turbostat 19.08.31 is available
Subsystems (2)
Len Brown
maintainerThanks for the patches everybody!
You can download the latest turbostat (and x86_energy_perf_policy) for testing here:
git://git.kernel.org/pub/scm/linux/kernel/git/lenb/linux.git turbostat
tools/power/x86/turbostat/Makefile | 3 +- tools/power/x86/turbostat/turbostat.c | 101 ++++++++++++++------- tools/power/x86/x86_energy_perf_policy/Makefile | 3 +- .../x86_energy_perf_policy.8 | 2 +- .../x86_energy_perf_policy.c | 28 +++--- 5 files changed, 90 insertions(+), 47 deletions(-)
[PATCH 01/19] tools/power x86_energy_perf_policy: Fix "uninitialized variable" warnings at -O2 [PATCH 02/19] tools/power/x86: Enable compiler optimisations and Fortify by default [PATCH 03/19] tools/power: Fix typo in man page
[PATCH 04/19] tools/power x86_energy_perf_policy: Fix argument parsing
[PATCH 05/19] tools/power turbostat: remove duplicate pc10 column
[PATCH 06/19] tools/power turbostat: Make interval calculation per thread to reduce jitter [PATCH 07/19] tools/power turbostat: fix leak of file descriptor on error return path [PATCH 08/19] tools/power turbostat: fix file descriptor leaks
[PATCH 09/19] tools/power turbostat: fix buffer overrun
Prarit, Naoya,
For now, I went with your initial 1-line fix,
as I currently don't have an appetite for a 400-line patch when 1-line spending 1KB / CPU will float the boat.
Yeah, it is a bummer that this "bundle formatted output into a single buffered write" optimiation is even necessary... you'd think that stdio could be smarter to automatically buffer multiple prints into a single write...
[PATCH 10/19] tools/power turbostat: add Jacobsville support [PATCH 11/19] tools/power turbostat: Fix Haswell Core systems [PATCH 12/19] tools/power turbostat: rename has_hsw_msrs() [PATCH 13/19] tools/power turbostat: Add Ice Lake NNPI support [PATCH 14/19] tools/power turbostat: read from pipes too [PATCH 15/19] tools/power turbostat: do not enforce 1ms [PATCH 16/19] tools/power turbostat: Fix CPU%C1 display value
[PATCH 17/19] tools/power turbostat: Fix caller parameter of get_tdp_amd() [PATCH 18/19] tools/power turbostat: Add support for Hygon Fam 18h (Dhyana) RAPL
Although these were marked RFC instead of PATCH, I applied them, because they were otherwise properly signed, and seemed ready.
[PATCH 19/19] tools/power turbostat: update version number
cheers,
-Len
Len Brown
maintainerPATCH 1/19 tools/power x86_energy_perf_policy: Fix "uninitialized variable" warnings at -O2
From: Ben Hutchings <[email protected]>
x86_energy_perf_policy first uses __get_cpuid() to check the maximum CPUID level and exits if it is too low. It then assumes that later calls will succeed (which I think is architecturally guaranteed). It also assumes that CPUID works at all (which is not guaranteed on x86_32).
If optimisations are enabled, gcc warns about potentially uninitialized variables. Fix this by adding an exit-on-error after every call to __get_cpuid() instead of just checking the maximum level.
--- .../x86_energy_perf_policy.c | 26 +++++++++++-------- 1 file changed, 15 insertions(+), 11 deletions(-)
tools/power/x86/x86_energy_perf_policy/x86_energy_perf_policy.c +15 -11
@@ -1259,6 +1259,15 @@ void probe_dev_msr(void)if (system ("/sbin/modprobe msr > /dev/null 2>&1" ))err (-5 ,"no /dev/cpu/0/msr, Try \"# modprobe msr\" " ); } + +static void get_cpuid_or_exit (unsigned int leaf, +unsigned int *eax,unsigned int *ebx, +unsigned int *ecx,unsigned int *edx) +{ +if (!__get_cpuid (leaf, eax, ebx, ecx, edx)) +errx (1 ,"Processor not supported\n" ); +} +/* * early_cpuid() * initialize turbo_is_enabled, has_hwp, has_epb
@@ -1266,15 +1275,10 @@ void probe_dev_msr(void)*/ void early_cpuid (void ) { -unsigned int eax, ebx, ecx, edx, max_level; +unsigned int eax, ebx, ecx, edx;unsigned int fms, family, model; -__get_cpuid (0 , &max_level, &ebx, &ecx, &edx); - -if (max_level <6 ) -errx (1 ,"Processor not supported\n" ); - -__get_cpuid (1 , &fms, &ebx, &ecx, &edx); +get_cpuid_or_exit (1 , &fms, &ebx, &ecx, &edx); family = (fms >>8 ) &0xf ; model = (fms >>4 ) &0xf ;if (family ==6 || family ==0xf )
@@ -1288,7 +1292,7 @@ void early_cpuid(void) bdx_highest_ratio = msr &0xFF ; } -__get_cpuid (0x6 , &eax, &ebx, &ecx, &edx); +get_cpuid_or_exit (0x6 , &eax, &ebx, &ecx, &edx); turbo_is_enabled = (eax >>1 ) &1 ; has_hwp = (eax >>7 ) &1 ; has_epb = (ecx >>3 ) &1 ;
@@ -1306,7 +1310,7 @@ void parse_cpuid(void) eax = ebx = ecx = edx =0 ; -__get_cpuid (0 , &max_level, &ebx, &ecx, &edx); +get_cpuid_or_exit (0 , &max_level, &ebx, &ecx, &edx);if (ebx ==0x756e6547 && edx ==0x49656e69 && ecx ==0x6c65746e ) genuine_intel =1 ;
@@ -1315,7 +1319,7 @@ void parse_cpuid(void)fprintf (stderr,"CPUID(0): %.4s%.4s%.4s " , (char *)&ebx, (char *)&edx, (char *)&ecx); -__get_cpuid (1 , &fms, &ebx, &ecx, &edx); +get_cpuid_or_exit (1 , &fms, &ebx, &ecx, &edx); family = (fms >>8 ) &0xf ; model = (fms >>4 ) &0xf ; stepping = fms &0xf ;
@@ -1340,7 +1344,7 @@ void parse_cpuid(void)errx (1 ,"CPUID: no MSR" ); -__get_cpuid (0x6 , &eax, &ebx, &ecx, &edx); +get_cpuid_or_exit (0x6 , &eax, &ebx, &ecx, &edx);/* turbo_is_enabled already set */ /* has_hwp already set */ has_hwp_notify = eax & (1 <<8 );
Signature · 1 line
2.20.1
Len Brown
maintainerPATCH 2/19 tools/power/x86: Enable compiler optimisations and Fortify by default
From: Ben Hutchings <[email protected]>
Compiling without optimisations is silly, especially since some warnings depend on the optimiser. Use -O2.
Fortify adds warnings for unchecked I/O (among other things), which seems to be a good idea for user-space code. Enable that too.
--- tools/power/x86/turbostat/Makefile | 3 ++- tools/power/x86/x86_energy_perf_policy/Makefile | 3 ++- 2 files changed, 4 insertions(+), 2 deletions(-)
tools/power/x86/turbostat/Makefile +2 -1
@@ -9,9 +9,10 @@ ifeq ("$(origin O)", "command line") endif turbostat : turbostat.c -override CFLAGS += -Wall -I../../../include +override CFLAGS += -O2 -Wall -I../../../include override CFLAGS += -DMSRHEADER='"../../../../arch/x86/include/asm/msr-index.h"' override CFLAGS += -DINTEL_FAMILY_HEADER='"../../../../arch/x86/include/asm/intel-family.h"' +override CFLAGS += -D_FORTIFY_SOURCE=2 %: %.c @mkdir -p $(BUILD_OUTPUT)
tools/power/x86/x86_energy_perf_policy/Makefile +2 -1
@@ -9,8 +9,9 @@ ifeq ("$(origin O)", "command line") endif x86_energy_perf_policy : x86_energy_perf_policy.c -override CFLAGS += -Wall -I../../../include +override CFLAGS += -O2 -Wall -I../../../include override CFLAGS += -DMSRHEADER='"../../../../arch/x86/include/asm/msr-index.h"' +override CFLAGS += -D_FORTIFY_SOURCE=2 %: %.c @mkdir -p $(BUILD_OUTPUT)
Signature · 1 line
2.20.1
Len Brown
maintainerPATCH 3/19 tools/power: Fix typo in man page
In reply to Len BrownFrom: Matt Lupfer <[email protected]>
From context, we mean EPB (Enegry Performance Bias).
--- tools/power/x86/x86_energy_perf_policy/x86_energy_perf_policy.8 | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
tools/power/x86/x86_energy_perf_policy/x86_energy_perf_policy.8 +1 -1
@@ -40,7 +40,7 @@ in the same processor package. Hardware P-States (HWP) are effectively an expansion of hardware P-state control from the opportunistic turbo-mode P-state range to include the entire range of available P-states. -On Broadwell Xeon, the initial HWP implementation, EBP influenced HWP. +On Broadwell Xeon, the initial HWP implementation, EPB influenced HWP. That influence was removed in subsequent generations, where it was moved to the Energy_Performance_Preference (EPP) field in
Signature · 1 line
2.20.1
Len Brown
maintainerPATCH 4/19 tools/power x86_energy_perf_policy: Fix argument parsing
In reply to Len BrownFrom: "Zephaniah E. Loss-Cutler-Hull" <[email protected]>
The -w argument in x86_energy_perf_policy currently triggers an unconditional segfault.
This is because the argument string reads: "+a:c:dD:E:e:f:m:M:rt:u:vw" and yet the argument handler expects an argument.
When parse_optarg_string is called with a null argument, we then proceed to crash in strncmp, not horribly friendly.
The man page describes -w as taking an argument, the long form (--hwp-window) is correctly marked as taking a required argument, and the code expects it.
As such, this patch simply marks the short form (-w) as requiring an argument.
--- tools/power/x86/x86_energy_perf_policy/x86_energy_perf_policy.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
tools/power/x86/x86_energy_perf_policy/x86_energy_perf_policy.c +1 -1
@@ -545,7 +545,7 @@ void cmdline(int argc, char **argv) progname = argv[0 ]; -while ((opt =getopt_long_only (argc, argv,"+a:c:dD:E:e:f:m:M:rt:u:vw" , +while ((opt =getopt_long_only (argc, argv,"+a:c:dD:E:e:f:m:M:rt:u:vw:" , long_options, &option_index)) !=-1 ) {switch (opt) {case 'a' :
Signature · 1 line
2.20.1
Len Brown
maintainerPATCH 5/19 tools/power turbostat: remove duplicate pc10 column
In reply to Len BrownFrom: Len Brown <[email protected]>
Remove the duplicate pc10 column.
--- tools/power/x86/turbostat/turbostat.c | 1 - 1 file changed, 1 deletion(-)
tools/power/x86/turbostat/turbostat.c +0 -1
@@ -849,7 +849,6 @@ int dump_counters(struct thread_data *t, struct core_data *c, outp +=sprintf (outp,"pc8: %016llX\n" , p->pc8); outp +=sprintf (outp,"pc9: %016llX\n" , p->pc9); outp +=sprintf (outp,"pc10: %016llX\n" , p->pc10); - outp +=sprintf (outp,"pc10: %016llX\n" , p->pc10); outp +=sprintf (outp,"cpu_lpi: %016llX\n" , p->cpu_lpi); outp +=sprintf (outp,"sys_lpi: %016llX\n" , p->sys_lpi); outp +=sprintf (outp,"Joules PKG: %0X\n" , p->energy_pkg);
Signature · 1 line
2.20.1
Len Brown
maintainerPATCH 6/19 tools/power turbostat: Make interval calculation per thread to reduce jitter
In reply to Len BrownFrom: Yazen Ghannam <[email protected]>
Turbostat currently normalizes TSC and other values by dividing by an interval. This interval is the delta between the start of one global (all counters on all CPUs) sampling and the start of another. However, this introduces a lot of jitter into the data.
In order to reduce jitter, the interval calculation should be based on timestamps taken per thread and close to the start of the thread's sampling.
Define a per thread time value to hold the delta between samples taken on the thread.
Use the timestamp taken at the beginning of sampling to calculate the delta.
Move the thread's beginning timestamp to after the CPU migration to avoid jitter due to the migration.
Use the global time delta for the average time delta.
--- tools/power/x86/turbostat/turbostat.c | 13 ++++++++++--- 1 file changed, 10 insertions(+), 3 deletions(-)
tools/power/x86/turbostat/turbostat.c +10 -3
@@ -166,6 +166,7 @@ size_t cpu_present_setsize, cpu_affinity_setsize, cpu_subset_size;struct thread_data {struct timeval tv_begin;struct timeval tv_end; +struct timeval tv_delta;unsigned long long tsc;unsigned long long aperf;unsigned long long mperf;
@@ -910,7 +911,7 @@ int format_counters(struct thread_data *t, struct core_data *c,if (DO_BIC (BIC_TOD )) outp +=sprintf (outp,"%10ld.%06ld\t" , t->tv_end.tv_sec, t->tv_end.tv_usec); - interval_float = tv_delta.tv_sec + tv_delta.tv_usec/1000000.0 ; + interval_float = t->tv_delta.tv_sec + t->tv_delta.tv_usec/1000000.0 ; tsc = t->tsc * tsc_tweak;
@@ -1308,6 +1309,7 @@ delta_thread(struct thread_data *new, struct thread_data *old,* over-write old w/ new so we can print end of interval values */ +timersub (&new->tv_begin, &old->tv_begin, &old->tv_delta); old->tv_begin = new->tv_begin; old->tv_end = new->tv_end;
@@ -1403,6 +1405,8 @@ void clear_counters(struct thread_data *t, struct core_data *c, struct pkg_data t->tv_begin.tv_usec =0 ; t->tv_end.tv_sec =0 ; t->tv_end.tv_usec =0 ; + t->tv_delta.tv_sec =0 ; + t->tv_delta.tv_usec =0 ; t->tsc =0 ; t->aperf =0 ;
@@ -1572,6 +1576,9 @@ void compute_average(struct thread_data *t, struct core_data *c,for_all_cpus (sum_counters, t, c, p); +/* Use the global time delta for the average. */ + average.threads.tv_delta = tv_delta; + average.threads.tsc /= topo.num_cpus; average.threads.aperf /= topo.num_cpus; average.threads.mperf /= topo.num_cpus;
@@ -1761,13 +1768,13 @@ int get_counters(struct thread_data *t, struct core_data *c, struct pkg_data *p)struct msr_counter *mp;int i; -gettimeofday (&t->tv_begin, (struct timezone *)NULL ); -if (cpu_migrate (cpu)) {fprintf (outf,"Could not migrate to CPU %d\n" , cpu);return -1 ; } +gettimeofday (&t->tv_begin, (struct timezone *)NULL ); +if (first_counter_read)get_apic_id (t); retry:
Signature · 1 line
2.20.1
Len Brown
maintainerPATCH 7/19 tools/power turbostat: fix leak of file descriptor on error return path
In reply to Len BrownFrom: Colin Ian King <[email protected]>
Currently the error return path does not close the file fp and leaks a file descriptor. Fix this by closing the file.
--- tools/power/x86/turbostat/turbostat.c | 1 + 1 file changed, 1 insertion(+)
tools/power/x86/turbostat/turbostat.c +1 -0
@@ -2944,6 +2944,7 @@ int snapshot_sys_lpi_us(void)if (retval !=1 ) {fprintf (stderr,"Disabling Low Power Idle System output\n" );BIC_NOT_PRESENT (BIC_SYS_LPI ); +fclose (fp);return -1 ; }fclose (fp);
Signature · 1 line
2.20.1
Len Brown
maintainerPATCH 8/19 tools/power turbostat: fix file descriptor leaks
In reply to Len BrownFrom: "Gustavo A. R. Silva" <[email protected]>
Fix file descriptor leaks by closing fp before return.
Addresses-Coverity-ID: 1444591 ("Resource leak")
Addresses-Coverity-ID: 1444592 ("Resource leak")
--- tools/power/x86/turbostat/turbostat.c | 1 + 1 file changed, 1 insertion(+)
tools/power/x86/turbostat/turbostat.c +1 -0
@@ -2918,6 +2918,7 @@ int snapshot_cpu_lpi_us(void)if (retval !=1 ) {fprintf (stderr,"Disabling Low Power Idle CPU output\n" );BIC_NOT_PRESENT (BIC_CPU_LPI ); +fclose (fp);return -1 ; }
Signature · 1 line
2.20.1
Len Brown
maintainerPATCH 9/19 tools/power turbostat: fix buffer overrun
In reply to Len BrownFrom: Naoya Horiguchi <[email protected]>
turbostat could be terminated by general protection fault on some latest hardwares which (for example) support 9 levels of C-states and show 18 "tADDED" lines. That bloats the total output and finally causes buffer overrun. So let's extend the buffer to avoid this.
--- tools/power/x86/turbostat/turbostat.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
tools/power/x86/turbostat/turbostat.c +1 -1
@@ -5131,7 +5131,7 @@ int initialize_counters(int cpu_id)void allocate_output_buffer () { - output_buffer =calloc (1 , (1 + topo.num_cpus) *1024 ); + output_buffer =calloc (1 , (1 + topo.num_cpus) *2048 ); outp = output_buffer;if (outp ==NULL )err (-1 ,"calloc output buffer" );
Signature · 1 line
2.20.1
Len Brown
maintainerPATCH 10/19 tools/power turbostat: add Jacobsville support
In reply to Len BrownFrom: Zhang Rui <[email protected]>
Jacobsville behaves like Denverton.
--- tools/power/x86/turbostat/turbostat.c | 3 +++ 1 file changed, 3 insertions(+)
tools/power/x86/turbostat/turbostat.c +3 -0
@@ -4591,6 +4591,9 @@ unsigned int intel_model_duplicates(unsigned int model)case INTEL_FAM6_ICELAKE_MOBILE :return INTEL_FAM6_CANNONLAKE_MOBILE ; + +case INTEL_FAM6_ATOM_TREMONT_X : +return INTEL_FAM6_ATOM_GOLDMONT_X ; }return model; }
Signature · 1 line
2.20.1
Len Brown
maintainerPATCH 11/19 tools/power turbostat: Fix Haswell Core systems
In reply to Len BrownFrom: Len Brown <[email protected]>
turbostat: cpu0: msr offset 0x630 read failed: Input/output error
because Haswell Core does not have C8-C10.
Output C8-C10 only on Haswell ULT.
--- tools/power/x86/turbostat/turbostat.c | 10 ++++++---- 1 file changed, 6 insertions(+), 4 deletions(-)
tools/power/x86/turbostat/turbostat.c +6 -4
@@ -3217,6 +3217,7 @@ int probe_nhm_msrs(unsigned int family, unsigned int model)break ;case INTEL_FAM6_HASWELL_CORE :/* HSW */ case INTEL_FAM6_HASWELL_X :/* HSX */ +case INTEL_FAM6_HASWELL_ULT :/* HSW */ case INTEL_FAM6_HASWELL_GT3E :/* HSW */ case INTEL_FAM6_BROADWELL_CORE :/* BDW */ case INTEL_FAM6_BROADWELL_GT3E :/* BDW */
@@ -3413,6 +3414,7 @@ int has_config_tdp(unsigned int family, unsigned int model)case INTEL_FAM6_IVYBRIDGE :/* IVB */ case INTEL_FAM6_HASWELL_CORE :/* HSW */ case INTEL_FAM6_HASWELL_X :/* HSX */ +case INTEL_FAM6_HASWELL_ULT :/* HSW */ case INTEL_FAM6_HASWELL_GT3E :/* HSW */ case INTEL_FAM6_BROADWELL_CORE :/* BDW */ case INTEL_FAM6_BROADWELL_GT3E :/* BDW */
@@ -3849,6 +3851,7 @@ void rapl_probe_intel(unsigned int family, unsigned int model)case INTEL_FAM6_SANDYBRIDGE :case INTEL_FAM6_IVYBRIDGE :case INTEL_FAM6_HASWELL_CORE :/* HSW */ +case INTEL_FAM6_HASWELL_ULT :/* HSW */ case INTEL_FAM6_HASWELL_GT3E :/* HSW */ case INTEL_FAM6_BROADWELL_CORE :/* BDW */ case INTEL_FAM6_BROADWELL_GT3E :/* BDW */
@@ -4040,6 +4043,7 @@ void perf_limit_reasons_probe(unsigned int family, unsigned int model)switch (model) {case INTEL_FAM6_HASWELL_CORE :/* HSW */ +case INTEL_FAM6_HASWELL_ULT :/* HSW */ case INTEL_FAM6_HASWELL_GT3E :/* HSW */ do_gfx_perf_limit_reasons =1 ;case INTEL_FAM6_HASWELL_X :/* HSX */
@@ -4259,6 +4263,7 @@ int has_snb_msrs(unsigned int family, unsigned int model)case INTEL_FAM6_IVYBRIDGE_X :/* IVB Xeon */ case INTEL_FAM6_HASWELL_CORE :/* HSW */ case INTEL_FAM6_HASWELL_X :/* HSW */ +case INTEL_FAM6_HASWELL_ULT :/* HSW */ case INTEL_FAM6_HASWELL_GT3E :/* HSW */ case INTEL_FAM6_BROADWELL_CORE :/* BDW */ case INTEL_FAM6_BROADWELL_GT3E :/* BDW */
@@ -4292,7 +4297,7 @@ int has_hsw_msrs(unsigned int family, unsigned int model)return 0 ;switch (model) { -case INTEL_FAM6_HASWELL_CORE : +case INTEL_FAM6_HASWELL_ULT :/* HSW */ case INTEL_FAM6_BROADWELL_CORE :/* BDW */ case INTEL_FAM6_SKYLAKE_MOBILE :/* SKL */ case INTEL_FAM6_CANNONLAKE_MOBILE :/* CNL */
@@ -4576,9 +4581,6 @@ unsigned int intel_model_duplicates(unsigned int model)case INTEL_FAM6_XEON_PHI_KNM :return INTEL_FAM6_XEON_PHI_KNL ; -case INTEL_FAM6_HASWELL_ULT : -return INTEL_FAM6_HASWELL_CORE ; -case INTEL_FAM6_BROADWELL_X :case INTEL_FAM6_BROADWELL_XEON_D :/* BDX-DE */ return INTEL_FAM6_BROADWELL_X ;
Signature · 1 line
2.20.1
Len Brown
maintainerPATCH 12/19 tools/power turbostat: rename has_hsw_msrs()
In reply to Len BrownFrom: Len Brown <[email protected]>
Perhaps if this more descriptive name had been used, then we wouldn't have had the HSW ULT vs HSW CORE bug, fixed by the previous commit.
--- tools/power/x86/turbostat/turbostat.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-)
tools/power/x86/turbostat/turbostat.c +4 -4
@@ -4280,7 +4280,7 @@ int has_snb_msrs(unsigned int family, unsigned int model) }/* -* HSW adds support for additional MSRs: +* HSW ULT added support for C8/C9/C10 MSRs: * * MSR_PKG_C8_RESIDENCY 0x00000630 * MSR_PKG_C9_RESIDENCY 0x00000631
@@ -4291,7 +4291,7 @@ int has_snb_msrs(unsigned int family, unsigned int model)* MSR_PKGC10_IRTL 0x00000635 * */ -int has_hsw_msrs (unsigned int family,unsigned int model) +int has_c8910_msrs (unsigned int family,unsigned int model) {if (!genuine_intel)return 0 ;
@@ -4833,12 +4833,12 @@ void process_cpuid()BIC_NOT_PRESENT (BIC_CPU_c7);BIC_NOT_PRESENT (BIC_Pkgpc7); } -if (has_hsw_msrs (family, model)) { +if (has_c8910_msrs (family, model)) {BIC_PRESENT (BIC_Pkgpc8);BIC_PRESENT (BIC_Pkgpc9);BIC_PRESENT (BIC_Pkgpc10); } - do_irtl_hsw =has_hsw_msrs (family, model); + do_irtl_hsw =has_c8910_msrs (family, model);if (has_skl_msrs (family, model)) {BIC_PRESENT (BIC_Totl_c0);BIC_PRESENT (BIC_Any_c0);
Signature · 1 line
2.20.1
Len Brown
maintainerPATCH 13/19 tools/power turbostat: Add Ice Lake NNPI support
In reply to Len BrownFrom: Rajneesh Bhardwaj <[email protected]>
This enables turbostat utility on Ice Lake NNPI SoC.
--- tools/power/x86/turbostat/turbostat.c | 1 + 1 file changed, 1 insertion(+)
tools/power/x86/turbostat/turbostat.c +1 -0
@@ -4592,6 +4592,7 @@ unsigned int intel_model_duplicates(unsigned int model)return INTEL_FAM6_SKYLAKE_MOBILE ;case INTEL_FAM6_ICELAKE_MOBILE : +case INTEL_FAM6_ICELAKE_NNPI :return INTEL_FAM6_CANNONLAKE_MOBILE ;case INTEL_FAM6_ATOM_TREMONT_X :
Signature · 1 line
2.20.1
Len Brown
maintainerPATCH 14/19 tools/power turbostat: read from pipes too
In reply to Len BrownFrom: Artem Bityutskiy <[email protected]>
Commit '47936f944e78 tools/power turbostat: fix printing on input' make a valid fix, but it completely disabled piped stdin support, which is a valuable use-case. Indeed, if stdin is a pipe, turbostat won't read anything from it, so it becomes impossible to get turbostat output at user-defined moments, instead of the regular intervals.
There is no reason why this should works for terminals, but not for pipes. This patch improves the situation. Instead of ignoring pipes, we read data from them but gracefully handle the EOF case.
--- tools/power/x86/turbostat/turbostat.c | 20 ++++++++++++++++---- 1 file changed, 16 insertions(+), 4 deletions(-)
tools/power/x86/turbostat/turbostat.c +16 -4
@@ -100,6 +100,7 @@ unsigned int has_hwp_epp; /* IA32_HWP_REQUEST[bits 31:24] */unsigned int has_hwp_pkg;/* IA32_HWP_REQUEST_PKG */ unsigned int has_misc_feature_control;unsigned int first_counter_read =1 ; +int ignore_stdin;#define RAPL_PKG (1 << 0)/* 0x610 MSR_PKG_POWER_LIMIT */
@@ -3013,26 +3014,37 @@ void setup_signal_handler(void)void do_sleep (void ) { -struct timeval select_timeout; +struct timeval tout; +struct timespec rest;fd_set readfds;int retval;FD_ZERO (&readfds);FD_SET (0 , &readfds); -if (!isatty (fileno (stdin))) { +if (ignore_stdin) {nanosleep (&interval_ts,NULL );return ; } - select_timeout = interval_tv; - retval =select (1 , &readfds,NULL ,NULL , &select_timeout); + tout = interval_tv; + retval =select (1 , &readfds,NULL ,NULL , &tout);if (retval ==1 ) {switch (getc (stdin)) {case 'q' : exit_requested =1 ;break ; +case EOF : +/* +* 'stdin' is a pipe closed on the other end. There +* won't be any further input. +*/ + ignore_stdin =1 ; +/* Sleep the rest of the time */ + rest.tv_sec = (tout.tv_sec + tout.tv_usec /1000000 ); + rest.tv_nsec = (tout.tv_usec %1000000 ) *1000 ; +nanosleep (&rest,NULL ); }/* make sure this manually-invoked interval is at least 1ms long */ nanosleep (&one_msec,NULL );
Signature · 1 line
2.20.1
Len Brown
maintainerPATCH 15/19 tools/power turbostat: do not enforce 1ms
In reply to Len BrownFrom: Artem Bityutskiy <[email protected]>
Turbostat works by taking a snapshot of counters, sleeping, taking another snapshot, calculating deltas, and printing out the table.
The sleep time is controlled via -i option or by user sending a signal or a character to stdin. In the latter case, turbostat always adds 1 ms sleep before it reads the counters, in order to avoid larger imprecisions in the results in prints.
While the 1 ms delay may be a good idea for a "dumb" user, it is a problem for an "aware" user. I do thousands and thousands of measurements over a short period of time (like 2ms), and turbostat unconditionally adds a 1ms to my interval, so I cannot get what I really need.
This patch removes the unconditional 1ms sleep. This is an expert user tool, after all, and non-experts will unlikely ever use it in the non-fixed interval mode anyway, so I think it is OK to remove the 1ms delay.
--- tools/power/x86/turbostat/turbostat.c | 5 ----- 1 file changed, 5 deletions(-)
tools/power/x86/turbostat/turbostat.c +0 -5
@@ -39,7 +39,6 @@ FILE *outf;int *fd_percpu;struct timeval interval_tv = {5 ,0 };struct timespec interval_ts = {5 ,0 }; -struct timespec one_msec = {0 ,1000000 };unsigned int num_iterations;unsigned int debug;unsigned int quiet;
@@ -2994,8 +2993,6 @@ static void signal_handler (int signal)fprintf (stderr,"SIGUSR1\n" );break ; } -/* make sure this manually-invoked interval is at least 1ms long */ -nanosleep (&one_msec,NULL ); }void setup_signal_handler (void )
@@ -3046,8 +3043,6 @@ void do_sleep(void) rest.tv_nsec = (tout.tv_usec %1000000 ) *1000 ;nanosleep (&rest,NULL ); } -/* make sure this manually-invoked interval is at least 1ms long */ -nanosleep (&one_msec,NULL ); } }
Signature · 1 line
2.20.1
Len Brown
maintainerPATCH 16/19 tools/power turbostat: Fix CPU%C1 display value
In reply to Len BrownFrom: Srinivas Pandruvada <[email protected]>
In some case C1% will be wrong value, when platform doesn't have MSR for C1 residency.
For example:
Core CPU CPU%c1 - - 100.00 0 0 100.00 0 2 100.00 1 1 100.00 1 3 100.00
But adding Busy% will fix this
Core CPU Busy% CPU%c1 - - 99.77 0.23 0 0 99.77 0.23 0 2 99.77 0.23 1 1 99.77 0.23 1 3 99.77 0.23
This issue can be reproduced on most of the recent systems including Broadwell, Skylake and later.
This is because if we don't select Busy% or Avg_MHz or Bzy_MHz then mperf value will not be read from MSR, so it will be 0. But this is required for C1% calculation when MSR for C1 residency is not present. Same is true for C3, C6 and C7 column selection.
So add another define DO_BIC_READ(), which doesn't depend on user column selection and use for mperf, C3, C6 and C7 related counters. So when there is no platform support for C1 residency counters, we still read these counters, if the CPU has support and user selected display of CPU%c1.
--- tools/power/x86/turbostat/turbostat.c | 23 +++++++++++++++++------ 1 file changed, 17 insertions(+), 6 deletions(-)
tools/power/x86/turbostat/turbostat.c +17 -6
@@ -507,6 +507,7 @@ unsigned long long bic_enabled = (0xFFFFFFFFFFFFFFFFULL & ~BIC_DISABLED_BY_DEFAUunsigned long long bic_present =BIC_USEC |BIC_TOD | BIC_sysfs |BIC_APIC |BIC_X2APIC ;#define DO_BIC (COUNTER_NAME ) (bic_enabled & bic_present & COUNTER_NAME) +#define DO_BIC_READ (COUNTER_NAME ) (bic_present & COUNTER_NAME)#define ENABLE_BIC (COUNTER_NAME ) (bic_enabled |= COUNTER_NAME)#define BIC_PRESENT (COUNTER_BIT ) (bic_present |= COUNTER_BIT)#define BIC_NOT_PRESENT (COUNTER_BIT ) (bic_present &= ~COUNTER_BIT)
@@ -1287,6 +1288,14 @@ delta_core(struct core_data *new, struct core_data *old) } } +int soft_c1_residency_display (int bic) +{ +if (!DO_BIC (BIC_CPU_c1) || use_c1_residency_msr) +return 0 ; + +return DO_BIC_READ (bic); +} +/* * old = new - old */
@@ -1323,7 +1332,8 @@ delta_thread(struct thread_data *new, struct thread_data *old, old->c1 = new->c1 - old->c1; -if (DO_BIC (BIC_Avg_MHz) ||DO_BIC (BIC_Busy) ||DO_BIC (BIC_Bzy_MHz)) { +if (DO_BIC (BIC_Avg_MHz) ||DO_BIC (BIC_Busy) ||DO_BIC (BIC_Bzy_MHz) || +soft_c1_residency_display (BIC_Avg_MHz)) {if ((new->aperf > old->aperf) && (new->mperf > old->mperf)) { old->aperf = new->aperf - old->aperf; old->mperf = new->mperf - old->mperf;
@@ -1780,7 +1790,8 @@ int get_counters(struct thread_data *t, struct core_data *c, struct pkg_data *p) retry: t->tsc =rdtsc ();/* we are running on local CPU of interest */ -if (DO_BIC (BIC_Avg_MHz) ||DO_BIC (BIC_Busy) ||DO_BIC (BIC_Bzy_MHz)) { +if (DO_BIC (BIC_Avg_MHz) ||DO_BIC (BIC_Busy) ||DO_BIC (BIC_Bzy_MHz) || +soft_c1_residency_display (BIC_Avg_MHz)) {unsigned long long tsc_before, tsc_between, tsc_after, aperf_time, mperf_time;/*
@@ -1857,20 +1868,20 @@ int get_counters(struct thread_data *t, struct core_data *c, struct pkg_data *p)if (!(t->flags &CPU_IS_FIRST_THREAD_IN_CORE )) goto done; -if (DO_BIC (BIC_CPU_c3)) { +if (DO_BIC (BIC_CPU_c3) ||soft_c1_residency_display (BIC_CPU_c3)) {if (get_msr (cpu,MSR_CORE_C3_RESIDENCY , &c->c3))return -6 ; } -if (DO_BIC (BIC_CPU_c6) && !do_knl_cstates) { +if ((DO_BIC (BIC_CPU_c6) ||soft_c1_residency_display (BIC_CPU_c6)) && !do_knl_cstates) {if (get_msr (cpu,MSR_CORE_C6_RESIDENCY , &c->c6))return -7 ; - }else if (do_knl_cstates) { + }else if (do_knl_cstates ||soft_c1_residency_display (BIC_CPU_c6)) {if (get_msr (cpu,MSR_KNL_CORE_C6_RESIDENCY , &c->c6))return -7 ; } -if (DO_BIC (BIC_CPU_c7)) +if (DO_BIC (BIC_CPU_c7) ||soft_c1_residency_display (BIC_CPU_c7))if (get_msr (cpu,MSR_CORE_C7_RESIDENCY , &c->c7))return -8 ;
Signature · 1 line
2.20.1
Len Brown
maintainerPATCH 17/19 tools/power turbostat: Fix caller parameter of get_tdp_amd()
In reply to Len BrownFrom: Pu Wen <[email protected]>
Commit 9392bd98bba760be96ee ("tools/power turbostat: Add support for AMD Fam 17h (Zen) RAPL") add a function get_tdp_amd(), the parameter is CPU family. But the rapl_probe_amd() function use wrong model parameter. Fix the wrong caller parameter of get_tdp_amd() to use family.
--- tools/power/x86/turbostat/turbostat.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
tools/power/x86/turbostat/turbostat.c +1 -1
@@ -4031,7 +4031,7 @@ void rapl_probe_amd(unsigned int family, unsigned int model) rapl_energy_units =ldexp (1.0 , -(msr >>8 &0x1f )); rapl_power_units =ldexp (1.0 , -(msr &0xf )); - tdp =get_tdp_amd (model); + tdp =get_tdp_amd (family); rapl_joule_counter_range =0xFFFFFFFF * rapl_energy_units / tdp;if (!quiet)
Signature · 1 line
2.20.1
Len Brown
maintainerPATCH 18/19 tools/power turbostat: Add support for Hygon Fam 18h (Dhyana) RAPL
In reply to Len BrownFrom: Pu Wen <[email protected]>
Commit 9392bd98bba760be96ee ("tools/power turbostat: Add support for AMD Fam 17h (Zen) RAPL") and the commit 3316f99a9f1b68c578c5 ("tools/power
turbostat: Also read package power on AMD F17h (Zen)") add AMD Fam 17h RAPL support.
Hygon Family 18h(Dhyana) support RAPL in bit 14 of CPUID 0x80000007 EDX, and has MSRs RAPL_PWR_UNIT/CORE_ENERGY_STAT/PKG_ENERGY_STAT. So add Hygon Dhyana Family 18h support for RAPL.
Already tested on Hygon multi-node systems and it shows correct per-core energy usage and the total package power.
--- tools/power/x86/turbostat/turbostat.c | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-)
tools/power/x86/turbostat/turbostat.c +7 -2
@@ -59,6 +59,7 @@ unsigned int do_irtl_hsw;unsigned int units =1000000 ;/* MHz etc */ unsigned int genuine_intel;unsigned int authentic_amd; +unsigned int hygon_genuine;unsigned int max_level, max_extended_level;unsigned int has_invariant_tsc;unsigned int do_nhm_platform_info;
@@ -1730,7 +1731,7 @@ void get_apic_id(struct thread_data *t)if (!DO_BIC (BIC_X2APIC ))return ; -if (authentic_amd) { +if (authentic_amd || hygon_genuine) {unsigned int topology_extensions;if (max_extended_level <0x8000001e )
@@ -3831,6 +3832,7 @@ double get_tdp_amd(unsigned int family) {switch (family) {case 0x17 : +case 0x18 :default :/* This is the max stock TDP of HEDT/Server Fam17h chips */ return 250.0 ;
@@ -4011,6 +4013,7 @@ void rapl_probe_amd(unsigned int family, unsigned int model)switch (family) {case 0x17 :/* Zen, Zen+ */ +case 0x18 :/* Hygon Dhyana */ do_rapl =RAPL_AMD_F17H |RAPL_PER_CORE_ENERGY ;if (rapl_joules) {BIC_PRESENT (BIC_Pkg_J);
@@ -4047,7 +4050,7 @@ void rapl_probe(unsigned int family, unsigned int model) {if (genuine_intel)rapl_probe_intel (family, model); -if (authentic_amd) +if (authentic_amd || hygon_genuine)rapl_probe_amd (family, model); }
@@ -4632,6 +4635,8 @@ void process_cpuid() genuine_intel =1 ; else if (ebx == 0x68747541 && ecx == 0x444d4163 && edx == 0x69746e65) authentic_amd =1 ; + else if (ebx == 0x6f677948 && ecx == 0x656e6975 && edx == 0x6e65476e) + hygon_genuine =1 ;if (!quiet) fprintf(outf, "CPUID(0): %.4s%.4s%.4s ",
Signature · 1 line
2.20.1
Len Brown
maintainerPATCH 19/19 tools/power turbostat: update version number
In reply to Len BrownFrom: Len Brown <[email protected]>
Today is 19.08.31, at least in some parts of the world.
--- tools/power/x86/turbostat/turbostat.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
tools/power/x86/turbostat/turbostat.c +1 -1
@@ -5306,7 +5306,7 @@ int get_and_dump_counters(void) }void print_version () { -fprintf (outf,"turbostat version 19.03.20" +fprintf (outf,"turbostat version 19.08.31" " - Len Brown <[email protected]>\n" ); }
Signature · 1 line
2.20.1