PM_QOS updates for 2.6.34-rc1
PATCH 1/1 PM_QOS updates for 2.6.34-rc1
The following patch series implements to following:
1) pm_qos now uses a handle based implementation. No more string compares while walking lists.
2) renames the exports to using more accurate names.
3) updates kernel clients of pm_qos per this change
4) added ascii interface to pmqos recognizing strings formatted in 10 charactor format "0x12345678"
5) added new pm_qos class, "system_bus_throughput"
I would like to see this series make it into 2.6.34 if its not too late.
I was able to test it on my laptop and comments welcome.
Thank you!
--mgross
Quoted 10 lines · On Sat 2010-03-13 16:34:25, 640E9920 wrote:
The following patch series implements to following:
1)pm_qosnow uses a handle based implementation. No more string compares while walking lists.
2) renames the exports to using more accurate names.
3) updates kernel clients ofpm_qosper this change
4) added ascii interface to pmqos recognizing strings formatted in 10 charactor format "0x12345678"
5) added newpm_qosclass, "system_bus_throughput"I would like to see this series make it into 2.6.34 if its not too late.
I believe it is... -rc1 is past us.
Also... the way the patches are structured... that will break bisect, no?
Pavel
Signature · 2 lines
(english) http://www.livejournal.com/~pavelmachek (cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html
Quoted 10 lines · On Sat 2010-03-13 16:34:25, 640E9920 wrote:
The following patch series implements to following:
1)pm_qosnow uses a handle based implementation. No more string compares while walking lists.
2) renames the exports to using more accurate names.
3) updates kernel clients ofpm_qosper this change
4) added ascii interface to pmqos recognizing strings formatted in 10 charactor format "0x12345678"
5) added newpm_qosclass, "system_bus_throughput"I would like to see this series make it into 2.6.34 if its not too late.
I believe it is... -rc1 is past us.
Also... the way the patches are structured... that will break bisect, no?
Pavel
Signature · 2 lines
(english) http://www.livejournal.com/~pavelmachek (cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html
Quoted 10 lines · Pavel Machek
Quoted 9 lines · On Sat 2010-03-13 16:34:25, 640E9920 wrote:
The following patch series implements to following:
1)pm_qosnow uses a handle based implementation. No more string compares while walking lists.
2) renames the exports to using more accurate names.
3) updates kernel clients ofpm_qosper this change
4) added ascii interface to pmqos recognizing strings formatted in 10 charactor format "0x12345678"
5) added newpm_qosclass, "system_bus_throughput"
Quoted 3 lines
Quoted 1 line
I would like to see this series make it into 2.6.34 if its not too late.
I believe it is... -rc1 is past us.
It definitely is.
Quoted 1 line
Also... the way the patches are structured... that will break bisect, no?
Not only that. Some changelogs would be heartily welcome. In short, what change the particular patch is going to make and why it's useful.
Rafael
Quoted 10 lines · Pavel Machek
Quoted 9 lines · On Sat 2010-03-13 16:34:25, 640E9920 wrote:
The following patch series implements to following:
1)pm_qosnow uses a handle based implementation. No more string compares while walking lists.
2) renames the exports to using more accurate names.
3) updates kernel clients ofpm_qosper this change
4) added ascii interface to pmqos recognizing strings formatted in 10 charactor format "0x12345678"
5) added newpm_qosclass, "system_bus_throughput"
Quoted 3 lines
Quoted 1 line
I would like to see this series make it into 2.6.34 if its not too late.
I believe it is... -rc1 is past us.
It definitely is.
Quoted 1 line
Also... the way the patches are structured... that will break bisect, no?
Not only that. Some changelogs would be heartily welcome. In short, what change the particular patch is going to make and why it's useful.
Rafael
Quoted 10 lines · Pavel Machek
Quoted 9 lines · On Sat 2010-03-13 16:34:25, 640E9920 wrote:
The following patch series implements to following:
1)pm_qosnow uses a handle based implementation. No more string compares while walking lists.
2) renames the exports to using more accurate names.
3) updates kernel clients ofpm_qosper this change
4) added ascii interface to pmqos recognizing strings formatted in 10 charactor format "0x12345678"
5) added newpm_qosclass, "system_bus_throughput"
Quoted 3 lines
Quoted 1 line
I would like to see this series make it into 2.6.34 if its not too late.
I believe it is... -rc1 is past us.
crap. I'll guess I'll redo them for next then.
Quoted 2 lines
Also... the way the patches are structured... that will break bisect, no?
Um not sure, how would you recomend I structure them?
--mgross
Quoted 4 lines
Pavel
Signature · 2 lines
(english) http://www.livejournal.com/~pavelmachek (cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html
Quoted 10 lines · Pavel Machek
Quoted 9 lines · On Sat 2010-03-13 16:34:25, 640E9920 wrote:
The following patch series implements to following:
1)pm_qosnow uses a handle based implementation. No more string compares while walking lists.
2) renames the exports to using more accurate names.
3) updates kernel clients ofpm_qosper this change
4) added ascii interface to pmqos recognizing strings formatted in 10 charactor format "0x12345678"
5) added newpm_qosclass, "system_bus_throughput"
Quoted 3 lines
Quoted 1 line
I would like to see this series make it into 2.6.34 if its not too late.
I believe it is... -rc1 is past us.
crap. I'll guess I'll redo them for next then.
Quoted 2 lines
Also... the way the patches are structured... that will break bisect, no?
Um not sure, how would you recomend I structure them?
--mgross
Quoted 4 lines
Pavel
Signature · 2 lines
(english) http://www.livejournal.com/~pavelmachek (cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html
Quoted 15 lines · mark gross
Quoted 10 lines · Pavel Machek
Quoted 9 lines · On Sat 2010-03-13 16:34:25, 640E9920 wrote:
The following patch series implements to following:
1)pm_qosnow uses a handle based implementation. No more string compares while walking lists.
2) renames the exports to using more accurate names.
3) updates kernel clients ofpm_qosper this change
4) added ascii interface to pmqos recognizing strings formatted in 10 charactor format "0x12345678"
5) added newpm_qosclass, "system_bus_throughput"Quoted 3 lines
Quoted 1 line
I would like to see this series make it into 2.6.34 if its not too late.
I believe it is... -rc1 is past us.
Quoted 8 lines
crap. I'll guess I'll redo them for next then.
Quoted 2 lines
Also... the way the patches are structured... that will break bisect, no?
Um not sure, how would you recomend I structure them?
Generally speaking, the kernel should build and should work correctly after applying each consecutive patch in the series.
Rafael
Quoted 20 lines · Rafael J. Wysocki
Quoted 19 lines · mark gross
Quoted 10 lines · Pavel Machek
Quoted 9 lines · On Sat 2010-03-13 16:34:25, 640E9920 wrote:
The following patch series implements to following:
1)pm_qosnow uses a handle based implementation. No more string compares while walking lists.
2) renames the exports to using more accurate names.
3) updates kernel clients ofpm_qosper this change
4) added ascii interface to pmqos recognizing strings formatted in 10 charactor format "0x12345678"
5) added newpm_qosclass, "system_bus_throughput"Quoted 3 lines
Quoted 1 line
I would like to see this series make it into 2.6.34 if its not too late.
I believe it is... -rc1 is past us.
crap. I'll guess I'll redo them for next then.
Quoted 7 lines
Quoted 4 lines
Quoted 2 lines
Also... the way the patches are structured... that will break bisect, no?
Um not sure, how would you recomend I structure them?
Generally speaking, the kernel should build and should work correctly after applying each consecutive patch in the series.
I think you would need to have the handle based method along side the old method until all the kernel clients are converted.. Then remove the old method ..
Daniel
Quoted 20 lines · Rafael J. Wysocki
Quoted 19 lines · mark gross
Quoted 10 lines · Pavel Machek
Quoted 9 lines · On Sat 2010-03-13 16:34:25, 640E9920 wrote:
The following patch series implements to following:
1)pm_qosnow uses a handle based implementation. No more string compares while walking lists.
2) renames the exports to using more accurate names.
3) updates kernel clients ofpm_qosper this change
4) added ascii interface to pmqos recognizing strings formatted in 10 charactor format "0x12345678"
5) added newpm_qosclass, "system_bus_throughput"Quoted 3 lines
Quoted 1 line
I would like to see this series make it into 2.6.34 if its not too late.
I believe it is... -rc1 is past us.
crap. I'll guess I'll redo them for next then.
Quoted 7 lines
Quoted 4 lines
Quoted 2 lines
Also... the way the patches are structured... that will break bisect, no?
Um not sure, how would you recomend I structure them?
Generally speaking, the kernel should build and should work correctly after applying each consecutive patch in the series.
I think you would need to have the handle based method along side the old method until all the kernel clients are converted.. Then remove the old method ..
Daniel
Quoted 15 lines · mark gross
Quoted 10 lines · Pavel Machek
Quoted 9 lines · On Sat 2010-03-13 16:34:25, 640E9920 wrote:
The following patch series implements to following:
1)pm_qosnow uses a handle based implementation. No more string compares while walking lists.
2) renames the exports to using more accurate names.
3) updates kernel clients ofpm_qosper this change
4) added ascii interface to pmqos recognizing strings formatted in 10 charactor format "0x12345678"
5) added newpm_qosclass, "system_bus_throughput"Quoted 3 lines
Quoted 1 line
I would like to see this series make it into 2.6.34 if its not too late.
I believe it is... -rc1 is past us.
Quoted 8 lines
crap. I'll guess I'll redo them for next then.
Quoted 2 lines
Also... the way the patches are structured... that will break bisect, no?
Um not sure, how would you recomend I structure them?
Generally speaking, the kernel should build and should work correctly after applying each consecutive patch in the series.
Rafael