Message ID | TYCP286MB214616F6AE3E088C5D781A2EC6C8A@TYCP286MB2146.JPNP286.PROD.OUTLOOK.COM (mailing list archive) |
---|---|
State | Accepted, archived |
Delegated to: | Len Brown |
Headers | show |
Series | tools/power turbostat: Fix Bzy_MHz calculation equation | expand |
Ping On 2023/10/7 13:46, Peng Liu wrote: > From: Peng Liu <liupeng17@lenovo.com> > > To calculate Bzy_MHz, TSC_delta should multiply APERF_delta instead > of dividing it. > > Signed-off-by: Peng Liu <liupeng17@lenovo.com> > --- > tools/power/x86/turbostat/turbostat.8 | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git a/tools/power/x86/turbostat/turbostat.8 b/tools/power/x86/turbostat/turbostat.8 > index 8f08c3fd498d..1ba6340d3b3d 100644 > --- a/tools/power/x86/turbostat/turbostat.8 > +++ b/tools/power/x86/turbostat/turbostat.8 > @@ -370,7 +370,7 @@ below the processor's base frequency. > > Busy% = MPERF_delta/TSC_delta > > -Bzy_MHz = TSC_delta/APERF_delta/MPERF_delta/measurement_interval > +Bzy_MHz = TSC_delta*APERF_delta/MPERF_delta/measurement_interval > > Note that these calculations depend on TSC_delta, so they > are not reliable during intervals when TSC_MHz is not running at the base frequency.
Hi Peng and Wyes, Just for your information about your recent "ping"s about your turbostat patches: Please be aware that there is rarely any reply. There tends to be about 2 turbostat updates released per year. Your patches are in patchworks [1], as are some others, including an old one of mine from 2023.04.03 Eventually there will be an update, and hopefully it will include all our patches. [1] https://patchwork.kernel.org/project/linux-pm/list/?series=&submitter=&state=&q=&archive=&delegate=107 ... Doug On 2023.10.29 08:12 Peng Liu wrote: > Ping > > On 2023/10/7 13:46, Peng Liu wrote: >> From: Peng Liu <liupeng17@lenovo.com> >> >> To calculate Bzy_MHz, TSC_delta should multiply APERF_delta instead >> of dividing it. >> >> Signed-off-by: Peng Liu <liupeng17@lenovo.com> >> --- >> tools/power/x86/turbostat/turbostat.8 | 2 +- >> 1 file changed, 1 insertion(+), 1 deletion(-) >> >> diff --git a/tools/power/x86/turbostat/turbostat.8 b/tools/power/x86/turbostat/turbostat.8 >> index 8f08c3fd498d..1ba6340d3b3d 100644 >> --- a/tools/power/x86/turbostat/turbostat.8 >> +++ b/tools/power/x86/turbostat/turbostat.8 >> @@ -370,7 +370,7 @@ below the processor's base frequency. >> >> Busy% = MPERF_delta/TSC_delta >> >> -Bzy_MHz = TSC_delta/APERF_delta/MPERF_delta/measurement_interval >> +Bzy_MHz = TSC_delta*APERF_delta/MPERF_delta/measurement_interval >> >> Note that these calculations depend on TSC_delta, so they >> are not reliable during intervals when TSC_MHz is not running at the base frequency.
Hi Doug, Thanks for your information. I wonder whether a submitter will be notified if his/her patch is not accepted. Can I assume that my patch is accepted if no reply is received? Regards, Peng On 2023/11/2 23:31, Doug Smythies wrote: > Hi Peng and Wyes, > > Just for your information about your recent "ping"s about your turbostat patches: > > Please be aware that there is rarely any reply. There tends to be about 2 turbostat updates released per year. Your patches are in patchworks [1], as are some others, including an old one of mine from 2023.04.03 > Eventually there will be an update, and hopefully it will include all our patches. > > [1] https://patchwork.kernel.org/project/linux-pm/list/?series=&submitter=&state=&q=&archive=&delegate=107 > > ... Doug > > On 2023.10.29 08:12 Peng Liu wrote: > >> Ping >> >> On 2023/10/7 13:46, Peng Liu wrote: >>> From: Peng Liu <liupeng17@lenovo.com> >>> >>> To calculate Bzy_MHz, TSC_delta should multiply APERF_delta instead >>> of dividing it. >>> >>> Signed-off-by: Peng Liu <liupeng17@lenovo.com> >>> --- >>> tools/power/x86/turbostat/turbostat.8 | 2 +- >>> 1 file changed, 1 insertion(+), 1 deletion(-) >>> >>> diff --git a/tools/power/x86/turbostat/turbostat.8 b/tools/power/x86/turbostat/turbostat.8 >>> index 8f08c3fd498d..1ba6340d3b3d 100644 >>> --- a/tools/power/x86/turbostat/turbostat.8 >>> +++ b/tools/power/x86/turbostat/turbostat.8 >>> @@ -370,7 +370,7 @@ below the processor's base frequency. >>> >>> Busy% = MPERF_delta/TSC_delta >>> >>> -Bzy_MHz = TSC_delta/APERF_delta/MPERF_delta/measurement_interval >>> +Bzy_MHz = TSC_delta*APERF_delta/MPERF_delta/measurement_interval >>> >>> Note that these calculations depend on TSC_delta, so they >>> are not reliable during intervals when TSC_MHz is not running at the base frequency.
On 02 Nov 08:31, Doug Smythies wrote: > Hi Peng and Wyes, > > Just for your information about your recent "ping"s about your turbostat patches: > > Please be aware that there is rarely any reply. There tends to be about 2 turbostat updates released per year. Your patches are in patchworks [1], as are some others, including an old one of mine from 2023.04.03 > Eventually there will be an update, and hopefully it will include all our patches. Thanks Doug for the information. Regards, Wyes > > [1] https://patchwork.kernel.org/project/linux-pm/list/?series=&submitter=&state=&q=&archive=&delegate=107 > > ... Doug > > On 2023.10.29 08:12 Peng Liu wrote: > > > Ping > > > > On 2023/10/7 13:46, Peng Liu wrote: > >> From: Peng Liu <liupeng17@lenovo.com> > >> > >> To calculate Bzy_MHz, TSC_delta should multiply APERF_delta instead > >> of dividing it. > >> > >> Signed-off-by: Peng Liu <liupeng17@lenovo.com> > >> --- > >> tools/power/x86/turbostat/turbostat.8 | 2 +- > >> 1 file changed, 1 insertion(+), 1 deletion(-) > >> > >> diff --git a/tools/power/x86/turbostat/turbostat.8 b/tools/power/x86/turbostat/turbostat.8 > >> index 8f08c3fd498d..1ba6340d3b3d 100644 > >> --- a/tools/power/x86/turbostat/turbostat.8 > >> +++ b/tools/power/x86/turbostat/turbostat.8 > >> @@ -370,7 +370,7 @@ below the processor's base frequency. > >> > >> Busy% = MPERF_delta/TSC_delta > >> > >> -Bzy_MHz = TSC_delta/APERF_delta/MPERF_delta/measurement_interval > >> +Bzy_MHz = TSC_delta*APERF_delta/MPERF_delta/measurement_interval > >> > >> Note that these calculations depend on TSC_delta, so they > >> are not reliable during intervals when TSC_MHz is not running at the base frequency. >
Peng Liu, Thanks for the close read of the documentation, and the patch. I've applied it, though I re-wrote the commit message (below) to make it clear it is a documentation fix, as the code was already correct. thanks, -Len ps. I would have replied to your original e-mail, but there was a period where gmail dropped mail due to space limitations. No worries, as Doug said, everything is in patchwork. tools/power turbostat: Fix Bzy_MHz documentation typo The code calculates Bzy_MHz by multiplying TSC_delta * APERF_delta/MPERF_delta The man page erroneously showed that TSC_delta was divided. On Thu, Nov 2, 2023 at 11:31 AM Doug Smythies <dsmythies@telus.net> wrote: > > Hi Peng and Wyes, > > Just for your information about your recent "ping"s about your turbostat patches: > > Please be aware that there is rarely any reply. There tends to be about 2 turbostat updates released per year. Your patches are in patchworks [1], as are some others, including an old one of mine from 2023.04.03 > Eventually there will be an update, and hopefully it will include all our patches. > > [1] https://patchwork.kernel.org/project/linux-pm/list/?series=&submitter=&state=&q=&archive=&delegate=107 > > ... Doug > > On 2023.10.29 08:12 Peng Liu wrote: > > > Ping > > > > On 2023/10/7 13:46, Peng Liu wrote: > >> From: Peng Liu <liupeng17@lenovo.com> > >> > >> To calculate Bzy_MHz, TSC_delta should multiply APERF_delta instead > >> of dividing it. > >> > >> Signed-off-by: Peng Liu <liupeng17@lenovo.com> > >> --- > >> tools/power/x86/turbostat/turbostat.8 | 2 +- > >> 1 file changed, 1 insertion(+), 1 deletion(-) > >> > >> diff --git a/tools/power/x86/turbostat/turbostat.8 b/tools/power/x86/turbostat/turbostat.8 > >> index 8f08c3fd498d..1ba6340d3b3d 100644 > >> --- a/tools/power/x86/turbostat/turbostat.8 > >> +++ b/tools/power/x86/turbostat/turbostat.8 > >> @@ -370,7 +370,7 @@ below the processor's base frequency. > >> > >> Busy% = MPERF_delta/TSC_delta > >> > >> -Bzy_MHz = TSC_delta/APERF_delta/MPERF_delta/measurement_interval > >> +Bzy_MHz = TSC_delta*APERF_delta/MPERF_delta/measurement_interval > >> > >> Note that these calculations depend on TSC_delta, so they > >> are not reliable during intervals when TSC_MHz is not running at the base frequency. >
Len, The new commit message is better. Thanks for your help
diff --git a/tools/power/x86/turbostat/turbostat.8 b/tools/power/x86/turbostat/turbostat.8 index 8f08c3fd498d..1ba6340d3b3d 100644 --- a/tools/power/x86/turbostat/turbostat.8 +++ b/tools/power/x86/turbostat/turbostat.8 @@ -370,7 +370,7 @@ below the processor's base frequency. Busy% = MPERF_delta/TSC_delta -Bzy_MHz = TSC_delta/APERF_delta/MPERF_delta/measurement_interval +Bzy_MHz = TSC_delta*APERF_delta/MPERF_delta/measurement_interval Note that these calculations depend on TSC_delta, so they are not reliable during intervals when TSC_MHz is not running at the base frequency.