| Message ID | 20130518014724.GA4006@linux.vnet.ibm.com (mailing list archive) |
|---|---|
| State | Not Applicable, archived |
| Headers | show |
On 05/18/13 03:47, Paul E. McKenney wrote: > On Fri, May 17, 2013 at 10:49:16PM +0200, Arend van Spriel wrote: >> On 05/14/2013 10:38 PM, Arend van Spriel wrote: >>> On 05/14/2013 10:34 PM, Rafael J. Wysocki wrote: >>>> On Tuesday, May 14, 2013 03:03:58 PM Arend van Spriel wrote: >>>>> Laptop hangs pretty soon after booting. Workaround for me was to turn >>>>> off ACPI on kernel command line, ie. acpi=off >>>>> >>>>> Attached is my kernel configuration. >>>> >>>> Well, I have no idea what may be the source of this. >>> >>> Me neither. >>> >>>> Any chance to bisect? >>> >>> can try. Rebuilding kernels can be a drag :-p >>> >>> Regards, >>> Arend >> >> Hi Rafael, >> >> The bisect result found commit c0f4dfd4f as culprit. I also verified >> that the merge into Linus' tree introduced the issue upstream, ie. >> 6c24499 works and 1f889ec does not. I tried to revert the commit, >> but that was a bit risky (conflict to resolve) and did not give me a >> stable kernel. The only thing that works is selecting acpi=off. I >> tried also nohz=off but that resulted in /init failure. >> >> Regards, >> Arend >> >> ---8<---------------------------------------------------------------- >> >> commit 1f889ec62c3f0d8913f3c32f9aff2a1e15099346 >> Merge: 6c24499 8fcfae3 >> >> commit 8fcfae31719c0a6c03f2cf63f815b46d378d8be4 >> Merge: d02a9a8 6d87669 >> >> commit 6d87669357936bffa1e8fea7a4e7743e76905736 >> Merge: 3f944ad 81e5949 910ee45 >> >> commit c0f4dfd4f90f1667d234d21f15153ea09a2eaa66 >> Author: Paul E. McKenney<paul.mckenney@linaro.org> >> Date: Fri Dec 28 11:30:36 2012 -0800 >> >> rcu: Make RCU_FAST_NO_HZ take advantage of numbered callbacks >> >> Because RCU callbacks are now associated with the number >> of the grace period that they must wait for, CPUs can now >> take advance callbacks corresponding to grace periods that >> ended while a given CPU was in dyntick-idle mode. This >> eliminates the need to try forcing the RCU state machine >> while entering idle, thus reducing the CPU intensiveness >> of RCU_FAST_NO_HZ, which should increase its energy >> efficiency. >> >> Signed-off-by: Paul E. McKenney<paul.mckenney@linaro.org> >> Signed-off-by: Paul E. McKenney<paulmck@linux.vnet.ibm.com> > > Hello, Arend, > > Thank you for tracking this down! Could you please try out the following > patch? Thanks, Paul That did the trick. Made me bumping into another regression in the driver I am maintaining so more fun to do :-) Looking at the commit message I guess it is already in some tree. Otherwise I would have said: "You may add Tested-by: Arend van Spriel <arend@broadcom.com>" > ------------------------------------------------------------------------ > > rcu: Fix comparison sense in rcu_needs_cpu() > > Commit c0f4dfd4f (rcu: Make RCU_FAST_NO_HZ take advantage of numbered > callbacks) introduced a bug that can result in excessively long grace > periods. This bug reverse the senes of the "if" statement checking Oh, there is a typo in the commit message: sense iso senes. > for lazy callbacks, so that RCU takes a lazy approach when there are > in fact non-lazy callbacks. This can result in excessive boot, suspend, > and resume times. > > This commit therefore fixes the sense of this "if" statement. > > Reported-by: Borislav Petkov<bp@alien8.de> > Reported-by: Bjørn Mork<bjorn@mork.no> > Reported-by: Joerg Roedel<joro@8bytes.org> > Signed-off-by: Paul E. McKenney<paulmck@linux.vnet.ibm.com> > Tested-by: Bjørn Mork<bjorn@mork.no> > Tested-by: Joerg Roedel<joro@8bytes.org> Gr. AvS -- To unsubscribe from this list: send the line "unsubscribe linux-acpi" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Sat, May 18, 2013 at 09:14:25AM +0200, Arend van Spriel wrote: > On 05/18/13 03:47, Paul E. McKenney wrote: > >On Fri, May 17, 2013 at 10:49:16PM +0200, Arend van Spriel wrote: > >>On 05/14/2013 10:38 PM, Arend van Spriel wrote: > >>>On 05/14/2013 10:34 PM, Rafael J. Wysocki wrote: > >>>>On Tuesday, May 14, 2013 03:03:58 PM Arend van Spriel wrote: > >>>>>Laptop hangs pretty soon after booting. Workaround for me was to turn > >>>>>off ACPI on kernel command line, ie. acpi=off > >>>>> > >>>>>Attached is my kernel configuration. > >>>> > >>>>Well, I have no idea what may be the source of this. > >>> > >>>Me neither. > >>> > >>>>Any chance to bisect? > >>> > >>>can try. Rebuilding kernels can be a drag :-p > >>> > >>>Regards, > >>>Arend > >> > >>Hi Rafael, > >> > >>The bisect result found commit c0f4dfd4f as culprit. I also verified > >>that the merge into Linus' tree introduced the issue upstream, ie. > >>6c24499 works and 1f889ec does not. I tried to revert the commit, > >>but that was a bit risky (conflict to resolve) and did not give me a > >>stable kernel. The only thing that works is selecting acpi=off. I > >>tried also nohz=off but that resulted in /init failure. > >> > >>Regards, > >>Arend > >> > >>---8<---------------------------------------------------------------- > >> > >>commit 1f889ec62c3f0d8913f3c32f9aff2a1e15099346 > >> Merge: 6c24499 8fcfae3 > >> > >> commit 8fcfae31719c0a6c03f2cf63f815b46d378d8be4 > >> Merge: d02a9a8 6d87669 > >> > >> commit 6d87669357936bffa1e8fea7a4e7743e76905736 > >> Merge: 3f944ad 81e5949 910ee45 > >> > >> commit c0f4dfd4f90f1667d234d21f15153ea09a2eaa66 > >> Author: Paul E. McKenney<paul.mckenney@linaro.org> > >> Date: Fri Dec 28 11:30:36 2012 -0800 > >> > >> rcu: Make RCU_FAST_NO_HZ take advantage of numbered callbacks > >> > >> Because RCU callbacks are now associated with the number > >> of the grace period that they must wait for, CPUs can now > >> take advance callbacks corresponding to grace periods that > >> ended while a given CPU was in dyntick-idle mode. This > >> eliminates the need to try forcing the RCU state machine > >> while entering idle, thus reducing the CPU intensiveness > >> of RCU_FAST_NO_HZ, which should increase its energy > >> efficiency. > >> > >> Signed-off-by: Paul E. McKenney<paul.mckenney@linaro.org> > >> Signed-off-by: Paul E. McKenney<paulmck@linux.vnet.ibm.com> > > > >Hello, Arend, > > > >Thank you for tracking this down! Could you please try out the following > >patch? > > Thanks, Paul > > That did the trick. Made me bumping into another regression in the > driver I am maintaining so more fun to do :-) > > Looking at the commit message I guess it is already in some tree. > Otherwise I would have said: "You may add > > Tested-by: Arend van Spriel <arend@broadcom.com>" Thank you for trying it out, and glad that it worked for you! And I did already send the pull request. But if I need to re-send it, I will update to add your Tested-by. Thanx, Paul > >------------------------------------------------------------------------ > > > >rcu: Fix comparison sense in rcu_needs_cpu() > > > >Commit c0f4dfd4f (rcu: Make RCU_FAST_NO_HZ take advantage of numbered > >callbacks) introduced a bug that can result in excessively long grace > >periods. This bug reverse the senes of the "if" statement checking > > Oh, there is a typo in the commit message: sense iso senes. > > >for lazy callbacks, so that RCU takes a lazy approach when there are > >in fact non-lazy callbacks. This can result in excessive boot, suspend, > >and resume times. > > > >This commit therefore fixes the sense of this "if" statement. > > > >Reported-by: Borislav Petkov<bp@alien8.de> > >Reported-by: Bjørn Mork<bjorn@mork.no> > >Reported-by: Joerg Roedel<joro@8bytes.org> > >Signed-off-by: Paul E. McKenney<paulmck@linux.vnet.ibm.com> > >Tested-by: Bjørn Mork<bjorn@mork.no> > >Tested-by: Joerg Roedel<joro@8bytes.org> > > Gr. AvS > -- To unsubscribe from this list: send the line "unsubscribe linux-acpi" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
diff --git a/kernel/rcutree_plugin.h b/kernel/rcutree_plugin.h index 170814d..6d939a6 100644 --- a/kernel/rcutree_plugin.h +++ b/kernel/rcutree_plugin.h @@ -1667,7 +1667,7 @@ int rcu_needs_cpu(int cpu, unsigned long *dj) rdtp->last_accelerate = jiffies; /* Request timer delay depending on laziness, and round. */ - if (rdtp->all_lazy) { + if (!rdtp->all_lazy) { *dj = round_up(rcu_idle_gp_delay + jiffies, rcu_idle_gp_delay) - jiffies; } else {