| Message ID | 1469049004-19069-1-git-send-email-ahs3@redhat.com (mailing list archive) |
|---|---|
| State | Accepted, archived |
| Delegated to: | Rafael Wysocki |
| Headers | show |
On 07/20/2016 03:10 PM, Al Stone wrote: > When CPPC is being used by ACPI on arm64, user space tools such as > cpupower report CPU frequency values from sysfs that are incorrect. > > What the driver was doing was reporting the values given by ACPI tables > in whatever scale was used to provide them. However, the ACPI spec > defines the CPPC values as unitless abstract numbers. Internal kernel > structures such as struct perf_cap, in contrast, expect these values > to be in KHz. When these struct values get reported via sysfs, the > user space tools also assume they are in KHz, causing them to report > incorrect values (for example, reporting a CPU frequency of 1MHz when > it should be 1.8GHz). > [snip...] Ping? If anyone has sent comments, I have not received them. Thanks in advance for the time and effort. I know merge window can be a very busy time.
[+ Ashwin's new email id..] On 20-07-16, 15:10, Al Stone wrote: > When CPPC is being used by ACPI on arm64, user space tools such as > cpupower report CPU frequency values from sysfs that are incorrect. > > What the driver was doing was reporting the values given by ACPI tables > in whatever scale was used to provide them. However, the ACPI spec > defines the CPPC values as unitless abstract numbers. Internal kernel > structures such as struct perf_cap, in contrast, expect these values > to be in KHz. When these struct values get reported via sysfs, the > user space tools also assume they are in KHz, causing them to report > incorrect values (for example, reporting a CPU frequency of 1MHz when > it should be 1.8GHz). > > The downside is that this approach has some assumptions: > > (1) It relies on SMBIOS3 being used, *and* that the Max Frequency > value for a processor is set to a non-zero value. > > (2) It assumes that all processors run at the same speed, or that > the CPPC values have all been scaled to reflect relative speed. > This patch retrieves the largest CPU Max Frequency from a type 4 DMI > record that it can find. This may not be an issue, however, as a > sampling of DMI data on x86 and arm64 indicates there is often only > one such record regardless. Since CPPC is relatively new, it is > unclear if the ACPI ASL will always be written to reflect any sort > of relative performance of processors of differing speeds. > > (3) It assumes that performance and frequency both scale linearly. > > For arm64 servers, this may be sufficient, but it does rely on > firmware values being set correctly. Hence, other approaches will > be considered in the future. > > This has been tested on three arm64 servers, with and without DMI, with > and without CPPC support. > > Changes for v5: > -- Move code to cpufreq/cppc_cpufreq.c from acpi/cppc_acpi.c to keep > frequency-related code together, and keep the CPPC abstract scale > in ACPI (Prashanth Prakash) > -- Fix the scaling to remove the incorrect assumption that frequency > was always a range from zero to max; as a practical matter, it is > not (Prasanth Prakash); this also allowed us to remove an over- > engineered function to do this math. > > Changes for v4: > -- Replaced magic constants with #defines (Rafael Wysocki) > -- Renamed cppc_unitless_to_khz() to cppc_to_khz() (Rafael Wysocki) > -- Replaced hidden initialization with a clearer form (Rafael Wysocki) > -- Instead of picking up the first Max Speed value from DMI, we will > now get the largest Max Speed; still an approximation, but slightly > less subject to error (Rafael Wysocki) > -- Kconfig for cppc_cpufreq now depends on DMI, instead of selecting > it, in order to make sure DMI is set up properly (Rafael Wysocki) > > Changes for v3: > -- Added clarifying commentary re short-term vs long-term fix (Alexey > Klimov) > -- Added range checking code to ensure proper arithmetic occurs, > especially no division by zero (Alexey Klimov) > > Changes for v2: > -- Corrected thinko: needed to have DEPENDS on DMI in Kconfig.arm, > not SELECT DMI (found by build daemon) > > Signed-off-by: Al Stone <ahs3@redhat.com> > Signed-off-by: Prashanth Prakash <pprakash@codeaurora.org> > --- > drivers/cpufreq/cppc_cpufreq.c | 53 ++++++++++++++++++++++++++++++++++++++---- > 1 file changed, 49 insertions(+), 4 deletions(-) > > diff --git a/drivers/cpufreq/cppc_cpufreq.c b/drivers/cpufreq/cppc_cpufreq.c > index 8882b8e..6debc18 100644 > --- a/drivers/cpufreq/cppc_cpufreq.c > +++ b/drivers/cpufreq/cppc_cpufreq.c > @@ -19,10 +19,19 @@ > #include <linux/delay.h> > #include <linux/cpu.h> > #include <linux/cpufreq.h> > +#include <linux/dmi.h> > #include <linux/vmalloc.h> > > +#include <asm/unaligned.h> > + > #include <acpi/cppc_acpi.h> > > +/* Minimum struct length needed for the DMI processor entry we want */ > +#define DMI_ENTRY_PROCESSOR_MIN_LENGTH 48 > + > +/* Offest in the DMI processor structure for the max frequency */ > +#define DMI_PROCESSOR_MAX_SPEED 0x14 > + > /* > * These structs contain information parsed from per CPU > * ACPI _CPC structures. > @@ -32,6 +41,39 @@ > */ > static struct cpudata **all_cpu_data; > > +/* Capture the max KHz from DMI */ > +static u64 cppc_dmi_max_khz; > + > +/* Callback function used to retrieve the max frequency from DMI */ > +static void cppc_find_dmi_mhz(const struct dmi_header *dm, void *private) > +{ > + const u8 *dmi_data = (const u8 *)dm; > + u16 *mhz = (u16 *)private; > + > + if (dm->type == DMI_ENTRY_PROCESSOR && > + dm->length >= DMI_ENTRY_PROCESSOR_MIN_LENGTH) { > + u16 val = (u16)get_unaligned((const u16 *) > + (dmi_data + DMI_PROCESSOR_MAX_SPEED)); > + *mhz = val > *mhz ? val : *mhz; > + } > +} > + > +/* Look up the max frequency in DMI */ > +static u64 cppc_get_dmi_max_khz(void) > +{ > + u16 mhz = 0; > + > + dmi_walk(cppc_find_dmi_mhz, &mhz); > + > + /* > + * Real stupid fallback value, just in case there is no > + * actual value set. > + */ > + mhz = mhz ? mhz : 1; > + > + return (1000 * mhz); > +} > + > static int cppc_cpufreq_set_target(struct cpufreq_policy *policy, > unsigned int target_freq, > unsigned int relation) > @@ -42,7 +84,7 @@ static int cppc_cpufreq_set_target(struct cpufreq_policy *policy, > > cpu = all_cpu_data[policy->cpu]; > > - cpu->perf_ctrls.desired_perf = target_freq; > + cpu->perf_ctrls.desired_perf = target_freq * policy->max / cppc_dmi_max_khz; > freqs.old = policy->cur; > freqs.new = target_freq; > > @@ -94,8 +136,10 @@ static int cppc_cpufreq_cpu_init(struct cpufreq_policy *policy) > return ret; > } > > - policy->min = cpu->perf_caps.lowest_perf; > - policy->max = cpu->perf_caps.highest_perf; > + cppc_dmi_max_khz = cppc_get_dmi_max_khz(); > + > + policy->min = cpu->perf_caps.lowest_perf * cppc_dmi_max_khz / cpu->perf_caps.highest_perf; > + policy->max = cppc_dmi_max_khz; > policy->cpuinfo.min_freq = policy->min; > policy->cpuinfo.max_freq = policy->max; > policy->shared_type = cpu->shared_type; > @@ -112,7 +156,8 @@ static int cppc_cpufreq_cpu_init(struct cpufreq_policy *policy) > cpu->cur_policy = policy; > > /* Set policy->cur to max now. The governors will adjust later. */ > - policy->cur = cpu->perf_ctrls.desired_perf = cpu->perf_caps.highest_perf; > + policy->cur = cppc_dmi_max_khz; > + cpu->perf_ctrls.desired_perf = cpu->perf_caps.highest_perf; > > ret = cppc_set_perf(cpu_num, &cpu->perf_ctrls); > if (ret) > -- > 2.7.4
On 08/01/2016 02:31 PM, Viresh Kumar wrote: > [+ Ashwin's new email id..] > > On 20-07-16, 15:10, Al Stone wrote: >> When CPPC is being used by ACPI on arm64, user space tools such as >> cpupower report CPU frequency values from sysfs that are incorrect. >> >> What the driver was doing was reporting the values given by ACPI tables >> in whatever scale was used to provide them. However, the ACPI spec >> defines the CPPC values as unitless abstract numbers. Internal kernel >> structures such as struct perf_cap, in contrast, expect these values >> to be in KHz. When these struct values get reported via sysfs, the >> user space tools also assume they are in KHz, causing them to report >> incorrect values (for example, reporting a CPU frequency of 1MHz when >> it should be 1.8GHz). >> >> The downside is that this approach has some assumptions: >> >> (1) It relies on SMBIOS3 being used, *and* that the Max Frequency >> value for a processor is set to a non-zero value. >> >> (2) It assumes that all processors run at the same speed, or that >> the CPPC values have all been scaled to reflect relative speed. >> This patch retrieves the largest CPU Max Frequency from a type 4 DMI >> record that it can find. This may not be an issue, however, as a >> sampling of DMI data on x86 and arm64 indicates there is often only >> one such record regardless. Since CPPC is relatively new, it is >> unclear if the ACPI ASL will always be written to reflect any sort >> of relative performance of processors of differing speeds. >> >> (3) It assumes that performance and frequency both scale linearly. >> >> For arm64 servers, this may be sufficient, but it does rely on >> firmware values being set correctly. Hence, other approaches will >> be considered in the future. >> >> This has been tested on three arm64 servers, with and without DMI, with >> and without CPPC support. >> >> Changes for v5: >> -- Move code to cpufreq/cppc_cpufreq.c from acpi/cppc_acpi.c to keep >> frequency-related code together, and keep the CPPC abstract scale >> in ACPI (Prashanth Prakash) >> -- Fix the scaling to remove the incorrect assumption that frequency >> was always a range from zero to max; as a practical matter, it is >> not (Prasanth Prakash); this also allowed us to remove an over- >> engineered function to do this math. >> >> Changes for v4: >> -- Replaced magic constants with #defines (Rafael Wysocki) >> -- Renamed cppc_unitless_to_khz() to cppc_to_khz() (Rafael Wysocki) >> -- Replaced hidden initialization with a clearer form (Rafael Wysocki) >> -- Instead of picking up the first Max Speed value from DMI, we will >> now get the largest Max Speed; still an approximation, but slightly >> less subject to error (Rafael Wysocki) >> -- Kconfig for cppc_cpufreq now depends on DMI, instead of selecting >> it, in order to make sure DMI is set up properly (Rafael Wysocki) >> >> Changes for v3: >> -- Added clarifying commentary re short-term vs long-term fix (Alexey >> Klimov) >> -- Added range checking code to ensure proper arithmetic occurs, >> especially no division by zero (Alexey Klimov) >> >> Changes for v2: >> -- Corrected thinko: needed to have DEPENDS on DMI in Kconfig.arm, >> not SELECT DMI (found by build daemon) >> >> Signed-off-by: Al Stone <ahs3@redhat.com> >> Signed-off-by: Prashanth Prakash <pprakash@codeaurora.org> >> --- >> drivers/cpufreq/cppc_cpufreq.c | 53 ++++++++++++++++++++++++++++++++++++++---- >> 1 file changed, 49 insertions(+), 4 deletions(-) >> >> diff --git a/drivers/cpufreq/cppc_cpufreq.c b/drivers/cpufreq/cppc_cpufreq.c >> index 8882b8e..6debc18 100644 >> --- a/drivers/cpufreq/cppc_cpufreq.c >> +++ b/drivers/cpufreq/cppc_cpufreq.c >> @@ -19,10 +19,19 @@ >> #include <linux/delay.h> >> #include <linux/cpu.h> >> #include <linux/cpufreq.h> >> +#include <linux/dmi.h> >> #include <linux/vmalloc.h> >> >> +#include <asm/unaligned.h> >> + >> #include <acpi/cppc_acpi.h> >> >> +/* Minimum struct length needed for the DMI processor entry we want */ >> +#define DMI_ENTRY_PROCESSOR_MIN_LENGTH 48 >> + >> +/* Offest in the DMI processor structure for the max frequency */ >> +#define DMI_PROCESSOR_MAX_SPEED 0x14 >> + >> /* >> * These structs contain information parsed from per CPU >> * ACPI _CPC structures. >> @@ -32,6 +41,39 @@ >> */ >> static struct cpudata **all_cpu_data; >> >> +/* Capture the max KHz from DMI */ >> +static u64 cppc_dmi_max_khz; >> + >> +/* Callback function used to retrieve the max frequency from DMI */ >> +static void cppc_find_dmi_mhz(const struct dmi_header *dm, void *private) >> +{ >> + const u8 *dmi_data = (const u8 *)dm; >> + u16 *mhz = (u16 *)private; >> + >> + if (dm->type == DMI_ENTRY_PROCESSOR && >> + dm->length >= DMI_ENTRY_PROCESSOR_MIN_LENGTH) { >> + u16 val = (u16)get_unaligned((const u16 *) >> + (dmi_data + DMI_PROCESSOR_MAX_SPEED)); >> + *mhz = val > *mhz ? val : *mhz; >> + } >> +} >> + >> +/* Look up the max frequency in DMI */ >> +static u64 cppc_get_dmi_max_khz(void) >> +{ >> + u16 mhz = 0; >> + >> + dmi_walk(cppc_find_dmi_mhz, &mhz); >> + >> + /* >> + * Real stupid fallback value, just in case there is no >> + * actual value set. >> + */ >> + mhz = mhz ? mhz : 1; >> + >> + return (1000 * mhz); >> +} >> + >> static int cppc_cpufreq_set_target(struct cpufreq_policy *policy, >> unsigned int target_freq, >> unsigned int relation) >> @@ -42,7 +84,7 @@ static int cppc_cpufreq_set_target(struct cpufreq_policy *policy, >> >> cpu = all_cpu_data[policy->cpu]; >> >> - cpu->perf_ctrls.desired_perf = target_freq; >> + cpu->perf_ctrls.desired_perf = target_freq * policy->max / cppc_dmi_max_khz; >> freqs.old = policy->cur; >> freqs.new = target_freq; >> >> @@ -94,8 +136,10 @@ static int cppc_cpufreq_cpu_init(struct cpufreq_policy *policy) >> return ret; >> } >> >> - policy->min = cpu->perf_caps.lowest_perf; >> - policy->max = cpu->perf_caps.highest_perf; >> + cppc_dmi_max_khz = cppc_get_dmi_max_khz(); >> + >> + policy->min = cpu->perf_caps.lowest_perf * cppc_dmi_max_khz / cpu->perf_caps.highest_perf; >> + policy->max = cppc_dmi_max_khz; >> policy->cpuinfo.min_freq = policy->min; >> policy->cpuinfo.max_freq = policy->max; >> policy->shared_type = cpu->shared_type; >> @@ -112,7 +156,8 @@ static int cppc_cpufreq_cpu_init(struct cpufreq_policy *policy) >> cpu->cur_policy = policy; >> >> /* Set policy->cur to max now. The governors will adjust later. */ >> - policy->cur = cpu->perf_ctrls.desired_perf = cpu->perf_caps.highest_perf; >> + policy->cur = cppc_dmi_max_khz; >> + cpu->perf_ctrls.desired_perf = cpu->perf_caps.highest_perf; >> >> ret = cppc_set_perf(cpu_num, &cpu->perf_ctrls); >> if (ret) >> -- >> 2.7.4 > Another gentle ping -- any comments? Can this get pulled in now? Thanks.
Maybe a top-post will get attention.... Yet another ping; this was first submitted on 20 July, and has received no comments. It has now been a month and other architectures are starting to use CPPC so they will run into the same errors that this fixes. Can I get an ACK, NAK, or further instructions, please? Also adding Rafael on the ACPI side, just in case, since he's also reviewing the Intel patches on the linux-acpi mailing list that are adding CPPC usage. On 08/11/2016 12:15 PM, Al Stone wrote: > On 08/01/2016 02:31 PM, Viresh Kumar wrote: >> [+ Ashwin's new email id..] >> >> On 20-07-16, 15:10, Al Stone wrote: >>> When CPPC is being used by ACPI on arm64, user space tools such as >>> cpupower report CPU frequency values from sysfs that are incorrect. >>> >>> What the driver was doing was reporting the values given by ACPI tables >>> in whatever scale was used to provide them. However, the ACPI spec >>> defines the CPPC values as unitless abstract numbers. Internal kernel >>> structures such as struct perf_cap, in contrast, expect these values >>> to be in KHz. When these struct values get reported via sysfs, the >>> user space tools also assume they are in KHz, causing them to report >>> incorrect values (for example, reporting a CPU frequency of 1MHz when >>> it should be 1.8GHz). >>> >>> The downside is that this approach has some assumptions: >>> >>> (1) It relies on SMBIOS3 being used, *and* that the Max Frequency >>> value for a processor is set to a non-zero value. >>> >>> (2) It assumes that all processors run at the same speed, or that >>> the CPPC values have all been scaled to reflect relative speed. >>> This patch retrieves the largest CPU Max Frequency from a type 4 DMI >>> record that it can find. This may not be an issue, however, as a >>> sampling of DMI data on x86 and arm64 indicates there is often only >>> one such record regardless. Since CPPC is relatively new, it is >>> unclear if the ACPI ASL will always be written to reflect any sort >>> of relative performance of processors of differing speeds. >>> >>> (3) It assumes that performance and frequency both scale linearly. >>> >>> For arm64 servers, this may be sufficient, but it does rely on >>> firmware values being set correctly. Hence, other approaches will >>> be considered in the future. >>> >>> This has been tested on three arm64 servers, with and without DMI, with >>> and without CPPC support. >>> >>> Changes for v5: >>> -- Move code to cpufreq/cppc_cpufreq.c from acpi/cppc_acpi.c to keep >>> frequency-related code together, and keep the CPPC abstract scale >>> in ACPI (Prashanth Prakash) >>> -- Fix the scaling to remove the incorrect assumption that frequency >>> was always a range from zero to max; as a practical matter, it is >>> not (Prasanth Prakash); this also allowed us to remove an over- >>> engineered function to do this math. >>> >>> Changes for v4: >>> -- Replaced magic constants with #defines (Rafael Wysocki) >>> -- Renamed cppc_unitless_to_khz() to cppc_to_khz() (Rafael Wysocki) >>> -- Replaced hidden initialization with a clearer form (Rafael Wysocki) >>> -- Instead of picking up the first Max Speed value from DMI, we will >>> now get the largest Max Speed; still an approximation, but slightly >>> less subject to error (Rafael Wysocki) >>> -- Kconfig for cppc_cpufreq now depends on DMI, instead of selecting >>> it, in order to make sure DMI is set up properly (Rafael Wysocki) >>> >>> Changes for v3: >>> -- Added clarifying commentary re short-term vs long-term fix (Alexey >>> Klimov) >>> -- Added range checking code to ensure proper arithmetic occurs, >>> especially no division by zero (Alexey Klimov) >>> >>> Changes for v2: >>> -- Corrected thinko: needed to have DEPENDS on DMI in Kconfig.arm, >>> not SELECT DMI (found by build daemon) >>> >>> Signed-off-by: Al Stone <ahs3@redhat.com> >>> Signed-off-by: Prashanth Prakash <pprakash@codeaurora.org> >>> --- >>> drivers/cpufreq/cppc_cpufreq.c | 53 ++++++++++++++++++++++++++++++++++++++---- >>> 1 file changed, 49 insertions(+), 4 deletions(-) >>> >>> diff --git a/drivers/cpufreq/cppc_cpufreq.c b/drivers/cpufreq/cppc_cpufreq.c >>> index 8882b8e..6debc18 100644 >>> --- a/drivers/cpufreq/cppc_cpufreq.c >>> +++ b/drivers/cpufreq/cppc_cpufreq.c >>> @@ -19,10 +19,19 @@ >>> #include <linux/delay.h> >>> #include <linux/cpu.h> >>> #include <linux/cpufreq.h> >>> +#include <linux/dmi.h> >>> #include <linux/vmalloc.h> >>> >>> +#include <asm/unaligned.h> >>> + >>> #include <acpi/cppc_acpi.h> >>> >>> +/* Minimum struct length needed for the DMI processor entry we want */ >>> +#define DMI_ENTRY_PROCESSOR_MIN_LENGTH 48 >>> + >>> +/* Offest in the DMI processor structure for the max frequency */ >>> +#define DMI_PROCESSOR_MAX_SPEED 0x14 >>> + >>> /* >>> * These structs contain information parsed from per CPU >>> * ACPI _CPC structures. >>> @@ -32,6 +41,39 @@ >>> */ >>> static struct cpudata **all_cpu_data; >>> >>> +/* Capture the max KHz from DMI */ >>> +static u64 cppc_dmi_max_khz; >>> + >>> +/* Callback function used to retrieve the max frequency from DMI */ >>> +static void cppc_find_dmi_mhz(const struct dmi_header *dm, void *private) >>> +{ >>> + const u8 *dmi_data = (const u8 *)dm; >>> + u16 *mhz = (u16 *)private; >>> + >>> + if (dm->type == DMI_ENTRY_PROCESSOR && >>> + dm->length >= DMI_ENTRY_PROCESSOR_MIN_LENGTH) { >>> + u16 val = (u16)get_unaligned((const u16 *) >>> + (dmi_data + DMI_PROCESSOR_MAX_SPEED)); >>> + *mhz = val > *mhz ? val : *mhz; >>> + } >>> +} >>> + >>> +/* Look up the max frequency in DMI */ >>> +static u64 cppc_get_dmi_max_khz(void) >>> +{ >>> + u16 mhz = 0; >>> + >>> + dmi_walk(cppc_find_dmi_mhz, &mhz); >>> + >>> + /* >>> + * Real stupid fallback value, just in case there is no >>> + * actual value set. >>> + */ >>> + mhz = mhz ? mhz : 1; >>> + >>> + return (1000 * mhz); >>> +} >>> + >>> static int cppc_cpufreq_set_target(struct cpufreq_policy *policy, >>> unsigned int target_freq, >>> unsigned int relation) >>> @@ -42,7 +84,7 @@ static int cppc_cpufreq_set_target(struct cpufreq_policy *policy, >>> >>> cpu = all_cpu_data[policy->cpu]; >>> >>> - cpu->perf_ctrls.desired_perf = target_freq; >>> + cpu->perf_ctrls.desired_perf = target_freq * policy->max / cppc_dmi_max_khz; >>> freqs.old = policy->cur; >>> freqs.new = target_freq; >>> >>> @@ -94,8 +136,10 @@ static int cppc_cpufreq_cpu_init(struct cpufreq_policy *policy) >>> return ret; >>> } >>> >>> - policy->min = cpu->perf_caps.lowest_perf; >>> - policy->max = cpu->perf_caps.highest_perf; >>> + cppc_dmi_max_khz = cppc_get_dmi_max_khz(); >>> + >>> + policy->min = cpu->perf_caps.lowest_perf * cppc_dmi_max_khz / cpu->perf_caps.highest_perf; >>> + policy->max = cppc_dmi_max_khz; >>> policy->cpuinfo.min_freq = policy->min; >>> policy->cpuinfo.max_freq = policy->max; >>> policy->shared_type = cpu->shared_type; >>> @@ -112,7 +156,8 @@ static int cppc_cpufreq_cpu_init(struct cpufreq_policy *policy) >>> cpu->cur_policy = policy; >>> >>> /* Set policy->cur to max now. The governors will adjust later. */ >>> - policy->cur = cpu->perf_ctrls.desired_perf = cpu->perf_caps.highest_perf; >>> + policy->cur = cppc_dmi_max_khz; >>> + cpu->perf_ctrls.desired_perf = cpu->perf_caps.highest_perf; >>> >>> ret = cppc_set_perf(cpu_num, &cpu->perf_ctrls); >>> if (ret) >>> -- >>> 2.7.4 >> > > Another gentle ping -- any comments? Can this get pulled in now? > > Thanks. >
Hi Al, On Mon, Aug 22, 2016 at 10:16 AM, Al Stone <ahs3@redhat.com> wrote: > Maybe a top-post will get attention.... > > Yet another ping; this was first submitted on 20 July, and has received > no comments. It has now been a month and other architectures are starting > to use CPPC so they will run into the same errors that this fixes. Can > I get an ACK, NAK, or further instructions, please? > Apologies for the delay. I thought this patch was merged already. > Also adding Rafael on the ACPI side, just in case, since he's also reviewing > the Intel patches on the linux-acpi mailing list that are adding CPPC usage. > > On 08/11/2016 12:15 PM, Al Stone wrote: >> On 08/01/2016 02:31 PM, Viresh Kumar wrote: >>> [+ Ashwin's new email id..] >>> >>> On 20-07-16, 15:10, Al Stone wrote: >>>> When CPPC is being used by ACPI on arm64, user space tools such as >>>> cpupower report CPU frequency values from sysfs that are incorrect. >>>> >>>> What the driver was doing was reporting the values given by ACPI tables >>>> in whatever scale was used to provide them. However, the ACPI spec >>>> defines the CPPC values as unitless abstract numbers. Internal kernel >>>> structures such as struct perf_cap, in contrast, expect these values >>>> to be in KHz. When these struct values get reported via sysfs, the >>>> user space tools also assume they are in KHz, causing them to report >>>> incorrect values (for example, reporting a CPU frequency of 1MHz when >>>> it should be 1.8GHz). >>>> >>>> The downside is that this approach has some assumptions: >>>> >>>> (1) It relies on SMBIOS3 being used, *and* that the Max Frequency >>>> value for a processor is set to a non-zero value. >>>> >>>> (2) It assumes that all processors run at the same speed, or that >>>> the CPPC values have all been scaled to reflect relative speed. >>>> This patch retrieves the largest CPU Max Frequency from a type 4 DMI >>>> record that it can find. This may not be an issue, however, as a >>>> sampling of DMI data on x86 and arm64 indicates there is often only >>>> one such record regardless. Since CPPC is relatively new, it is >>>> unclear if the ACPI ASL will always be written to reflect any sort >>>> of relative performance of processors of differing speeds. >>>> >>>> (3) It assumes that performance and frequency both scale linearly. >>>> >>>> For arm64 servers, this may be sufficient, but it does rely on >>>> firmware values being set correctly. Hence, other approaches will >>>> be considered in the future. >>>> >>>> This has been tested on three arm64 servers, with and without DMI, with >>>> and without CPPC support. >>>> >>>> Changes for v5: >>>> -- Move code to cpufreq/cppc_cpufreq.c from acpi/cppc_acpi.c to keep >>>> frequency-related code together, and keep the CPPC abstract scale >>>> in ACPI (Prashanth Prakash) >>>> -- Fix the scaling to remove the incorrect assumption that frequency >>>> was always a range from zero to max; as a practical matter, it is >>>> not (Prasanth Prakash); this also allowed us to remove an over- >>>> engineered function to do this math. >>>> This addresses my previous feedback. So FWIW, Acked-by: Ashwin Chaugule <ashwinch@google.com> Cheers, Ashwin. -- To unsubscribe from this list: send the line "unsubscribe linux-pm" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On 08/22/2016 11:45 AM, Ashwin Chaugule wrote: > Hi Al, > > On Mon, Aug 22, 2016 at 10:16 AM, Al Stone <ahs3@redhat.com> wrote: >> Maybe a top-post will get attention.... >> >> Yet another ping; this was first submitted on 20 July, and has received >> no comments. It has now been a month and other architectures are starting >> to use CPPC so they will run into the same errors that this fixes. Can >> I get an ACK, NAK, or further instructions, please? >> > > Apologies for the delay. I thought this patch was merged already. I've looked in linux-next and linux-pm; I could have missed it, but I didn't see it...my bad, if I did. >> Also adding Rafael on the ACPI side, just in case, since he's also reviewing >> the Intel patches on the linux-acpi mailing list that are adding CPPC usage. >> >> On 08/11/2016 12:15 PM, Al Stone wrote: >>> On 08/01/2016 02:31 PM, Viresh Kumar wrote: >>>> [+ Ashwin's new email id..] >>>> >>>> On 20-07-16, 15:10, Al Stone wrote: >>>>> When CPPC is being used by ACPI on arm64, user space tools such as >>>>> cpupower report CPU frequency values from sysfs that are incorrect. >>>>> >>>>> What the driver was doing was reporting the values given by ACPI tables >>>>> in whatever scale was used to provide them. However, the ACPI spec >>>>> defines the CPPC values as unitless abstract numbers. Internal kernel >>>>> structures such as struct perf_cap, in contrast, expect these values >>>>> to be in KHz. When these struct values get reported via sysfs, the >>>>> user space tools also assume they are in KHz, causing them to report >>>>> incorrect values (for example, reporting a CPU frequency of 1MHz when >>>>> it should be 1.8GHz). >>>>> >>>>> The downside is that this approach has some assumptions: >>>>> >>>>> (1) It relies on SMBIOS3 being used, *and* that the Max Frequency >>>>> value for a processor is set to a non-zero value. >>>>> >>>>> (2) It assumes that all processors run at the same speed, or that >>>>> the CPPC values have all been scaled to reflect relative speed. >>>>> This patch retrieves the largest CPU Max Frequency from a type 4 DMI >>>>> record that it can find. This may not be an issue, however, as a >>>>> sampling of DMI data on x86 and arm64 indicates there is often only >>>>> one such record regardless. Since CPPC is relatively new, it is >>>>> unclear if the ACPI ASL will always be written to reflect any sort >>>>> of relative performance of processors of differing speeds. >>>>> >>>>> (3) It assumes that performance and frequency both scale linearly. >>>>> >>>>> For arm64 servers, this may be sufficient, but it does rely on >>>>> firmware values being set correctly. Hence, other approaches will >>>>> be considered in the future. >>>>> >>>>> This has been tested on three arm64 servers, with and without DMI, with >>>>> and without CPPC support. >>>>> >>>>> Changes for v5: >>>>> -- Move code to cpufreq/cppc_cpufreq.c from acpi/cppc_acpi.c to keep >>>>> frequency-related code together, and keep the CPPC abstract scale >>>>> in ACPI (Prashanth Prakash) >>>>> -- Fix the scaling to remove the incorrect assumption that frequency >>>>> was always a range from zero to max; as a practical matter, it is >>>>> not (Prasanth Prakash); this also allowed us to remove an over- >>>>> engineered function to do this math. >>>>> > > This addresses my previous feedback. So FWIW, Acked-by: Ashwin > Chaugule <ashwinch@google.com> > > Cheers, > Ashwin. > Thanks, Ashwin.
T24gTW9uLCAyMDE2LTA4LTIyIGF0IDEyOjEyIC0wNjAwLCBBbCBTdG9uZSB3cm90ZToNCj4gT24g MDgvMjIvMjAxNiAxMTo0NSBBTSwgQXNod2luIENoYXVndWxlIHdyb3RlOg0KPiA+IA0KPiA+IEhp IEFsLA0KPiA+IA0KPiA+IE9uIE1vbiwgQXVnIDIyLCAyMDE2IGF0IDEwOjE2IEFNLCBBbCBTdG9u ZSA8YWhzM0ByZWRoYXQuY29tPiB3cm90ZToNCj4gPiA+IA0KPiA+ID4gTWF5YmUgYSB0b3AtcG9z dCB3aWxsIGdldCBhdHRlbnRpb24uLi4uDQo+ID4gPiANCj4gPiA+IFlldCBhbm90aGVyIHBpbmc7 IHRoaXMgd2FzIGZpcnN0IHN1Ym1pdHRlZCBvbiAyMCBKdWx5LCBhbmQgaGFzDQo+ID4gPiByZWNl aXZlZA0KPiA+ID4gbm8gY29tbWVudHMuwqDCoEl0IGhhcyBub3cgYmVlbiBhIG1vbnRoIGFuZCBv dGhlciBhcmNoaXRlY3R1cmVzIGFyZQ0KPiA+ID4gc3RhcnRpbmcNCj4gPiA+IHRvIHVzZSBDUFBD IHNvIHRoZXkgd2lsbCBydW4gaW50byB0aGUgc2FtZSBlcnJvcnMgdGhhdCB0aGlzDQo+ID4gPiBm aXhlcy7CoMKgQ2FuDQo+ID4gPiBJIGdldCBhbiBBQ0ssIE5BSywgb3IgZnVydGhlciBpbnN0cnVj dGlvbnMsIHBsZWFzZT8NCj4gPiA+IA0KPiA+IA0KPiA+IEFwb2xvZ2llcyBmb3IgdGhlIGRlbGF5 LiBJIHRob3VnaHQgdGhpcyBwYXRjaCB3YXMgbWVyZ2VkIGFscmVhZHkuDQo+IA0KPiBJJ3ZlIGxv b2tlZCBpbiBsaW51eC1uZXh0IGFuZCBsaW51eC1wbTsgSSBjb3VsZCBoYXZlIG1pc3NlZCBpdCwg YnV0IEkNCj4gZGlkbid0DQo+IHNlZSBpdC4uLm15IGJhZCwgaWYgSSBkaWQuDQo+IA0KPiA+IA0K PiA+ID4gDQo+ID4gPiBBbHNvIGFkZGluZyBSYWZhZWwgb24gdGhlIEFDUEkgc2lkZSwganVzdCBp biBjYXNlLCBzaW5jZSBoZSdzDQo+ID4gPiBhbHNvIHJldmlld2luZw0KPiA+ID4gdGhlIEludGVs IHBhdGNoZXMgb24gdGhlIGxpbnV4LWFjcGkgbWFpbGluZyBsaXN0IHRoYXQgYXJlIGFkZGluZw0K PiA+ID4gQ1BQQyB1c2FnZS4NCj4gPiA+IA0KPiA+ID4gT24gMDgvMTEvMjAxNiAxMjoxNSBQTSwg QWwgU3RvbmUgd3JvdGU6DQo+ID4gPiA+IA0KPiA+ID4gPiBPbiAwOC8wMS8yMDE2IDAyOjMxIFBN LCBWaXJlc2ggS3VtYXIgd3JvdGU6DQo+ID4gPiA+ID4gDQo+ID4gPiA+ID4gWysgQXNod2luJ3Mg bmV3IGVtYWlsIGlkLi5dDQo+ID4gPiA+ID4gDQo+ID4gPiA+ID4gT24gMjAtMDctMTYsIDE1OjEw LCBBbCBTdG9uZSB3cm90ZToNCj4gPiA+ID4gPiA+IA0KPiA+ID4gPiA+ID4gV2hlbiBDUFBDIGlz IGJlaW5nIHVzZWQgYnkgQUNQSSBvbiBhcm02NCwgdXNlciBzcGFjZSB0b29scw0KPiA+ID4gPiA+ ID4gc3VjaCBhcw0KPiA+ID4gPiA+ID4gY3B1cG93ZXIgcmVwb3J0IENQVSBmcmVxdWVuY3kgdmFs dWVzIGZyb20gc3lzZnMgdGhhdCBhcmUNCj4gPiA+ID4gPiA+IGluY29ycmVjdC4NCj4gPiA+ID4g PiA+IA0KPiA+ID4gPiA+ID4gV2hhdCB0aGUgZHJpdmVyIHdhcyBkb2luZyB3YXMgcmVwb3J0aW5n IHRoZSB2YWx1ZXMgZ2l2ZW4gYnkNCj4gPiA+ID4gPiA+IEFDUEkgdGFibGVzDQo+ID4gPiA+ID4g PiBpbiB3aGF0ZXZlciBzY2FsZSB3YXMgdXNlZCB0byBwcm92aWRlIHRoZW0uwqDCoEhvd2V2ZXIs IHRoZQ0KPiA+ID4gPiA+ID4gQUNQSSBzcGVjDQo+ID4gPiA+ID4gPiBkZWZpbmVzIHRoZSBDUFBD IHZhbHVlcyBhcyB1bml0bGVzcyBhYnN0cmFjdA0KPiA+ID4gPiA+ID4gbnVtYmVycy7CoMKgSW50 ZXJuYWwga2VybmVsDQo+ID4gPiA+ID4gPiBzdHJ1Y3R1cmVzIHN1Y2ggYXMgc3RydWN0IHBlcmZf Y2FwLCBpbiBjb250cmFzdCwgZXhwZWN0DQo+ID4gPiA+ID4gPiB0aGVzZSB2YWx1ZXMNCj4gPiA+ ID4gPiA+IHRvIGJlIGluIEtIei7CoMKgV2hlbiB0aGVzZSBzdHJ1Y3QgdmFsdWVzIGdldCByZXBv cnRlZCB2aWENCj4gPiA+ID4gPiA+IHN5c2ZzLCB0aGUNCj4gPiA+ID4gPiA+IHVzZXIgc3BhY2Ug dG9vbHMgYWxzbyBhc3N1bWUgdGhleSBhcmUgaW4gS0h6LCBjYXVzaW5nIHRoZW0NCj4gPiA+ID4g PiA+IHRvIHJlcG9ydA0KPiA+ID4gPiA+ID4gaW5jb3JyZWN0IHZhbHVlcyAoZm9yIGV4YW1wbGUs IHJlcG9ydGluZyBhIENQVSBmcmVxdWVuY3kgb2YNCj4gPiA+ID4gPiA+IDFNSHogd2hlbg0KPiA+ ID4gPiA+ID4gaXQgc2hvdWxkIGJlIDEuOEdIeikuDQo+ID4gPiA+ID4gPiANCj4gPiA+ID4gPiA+ IFRoZSBkb3duc2lkZSBpcyB0aGF0IHRoaXMgYXBwcm9hY2ggaGFzIHNvbWUgYXNzdW1wdGlvbnM6 DQo+ID4gPiA+ID4gPiANCj4gPiA+ID4gPiA+IMKgwqDCoCgxKSBJdCByZWxpZXMgb24gU01CSU9T MyBiZWluZyB1c2VkLCAqYW5kKiB0aGF0IHRoZSBNYXgNCj4gPiA+ID4gPiA+IEZyZXF1ZW5jeQ0K PiA+ID4gPiA+ID4gwqDCoMKgdmFsdWUgZm9yIGEgcHJvY2Vzc29yIGlzIHNldCB0byBhIG5vbi16 ZXJvIHZhbHVlLg0KPiA+ID4gPiA+ID4gDQo+ID4gPiA+ID4gPiDCoMKgwqAoMikgSXQgYXNzdW1l cyB0aGF0IGFsbCBwcm9jZXNzb3JzIHJ1biBhdCB0aGUgc2FtZQ0KPiA+ID4gPiA+ID4gc3BlZWQs IG9yIHRoYXQNCj4gPiA+ID4gPiA+IMKgwqDCoHRoZSBDUFBDIHZhbHVlcyBoYXZlIGFsbCBiZWVu IHNjYWxlZCB0byByZWZsZWN0IHJlbGF0aXZlDQo+ID4gPiA+ID4gPiBzcGVlZC4NCj4gPiA+ID4g PiA+IMKgwqDCoFRoaXMgcGF0Y2ggcmV0cmlldmVzIHRoZSBsYXJnZXN0IENQVSBNYXggRnJlcXVl bmN5IGZyb20NCj4gPiA+ID4gPiA+IGEgdHlwZSA0IERNSQ0KPiA+ID4gPiA+ID4gwqDCoMKgcmVj b3JkIHRoYXQgaXQgY2FuIGZpbmQuwqDCoFRoaXMgbWF5IG5vdCBiZSBhbiBpc3N1ZSwNCj4gPiA+ ID4gPiA+IGhvd2V2ZXIsIGFzIGENCj4gPiA+ID4gPiA+IMKgwqDCoHNhbXBsaW5nIG9mIERNSSBk YXRhIG9uIHg4NiBhbmQgYXJtNjQgaW5kaWNhdGVzIHRoZXJlIGlzDQo+ID4gPiA+ID4gPiBvZnRl biBvbmx5DQo+ID4gPiA+ID4gPiDCoMKgwqBvbmUgc3VjaCByZWNvcmQgcmVnYXJkbGVzcy7CoMKg U2luY2UgQ1BQQyBpcyByZWxhdGl2ZWx5DQo+ID4gPiA+ID4gPiBuZXcsIGl0IGlzDQo+ID4gPiA+ ID4gPiDCoMKgwqB1bmNsZWFyIGlmIHRoZSBBQ1BJIEFTTCB3aWxsIGFsd2F5cyBiZSB3cml0dGVu IHRvDQo+ID4gPiA+ID4gPiByZWZsZWN0IGFueSBzb3J0DQo+ID4gPiA+ID4gPiDCoMKgwqBvZiBy ZWxhdGl2ZSBwZXJmb3JtYW5jZSBvZiBwcm9jZXNzb3JzIG9mIGRpZmZlcmluZw0KPiA+ID4gPiA+ ID4gc3BlZWRzLg0KPiA+ID4gPiA+ID4gDQo+ID4gPiA+ID4gPiDCoMKgwqAoMykgSXQgYXNzdW1l cyB0aGF0IHBlcmZvcm1hbmNlIGFuZCBmcmVxdWVuY3kgYm90aCBzY2FsZQ0KPiA+ID4gPiA+ID4g bGluZWFybHkuDQo+ID4gPiA+ID4gPiANCj4gPiA+ID4gPiA+IEZvciBhcm02NCBzZXJ2ZXJzLCB0 aGlzIG1heSBiZSBzdWZmaWNpZW50LCBidXQgaXQgZG9lcyByZWx5DQo+ID4gPiA+ID4gPiBvbg0K PiA+ID4gPiA+ID4gZmlybXdhcmUgdmFsdWVzIGJlaW5nIHNldCBjb3JyZWN0bHkuwqDCoEhlbmNl LCBvdGhlcg0KPiA+ID4gPiA+ID4gYXBwcm9hY2hlcyB3aWxsDQo+ID4gPiA+ID4gPiBiZSBjb25z aWRlcmVkIGluIHRoZSBmdXR1cmUuDQo+ID4gPiA+ID4gPiANCj4gPiA+ID4gPiA+IFRoaXMgaGFz IGJlZW4gdGVzdGVkIG9uIHRocmVlIGFybTY0IHNlcnZlcnMsIHdpdGggYW5kDQo+ID4gPiA+ID4g PiB3aXRob3V0IERNSSwgd2l0aA0KPiA+ID4gPiA+ID4gYW5kIHdpdGhvdXQgQ1BQQyBzdXBwb3J0 Lg0KPiA+ID4gPiA+ID4gDQo+ID4gPiA+ID4gPiBDaGFuZ2VzIGZvciB2NToNCj4gPiA+ID4gPiA+ IMKgwqDCoMKgLS0gTW92ZSBjb2RlIHRvIGNwdWZyZXEvY3BwY19jcHVmcmVxLmMgZnJvbQ0KPiA+ ID4gPiA+ID4gYWNwaS9jcHBjX2FjcGkuYyB0byBrZWVwDQo+ID4gPiA+ID4gPiDCoMKgwqDCoMKg wqDCoGZyZXF1ZW5jeS1yZWxhdGVkIGNvZGUgdG9nZXRoZXIsIGFuZCBrZWVwIHRoZSBDUFBDDQo+ ID4gPiA+ID4gPiBhYnN0cmFjdCBzY2FsZQ0KPiA+ID4gPiA+ID4gwqDCoMKgwqDCoMKgwqBpbiBB Q1BJIChQcmFzaGFudGggUHJha2FzaCkNCj4gPiA+ID4gPiA+IMKgwqDCoMKgLS0gRml4IHRoZSBz Y2FsaW5nIHRvIHJlbW92ZSB0aGUgaW5jb3JyZWN0IGFzc3VtcHRpb24NCj4gPiA+ID4gPiA+IHRo YXQgZnJlcXVlbmN5DQo+ID4gPiA+ID4gPiDCoMKgwqDCoMKgwqDCoHdhcyBhbHdheXMgYSByYW5n ZSBmcm9tIHplcm8gdG8gbWF4OyBhcyBhIHByYWN0aWNhbA0KPiA+ID4gPiA+ID4gbWF0dGVyLCBp dCBpcw0KPiA+ID4gPiA+ID4gwqDCoMKgwqDCoMKgwqBub3QgKFByYXNhbnRoIFByYWthc2gpOyB0 aGlzIGFsc28gYWxsb3dlZCB1cyB0bw0KPiA+ID4gPiA+ID4gcmVtb3ZlIGFuIG92ZXItDQo+ID4g PiA+ID4gPiDCoMKgwqDCoMKgwqDCoGVuZ2luZWVyZWQgZnVuY3Rpb24gdG8gZG8gdGhpcyBtYXRo Lg0KPiA+ID4gPiA+ID4gDQoNCkluIHg4NiB3aGVuIENQUEMgaXMgdXNlZCwgdGhlIHVuaXQgaXMg cmVhbGx5IHVuaXQtbGVzcyBpbiBDUFBDIHRhYmxlcy4NClRoaXMgbWVhbnMgdGhhdMKgY3B1LT5w ZXJmX2NhcHMuaGlnaGVzdF9wZXJmIGNhbiBiZSBqdXN0IDB4ZmYsIGluc3RlYWQNCm9mIHNvbWUg c2NhbGVkIGNwcGMgbWF4IHBlcmZvcm1hbmNlIGNvcnJlc3BvbmRpbmcgdG8gbWF4IE1IWiB0aGUN CnByb2Nlc3NvciBjYW4gc3VwcG9ydC4gVGhpcyBhbGxvd3MgdGhlIHByb2Nlc3NvciB0byBjYXAg YXQgbWF4IHdoaWNoIGl0DQpjYW4gZGVsaXZlci4NCklzIHRoaXMgY2FzZSBub3QgcG9zc2libGUg Zm9yIEFSTSBTb0NzPw0KDQpUaGFua3MsDQpTcmluaXZhcw0K -- To unsubscribe from this list: send the line "unsubscribe linux-pm" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On 08/22/2016 10:31 PM, Pandruvada, Srinivas wrote: > On Mon, 2016-08-22 at 12:12 -0600, Al Stone wrote: >> On 08/22/2016 11:45 AM, Ashwin Chaugule wrote: >>> >>> Hi Al, >>> >>> On Mon, Aug 22, 2016 at 10:16 AM, Al Stone <ahs3@redhat.com> wrote: >>>> >>>> Maybe a top-post will get attention.... >>>> >>>> Yet another ping; this was first submitted on 20 July, and has >>>> received >>>> no comments. It has now been a month and other architectures are >>>> starting >>>> to use CPPC so they will run into the same errors that this >>>> fixes. Can >>>> I get an ACK, NAK, or further instructions, please? >>>> >>> >>> Apologies for the delay. I thought this patch was merged already. >> >> I've looked in linux-next and linux-pm; I could have missed it, but I >> didn't >> see it...my bad, if I did. >> >>> >>>> >>>> Also adding Rafael on the ACPI side, just in case, since he's >>>> also reviewing >>>> the Intel patches on the linux-acpi mailing list that are adding >>>> CPPC usage. >>>> >>>> On 08/11/2016 12:15 PM, Al Stone wrote: >>>>> >>>>> On 08/01/2016 02:31 PM, Viresh Kumar wrote: >>>>>> >>>>>> [+ Ashwin's new email id..] >>>>>> >>>>>> On 20-07-16, 15:10, Al Stone wrote: >>>>>>> >>>>>>> When CPPC is being used by ACPI on arm64, user space tools >>>>>>> such as >>>>>>> cpupower report CPU frequency values from sysfs that are >>>>>>> incorrect. >>>>>>> >>>>>>> What the driver was doing was reporting the values given by >>>>>>> ACPI tables >>>>>>> in whatever scale was used to provide them. However, the >>>>>>> ACPI spec >>>>>>> defines the CPPC values as unitless abstract >>>>>>> numbers. Internal kernel >>>>>>> structures such as struct perf_cap, in contrast, expect >>>>>>> these values >>>>>>> to be in KHz. When these struct values get reported via >>>>>>> sysfs, the >>>>>>> user space tools also assume they are in KHz, causing them >>>>>>> to report >>>>>>> incorrect values (for example, reporting a CPU frequency of >>>>>>> 1MHz when >>>>>>> it should be 1.8GHz). >>>>>>> >>>>>>> The downside is that this approach has some assumptions: >>>>>>> >>>>>>> (1) It relies on SMBIOS3 being used, *and* that the Max >>>>>>> Frequency >>>>>>> value for a processor is set to a non-zero value. >>>>>>> >>>>>>> (2) It assumes that all processors run at the same >>>>>>> speed, or that >>>>>>> the CPPC values have all been scaled to reflect relative >>>>>>> speed. >>>>>>> This patch retrieves the largest CPU Max Frequency from >>>>>>> a type 4 DMI >>>>>>> record that it can find. This may not be an issue, >>>>>>> however, as a >>>>>>> sampling of DMI data on x86 and arm64 indicates there is >>>>>>> often only >>>>>>> one such record regardless. Since CPPC is relatively >>>>>>> new, it is >>>>>>> unclear if the ACPI ASL will always be written to >>>>>>> reflect any sort >>>>>>> of relative performance of processors of differing >>>>>>> speeds. >>>>>>> >>>>>>> (3) It assumes that performance and frequency both scale >>>>>>> linearly. >>>>>>> >>>>>>> For arm64 servers, this may be sufficient, but it does rely >>>>>>> on >>>>>>> firmware values being set correctly. Hence, other >>>>>>> approaches will >>>>>>> be considered in the future. >>>>>>> >>>>>>> This has been tested on three arm64 servers, with and >>>>>>> without DMI, with >>>>>>> and without CPPC support. >>>>>>> >>>>>>> Changes for v5: >>>>>>> -- Move code to cpufreq/cppc_cpufreq.c from >>>>>>> acpi/cppc_acpi.c to keep >>>>>>> frequency-related code together, and keep the CPPC >>>>>>> abstract scale >>>>>>> in ACPI (Prashanth Prakash) >>>>>>> -- Fix the scaling to remove the incorrect assumption >>>>>>> that frequency >>>>>>> was always a range from zero to max; as a practical >>>>>>> matter, it is >>>>>>> not (Prasanth Prakash); this also allowed us to >>>>>>> remove an over- >>>>>>> engineered function to do this math. >>>>>>> > > In x86 when CPPC is used, the unit is really unit-less in CPPC tables. > This means that cpu->perf_caps.highest_perf can be just 0xff, instead > of some scaled cppc max performance corresponding to max MHZ the > processor can support. This allows the processor to cap at max which it > can deliver. > Is this case not possible for ARM SoCs? > > Thanks, > Srinivas > If I understand the question properly, I don't think it matters, and I don't think that will work for x86 either. This patch is meant to allow CPPC to continue operating solely based on the abstract scale provided by the ACPI tables; this should be true regardless of architecture. Any actual processor performance changes are still guided solely by the CPPC scale provided in the tables, and not the values in the cpu->perf_caps struct. Assuming I understand the kernel code, the values in cpu->perf_caps -- in this case -- are really just for reporting to user space via sysfs, which is the root of the problem: user space expects frequencies, and we have none when using CPPC so we have to provide an approximation. In those circumstances, I think a value of 0xff would be kind of confusing in sysfs, since it's basically saying the CPU is operating at a frequency equal to the largest integer value. To be fair, this is how the ARM processor implements CPPC; I have not examined in detail the newly submitted x86 patches to use CPPC so I cannot comment on those. This patch was written well before those showed up.
On Tue, 2016-08-23 at 10:14 -0600, Al Stone wrote: > > > [...] > > In x86 when CPPC is used, the unit is really unit-less in CPPC > > tables. > > This means that cpu->perf_caps.highest_perf can be just 0xff, > > instead > > of some scaled cppc max performance corresponding to max MHZ the > > processor can support. This allows the processor to cap at max > > which it > > can deliver. > > Is this case not possible for ARM SoCs? > > [...] > If I understand the question properly, I don't think it matters, and > I don't > think that will work for x86 either. > > This patch is meant to allow CPPC to continue operating solely based > on the > abstract scale provided by the ACPI tables; this should be true > regardless > of architecture. Any actual processor performance changes are still > guided > solely by the CPPC scale provided in the tables, and not the values > in the > cpu->perf_caps struct. > > Assuming I understand the kernel code, the values in cpu->perf_caps > -- in this > case -- are really just for reporting to user space via sysfs, which > is the > root of the problem: user space expects frequencies, and we have none > when using > CPPC so we have to provide an approximation. In those circumstances, > I think a > value of 0xff would be kind of confusing in sysfs, since it's > basically saying > the CPU is operating at a frequency equal to the largest integer > value. > > To be fair, this is how the ARM processor implements CPPC; I have not > examined > in detail the newly submitted x86 patches to use CPPC so I cannot > comment on > those. This patch was written well before those showed up. Currently we are not using cppc-cpufreq driver, so not will not directly impact (you are not changing acpi-cpufreq source, which we are using). x86 has other way to get max/min cpufreq policy frequencies using MSRs. So you may choose to ignore my comments here, as long as your changes are limited to drivers/cpufreq/cppc_cpufreq.c When you are doing: policy->min = cpu->perf_caps.lowest_perf * cppc_dmi_max_khz / cpu- >perf_caps.highest_perf; Aren't you assuming that scale from max to min performance is only related to frequency? It is possible that many points in between can be same frequency with multiple voltages. As per spec " The platform may choose to use a single metric such as processor frequency, or it may choose to blend multiple hardware metrics to create a synthetic measure of performance". Thanks, Srinivas >
On 08/25/2016 04:00 PM, Pandruvada, Srinivas wrote: > On Tue, 2016-08-23 at 10:14 -0600, Al Stone wrote: >>>> > > [...] > >>> In x86 when CPPC is used, the unit is really unit-less in CPPC >>> tables. >>> This means that cpu->perf_caps.highest_perf can be just 0xff, >>> instead >>> of some scaled cppc max performance corresponding to max MHZ the >>> processor can support. This allows the processor to cap at max >>> which it >>> can deliver. >>> Is this case not possible for ARM SoCs? >>> > [...] >> If I understand the question properly, I don't think it matters, and >> I don't >> think that will work for x86 either. >> >> This patch is meant to allow CPPC to continue operating solely based >> on the >> abstract scale provided by the ACPI tables; this should be true >> regardless >> of architecture. Any actual processor performance changes are still >> guided >> solely by the CPPC scale provided in the tables, and not the values >> in the >> cpu->perf_caps struct. >> >> Assuming I understand the kernel code, the values in cpu->perf_caps >> -- in this >> case -- are really just for reporting to user space via sysfs, which >> is the >> root of the problem: user space expects frequencies, and we have none >> when using >> CPPC so we have to provide an approximation. In those circumstances, >> I think a >> value of 0xff would be kind of confusing in sysfs, since it's >> basically saying >> the CPU is operating at a frequency equal to the largest integer >> value. >> >> To be fair, this is how the ARM processor implements CPPC; I have not >> examined >> in detail the newly submitted x86 patches to use CPPC so I cannot >> comment on >> those. This patch was written well before those showed up. > Currently we are not using cppc-cpufreq driver, so not will not > directly impact (you are not changing acpi-cpufreq source, which we > are using). > x86 has other way to get max/min cpufreq policy frequencies using MSRs. > So you may choose to ignore my comments here, as long as your changes > are limited to drivers/cpufreq/cppc_cpufreq.c Ah, okay. Yes, the changes are intentionaly limited to cppc_cpufreq.c. > When you are doing: > policy->min = cpu->perf_caps.lowest_perf * cppc_dmi_max_khz / cpu- >> perf_caps.highest_perf; > > Aren't you assuming that scale from max to min performance is only > related to frequency? It is possible that many points in between can be > same frequency with multiple voltages. > As per spec " The platform may choose to use a single metric such as > processor frequency, or it may choose to blend multiple hardware > metrics to create a synthetic measure of performance". There are a couple of assumptions: (1) that the CPPC scale is a linear scale, and that (2) there is a direct correlation to the frequency. And this is why longer term, we have to separate performance reporting from the frequency -- any correlation between them is suspect on most modern processors. We know a priori that these assumptions are only approximations, at best. However, this is the only information we currently have, so we have to make a best guess...um, I mean, "apply heuristics". I'm treating this as two problems, really: the first is the immediate term where we need to make sure user space tools don't report complete garbage, which this patch tries to address. The second is the much larger problem of changing the way the kernel reports performance in general, and fixing the user space tools that rely on the info being reported; I'm still thinking through those patches (suggestions are always welcome).
On Wednesday, July 20, 2016 03:10:04 PM Al Stone wrote: > When CPPC is being used by ACPI on arm64, user space tools such as > cpupower report CPU frequency values from sysfs that are incorrect. > > What the driver was doing was reporting the values given by ACPI tables > in whatever scale was used to provide them. However, the ACPI spec > defines the CPPC values as unitless abstract numbers. Internal kernel > structures such as struct perf_cap, in contrast, expect these values > to be in KHz. When these struct values get reported via sysfs, the > user space tools also assume they are in KHz, causing them to report > incorrect values (for example, reporting a CPU frequency of 1MHz when > it should be 1.8GHz). > > The downside is that this approach has some assumptions: > > (1) It relies on SMBIOS3 being used, *and* that the Max Frequency > value for a processor is set to a non-zero value. > > (2) It assumes that all processors run at the same speed, or that > the CPPC values have all been scaled to reflect relative speed. > This patch retrieves the largest CPU Max Frequency from a type 4 DMI > record that it can find. This may not be an issue, however, as a > sampling of DMI data on x86 and arm64 indicates there is often only > one such record regardless. Since CPPC is relatively new, it is > unclear if the ACPI ASL will always be written to reflect any sort > of relative performance of processors of differing speeds. > > (3) It assumes that performance and frequency both scale linearly. > > For arm64 servers, this may be sufficient, but it does rely on > firmware values being set correctly. Hence, other approaches will > be considered in the future. > > This has been tested on three arm64 servers, with and without DMI, with > and without CPPC support. > > Changes for v5: > -- Move code to cpufreq/cppc_cpufreq.c from acpi/cppc_acpi.c to keep > frequency-related code together, and keep the CPPC abstract scale > in ACPI (Prashanth Prakash) > -- Fix the scaling to remove the incorrect assumption that frequency > was always a range from zero to max; as a practical matter, it is > not (Prasanth Prakash); this also allowed us to remove an over- > engineered function to do this math. > > Changes for v4: > -- Replaced magic constants with #defines (Rafael Wysocki) > -- Renamed cppc_unitless_to_khz() to cppc_to_khz() (Rafael Wysocki) > -- Replaced hidden initialization with a clearer form (Rafael Wysocki) > -- Instead of picking up the first Max Speed value from DMI, we will > now get the largest Max Speed; still an approximation, but slightly > less subject to error (Rafael Wysocki) > -- Kconfig for cppc_cpufreq now depends on DMI, instead of selecting > it, in order to make sure DMI is set up properly (Rafael Wysocki) > > Changes for v3: > -- Added clarifying commentary re short-term vs long-term fix (Alexey > Klimov) > -- Added range checking code to ensure proper arithmetic occurs, > especially no division by zero (Alexey Klimov) > > Changes for v2: > -- Corrected thinko: needed to have DEPENDS on DMI in Kconfig.arm, > not SELECT DMI (found by build daemon) > > Signed-off-by: Al Stone <ahs3@redhat.com> > Signed-off-by: Prashanth Prakash <pprakash@codeaurora.org> Applied. Thanks, Rafael -- To unsubscribe from this list: send the line "unsubscribe linux-pm" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On 09/13/2016 07:09 PM, Rafael J. Wysocki wrote: > On Wednesday, July 20, 2016 03:10:04 PM Al Stone wrote: >> When CPPC is being used by ACPI on arm64, user space tools such as >> cpupower report CPU frequency values from sysfs that are incorrect. >> >> What the driver was doing was reporting the values given by ACPI tables >> in whatever scale was used to provide them. However, the ACPI spec >> defines the CPPC values as unitless abstract numbers. Internal kernel >> structures such as struct perf_cap, in contrast, expect these values >> to be in KHz. When these struct values get reported via sysfs, the >> user space tools also assume they are in KHz, causing them to report >> incorrect values (for example, reporting a CPU frequency of 1MHz when >> it should be 1.8GHz). >> >> The downside is that this approach has some assumptions: >> >> (1) It relies on SMBIOS3 being used, *and* that the Max Frequency >> value for a processor is set to a non-zero value. >> >> (2) It assumes that all processors run at the same speed, or that >> the CPPC values have all been scaled to reflect relative speed. >> This patch retrieves the largest CPU Max Frequency from a type 4 DMI >> record that it can find. This may not be an issue, however, as a >> sampling of DMI data on x86 and arm64 indicates there is often only >> one such record regardless. Since CPPC is relatively new, it is >> unclear if the ACPI ASL will always be written to reflect any sort >> of relative performance of processors of differing speeds. >> >> (3) It assumes that performance and frequency both scale linearly. >> >> For arm64 servers, this may be sufficient, but it does rely on >> firmware values being set correctly. Hence, other approaches will >> be considered in the future. >> >> This has been tested on three arm64 servers, with and without DMI, with >> and without CPPC support. >> >> Changes for v5: >> -- Move code to cpufreq/cppc_cpufreq.c from acpi/cppc_acpi.c to keep >> frequency-related code together, and keep the CPPC abstract scale >> in ACPI (Prashanth Prakash) >> -- Fix the scaling to remove the incorrect assumption that frequency >> was always a range from zero to max; as a practical matter, it is >> not (Prasanth Prakash); this also allowed us to remove an over- >> engineered function to do this math. >> >> Changes for v4: >> -- Replaced magic constants with #defines (Rafael Wysocki) >> -- Renamed cppc_unitless_to_khz() to cppc_to_khz() (Rafael Wysocki) >> -- Replaced hidden initialization with a clearer form (Rafael Wysocki) >> -- Instead of picking up the first Max Speed value from DMI, we will >> now get the largest Max Speed; still an approximation, but slightly >> less subject to error (Rafael Wysocki) >> -- Kconfig for cppc_cpufreq now depends on DMI, instead of selecting >> it, in order to make sure DMI is set up properly (Rafael Wysocki) >> >> Changes for v3: >> -- Added clarifying commentary re short-term vs long-term fix (Alexey >> Klimov) >> -- Added range checking code to ensure proper arithmetic occurs, >> especially no division by zero (Alexey Klimov) >> >> Changes for v2: >> -- Corrected thinko: needed to have DEPENDS on DMI in Kconfig.arm, >> not SELECT DMI (found by build daemon) >> >> Signed-off-by: Al Stone <ahs3@redhat.com> >> Signed-off-by: Prashanth Prakash <pprakash@codeaurora.org> > > Applied. > > Thanks, > Rafael > I've been on vacation so just now am seeing this. Thanks, Rafael!
diff --git a/drivers/cpufreq/cppc_cpufreq.c b/drivers/cpufreq/cppc_cpufreq.c index 8882b8e..6debc18 100644 --- a/drivers/cpufreq/cppc_cpufreq.c +++ b/drivers/cpufreq/cppc_cpufreq.c @@ -19,10 +19,19 @@ #include <linux/delay.h> #include <linux/cpu.h> #include <linux/cpufreq.h> +#include <linux/dmi.h> #include <linux/vmalloc.h> +#include <asm/unaligned.h> + #include <acpi/cppc_acpi.h> +/* Minimum struct length needed for the DMI processor entry we want */ +#define DMI_ENTRY_PROCESSOR_MIN_LENGTH 48 + +/* Offest in the DMI processor structure for the max frequency */ +#define DMI_PROCESSOR_MAX_SPEED 0x14 + /* * These structs contain information parsed from per CPU * ACPI _CPC structures. @@ -32,6 +41,39 @@ */ static struct cpudata **all_cpu_data; +/* Capture the max KHz from DMI */ +static u64 cppc_dmi_max_khz; + +/* Callback function used to retrieve the max frequency from DMI */ +static void cppc_find_dmi_mhz(const struct dmi_header *dm, void *private) +{ + const u8 *dmi_data = (const u8 *)dm; + u16 *mhz = (u16 *)private; + + if (dm->type == DMI_ENTRY_PROCESSOR && + dm->length >= DMI_ENTRY_PROCESSOR_MIN_LENGTH) { + u16 val = (u16)get_unaligned((const u16 *) + (dmi_data + DMI_PROCESSOR_MAX_SPEED)); + *mhz = val > *mhz ? val : *mhz; + } +} + +/* Look up the max frequency in DMI */ +static u64 cppc_get_dmi_max_khz(void) +{ + u16 mhz = 0; + + dmi_walk(cppc_find_dmi_mhz, &mhz); + + /* + * Real stupid fallback value, just in case there is no + * actual value set. + */ + mhz = mhz ? mhz : 1; + + return (1000 * mhz); +} + static int cppc_cpufreq_set_target(struct cpufreq_policy *policy, unsigned int target_freq, unsigned int relation) @@ -42,7 +84,7 @@ static int cppc_cpufreq_set_target(struct cpufreq_policy *policy, cpu = all_cpu_data[policy->cpu]; - cpu->perf_ctrls.desired_perf = target_freq; + cpu->perf_ctrls.desired_perf = target_freq * policy->max / cppc_dmi_max_khz; freqs.old = policy->cur; freqs.new = target_freq; @@ -94,8 +136,10 @@ static int cppc_cpufreq_cpu_init(struct cpufreq_policy *policy) return ret; } - policy->min = cpu->perf_caps.lowest_perf; - policy->max = cpu->perf_caps.highest_perf; + cppc_dmi_max_khz = cppc_get_dmi_max_khz(); + + policy->min = cpu->perf_caps.lowest_perf * cppc_dmi_max_khz / cpu->perf_caps.highest_perf; + policy->max = cppc_dmi_max_khz; policy->cpuinfo.min_freq = policy->min; policy->cpuinfo.max_freq = policy->max; policy->shared_type = cpu->shared_type; @@ -112,7 +156,8 @@ static int cppc_cpufreq_cpu_init(struct cpufreq_policy *policy) cpu->cur_policy = policy; /* Set policy->cur to max now. The governors will adjust later. */ - policy->cur = cpu->perf_ctrls.desired_perf = cpu->perf_caps.highest_perf; + policy->cur = cppc_dmi_max_khz; + cpu->perf_ctrls.desired_perf = cpu->perf_caps.highest_perf; ret = cppc_set_perf(cpu_num, &cpu->perf_ctrls); if (ret)
When CPPC is being used by ACPI on arm64, user space tools such as cpupower report CPU frequency values from sysfs that are incorrect. What the driver was doing was reporting the values given by ACPI tables in whatever scale was used to provide them. However, the ACPI spec defines the CPPC values as unitless abstract numbers. Internal kernel structures such as struct perf_cap, in contrast, expect these values to be in KHz. When these struct values get reported via sysfs, the user space tools also assume they are in KHz, causing them to report incorrect values (for example, reporting a CPU frequency of 1MHz when it should be 1.8GHz). The downside is that this approach has some assumptions: (1) It relies on SMBIOS3 being used, *and* that the Max Frequency value for a processor is set to a non-zero value. (2) It assumes that all processors run at the same speed, or that the CPPC values have all been scaled to reflect relative speed. This patch retrieves the largest CPU Max Frequency from a type 4 DMI record that it can find. This may not be an issue, however, as a sampling of DMI data on x86 and arm64 indicates there is often only one such record regardless. Since CPPC is relatively new, it is unclear if the ACPI ASL will always be written to reflect any sort of relative performance of processors of differing speeds. (3) It assumes that performance and frequency both scale linearly. For arm64 servers, this may be sufficient, but it does rely on firmware values being set correctly. Hence, other approaches will be considered in the future. This has been tested on three arm64 servers, with and without DMI, with and without CPPC support. Changes for v5: -- Move code to cpufreq/cppc_cpufreq.c from acpi/cppc_acpi.c to keep frequency-related code together, and keep the CPPC abstract scale in ACPI (Prashanth Prakash) -- Fix the scaling to remove the incorrect assumption that frequency was always a range from zero to max; as a practical matter, it is not (Prasanth Prakash); this also allowed us to remove an over- engineered function to do this math. Changes for v4: -- Replaced magic constants with #defines (Rafael Wysocki) -- Renamed cppc_unitless_to_khz() to cppc_to_khz() (Rafael Wysocki) -- Replaced hidden initialization with a clearer form (Rafael Wysocki) -- Instead of picking up the first Max Speed value from DMI, we will now get the largest Max Speed; still an approximation, but slightly less subject to error (Rafael Wysocki) -- Kconfig for cppc_cpufreq now depends on DMI, instead of selecting it, in order to make sure DMI is set up properly (Rafael Wysocki) Changes for v3: -- Added clarifying commentary re short-term vs long-term fix (Alexey Klimov) -- Added range checking code to ensure proper arithmetic occurs, especially no division by zero (Alexey Klimov) Changes for v2: -- Corrected thinko: needed to have DEPENDS on DMI in Kconfig.arm, not SELECT DMI (found by build daemon) Signed-off-by: Al Stone <ahs3@redhat.com> Signed-off-by: Prashanth Prakash <pprakash@codeaurora.org> --- drivers/cpufreq/cppc_cpufreq.c | 53 ++++++++++++++++++++++++++++++++++++++---- 1 file changed, 49 insertions(+), 4 deletions(-)