to the cube rotate function. Just one more challenge for the driver team ;-)

This is the tricky part of power management. You need to slow the clock at the right times otherwise you can negatively impact the user experience. If you are going to run with reduced clocks you need to take into account the bandwidth requirements (display controllers, 2D engine, 3D engine, overlays, etc.) and latency requirements (frame has to be rendered in this amount of time). That's why the best place for power management will be in the new kms drm since only there does the driver have the full view of the hw (modesetting and command streams) so that it can change the clocks on demand in response to the current demands of the chip. Alex -- To unsubscribe, e-mail: radeonhd+unsubscribe@opensuse.org For additional commands, e-mail: radeonhd+help@opensuse.org

Conn Clark

15:50

New subject: [PATCH] PM patches: logic fix, feature, cleaning, memory downclocking

2009/5/29 Alex Deucher :

...

On Fri, May 29, 2009 at 1:57 AM, David C. Rankin, J.D.,P.E. wrote:

...
On or about Friday 29 May 2009 at approximately 00:41:01 David C. Rankin, J.D.,P.E. composed:

...
On or about Friday 29 May 2009 at approximately 00:01:26 Rafał Miłecki

composed:

...
2009/5/29 David C. Rankin, J.D.,P.E. :

...
Rafał,

That's what I needed to know. I'm am not familiar with 'git' and the

git checkout -t -b downclocking origin/master git am ~/archlinux/radeonhd/patches/*patch

parts were the parts I knew nothing about. I'll pull the patches, apply and report back. Something really has to give here. The heat is unbearable. To get the actual temp data, I have a digital multimeter with a great surface temperature probe on it ( Craftsman Model No. 82400 ) and I took the temperature readings on the left palm rest and left fan discharge of my laptop with the radeonhd driver active. The results:

Left Palm Rest Temp: 96 Deg. F

Left Fan Discharge: 147 Deg. F

The inside of this box is cooking.... That's with:

Option "ForceLowPowerMode" Option "LowPowerModeEngineClock" "140000"

I'll keep my fingers crossed and report back.

Nice, tests with your multimeter would be absolutely great. You may also try to reduce EngineClock later. For example change it from 140000 to 120000. If it doesn't cause any corruptions you can still try decreasing that. Don't worry about values you may use 139999, 139998, 139997, end so on.AtomBIOS will round it to something acceptable.

-- Rafał Miłecki

Uh oh,

I have some good news and I have some bad news. I did some additional testing with compiz and on cube rotate, I can't view the cube-caps any longer (the cube keeps jumping around) and when trying to rotate the cube, it will not rotate all the way around. It gets roughly 1/2 - 3/4 way around and then stops and you are left on desktop 3 or 4. Strange, I haven't seen that with the radeonhd driver before.

For testing purposes, I backed out the downclocking by setting:

#if 0

in src/rhd_pm.c and compiz returned to normal. I don't know why it makes a difference, but it makes a real difference to the cube rotate function. Just one more challenge for the driver team ;-)

This is the tricky part of power management. You need to slow the clock at the right times otherwise you can negatively impact the user experience. If you are going to run with reduced clocks you need to take into account the bandwidth requirements (display controllers, 2D engine, 3D engine, overlays, etc.) and latency requirements (frame has to be rendered in this amount of time). That's why the best place for power management will be in the new kms drm since only there does the driver have the full view of the hw (modesetting and command streams) so that it can change the clocks on demand in response to the current demands of the chip.

Alex -- To unsubscribe, e-mail: radeonhd+unsubscribe@opensuse.org For additional commands, e-mail: radeonhd+help@opensuse.org

I found the best way of determining the effectiveness of any power saving technique is to test it on a laptop running on a battery and look at the difference in the discharge rate of the battery with the following command. cat /proc/acpi/battery/BAT1/state I determined with the power management that is in the master branch now it saved 100mA of current or about 1.2 watts. I would expect with the memory downclocking it will be much more. -- Conn Conn O. Clark Observation: In formal computer science advances are made by standing on the shoulders of giants. Linux has proved that if there are enough of you, you can advance just as far by stepping on each others toes. -- To unsubscribe, e-mail: radeonhd+unsubscribe@opensuse.org For additional commands, e-mail: radeonhd+help@opensuse.org

Matthias Hopf

3 Jun 3 Jun

16:42

New subject: [PATCH] PM patches: logic fix, feature, cleaning, memory downclocking

On May 29, 09 09:40:13 -0400, Alex Deucher wrote:

...

This is the tricky part of power management. You need to slow the clock at the right times otherwise you can negatively impact the user experience. If you are going to run with reduced clocks you need to take into account the bandwidth requirements (display controllers, 2D engine, 3D engine, overlays, etc.) and latency requirements (frame has to be rendered in this amount of time). That's why the best place for

Also note that we're doing *no* bandwidth requirement calculation at the moment. They tend to be tricky. So the best we can do for calculating memory speeds is guessing. Which is a contradiction by itself.

...

power management will be in the new kms drm since only there does the driver have the full view of the hw (modesetting and command streams) so that it can change the clocks on demand in response to the current demands of the chip.

Also note that AFAIK there is no interface for that defined yet. Matthias -- Matthias Hopf __ __ __ Maxfeldstr. 5 / 90409 Nuernberg (_ | | (_ |__ mat@mshopf.de Phone +49-911-74053-715 __) |_| __) |__ R & D www.mshopf.de -- To unsubscribe, e-mail: radeonhd+unsubscribe@opensuse.org For additional commands, e-mail: radeonhd+help@opensuse.org

Rafał Miłecki

30 May 30 May

22:01

New subject: [PATCH] PM patches: logic fix, feature, cleaning, memory downclocking

2009/5/29 David C. Rankin, J.D.,P.E. :

...

I have some good news and I have some bad news. I did some additional testing with compiz and on cube rotate, I can't view the cube-caps any longer (the cube keeps jumping around) and when trying to rotate the cube, it will not rotate all the way around. It gets roughly 1/2 - 3/4 way around and then stops and you are left on desktop 3 or 4. Strange, I haven't seen that with the radeonhd driver before.

Thanks for your testing. It really shows us that downclocking Memory even by 50% can cause problems. I was wondering if this can happenw thi 50% only. So the only solution is to implement real power saving in KMS as Alex explained. We should calculate all clocks using current state of usage: outputs, engines, 2D, 3D. Anyway, can ANYONE comment my patches? I belive first and third should really be applied. Second and fourth optionally (as the allow to manually enable additional, not always working feature). Does someone care about radeonhd development? -- Rafał Miłecki -- To unsubscribe, e-mail: radeonhd+unsubscribe@opensuse.org For additional commands, e-mail: radeonhd+help@opensuse.org

Matthias Hopf

3 Jun 3 Jun

17:00

New subject: [PATCH] PM patches: logic fix, feature, cleaning, memory downclocking

On May 31, 09 00:01:38 +0200, Rafał Miłecki wrote:

...

Anyway, can ANYONE comment my patches? I belive first and third should really be applied. Second and fourth optionally (as the allow to manually enable additional, not always working feature). Does someone care about radeonhd development?

They are looking good - I will apply them all in a minute. As you determined yourself, memory clock changes will need more testing, also I have a tool here I have to revalidate for testing on R6xx. CU Matthias -- Matthias Hopf __ __ __ Maxfeldstr. 5 / 90409 Nuernberg (_ | | (_ |__ mat@mshopf.de Phone +49-911-74053-715 __) |_| __) |__ R & D www.mshopf.de -- To unsubscribe, e-mail: radeonhd+unsubscribe@opensuse.org For additional commands, e-mail: radeonhd+help@opensuse.org

David C. Rankin

29 May 29 May

16:58

New subject: [PATCH] PM patches: logic fix, feature, cleaning, memory downclocking

Rafał Miłecki wrote:

...

2009/5/29 David C. Rankin, J.D.,P.E. :

...
Rafał,

That's what I needed to know. I'm am not familiar with 'git' and the

git checkout -t -b downclocking origin/master git am ~/archlinux/radeonhd/patches/*patch

parts were the parts I knew nothing about. I'll pull the patches, apply and report back. Something really has to give here. The heat is unbearable. To get the actual temp data, I have a digital multimeter with a great surface temperature probe on it ( Craftsman Model No. 82400 ) and I took the temperature readings on the left palm rest and left fan discharge of my laptop with the radeonhd driver active. The results:

Left Palm Rest Temp: 96 Deg. F

Left Fan Discharge: 147 Deg. F

The inside of this box is cooking.... That's with:

Option "ForceLowPowerMode" Option "LowPowerModeEngineClock" "140000"

I'll keep my fingers crossed and report back.

Nice, tests with your multimeter would be absolutely great. You may also try to reduce EngineClock later. For example change it from 140000 to 120000. If it doesn't cause any corruptions you can still try decreasing that. Don't worry about values you may use 139999, 139998, 139997, end so on.AtomBIOS will round it to something acceptable.

Rafał, I have created a baseline temperature comparison for testing purposes. It is located at: http://www.3111skyline.com/download/linux/radeonhd/temp-comparison Here is the baseline: Temperature Comparison Between the ATI/fglrx driver and OSS radeonhd driver Hardware: Toshiba P205D; AMD Turion(tm) 64 X2 Mobile Technology TL-58 ATI Technologies Inc RS690M [Radeon X1200 Series] RAM: 4G Driver/Operating Systems: fglrx (Ver. 8.532; 8-9 Release) - openSuSE 11.0; KDE3; Compiz radeonhd - Archlinux; KDE3; Compiz git checkout -t -b downclocking origin/master git am patches/*patch vi src/rhd_pm.c (set: #if 1) Log confirmation: (II) RADEONHD: version 1.2.5, built from git branch downclocking, commit 0de8b484 + changes Operating Temperatures: fglrx - openSuSE Left Palm Rest Temp: 89 Deg. F Left Fan Discharge: 122 Deg. F radeonhd - Archlinux Left Palm Rest Temp: 96 Deg. F Left Fan Discharge: 147 Deg. F -- David C. Rankin, J.D.,P.E. Rankin Law Firm, PLLC 510 Ochiltree Street Nacogdoches, Texas 75961 Telephone: (936) 715-9333 Facsimile: (936) 715-9339 www.rankinlawfirm.com -- To unsubscribe, e-mail: radeonhd+unsubscribe@opensuse.org For additional commands, e-mail: radeonhd+help@opensuse.org

Bridgman, John

20:04

New subject: [PATCH] PM patches: logic fix, feature, cleaning, memory downclocking

Does anyone know if radeonhd turns on clock gating for this chip, either by default or with an option ? -----Original Message----- From: David C. Rankin [mailto:drankinatty@suddenlinkmail.com] Sent: Friday, May 29, 2009 12:58 PM To: radeon Subject: Re: [radeonhd] [PATCH] PM patches: logic fix, feature, cleaning, memory downclocking Rafał Miłecki wrote:

...

2009/5/29 David C. Rankin, J.D.,P.E. :

...
Rafał,

That's what I needed to know. I'm am not familiar with 'git' and the

git checkout -t -b downclocking origin/master git am ~/archlinux/radeonhd/patches/*patch

parts were the parts I knew nothing about. I'll pull the patches, apply and report back. Something really has to give here. The heat is unbearable. To get the actual temp data, I have a digital multimeter with a great surface temperature probe on it ( Craftsman Model No. 82400 ) and I took the temperature readings on the left palm rest and left fan discharge of my laptop with the radeonhd driver active. The results:

Left Palm Rest Temp: 96 Deg. F

Left Fan Discharge: 147 Deg. F

The inside of this box is cooking.... That's with:

Option "ForceLowPowerMode" Option "LowPowerModeEngineClock" "140000"

I'll keep my fingers crossed and report back.

Nice, tests with your multimeter would be absolutely great. You may also try to reduce EngineClock later. For example change it from 140000 to 120000. If it doesn't cause any corruptions you can still try decreasing that. Don't worry about values you may use 139999, 139998, 139997, end so on.AtomBIOS will round it to something acceptable.

Yang Zhao

20:35

New subject: [PATCH] PM patches: logic fix, feature, cleaning, memory downclocking

2009/5/29 Bridgman, John :

...

Does anyone know if radeonhd turns on clock gating for [RS690M], either by default or with an option ?

No clock gating controls at all for r5xx. I have old code I can forward port, though... -- Yang Zhao http://yangman.ca -- To unsubscribe, e-mail: radeonhd+unsubscribe@opensuse.org For additional commands, e-mail: radeonhd+help@opensuse.org

Bridgman, John

21:42

Alex updated the old clock gating code in radeon (xf86-video-ati) recently, might be worth looking at that also. -----Original Message----- From: yangman@gmail.com [mailto:yangman@gmail.com] On Behalf Of Yang Zhao Sent: Friday, May 29, 2009 4:35 PM To: Bridgman, John Cc: radeon Subject: Re: [radeonhd] [PATCH] PM patches: logic fix, feature, cleaning, memory downclocking 2009/5/29 Bridgman, John :

...

Does anyone know if radeonhd turns on clock gating for [RS690M], either by default or with an option ?

Matthias Hopf

3 Jun 3 Jun

17:02

New subject: [PATCH] PM patches: logic fix, feature, cleaning, memory downclocking

On May 29, 09 17:42:53 -0400, John Bridgman wrote:

...

Alex updated the old clock gating code in radeon (xf86-video-ati) recently, might be worth looking at that also.

Yes, that could really be a big gain. Matthias -- Matthias Hopf __ __ __ Maxfeldstr. 5 / 90409 Nuernberg (_ | | (_ |__ mat@mshopf.de Phone +49-911-74053-715 __) |_| __) |__ R & D www.mshopf.de -- To unsubscribe, e-mail: radeonhd+unsubscribe@opensuse.org For additional commands, e-mail: radeonhd+help@opensuse.org

Conn Clark

26 May 26 May

15:25

New subject: [PATCH] PM patches: logic fix, feature, cleaning, memory downclocking

2009/5/22 Rafał Miłecki :

...

Well, topic says everything :)

First 3 patches are quite nice, stable changes. Not sure if you will want to include 4th patch which adds code (commented out for now) for almost untested memory downclocking. I suggest commiting this, then we can ask ppl to test this. If you prefer other way, let me know.

-- Rafał Miłecki

Any chance that the future memory down clocking patch will be variable like the GPU clocking? I found I can over clock my GPU a little using the pm stuff and would like to try the same with my video ram. Thanks for the patch a few more minutes of battery time is always welcome :) -- Conn Conn O. Clark Observation: In formal computer science advances are made by standing on the shoulders of giants. Linux has proved that if there are enough of you, you can advance just as far by stepping on each others toes. -- To unsubscribe, e-mail: radeonhd+unsubscribe@opensuse.org For additional commands, e-mail: radeonhd+help@opensuse.org

Rafał Miłecki

19:26

New subject: [PATCH] PM patches: logic fix, feature, cleaning, memory downclocking

W dniu 26 maja 2009 17:25 użytkownik Conn Clark napisał:

...

2009/5/22 Rafał Miłecki :

...
Well, topic says everything :)

First 3 patches are quite nice, stable changes. Not sure if you will want to include 4th patch which adds code (commented out for now) for almost untested memory downclocking. I suggest commiting this, then we can ask ppl to test this. If you prefer other way, let me know.

-- Rafał Miłecki

Any chance that the future memory down clocking patch will be variable like the GPU clocking? I found I can over clock my GPU a little using the pm stuff and would like to try the same with my video ram.

Thanks for the patch a few more minutes of battery time is always welcome :)

If developers will accept such a patch (and will first accept the serie I sent) I can prepare it :) -- Rafał Miłecki -- To unsubscribe, e-mail: radeonhd+unsubscribe@opensuse.org For additional commands, e-mail: radeonhd+help@opensuse.org

Yang Zhao

30 May 30 May

23:06

New subject: [PATCH] PM patches: logic fix, feature, cleaning, memory downclocking

2009/5/22 Rafał Miłecki :

...

First 3 patches are quite nice, stable changes. Not sure if you will want to include 4th patch which adds code (commented out for now) for almost untested memory downclocking. I suggest commiting this, then we can ask ppl to test this. If you prefer other way, let me know.

They all look good to me. Acking 1-3 for commit. IIRC Matthias is currently on vacation until next week. Let's wait on 0004 until we have some concrete numbers that say it's beneficial to have. -- Yang Zhao http://yangman.ca -- To unsubscribe, e-mail: radeonhd+unsubscribe@opensuse.org For additional commands, e-mail: radeonhd+help@opensuse.org

5442

Age (days ago)

5454

Last active (days ago)

List overview

Download

20 comments

8 participants

participants (8)

Alex Deucher
Bridgman, John
Conn Clark
David C. Rankin
David C. Rankin, J.D.,P.E.
Matthias Hopf
Rafał Miłecki
Yang Zhao