| Message ID | 1409548719.13507.13.camel@localhost.localdomain (mailing list archive) |
|---|---|
| State | Superseded, archived |
| Headers | show |
On 1 September 2014 10:48, Shilpa Bhat <shilpa.bhat@linux.vnet.ibm.com> wrote: > Hi Viresh, > On Fri, 2014-08-29 at 05:33 +0530, Viresh Kumar wrote: >> On 28 August 2014 19:36, Shilpasri G Bhat >> <shilpa.bhat@linux.vnet.ibm.com> wrote: >> > >> > Changes v1->v2: >> > Invoke .target() driver callback to set the cpus to nominal frequency >> > in reboot notifier, instead of calling cpufreq_suspend() as suggested >> > by Viresh Kumar. >> > Modified the commit message. >> >> This changelog will get commited, is this what you want? You skipped replying to this and commited the same mistake again.. > > This patch ensures the cpus to kexec/reboot at nominal frequency. > Nominal frequency is the highest cpu frequency on PowerPC at > which the cores can run without getting throttled. > > If the host kernel had set the cpus to a low pstate and then it > kexecs/reboots to a cpufreq disabled kernel it would cause the target > kernel to perform poorly. It will also increase the boot up time of > the target kernel. So set the cpus to high pstate, in this case to > nominal frequency before rebooting to avoid such scenarios. > > The reboot notifier will set the cpus to nominal frequncy. > > Changes v1->v2: > Invoke .target() driver callback to set the cpus to nominal frequency > in reboot notifier, instead of calling cpufreq_suspend() as suggested > by Viresh Kumar. > Modified the commit message. This would get commited in git changelogs and probably you don't want this. So please place them after the three dash '-' symbols below.. > Signed-off-by: Shilpasri G Bhat <shilpa.bhat@linux.vnet.ibm.com> > Suggested-by: Viresh Kumar <viresh.kumar@linaro.org> > Reviewed-by: Preeti U Murthy <preeti@linux.vnet.ibm.com> > --- HERE>>>>>>>>>>> > drivers/cpufreq/powernv-cpufreq.c | 35 +++++++++++++++++++++++++++++++++++ > 1 file changed, 35 insertions(+) I have already Acked this, so you could have added that yourself on this resend.
From: Shilpa Bhat > Hi Viresh, > On Fri, 2014-08-29 at 05:33 +0530, Viresh Kumar wrote: > > On 28 August 2014 19:36, Shilpasri G Bhat > > <shilpa.bhat@linux.vnet.ibm.com> wrote: > > > > > > Changes v1->v2: > > > Invoke .target() driver callback to set the cpus to nominal frequency > > > in reboot notifier, instead of calling cpufreq_suspend() as suggested > > > by Viresh Kumar. > > > Modified the commit message. > > > > This changelog will get commited, is this what you want? > > > > + if (unlikely(rebooting) && new_index != get_nominal_index()) > > > + return -EBUSY; > > > > Have you placed the unlikely only around 'rebooting' intentionally or > > should it cover whole if statement? > > > > Yes unlikely() should cover the whole if statement... Actually it probably shouldn't. You need to look at the generated code with each different set of 'unlikely()' to see how gcc processes them. In this case, if 'rebooting' is false you want to 'fall through' on a statically predicted 'not taken' branch. You don't ever care about the second clause. With an 'unlikely' covering the entire statement gcc could easily add a forwards conditional branch (that will be mis-predicted) for the 'rebooting' test. (Yes, I spent a lot of time getting gcc to generate branches that were correctly statically predicted for some code where every cycle mattered.) David
On 09/01/2014 02:42 PM, David Laight wrote: >> Yes unlikely() should cover the whole if statement... > > Actually it probably shouldn't. > You need to look at the generated code with each different set of 'unlikely()' > to see how gcc processes them. > In this case, if 'rebooting' is false you want to 'fall through' on a statically > predicted 'not taken' branch. You don't ever care about the second clause. > With an 'unlikely' covering the entire statement gcc could easily add a > forwards conditional branch (that will be mis-predicted) for the 'rebooting' test. > > (Yes, I spent a lot of time getting gcc to generate branches that were > correctly statically predicted for some code where every cycle mattered.) > > David > Hi David, The objdup with an 'unlikely()' covering the entire if statement is as follows: if (unlikely(rebooting && new_index != get_nominal_index())) return -EBUSY; 1ac: 2f 89 00 00 cmpwi cr7,r9,0 /* compare rebooting,0 */ 1b0: 40 de 00 4c bne- cr7,1fc <.powernv_cpufreq_target_index+0x7c> The '-' in the instruction bne- specifies an unlikely branch. So gcc has processed the first clause to be identified as an unlikely branch i.e, branch to <1fc> (to test the second clause) is unlikely on 'rebooting' not equal to 0. 1b4: 1f ff 00 0c mulli r31,r31,12 . . <--- Set the frequency and return ---> . . 1fc: 3d 22 00 00 addis r9,r2,0 /* test the second clause */ 200: 3d 02 00 00 addis r8,r2,0 204: 81 49 00 00 lwz r10,0(r9) 208: 81 28 00 00 lwz r9,0(r8) 20c: 7d 29 50 50 subf r9,r9,r10 210: 7f 89 f8 00 cmpw cr7,r9,r31 /* compare new_index,nominal_index */ 214: 41 9e ff a0 beq+ cr7,1b4 <.powernv_cpufreq_target_index+0x34> The '+' in the instruction beq+ specifies a likely branch. The second clause unlikely(new_index != get_nominal_index()) is processed to likely(new_index == get_nominal_index()). 218: 38 60 ff f0 li r3,-16 /* return -EBUSY */ 21c: 4b ff ff cc b 1e8 <.powernv_cpufreq_target_index+0x68> So unlikely() covering the entire statement will not lead to a branch mis-prediction for the 'rebooting' test. Having unlikely to cover both 'rebooting' and the second clause we can avoid the branch miss prediction for the second clause. This is advantageous for the code path powernv_cpufreq_target_index(policy,nominal_index) which will be invoked by the reboot_notifier. Thanks and Regards, Shilpa
diff --git a/drivers/cpufreq/powernv-cpufreq.c b/drivers/cpufreq/powernv-cpufreq.c index 379c083..f8b83c8 100644 --- a/drivers/cpufreq/powernv-cpufreq.c +++ b/drivers/cpufreq/powernv-cpufreq.c @@ -26,6 +26,7 @@ #include <linux/cpufreq.h> #include <linux/smp.h> #include <linux/of.h> +#include <linux/reboot.h> #include <asm/cputhreads.h> #include <asm/firmware.h> @@ -35,6 +36,7 @@ #define POWERNV_MAX_PSTATES 256 static struct cpufreq_frequency_table powernv_freqs[POWERNV_MAX_PSTATES+1]; +static bool rebooting; /* * Note: The set of pstates consists of contiguous integers, the @@ -284,6 +286,15 @@ static void set_pstate(void *freq_data) } /* + * get_nominal_index: Returns the index corresponding to the nominal + * pstate in the cpufreq table + */ +static inline unsigned int get_nominal_index(void) +{ + return powernv_pstate_info.max - powernv_pstate_info.nominal; +} + +/* * powernv_cpufreq_target_index: Sets the frequency corresponding to * the cpufreq table entry indexed by new_index on the cpus in the * mask policy->cpus @@ -293,6 +304,9 @@ static int powernv_cpufreq_target_index(struct cpufreq_policy *policy, { struct powernv_smp_call_data freq_data; + if (unlikely(rebooting && new_index != get_nominal_index())) + return -EBUSY; + freq_data.pstate_id = powernv_freqs[new_index].driver_data; /* @@ -317,6 +331,25 @@ static int powernv_cpufreq_cpu_init(struct cpufreq_policy *policy) return cpufreq_table_validate_and_show(policy, powernv_freqs); } +static int powernv_cpufreq_reboot_notifier(struct notifier_block *nb, + unsigned long action, void *unused) +{ + int cpu; + struct cpufreq_policy cpu_policy; + + rebooting = true; + for_each_online_cpu(cpu) { + cpufreq_get_policy(&cpu_policy, cpu); + powernv_cpufreq_target_index(&cpu_policy, get_nominal_index()); + } + + return NOTIFY_DONE; +} + +static struct notifier_block powernv_cpufreq_reboot_nb = { + .notifier_call = powernv_cpufreq_reboot_notifier, +}; + static struct cpufreq_driver powernv_cpufreq_driver = { .name = "powernv-cpufreq", .flags = CPUFREQ_CONST_LOOPS, @@ -342,12 +375,14 @@ static int __init powernv_cpufreq_init(void) return rc; } + register_reboot_notifier(&powernv_cpufreq_reboot_nb); return cpufreq_register_driver(&powernv_cpufreq_driver); } module_init(powernv_cpufreq_init); static void __exit powernv_cpufreq_exit(void) { + unregister_reboot_notifier(&powernv_cpufreq_reboot_nb); cpufreq_unregister_driver(&powernv_cpufreq_driver); } module_exit(powernv_cpufreq_exit);