From ba99a4fc8c24dc7d35f18edb6e3b0a65345fbfa3 Mon Sep 17 00:00:00 2001 From: Zhang Rui Date: Thu, 14 Nov 2024 15:59:43 +0800 Subject: [PATCH 01/16] tools/power turbostat: Remove unnecessary fflush() call The graphics sysfs knobs are read-only, making the use of fflush() before reading them redundant. Remove the unnecessary fflush() call. Signed-off-by: Zhang Rui Signed-off-by: Len Brown --- tools/power/x86/turbostat/turbostat.c | 7 +++---- 1 file changed, 3 insertions(+), 4 deletions(-) diff --git a/tools/power/x86/turbostat/turbostat.c b/tools/power/x86/turbostat/turbostat.c index ae841baeca85..c0596ccf92cd 100644 --- a/tools/power/x86/turbostat/turbostat.c +++ b/tools/power/x86/turbostat/turbostat.c @@ -5780,12 +5780,11 @@ int snapshot_graphics(int idx) case GFX_ACTMHz: case SAM_MHz: case SAM_ACTMHz: - if (gfx_info[idx].fp == NULL) { + if (gfx_info[idx].fp == NULL) gfx_info[idx].fp = fopen_or_die(gfx_info[idx].path, "r"); - } else { + else rewind(gfx_info[idx].fp); - fflush(gfx_info[idx].fp); - } + retval = fscanf(gfx_info[idx].fp, "%d", &gfx_info[idx].val); if (retval != 1) err(1, "MHz"); -- 2.50.1 From d071004e623b7433573019d67cba79e345d83006 Mon Sep 17 00:00:00 2001 From: Zhang Rui Date: Thu, 14 Nov 2024 15:59:44 +0800 Subject: [PATCH 02/16] tools/power turbostat: Consolidate graphics sysfs access Currently, there is an inconsistency in how graphics sysfs knobs are accessed: graphics residency sysfs knobs are opened and closed for each read, while graphics frequency sysfs knobs are opened once and remain open until turbostat exits. This inconsistency is confusing and adds unnecessary code complexity. Consolidate the access method by opening the sysfs files once and reusing the file pointers for subsequent accesses. This approach simplifies the code and ensures a consistent method for accessing graphics sysfs knobs. Signed-off-by: Zhang Rui Signed-off-by: Len Brown --- tools/power/x86/turbostat/turbostat.c | 15 ++++++--------- 1 file changed, 6 insertions(+), 9 deletions(-) diff --git a/tools/power/x86/turbostat/turbostat.c b/tools/power/x86/turbostat/turbostat.c index c0596ccf92cd..e5b100b8db24 100644 --- a/tools/power/x86/turbostat/turbostat.c +++ b/tools/power/x86/turbostat/turbostat.c @@ -5764,27 +5764,24 @@ int snapshot_proc_interrupts(void) */ int snapshot_graphics(int idx) { - FILE *fp; int retval; + if (gfx_info[idx].fp == NULL) + gfx_info[idx].fp = fopen_or_die(gfx_info[idx].path, "r"); + else + rewind(gfx_info[idx].fp); + switch (idx) { case GFX_rc6: case SAM_mc6: - fp = fopen_or_die(gfx_info[idx].path, "r"); - retval = fscanf(fp, "%lld", &gfx_info[idx].val_ull); + retval = fscanf(gfx_info[idx].fp, "%lld", &gfx_info[idx].val_ull); if (retval != 1) err(1, "rc6"); - fclose(fp); return 0; case GFX_MHz: case GFX_ACTMHz: case SAM_MHz: case SAM_ACTMHz: - if (gfx_info[idx].fp == NULL) - gfx_info[idx].fp = fopen_or_die(gfx_info[idx].path, "r"); - else - rewind(gfx_info[idx].fp); - retval = fscanf(gfx_info[idx].fp, "%d", &gfx_info[idx].val); if (retval != 1) err(1, "MHz"); -- 2.50.1 From c7538f33853b11d0ff2a81efb78bde125d1fc49f Mon Sep 17 00:00:00 2001 From: Zhang Rui Date: Thu, 14 Nov 2024 15:59:45 +0800 Subject: [PATCH 03/16] tools/power turbostat: Cache graphics sysfs file descriptors during probe Snapshots of the graphics sysfs knobs are taken based on file descriptors. To optimize this process, open the files and cache the file descriptors during the graphics probe phase. As a result, the previously cached pathnames become redundant and are removed. This change aims to streamline the code without altering its functionality. No functional change intended. Signed-off-by: Zhang Rui Signed-off-by: Len Brown --- tools/power/x86/turbostat/turbostat.c | 82 +++++++++++---------------- 1 file changed, 32 insertions(+), 50 deletions(-) diff --git a/tools/power/x86/turbostat/turbostat.c b/tools/power/x86/turbostat/turbostat.c index e5b100b8db24..28513172ffce 100644 --- a/tools/power/x86/turbostat/turbostat.c +++ b/tools/power/x86/turbostat/turbostat.c @@ -376,7 +376,6 @@ enum gfx_sysfs_idx { }; struct gfx_sysfs_info { - const char *path; FILE *fp; unsigned int val; unsigned long long val_ull; @@ -5766,10 +5765,7 @@ int snapshot_graphics(int idx) { int retval; - if (gfx_info[idx].fp == NULL) - gfx_info[idx].fp = fopen_or_die(gfx_info[idx].path, "r"); - else - rewind(gfx_info[idx].fp); + rewind(gfx_info[idx].fp); switch (idx) { case GFX_rc6: @@ -6474,6 +6470,12 @@ static void probe_intel_uncore_frequency(void) probe_intel_uncore_frequency_legacy(); } +static void set_graphics_fp(char *path, int idx) +{ + if (!access(path, R_OK)) + gfx_info[idx].fp = fopen_or_die(path, "r"); +} + static void probe_graphics(void) { /* Xe graphics sysfs knobs */ @@ -6481,7 +6483,6 @@ static void probe_graphics(void) FILE *fp; char buf[8]; bool gt0_is_gt; - int idx; fp = fopen("/sys/class/drm/card0/device/tile0/gt0/gtidle/name", "r"); if (!fp) @@ -6500,28 +6501,17 @@ static void probe_graphics(void) else goto next; - idx = gt0_is_gt ? GFX_rc6 : SAM_mc6; - gfx_info[idx].path = "/sys/class/drm/card0/device/tile0/gt0/gtidle/idle_residency_ms"; + set_graphics_fp("/sys/class/drm/card0/device/tile0/gt0/gtidle/idle_residency_ms", gt0_is_gt ? GFX_rc6 : SAM_mc6); - idx = gt0_is_gt ? GFX_MHz : SAM_MHz; - if (!access("/sys/class/drm/card0/device/tile0/gt0/freq0/cur_freq", R_OK)) - gfx_info[idx].path = "/sys/class/drm/card0/device/tile0/gt0/freq0/cur_freq"; + set_graphics_fp("/sys/class/drm/card0/device/tile0/gt0/freq0/cur_freq", gt0_is_gt ? GFX_MHz : SAM_MHz); - idx = gt0_is_gt ? GFX_ACTMHz : SAM_ACTMHz; - if (!access("/sys/class/drm/card0/device/tile0/gt0/freq0/act_freq", R_OK)) - gfx_info[idx].path = "/sys/class/drm/card0/device/tile0/gt0/freq0/act_freq"; + set_graphics_fp("/sys/class/drm/card0/device/tile0/gt0/freq0/act_freq", gt0_is_gt ? GFX_ACTMHz : SAM_ACTMHz); - idx = gt0_is_gt ? SAM_mc6 : GFX_rc6; - if (!access("/sys/class/drm/card0/device/tile0/gt1/gtidle/idle_residency_ms", R_OK)) - gfx_info[idx].path = "/sys/class/drm/card0/device/tile0/gt1/gtidle/idle_residency_ms"; + set_graphics_fp("/sys/class/drm/card0/device/tile0/gt1/gtidle/idle_residency_ms", gt0_is_gt ? SAM_mc6 : GFX_rc6); - idx = gt0_is_gt ? SAM_MHz : GFX_MHz; - if (!access("/sys/class/drm/card0/device/tile0/gt1/freq0/cur_freq", R_OK)) - gfx_info[idx].path = "/sys/class/drm/card0/device/tile0/gt1/freq0/cur_freq"; + set_graphics_fp("/sys/class/drm/card0/device/tile0/gt1/freq0/cur_freq", gt0_is_gt ? SAM_MHz : GFX_MHz); - idx = gt0_is_gt ? SAM_ACTMHz : GFX_ACTMHz; - if (!access("/sys/class/drm/card0/device/tile0/gt1/freq0/act_freq", R_OK)) - gfx_info[idx].path = "/sys/class/drm/card0/device/tile0/gt1/freq0/act_freq"; + set_graphics_fp("/sys/class/drm/card0/device/tile0/gt1/freq0/act_freq", gt0_is_gt ? SAM_ACTMHz : GFX_ACTMHz); goto end; } @@ -6529,52 +6519,44 @@ static void probe_graphics(void) next: /* New i915 graphics sysfs knobs */ if (!access("/sys/class/drm/card0/gt/gt0/rc6_residency_ms", R_OK)) { - gfx_info[GFX_rc6].path = "/sys/class/drm/card0/gt/gt0/rc6_residency_ms"; + set_graphics_fp("/sys/class/drm/card0/gt/gt0/rc6_residency_ms", GFX_rc6); - if (!access("/sys/class/drm/card0/gt/gt0/rps_cur_freq_mhz", R_OK)) - gfx_info[GFX_MHz].path = "/sys/class/drm/card0/gt/gt0/rps_cur_freq_mhz"; + set_graphics_fp("/sys/class/drm/card0/gt/gt0/rps_cur_freq_mhz", GFX_MHz); - if (!access("/sys/class/drm/card0/gt/gt0/rps_act_freq_mhz", R_OK)) - gfx_info[GFX_ACTMHz].path = "/sys/class/drm/card0/gt/gt0/rps_act_freq_mhz"; + set_graphics_fp("/sys/class/drm/card0/gt/gt0/rps_act_freq_mhz", GFX_ACTMHz); - if (!access("/sys/class/drm/card0/gt/gt1/rc6_residency_ms", R_OK)) - gfx_info[SAM_mc6].path = "/sys/class/drm/card0/gt/gt1/rc6_residency_ms"; + set_graphics_fp("/sys/class/drm/card0/gt/gt1/rc6_residency_ms", SAM_mc6); - if (!access("/sys/class/drm/card0/gt/gt1/rps_cur_freq_mhz", R_OK)) - gfx_info[SAM_MHz].path = "/sys/class/drm/card0/gt/gt1/rps_cur_freq_mhz"; + set_graphics_fp("/sys/class/drm/card0/gt/gt1/rps_cur_freq_mhz", SAM_MHz); - if (!access("/sys/class/drm/card0/gt/gt1/rps_act_freq_mhz", R_OK)) - gfx_info[SAM_ACTMHz].path = "/sys/class/drm/card0/gt/gt1/rps_act_freq_mhz"; + set_graphics_fp("/sys/class/drm/card0/gt/gt1/rps_act_freq_mhz", SAM_ACTMHz); goto end; } /* Fall back to traditional i915 graphics sysfs knobs */ - if (!access("/sys/class/drm/card0/power/rc6_residency_ms", R_OK)) - gfx_info[GFX_rc6].path = "/sys/class/drm/card0/power/rc6_residency_ms"; + set_graphics_fp("/sys/class/drm/card0/power/rc6_residency_ms", GFX_rc6); - if (!access("/sys/class/drm/card0/gt_cur_freq_mhz", R_OK)) - gfx_info[GFX_MHz].path = "/sys/class/drm/card0/gt_cur_freq_mhz"; - else if (!access("/sys/class/graphics/fb0/device/drm/card0/gt_cur_freq_mhz", R_OK)) - gfx_info[GFX_MHz].path = "/sys/class/graphics/fb0/device/drm/card0/gt_cur_freq_mhz"; + set_graphics_fp("/sys/class/drm/card0/gt_cur_freq_mhz", GFX_MHz); + if (!gfx_info[GFX_MHz].fp) + set_graphics_fp("/sys/class/graphics/fb0/device/drm/card0/gt_cur_freq_mhz", GFX_MHz); - if (!access("/sys/class/drm/card0/gt_act_freq_mhz", R_OK)) - gfx_info[GFX_ACTMHz].path = "/sys/class/drm/card0/gt_act_freq_mhz"; - else if (!access("/sys/class/graphics/fb0/device/drm/card0/gt_act_freq_mhz", R_OK)) - gfx_info[GFX_ACTMHz].path = "/sys/class/graphics/fb0/device/drm/card0/gt_act_freq_mhz"; + set_graphics_fp("/sys/class/drm/card0/gt_act_freq_mhz", GFX_ACTMHz); + if (!gfx_info[GFX_ACTMHz].fp) + set_graphics_fp("/sys/class/graphics/fb0/device/drm/card0/gt_act_freq_mhz", GFX_ACTMHz); end: - if (gfx_info[GFX_rc6].path) + if (gfx_info[GFX_rc6].fp) BIC_PRESENT(BIC_GFX_rc6); - if (gfx_info[GFX_MHz].path) + if (gfx_info[GFX_MHz].fp) BIC_PRESENT(BIC_GFXMHz); - if (gfx_info[GFX_ACTMHz].path) + if (gfx_info[GFX_ACTMHz].fp) BIC_PRESENT(BIC_GFXACTMHz); - if (gfx_info[SAM_mc6].path) + if (gfx_info[SAM_mc6].fp) BIC_PRESENT(BIC_SAM_mc6); - if (gfx_info[SAM_MHz].path) + if (gfx_info[SAM_MHz].fp) BIC_PRESENT(BIC_SAMMHz); - if (gfx_info[SAM_ACTMHz].path) + if (gfx_info[SAM_ACTMHz].fp) BIC_PRESENT(BIC_SAMACTMHz); } -- 2.50.1 From 03109e2f0d18dcb84218bd91c4fbf864193ca934 Mon Sep 17 00:00:00 2001 From: Zhang Rui Date: Thu, 14 Nov 2024 15:59:46 +0800 Subject: [PATCH 04/16] tools/power turbostat: Add support for /sys/class/drm/card1 On some machines, the graphics device is enumerated as /sys/class/drm/card1 instead of /sys/class/drm/card0. The current implementation does not handle this scenario, resulting in the loss of graphics C6 residency and frequency information. Add support for /sys/class/drm/card1, ensuring that turbostat can retrieve and display the graphics columns for these platforms. Signed-off-by: Zhang Rui Signed-off-by: Len Brown --- tools/power/x86/turbostat/turbostat.c | 38 ++++++++++++++++++++------- 1 file changed, 29 insertions(+), 9 deletions(-) diff --git a/tools/power/x86/turbostat/turbostat.c b/tools/power/x86/turbostat/turbostat.c index 28513172ffce..b250676c174e 100644 --- a/tools/power/x86/turbostat/turbostat.c +++ b/tools/power/x86/turbostat/turbostat.c @@ -6476,8 +6476,14 @@ static void set_graphics_fp(char *path, int idx) gfx_info[idx].fp = fopen_or_die(path, "r"); } +/* Enlarge this if there are /sys/class/drm/card2 ... */ +#define GFX_MAX_CARDS 2 + static void probe_graphics(void) { + char path[PATH_MAX]; + int i; + /* Xe graphics sysfs knobs */ if (!access("/sys/class/drm/card0/device/tile0/gt0/gtidle/idle_residency_ms", R_OK)) { FILE *fp; @@ -6518,22 +6524,36 @@ static void probe_graphics(void) next: /* New i915 graphics sysfs knobs */ - if (!access("/sys/class/drm/card0/gt/gt0/rc6_residency_ms", R_OK)) { - set_graphics_fp("/sys/class/drm/card0/gt/gt0/rc6_residency_ms", GFX_rc6); + for (i = 0; i < GFX_MAX_CARDS; i++) { + snprintf(path, PATH_MAX, "/sys/class/drm/card%d/gt/gt0/rc6_residency_ms", i); + if (!access(path, R_OK)) + break; + } - set_graphics_fp("/sys/class/drm/card0/gt/gt0/rps_cur_freq_mhz", GFX_MHz); + if (i == GFX_MAX_CARDS) + goto legacy_i915; - set_graphics_fp("/sys/class/drm/card0/gt/gt0/rps_act_freq_mhz", GFX_ACTMHz); + snprintf(path, PATH_MAX, "/sys/class/drm/card%d/gt/gt0/rc6_residency_ms", i); + set_graphics_fp(path, GFX_rc6); - set_graphics_fp("/sys/class/drm/card0/gt/gt1/rc6_residency_ms", SAM_mc6); + snprintf(path, PATH_MAX, "/sys/class/drm/card%d/gt/gt0/rps_cur_freq_mhz", i); + set_graphics_fp(path, GFX_MHz); - set_graphics_fp("/sys/class/drm/card0/gt/gt1/rps_cur_freq_mhz", SAM_MHz); + snprintf(path, PATH_MAX, "/sys/class/drm/card%d/gt/gt0/rps_act_freq_mhz", i); + set_graphics_fp(path, GFX_ACTMHz); - set_graphics_fp("/sys/class/drm/card0/gt/gt1/rps_act_freq_mhz", SAM_ACTMHz); + snprintf(path, PATH_MAX, "/sys/class/drm/card%d/gt/gt1/rc6_residency_ms", i); + set_graphics_fp(path, SAM_mc6); - goto end; - } + snprintf(path, PATH_MAX, "/sys/class/drm/card%d/gt/gt1/rps_cur_freq_mhz", i); + set_graphics_fp(path, SAM_MHz); + + snprintf(path, PATH_MAX, "/sys/class/drm/card%d/gt/gt1/rps_act_freq_mhz", i); + set_graphics_fp(path, SAM_ACTMHz); + + goto end; +legacy_i915: /* Fall back to traditional i915 graphics sysfs knobs */ set_graphics_fp("/sys/class/drm/card0/power/rc6_residency_ms", GFX_rc6); -- 2.50.1 From bcfab87108b33f20d847fd71a2a93114dd2ce83e Mon Sep 17 00:00:00 2001 From: Patryk Wlazlyn Date: Thu, 24 Oct 2024 15:17:45 +0200 Subject: [PATCH 05/16] tools/power turbostat: Force --no-perf in --dump mode Force the --no-perf early to prevent using it as a source. User asks for raw values, but perf returns them relative to the opening of the file descriptor. Signed-off-by: Patryk Wlazlyn Signed-off-by: Len Brown --- tools/power/x86/turbostat/turbostat.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/tools/power/x86/turbostat/turbostat.c b/tools/power/x86/turbostat/turbostat.c index b250676c174e..1fed799a5537 100644 --- a/tools/power/x86/turbostat/turbostat.c +++ b/tools/power/x86/turbostat/turbostat.c @@ -9897,6 +9897,12 @@ void cmdline(int argc, char **argv) break; case 'D': dump_only++; + /* + * Force the no_perf early to prevent using it as a source. + * User asks for raw values, but perf returns them relative + * to the opening of the file descriptor. + */ + no_perf = 1; break; case 'e': /* --enable specified counter */ -- 2.50.1 From 1da0daf746342dfdc114e4dc8fbf3ece28666d4f Mon Sep 17 00:00:00 2001 From: Patryk Wlazlyn Date: Wed, 13 Nov 2024 15:48:22 +0100 Subject: [PATCH 06/16] tools/power turbostat: Fix child's argument forwarding Add '+' to optstring when early scanning for --no-msr and --no-perf. It causes option processing to stop as soon as a nonoption argument is encountered, effectively skipping child's arguments. Fixes: 3e4048466c39 ("tools/power turbostat: Add --no-msr option") Signed-off-by: Patryk Wlazlyn 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 1fed799a5537..9025c2945737 100644 --- a/tools/power/x86/turbostat/turbostat.c +++ b/tools/power/x86/turbostat/turbostat.c @@ -9873,7 +9873,7 @@ void cmdline(int argc, char **argv) * Parse some options early, because they may make other options invalid, * like adding the MSR counter with --add and at the same time using --no-msr. */ - while ((opt = getopt_long_only(argc, argv, "MPn:", long_options, &option_index)) != -1) { + while ((opt = getopt_long_only(argc, argv, "+MPn:", long_options, &option_index)) != -1) { switch (opt) { case 'M': no_msr = 1; -- 2.50.1 From e5f687b89bc2a892256f48e14a568970b59a4812 Mon Sep 17 00:00:00 2001 From: Patryk Wlazlyn Date: Wed, 2 Oct 2024 15:05:15 +0200 Subject: [PATCH 07/16] tools/power turbostat: Add RAPL psys as a built-in counter Introduce the counter as a part of global, platform counters structure. We open the counter for only one cpu, but otherwise treat it as an ordinary RAPL counter, allowing for grouped perf read. The counter is disabled by default, because it's interpretation may require additional, platform specific information, making it unsuitable for general use. Signed-off-by: Patryk Wlazlyn Signed-off-by: Len Brown --- tools/power/x86/turbostat/turbostat.8 | 2 + tools/power/x86/turbostat/turbostat.c | 93 ++++++++++++++++++++++++--- 2 files changed, 85 insertions(+), 10 deletions(-) diff --git a/tools/power/x86/turbostat/turbostat.8 b/tools/power/x86/turbostat/turbostat.8 index 56c7ff6efcda..95eb02346d3a 100644 --- a/tools/power/x86/turbostat/turbostat.8 +++ b/tools/power/x86/turbostat/turbostat.8 @@ -190,6 +190,8 @@ The system configuration dump (if --quiet is not used) is followed by statistics .PP \fBRAMWatt\fP Watts consumed by the DRAM DIMMS -- available only on server processors. .PP +\fBSysWatt\fP Watts consumed by the whole platform (RAPL PSYS). Disabled by default. May require platform specific information to interpret the data, making it not suitable for general use. +.PP \fBPKG_%\fP percent of the interval that RAPL throttling was active on the Package. Note that the system summary is the sum of the package throttling time, and thus may be higher than 100% on a multi-package system. Note that the meaning of this field is model specific. For example, some hardware increments this counter when RAPL responds to thermal limits, but does not increment this counter when RAPL responds to power limits. Comparing PkgWatt and PkgTmp to system limits is necessary. .PP \fBRAM_%\fP percent of the interval that RAPL throttling was active on DRAM. diff --git a/tools/power/x86/turbostat/turbostat.c b/tools/power/x86/turbostat/turbostat.c index 9025c2945737..88c7f896c5b2 100644 --- a/tools/power/x86/turbostat/turbostat.c +++ b/tools/power/x86/turbostat/turbostat.c @@ -200,6 +200,8 @@ struct msr_counter bic[] = { { 0x0, "SAMMHz", NULL, 0, 0, 0, NULL, 0 }, { 0x0, "SAMAMHz", NULL, 0, 0, 0, NULL, 0 }, { 0x0, "Die%c6", NULL, 0, 0, 0, NULL, 0 }, + { 0x0, "SysWatt", NULL, 0, 0, 0, NULL, 0 }, + { 0x0, "Sys_J", NULL, 0, 0, 0, NULL, 0 }, }; #define MAX_BIC (sizeof(bic) / sizeof(struct msr_counter)) @@ -262,6 +264,8 @@ struct msr_counter bic[] = { #define BIC_SAMMHz (1ULL << 56) #define BIC_SAMACTMHz (1ULL << 57) #define BIC_Diec6 (1ULL << 58) +#define BIC_SysWatt (1ULL << 59) +#define BIC_Sys_J (1ULL << 60) #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__) @@ -269,7 +273,7 @@ struct msr_counter bic[] = { #define BIC_IDLE (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_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_SysWatt | BIC_Sys_J) 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; @@ -507,12 +511,15 @@ enum rapl_msrs { RAPL_AMD_PWR_UNIT = BIT(14), /* 0xc0010299 MSR_AMD_RAPL_POWER_UNIT */ RAPL_AMD_CORE_ENERGY_STAT = BIT(15), /* 0xc001029a MSR_AMD_CORE_ENERGY_STATUS */ RAPL_AMD_PKG_ENERGY_STAT = BIT(16), /* 0xc001029b MSR_AMD_PKG_ENERGY_STATUS */ + RAPL_PLATFORM_ENERGY_LIMIT = BIT(17), /* 0x64c MSR_PLATFORM_ENERGY_LIMIT */ + RAPL_PLATFORM_ENERGY_STATUS = BIT(18), /* 0x64d MSR_PLATFORM_ENERGY_STATUS */ }; #define RAPL_PKG (RAPL_PKG_ENERGY_STATUS | RAPL_PKG_POWER_LIMIT) #define RAPL_DRAM (RAPL_DRAM_ENERGY_STATUS | RAPL_DRAM_POWER_LIMIT) #define RAPL_CORE (RAPL_CORE_ENERGY_STATUS | RAPL_CORE_POWER_LIMIT) #define RAPL_GFX (RAPL_GFX_POWER_LIMIT | RAPL_GFX_ENERGY_STATUS) +#define RAPL_PSYS (RAPL_PLATFORM_ENERGY_STATUS | RAPL_PLATFORM_ENERGY_LIMIT) #define RAPL_PKG_ALL (RAPL_PKG | RAPL_PKG_PERF_STATUS | RAPL_PKG_POWER_INFO) #define RAPL_DRAM_ALL (RAPL_DRAM | RAPL_DRAM_PERF_STATUS | RAPL_DRAM_POWER_INFO) @@ -713,7 +720,7 @@ static const struct platform_features skl_features = { .has_ext_cst_msrs = 1, .trl_msrs = TRL_BASE, .tcc_offset_bits = 6, - .rapl_msrs = RAPL_PKG_ALL | RAPL_CORE_ALL | RAPL_DRAM | RAPL_DRAM_PERF_STATUS | RAPL_GFX, + .rapl_msrs = RAPL_PKG_ALL | RAPL_CORE_ALL | RAPL_DRAM | RAPL_DRAM_PERF_STATUS | RAPL_GFX | RAPL_PSYS, .enable_tsc_tweak = 1, }; @@ -730,7 +737,7 @@ static const struct platform_features cnl_features = { .has_ext_cst_msrs = 1, .trl_msrs = TRL_BASE, .tcc_offset_bits = 6, - .rapl_msrs = RAPL_PKG_ALL | RAPL_CORE_ALL | RAPL_DRAM | RAPL_DRAM_PERF_STATUS | RAPL_GFX, + .rapl_msrs = RAPL_PKG_ALL | RAPL_CORE_ALL | RAPL_DRAM | RAPL_DRAM_PERF_STATUS | RAPL_GFX | RAPL_PSYS, .enable_tsc_tweak = 1, }; @@ -797,7 +804,7 @@ static const struct platform_features icx_features = { .has_irtl_msrs = 1, .has_cst_prewake_bit = 1, .trl_msrs = TRL_BASE | TRL_CORECOUNT, - .rapl_msrs = RAPL_PKG_ALL | RAPL_DRAM_ALL, + .rapl_msrs = RAPL_PKG_ALL | RAPL_DRAM_ALL | RAPL_PSYS, .has_fixed_rapl_unit = 1, }; @@ -813,7 +820,7 @@ static const struct platform_features spr_features = { .has_irtl_msrs = 1, .has_cst_prewake_bit = 1, .trl_msrs = TRL_BASE | TRL_CORECOUNT, - .rapl_msrs = RAPL_PKG_ALL | RAPL_DRAM_ALL, + .rapl_msrs = RAPL_PKG_ALL | RAPL_DRAM_ALL | RAPL_PSYS, }; static const struct platform_features srf_features = { @@ -829,7 +836,7 @@ static const struct platform_features srf_features = { .has_irtl_msrs = 1, .has_cst_prewake_bit = 1, .trl_msrs = TRL_BASE | TRL_CORECOUNT, - .rapl_msrs = RAPL_PKG_ALL | RAPL_DRAM_ALL, + .rapl_msrs = RAPL_PKG_ALL | RAPL_DRAM_ALL | RAPL_PSYS, }; static const struct platform_features grr_features = { @@ -845,7 +852,7 @@ static const struct platform_features grr_features = { .has_irtl_msrs = 1, .has_cst_prewake_bit = 1, .trl_msrs = TRL_BASE | TRL_CORECOUNT, - .rapl_msrs = RAPL_PKG_ALL | RAPL_DRAM_ALL, + .rapl_msrs = RAPL_PKG_ALL | RAPL_DRAM_ALL | RAPL_PSYS, }; static const struct platform_features slv_features = { @@ -1108,6 +1115,7 @@ enum rapl_rci_index { RAPL_RCI_INDEX_PKG_PERF_STATUS = 4, RAPL_RCI_INDEX_DRAM_PERF_STATUS = 5, RAPL_RCI_INDEX_CORE_ENERGY = 6, + RAPL_RCI_INDEX_ENERGY_PLATFORM = 7, NUM_RAPL_COUNTERS, }; @@ -1134,6 +1142,7 @@ struct rapl_counter_info_t { struct rapl_counter_info_t *rapl_counter_info_perdomain; unsigned int rapl_counter_info_perdomain_size; +#define RAPL_COUNTER_FLAG_PLATFORM_COUNTER (1u << 0) #define RAPL_COUNTER_FLAG_USE_MSR_SUM (1u << 1) struct rapl_counter_arch_info { @@ -1255,6 +1264,19 @@ static const struct rapl_counter_arch_info rapl_counter_arch_infos[] = { .compat_scale = 1.0, .flags = 0, }, + { + .feature_mask = RAPL_PSYS, + .perf_subsys = "power", + .perf_name = "energy-psys", + .msr = MSR_PLATFORM_ENERGY_STATUS, + .msr_mask = 0x00000000FFFFFFFF, + .msr_shift = 0, + .platform_rapl_msr_scale = &rapl_energy_units, + .rci_index = RAPL_RCI_INDEX_ENERGY_PLATFORM, + .bic = BIC_SysWatt | BIC_Sys_J, + .compat_scale = 1.0, + .flags = RAPL_COUNTER_FLAG_PLATFORM_COUNTER | RAPL_COUNTER_FLAG_USE_MSR_SUM, + }, }; struct rapl_counter { @@ -1682,6 +1704,7 @@ enum { IDX_PP1_ENERGY, IDX_PKG_PERF, IDX_DRAM_PERF, + IDX_PSYS_ENERGY, IDX_COUNT, }; @@ -1726,6 +1749,9 @@ off_t idx_to_offset(int idx) case IDX_DRAM_PERF: offset = MSR_DRAM_PERF_STATUS; break; + case IDX_PSYS_ENERGY: + offset = MSR_PLATFORM_ENERGY_STATUS; + break; default: offset = -1; } @@ -1756,6 +1782,9 @@ int offset_to_idx(off_t offset) case MSR_DRAM_PERF_STATUS: idx = IDX_DRAM_PERF; break; + case MSR_PLATFORM_ENERGY_STATUS: + idx = IDX_PSYS_ENERGY; + break; default: idx = -1; } @@ -1777,6 +1806,8 @@ int idx_valid(int idx) return platform->rapl_msrs & RAPL_PKG_PERF_STATUS; case IDX_DRAM_PERF: return platform->rapl_msrs & RAPL_DRAM_PERF_STATUS; + case IDX_PSYS_ENERGY: + return platform->rapl_msrs & RAPL_PSYS; default: return 0; } @@ -1848,6 +1879,10 @@ struct system_summary { struct pkg_data packages; } average; +struct platform_counters { + struct rapl_counter energy_psys; /* MSR_PLATFORM_ENERGY_STATUS */ +} platform_counters_odd, platform_counters_even; + struct cpu_topology { int physical_package_id; int die_id; @@ -2512,6 +2547,11 @@ void print_header(char *delim) ppmt = ppmt->next; } + if (DO_BIC(BIC_SysWatt)) + outp += sprintf(outp, "%sSysWatt", (printed++ ? delim : "")); + if (DO_BIC(BIC_Sys_J)) + outp += sprintf(outp, "%sSys_J", (printed++ ? delim : "")); + outp += sprintf(outp, "\n"); } @@ -2519,6 +2559,7 @@ int dump_counters(struct thread_data *t, struct core_data *c, struct pkg_data *p { int i; struct msr_counter *mp; + struct platform_counters *pplat_cnt = p == package_odd ? &platform_counters_odd : &platform_counters_even; outp += sprintf(outp, "t %p, c %p, p %p\n", t, c, p); @@ -2590,6 +2631,7 @@ int dump_counters(struct thread_data *t, struct core_data *c, struct pkg_data *p outp += sprintf(outp, "Joules COR: %0llX\n", p->energy_cores.raw_value); outp += sprintf(outp, "Joules GFX: %0llX\n", p->energy_gfx.raw_value); outp += sprintf(outp, "Joules RAM: %0llX\n", p->energy_dram.raw_value); + outp += sprintf(outp, "Joules PSYS: %0llX\n", pplat_cnt->energy_psys.raw_value); outp += sprintf(outp, "Throttle PKG: %0llX\n", p->rapl_pkg_perf_status.raw_value); outp += sprintf(outp, "Throttle RAM: %0llX\n", p->rapl_dram_perf_status.raw_value); outp += sprintf(outp, "PTM: %dC\n", p->pkg_temp_c); @@ -2628,6 +2670,9 @@ double rapl_counter_get_value(const struct rapl_counter *c, enum rapl_unit desir */ int format_counters(struct thread_data *t, struct core_data *c, struct pkg_data *p) { + static int count; + + struct platform_counters *pplat_cnt = NULL; double interval_float, tsc; char *fmt8; int i; @@ -2637,6 +2682,11 @@ int format_counters(struct thread_data *t, struct core_data *c, struct pkg_data char *delim = "\t"; int printed = 0; + if (t == &average.threads) { + pplat_cnt = count & 1 ? &platform_counters_odd : &platform_counters_even; + ++count; + } + /* if showing only 1st thread in core and this isn't one, bail out */ if (show_core_only && !is_cpu_first_thread_in_core(t, c, p)) return 0; @@ -3093,6 +3143,13 @@ int format_counters(struct thread_data *t, struct core_data *c, struct pkg_data } } + if (DO_BIC(BIC_SysWatt) && (t == &average.threads)) + outp += sprintf(outp, fmt8, (printed++ ? delim : ""), + rapl_counter_get_value(&pplat_cnt->energy_psys, RAPL_UNIT_WATTS, interval_float)); + if (DO_BIC(BIC_Sys_J) && (t == &average.threads)) + outp += sprintf(outp, fmt8, (printed++ ? delim : ""), + rapl_counter_get_value(&pplat_cnt->energy_psys, RAPL_UNIT_JOULES, interval_float)); + done: if (*(outp - 1) != '\n') outp += sprintf(outp, "\n"); @@ -3400,6 +3457,11 @@ int delta_cpu(struct thread_data *t, struct core_data *c, return retval; } +void delta_platform(struct platform_counters *new, struct platform_counters *old) +{ + old->energy_psys.raw_value = new->energy_psys.raw_value - old->energy_psys.raw_value; +} + void rapl_counter_clear(struct rapl_counter *c) { c->raw_value = 0; @@ -4129,6 +4191,9 @@ static size_t cstate_counter_info_count_perf(const struct cstate_counter_info_t void write_rapl_counter(struct rapl_counter *rc, struct rapl_counter_info_t *rci, unsigned int idx) { + if (rci->source[idx] == COUNTER_SOURCE_NONE) + return; + rc->raw_value = rci->data[idx]; rc->unit = rci->unit[idx]; rc->scale = rci->scale[idx]; @@ -4136,6 +4201,7 @@ void write_rapl_counter(struct rapl_counter *rc, struct rapl_counter_info_t *rci int get_rapl_counters(int cpu, unsigned int domain, struct core_data *c, struct pkg_data *p) { + struct platform_counters *pplat_cnt = p == package_odd ? &platform_counters_odd : &platform_counters_even; unsigned long long perf_data[NUM_RAPL_COUNTERS + 1]; struct rapl_counter_info_t *rci; @@ -4163,6 +4229,7 @@ int get_rapl_counters(int cpu, unsigned int domain, struct core_data *c, struct for (unsigned int i = 0, pi = 1; i < NUM_RAPL_COUNTERS; ++i) { switch (rci->source[i]) { case COUNTER_SOURCE_NONE: + rci->data[i] = 0; break; case COUNTER_SOURCE_PERF: @@ -4201,7 +4268,7 @@ int get_rapl_counters(int cpu, unsigned int domain, struct core_data *c, struct } } - BUILD_BUG_ON(NUM_RAPL_COUNTERS != 7); + BUILD_BUG_ON(NUM_RAPL_COUNTERS != 8); write_rapl_counter(&p->energy_pkg, rci, RAPL_RCI_INDEX_ENERGY_PKG); write_rapl_counter(&p->energy_cores, rci, RAPL_RCI_INDEX_ENERGY_CORES); write_rapl_counter(&p->energy_dram, rci, RAPL_RCI_INDEX_DRAM); @@ -4209,6 +4276,7 @@ int get_rapl_counters(int cpu, unsigned int domain, struct core_data *c, struct write_rapl_counter(&p->rapl_pkg_perf_status, rci, RAPL_RCI_INDEX_PKG_PERF_STATUS); write_rapl_counter(&p->rapl_dram_perf_status, rci, RAPL_RCI_INDEX_DRAM_PERF_STATUS); write_rapl_counter(&c->core_energy, rci, RAPL_RCI_INDEX_CORE_ENERGY); + write_rapl_counter(&pplat_cnt->energy_psys, rci, RAPL_RCI_INDEX_ENERGY_PLATFORM); return 0; } @@ -6144,6 +6212,7 @@ restart: re_initialize(); goto restart; } + delta_platform(&platform_counters_odd, &platform_counters_even); compute_average(EVEN_COUNTERS); format_all_counters(EVEN_COUNTERS); flush_output_stdout(); @@ -6167,6 +6236,7 @@ restart: re_initialize(); goto restart; } + delta_platform(&platform_counters_even, &platform_counters_odd); compute_average(ODD_COUNTERS); format_all_counters(ODD_COUNTERS); flush_output_stdout(); @@ -6945,8 +7015,8 @@ void rapl_probe_intel(void) unsigned long long msr; unsigned int time_unit; double tdp; - const unsigned long long bic_watt_bits = BIC_PkgWatt | BIC_CorWatt | BIC_RAMWatt | BIC_GFXWatt; - const unsigned long long bic_joules_bits = BIC_Pkg_J | BIC_Cor_J | BIC_RAM_J | BIC_GFX_J; + const unsigned long long bic_watt_bits = BIC_SysWatt | BIC_PkgWatt | BIC_CorWatt | BIC_RAMWatt | BIC_GFXWatt; + const unsigned long long bic_joules_bits = BIC_Sys_J | BIC_Pkg_J | BIC_Cor_J | BIC_RAM_J | BIC_GFX_J; if (rapl_joules) bic_enabled &= ~bic_watt_bits; @@ -7606,6 +7676,9 @@ void rapl_perf_init(void) domain_visited[next_domain] = 1; + if ((cai->flags & RAPL_COUNTER_FLAG_PLATFORM_COUNTER) && (cpu != base_cpu)) + continue; + struct rapl_counter_info_t *rci = &rapl_counter_info_perdomain[next_domain]; /* Check if the counter is enabled and accessible */ -- 2.50.1 From 86d237734091201d2ab2c1d2e1063893621c770f Mon Sep 17 00:00:00 2001 From: Len Brown Date: Sat, 30 Nov 2024 16:22:00 -0500 Subject: [PATCH 08/16] tools/power turbostat: 2024.11.30 since 2024.07.26: assorted minor bug fixes assorted platform specific tweaks initial RAPL PSYS (SysWatt) support Signed-off-by: Len Brown --- tools/power/x86/turbostat/turbostat.8 | 2 +- tools/power/x86/turbostat/turbostat.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/tools/power/x86/turbostat/turbostat.8 b/tools/power/x86/turbostat/turbostat.8 index 95eb02346d3a..a7f7ed01421c 100644 --- a/tools/power/x86/turbostat/turbostat.8 +++ b/tools/power/x86/turbostat/turbostat.8 @@ -190,7 +190,7 @@ The system configuration dump (if --quiet is not used) is followed by statistics .PP \fBRAMWatt\fP Watts consumed by the DRAM DIMMS -- available only on server processors. .PP -\fBSysWatt\fP Watts consumed by the whole platform (RAPL PSYS). Disabled by default. May require platform specific information to interpret the data, making it not suitable for general use. +\fBSysWatt\fP Watts consumed by the whole platform (RAPL PSYS). Disabled by default. Enable with --enable SysWatt. .PP \fBPKG_%\fP percent of the interval that RAPL throttling was active on the Package. Note that the system summary is the sum of the package throttling time, and thus may be higher than 100% on a multi-package system. Note that the meaning of this field is model specific. For example, some hardware increments this counter when RAPL responds to thermal limits, but does not increment this counter when RAPL responds to power limits. Comparing PkgWatt and PkgTmp to system limits is necessary. .PP diff --git a/tools/power/x86/turbostat/turbostat.c b/tools/power/x86/turbostat/turbostat.c index 88c7f896c5b2..58a487c225a7 100644 --- a/tools/power/x86/turbostat/turbostat.c +++ b/tools/power/x86/turbostat/turbostat.c @@ -9236,7 +9236,7 @@ int get_and_dump_counters(void) void print_version() { - fprintf(outf, "turbostat version 2024.07.26 - Len Brown \n"); + fprintf(outf, "turbostat version 2024.11.30 - Len Brown \n"); } #define COMMAND_LINE_SIZE 2048 -- 2.50.1 From f69e63756f7822fcdad8a34f9967e8b243e883ee Mon Sep 17 00:00:00 2001 From: "Dr. David Alan Gilbert" Date: Wed, 2 Oct 2024 18:31:47 +0100 Subject: [PATCH 09/16] printf: Remove unused 'bprintf' bprintf() is unused. Remove it. It was added in the commit 4370aa4aa753 ("vsprintf: add binary printf") but as far as I can see was never used, unlike the other two functions in that patch. Link: https://lore.kernel.org/20241002173147.210107-1-linux@treblig.org Reviewed-by: Andy Shevchenko Acked-by: Petr Mladek Signed-off-by: Dr. David Alan Gilbert Signed-off-by: Steven Rostedt (Google) --- include/linux/string.h | 1 - lib/vsprintf.c | 23 ----------------------- 2 files changed, 24 deletions(-) diff --git a/include/linux/string.h b/include/linux/string.h index 0dd27afcfaf7..493ac4862c77 100644 --- a/include/linux/string.h +++ b/include/linux/string.h @@ -335,7 +335,6 @@ int __sysfs_match_string(const char * const *array, size_t n, const char *s); #ifdef CONFIG_BINARY_PRINTF int vbin_printf(u32 *bin_buf, size_t size, const char *fmt, va_list args); int bstr_printf(char *buf, size_t size, const char *fmt, const u32 *bin_buf); -int bprintf(u32 *bin_buf, size_t size, const char *fmt, ...) __printf(3, 4); #endif extern ssize_t memory_read_from_buffer(void *to, size_t count, loff_t *ppos, diff --git a/lib/vsprintf.c b/lib/vsprintf.c index 6ac02bbb7df1..9d3dac38a3f4 100644 --- a/lib/vsprintf.c +++ b/lib/vsprintf.c @@ -3428,29 +3428,6 @@ out: } EXPORT_SYMBOL_GPL(bstr_printf); -/** - * bprintf - Parse a format string and place args' binary value in a buffer - * @bin_buf: The buffer to place args' binary value - * @size: The size of the buffer(by words(32bits), not characters) - * @fmt: The format string to use - * @...: Arguments for the format string - * - * The function returns the number of words(u32) written - * into @bin_buf. - */ -int bprintf(u32 *bin_buf, size_t size, const char *fmt, ...) -{ - va_list args; - int ret; - - va_start(args, fmt); - ret = vbin_printf(bin_buf, size, fmt, args); - va_end(args); - - return ret; -} -EXPORT_SYMBOL_GPL(bprintf); - #endif /* CONFIG_BINARY_PRINTF */ /** -- 2.50.1 From 9022ed0e7e65734d83a0648648589b9fbea8e8c9 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Sun, 1 Dec 2024 09:23:33 -0800 Subject: [PATCH 10/16] strscpy: write destination buffer only once The point behind strscpy() was to once and for all avoid all the problems with 'strncpy()' and later broken "fixed" versions like strlcpy() that just made things worse. So strscpy not only guarantees NUL-termination (unlike strncpy), it also doesn't do unnecessary padding at the destination. But at the same time also avoids byte-at-a-time reads and writes by _allowing_ some extra NUL writes - within the size, of course - so that the whole copy can be done with word operations. It is also stable in the face of a mutable source string: it explicitly does not read the source buffer multiple times (so an implementation using "strnlen()+memcpy()" would be wrong), and does not read the source buffer past the size (like the mis-design that is strlcpy does). Finally, the return value is designed to be simple and unambiguous: if the string cannot be copied fully, it returns an actual negative error, making error handling clearer and simpler (and the caller already knows the size of the buffer). Otherwise it returns the string length of the result. However, there was one final stability issue that can be important to callers: the stability of the destination buffer. In particular, the same way we shouldn't read the source buffer more than once, we should avoid doing multiple writes to the destination buffer: first writing a potentially non-terminated string, and then terminating it with NUL at the end does not result in a stable result buffer. Yes, it gives the right result in the end, but if the rule for the destination buffer was that it is _always_ NUL-terminated even when accessed concurrently with updates, the final byte of the buffer needs to always _stay_ as a NUL byte. [ Note that "final byte is NUL" here is literally about the final byte in the destination array, not the terminating NUL at the end of the string itself. There is no attempt to try to make concurrent reads and writes give any kind of consistent string length or contents, but we do want to guarantee that there is always at least that final terminating NUL character at the end of the destination array if it existed before ] This is relevant in the kernel for the tsk->comm[] array, for example. Even without locking (for either readers or writers), we want to know that while the buffer contents may be garbled, it is always a valid C string and always has a NUL character at 'comm[TASK_COMM_LEN-1]' (and never has any "out of thin air" data). So avoid any "copy possibly non-terminated string, and terminate later" behavior, and write the destination buffer only once. Signed-off-by: Linus Torvalds --- lib/string.c | 23 +++++++++++++++++------ 1 file changed, 17 insertions(+), 6 deletions(-) diff --git a/lib/string.c b/lib/string.c index 76327b51e36f..eb4486ed40d2 100644 --- a/lib/string.c +++ b/lib/string.c @@ -104,6 +104,12 @@ char *strncpy(char *dest, const char *src, size_t count) EXPORT_SYMBOL(strncpy); #endif +#ifdef __BIG_ENDIAN +# define ALLBUTLAST_BYTE_MASK (~255ul) +#else +# define ALLBUTLAST_BYTE_MASK (~0ul >> 8) +#endif + ssize_t sized_strscpy(char *dest, const char *src, size_t count) { const struct word_at_a_time constants = WORD_AT_A_TIME_CONSTANTS; @@ -147,13 +153,18 @@ ssize_t sized_strscpy(char *dest, const char *src, size_t count) *(unsigned long *)(dest+res) = c & zero_bytemask(data); return res + find_zero(data); } + count -= sizeof(unsigned long); + if (unlikely(!count)) { + c &= ALLBUTLAST_BYTE_MASK; + *(unsigned long *)(dest+res) = c; + return -E2BIG; + } *(unsigned long *)(dest+res) = c; res += sizeof(unsigned long); - count -= sizeof(unsigned long); max -= sizeof(unsigned long); } - while (count) { + while (count > 1) { char c; c = src[res]; @@ -164,11 +175,11 @@ ssize_t sized_strscpy(char *dest, const char *src, size_t count) count--; } - /* Hit buffer length without finding a NUL; force NUL-termination. */ - if (res) - dest[res-1] = '\0'; + /* Force NUL-termination. */ + dest[res] = '\0'; - return -E2BIG; + /* Return E2BIG if the source didn't stop */ + return src[res] ? -E2BIG : res; } EXPORT_SYMBOL(sized_strscpy); -- 2.50.1 From 40384c840ea1944d7c5a392e8975ed088ecf0b37 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Sun, 1 Dec 2024 14:28:56 -0800 Subject: [PATCH 11/16] Linux 6.13-rc1 --- Makefile | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/Makefile b/Makefile index e34a97473fb6..93ab62cef244 100644 --- a/Makefile +++ b/Makefile @@ -1,8 +1,8 @@ # SPDX-License-Identifier: GPL-2.0 VERSION = 6 -PATCHLEVEL = 12 +PATCHLEVEL = 13 SUBLEVEL = 0 -EXTRAVERSION = +EXTRAVERSION = -rc1 NAME = Baby Opossum Posse # *DOCUMENTATION* -- 2.50.1 From 5daececd4ff533ab316ab360aba0bda1bf01961d Mon Sep 17 00:00:00 2001 From: "Borislav Petkov (AMD)" Date: Sat, 30 Nov 2024 13:26:44 +0100 Subject: [PATCH 12/16] x86/boot/compressed: Remove unused header includes from kaslr.c Nothing is using the linux/ namespace headers anymore. Remove them. Signed-off-by: Borislav Petkov (AMD) Link: https://lore.kernel.org/r/20241130122644.GAZ0sEhD3Bm_9ZAIuc@fat_crate.local --- arch/x86/boot/compressed/kaslr.c | 4 ---- 1 file changed, 4 deletions(-) diff --git a/arch/x86/boot/compressed/kaslr.c b/arch/x86/boot/compressed/kaslr.c index f4d82379bf44..f03d59ea6e40 100644 --- a/arch/x86/boot/compressed/kaslr.c +++ b/arch/x86/boot/compressed/kaslr.c @@ -25,10 +25,6 @@ #include "efi.h" #include -#include -#include -#include -#include #include #include -- 2.50.1 From 095ac6fa19500fecd7c62e755dee45bb303d4d43 Mon Sep 17 00:00:00 2001 From: Baoquan He Date: Sat, 23 Nov 2024 19:42:19 +0800 Subject: [PATCH 13/16] x86/ioremap: Simplify setup_data mapping variants memremap_is_setup_data() and early_memremap_is_setup_data() share completely the same process and handling, except for the differing memremap/unmap invocations. Add a helper __memremap_is_setup_data() extracting the common part and simplify a lot of code while at it. Mark __memremap_is_setup_data() as __ref to suppress this section mismatch warning: WARNING: modpost: vmlinux: section mismatch in reference: __memremap_is_setup_data+0x5f (section: .text) -> early_memunmap (section: .init.text) [ bp: Massage a bit. ] Signed-off-by: Baoquan He Signed-off-by: Borislav Petkov (AMD) Link: https://lore.kernel.org/r/20241123114221.149383-2-bhe@redhat.com --- arch/x86/mm/ioremap.c | 106 ++++++++++++++---------------------------- 1 file changed, 35 insertions(+), 71 deletions(-) diff --git a/arch/x86/mm/ioremap.c b/arch/x86/mm/ioremap.c index 8d29163568a7..fe44e8180bdd 100644 --- a/arch/x86/mm/ioremap.c +++ b/arch/x86/mm/ioremap.c @@ -632,42 +632,54 @@ static bool memremap_is_efi_data(resource_size_t phys_addr, * Examine the physical address to determine if it is boot data by checking * it against the boot params setup_data chain. */ -static bool memremap_is_setup_data(resource_size_t phys_addr, - unsigned long size) +static bool __ref __memremap_is_setup_data(resource_size_t phys_addr, bool early) { + unsigned int setup_data_sz = sizeof(struct setup_data); struct setup_indirect *indirect; struct setup_data *data; u64 paddr, paddr_next; paddr = boot_params.hdr.setup_data; while (paddr) { - unsigned int len; + unsigned int len, size; if (phys_addr == paddr) return true; - data = memremap(paddr, sizeof(*data), - MEMREMAP_WB | MEMREMAP_DEC); + if (early) + data = early_memremap_decrypted(paddr, setup_data_sz); + else + data = memremap(paddr, setup_data_sz, MEMREMAP_WB | MEMREMAP_DEC); if (!data) { - pr_warn("failed to memremap setup_data entry\n"); + pr_warn("failed to remap setup_data entry\n"); return false; } + size = setup_data_sz; + paddr_next = data->next; len = data->len; if ((phys_addr > paddr) && - (phys_addr < (paddr + sizeof(struct setup_data) + len))) { - memunmap(data); + (phys_addr < (paddr + setup_data_sz + len))) { + if (early) + early_memunmap(data, setup_data_sz); + else + memunmap(data); return true; } if (data->type == SETUP_INDIRECT) { - memunmap(data); - data = memremap(paddr, sizeof(*data) + len, - MEMREMAP_WB | MEMREMAP_DEC); + size += len; + if (early) { + early_memunmap(data, setup_data_sz); + data = early_memremap_decrypted(paddr, size); + } else { + memunmap(data); + data = memremap(paddr, size, MEMREMAP_WB | MEMREMAP_DEC); + } if (!data) { - pr_warn("failed to memremap indirect setup_data\n"); + pr_warn("failed to remap indirect setup_data\n"); return false; } @@ -679,7 +691,10 @@ static bool memremap_is_setup_data(resource_size_t phys_addr, } } - memunmap(data); + if (early) + early_memunmap(data, size); + else + memunmap(data); if ((phys_addr > paddr) && (phys_addr < (paddr + len))) return true; @@ -690,67 +705,16 @@ static bool memremap_is_setup_data(resource_size_t phys_addr, return false; } -/* - * Examine the physical address to determine if it is boot data by checking - * it against the boot params setup_data chain (early boot version). - */ +static bool memremap_is_setup_data(resource_size_t phys_addr, + unsigned long size) +{ + return __memremap_is_setup_data(phys_addr, false); +} + static bool __init early_memremap_is_setup_data(resource_size_t phys_addr, unsigned long size) { - struct setup_indirect *indirect; - struct setup_data *data; - u64 paddr, paddr_next; - - paddr = boot_params.hdr.setup_data; - while (paddr) { - unsigned int len, size; - - if (phys_addr == paddr) - return true; - - data = early_memremap_decrypted(paddr, sizeof(*data)); - if (!data) { - pr_warn("failed to early memremap setup_data entry\n"); - return false; - } - - size = sizeof(*data); - - paddr_next = data->next; - len = data->len; - - if ((phys_addr > paddr) && - (phys_addr < (paddr + sizeof(struct setup_data) + len))) { - early_memunmap(data, sizeof(*data)); - return true; - } - - if (data->type == SETUP_INDIRECT) { - size += len; - early_memunmap(data, sizeof(*data)); - data = early_memremap_decrypted(paddr, size); - if (!data) { - pr_warn("failed to early memremap indirect setup_data\n"); - return false; - } - - indirect = (struct setup_indirect *)data->data; - - if (indirect->type != SETUP_INDIRECT) { - paddr = indirect->addr; - len = indirect->len; - } - } - - early_memunmap(data, size); - - if ((phys_addr > paddr) && (phys_addr < (paddr + len))) - return true; - - paddr = paddr_next; - } - - return false; + return __memremap_is_setup_data(phys_addr, true); } /* -- 2.50.1 From 525077ae7145cc868b69282f85bed2be8ecd1ed5 Mon Sep 17 00:00:00 2001 From: Baoquan He Date: Sat, 23 Nov 2024 19:42:21 +0800 Subject: [PATCH 14/16] x86/ioremap: Remove unused size parameter in remapping functions The size parameter of functions memremap_is_efi_data(), memremap_is_setup_data() and early_memremap_is_setup_data() is not used. Remove it. [ bp: Massage commit message. ] Signed-off-by: Baoquan He Signed-off-by: Borislav Petkov (AMD) Link: https://lore.kernel.org/r/20241123114221.149383-4-bhe@redhat.com --- arch/x86/mm/ioremap.c | 17 +++++++---------- 1 file changed, 7 insertions(+), 10 deletions(-) diff --git a/arch/x86/mm/ioremap.c b/arch/x86/mm/ioremap.c index fe44e8180bdd..38ff7791a9c7 100644 --- a/arch/x86/mm/ioremap.c +++ b/arch/x86/mm/ioremap.c @@ -593,8 +593,7 @@ static bool memremap_should_map_decrypted(resource_size_t phys_addr, * Examine the physical address to determine if it is EFI data. Check * it against the boot params structure and EFI tables and memory types. */ -static bool memremap_is_efi_data(resource_size_t phys_addr, - unsigned long size) +static bool memremap_is_efi_data(resource_size_t phys_addr) { u64 paddr; @@ -705,14 +704,12 @@ static bool __ref __memremap_is_setup_data(resource_size_t phys_addr, bool early return false; } -static bool memremap_is_setup_data(resource_size_t phys_addr, - unsigned long size) +static bool memremap_is_setup_data(resource_size_t phys_addr) { return __memremap_is_setup_data(phys_addr, false); } -static bool __init early_memremap_is_setup_data(resource_size_t phys_addr, - unsigned long size) +static bool __init early_memremap_is_setup_data(resource_size_t phys_addr) { return __memremap_is_setup_data(phys_addr, true); } @@ -735,8 +732,8 @@ bool arch_memremap_can_ram_remap(resource_size_t phys_addr, unsigned long size, return false; if (cc_platform_has(CC_ATTR_HOST_MEM_ENCRYPT)) { - if (memremap_is_setup_data(phys_addr, size) || - memremap_is_efi_data(phys_addr, size)) + if (memremap_is_setup_data(phys_addr) || + memremap_is_efi_data(phys_addr)) return false; } @@ -761,8 +758,8 @@ pgprot_t __init early_memremap_pgprot_adjust(resource_size_t phys_addr, encrypted_prot = true; if (cc_platform_has(CC_ATTR_HOST_MEM_ENCRYPT)) { - if (early_memremap_is_setup_data(phys_addr, size) || - memremap_is_efi_data(phys_addr, size)) + if (early_memremap_is_setup_data(phys_addr) || + memremap_is_efi_data(phys_addr)) encrypted_prot = false; } -- 2.50.1 From ab0e7f20768af59fe161d71cc5d1de384f2a9da8 Mon Sep 17 00:00:00 2001 From: "Borislav Petkov (AMD)" Date: Mon, 2 Dec 2024 20:00:10 +0100 Subject: [PATCH 15/16] Documentation: Merge x86-specific boot options doc into kernel-parameters.txt Documentation/arch/x86/x86_64/boot-options.rst is causing unnecessary confusion by being a second place where one can put x86 boot options. Move them into the main one. Drop removed ones like "acpi=ht", while at it. Signed-off-by: Borislav Petkov (AMD) Reviewed-by: Sohil Mehta Link: https://lore.kernel.org/r/20241202190011.11979-1-bp@kernel.org --- .../admin-guide/kernel-parameters.rst | 3 - .../admin-guide/kernel-parameters.txt | 237 ++++++++++++- .../arch/x86/x86_64/boot-options.rst | 312 ------------------ .../arch/x86/x86_64/fake-numa-for-cpusets.rst | 2 +- Documentation/arch/x86/x86_64/index.rst | 1 - arch/x86/Kconfig.debug | 2 +- arch/x86/kernel/pci-dma.c | 4 - 7 files changed, 227 insertions(+), 334 deletions(-) delete mode 100644 Documentation/arch/x86/x86_64/boot-options.rst diff --git a/Documentation/admin-guide/kernel-parameters.rst b/Documentation/admin-guide/kernel-parameters.rst index 59931f21c974..39d0e7ff0965 100644 --- a/Documentation/admin-guide/kernel-parameters.rst +++ b/Documentation/admin-guide/kernel-parameters.rst @@ -194,8 +194,6 @@ is applicable:: WDT Watchdog support is enabled. X86-32 X86-32, aka i386 architecture is enabled. X86-64 X86-64 architecture is enabled. - More X86-64 boot options can be found in - Documentation/arch/x86/x86_64/boot-options.rst. X86 Either 32-bit or 64-bit x86 (same as X86-32+X86-64) X86_UV SGI UV support is enabled. XEN Xen support is enabled @@ -213,7 +211,6 @@ Do not modify the syntax of boot loader parameters without extreme need or coordination with . There are also arch-specific kernel-parameters not documented here. -See for example . Note that ALL kernel parameters listed below are CASE SENSITIVE, and that a trailing = on the name of any parameter states that that parameter will diff --git a/Documentation/admin-guide/kernel-parameters.txt b/Documentation/admin-guide/kernel-parameters.txt index dc663c0ca670..bf7b3568d1a5 100644 --- a/Documentation/admin-guide/kernel-parameters.txt +++ b/Documentation/admin-guide/kernel-parameters.txt @@ -21,6 +21,10 @@ strictly ACPI specification compliant. rsdt -- prefer RSDT over (default) XSDT copy_dsdt -- copy DSDT to memory + nocmcff -- Disable firmware first mode for corrected + errors. This disables parsing the HEST CMC error + source to check if firmware has set the FF flag. This + may result in duplicate corrected error reports. nospcr -- disable console in ACPI SPCR table as default _serial_ console on ARM64 For ARM64, ONLY "acpi=off", "acpi=on", "acpi=force" or @@ -405,6 +409,8 @@ not play well with APC CPU idle - disable it if you have APC and your system crashes randomly. + apic [APIC,X86-64] Use IO-APIC. Default. + apic= [APIC,X86,EARLY] Advanced Programmable Interrupt Controller Change the output verbosity while booting Format: { quiet (default) | verbose | debug } @@ -424,6 +430,10 @@ useful so that a dump capture kernel won't be shot down by NMI + apicpmtimer Do APIC timer calibration using the pmtimer. Implies + apicmaintimer. Useful when your PIT timer is totally + broken. + autoconf= [IPV6] See Documentation/networking/ipv6.rst. @@ -1726,6 +1736,8 @@ off: Disable GDS mitigation. + gbpages [X86] Use GB pages for kernel direct mappings. + gcov_persist= [GCOV] When non-zero (default), profiling data for kernel modules is saved and remains accessible via debugfs, even when the module is unloaded/reloaded. @@ -2008,12 +2020,21 @@ idle= [X86,EARLY] Format: idle=poll, idle=halt, idle=nomwait - Poll forces a polling idle loop that can slightly - improve the performance of waking up a idle CPU, but - will use a lot of power and make the system run hot. - Not recommended. + + idle=poll: Don't do power saving in the idle loop + using HLT, but poll for rescheduling event. This will + make the CPUs eat a lot more power, but may be useful + to get slightly better performance in multiprocessor + benchmarks. It also makes some profiling using + performance counters more accurate. Please note that + on systems with MONITOR/MWAIT support (like Intel + EM64T CPUs) this option has no performance advantage + over the normal idle loop. It may also interact badly + with hyperthreading. + idle=halt: Halt is forced to be used for CPU idle. In such case C2/C3 won't be used again. + idle=nomwait: Disable mwait for CPU C-states idxd.sva= [HW] @@ -2311,20 +2332,73 @@ relaxed iommu= [X86,EARLY] + off + Don't initialize and use any kind of IOMMU. + force + Force the use of the hardware IOMMU even when + it is not actually needed (e.g. because < 3 GB + memory). + noforce + Don't force hardware IOMMU usage when it is not + needed. (default). + biomerge panic nopanic merge nomerge + soft - pt [X86] - nopt [X86] - nobypass [PPC/POWERNV] + Use software bounce buffering (SWIOTLB) (default for + Intel machines). This can be used to prevent the usage + of an available hardware IOMMU. + + [X86] + pt + [X86] + nopt + [PPC/POWERNV] + nobypass Disable IOMMU bypass, using IOMMU for PCI devices. + [X86] + AMD Gart HW IOMMU-specific options: + + + Set the size of the remapping area in bytes. + + allowed + Overwrite iommu off workarounds for specific chipsets + + fullflush + Flush IOMMU on each allocation (default). + + nofullflush + Don't use IOMMU fullflush. + + memaper[=] + Allocate an own aperture over RAM with size + 32MB< series board detected. + Selecting for reboots." In the case where you + think the quirk is in error (e.g. you have newer BIOS, + or newer board) using this option will ignore the + built-in quirk table, and use the generic default + reboot actions. + + efi + Use efi reset_system runtime service. If EFI is not + configured or the EFI reset does not work, the reboot + path attempts the reset using the keyboard controller. + + force + Don't stop other CPUs on reboot. This can make reboot + more reliable in some cases. + + kbd + Use the keyboard controller. cold reset (default) + + pci + Use a write to the PCI config space register 0xcf9 to + trigger reboot. + + triple + Force a triple fault (init) + + warm + Don't set the cold reboot flag + + Using warm reset will be much faster especially on big + memory systems because the BIOS will not go through + the memory check. Disadvantage is that not all + hardware will be completely reinitialized on reboot so + there may be boot problems on some systems. + + refscale.holdoff= [KNL] Set test-start holdoff period. The purpose of this parameter is to delay the start of the @@ -6101,7 +6305,16 @@ serialnumber [BUGS=X86-32] - sev=option[,option...] [X86-64] See Documentation/arch/x86/x86_64/boot-options.rst + sev=option[,option...] [X86-64] + + debug + Enable debug messages. + + nosnp + Do not enable SEV-SNP (applies to host/hypervisor + only). Setting 'nosnp' avoids the RMP check overhead + in memory accesses when users do not want to run + SEV-SNP guests. shapers= [NET] Maximal number of shapers. diff --git a/Documentation/arch/x86/x86_64/boot-options.rst b/Documentation/arch/x86/x86_64/boot-options.rst deleted file mode 100644 index d69e3cfbdba5..000000000000 --- a/Documentation/arch/x86/x86_64/boot-options.rst +++ /dev/null @@ -1,312 +0,0 @@ -.. SPDX-License-Identifier: GPL-2.0 - -=========================== -AMD64 Specific Boot Options -=========================== - -There are many others (usually documented in driver documentation), but -only the AMD64 specific ones are listed here. - -Machine check -============= -Please see Documentation/arch/x86/x86_64/machinecheck.rst for sysfs runtime tunables. - - mce=off - Disable machine check - mce=no_cmci - Disable CMCI(Corrected Machine Check Interrupt) that - Intel processor supports. Usually this disablement is - not recommended, but it might be handy if your hardware - is misbehaving. - Note that you'll get more problems without CMCI than with - due to the shared banks, i.e. you might get duplicated - error logs. - mce=dont_log_ce - Don't make logs for corrected errors. All events reported - as corrected are silently cleared by OS. - This option will be useful if you have no interest in any - of corrected errors. - mce=ignore_ce - Disable features for corrected errors, e.g. polling timer - and CMCI. All events reported as corrected are not cleared - by OS and remained in its error banks. - Usually this disablement is not recommended, however if - there is an agent checking/clearing corrected errors - (e.g. BIOS or hardware monitoring applications), conflicting - with OS's error handling, and you cannot deactivate the agent, - then this option will be a help. - mce=no_lmce - Do not opt-in to Local MCE delivery. Use legacy method - to broadcast MCEs. - mce=bootlog - Enable logging of machine checks left over from booting. - Disabled by default on AMD Fam10h and older because some BIOS - leave bogus ones. - If your BIOS doesn't do that it's a good idea to enable though - to make sure you log even machine check events that result - in a reboot. On Intel systems it is enabled by default. - mce=nobootlog - Disable boot machine check logging. - mce=monarchtimeout (number) - monarchtimeout: - Sets the time in us to wait for other CPUs on machine checks. 0 - to disable. - mce=bios_cmci_threshold - Don't overwrite the bios-set CMCI threshold. This boot option - prevents Linux from overwriting the CMCI threshold set by the - bios. Without this option, Linux always sets the CMCI - threshold to 1. Enabling this may make memory predictive failure - analysis less effective if the bios sets thresholds for memory - errors since we will not see details for all errors. - mce=recovery - Force-enable recoverable machine check code paths - - nomce (for compatibility with i386) - same as mce=off - - Everything else is in sysfs now. - -APICs -===== - - apic - Use IO-APIC. Default - - noapic - Don't use the IO-APIC. - - disableapic - Don't use the local APIC - - nolapic - Don't use the local APIC (alias for i386 compatibility) - - pirq=... - See Documentation/arch/x86/i386/IO-APIC.rst - - noapictimer - Don't set up the APIC timer - - no_timer_check - Don't check the IO-APIC timer. This can work around - problems with incorrect timer initialization on some boards. - - apicpmtimer - Do APIC timer calibration using the pmtimer. Implies - apicmaintimer. Useful when your PIT timer is totally broken. - -Timing -====== - - notsc - Deprecated, use tsc=unstable instead. - - nohpet - Don't use the HPET timer. - -Idle loop -========= - - idle=poll - Don't do power saving in the idle loop using HLT, but poll for rescheduling - event. This will make the CPUs eat a lot more power, but may be useful - to get slightly better performance in multiprocessor benchmarks. It also - makes some profiling using performance counters more accurate. - Please note that on systems with MONITOR/MWAIT support (like Intel EM64T - CPUs) this option has no performance advantage over the normal idle loop. - It may also interact badly with hyperthreading. - -Rebooting -========= - - reboot=b[ios] | t[riple] | k[bd] | a[cpi] | e[fi] | p[ci] [, [w]arm | [c]old] - bios - Use the CPU reboot vector for warm reset - warm - Don't set the cold reboot flag - cold - Set the cold reboot flag - triple - Force a triple fault (init) - kbd - Use the keyboard controller. cold reset (default) - acpi - Use the ACPI RESET_REG in the FADT. If ACPI is not configured or - the ACPI reset does not work, the reboot path attempts the reset - using the keyboard controller. - efi - Use efi reset_system runtime service. If EFI is not configured or - the EFI reset does not work, the reboot path attempts the reset using - the keyboard controller. - pci - Use a write to the PCI config space register 0xcf9 to trigger reboot. - - Using warm reset will be much faster especially on big memory - systems because the BIOS will not go through the memory check. - Disadvantage is that not all hardware will be completely reinitialized - on reboot so there may be boot problems on some systems. - - reboot=force - Don't stop other CPUs on reboot. This can make reboot more reliable - in some cases. - - reboot=default - There are some built-in platform specific "quirks" - you may see: - "reboot: series board detected. Selecting for reboots." - In the case where you think the quirk is in error (e.g. you have - newer BIOS, or newer board) using this option will ignore the built-in - quirk table, and use the generic default reboot actions. - -NUMA -==== - - numa=off - Only set up a single NUMA node spanning all memory. - - numa=noacpi - Don't parse the SRAT table for NUMA setup - - numa=nohmat - Don't parse the HMAT table for NUMA setup, or soft-reserved memory - partitioning. - -ACPI -==== - - acpi=off - Don't enable ACPI - acpi=ht - Use ACPI boot table parsing, but don't enable ACPI interpreter - acpi=force - Force ACPI on (currently not needed) - acpi=strict - Disable out of spec ACPI workarounds. - acpi_sci={edge,level,high,low} - Set up ACPI SCI interrupt. - acpi=noirq - Don't route interrupts - acpi=nocmcff - Disable firmware first mode for corrected errors. This - disables parsing the HEST CMC error source to check if - firmware has set the FF flag. This may result in - duplicate corrected error reports. - -PCI -=== - - pci=off - Don't use PCI - pci=conf1 - Use conf1 access. - pci=conf2 - Use conf2 access. - pci=rom - Assign ROMs. - pci=assign-busses - Assign busses - pci=irqmask=MASK - Set PCI interrupt mask to MASK - pci=lastbus=NUMBER - Scan up to NUMBER busses, no matter what the mptable says. - pci=noacpi - Don't use ACPI to set up PCI interrupt routing. - -IOMMU (input/output memory management unit) -=========================================== -Multiple x86-64 PCI-DMA mapping implementations exist, for example: - - 1. : use no hardware/software IOMMU at all - (e.g. because you have < 3 GB memory). - Kernel boot message: "PCI-DMA: Disabling IOMMU" - - 2. : AMD GART based hardware IOMMU. - Kernel boot message: "PCI-DMA: using GART IOMMU" - - 3. : Software IOMMU implementation. Used - e.g. if there is no hardware IOMMU in the system and it is need because - you have >3GB memory or told the kernel to us it (iommu=soft)) - Kernel boot message: "PCI-DMA: Using software bounce buffering - for IO (SWIOTLB)" - -:: - - iommu=[][,noagp][,off][,force][,noforce] - [,memaper[=]][,merge][,fullflush][,nomerge] - [,noaperture] - -General iommu options: - - off - Don't initialize and use any kind of IOMMU. - noforce - Don't force hardware IOMMU usage when it is not needed. (default). - force - Force the use of the hardware IOMMU even when it is - not actually needed (e.g. because < 3 GB memory). - soft - Use software bounce buffering (SWIOTLB) (default for - Intel machines). This can be used to prevent the usage - of an available hardware IOMMU. - -iommu options only relevant to the AMD GART hardware IOMMU: - - - Set the size of the remapping area in bytes. - allowed - Overwrite iommu off workarounds for specific chipsets. - fullflush - Flush IOMMU on each allocation (default). - nofullflush - Don't use IOMMU fullflush. - memaper[=] - Allocate an own aperture over RAM with size 32MB<[,force,noforce] - - Prereserve that many 2K slots for the software IO bounce buffering. - force - Force all IO through the software TLB. - noforce - Do not initialize the software TLB. - - -Miscellaneous -============= - - nogbpages - Do not use GB pages for kernel direct mappings. - gbpages - Use GB pages for kernel direct mappings. - - -AMD SEV (Secure Encrypted Virtualization) -========================================= -Options relating to AMD SEV, specified via the following format: - -:: - - sev=option1[,option2] - -The available options are: - - debug - Enable debug messages. - - nosnp - Do not enable SEV-SNP (applies to host/hypervisor only). Setting - 'nosnp' avoids the RMP check overhead in memory accesses when - users do not want to run SEV-SNP guests. diff --git a/Documentation/arch/x86/x86_64/fake-numa-for-cpusets.rst b/Documentation/arch/x86/x86_64/fake-numa-for-cpusets.rst index ba74617d4999..970ee94eb551 100644 --- a/Documentation/arch/x86/x86_64/fake-numa-for-cpusets.rst +++ b/Documentation/arch/x86/x86_64/fake-numa-for-cpusets.rst @@ -18,7 +18,7 @@ For more information on the features of cpusets, see Documentation/admin-guide/cgroup-v1/cpusets.rst. There are a number of different configurations you can use for your needs. For more information on the numa=fake command line option and its various ways of -configuring fake nodes, see Documentation/arch/x86/x86_64/boot-options.rst. +configuring fake nodes, see Documentation/admin-guide/kernel-parameters.txt For the purposes of this introduction, we'll assume a very primitive NUMA emulation setup of "numa=fake=4*512,". This will split our system memory into diff --git a/Documentation/arch/x86/x86_64/index.rst b/Documentation/arch/x86/x86_64/index.rst index ad15e9bd623f..a0261957a08a 100644 --- a/Documentation/arch/x86/x86_64/index.rst +++ b/Documentation/arch/x86/x86_64/index.rst @@ -7,7 +7,6 @@ x86_64 Support .. toctree:: :maxdepth: 2 - boot-options uefi mm 5level-paging diff --git a/arch/x86/Kconfig.debug b/arch/x86/Kconfig.debug index 74777a97e394..1eb4d23cdaae 100644 --- a/arch/x86/Kconfig.debug +++ b/arch/x86/Kconfig.debug @@ -97,7 +97,7 @@ config IOMMU_DEBUG code. When you use it make sure you have a big enough IOMMU/AGP aperture. Most of the options enabled by this can be set more finegrained using the iommu= command line - options. See Documentation/arch/x86/x86_64/boot-options.rst for more + options. See Documentation/admin-guide/kernel-parameters.txt for more details. config IOMMU_LEAK diff --git a/arch/x86/kernel/pci-dma.c b/arch/x86/kernel/pci-dma.c index f323d83e40a7..6267363e0189 100644 --- a/arch/x86/kernel/pci-dma.c +++ b/arch/x86/kernel/pci-dma.c @@ -108,10 +108,6 @@ void __init pci_iommu_alloc(void) swiotlb_init(x86_swiotlb_enable, x86_swiotlb_flags); } -/* - * See for the iommu kernel - * parameter documentation. - */ static __init int iommu_setup(char *p) { iommu_merge = 1; -- 2.50.1 From 13148e22c151e871c1c00bab519f39cc6f6ea37a Mon Sep 17 00:00:00 2001 From: "Borislav Petkov (AMD)" Date: Mon, 2 Dec 2024 20:00:11 +0100 Subject: [PATCH 16/16] x86/apic: Remove "disablelapic" cmdline option The convention is "no" and there already is "nolapic". Drop the disable one. Signed-off-by: Borislav Petkov (AMD) Reviewed-by: Sohil Mehta Link: https://lore.kernel.org/r/20241202190011.11979-2-bp@kernel.org --- arch/x86/kernel/apic/apic.c | 9 +-------- arch/x86/kernel/cpu/topology.c | 2 +- 2 files changed, 2 insertions(+), 9 deletions(-) diff --git a/arch/x86/kernel/apic/apic.c b/arch/x86/kernel/apic/apic.c index c5fb28e6451a..1267b2691abc 100644 --- a/arch/x86/kernel/apic/apic.c +++ b/arch/x86/kernel/apic/apic.c @@ -2582,19 +2582,12 @@ int apic_is_clustered_box(void) /* * APIC command line parameters */ -static int __init setup_disableapic(char *arg) +static int __init setup_nolapic(char *arg) { apic_is_disabled = true; setup_clear_cpu_cap(X86_FEATURE_APIC); return 0; } -early_param("disableapic", setup_disableapic); - -/* same as disableapic, for compatibility */ -static int __init setup_nolapic(char *arg) -{ - return setup_disableapic(arg); -} early_param("nolapic", setup_nolapic); static int __init parse_lapic_timer_c2_ok(char *arg) diff --git a/arch/x86/kernel/cpu/topology.c b/arch/x86/kernel/cpu/topology.c index 621a151ccf7d..6ebed852a9ef 100644 --- a/arch/x86/kernel/cpu/topology.c +++ b/arch/x86/kernel/cpu/topology.c @@ -428,7 +428,7 @@ void __init topology_apply_cmdline_limits_early(void) { unsigned int possible = nr_cpu_ids; - /* 'maxcpus=0' 'nosmp' 'nolapic' 'disableapic' 'noapic' */ + /* 'maxcpus=0' 'nosmp' 'nolapic' 'noapic' */ if (!setup_max_cpus || ioapic_is_disabled || apic_is_disabled) possible = 1; -- 2.50.1