linux/arch/x86/events/intel/uncore.h

551 lines
16 KiB
C
Raw Normal View History

License cleanup: add SPDX GPL-2.0 license identifier to files with no license Many source files in the tree are missing licensing information, which makes it harder for compliance tools to determine the correct license. By default all files without license information are under the default license of the kernel, which is GPL version 2. Update the files which contain no license information with the 'GPL-2.0' SPDX license identifier. The SPDX identifier is a legally binding shorthand, which can be used instead of the full boiler plate text. This patch is based on work done by Thomas Gleixner and Kate Stewart and Philippe Ombredanne. How this work was done: Patches were generated and checked against linux-4.14-rc6 for a subset of the use cases: - file had no licensing information it it. - file was a */uapi/* one with no licensing information in it, - file was a */uapi/* one with existing licensing information, Further patches will be generated in subsequent months to fix up cases where non-standard license headers were used, and references to license had to be inferred by heuristics based on keywords. The analysis to determine which SPDX License Identifier to be applied to a file was done in a spreadsheet of side by side results from of the output of two independent scanners (ScanCode & Windriver) producing SPDX tag:value files created by Philippe Ombredanne. Philippe prepared the base worksheet, and did an initial spot review of a few 1000 files. The 4.13 kernel was the starting point of the analysis with 60,537 files assessed. Kate Stewart did a file by file comparison of the scanner results in the spreadsheet to determine which SPDX license identifier(s) to be applied to the file. She confirmed any determination that was not immediately clear with lawyers working with the Linux Foundation. Criteria used to select files for SPDX license identifier tagging was: - Files considered eligible had to be source code files. - Make and config files were included as candidates if they contained >5 lines of source - File already had some variant of a license header in it (even if <5 lines). All documentation files were explicitly excluded. The following heuristics were used to determine which SPDX license identifiers to apply. - when both scanners couldn't find any license traces, file was considered to have no license information in it, and the top level COPYING file license applied. For non */uapi/* files that summary was: SPDX license identifier # files ---------------------------------------------------|------- GPL-2.0 11139 and resulted in the first patch in this series. If that file was a */uapi/* path one, it was "GPL-2.0 WITH Linux-syscall-note" otherwise it was "GPL-2.0". Results of that was: SPDX license identifier # files ---------------------------------------------------|------- GPL-2.0 WITH Linux-syscall-note 930 and resulted in the second patch in this series. - if a file had some form of licensing information in it, and was one of the */uapi/* ones, it was denoted with the Linux-syscall-note if any GPL family license was found in the file or had no licensing in it (per prior point). Results summary: SPDX license identifier # files ---------------------------------------------------|------ GPL-2.0 WITH Linux-syscall-note 270 GPL-2.0+ WITH Linux-syscall-note 169 ((GPL-2.0 WITH Linux-syscall-note) OR BSD-2-Clause) 21 ((GPL-2.0 WITH Linux-syscall-note) OR BSD-3-Clause) 17 LGPL-2.1+ WITH Linux-syscall-note 15 GPL-1.0+ WITH Linux-syscall-note 14 ((GPL-2.0+ WITH Linux-syscall-note) OR BSD-3-Clause) 5 LGPL-2.0+ WITH Linux-syscall-note 4 LGPL-2.1 WITH Linux-syscall-note 3 ((GPL-2.0 WITH Linux-syscall-note) OR MIT) 3 ((GPL-2.0 WITH Linux-syscall-note) AND MIT) 1 and that resulted in the third patch in this series. - when the two scanners agreed on the detected license(s), that became the concluded license(s). - when there was disagreement between the two scanners (one detected a license but the other didn't, or they both detected different licenses) a manual inspection of the file occurred. - In most cases a manual inspection of the information in the file resulted in a clear resolution of the license that should apply (and which scanner probably needed to revisit its heuristics). - When it was not immediately clear, the license identifier was confirmed with lawyers working with the Linux Foundation. - If there was any question as to the appropriate license identifier, the file was flagged for further research and to be revisited later in time. In total, over 70 hours of logged manual review was done on the spreadsheet to determine the SPDX license identifiers to apply to the source files by Kate, Philippe, Thomas and, in some cases, confirmation by lawyers working with the Linux Foundation. Kate also obtained a third independent scan of the 4.13 code base from FOSSology, and compared selected files where the other two scanners disagreed against that SPDX file, to see if there was new insights. The Windriver scanner is based on an older version of FOSSology in part, so they are related. Thomas did random spot checks in about 500 files from the spreadsheets for the uapi headers and agreed with SPDX license identifier in the files he inspected. For the non-uapi files Thomas did random spot checks in about 15000 files. In initial set of patches against 4.14-rc6, 3 files were found to have copy/paste license identifier errors, and have been fixed to reflect the correct identifier. Additionally Philippe spent 10 hours this week doing a detailed manual inspection and review of the 12,461 patched files from the initial patch version early this week with: - a full scancode scan run, collecting the matched texts, detected license ids and scores - reviewing anything where there was a license detected (about 500+ files) to ensure that the applied SPDX license was correct - reviewing anything where there was no detection but the patch license was not GPL-2.0 WITH Linux-syscall-note to ensure that the applied SPDX license was correct This produced a worksheet with 20 files needing minor correction. This worksheet was then exported into 3 different .csv files for the different types of files to be modified. These .csv files were then reviewed by Greg. Thomas wrote a script to parse the csv files and add the proper SPDX tag to the file, in the format that the file expected. This script was further refined by Greg based on the output to detect more types of files automatically and to distinguish between header and source .c files (which need different comment types.) Finally Greg ran the script using the .csv files to generate the patches. Reviewed-by: Kate Stewart <kstewart@linuxfoundation.org> Reviewed-by: Philippe Ombredanne <pombredanne@nexb.com> Reviewed-by: Thomas Gleixner <tglx@linutronix.de> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2017-11-01 22:07:57 +08:00
/* SPDX-License-Identifier: GPL-2.0 */
#include <linux/slab.h>
#include <linux/pci.h>
#include <asm/apicdef.h>
#include <linux/io-64-nonatomic-lo-hi.h>
#include <linux/perf_event.h>
#include "../perf_event.h"
#define UNCORE_PMU_NAME_LEN 32
#define UNCORE_PMU_HRTIMER_INTERVAL (60LL * NSEC_PER_SEC)
#define UNCORE_SNB_IMC_HRTIMER_INTERVAL (5ULL * NSEC_PER_SEC)
#define UNCORE_FIXED_EVENT 0xff
#define UNCORE_PMC_IDX_MAX_GENERIC 8
perf/x86/intel/uncore: Add new data structures for free running counters There are a number of free running counters introduced for uncore, which provide highly valuable information to a wide array of customers. For example, Skylake Server has IIO free running counters to collect Input/Output x BW/Utilization. There is NO event available on the general purpose counters, that is exactly the same as the free running counters. The generic uncore code needs to be enhanced to support the new counters. In the uncore document, there is no event-code assigned to free running counters. Some events need to be defined to indicate the free running counters. The events are encoded as event-code + umask-code. The event-code for all free running counters is 0xff, which is the same as the fixed counters: - It has not been decided what code will be used for common events on future platforms. 0xff is the only one which will definitely not be used as any common event-code. - Cannot re-use current events on the general purpose counters. Because there is NO event available, that is exactly the same as the free running counters. - Even in the existing codes, the fixed counters for core, that have the same event-code, may count different things. Hence, it should not surprise the users if the free running counters that share the same event-code also count different things. Umask will be used to distinguish the counters. The umask-code is used to distinguish a fixed counter and a free running counter, and different types of free running counters. For fixed counters, the umask-code is 0x0X, where X indicates the index of the fixed counter, which starts from 0. - Compatible with the old event encoding. - Currently, there is only one fixed counter. There are still 15 reserved spaces for extension. For free running counters, the umask-code uses the rest of the space. It would follow the format of 0xXY: - X stands for the type of free running counters, which starts from 1. - Y stands for the index of free running counters of same type, which starts from 0. - The free running counters do different thing. It can be categorized to several types, according to the MSR location, bit width and definition. E.g. there are three types of IIO free running counters on Skylake server to monitor IO CLOCKS, BANDWIDTH and UTILIZATION on different ports. It makes it easy to locate the free running counter of a specific type. - So far, there are at most 8 counters of each type. There are still 8 reserved spaces for extension. Introducing a new index to indicate the free running counters. Only one index is enough for all free running counters. Because the free running counters are always active, and the event and free running counter are always 1:1 mapped, it does not need extra index to indicate the assigned counter. Introducing a new data structure to store free running counters related information for each type. It includes the number of counters, bit width, base address, offset between counters and offset between boxes. Introducing several inline helpers to check index for fixed counter and free running counter, validate free running counter event, and retrieve the free running counter information according to box and event. Signed-off-by: Kan Liang <kan.liang@intel.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Reviewed-by: Thomas Gleixner <tglx@linutronix.de> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: acme@kernel.org Cc: eranian@google.com Link: http://lkml.kernel.org/r/1525371913-10597-4-git-send-email-kan.liang@intel.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2018-05-04 02:25:09 +08:00
#define UNCORE_PMC_IDX_MAX_FIXED 1
#define UNCORE_PMC_IDX_MAX_FREERUNNING 1
#define UNCORE_PMC_IDX_FIXED UNCORE_PMC_IDX_MAX_GENERIC
perf/x86/intel/uncore: Add new data structures for free running counters There are a number of free running counters introduced for uncore, which provide highly valuable information to a wide array of customers. For example, Skylake Server has IIO free running counters to collect Input/Output x BW/Utilization. There is NO event available on the general purpose counters, that is exactly the same as the free running counters. The generic uncore code needs to be enhanced to support the new counters. In the uncore document, there is no event-code assigned to free running counters. Some events need to be defined to indicate the free running counters. The events are encoded as event-code + umask-code. The event-code for all free running counters is 0xff, which is the same as the fixed counters: - It has not been decided what code will be used for common events on future platforms. 0xff is the only one which will definitely not be used as any common event-code. - Cannot re-use current events on the general purpose counters. Because there is NO event available, that is exactly the same as the free running counters. - Even in the existing codes, the fixed counters for core, that have the same event-code, may count different things. Hence, it should not surprise the users if the free running counters that share the same event-code also count different things. Umask will be used to distinguish the counters. The umask-code is used to distinguish a fixed counter and a free running counter, and different types of free running counters. For fixed counters, the umask-code is 0x0X, where X indicates the index of the fixed counter, which starts from 0. - Compatible with the old event encoding. - Currently, there is only one fixed counter. There are still 15 reserved spaces for extension. For free running counters, the umask-code uses the rest of the space. It would follow the format of 0xXY: - X stands for the type of free running counters, which starts from 1. - Y stands for the index of free running counters of same type, which starts from 0. - The free running counters do different thing. It can be categorized to several types, according to the MSR location, bit width and definition. E.g. there are three types of IIO free running counters on Skylake server to monitor IO CLOCKS, BANDWIDTH and UTILIZATION on different ports. It makes it easy to locate the free running counter of a specific type. - So far, there are at most 8 counters of each type. There are still 8 reserved spaces for extension. Introducing a new index to indicate the free running counters. Only one index is enough for all free running counters. Because the free running counters are always active, and the event and free running counter are always 1:1 mapped, it does not need extra index to indicate the assigned counter. Introducing a new data structure to store free running counters related information for each type. It includes the number of counters, bit width, base address, offset between counters and offset between boxes. Introducing several inline helpers to check index for fixed counter and free running counter, validate free running counter event, and retrieve the free running counter information according to box and event. Signed-off-by: Kan Liang <kan.liang@intel.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Reviewed-by: Thomas Gleixner <tglx@linutronix.de> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: acme@kernel.org Cc: eranian@google.com Link: http://lkml.kernel.org/r/1525371913-10597-4-git-send-email-kan.liang@intel.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2018-05-04 02:25:09 +08:00
#define UNCORE_PMC_IDX_FREERUNNING (UNCORE_PMC_IDX_FIXED + \
UNCORE_PMC_IDX_MAX_FIXED)
#define UNCORE_PMC_IDX_MAX (UNCORE_PMC_IDX_FREERUNNING + \
UNCORE_PMC_IDX_MAX_FREERUNNING)
perf/x86/intel/uncore: Locate specific box by checking full device info Some platforms, e.g. Knights Landing, use a common PCI device ID for multiple instances of an uncore PMU device type. So it is impossible to locate the specific instances only by PCI device ID. The current code specially handles Knights Landing by arbitrarily pointing an instance to an unused uncore box. However, we still have no idea which uncore device is mapped to which box. Furthermore, there could be more platforms which use a common PCI device ID for uncore devices. We have to specially handle them one by one. This patch records full device information (slot, func, and device ID) in id_table[]. So the probe function can point the instance to a specific uncore box by checking the full device information. Tested-by: Lukasz Odzioba <lukasz.odzioba@intel.com> Signed-off-by: Kan Liang <kan.liang@intel.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Acked-by: tglx@linutronix.de Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stephane Eranian <eranian@google.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Vince Weaver <vincent.weaver@maine.edu> Cc: bp@suse.de Cc: harish.chegondi@intel.com Cc: hubert.chrzaniuk@intel.com Cc: lawrence.f.meadows@intel.com Link: http://lkml.kernel.org/r/1463379504-39003-1-git-send-email-kan.liang@intel.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2016-05-16 14:18:24 +08:00
#define UNCORE_PCI_DEV_FULL_DATA(dev, func, type, idx) \
((dev << 24) | (func << 16) | (type << 8) | idx)
#define UNCORE_PCI_DEV_DATA(type, idx) ((type << 8) | idx)
perf/x86/intel/uncore: Locate specific box by checking full device info Some platforms, e.g. Knights Landing, use a common PCI device ID for multiple instances of an uncore PMU device type. So it is impossible to locate the specific instances only by PCI device ID. The current code specially handles Knights Landing by arbitrarily pointing an instance to an unused uncore box. However, we still have no idea which uncore device is mapped to which box. Furthermore, there could be more platforms which use a common PCI device ID for uncore devices. We have to specially handle them one by one. This patch records full device information (slot, func, and device ID) in id_table[]. So the probe function can point the instance to a specific uncore box by checking the full device information. Tested-by: Lukasz Odzioba <lukasz.odzioba@intel.com> Signed-off-by: Kan Liang <kan.liang@intel.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Acked-by: tglx@linutronix.de Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stephane Eranian <eranian@google.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Vince Weaver <vincent.weaver@maine.edu> Cc: bp@suse.de Cc: harish.chegondi@intel.com Cc: hubert.chrzaniuk@intel.com Cc: lawrence.f.meadows@intel.com Link: http://lkml.kernel.org/r/1463379504-39003-1-git-send-email-kan.liang@intel.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2016-05-16 14:18:24 +08:00
#define UNCORE_PCI_DEV_DEV(data) ((data >> 24) & 0xff)
#define UNCORE_PCI_DEV_FUNC(data) ((data >> 16) & 0xff)
#define UNCORE_PCI_DEV_TYPE(data) ((data >> 8) & 0xff)
#define UNCORE_PCI_DEV_IDX(data) (data & 0xff)
#define UNCORE_EXTRA_PCI_DEV 0xff
perf/x86/intel/uncore: Fix hardcoded index of Broadwell extra PCI devices Masayoshi Mizuma reported that a warning message is shown while a CPU is hot-removed on Broadwell servers: WARNING: CPU: 126 PID: 6 at arch/x86/events/intel/uncore.c:988 uncore_pci_remove+0x10b/0x150 Call Trace: pci_device_remove+0x42/0xd0 device_release_driver_internal+0x148/0x220 pci_stop_bus_device+0x76/0xa0 pci_stop_root_bus+0x44/0x60 acpi_pci_root_remove+0x1f/0x80 acpi_bus_trim+0x57/0x90 acpi_bus_trim+0x2e/0x90 acpi_device_hotplug+0x2bc/0x4b0 acpi_hotplug_work_fn+0x1a/0x30 process_one_work+0x174/0x3a0 worker_thread+0x4c/0x3d0 kthread+0xf8/0x130 This bug was introduced by: commit 15a3e845b01c ("perf/x86/intel/uncore: Fix SBOX support for Broadwell CPUs") The index of "QPI Port 2 filter" was hardcode to 2, but this conflicts with the index of "PCU.3" which is "HSWEP_PCI_PCU_3", which equals to 2 as well. To fix the conflict, the hardcoded index needs to be cleaned up: - introduce a new enumerator "BDX_PCI_QPI_PORT2_FILTER" for "QPI Port 2 filter" on Broadwell, - increase UNCORE_EXTRA_PCI_DEV_MAX by one, - clean up the hardcoded index. Debugged-by: Masayoshi Mizuma <m.mizuma@jp.fujitsu.com> Suggested-by: Ingo Molnar <mingo@kernel.org> Reported-by: Masayoshi Mizuma <m.mizuma@jp.fujitsu.com> Tested-by: Masayoshi Mizuma <m.mizuma@jp.fujitsu.com> Signed-off-by: Kan Liang <kan.liang@linux.intel.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stephane Eranian <eranian@google.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Vince Weaver <vincent.weaver@maine.edu> Cc: msys.mizuma@gmail.com Cc: stable@vger.kernel.org Fixes: 15a3e845b01c ("perf/x86/intel/uncore: Fix SBOX support for Broadwell CPUs") Link: http://lkml.kernel.org/r/1532953688-15008-1-git-send-email-kan.liang@linux.intel.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2018-07-30 20:28:08 +08:00
#define UNCORE_EXTRA_PCI_DEV_MAX 4
#define UNCORE_EVENT_CONSTRAINT(c, n) EVENT_CONSTRAINT(c, n, 0xff)
struct pci_extra_dev {
struct pci_dev *dev[UNCORE_EXTRA_PCI_DEV_MAX];
};
struct intel_uncore_ops;
struct intel_uncore_pmu;
struct intel_uncore_box;
struct uncore_event_desc;
perf/x86/intel/uncore: Add new data structures for free running counters There are a number of free running counters introduced for uncore, which provide highly valuable information to a wide array of customers. For example, Skylake Server has IIO free running counters to collect Input/Output x BW/Utilization. There is NO event available on the general purpose counters, that is exactly the same as the free running counters. The generic uncore code needs to be enhanced to support the new counters. In the uncore document, there is no event-code assigned to free running counters. Some events need to be defined to indicate the free running counters. The events are encoded as event-code + umask-code. The event-code for all free running counters is 0xff, which is the same as the fixed counters: - It has not been decided what code will be used for common events on future platforms. 0xff is the only one which will definitely not be used as any common event-code. - Cannot re-use current events on the general purpose counters. Because there is NO event available, that is exactly the same as the free running counters. - Even in the existing codes, the fixed counters for core, that have the same event-code, may count different things. Hence, it should not surprise the users if the free running counters that share the same event-code also count different things. Umask will be used to distinguish the counters. The umask-code is used to distinguish a fixed counter and a free running counter, and different types of free running counters. For fixed counters, the umask-code is 0x0X, where X indicates the index of the fixed counter, which starts from 0. - Compatible with the old event encoding. - Currently, there is only one fixed counter. There are still 15 reserved spaces for extension. For free running counters, the umask-code uses the rest of the space. It would follow the format of 0xXY: - X stands for the type of free running counters, which starts from 1. - Y stands for the index of free running counters of same type, which starts from 0. - The free running counters do different thing. It can be categorized to several types, according to the MSR location, bit width and definition. E.g. there are three types of IIO free running counters on Skylake server to monitor IO CLOCKS, BANDWIDTH and UTILIZATION on different ports. It makes it easy to locate the free running counter of a specific type. - So far, there are at most 8 counters of each type. There are still 8 reserved spaces for extension. Introducing a new index to indicate the free running counters. Only one index is enough for all free running counters. Because the free running counters are always active, and the event and free running counter are always 1:1 mapped, it does not need extra index to indicate the assigned counter. Introducing a new data structure to store free running counters related information for each type. It includes the number of counters, bit width, base address, offset between counters and offset between boxes. Introducing several inline helpers to check index for fixed counter and free running counter, validate free running counter event, and retrieve the free running counter information according to box and event. Signed-off-by: Kan Liang <kan.liang@intel.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Reviewed-by: Thomas Gleixner <tglx@linutronix.de> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: acme@kernel.org Cc: eranian@google.com Link: http://lkml.kernel.org/r/1525371913-10597-4-git-send-email-kan.liang@intel.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2018-05-04 02:25:09 +08:00
struct freerunning_counters;
struct intel_uncore_type {
const char *name;
int num_counters;
int num_boxes;
int perf_ctr_bits;
int fixed_ctr_bits;
perf/x86/intel/uncore: Add new data structures for free running counters There are a number of free running counters introduced for uncore, which provide highly valuable information to a wide array of customers. For example, Skylake Server has IIO free running counters to collect Input/Output x BW/Utilization. There is NO event available on the general purpose counters, that is exactly the same as the free running counters. The generic uncore code needs to be enhanced to support the new counters. In the uncore document, there is no event-code assigned to free running counters. Some events need to be defined to indicate the free running counters. The events are encoded as event-code + umask-code. The event-code for all free running counters is 0xff, which is the same as the fixed counters: - It has not been decided what code will be used for common events on future platforms. 0xff is the only one which will definitely not be used as any common event-code. - Cannot re-use current events on the general purpose counters. Because there is NO event available, that is exactly the same as the free running counters. - Even in the existing codes, the fixed counters for core, that have the same event-code, may count different things. Hence, it should not surprise the users if the free running counters that share the same event-code also count different things. Umask will be used to distinguish the counters. The umask-code is used to distinguish a fixed counter and a free running counter, and different types of free running counters. For fixed counters, the umask-code is 0x0X, where X indicates the index of the fixed counter, which starts from 0. - Compatible with the old event encoding. - Currently, there is only one fixed counter. There are still 15 reserved spaces for extension. For free running counters, the umask-code uses the rest of the space. It would follow the format of 0xXY: - X stands for the type of free running counters, which starts from 1. - Y stands for the index of free running counters of same type, which starts from 0. - The free running counters do different thing. It can be categorized to several types, according to the MSR location, bit width and definition. E.g. there are three types of IIO free running counters on Skylake server to monitor IO CLOCKS, BANDWIDTH and UTILIZATION on different ports. It makes it easy to locate the free running counter of a specific type. - So far, there are at most 8 counters of each type. There are still 8 reserved spaces for extension. Introducing a new index to indicate the free running counters. Only one index is enough for all free running counters. Because the free running counters are always active, and the event and free running counter are always 1:1 mapped, it does not need extra index to indicate the assigned counter. Introducing a new data structure to store free running counters related information for each type. It includes the number of counters, bit width, base address, offset between counters and offset between boxes. Introducing several inline helpers to check index for fixed counter and free running counter, validate free running counter event, and retrieve the free running counter information according to box and event. Signed-off-by: Kan Liang <kan.liang@intel.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Reviewed-by: Thomas Gleixner <tglx@linutronix.de> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: acme@kernel.org Cc: eranian@google.com Link: http://lkml.kernel.org/r/1525371913-10597-4-git-send-email-kan.liang@intel.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2018-05-04 02:25:09 +08:00
int num_freerunning_types;
unsigned perf_ctr;
unsigned event_ctl;
unsigned event_mask;
unsigned event_mask_ext;
unsigned fixed_ctr;
unsigned fixed_ctl;
unsigned box_ctl;
union {
unsigned msr_offset;
unsigned mmio_offset;
};
unsigned num_shared_regs:8;
unsigned single_fixed:1;
unsigned pair_ctr_ctl:1;
unsigned *msr_offsets;
struct event_constraint unconstrainted;
struct event_constraint *constraints;
struct intel_uncore_pmu *pmus;
struct intel_uncore_ops *ops;
struct uncore_event_desc *event_descs;
perf/x86/intel/uncore: Add new data structures for free running counters There are a number of free running counters introduced for uncore, which provide highly valuable information to a wide array of customers. For example, Skylake Server has IIO free running counters to collect Input/Output x BW/Utilization. There is NO event available on the general purpose counters, that is exactly the same as the free running counters. The generic uncore code needs to be enhanced to support the new counters. In the uncore document, there is no event-code assigned to free running counters. Some events need to be defined to indicate the free running counters. The events are encoded as event-code + umask-code. The event-code for all free running counters is 0xff, which is the same as the fixed counters: - It has not been decided what code will be used for common events on future platforms. 0xff is the only one which will definitely not be used as any common event-code. - Cannot re-use current events on the general purpose counters. Because there is NO event available, that is exactly the same as the free running counters. - Even in the existing codes, the fixed counters for core, that have the same event-code, may count different things. Hence, it should not surprise the users if the free running counters that share the same event-code also count different things. Umask will be used to distinguish the counters. The umask-code is used to distinguish a fixed counter and a free running counter, and different types of free running counters. For fixed counters, the umask-code is 0x0X, where X indicates the index of the fixed counter, which starts from 0. - Compatible with the old event encoding. - Currently, there is only one fixed counter. There are still 15 reserved spaces for extension. For free running counters, the umask-code uses the rest of the space. It would follow the format of 0xXY: - X stands for the type of free running counters, which starts from 1. - Y stands for the index of free running counters of same type, which starts from 0. - The free running counters do different thing. It can be categorized to several types, according to the MSR location, bit width and definition. E.g. there are three types of IIO free running counters on Skylake server to monitor IO CLOCKS, BANDWIDTH and UTILIZATION on different ports. It makes it easy to locate the free running counter of a specific type. - So far, there are at most 8 counters of each type. There are still 8 reserved spaces for extension. Introducing a new index to indicate the free running counters. Only one index is enough for all free running counters. Because the free running counters are always active, and the event and free running counter are always 1:1 mapped, it does not need extra index to indicate the assigned counter. Introducing a new data structure to store free running counters related information for each type. It includes the number of counters, bit width, base address, offset between counters and offset between boxes. Introducing several inline helpers to check index for fixed counter and free running counter, validate free running counter event, and retrieve the free running counter information according to box and event. Signed-off-by: Kan Liang <kan.liang@intel.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Reviewed-by: Thomas Gleixner <tglx@linutronix.de> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: acme@kernel.org Cc: eranian@google.com Link: http://lkml.kernel.org/r/1525371913-10597-4-git-send-email-kan.liang@intel.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2018-05-04 02:25:09 +08:00
struct freerunning_counters *freerunning;
const struct attribute_group *attr_groups[4];
struct pmu *pmu; /* for custom pmu ops */
};
#define pmu_group attr_groups[0]
#define format_group attr_groups[1]
#define events_group attr_groups[2]
struct intel_uncore_ops {
void (*init_box)(struct intel_uncore_box *);
void (*exit_box)(struct intel_uncore_box *);
void (*disable_box)(struct intel_uncore_box *);
void (*enable_box)(struct intel_uncore_box *);
void (*disable_event)(struct intel_uncore_box *, struct perf_event *);
void (*enable_event)(struct intel_uncore_box *, struct perf_event *);
u64 (*read_counter)(struct intel_uncore_box *, struct perf_event *);
int (*hw_config)(struct intel_uncore_box *, struct perf_event *);
struct event_constraint *(*get_constraint)(struct intel_uncore_box *,
struct perf_event *);
void (*put_constraint)(struct intel_uncore_box *, struct perf_event *);
};
struct intel_uncore_pmu {
struct pmu pmu;
char name[UNCORE_PMU_NAME_LEN];
int pmu_idx;
int func_id;
bool registered;
atomic_t activeboxes;
struct intel_uncore_type *type;
struct intel_uncore_box **boxes;
};
struct intel_uncore_extra_reg {
raw_spinlock_t lock;
u64 config, config1, config2;
atomic_t ref;
};
struct intel_uncore_box {
int pci_phys_id;
int dieid; /* Logical die ID */
int n_active; /* number of active events */
int n_events;
int cpu; /* cpu to collect events */
unsigned long flags;
atomic_t refcnt;
struct perf_event *events[UNCORE_PMC_IDX_MAX];
struct perf_event *event_list[UNCORE_PMC_IDX_MAX];
perf/x86: Fix event/group validation Commit 43b4578071c0 ("perf/x86: Reduce stack usage of x86_schedule_events()") violated the rule that 'fake' scheduling; as used for event/group validation; should not change the event state. This went mostly un-noticed because repeated calls of x86_pmu::get_event_constraints() would give the same result. And x86_pmu::put_event_constraints() would mostly not do anything. Commit e979121b1b15 ("perf/x86/intel: Implement cross-HT corruption bug workaround") made the situation much worse by actually setting the event->hw.constraint value to NULL, so when validation and actual scheduling interact we get NULL ptr derefs. Fix it by removing the constraint pointer from the event and move it back to an array, this time in cpuc instead of on the stack. validate_group() x86_schedule_events() event->hw.constraint = c; # store <context switch> perf_task_event_sched_in() ... x86_schedule_events(); event->hw.constraint = c2; # store ... put_event_constraints(event); # assume failure to schedule intel_put_event_constraints() event->hw.constraint = NULL; <context switch end> c = event->hw.constraint; # read -> NULL if (!test_bit(hwc->idx, c->idxmsk)) # <- *BOOM* NULL deref This in particular is possible when the event in question is a cpu-wide event and group-leader, where the validate_group() tries to add an event to the group. Reported-by: Vince Weaver <vincent.weaver@maine.edu> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Cc: Andrew Hunter <ahh@google.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Maria Dimakopoulou <maria.n.dimakopoulou@gmail.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Fixes: 43b4578071c0 ("perf/x86: Reduce stack usage of x86_schedule_events()") Fixes: e979121b1b15 ("perf/x86/intel: Implement cross-HT corruption bug workaround") Signed-off-by: Ingo Molnar <mingo@kernel.org>
2015-05-21 16:57:13 +08:00
struct event_constraint *event_constraint[UNCORE_PMC_IDX_MAX];
unsigned long active_mask[BITS_TO_LONGS(UNCORE_PMC_IDX_MAX)];
u64 tags[UNCORE_PMC_IDX_MAX];
struct pci_dev *pci_dev;
struct intel_uncore_pmu *pmu;
u64 hrtimer_duration; /* hrtimer timeout for this box */
struct hrtimer hrtimer;
struct list_head list;
struct list_head active_list;
void __iomem *io_addr;
struct intel_uncore_extra_reg shared_regs[0];
};
/* CFL uncore 8th cbox MSRs */
#define CFL_UNC_CBO_7_PERFEVTSEL0 0xf70
#define CFL_UNC_CBO_7_PER_CTR0 0xf76
#define UNCORE_BOX_FLAG_INITIATED 0
/* event config registers are 8-byte apart */
#define UNCORE_BOX_FLAG_CTL_OFFS8 1
/* CFL 8th CBOX has different MSR space */
#define UNCORE_BOX_FLAG_CFL8_CBOX_MSR_OFFS 2
struct uncore_event_desc {
struct kobj_attribute attr;
const char *config;
};
perf/x86/intel/uncore: Add new data structures for free running counters There are a number of free running counters introduced for uncore, which provide highly valuable information to a wide array of customers. For example, Skylake Server has IIO free running counters to collect Input/Output x BW/Utilization. There is NO event available on the general purpose counters, that is exactly the same as the free running counters. The generic uncore code needs to be enhanced to support the new counters. In the uncore document, there is no event-code assigned to free running counters. Some events need to be defined to indicate the free running counters. The events are encoded as event-code + umask-code. The event-code for all free running counters is 0xff, which is the same as the fixed counters: - It has not been decided what code will be used for common events on future platforms. 0xff is the only one which will definitely not be used as any common event-code. - Cannot re-use current events on the general purpose counters. Because there is NO event available, that is exactly the same as the free running counters. - Even in the existing codes, the fixed counters for core, that have the same event-code, may count different things. Hence, it should not surprise the users if the free running counters that share the same event-code also count different things. Umask will be used to distinguish the counters. The umask-code is used to distinguish a fixed counter and a free running counter, and different types of free running counters. For fixed counters, the umask-code is 0x0X, where X indicates the index of the fixed counter, which starts from 0. - Compatible with the old event encoding. - Currently, there is only one fixed counter. There are still 15 reserved spaces for extension. For free running counters, the umask-code uses the rest of the space. It would follow the format of 0xXY: - X stands for the type of free running counters, which starts from 1. - Y stands for the index of free running counters of same type, which starts from 0. - The free running counters do different thing. It can be categorized to several types, according to the MSR location, bit width and definition. E.g. there are three types of IIO free running counters on Skylake server to monitor IO CLOCKS, BANDWIDTH and UTILIZATION on different ports. It makes it easy to locate the free running counter of a specific type. - So far, there are at most 8 counters of each type. There are still 8 reserved spaces for extension. Introducing a new index to indicate the free running counters. Only one index is enough for all free running counters. Because the free running counters are always active, and the event and free running counter are always 1:1 mapped, it does not need extra index to indicate the assigned counter. Introducing a new data structure to store free running counters related information for each type. It includes the number of counters, bit width, base address, offset between counters and offset between boxes. Introducing several inline helpers to check index for fixed counter and free running counter, validate free running counter event, and retrieve the free running counter information according to box and event. Signed-off-by: Kan Liang <kan.liang@intel.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Reviewed-by: Thomas Gleixner <tglx@linutronix.de> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: acme@kernel.org Cc: eranian@google.com Link: http://lkml.kernel.org/r/1525371913-10597-4-git-send-email-kan.liang@intel.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2018-05-04 02:25:09 +08:00
struct freerunning_counters {
unsigned int counter_base;
unsigned int counter_offset;
unsigned int box_offset;
unsigned int num_counters;
unsigned int bits;
};
struct pci2phy_map {
struct list_head list;
int segment;
int pbus_to_physid[256];
};
struct pci2phy_map *__find_pci2phy_map(int segment);
int uncore_pcibus_to_physid(struct pci_bus *bus);
ssize_t uncore_event_show(struct kobject *kobj,
struct kobj_attribute *attr, char *buf);
#define INTEL_UNCORE_EVENT_DESC(_name, _config) \
{ \
.attr = __ATTR(_name, 0444, uncore_event_show, NULL), \
.config = _config, \
}
#define DEFINE_UNCORE_FORMAT_ATTR(_var, _name, _format) \
static ssize_t __uncore_##_var##_show(struct kobject *kobj, \
struct kobj_attribute *attr, \
char *page) \
{ \
BUILD_BUG_ON(sizeof(_format) >= PAGE_SIZE); \
return sprintf(page, _format "\n"); \
} \
static struct kobj_attribute format_attr_##_var = \
__ATTR(_name, 0444, __uncore_##_var##_show, NULL)
perf/x86/intel/uncore: Add new data structures for free running counters There are a number of free running counters introduced for uncore, which provide highly valuable information to a wide array of customers. For example, Skylake Server has IIO free running counters to collect Input/Output x BW/Utilization. There is NO event available on the general purpose counters, that is exactly the same as the free running counters. The generic uncore code needs to be enhanced to support the new counters. In the uncore document, there is no event-code assigned to free running counters. Some events need to be defined to indicate the free running counters. The events are encoded as event-code + umask-code. The event-code for all free running counters is 0xff, which is the same as the fixed counters: - It has not been decided what code will be used for common events on future platforms. 0xff is the only one which will definitely not be used as any common event-code. - Cannot re-use current events on the general purpose counters. Because there is NO event available, that is exactly the same as the free running counters. - Even in the existing codes, the fixed counters for core, that have the same event-code, may count different things. Hence, it should not surprise the users if the free running counters that share the same event-code also count different things. Umask will be used to distinguish the counters. The umask-code is used to distinguish a fixed counter and a free running counter, and different types of free running counters. For fixed counters, the umask-code is 0x0X, where X indicates the index of the fixed counter, which starts from 0. - Compatible with the old event encoding. - Currently, there is only one fixed counter. There are still 15 reserved spaces for extension. For free running counters, the umask-code uses the rest of the space. It would follow the format of 0xXY: - X stands for the type of free running counters, which starts from 1. - Y stands for the index of free running counters of same type, which starts from 0. - The free running counters do different thing. It can be categorized to several types, according to the MSR location, bit width and definition. E.g. there are three types of IIO free running counters on Skylake server to monitor IO CLOCKS, BANDWIDTH and UTILIZATION on different ports. It makes it easy to locate the free running counter of a specific type. - So far, there are at most 8 counters of each type. There are still 8 reserved spaces for extension. Introducing a new index to indicate the free running counters. Only one index is enough for all free running counters. Because the free running counters are always active, and the event and free running counter are always 1:1 mapped, it does not need extra index to indicate the assigned counter. Introducing a new data structure to store free running counters related information for each type. It includes the number of counters, bit width, base address, offset between counters and offset between boxes. Introducing several inline helpers to check index for fixed counter and free running counter, validate free running counter event, and retrieve the free running counter information according to box and event. Signed-off-by: Kan Liang <kan.liang@intel.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Reviewed-by: Thomas Gleixner <tglx@linutronix.de> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: acme@kernel.org Cc: eranian@google.com Link: http://lkml.kernel.org/r/1525371913-10597-4-git-send-email-kan.liang@intel.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2018-05-04 02:25:09 +08:00
static inline bool uncore_pmc_fixed(int idx)
{
return idx == UNCORE_PMC_IDX_FIXED;
}
static inline bool uncore_pmc_freerunning(int idx)
{
return idx == UNCORE_PMC_IDX_FREERUNNING;
}
static inline
unsigned int uncore_mmio_box_ctl(struct intel_uncore_box *box)
{
return box->pmu->type->box_ctl +
box->pmu->type->mmio_offset * box->pmu->pmu_idx;
}
static inline unsigned uncore_pci_box_ctl(struct intel_uncore_box *box)
{
return box->pmu->type->box_ctl;
}
static inline unsigned uncore_pci_fixed_ctl(struct intel_uncore_box *box)
{
return box->pmu->type->fixed_ctl;
}
static inline unsigned uncore_pci_fixed_ctr(struct intel_uncore_box *box)
{
return box->pmu->type->fixed_ctr;
}
static inline
unsigned uncore_pci_event_ctl(struct intel_uncore_box *box, int idx)
{
if (test_bit(UNCORE_BOX_FLAG_CTL_OFFS8, &box->flags))
return idx * 8 + box->pmu->type->event_ctl;
return idx * 4 + box->pmu->type->event_ctl;
}
static inline
unsigned uncore_pci_perf_ctr(struct intel_uncore_box *box, int idx)
{
return idx * 8 + box->pmu->type->perf_ctr;
}
static inline unsigned uncore_msr_box_offset(struct intel_uncore_box *box)
{
struct intel_uncore_pmu *pmu = box->pmu;
return pmu->type->msr_offsets ?
pmu->type->msr_offsets[pmu->pmu_idx] :
pmu->type->msr_offset * pmu->pmu_idx;
}
static inline unsigned uncore_msr_box_ctl(struct intel_uncore_box *box)
{
if (!box->pmu->type->box_ctl)
return 0;
return box->pmu->type->box_ctl + uncore_msr_box_offset(box);
}
static inline unsigned uncore_msr_fixed_ctl(struct intel_uncore_box *box)
{
if (!box->pmu->type->fixed_ctl)
return 0;
return box->pmu->type->fixed_ctl + uncore_msr_box_offset(box);
}
static inline unsigned uncore_msr_fixed_ctr(struct intel_uncore_box *box)
{
return box->pmu->type->fixed_ctr + uncore_msr_box_offset(box);
}
perf/x86/intel/uncore: Add new data structures for free running counters There are a number of free running counters introduced for uncore, which provide highly valuable information to a wide array of customers. For example, Skylake Server has IIO free running counters to collect Input/Output x BW/Utilization. There is NO event available on the general purpose counters, that is exactly the same as the free running counters. The generic uncore code needs to be enhanced to support the new counters. In the uncore document, there is no event-code assigned to free running counters. Some events need to be defined to indicate the free running counters. The events are encoded as event-code + umask-code. The event-code for all free running counters is 0xff, which is the same as the fixed counters: - It has not been decided what code will be used for common events on future platforms. 0xff is the only one which will definitely not be used as any common event-code. - Cannot re-use current events on the general purpose counters. Because there is NO event available, that is exactly the same as the free running counters. - Even in the existing codes, the fixed counters for core, that have the same event-code, may count different things. Hence, it should not surprise the users if the free running counters that share the same event-code also count different things. Umask will be used to distinguish the counters. The umask-code is used to distinguish a fixed counter and a free running counter, and different types of free running counters. For fixed counters, the umask-code is 0x0X, where X indicates the index of the fixed counter, which starts from 0. - Compatible with the old event encoding. - Currently, there is only one fixed counter. There are still 15 reserved spaces for extension. For free running counters, the umask-code uses the rest of the space. It would follow the format of 0xXY: - X stands for the type of free running counters, which starts from 1. - Y stands for the index of free running counters of same type, which starts from 0. - The free running counters do different thing. It can be categorized to several types, according to the MSR location, bit width and definition. E.g. there are three types of IIO free running counters on Skylake server to monitor IO CLOCKS, BANDWIDTH and UTILIZATION on different ports. It makes it easy to locate the free running counter of a specific type. - So far, there are at most 8 counters of each type. There are still 8 reserved spaces for extension. Introducing a new index to indicate the free running counters. Only one index is enough for all free running counters. Because the free running counters are always active, and the event and free running counter are always 1:1 mapped, it does not need extra index to indicate the assigned counter. Introducing a new data structure to store free running counters related information for each type. It includes the number of counters, bit width, base address, offset between counters and offset between boxes. Introducing several inline helpers to check index for fixed counter and free running counter, validate free running counter event, and retrieve the free running counter information according to box and event. Signed-off-by: Kan Liang <kan.liang@intel.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Reviewed-by: Thomas Gleixner <tglx@linutronix.de> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: acme@kernel.org Cc: eranian@google.com Link: http://lkml.kernel.org/r/1525371913-10597-4-git-send-email-kan.liang@intel.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2018-05-04 02:25:09 +08:00
/*
* In the uncore document, there is no event-code assigned to free running
* counters. Some events need to be defined to indicate the free running
* counters. The events are encoded as event-code + umask-code.
*
* The event-code for all free running counters is 0xff, which is the same as
* the fixed counters.
*
* The umask-code is used to distinguish a fixed counter and a free running
* counter, and different types of free running counters.
* - For fixed counters, the umask-code is 0x0X.
* X indicates the index of the fixed counter, which starts from 0.
* - For free running counters, the umask-code uses the rest of the space.
* It would bare the format of 0xXY.
* X stands for the type of free running counters, which starts from 1.
* Y stands for the index of free running counters of same type, which
* starts from 0.
*
* For example, there are three types of IIO free running counters on Skylake
* server, IO CLOCKS counters, BANDWIDTH counters and UTILIZATION counters.
* The event-code for all the free running counters is 0xff.
* 'ioclk' is the first counter of IO CLOCKS. IO CLOCKS is the first type,
* which umask-code starts from 0x10.
* So 'ioclk' is encoded as event=0xff,umask=0x10
* 'bw_in_port2' is the third counter of BANDWIDTH counters. BANDWIDTH is
* the second type, which umask-code starts from 0x20.
* So 'bw_in_port2' is encoded as event=0xff,umask=0x22
*/
static inline unsigned int uncore_freerunning_idx(u64 config)
{
return ((config >> 8) & 0xf);
}
#define UNCORE_FREERUNNING_UMASK_START 0x10
static inline unsigned int uncore_freerunning_type(u64 config)
{
return ((((config >> 8) - UNCORE_FREERUNNING_UMASK_START) >> 4) & 0xf);
}
static inline
unsigned int uncore_freerunning_counter(struct intel_uncore_box *box,
struct perf_event *event)
{
perf/x86/intel/uncore: Fix client IMC events return huge result The client IMC bandwidth events currently return very large values: $ perf stat -e uncore_imc/data_reads/ -e uncore_imc/data_writes/ -I 10000 -a 10.000117222 34,788.76 MiB uncore_imc/data_reads/ 10.000117222 8.26 MiB uncore_imc/data_writes/ 20.000374584 34,842.89 MiB uncore_imc/data_reads/ 20.000374584 10.45 MiB uncore_imc/data_writes/ 30.000633299 37,965.29 MiB uncore_imc/data_reads/ 30.000633299 323.62 MiB uncore_imc/data_writes/ 40.000891548 41,012.88 MiB uncore_imc/data_reads/ 40.000891548 6.98 MiB uncore_imc/data_writes/ 50.001142480 1,125,899,906,621,494.75 MiB uncore_imc/data_reads/ 50.001142480 6.97 MiB uncore_imc/data_writes/ The client IMC events are freerunning counters. They still use the old event encoding format (0x1 for data_read and 0x2 for data write). The counter bit width is calculated by common code, which assume that the standard encoding format is used for the freerunning counters. Error bit width information is calculated. The patch intends to convert the old client IMC event encoding to the standard encoding format. Current common code uses event->attr.config which directly copy from user space. We should not implicitly modify it for a converted event. The event->hw.config is used to replace the event->attr.config in common code. For client IMC events, the event->attr.config is used to calculate a converted event with standard encoding format in the custom event_init(). The converted event is stored in event->hw.config. For other events of freerunning counters, they already use the standard encoding format. The same value as event->attr.config is assigned to event->hw.config in common event_init(). Reported-by: Jin Yao <yao.jin@linux.intel.com> Tested-by: Jin Yao <yao.jin@linux.intel.com> Signed-off-by: Kan Liang <kan.liang@linux.intel.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Andy Lutomirski <luto@kernel.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Borislav Petkov <bp@alien8.de> Cc: Dave Hansen <dave.hansen@linux.intel.com> Cc: H. Peter Anvin <hpa@zytor.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Rik van Riel <riel@surriel.com> Cc: Stephane Eranian <eranian@google.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Vince Weaver <vincent.weaver@maine.edu> Cc: stable@kernel.org # v4.18+ Fixes: 9aae1780e7e8 ("perf/x86/intel/uncore: Clean up client IMC uncore") Link: https://lkml.kernel.org/r/20190227165729.1861-1-kan.liang@linux.intel.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2019-02-28 00:57:29 +08:00
unsigned int type = uncore_freerunning_type(event->hw.config);
unsigned int idx = uncore_freerunning_idx(event->hw.config);
perf/x86/intel/uncore: Add new data structures for free running counters There are a number of free running counters introduced for uncore, which provide highly valuable information to a wide array of customers. For example, Skylake Server has IIO free running counters to collect Input/Output x BW/Utilization. There is NO event available on the general purpose counters, that is exactly the same as the free running counters. The generic uncore code needs to be enhanced to support the new counters. In the uncore document, there is no event-code assigned to free running counters. Some events need to be defined to indicate the free running counters. The events are encoded as event-code + umask-code. The event-code for all free running counters is 0xff, which is the same as the fixed counters: - It has not been decided what code will be used for common events on future platforms. 0xff is the only one which will definitely not be used as any common event-code. - Cannot re-use current events on the general purpose counters. Because there is NO event available, that is exactly the same as the free running counters. - Even in the existing codes, the fixed counters for core, that have the same event-code, may count different things. Hence, it should not surprise the users if the free running counters that share the same event-code also count different things. Umask will be used to distinguish the counters. The umask-code is used to distinguish a fixed counter and a free running counter, and different types of free running counters. For fixed counters, the umask-code is 0x0X, where X indicates the index of the fixed counter, which starts from 0. - Compatible with the old event encoding. - Currently, there is only one fixed counter. There are still 15 reserved spaces for extension. For free running counters, the umask-code uses the rest of the space. It would follow the format of 0xXY: - X stands for the type of free running counters, which starts from 1. - Y stands for the index of free running counters of same type, which starts from 0. - The free running counters do different thing. It can be categorized to several types, according to the MSR location, bit width and definition. E.g. there are three types of IIO free running counters on Skylake server to monitor IO CLOCKS, BANDWIDTH and UTILIZATION on different ports. It makes it easy to locate the free running counter of a specific type. - So far, there are at most 8 counters of each type. There are still 8 reserved spaces for extension. Introducing a new index to indicate the free running counters. Only one index is enough for all free running counters. Because the free running counters are always active, and the event and free running counter are always 1:1 mapped, it does not need extra index to indicate the assigned counter. Introducing a new data structure to store free running counters related information for each type. It includes the number of counters, bit width, base address, offset between counters and offset between boxes. Introducing several inline helpers to check index for fixed counter and free running counter, validate free running counter event, and retrieve the free running counter information according to box and event. Signed-off-by: Kan Liang <kan.liang@intel.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Reviewed-by: Thomas Gleixner <tglx@linutronix.de> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: acme@kernel.org Cc: eranian@google.com Link: http://lkml.kernel.org/r/1525371913-10597-4-git-send-email-kan.liang@intel.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2018-05-04 02:25:09 +08:00
struct intel_uncore_pmu *pmu = box->pmu;
return pmu->type->freerunning[type].counter_base +
pmu->type->freerunning[type].counter_offset * idx +
pmu->type->freerunning[type].box_offset * pmu->pmu_idx;
}
static inline
unsigned uncore_msr_event_ctl(struct intel_uncore_box *box, int idx)
{
if (test_bit(UNCORE_BOX_FLAG_CFL8_CBOX_MSR_OFFS, &box->flags)) {
return CFL_UNC_CBO_7_PERFEVTSEL0 +
(box->pmu->type->pair_ctr_ctl ? 2 * idx : idx);
} else {
return box->pmu->type->event_ctl +
(box->pmu->type->pair_ctr_ctl ? 2 * idx : idx) +
uncore_msr_box_offset(box);
}
}
static inline
unsigned uncore_msr_perf_ctr(struct intel_uncore_box *box, int idx)
{
if (test_bit(UNCORE_BOX_FLAG_CFL8_CBOX_MSR_OFFS, &box->flags)) {
return CFL_UNC_CBO_7_PER_CTR0 +
(box->pmu->type->pair_ctr_ctl ? 2 * idx : idx);
} else {
return box->pmu->type->perf_ctr +
(box->pmu->type->pair_ctr_ctl ? 2 * idx : idx) +
uncore_msr_box_offset(box);
}
}
static inline
unsigned uncore_fixed_ctl(struct intel_uncore_box *box)
{
if (box->pci_dev || box->io_addr)
return uncore_pci_fixed_ctl(box);
else
return uncore_msr_fixed_ctl(box);
}
static inline
unsigned uncore_fixed_ctr(struct intel_uncore_box *box)
{
if (box->pci_dev || box->io_addr)
return uncore_pci_fixed_ctr(box);
else
return uncore_msr_fixed_ctr(box);
}
static inline
unsigned uncore_event_ctl(struct intel_uncore_box *box, int idx)
{
if (box->pci_dev || box->io_addr)
return uncore_pci_event_ctl(box, idx);
else
return uncore_msr_event_ctl(box, idx);
}
static inline
unsigned uncore_perf_ctr(struct intel_uncore_box *box, int idx)
{
if (box->pci_dev || box->io_addr)
return uncore_pci_perf_ctr(box, idx);
else
return uncore_msr_perf_ctr(box, idx);
}
static inline int uncore_perf_ctr_bits(struct intel_uncore_box *box)
{
return box->pmu->type->perf_ctr_bits;
}
static inline int uncore_fixed_ctr_bits(struct intel_uncore_box *box)
{
return box->pmu->type->fixed_ctr_bits;
}
perf/x86/intel/uncore: Add new data structures for free running counters There are a number of free running counters introduced for uncore, which provide highly valuable information to a wide array of customers. For example, Skylake Server has IIO free running counters to collect Input/Output x BW/Utilization. There is NO event available on the general purpose counters, that is exactly the same as the free running counters. The generic uncore code needs to be enhanced to support the new counters. In the uncore document, there is no event-code assigned to free running counters. Some events need to be defined to indicate the free running counters. The events are encoded as event-code + umask-code. The event-code for all free running counters is 0xff, which is the same as the fixed counters: - It has not been decided what code will be used for common events on future platforms. 0xff is the only one which will definitely not be used as any common event-code. - Cannot re-use current events on the general purpose counters. Because there is NO event available, that is exactly the same as the free running counters. - Even in the existing codes, the fixed counters for core, that have the same event-code, may count different things. Hence, it should not surprise the users if the free running counters that share the same event-code also count different things. Umask will be used to distinguish the counters. The umask-code is used to distinguish a fixed counter and a free running counter, and different types of free running counters. For fixed counters, the umask-code is 0x0X, where X indicates the index of the fixed counter, which starts from 0. - Compatible with the old event encoding. - Currently, there is only one fixed counter. There are still 15 reserved spaces for extension. For free running counters, the umask-code uses the rest of the space. It would follow the format of 0xXY: - X stands for the type of free running counters, which starts from 1. - Y stands for the index of free running counters of same type, which starts from 0. - The free running counters do different thing. It can be categorized to several types, according to the MSR location, bit width and definition. E.g. there are three types of IIO free running counters on Skylake server to monitor IO CLOCKS, BANDWIDTH and UTILIZATION on different ports. It makes it easy to locate the free running counter of a specific type. - So far, there are at most 8 counters of each type. There are still 8 reserved spaces for extension. Introducing a new index to indicate the free running counters. Only one index is enough for all free running counters. Because the free running counters are always active, and the event and free running counter are always 1:1 mapped, it does not need extra index to indicate the assigned counter. Introducing a new data structure to store free running counters related information for each type. It includes the number of counters, bit width, base address, offset between counters and offset between boxes. Introducing several inline helpers to check index for fixed counter and free running counter, validate free running counter event, and retrieve the free running counter information according to box and event. Signed-off-by: Kan Liang <kan.liang@intel.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Reviewed-by: Thomas Gleixner <tglx@linutronix.de> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: acme@kernel.org Cc: eranian@google.com Link: http://lkml.kernel.org/r/1525371913-10597-4-git-send-email-kan.liang@intel.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2018-05-04 02:25:09 +08:00
static inline
unsigned int uncore_freerunning_bits(struct intel_uncore_box *box,
struct perf_event *event)
{
perf/x86/intel/uncore: Fix client IMC events return huge result The client IMC bandwidth events currently return very large values: $ perf stat -e uncore_imc/data_reads/ -e uncore_imc/data_writes/ -I 10000 -a 10.000117222 34,788.76 MiB uncore_imc/data_reads/ 10.000117222 8.26 MiB uncore_imc/data_writes/ 20.000374584 34,842.89 MiB uncore_imc/data_reads/ 20.000374584 10.45 MiB uncore_imc/data_writes/ 30.000633299 37,965.29 MiB uncore_imc/data_reads/ 30.000633299 323.62 MiB uncore_imc/data_writes/ 40.000891548 41,012.88 MiB uncore_imc/data_reads/ 40.000891548 6.98 MiB uncore_imc/data_writes/ 50.001142480 1,125,899,906,621,494.75 MiB uncore_imc/data_reads/ 50.001142480 6.97 MiB uncore_imc/data_writes/ The client IMC events are freerunning counters. They still use the old event encoding format (0x1 for data_read and 0x2 for data write). The counter bit width is calculated by common code, which assume that the standard encoding format is used for the freerunning counters. Error bit width information is calculated. The patch intends to convert the old client IMC event encoding to the standard encoding format. Current common code uses event->attr.config which directly copy from user space. We should not implicitly modify it for a converted event. The event->hw.config is used to replace the event->attr.config in common code. For client IMC events, the event->attr.config is used to calculate a converted event with standard encoding format in the custom event_init(). The converted event is stored in event->hw.config. For other events of freerunning counters, they already use the standard encoding format. The same value as event->attr.config is assigned to event->hw.config in common event_init(). Reported-by: Jin Yao <yao.jin@linux.intel.com> Tested-by: Jin Yao <yao.jin@linux.intel.com> Signed-off-by: Kan Liang <kan.liang@linux.intel.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Andy Lutomirski <luto@kernel.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Borislav Petkov <bp@alien8.de> Cc: Dave Hansen <dave.hansen@linux.intel.com> Cc: H. Peter Anvin <hpa@zytor.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Rik van Riel <riel@surriel.com> Cc: Stephane Eranian <eranian@google.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Vince Weaver <vincent.weaver@maine.edu> Cc: stable@kernel.org # v4.18+ Fixes: 9aae1780e7e8 ("perf/x86/intel/uncore: Clean up client IMC uncore") Link: https://lkml.kernel.org/r/20190227165729.1861-1-kan.liang@linux.intel.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2019-02-28 00:57:29 +08:00
unsigned int type = uncore_freerunning_type(event->hw.config);
perf/x86/intel/uncore: Add new data structures for free running counters There are a number of free running counters introduced for uncore, which provide highly valuable information to a wide array of customers. For example, Skylake Server has IIO free running counters to collect Input/Output x BW/Utilization. There is NO event available on the general purpose counters, that is exactly the same as the free running counters. The generic uncore code needs to be enhanced to support the new counters. In the uncore document, there is no event-code assigned to free running counters. Some events need to be defined to indicate the free running counters. The events are encoded as event-code + umask-code. The event-code for all free running counters is 0xff, which is the same as the fixed counters: - It has not been decided what code will be used for common events on future platforms. 0xff is the only one which will definitely not be used as any common event-code. - Cannot re-use current events on the general purpose counters. Because there is NO event available, that is exactly the same as the free running counters. - Even in the existing codes, the fixed counters for core, that have the same event-code, may count different things. Hence, it should not surprise the users if the free running counters that share the same event-code also count different things. Umask will be used to distinguish the counters. The umask-code is used to distinguish a fixed counter and a free running counter, and different types of free running counters. For fixed counters, the umask-code is 0x0X, where X indicates the index of the fixed counter, which starts from 0. - Compatible with the old event encoding. - Currently, there is only one fixed counter. There are still 15 reserved spaces for extension. For free running counters, the umask-code uses the rest of the space. It would follow the format of 0xXY: - X stands for the type of free running counters, which starts from 1. - Y stands for the index of free running counters of same type, which starts from 0. - The free running counters do different thing. It can be categorized to several types, according to the MSR location, bit width and definition. E.g. there are three types of IIO free running counters on Skylake server to monitor IO CLOCKS, BANDWIDTH and UTILIZATION on different ports. It makes it easy to locate the free running counter of a specific type. - So far, there are at most 8 counters of each type. There are still 8 reserved spaces for extension. Introducing a new index to indicate the free running counters. Only one index is enough for all free running counters. Because the free running counters are always active, and the event and free running counter are always 1:1 mapped, it does not need extra index to indicate the assigned counter. Introducing a new data structure to store free running counters related information for each type. It includes the number of counters, bit width, base address, offset between counters and offset between boxes. Introducing several inline helpers to check index for fixed counter and free running counter, validate free running counter event, and retrieve the free running counter information according to box and event. Signed-off-by: Kan Liang <kan.liang@intel.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Reviewed-by: Thomas Gleixner <tglx@linutronix.de> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: acme@kernel.org Cc: eranian@google.com Link: http://lkml.kernel.org/r/1525371913-10597-4-git-send-email-kan.liang@intel.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2018-05-04 02:25:09 +08:00
return box->pmu->type->freerunning[type].bits;
}
static inline int uncore_num_freerunning(struct intel_uncore_box *box,
struct perf_event *event)
{
perf/x86/intel/uncore: Fix client IMC events return huge result The client IMC bandwidth events currently return very large values: $ perf stat -e uncore_imc/data_reads/ -e uncore_imc/data_writes/ -I 10000 -a 10.000117222 34,788.76 MiB uncore_imc/data_reads/ 10.000117222 8.26 MiB uncore_imc/data_writes/ 20.000374584 34,842.89 MiB uncore_imc/data_reads/ 20.000374584 10.45 MiB uncore_imc/data_writes/ 30.000633299 37,965.29 MiB uncore_imc/data_reads/ 30.000633299 323.62 MiB uncore_imc/data_writes/ 40.000891548 41,012.88 MiB uncore_imc/data_reads/ 40.000891548 6.98 MiB uncore_imc/data_writes/ 50.001142480 1,125,899,906,621,494.75 MiB uncore_imc/data_reads/ 50.001142480 6.97 MiB uncore_imc/data_writes/ The client IMC events are freerunning counters. They still use the old event encoding format (0x1 for data_read and 0x2 for data write). The counter bit width is calculated by common code, which assume that the standard encoding format is used for the freerunning counters. Error bit width information is calculated. The patch intends to convert the old client IMC event encoding to the standard encoding format. Current common code uses event->attr.config which directly copy from user space. We should not implicitly modify it for a converted event. The event->hw.config is used to replace the event->attr.config in common code. For client IMC events, the event->attr.config is used to calculate a converted event with standard encoding format in the custom event_init(). The converted event is stored in event->hw.config. For other events of freerunning counters, they already use the standard encoding format. The same value as event->attr.config is assigned to event->hw.config in common event_init(). Reported-by: Jin Yao <yao.jin@linux.intel.com> Tested-by: Jin Yao <yao.jin@linux.intel.com> Signed-off-by: Kan Liang <kan.liang@linux.intel.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Andy Lutomirski <luto@kernel.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Borislav Petkov <bp@alien8.de> Cc: Dave Hansen <dave.hansen@linux.intel.com> Cc: H. Peter Anvin <hpa@zytor.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Rik van Riel <riel@surriel.com> Cc: Stephane Eranian <eranian@google.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Vince Weaver <vincent.weaver@maine.edu> Cc: stable@kernel.org # v4.18+ Fixes: 9aae1780e7e8 ("perf/x86/intel/uncore: Clean up client IMC uncore") Link: https://lkml.kernel.org/r/20190227165729.1861-1-kan.liang@linux.intel.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2019-02-28 00:57:29 +08:00
unsigned int type = uncore_freerunning_type(event->hw.config);
perf/x86/intel/uncore: Add new data structures for free running counters There are a number of free running counters introduced for uncore, which provide highly valuable information to a wide array of customers. For example, Skylake Server has IIO free running counters to collect Input/Output x BW/Utilization. There is NO event available on the general purpose counters, that is exactly the same as the free running counters. The generic uncore code needs to be enhanced to support the new counters. In the uncore document, there is no event-code assigned to free running counters. Some events need to be defined to indicate the free running counters. The events are encoded as event-code + umask-code. The event-code for all free running counters is 0xff, which is the same as the fixed counters: - It has not been decided what code will be used for common events on future platforms. 0xff is the only one which will definitely not be used as any common event-code. - Cannot re-use current events on the general purpose counters. Because there is NO event available, that is exactly the same as the free running counters. - Even in the existing codes, the fixed counters for core, that have the same event-code, may count different things. Hence, it should not surprise the users if the free running counters that share the same event-code also count different things. Umask will be used to distinguish the counters. The umask-code is used to distinguish a fixed counter and a free running counter, and different types of free running counters. For fixed counters, the umask-code is 0x0X, where X indicates the index of the fixed counter, which starts from 0. - Compatible with the old event encoding. - Currently, there is only one fixed counter. There are still 15 reserved spaces for extension. For free running counters, the umask-code uses the rest of the space. It would follow the format of 0xXY: - X stands for the type of free running counters, which starts from 1. - Y stands for the index of free running counters of same type, which starts from 0. - The free running counters do different thing. It can be categorized to several types, according to the MSR location, bit width and definition. E.g. there are three types of IIO free running counters on Skylake server to monitor IO CLOCKS, BANDWIDTH and UTILIZATION on different ports. It makes it easy to locate the free running counter of a specific type. - So far, there are at most 8 counters of each type. There are still 8 reserved spaces for extension. Introducing a new index to indicate the free running counters. Only one index is enough for all free running counters. Because the free running counters are always active, and the event and free running counter are always 1:1 mapped, it does not need extra index to indicate the assigned counter. Introducing a new data structure to store free running counters related information for each type. It includes the number of counters, bit width, base address, offset between counters and offset between boxes. Introducing several inline helpers to check index for fixed counter and free running counter, validate free running counter event, and retrieve the free running counter information according to box and event. Signed-off-by: Kan Liang <kan.liang@intel.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Reviewed-by: Thomas Gleixner <tglx@linutronix.de> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: acme@kernel.org Cc: eranian@google.com Link: http://lkml.kernel.org/r/1525371913-10597-4-git-send-email-kan.liang@intel.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2018-05-04 02:25:09 +08:00
return box->pmu->type->freerunning[type].num_counters;
}
static inline int uncore_num_freerunning_types(struct intel_uncore_box *box,
struct perf_event *event)
{
return box->pmu->type->num_freerunning_types;
}
static inline bool check_valid_freerunning_event(struct intel_uncore_box *box,
struct perf_event *event)
{
perf/x86/intel/uncore: Fix client IMC events return huge result The client IMC bandwidth events currently return very large values: $ perf stat -e uncore_imc/data_reads/ -e uncore_imc/data_writes/ -I 10000 -a 10.000117222 34,788.76 MiB uncore_imc/data_reads/ 10.000117222 8.26 MiB uncore_imc/data_writes/ 20.000374584 34,842.89 MiB uncore_imc/data_reads/ 20.000374584 10.45 MiB uncore_imc/data_writes/ 30.000633299 37,965.29 MiB uncore_imc/data_reads/ 30.000633299 323.62 MiB uncore_imc/data_writes/ 40.000891548 41,012.88 MiB uncore_imc/data_reads/ 40.000891548 6.98 MiB uncore_imc/data_writes/ 50.001142480 1,125,899,906,621,494.75 MiB uncore_imc/data_reads/ 50.001142480 6.97 MiB uncore_imc/data_writes/ The client IMC events are freerunning counters. They still use the old event encoding format (0x1 for data_read and 0x2 for data write). The counter bit width is calculated by common code, which assume that the standard encoding format is used for the freerunning counters. Error bit width information is calculated. The patch intends to convert the old client IMC event encoding to the standard encoding format. Current common code uses event->attr.config which directly copy from user space. We should not implicitly modify it for a converted event. The event->hw.config is used to replace the event->attr.config in common code. For client IMC events, the event->attr.config is used to calculate a converted event with standard encoding format in the custom event_init(). The converted event is stored in event->hw.config. For other events of freerunning counters, they already use the standard encoding format. The same value as event->attr.config is assigned to event->hw.config in common event_init(). Reported-by: Jin Yao <yao.jin@linux.intel.com> Tested-by: Jin Yao <yao.jin@linux.intel.com> Signed-off-by: Kan Liang <kan.liang@linux.intel.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Andy Lutomirski <luto@kernel.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Borislav Petkov <bp@alien8.de> Cc: Dave Hansen <dave.hansen@linux.intel.com> Cc: H. Peter Anvin <hpa@zytor.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Rik van Riel <riel@surriel.com> Cc: Stephane Eranian <eranian@google.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Vince Weaver <vincent.weaver@maine.edu> Cc: stable@kernel.org # v4.18+ Fixes: 9aae1780e7e8 ("perf/x86/intel/uncore: Clean up client IMC uncore") Link: https://lkml.kernel.org/r/20190227165729.1861-1-kan.liang@linux.intel.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2019-02-28 00:57:29 +08:00
unsigned int type = uncore_freerunning_type(event->hw.config);
unsigned int idx = uncore_freerunning_idx(event->hw.config);
perf/x86/intel/uncore: Add new data structures for free running counters There are a number of free running counters introduced for uncore, which provide highly valuable information to a wide array of customers. For example, Skylake Server has IIO free running counters to collect Input/Output x BW/Utilization. There is NO event available on the general purpose counters, that is exactly the same as the free running counters. The generic uncore code needs to be enhanced to support the new counters. In the uncore document, there is no event-code assigned to free running counters. Some events need to be defined to indicate the free running counters. The events are encoded as event-code + umask-code. The event-code for all free running counters is 0xff, which is the same as the fixed counters: - It has not been decided what code will be used for common events on future platforms. 0xff is the only one which will definitely not be used as any common event-code. - Cannot re-use current events on the general purpose counters. Because there is NO event available, that is exactly the same as the free running counters. - Even in the existing codes, the fixed counters for core, that have the same event-code, may count different things. Hence, it should not surprise the users if the free running counters that share the same event-code also count different things. Umask will be used to distinguish the counters. The umask-code is used to distinguish a fixed counter and a free running counter, and different types of free running counters. For fixed counters, the umask-code is 0x0X, where X indicates the index of the fixed counter, which starts from 0. - Compatible with the old event encoding. - Currently, there is only one fixed counter. There are still 15 reserved spaces for extension. For free running counters, the umask-code uses the rest of the space. It would follow the format of 0xXY: - X stands for the type of free running counters, which starts from 1. - Y stands for the index of free running counters of same type, which starts from 0. - The free running counters do different thing. It can be categorized to several types, according to the MSR location, bit width and definition. E.g. there are three types of IIO free running counters on Skylake server to monitor IO CLOCKS, BANDWIDTH and UTILIZATION on different ports. It makes it easy to locate the free running counter of a specific type. - So far, there are at most 8 counters of each type. There are still 8 reserved spaces for extension. Introducing a new index to indicate the free running counters. Only one index is enough for all free running counters. Because the free running counters are always active, and the event and free running counter are always 1:1 mapped, it does not need extra index to indicate the assigned counter. Introducing a new data structure to store free running counters related information for each type. It includes the number of counters, bit width, base address, offset between counters and offset between boxes. Introducing several inline helpers to check index for fixed counter and free running counter, validate free running counter event, and retrieve the free running counter information according to box and event. Signed-off-by: Kan Liang <kan.liang@intel.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Reviewed-by: Thomas Gleixner <tglx@linutronix.de> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: acme@kernel.org Cc: eranian@google.com Link: http://lkml.kernel.org/r/1525371913-10597-4-git-send-email-kan.liang@intel.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2018-05-04 02:25:09 +08:00
return (type < uncore_num_freerunning_types(box, event)) &&
(idx < uncore_num_freerunning(box, event));
}
static inline int uncore_num_counters(struct intel_uncore_box *box)
{
return box->pmu->type->num_counters;
}
perf/x86/intel/uncore: Add new data structures for free running counters There are a number of free running counters introduced for uncore, which provide highly valuable information to a wide array of customers. For example, Skylake Server has IIO free running counters to collect Input/Output x BW/Utilization. There is NO event available on the general purpose counters, that is exactly the same as the free running counters. The generic uncore code needs to be enhanced to support the new counters. In the uncore document, there is no event-code assigned to free running counters. Some events need to be defined to indicate the free running counters. The events are encoded as event-code + umask-code. The event-code for all free running counters is 0xff, which is the same as the fixed counters: - It has not been decided what code will be used for common events on future platforms. 0xff is the only one which will definitely not be used as any common event-code. - Cannot re-use current events on the general purpose counters. Because there is NO event available, that is exactly the same as the free running counters. - Even in the existing codes, the fixed counters for core, that have the same event-code, may count different things. Hence, it should not surprise the users if the free running counters that share the same event-code also count different things. Umask will be used to distinguish the counters. The umask-code is used to distinguish a fixed counter and a free running counter, and different types of free running counters. For fixed counters, the umask-code is 0x0X, where X indicates the index of the fixed counter, which starts from 0. - Compatible with the old event encoding. - Currently, there is only one fixed counter. There are still 15 reserved spaces for extension. For free running counters, the umask-code uses the rest of the space. It would follow the format of 0xXY: - X stands for the type of free running counters, which starts from 1. - Y stands for the index of free running counters of same type, which starts from 0. - The free running counters do different thing. It can be categorized to several types, according to the MSR location, bit width and definition. E.g. there are three types of IIO free running counters on Skylake server to monitor IO CLOCKS, BANDWIDTH and UTILIZATION on different ports. It makes it easy to locate the free running counter of a specific type. - So far, there are at most 8 counters of each type. There are still 8 reserved spaces for extension. Introducing a new index to indicate the free running counters. Only one index is enough for all free running counters. Because the free running counters are always active, and the event and free running counter are always 1:1 mapped, it does not need extra index to indicate the assigned counter. Introducing a new data structure to store free running counters related information for each type. It includes the number of counters, bit width, base address, offset between counters and offset between boxes. Introducing several inline helpers to check index for fixed counter and free running counter, validate free running counter event, and retrieve the free running counter information according to box and event. Signed-off-by: Kan Liang <kan.liang@intel.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Reviewed-by: Thomas Gleixner <tglx@linutronix.de> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: acme@kernel.org Cc: eranian@google.com Link: http://lkml.kernel.org/r/1525371913-10597-4-git-send-email-kan.liang@intel.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2018-05-04 02:25:09 +08:00
static inline bool is_freerunning_event(struct perf_event *event)
{
u64 cfg = event->attr.config;
return ((cfg & UNCORE_FIXED_EVENT) == UNCORE_FIXED_EVENT) &&
(((cfg >> 8) & 0xff) >= UNCORE_FREERUNNING_UMASK_START);
}
/* Check and reject invalid config */
static inline int uncore_freerunning_hw_config(struct intel_uncore_box *box,
struct perf_event *event)
{
if (is_freerunning_event(event))
return 0;
return -EINVAL;
}
static inline void uncore_disable_event(struct intel_uncore_box *box,
struct perf_event *event)
{
box->pmu->type->ops->disable_event(box, event);
}
static inline void uncore_enable_event(struct intel_uncore_box *box,
struct perf_event *event)
{
box->pmu->type->ops->enable_event(box, event);
}
static inline u64 uncore_read_counter(struct intel_uncore_box *box,
struct perf_event *event)
{
return box->pmu->type->ops->read_counter(box, event);
}
Revert "perf/x86/intel/uncore: Move uncore_box_init() out of driver initialization" This reverts commit c05199e5a57a579fea1e8fa65e2b511ceb524ffc. Vince Weaver reported the following crash while perf fuzzing: [ 79.473121] kernel BUG at mm/vmalloc.c:1335! [ 79.694391] Call Trace: [ 79.696997] <IRQ> [ 79.699090] [<ffffffff811b2130>] get_vm_area_caller+0x40/0x50 [ 79.705505] [<ffffffff81039f4d>] ? snb_uncore_imc_init_box+0x6d/0x90 [ 79.712414] [<ffffffff810635e5>] __ioremap_caller+0x195/0x350 [ 79.718610] [<ffffffff81039f4d>] ? snb_uncore_imc_init_box+0x6d/0x90 [ 79.725462] [<ffffffff81427f6b>] ? debug_object_activate+0x14b/0x1e0 [ 79.732346] [<ffffffff810637b7>] ioremap_nocache+0x17/0x20 [ 79.738283] [<ffffffff81039f4d>] snb_uncore_imc_init_box+0x6d/0x90 [ 79.744945] [<ffffffff81039cf7>] snb_uncore_imc_event_start+0xb7/0x110 [ 79.752020] [<ffffffff81039d97>] snb_uncore_imc_event_add+0x47/0x60 [ 79.758832] [<ffffffff81162cbb>] event_sched_in.isra.85+0xfb/0x330 [ 79.765519] [<ffffffff81162f5f>] group_sched_in+0x6f/0x1e0 [ 79.771481] [<ffffffff8101df1a>] ? native_sched_clock+0x2a/0x90 [ 79.777858] [<ffffffff811637bc>] __perf_event_enable+0x25c/0x2a0 [ 79.784418] [<ffffffff810f3e69>] ? tick_nohz_irq_exit+0x29/0x30 [ 79.790820] [<ffffffff8115ef30>] ? cpu_clock_event_start+0x40/0x40 [ 79.797546] [<ffffffff8115ef80>] remote_function+0x50/0x60 [ 79.803535] [<ffffffff810f8cd1>] flush_smp_call_function_queue+0x81/0x180 [ 79.810840] [<ffffffff810f9763>] generic_smp_call_function_single_interrupt+0x13/0x60 [ 79.819328] [<ffffffff8104b5e8>] smp_trace_call_function_single_interrupt+0x38/0xc0 [ 79.827614] [<ffffffff816de9be>] trace_call_function_single_interrupt+0x6e/0x80 [ 79.835465] <EOI> [ 79.837543] [<ffffffff8156e8b5>] ? cpuidle_enter_state+0x65/0x160 [ 79.844377] [<ffffffff8156e8a1>] ? cpuidle_enter_state+0x51/0x160 [ 79.851015] [<ffffffff8156e9e7>] cpuidle_enter+0x17/0x20 [ 79.856791] [<ffffffff810b6e39>] cpu_startup_entry+0x399/0x440 [ 79.863165] [<ffffffff816c9ddb>] rest_init+0xbb/0xd0 The offending commit is clearly confused as it moves heavy initialization work into IPI context. Revert it. Reported-by: Vince Weaver <vincent.weaver@maine.edu> Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Cc: Kan Liang <kan.liang@intel.com> Cc: Arnaldo Carvalho de Melo <acme@kernel.org> Cc: Stephane Eranian <eranian@google.com> Cc: Yan, Zheng <zheng.z.yan@intel.com> Cc: linux-kernel@vger.kernel.org Signed-off-by: Ingo Molnar <mingo@kernel.org>
2015-06-09 17:40:28 +08:00
static inline void uncore_box_init(struct intel_uncore_box *box)
{
if (!test_and_set_bit(UNCORE_BOX_FLAG_INITIATED, &box->flags)) {
if (box->pmu->type->ops->init_box)
box->pmu->type->ops->init_box(box);
}
}
static inline void uncore_box_exit(struct intel_uncore_box *box)
{
if (test_and_clear_bit(UNCORE_BOX_FLAG_INITIATED, &box->flags)) {
if (box->pmu->type->ops->exit_box)
box->pmu->type->ops->exit_box(box);
}
}
static inline bool uncore_box_is_fake(struct intel_uncore_box *box)
{
return (box->dieid < 0);
}
static inline struct intel_uncore_pmu *uncore_event_to_pmu(struct perf_event *event)
{
return container_of(event->pmu, struct intel_uncore_pmu, pmu);
}
static inline struct intel_uncore_box *uncore_event_to_box(struct perf_event *event)
{
return event->pmu_private;
}
struct intel_uncore_box *uncore_pmu_to_box(struct intel_uncore_pmu *pmu, int cpu);
u64 uncore_msr_read_counter(struct intel_uncore_box *box, struct perf_event *event);
void uncore_mmio_exit_box(struct intel_uncore_box *box);
u64 uncore_mmio_read_counter(struct intel_uncore_box *box,
struct perf_event *event);
void uncore_pmu_start_hrtimer(struct intel_uncore_box *box);
void uncore_pmu_cancel_hrtimer(struct intel_uncore_box *box);
void uncore_pmu_event_start(struct perf_event *event, int flags);
void uncore_pmu_event_stop(struct perf_event *event, int flags);
int uncore_pmu_event_add(struct perf_event *event, int flags);
void uncore_pmu_event_del(struct perf_event *event, int flags);
void uncore_pmu_event_read(struct perf_event *event);
void uncore_perf_event_update(struct intel_uncore_box *box, struct perf_event *event);
struct event_constraint *
uncore_get_constraint(struct intel_uncore_box *box, struct perf_event *event);
void uncore_put_constraint(struct intel_uncore_box *box, struct perf_event *event);
u64 uncore_shared_reg_config(struct intel_uncore_box *box, int idx);
extern struct intel_uncore_type **uncore_msr_uncores;
extern struct intel_uncore_type **uncore_pci_uncores;
extern struct intel_uncore_type **uncore_mmio_uncores;
extern struct pci_driver *uncore_pci_driver;
extern raw_spinlock_t pci2phy_map_lock;
extern struct list_head pci2phy_map_head;
extern struct pci_extra_dev *uncore_extra_pci_dev;
extern struct event_constraint uncore_constraint_empty;
/* uncore_snb.c */
int snb_uncore_pci_init(void);
int ivb_uncore_pci_init(void);
int hsw_uncore_pci_init(void);
int bdw_uncore_pci_init(void);
int skl_uncore_pci_init(void);
void snb_uncore_cpu_init(void);
void nhm_uncore_cpu_init(void);
void skl_uncore_cpu_init(void);
void icl_uncore_cpu_init(void);
perf/x86/intel/uncore: Add Knights Landing uncore PMU support Knights Landing uncore performance monitoring (perfmon) is derived from Haswell-EP uncore perfmon with several differences. One notable difference is in PCI device IDs. Knights Landing uses common PCI device ID for multiple instances of an uncore PMU device type. In Haswell-EP, each instance of a PMU device type has a unique device ID. Knights Landing uncore components that have performance monitoring units are UBOX, CHA, EDC, MC, M2PCIe, IRP and PCU. Perfmon registers in EDC, MC, IRP, and M2PCIe reside in the PCIe configuration space. Perfmon registers in UBOX, CHA and PCU are accessed via the MSR interface. For more details, please refer to the public document: https://software.intel.com/sites/default/files/managed/15/8d/IntelXeonPhi%E2%84%A2x200ProcessorPerformanceMonitoringReferenceManual_Volume1_Registers_v0%206.pdf Signed-off-by: Harish Chegondi <harish.chegondi@intel.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Cc: Andi Kleen <andi.kleen@intel.com> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Harish Chegondi <harish.chegondi@gmail.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Kan Liang <kan.liang@intel.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Lukasz Anaczkowski <lukasz.anaczkowski@intel.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stephane Eranian <eranian@google.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Vince Weaver <vincent.weaver@maine.edu> Link: http://lkml.kernel.org/r/8ac513981264c3eb10343a3f523f19cc5a2d12fe.1449470704.git.harish.chegondi@intel.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2015-12-08 06:32:32 +08:00
int snb_pci2phy_map_init(int devid);
/* uncore_snbep.c */
int snbep_uncore_pci_init(void);
void snbep_uncore_cpu_init(void);
int ivbep_uncore_pci_init(void);
void ivbep_uncore_cpu_init(void);
int hswep_uncore_pci_init(void);
void hswep_uncore_cpu_init(void);
int bdx_uncore_pci_init(void);
void bdx_uncore_cpu_init(void);
perf/x86/intel/uncore: Add Knights Landing uncore PMU support Knights Landing uncore performance monitoring (perfmon) is derived from Haswell-EP uncore perfmon with several differences. One notable difference is in PCI device IDs. Knights Landing uses common PCI device ID for multiple instances of an uncore PMU device type. In Haswell-EP, each instance of a PMU device type has a unique device ID. Knights Landing uncore components that have performance monitoring units are UBOX, CHA, EDC, MC, M2PCIe, IRP and PCU. Perfmon registers in EDC, MC, IRP, and M2PCIe reside in the PCIe configuration space. Perfmon registers in UBOX, CHA and PCU are accessed via the MSR interface. For more details, please refer to the public document: https://software.intel.com/sites/default/files/managed/15/8d/IntelXeonPhi%E2%84%A2x200ProcessorPerformanceMonitoringReferenceManual_Volume1_Registers_v0%206.pdf Signed-off-by: Harish Chegondi <harish.chegondi@intel.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Cc: Andi Kleen <andi.kleen@intel.com> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Harish Chegondi <harish.chegondi@gmail.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Kan Liang <kan.liang@intel.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Lukasz Anaczkowski <lukasz.anaczkowski@intel.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stephane Eranian <eranian@google.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Vince Weaver <vincent.weaver@maine.edu> Link: http://lkml.kernel.org/r/8ac513981264c3eb10343a3f523f19cc5a2d12fe.1449470704.git.harish.chegondi@intel.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2015-12-08 06:32:32 +08:00
int knl_uncore_pci_init(void);
void knl_uncore_cpu_init(void);
int skx_uncore_pci_init(void);
void skx_uncore_cpu_init(void);
int snr_uncore_pci_init(void);
void snr_uncore_cpu_init(void);
void snr_uncore_mmio_init(void);
/* uncore_nhmex.c */
void nhmex_uncore_cpu_init(void);