From 3ae8508663372b93c5556a887e96ed0ca5df0711 Mon Sep 17 00:00:00 2001 From: Len Brown Date: Sun, 6 Apr 2025 12:23:22 -0400 Subject: [PATCH 01/16] tools/power turbostat: Document GNR UncMHz domain convention Document that on Intel Granite Rapids Systems, Uncore domains 0-2 are CPU domains, and uncore domains 3-4 are IO domains. Signed-off-by: Len Brown --- tools/power/x86/turbostat/turbostat.8 | 1 + 1 file changed, 1 insertion(+) diff --git a/tools/power/x86/turbostat/turbostat.8 b/tools/power/x86/turbostat/turbostat.8 index 144565151e1e..e86493880c16 100644 --- a/tools/power/x86/turbostat/turbostat.8 +++ b/tools/power/x86/turbostat/turbostat.8 @@ -205,6 +205,7 @@ The system configuration dump (if --quiet is not used) is followed by statistics \fBUncMHz\fP per-package uncore MHz, instantaneous sample. .PP \fBUMHz1.0\fP per-package uncore MHz for domain=1 and fabric_cluster=0, instantaneous sample. System summary is the average of all packages. +Intel Granite Rapids systems use domains 0-2 for CPUs, and 3-4 for IO, with cluster always 0. For the "--show" and "--hide" options, use "UncMHz" to operate on all UMHz*.* as a group. .SH TOO MUCH INFORMATION EXAMPLE By default, turbostat dumps all possible information -- a system configuration header, followed by columns for all counters. -- 2.51.0 From f8b136ef2605c1bf62020462d10e35228760aa19 Mon Sep 17 00:00:00 2001 From: Zhang Rui Date: Wed, 19 Mar 2025 08:53:07 +0800 Subject: [PATCH 02/16] tools/power turbostat: Restore GFX sysfs fflush() call Do fflush() to discard the buffered data, before each read of the graphics sysfs knobs. Fixes: ba99a4fc8c24 ("tools/power turbostat: Remove unnecessary fflush() call") Signed-off-by: Zhang Rui Signed-off-by: Len Brown --- tools/power/x86/turbostat/turbostat.c | 1 + 1 file changed, 1 insertion(+) diff --git a/tools/power/x86/turbostat/turbostat.c b/tools/power/x86/turbostat/turbostat.c index 70e17d4ad9b6..c9a34c16c7a8 100644 --- a/tools/power/x86/turbostat/turbostat.c +++ b/tools/power/x86/turbostat/turbostat.c @@ -6039,6 +6039,7 @@ int snapshot_graphics(int idx) int retval; rewind(gfx_info[idx].fp); + fflush(gfx_info[idx].fp); switch (idx) { case GFX_rc6: -- 2.51.0 From 994633894f208a0151baaee1688ab3c431912553 Mon Sep 17 00:00:00 2001 From: Len Brown Date: Sun, 6 Apr 2025 12:53:18 -0400 Subject: [PATCH 03/16] tools/power turbostat: re-factor sysfs code Probe cpuidle "sysfs" residency and counts separately, since soon we will make one disabled on, and the other disabled off. Clarify that some BIC (build-in-counters) are actually "groups". since we're about to re-name some of those groups. no functional change. Signed-off-by: Len Brown --- tools/power/x86/turbostat/turbostat.c | 31 ++++++++++++++++++--------- 1 file changed, 21 insertions(+), 10 deletions(-) diff --git a/tools/power/x86/turbostat/turbostat.c b/tools/power/x86/turbostat/turbostat.c index c9a34c16c7a8..df0391bedcde 100644 --- a/tools/power/x86/turbostat/turbostat.c +++ b/tools/power/x86/turbostat/turbostat.c @@ -273,10 +273,10 @@ struct msr_counter bic[] = { #define BIC_NMI (1ULL << 61) #define BIC_CPU_c1e (1ULL << 62) -#define BIC_TOPOLOGY (BIC_Package | BIC_Node | BIC_CoreCnt | BIC_PkgCnt | BIC_Core | BIC_CPU | BIC_Die) -#define BIC_THERMAL_PWR (BIC_CoreTmp | BIC_PkgTmp | BIC_PkgWatt | BIC_CorWatt | BIC_GFXWatt | BIC_RAMWatt | BIC_PKG__ | BIC_RAM__ | BIC_SysWatt) -#define BIC_FREQUENCY (BIC_Avg_MHz | BIC_Busy | BIC_Bzy_MHz | BIC_TSC_MHz | BIC_GFXMHz | BIC_GFXACTMHz | BIC_SAMMHz | BIC_SAMACTMHz | BIC_UNCORE_MHZ) -#define BIC_IDLE (BIC_Busy | BIC_sysfs | BIC_CPU_c1 | BIC_CPU_c3 | BIC_CPU_c6 | BIC_CPU_c7 | BIC_GFX_rc6 | BIC_Pkgpc2 | BIC_Pkgpc3 | BIC_Pkgpc6 | BIC_Pkgpc7 | BIC_Pkgpc8 | BIC_Pkgpc9 | BIC_Pkgpc10 | BIC_CPU_LPI | BIC_SYS_LPI | BIC_Mod_c6 | BIC_Totl_c0 | BIC_Any_c0 | BIC_GFX_c0 | BIC_CPUGFX | BIC_SAM_mc6 | BIC_Diec6) +#define BIC_GROUP_TOPOLOGY (BIC_Package | BIC_Node | BIC_CoreCnt | BIC_PkgCnt | BIC_Core | BIC_CPU | BIC_Die) +#define BIC_GROUP_THERMAL_PWR (BIC_CoreTmp | BIC_PkgTmp | BIC_PkgWatt | BIC_CorWatt | BIC_GFXWatt | BIC_RAMWatt | BIC_PKG__ | BIC_RAM__ | BIC_SysWatt) +#define BIC_GROUP_FREQUENCY (BIC_Avg_MHz | BIC_Busy | BIC_Bzy_MHz | BIC_TSC_MHz | BIC_GFXMHz | BIC_GFXACTMHz | BIC_SAMMHz | BIC_SAMACTMHz | BIC_UNCORE_MHZ) +#define BIC_GROUP_IDLE (BIC_Busy | BIC_sysfs | BIC_CPU_c1 | BIC_CPU_c3 | BIC_CPU_c6 | BIC_CPU_c7 | BIC_GFX_rc6 | BIC_Pkgpc2 | BIC_Pkgpc3 | BIC_Pkgpc6 | BIC_Pkgpc7 | BIC_Pkgpc8 | BIC_Pkgpc9 | BIC_Pkgpc10 | BIC_CPU_LPI | BIC_SYS_LPI | BIC_Mod_c6 | BIC_Totl_c0 | BIC_Any_c0 | BIC_GFX_c0 | BIC_CPUGFX | BIC_SAM_mc6 | BIC_Diec6) #define BIC_OTHER (BIC_IRQ | BIC_NMI | BIC_SMI | BIC_ThreadC | BIC_CoreTmp | BIC_IPC) #define BIC_DISABLED_BY_DEFAULT (BIC_USEC | BIC_TOD | BIC_APIC | BIC_X2APIC) @@ -2354,16 +2354,16 @@ unsigned long long bic_lookup(char *name_list, enum show_hide_mode mode) retval |= ~0; break; } else if (!strcmp(name_list, "topology")) { - retval |= BIC_TOPOLOGY; + retval |= BIC_GROUP_TOPOLOGY; break; } else if (!strcmp(name_list, "power")) { - retval |= BIC_THERMAL_PWR; + retval |= BIC_GROUP_THERMAL_PWR; break; } else if (!strcmp(name_list, "idle")) { - retval |= BIC_IDLE; + retval |= BIC_GROUP_IDLE; break; } else if (!strcmp(name_list, "frequency")) { - retval |= BIC_FREQUENCY; + retval |= BIC_GROUP_FREQUENCY; break; } else if (!strcmp(name_list, "other")) { retval |= BIC_OTHER; @@ -10260,7 +10260,7 @@ int is_deferred_skip(char *name) return 0; } -void probe_sysfs(void) +void probe_cpuidle_residency(void) { char path[64]; char name_buf[16]; @@ -10304,6 +10304,16 @@ void probe_sysfs(void) if (state < min_state) min_state = state; } +} + +void probe_cpuidle_counts(void) +{ + char path[64]; + char name_buf[16]; + FILE *input; + int state; + int min_state = 1024, max_state = 0; + char *sp; for (state = 10; state >= 0; --state) { @@ -10602,7 +10612,8 @@ skip_cgroup_setting: print_bootcmd(); } - probe_sysfs(); + probe_cpuidle_residency(); + probe_cpuidle_counts(); if (!getuid()) set_rlimit(); -- 2.51.0 From 6f110a5e4f9977c31ce76fefbfef6fd4eab6bfb7 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Sun, 6 Apr 2025 10:00:04 -0700 Subject: [PATCH 04/16] Disable SLUB_TINY for build testing ... and don't error out so hard on missing module descriptions. Before commit 6c6c1fc09de3 ("modpost: require a MODULE_DESCRIPTION()") we used to warn about missing module descriptions, but only when building with extra warnigns (ie 'W=1'). After that commit the warning became an unconditional hard error. And it turns out not all modules have been converted despite the claims to the contrary. As reported by Damian Tometzki, the slub KUnit test didn't have a module description, and apparently nobody ever really noticed. The reason nobody noticed seems to be that the slub KUnit tests get disabled by SLUB_TINY, which also ends up disabling a lot of other code, both in tests and in slub itself. And so anybody doing full build tests didn't actually see this failre. So let's disable SLUB_TINY for build-only tests, since it clearly ends up limiting build coverage. Also turn the missing module descriptions error back into a warning, but let's keep it around for non-'W=1' builds. Reported-by: Damian Tometzki Link: https://lore.kernel.org/all/01070196099fd059-e8463438-7b1b-4ec8-816d-173874be9966-000000@eu-central-1.amazonses.com/ Cc: Masahiro Yamada Cc: Jeff Johnson Fixes: 6c6c1fc09de3 ("modpost: require a MODULE_DESCRIPTION()") Signed-off-by: Linus Torvalds --- mm/Kconfig | 2 +- scripts/mod/modpost.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/mm/Kconfig b/mm/Kconfig index d3fb3762887b..e113f713b493 100644 --- a/mm/Kconfig +++ b/mm/Kconfig @@ -201,7 +201,7 @@ config KVFREE_RCU_BATCHED config SLUB_TINY bool "Configure for minimal memory footprint" - depends on EXPERT + depends on EXPERT && !COMPILE_TEST select SLAB_MERGE_DEFAULT help Configures the slab allocator in a way to achieve minimal memory diff --git a/scripts/mod/modpost.c b/scripts/mod/modpost.c index 92627e8d0e16..be89921d60b6 100644 --- a/scripts/mod/modpost.c +++ b/scripts/mod/modpost.c @@ -1603,7 +1603,7 @@ static void read_symbols(const char *modname) } if (!get_modinfo(&info, "description")) - error("missing MODULE_DESCRIPTION() in %s\n", modname); + warn("missing MODULE_DESCRIPTION() in %s\n", modname); } for (sym = info.symtab_start; sym < info.symtab_stop; sym++) { -- 2.51.0 From ec4acd3166d8a7a03b059d01b9c6f11a658e833f Mon Sep 17 00:00:00 2001 From: Len Brown Date: Sun, 6 Apr 2025 14:29:57 -0400 Subject: [PATCH 05/16] tools/power turbostat: disable "cpuidle" invocation counters, by default Create "pct_idle" counter group, the sofware notion of residency so it can now be singled out, independent of other counter groups. Create "cpuidle" group, the cpuidle invocation counts. Disable "cpuidle", by default. Create "swidle" = "cpuidle" + "pct_idle". Undocument "sysfs", the old name for "swidle", but keep it working for backwards compatibilty. Create "hwidle", all the HW idle counters Modify "idle", enabled by default "idle" = "hwidle" + "pct_idle" (and now excludes "cpuidle") Signed-off-by: Len Brown --- tools/power/x86/turbostat/turbostat.8 | 12 +++++----- tools/power/x86/turbostat/turbostat.c | 34 +++++++++++++++++++++------ 2 files changed, 33 insertions(+), 13 deletions(-) diff --git a/tools/power/x86/turbostat/turbostat.8 b/tools/power/x86/turbostat/turbostat.8 index e86493880c16..b74ed916057e 100644 --- a/tools/power/x86/turbostat/turbostat.8 +++ b/tools/power/x86/turbostat/turbostat.8 @@ -100,7 +100,7 @@ The column name "all" can be used to enable all disabled-by-default built-in cou .PP \fB--show column\fP show only the specified built-in columns. May be invoked multiple times, or with a comma-separated list of column names. .PP -\fB--show CATEGORY --hide CATEGORY\fP Show and hide also accept a single CATEGORY of columns: "all", "topology", "idle", "frequency", "power", "sysfs", "other". +\fB--show CATEGORY --hide CATEGORY\fP Show and hide also accept a single CATEGORY of columns: "all", "topology", "idle", "frequency", "power", "cpuidle", "hwidle", "swidle", "other". "idle" (enabled by default), includes "hwidle" and "idle_pct". "cpuidle" (default disabled) includes cpuidle software invocation counters. "swidle" includes "cpuidle" plus "idle_pct". "hwidle" includes only hardware based idle residency counters. Older versions of turbostat used the term "sysfs" for what is now "swidle". .PP \fB--Dump\fP displays the raw counter values. .PP @@ -158,15 +158,15 @@ The system configuration dump (if --quiet is not used) is followed by statistics .PP \fBSMI\fP The number of System Management Interrupts serviced CPU during the measurement interval. While this counter is actually per-CPU, SMI are triggered on all processors, so the number should be the same for all CPUs. .PP -\fBC1, C2, C3...\fP The number times Linux requested the C1, C2, C3 idle state during the measurement interval. The system summary line shows the sum for all CPUs. These are C-state names as exported in /sys/devices/system/cpu/cpu*/cpuidle/state*/name. While their names are generic, their attributes are processor specific. They the system description section of output shows what MWAIT sub-states they are mapped to on each system. +\fBC1, C2, C3...\fP The number times Linux requested the C1, C2, C3 idle state during the measurement interval. The system summary line shows the sum for all CPUs. These are C-state names as exported in /sys/devices/system/cpu/cpu*/cpuidle/state*/name. While their names are generic, their attributes are processor specific. They the system description section of output shows what MWAIT sub-states they are mapped to on each system. These counters are in the "cpuidle" group, which is disabled, by default. .PP -\fBC1+, C2+, C3+...\fP The idle governor idle state misprediction statistics. Inidcates the number times Linux requested the C1, C2, C3 idle state during the measurement interval, but should have requested a deeper idle state (if it exists and enabled). These statistics come from the /sys/devices/system/cpu/cpu*/cpuidle/state*/below file. +\fBC1+, C2+, C3+...\fP The idle governor idle state misprediction statistics. Inidcates the number times Linux requested the C1, C2, C3 idle state during the measurement interval, but should have requested a deeper idle state (if it exists and enabled). These statistics come from the /sys/devices/system/cpu/cpu*/cpuidle/state*/below file. These counters are in the "cpuidle" group, which is disabled, by default. .PP -\fBC1-, C2-, C3-...\fP The idle governor idle state misprediction statistics. Inidcates the number times Linux requested the C1, C2, C3 idle state during the measurement interval, but should have requested a shallower idle state (if it exists and enabled). These statistics come from the /sys/devices/system/cpu/cpu*/cpuidle/state*/above file. +\fBC1-, C2-, C3-...\fP The idle governor idle state misprediction statistics. Inidcates the number times Linux requested the C1, C2, C3 idle state during the measurement interval, but should have requested a shallower idle state (if it exists and enabled). These statistics come from the /sys/devices/system/cpu/cpu*/cpuidle/state*/above file. These counters are in the "cpuidle" group, which is disabled, by default. .PP -\fBC1%, C2%, C3%\fP The residency percentage that Linux requested C1, C2, C3.... The system summary is the average of all CPUs in the system. Note that these are software, reflecting what was requested. The hardware counters reflect what was actually achieved. +\fBC1%, C2%, C3%\fP The residency percentage that Linux requested C1, C2, C3.... The system summary is the average of all CPUs in the system. Note that these are software, reflecting what was requested. The hardware counters reflect what was actually achieved. These counters are in the "pct_idle" group, which is enabled by default. .PP -\fBCPU%c1, CPU%c3, CPU%c6, CPU%c7\fP show the percentage residency in hardware core idle states. These numbers are from hardware residency counters. +\fBCPU%c1, CPU%c3, CPU%c6, CPU%c7\fP show the percentage residency in hardware core idle states. These numbers are from hardware residency counters and are in the "hwidle" group, which is enabled, by default. .PP \fBCoreTmp\fP Degrees Celsius reported by the per-core Digital Thermal Sensor. .PP diff --git a/tools/power/x86/turbostat/turbostat.c b/tools/power/x86/turbostat/turbostat.c index df0391bedcde..ab184f95cdaf 100644 --- a/tools/power/x86/turbostat/turbostat.c +++ b/tools/power/x86/turbostat/turbostat.c @@ -153,7 +153,7 @@ struct msr_counter bic[] = { { 0x0, "TSC_MHz", NULL, 0, 0, 0, NULL, 0 }, { 0x0, "IRQ", NULL, 0, 0, 0, NULL, 0 }, { 0x0, "SMI", NULL, 32, 0, FORMAT_DELTA, NULL, 0 }, - { 0x0, "sysfs", NULL, 0, 0, 0, NULL, 0 }, + { 0x0, "cpuidle", NULL, 0, 0, 0, NULL, 0 }, { 0x0, "CPU%c1", NULL, 0, 0, 0, NULL, 0 }, { 0x0, "CPU%c3", NULL, 0, 0, 0, NULL, 0 }, { 0x0, "CPU%c6", NULL, 0, 0, 0, NULL, 0 }, @@ -206,6 +206,7 @@ struct msr_counter bic[] = { { 0x0, "Sys_J", NULL, 0, 0, 0, NULL, 0 }, { 0x0, "NMI", NULL, 0, 0, 0, NULL, 0 }, { 0x0, "CPU%c1e", NULL, 0, 0, 0, NULL, 0 }, + { 0x0, "pct_idle", NULL, 0, 0, 0, NULL, 0 }, }; #define MAX_BIC (sizeof(bic) / sizeof(struct msr_counter)) @@ -219,7 +220,7 @@ struct msr_counter bic[] = { #define BIC_TSC_MHz (1ULL << 7) #define BIC_IRQ (1ULL << 8) #define BIC_SMI (1ULL << 9) -#define BIC_sysfs (1ULL << 10) +#define BIC_cpuidle (1ULL << 10) #define BIC_CPU_c1 (1ULL << 11) #define BIC_CPU_c3 (1ULL << 12) #define BIC_CPU_c6 (1ULL << 13) @@ -272,17 +273,20 @@ struct msr_counter bic[] = { #define BIC_Sys_J (1ULL << 60) #define BIC_NMI (1ULL << 61) #define BIC_CPU_c1e (1ULL << 62) +#define BIC_pct_idle (1ULL << 63) #define BIC_GROUP_TOPOLOGY (BIC_Package | BIC_Node | BIC_CoreCnt | BIC_PkgCnt | BIC_Core | BIC_CPU | BIC_Die) #define BIC_GROUP_THERMAL_PWR (BIC_CoreTmp | BIC_PkgTmp | BIC_PkgWatt | BIC_CorWatt | BIC_GFXWatt | BIC_RAMWatt | BIC_PKG__ | BIC_RAM__ | BIC_SysWatt) #define BIC_GROUP_FREQUENCY (BIC_Avg_MHz | BIC_Busy | BIC_Bzy_MHz | BIC_TSC_MHz | BIC_GFXMHz | BIC_GFXACTMHz | BIC_SAMMHz | BIC_SAMACTMHz | BIC_UNCORE_MHZ) -#define BIC_GROUP_IDLE (BIC_Busy | BIC_sysfs | BIC_CPU_c1 | BIC_CPU_c3 | BIC_CPU_c6 | BIC_CPU_c7 | BIC_GFX_rc6 | BIC_Pkgpc2 | BIC_Pkgpc3 | BIC_Pkgpc6 | BIC_Pkgpc7 | BIC_Pkgpc8 | BIC_Pkgpc9 | BIC_Pkgpc10 | BIC_CPU_LPI | BIC_SYS_LPI | BIC_Mod_c6 | BIC_Totl_c0 | BIC_Any_c0 | BIC_GFX_c0 | BIC_CPUGFX | BIC_SAM_mc6 | BIC_Diec6) +#define BIC_GROUP_HW_IDLE (BIC_Busy | BIC_CPU_c1 | BIC_CPU_c3 | BIC_CPU_c6 | BIC_CPU_c7 | BIC_GFX_rc6 | BIC_Pkgpc2 | BIC_Pkgpc3 | BIC_Pkgpc6 | BIC_Pkgpc7 | BIC_Pkgpc8 | BIC_Pkgpc9 | BIC_Pkgpc10 | BIC_CPU_LPI | BIC_SYS_LPI | BIC_Mod_c6 | BIC_Totl_c0 | BIC_Any_c0 | BIC_GFX_c0 | BIC_CPUGFX | BIC_SAM_mc6 | BIC_Diec6) +#define BIC_GROUP_SW_IDLE (BIC_Busy | BIC_cpuidle | BIC_pct_idle ) +#define BIC_GROUP_IDLE (BIC_GROUP_HW_IDLE | BIC_pct_idle) #define BIC_OTHER (BIC_IRQ | BIC_NMI | BIC_SMI | BIC_ThreadC | BIC_CoreTmp | BIC_IPC) -#define BIC_DISABLED_BY_DEFAULT (BIC_USEC | BIC_TOD | BIC_APIC | BIC_X2APIC) +#define BIC_DISABLED_BY_DEFAULT (BIC_USEC | BIC_TOD | BIC_APIC | BIC_X2APIC | BIC_cpuidle) unsigned long long bic_enabled = (0xFFFFFFFFFFFFFFFFULL & ~BIC_DISABLED_BY_DEFAULT); -unsigned long long bic_present = BIC_USEC | BIC_TOD | BIC_sysfs | BIC_APIC | BIC_X2APIC; +unsigned long long bic_present = BIC_USEC | BIC_TOD | BIC_cpuidle | BIC_pct_idle | 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) @@ -2362,6 +2366,15 @@ unsigned long long bic_lookup(char *name_list, enum show_hide_mode mode) } else if (!strcmp(name_list, "idle")) { retval |= BIC_GROUP_IDLE; break; + } else if (!strcmp(name_list, "swidle")) { + retval |= BIC_GROUP_SW_IDLE; + break; + } else if (!strcmp(name_list, "sysfs")) { /* legacy compatibility */ + retval |= BIC_GROUP_SW_IDLE; + break; + } else if (!strcmp(name_list, "hwidle")) { + retval |= BIC_GROUP_HW_IDLE; + break; } else if (!strcmp(name_list, "frequency")) { retval |= BIC_GROUP_FREQUENCY; break; @@ -2372,6 +2385,7 @@ unsigned long long bic_lookup(char *name_list, enum show_hide_mode mode) } if (i == MAX_BIC) { + fprintf(stderr, "deferred %s\n", name_list); if (mode == SHOW_LIST) { deferred_add_names[deferred_add_index++] = name_list; if (deferred_add_index >= MAX_DEFERRED) { @@ -10269,6 +10283,9 @@ void probe_cpuidle_residency(void) int min_state = 1024, max_state = 0; char *sp; + if (!DO_BIC(BIC_pct_idle)) + return; + for (state = 10; state >= 0; --state) { sprintf(path, "/sys/devices/system/cpu/cpu%d/cpuidle/state%d/name", base_cpu, state); @@ -10291,7 +10308,7 @@ void probe_cpuidle_residency(void) sprintf(path, "cpuidle/state%d/time", state); - if (!DO_BIC(BIC_sysfs) && !is_deferred_add(name_buf)) + if (!DO_BIC(BIC_pct_idle) && !is_deferred_add(name_buf)) continue; if (is_deferred_skip(name_buf)) @@ -10315,6 +10332,9 @@ void probe_cpuidle_counts(void) int min_state = 1024, max_state = 0; char *sp; + if (!DO_BIC(BIC_cpuidle)) + return; + for (state = 10; state >= 0; --state) { sprintf(path, "/sys/devices/system/cpu/cpu%d/cpuidle/state%d/name", base_cpu, state); @@ -10327,7 +10347,7 @@ void probe_cpuidle_counts(void) remove_underbar(name_buf); - if (!DO_BIC(BIC_sysfs) && !is_deferred_add(name_buf)) + if (!DO_BIC(BIC_cpuidle) && !is_deferred_add(name_buf)) continue; if (is_deferred_skip(name_buf)) -- 2.51.0 From 03e00e373cab981ad808271b2650700cfa0fbda6 Mon Sep 17 00:00:00 2001 From: Len Brown Date: Sun, 6 Apr 2025 14:49:20 -0400 Subject: [PATCH 06/16] tools/power turbostat: v2025.05.06 Support up to 8192 processors Add cpuidle governor debug telemetry, disabled by default Update default output to exclude cpuidle invocation counts Bug fixes Signed-off-by: Len Brown --- tools/power/x86/turbostat/turbostat.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/power/x86/turbostat/turbostat.c b/tools/power/x86/turbostat/turbostat.c index ab184f95cdaf..1b9fdc1a7ee8 100644 --- a/tools/power/x86/turbostat/turbostat.c +++ b/tools/power/x86/turbostat/turbostat.c @@ -9594,7 +9594,7 @@ int get_and_dump_counters(void) void print_version() { - fprintf(outf, "turbostat version 2025.02.02 - Len Brown \n"); + fprintf(outf, "turbostat version 2025.04.06 - Len Brown \n"); } #define COMMAND_LINE_SIZE 2048 -- 2.51.0 From 0efdedb3358aa78102967f242379686f94315830 Mon Sep 17 00:00:00 2001 From: =?utf8?q?Thomas=20Wei=C3=9Fschuh?= Date: Wed, 2 Apr 2025 21:21:57 +0100 Subject: [PATCH 07/16] tools/include: make uapi/linux/types.h usable from assembly MIME-Version: 1.0 Content-Type: text/plain; charset=utf8 Content-Transfer-Encoding: 8bit The "real" linux/types.h UAPI header gracefully degrades to a NOOP when included from assembly code. Mirror this behaviour in the tools/ variant. Test for __ASSEMBLER__ over __ASSEMBLY__ as the former is provided by the toolchain automatically. Reported-by: Mark Brown Closes: https://lore.kernel.org/lkml/af553c62-ca2f-4956-932c-dd6e3a126f58@sirena.org.uk/ Fixes: c9fbaa879508 ("selftests: vDSO: parse_vdso: Use UAPI headers instead of libc headers") Signed-off-by: Thomas Weißschuh Link: https://patch.msgid.link/20250321-uapi-consistency-v1-1-439070118dc0@linutronix.de Signed-off-by: Mark Brown Reviewed-by: Mark Brown Signed-off-by: Linus Torvalds --- tools/include/uapi/linux/types.h | 3 +++ 1 file changed, 3 insertions(+) diff --git a/tools/include/uapi/linux/types.h b/tools/include/uapi/linux/types.h index 91fa51a9c31d..85aa327245c6 100644 --- a/tools/include/uapi/linux/types.h +++ b/tools/include/uapi/linux/types.h @@ -4,6 +4,8 @@ #include +#ifndef __ASSEMBLER__ + /* copied from linux:include/uapi/linux/types.h */ #define __bitwise typedef __u16 __bitwise __le16; @@ -20,4 +22,5 @@ typedef __u32 __bitwise __wsum; #define __aligned_be64 __be64 __attribute__((aligned(8))) #define __aligned_le64 __le64 __attribute__((aligned(8))) +#endif /* __ASSEMBLER__ */ #endif /* _UAPI_LINUX_TYPES_H */ -- 2.51.0 From 0af2f6be1b4281385b618cb86ad946eded089ac8 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Sun, 6 Apr 2025 13:11:33 -0700 Subject: [PATCH 08/16] Linux 6.15-rc1 --- Makefile | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/Makefile b/Makefile index e55726a71d95..38689a0c3605 100644 --- a/Makefile +++ b/Makefile @@ -1,8 +1,8 @@ # SPDX-License-Identifier: GPL-2.0 VERSION = 6 -PATCHLEVEL = 14 +PATCHLEVEL = 15 SUBLEVEL = 0 -EXTRAVERSION = +EXTRAVERSION = -rc1 NAME = Baby Opossum Posse # *DOCUMENTATION* -- 2.51.0 From 6907e8093b3070d877ee607e5ceede60cfd08bde Mon Sep 17 00:00:00 2001 From: Heiko Stuebner Date: Fri, 11 Apr 2025 12:22:39 +0100 Subject: [PATCH 09/16] nvmem: rockchip-otp: Move read-offset into variant-data The RK3588 has an offset into the OTP area where the readable area begins and automatically adds this to the start address. Other variants are very much similar to rk3588, just with a different offset, so move that value into variant-data. To match the size in bytes, store this value also in bytes and not in number of blocks. Signed-off-by: Heiko Stuebner Tested-by: Nicolas Frattaroli Signed-off-by: Srinivas Kandagatla Link: https://lore.kernel.org/r/20250411112251.68002-2-srinivas.kandagatla@linaro.org Signed-off-by: Greg Kroah-Hartman --- drivers/nvmem/rockchip-otp.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/drivers/nvmem/rockchip-otp.c b/drivers/nvmem/rockchip-otp.c index ebc3f0b24166..3edfbfc2d722 100644 --- a/drivers/nvmem/rockchip-otp.c +++ b/drivers/nvmem/rockchip-otp.c @@ -59,7 +59,6 @@ #define RK3588_OTPC_AUTO_EN 0x08 #define RK3588_OTPC_INT_ST 0x84 #define RK3588_OTPC_DOUT0 0x20 -#define RK3588_NO_SECURE_OFFSET 0x300 #define RK3588_NBYTES 4 #define RK3588_BURST_NUM 1 #define RK3588_BURST_SHIFT 8 @@ -69,6 +68,7 @@ struct rockchip_data { int size; + int read_offset; const char * const *clks; int num_clks; nvmem_reg_read_t reg_read; @@ -196,7 +196,7 @@ static int rk3588_otp_read(void *context, unsigned int offset, addr_start = round_down(offset, RK3588_NBYTES) / RK3588_NBYTES; addr_end = round_up(offset + bytes, RK3588_NBYTES) / RK3588_NBYTES; addr_len = addr_end - addr_start; - addr_start += RK3588_NO_SECURE_OFFSET; + addr_start += otp->data->read_offset / RK3588_NBYTES; buf = kzalloc(array_size(addr_len, RK3588_NBYTES), GFP_KERNEL); if (!buf) @@ -280,6 +280,7 @@ static const char * const rk3588_otp_clocks[] = { static const struct rockchip_data rk3588_data = { .size = 0x400, + .read_offset = 0xc00, .clks = rk3588_otp_clocks, .num_clks = ARRAY_SIZE(rk3588_otp_clocks), .reg_read = rk3588_otp_read, -- 2.51.0 From 1b23c14c07326a095b93145ca9ea31cf53d4bde1 Mon Sep 17 00:00:00 2001 From: Heiko Stuebner Date: Fri, 11 Apr 2025 12:22:40 +0100 Subject: [PATCH 10/16] dt-bindings: nvmem: rockchip,otp: add missing limits for clock-names The clocks property correctly declares minItems and maxItems for its variants, but clock-names does not. Both properties are always used together, so should declare the same limits. Suggested-by: Krzysztof Kozlowski Signed-off-by: Heiko Stuebner Acked-by: Conor Dooley Tested-by: Nicolas Frattaroli Signed-off-by: Srinivas Kandagatla Link: https://lore.kernel.org/r/20250411112251.68002-3-srinivas.kandagatla@linaro.org Signed-off-by: Greg Kroah-Hartman --- Documentation/devicetree/bindings/nvmem/rockchip,otp.yaml | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/Documentation/devicetree/bindings/nvmem/rockchip,otp.yaml b/Documentation/devicetree/bindings/nvmem/rockchip,otp.yaml index a44d44b32809..3201ff8f9334 100644 --- a/Documentation/devicetree/bindings/nvmem/rockchip,otp.yaml +++ b/Documentation/devicetree/bindings/nvmem/rockchip,otp.yaml @@ -62,6 +62,8 @@ allOf: properties: clocks: maxItems: 3 + clock-names: + maxItems: 3 resets: maxItems: 1 reset-names: @@ -78,6 +80,8 @@ allOf: properties: clocks: minItems: 4 + clock-names: + minItems: 4 resets: minItems: 3 reset-names: -- 2.51.0 From 9165960606dff725174155472583efec60f25bab Mon Sep 17 00:00:00 2001 From: Heiko Stuebner Date: Fri, 11 Apr 2025 12:22:41 +0100 Subject: [PATCH 11/16] dt-bindings: nvmem: rockchip,otp: Add compatible for RK3576 Document the OTP memory found on Rockchip RK3576 SoC. The RK3576 uses the same set of clocks as the px30/rk3308 but has one reset more, so adapt the binding to handle this variant as well. Signed-off-by: Heiko Stuebner Acked-by: Conor Dooley Tested-by: Nicolas Frattaroli Signed-off-by: Srinivas Kandagatla Link: https://lore.kernel.org/r/20250411112251.68002-4-srinivas.kandagatla@linaro.org Signed-off-by: Greg Kroah-Hartman --- .../bindings/nvmem/rockchip,otp.yaml | 21 +++++++++++++++++++ 1 file changed, 21 insertions(+) diff --git a/Documentation/devicetree/bindings/nvmem/rockchip,otp.yaml b/Documentation/devicetree/bindings/nvmem/rockchip,otp.yaml index 3201ff8f9334..dc89020b0950 100644 --- a/Documentation/devicetree/bindings/nvmem/rockchip,otp.yaml +++ b/Documentation/devicetree/bindings/nvmem/rockchip,otp.yaml @@ -14,6 +14,7 @@ properties: enum: - rockchip,px30-otp - rockchip,rk3308-otp + - rockchip,rk3576-otp - rockchip,rk3588-otp reg: @@ -70,6 +71,26 @@ allOf: items: - const: phy + - if: + properties: + compatible: + contains: + enum: + - rockchip,rk3576-otp + then: + properties: + clocks: + maxItems: 3 + clock-names: + maxItems: 3 + resets: + minItems: 2 + maxItems: 2 + reset-names: + items: + - const: otp + - const: apb + - if: properties: compatible: -- 2.51.0 From 50d75a13a9ce880a5ef07a4ccc63ba561cc2e69a Mon Sep 17 00:00:00 2001 From: Heiko Stuebner Date: Fri, 11 Apr 2025 12:22:42 +0100 Subject: [PATCH 12/16] nvmem: rockchip-otp: add rk3576 variant data The variant works very similar to the rk3588, just with a different read-offset and size. Signed-off-by: Heiko Stuebner Tested-by: Nicolas Frattaroli Signed-off-by: Srinivas Kandagatla Link: https://lore.kernel.org/r/20250411112251.68002-5-srinivas.kandagatla@linaro.org Signed-off-by: Greg Kroah-Hartman --- drivers/nvmem/rockchip-otp.c | 12 ++++++++++++ 1 file changed, 12 insertions(+) diff --git a/drivers/nvmem/rockchip-otp.c b/drivers/nvmem/rockchip-otp.c index 3edfbfc2d722..d88f12c53242 100644 --- a/drivers/nvmem/rockchip-otp.c +++ b/drivers/nvmem/rockchip-otp.c @@ -274,6 +274,14 @@ static const struct rockchip_data px30_data = { .reg_read = px30_otp_read, }; +static const struct rockchip_data rk3576_data = { + .size = 0x100, + .read_offset = 0x700, + .clks = px30_otp_clocks, + .num_clks = ARRAY_SIZE(px30_otp_clocks), + .reg_read = rk3588_otp_read, +}; + static const char * const rk3588_otp_clocks[] = { "otp", "apb_pclk", "phy", "arb", }; @@ -295,6 +303,10 @@ static const struct of_device_id rockchip_otp_match[] = { .compatible = "rockchip,rk3308-otp", .data = &px30_data, }, + { + .compatible = "rockchip,rk3576-otp", + .data = &rk3576_data, + }, { .compatible = "rockchip,rk3588-otp", .data = &rk3588_data, -- 2.51.0 From f487438d370590193c5635ccbae1b1c51c5b273c Mon Sep 17 00:00:00 2001 From: Akhil P Oommen Date: Fri, 11 Apr 2025 12:22:43 +0100 Subject: [PATCH 13/16] dt-bindings: nvmem: qfprom: Add X1E80100 compatible Document compatible string for the QFPROM on X1E80100 platform. Signed-off-by: Akhil P Oommen Reviewed-by: Krzysztof Kozlowski Signed-off-by: Srinivas Kandagatla Link: https://lore.kernel.org/r/20250411112251.68002-6-srinivas.kandagatla@linaro.org Signed-off-by: Greg Kroah-Hartman --- Documentation/devicetree/bindings/nvmem/qcom,qfprom.yaml | 1 + 1 file changed, 1 insertion(+) diff --git a/Documentation/devicetree/bindings/nvmem/qcom,qfprom.yaml b/Documentation/devicetree/bindings/nvmem/qcom,qfprom.yaml index 39c209249c9c..4c332b44d35e 100644 --- a/Documentation/devicetree/bindings/nvmem/qcom,qfprom.yaml +++ b/Documentation/devicetree/bindings/nvmem/qcom,qfprom.yaml @@ -51,6 +51,7 @@ properties: - qcom,sm8450-qfprom - qcom,sm8550-qfprom - qcom,sm8650-qfprom + - qcom,x1e80100-qfprom - const: qcom,qfprom reg: -- 2.51.0 From 269e074da1882c296085a27f0742c47ac860072e Mon Sep 17 00:00:00 2001 From: =?utf8?q?Barnab=C3=A1s=20Cz=C3=A9m=C3=A1n?= Date: Fri, 11 Apr 2025 12:22:44 +0100 Subject: [PATCH 14/16] dt-bindings: nvmem: Add compatible for MS8937 MIME-Version: 1.0 Content-Type: text/plain; charset=utf8 Content-Transfer-Encoding: 8bit Document the QFPROM block found on MSM8937. Signed-off-by: Barnabás Czémán Reviewed-by: Krzysztof Kozlowski Signed-off-by: Srinivas Kandagatla Link: https://lore.kernel.org/r/20250411112251.68002-7-srinivas.kandagatla@linaro.org Signed-off-by: Greg Kroah-Hartman --- Documentation/devicetree/bindings/nvmem/qcom,qfprom.yaml | 1 + 1 file changed, 1 insertion(+) diff --git a/Documentation/devicetree/bindings/nvmem/qcom,qfprom.yaml b/Documentation/devicetree/bindings/nvmem/qcom,qfprom.yaml index 4c332b44d35e..e1a8856ccba4 100644 --- a/Documentation/devicetree/bindings/nvmem/qcom,qfprom.yaml +++ b/Documentation/devicetree/bindings/nvmem/qcom,qfprom.yaml @@ -28,6 +28,7 @@ properties: - qcom,msm8226-qfprom - qcom,msm8916-qfprom - qcom,msm8917-qfprom + - qcom,msm8937-qfprom - qcom,msm8974-qfprom - qcom,msm8976-qfprom - qcom,msm8996-qfprom -- 2.51.0 From eed6d954542fb55c814dd54b7fcc1b515bd76464 Mon Sep 17 00:00:00 2001 From: Dmitry Baryshkov Date: Fri, 11 Apr 2025 12:22:45 +0100 Subject: [PATCH 15/16] dt-bindings: nvmem: fixed-cell: increase bits start value to 31 If NVMEM uses a data stride bigger than a byte, the starting bit of the cell might be bigger than a byte (e.g. if the data comes in the second byte of the 4-byte word). Allow the staring bit to be 8 or greater to reflect such usecases. Signed-off-by: Dmitry Baryshkov Signed-off-by: Srinivas Kandagatla Link: https://lore.kernel.org/r/20250411112251.68002-8-srinivas.kandagatla@linaro.org Signed-off-by: Greg Kroah-Hartman --- Documentation/devicetree/bindings/nvmem/layouts/fixed-cell.yaml | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/Documentation/devicetree/bindings/nvmem/layouts/fixed-cell.yaml b/Documentation/devicetree/bindings/nvmem/layouts/fixed-cell.yaml index 8b3826243ddd..38e3ad50ff4f 100644 --- a/Documentation/devicetree/bindings/nvmem/layouts/fixed-cell.yaml +++ b/Documentation/devicetree/bindings/nvmem/layouts/fixed-cell.yaml @@ -27,7 +27,7 @@ properties: $ref: /schemas/types.yaml#/definitions/uint32-array items: - minimum: 0 - maximum: 7 + maximum: 31 description: Offset in bit within the address range specified by reg. - minimum: 1 -- 2.51.0 From 7a06ef75107799675ea6e4d73b9df37e18e352a8 Mon Sep 17 00:00:00 2001 From: Dmitry Baryshkov Date: Fri, 11 Apr 2025 12:22:46 +0100 Subject: [PATCH 16/16] nvmem: core: fix bit offsets of more than one byte If the NVMEM specifies a stride to access data, reading particular cell might require bit offset that is bigger than one byte. Rework NVMEM core code to support bit offsets of more than 8 bits. Signed-off-by: Dmitry Baryshkov Signed-off-by: Srinivas Kandagatla Link: https://lore.kernel.org/r/20250411112251.68002-9-srinivas.kandagatla@linaro.org Signed-off-by: Greg Kroah-Hartman --- drivers/nvmem/core.c | 24 +++++++++++++++++------- 1 file changed, 17 insertions(+), 7 deletions(-) diff --git a/drivers/nvmem/core.c b/drivers/nvmem/core.c index fff85bbf0ecd..7872903c08a1 100644 --- a/drivers/nvmem/core.c +++ b/drivers/nvmem/core.c @@ -837,7 +837,9 @@ static int nvmem_add_cells_from_dt(struct nvmem_device *nvmem, struct device_nod if (addr && len == (2 * sizeof(u32))) { info.bit_offset = be32_to_cpup(addr++); info.nbits = be32_to_cpup(addr); - if (info.bit_offset >= BITS_PER_BYTE || info.nbits < 1) { + if (info.bit_offset >= BITS_PER_BYTE * info.bytes || + info.nbits < 1 || + info.bit_offset + info.nbits > BITS_PER_BYTE * info.bytes) { dev_err(dev, "nvmem: invalid bits on %pOF\n", child); of_node_put(child); return -EINVAL; @@ -1630,21 +1632,29 @@ EXPORT_SYMBOL_GPL(nvmem_cell_put); static void nvmem_shift_read_buffer_in_place(struct nvmem_cell_entry *cell, void *buf) { u8 *p, *b; - int i, extra, bit_offset = cell->bit_offset; + int i, extra, bytes_offset; + int bit_offset = cell->bit_offset; p = b = buf; - if (bit_offset) { + + bytes_offset = bit_offset / BITS_PER_BYTE; + b += bytes_offset; + bit_offset %= BITS_PER_BYTE; + + if (bit_offset % BITS_PER_BYTE) { /* First shift */ - *b++ >>= bit_offset; + *p = *b++ >> bit_offset; /* setup rest of the bytes if any */ for (i = 1; i < cell->bytes; i++) { /* Get bits from next byte and shift them towards msb */ - *p |= *b << (BITS_PER_BYTE - bit_offset); + *p++ |= *b << (BITS_PER_BYTE - bit_offset); - p = b; - *b++ >>= bit_offset; + *p = *b++ >> bit_offset; } + } else if (p != b) { + memmove(p, b, cell->bytes - bytes_offset); + p += cell->bytes - 1; } else { /* point to the msb */ p += cell->bytes - 1; -- 2.51.0