/* * Copyright © 2015-2016 Intel Corporation * * Permission is hereby granted, free of charge, to any person obtaining a * copy of this software and associated documentation files (the "Software"), * to deal in the Software without restriction, including without limitation * the rights to use, copy, modify, merge, publish, distribute, sublicense, * and/or sell copies of the Software, and to permit persons to whom the * Software is furnished to do so, subject to the following conditions: * * The above copyright notice and this permission notice (including the next * paragraph) shall be included in all copies or substantial portions of the * Software. * * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS * IN THE SOFTWARE. * * Authors: * Robert Bragg */ /** * DOC: i915 Perf Overview * * Gen graphics supports a large number of performance counters that can help * driver and application developers understand and optimize their use of the * GPU. * * This i915 perf interface enables userspace to configure and open a file * descriptor representing a stream of GPU metrics which can then be read() as * a stream of sample records. * * The interface is particularly suited to exposing buffered metrics that are * captured by DMA from the GPU, unsynchronized with and unrelated to the CPU. * * Streams representing a single context are accessible to applications with a * corresponding drm file descriptor, such that OpenGL can use the interface * without special privileges. Access to system-wide metrics requires root * privileges by default, unless changed via the dev.i915.perf_event_paranoid * sysctl option. * */ /** * DOC: i915 Perf History and Comparison with Core Perf * * The interface was initially inspired by the core Perf infrastructure but * some notable differences are: * * i915 perf file descriptors represent a "stream" instead of an "event"; where * a perf event primarily corresponds to a single 64bit value, while a stream * might sample sets of tightly-coupled counters, depending on the * configuration. For example the Gen OA unit isn't designed to support * orthogonal configurations of individual counters; it's configured for a set * of related counters. Samples for an i915 perf stream capturing OA metrics * will include a set of counter values packed in a compact HW specific format. * The OA unit supports a number of different packing formats which can be * selected by the user opening the stream. Perf has support for grouping * events, but each event in the group is configured, validated and * authenticated individually with separate system calls. * * i915 perf stream configurations are provided as an array of u64 (key,value) * pairs, instead of a fixed struct with multiple miscellaneous config members, * interleaved with event-type specific members. * * i915 perf doesn't support exposing metrics via an mmap'd circular buffer. * The supported metrics are being written to memory by the GPU unsynchronized * with the CPU, using HW specific packing formats for counter sets. Sometimes * the constraints on HW configuration require reports to be filtered before it * would be acceptable to expose them to unprivileged applications - to hide * the metrics of other processes/contexts. For these use cases a read() based * interface is a good fit, and provides an opportunity to filter data as it * gets copied from the GPU mapped buffers to userspace buffers. * * * Issues hit with first prototype based on Core Perf * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ * * The first prototype of this driver was based on the core perf * infrastructure, and while we did make that mostly work, with some changes to * perf, we found we were breaking or working around too many assumptions baked * into perf's currently cpu centric design. * * In the end we didn't see a clear benefit to making perf's implementation and * interface more complex by changing design assumptions while we knew we still * wouldn't be able to use any existing perf based userspace tools. * * Also considering the Gen specific nature of the Observability hardware and * how userspace will sometimes need to combine i915 perf OA metrics with * side-band OA data captured via MI_REPORT_PERF_COUNT commands; we're * expecting the interface to be used by a platform specific userspace such as * OpenGL or tools. This is to say; we aren't inherently missing out on having * a standard vendor/architecture agnostic interface by not using perf. * * * For posterity, in case we might re-visit trying to adapt core perf to be * better suited to exposing i915 metrics these were the main pain points we * hit: * * - The perf based OA PMU driver broke some significant design assumptions: * * Existing perf pmus are used for profiling work on a cpu and we were * introducing the idea of _IS_DEVICE pmus with different security * implications, the need to fake cpu-related data (such as user/kernel * registers) to fit with perf's current design, and adding _DEVICE records * as a way to forward device-specific status records. * * The OA unit writes reports of counters into a circular buffer, without * involvement from the CPU, making our PMU driver the first of a kind. * * Given the way we were periodically forward data from the GPU-mapped, OA * buffer to perf's buffer, those bursts of sample writes looked to perf like * we were sampling too fast and so we had to subvert its throttling checks. * * Perf supports groups of counters and allows those to be read via * transactions internally but transactions currently seem designed to be * explicitly initiated from the cpu (say in response to a userspace read()) * and while we could pull a report out of the OA buffer we can't * trigger a report from the cpu on demand. * * Related to being report based; the OA counters are configured in HW as a * set while perf generally expects counter configurations to be orthogonal. * Although counters can be associated with a group leader as they are * opened, there's no clear precedent for being able to provide group-wide * configuration attributes (for example we want to let userspace choose the * OA unit report format used to capture all counters in a set, or specify a * GPU context to filter metrics on). We avoided using perf's grouping * feature and forwarded OA reports to userspace via perf's 'raw' sample * field. This suited our userspace well considering how coupled the counters * are when dealing with normalizing. It would be inconvenient to split * counters up into separate events, only to require userspace to recombine * them. For Mesa it's also convenient to be forwarded raw, periodic reports * for combining with the side-band raw reports it captures using * MI_REPORT_PERF_COUNT commands. * * - As a side note on perf's grouping feature; there was also some concern * that using PERF_FORMAT_GROUP as a way to pack together counter values * would quite drastically inflate our sample sizes, which would likely * lower the effective sampling resolutions we could use when the available * memory bandwidth is limited. * * With the OA unit's report formats, counters are packed together as 32 * or 40bit values, with the largest report size being 256 bytes. * * PERF_FORMAT_GROUP values are 64bit, but there doesn't appear to be a * documented ordering to the values, implying PERF_FORMAT_ID must also be * used to add a 64bit ID before each value; giving 16 bytes per counter. * * Related to counter orthogonality; we can't time share the OA unit, while * event scheduling is a central design idea within perf for allowing * userspace to open + enable more events than can be configured in HW at any * one time. The OA unit is not designed to allow re-configuration while in * use. We can't reconfigure the OA unit without losing internal OA unit * state which we can't access explicitly to save and restore. Reconfiguring * the OA unit is also relatively slow, involving ~100 register writes. From * userspace Mesa also depends on a stable OA configuration when emitting * MI_REPORT_PERF_COUNT commands and importantly the OA unit can't be * disabled while there are outstanding MI_RPC commands lest we hang the * command streamer. * * The contents of sample records aren't extensible by device drivers (i.e. * the sample_type bits). As an example; Sourab Gupta had been looking to * attach GPU timestamps to our OA samples. We were shoehorning OA reports * into sample records by using the 'raw' field, but it's tricky to pack more * than one thing into this field because events/core.c currently only lets a * pmu give a single raw data pointer plus len which will be copied into the * ring buffer. To include more than the OA report we'd have to copy the * report into an intermediate larger buffer. I'd been considering allowing a * vector of data+len values to be specified for copying the raw data, but * it felt like a kludge to being using the raw field for this purpose. * * - It felt like our perf based PMU was making some technical compromises * just for the sake of using perf: * * perf_event_open() requires events to either relate to a pid or a specific * cpu core, while our device pmu related to neither. Events opened with a * pid will be automatically enabled/disabled according to the scheduling of * that process - so not appropriate for us. When an event is related to a * cpu id, perf ensures pmu methods will be invoked via an inter process * interrupt on that core. To avoid invasive changes our userspace opened OA * perf events for a specific cpu. This was workable but it meant the * majority of the OA driver ran in atomic context, including all OA report * forwarding, which wasn't really necessary in our case and seems to make * our locking requirements somewhat complex as we handled the interaction * with the rest of the i915 driver. */ #include #include #include #include "gem/i915_gem_context.h" #include "gem/i915_gem_internal.h" #include "gt/intel_engine_pm.h" #include "gt/intel_engine_regs.h" #include "gt/intel_engine_user.h" #include "gt/intel_execlists_submission.h" #include "gt/intel_gpu_commands.h" #include "gt/intel_gt.h" #include "gt/intel_gt_clock_utils.h" #include "gt/intel_gt_regs.h" #include "gt/intel_lrc.h" #include "gt/intel_lrc_reg.h" #include "gt/intel_ring.h" #include "i915_drv.h" #include "i915_file_private.h" #include "i915_perf.h" #include "i915_perf_oa_regs.h" /* HW requires this to be a power of two, between 128k and 16M, though driver * is currently generally designed assuming the largest 16M size is used such * that the overflow cases are unlikely in normal operation. */ #define OA_BUFFER_SIZE SZ_16M #define OA_TAKEN(tail, head) ((tail - head) & (OA_BUFFER_SIZE - 1)) /** * DOC: OA Tail Pointer Race * * There's a HW race condition between OA unit tail pointer register updates and * writes to memory whereby the tail pointer can sometimes get ahead of what's * been written out to the OA buffer so far (in terms of what's visible to the * CPU). * * Although this can be observed explicitly while copying reports to userspace * by checking for a zeroed report-id field in tail reports, we want to account * for this earlier, as part of the oa_buffer_check_unlocked to avoid lots of * redundant read() attempts. * * We workaround this issue in oa_buffer_check_unlocked() by reading the reports * in the OA buffer, starting from the tail reported by the HW until we find a * report with its first 2 dwords not 0 meaning its previous report is * completely in memory and ready to be read. Those dwords are also set to 0 * once read and the whole buffer is cleared upon OA buffer initialization. The * first dword is the reason for this report while the second is the timestamp, * making the chances of having those 2 fields at 0 fairly unlikely. A more * detailed explanation is available in oa_buffer_check_unlocked(). * * Most of the implementation details for this workaround are in * oa_buffer_check_unlocked() and _append_oa_reports() * * Note for posterity: previously the driver used to define an effective tail * pointer that lagged the real pointer by a 'tail margin' measured in bytes * derived from %OA_TAIL_MARGIN_NSEC and the configured sampling frequency. * This was flawed considering that the OA unit may also automatically generate * non-periodic reports (such as on context switch) or the OA unit may be * enabled without any periodic sampling. */ #define OA_TAIL_MARGIN_NSEC 100000ULL #define INVALID_TAIL_PTR 0xffffffff /* The default frequency for checking whether the OA unit has written new * reports to the circular OA buffer... */ #define DEFAULT_POLL_FREQUENCY_HZ 200 #define DEFAULT_POLL_PERIOD_NS (NSEC_PER_SEC / DEFAULT_POLL_FREQUENCY_HZ) /* for sysctl proc_dointvec_minmax of dev.i915.perf_stream_paranoid */ static u32 i915_perf_stream_paranoid = true; /* The maximum exponent the hardware accepts is 63 (essentially it selects one * of the 64bit timestamp bits to trigger reports from) but there's currently * no known use case for sampling as infrequently as once per 47 thousand years. * * Since the timestamps included in OA reports are only 32bits it seems * reasonable to limit the OA exponent where it's still possible to account for * overflow in OA report timestamps. */ #define OA_EXPONENT_MAX 31 #define INVALID_CTX_ID 0xffffffff /* On Gen8+ automatically triggered OA reports include a 'reason' field... */ #define OAREPORT_REASON_MASK 0x3f #define OAREPORT_REASON_MASK_EXTENDED 0x7f #define OAREPORT_REASON_SHIFT 19 #define OAREPORT_REASON_TIMER (1<<0) #define OAREPORT_REASON_CTX_SWITCH (1<<3) #define OAREPORT_REASON_CLK_RATIO (1<<5) /* For sysctl proc_dointvec_minmax of i915_oa_max_sample_rate * * The highest sampling frequency we can theoretically program the OA unit * with is always half the timestamp frequency: E.g. 6.25Mhz for Haswell. * * Initialized just before we register the sysctl parameter. */ static int oa_sample_rate_hard_limit; /* Theoretically we can program the OA unit to sample every 160ns but don't * allow that by default unless root... * * The default threshold of 100000Hz is based on perf's similar * kernel.perf_event_max_sample_rate sysctl parameter. */ static u32 i915_oa_max_sample_rate = 100000; /* XXX: beware if future OA HW adds new report formats that the current * code assumes all reports have a power-of-two size and ~(size - 1) can * be used as a mask to align the OA tail pointer. */ static const struct i915_oa_format oa_formats[I915_OA_FORMAT_MAX] = { [I915_OA_FORMAT_A13] = { 0, 64 }, [I915_OA_FORMAT_A29] = { 1, 128 }, [I915_OA_FORMAT_A13_B8_C8] = { 2, 128 }, /* A29_B8_C8 Disallowed as 192 bytes doesn't factor into buffer size */ [I915_OA_FORMAT_B4_C8] = { 4, 64 }, [I915_OA_FORMAT_A45_B8_C8] = { 5, 256 }, [I915_OA_FORMAT_B4_C8_A16] = { 6, 128 }, [I915_OA_FORMAT_C4_B8] = { 7, 64 }, [I915_OA_FORMAT_A12] = { 0, 64 }, [I915_OA_FORMAT_A12_B8_C8] = { 2, 128 }, [I915_OA_FORMAT_A32u40_A4u32_B8_C8] = { 5, 256 }, }; #define SAMPLE_OA_REPORT (1<<0) /** * struct perf_open_properties - for validated properties given to open a stream * @sample_flags: `DRM_I915_PERF_PROP_SAMPLE_*` properties are tracked as flags * @single_context: Whether a single or all gpu contexts should be monitored * @hold_preemption: Whether the preemption is disabled for the filtered * context * @ctx_handle: A gem ctx handle for use with @single_context * @metrics_set: An ID for an OA unit metric set advertised via sysfs * @oa_format: An OA unit HW report format * @oa_periodic: Whether to enable periodic OA unit sampling * @oa_period_exponent: The OA unit sampling period is derived from this * @engine: The engine (typically rcs0) being monitored by the OA unit * @has_sseu: Whether @sseu was specified by userspace * @sseu: internal SSEU configuration computed either from the userspace * specified configuration in the opening parameters or a default value * (see get_default_sseu_config()) * @poll_oa_period: The period in nanoseconds at which the CPU will check for OA * data availability * * As read_properties_unlocked() enumerates and validates the properties given * to open a stream of metrics the configuration is built up in the structure * which starts out zero initialized. */ struct perf_open_properties { u32 sample_flags; u64 single_context:1; u64 hold_preemption:1; u64 ctx_handle; /* OA sampling state */ int metrics_set; int oa_format; bool oa_periodic; int oa_period_exponent; struct intel_engine_cs *engine; bool has_sseu; struct intel_sseu sseu; u64 poll_oa_period; }; struct i915_oa_config_bo { struct llist_node node; struct i915_oa_config *oa_config; struct i915_vma *vma; }; static struct ctl_table_header *sysctl_header; static enum hrtimer_restart oa_poll_check_timer_cb(struct hrtimer *hrtimer); void i915_oa_config_release(struct kref *ref) { struct i915_oa_config *oa_config = container_of(ref, typeof(*oa_config), ref); kfree(oa_config->flex_regs); kfree(oa_config->b_counter_regs); kfree(oa_config->mux_regs); kfree_rcu(oa_config, rcu); } struct i915_oa_config * i915_perf_get_oa_config(struct i915_perf *perf, int metrics_set) { struct i915_oa_config *oa_config; rcu_read_lock(); oa_config = idr_find(&perf->metrics_idr, metrics_set); if (oa_config) oa_config = i915_oa_config_get(oa_config); rcu_read_unlock(); return oa_config; } static void free_oa_config_bo(struct i915_oa_config_bo *oa_bo) { i915_oa_config_put(oa_bo->oa_config); i915_vma_put(oa_bo->vma); kfree(oa_bo); } static u32 gen12_oa_hw_tail_read(struct i915_perf_stream *stream) { struct intel_uncore *uncore = stream->uncore; return intel_uncore_read(uncore, GEN12_OAG_OATAILPTR) & GEN12_OAG_OATAILPTR_MASK; } static u32 gen8_oa_hw_tail_read(struct i915_perf_stream *stream) { struct intel_uncore *uncore = stream->uncore; return intel_uncore_read(uncore, GEN8_OATAILPTR) & GEN8_OATAILPTR_MASK; } static u32 gen7_oa_hw_tail_read(struct i915_perf_stream *stream) { struct intel_uncore *uncore = stream->uncore; u32 oastatus1 = intel_uncore_read(uncore, GEN7_OASTATUS1); return oastatus1 & GEN7_OASTATUS1_TAIL_MASK; } /** * oa_buffer_check_unlocked - check for data and update tail ptr state * @stream: i915 stream instance * * This is either called via fops (for blocking reads in user ctx) or the poll * check hrtimer (atomic ctx) to check the OA buffer tail pointer and check * if there is data available for userspace to read. * * This function is central to providing a workaround for the OA unit tail * pointer having a race with respect to what data is visible to the CPU. * It is responsible for reading tail pointers from the hardware and giving * the pointers time to 'age' before they are made available for reading. * (See description of OA_TAIL_MARGIN_NSEC above for further details.) * * Besides returning true when there is data available to read() this function * also updates the tail, aging_tail and aging_timestamp in the oa_buffer * object. * * Note: It's safe to read OA config state here unlocked, assuming that this is * only called while the stream is enabled, while the global OA configuration * can't be modified. * * Returns: %true if the OA buffer contains data, else %false */ static bool oa_buffer_check_unlocked(struct i915_perf_stream *stream) { u32 gtt_offset = i915_ggtt_offset(stream->oa_buffer.vma); int report_size = stream->oa_buffer.format_size; unsigned long flags; bool pollin; u32 hw_tail; u64 now; /* We have to consider the (unlikely) possibility that read() errors * could result in an OA buffer reset which might reset the head and * tail state. */ spin_lock_irqsave(&stream->oa_buffer.ptr_lock, flags); hw_tail = stream->perf->ops.oa_hw_tail_read(stream); /* The tail pointer increases in 64 byte increments, * not in report_size steps... */ hw_tail &= ~(report_size - 1); now = ktime_get_mono_fast_ns(); if (hw_tail == stream->oa_buffer.aging_tail && (now - stream->oa_buffer.aging_timestamp) > OA_TAIL_MARGIN_NSEC) { /* If the HW tail hasn't move since the last check and the HW * tail has been aging for long enough, declare it the new * tail. */ stream->oa_buffer.tail = stream->oa_buffer.aging_tail; } else { u32 head, tail, aged_tail; /* NB: The head we observe here might effectively be a little * out of date. If a read() is in progress, the head could be * anywhere between this head and stream->oa_buffer.tail. */ head = stream->oa_buffer.head - gtt_offset; aged_tail = stream->oa_buffer.tail - gtt_offset; hw_tail -= gtt_offset; tail = hw_tail; /* Walk the stream backward until we find a report with dword 0 * & 1 not at 0. Since the circular buffer pointers progress by * increments of 64 bytes and that reports can be up to 256 * bytes long, we can't tell whether a report has fully landed * in memory before the first 2 dwords of the following report * have effectively landed. * * This is assuming that the writes of the OA unit land in * memory in the order they were written to. * If not : (╯°□°）╯︵ ┻━┻ */ while (OA_TAKEN(tail, aged_tail) >= report_size) { u32 *report32 = (void *)(stream->oa_buffer.vaddr + tail); if (report32[0] != 0 || report32[1] != 0) break; tail = (tail - report_size) & (OA_BUFFER_SIZE - 1); } if (OA_TAKEN(hw_tail, tail) > report_size && __ratelimit(&stream->perf->tail_pointer_race)) DRM_NOTE("unlanded report(s) head=0x%x " "tail=0x%x hw_tail=0x%x\n", head, tail, hw_tail); stream->oa_buffer.tail = gtt_offset + tail; stream->oa_buffer.aging_tail = gtt_offset + hw_tail; stream->oa_buffer.aging_timestamp = now; } pollin = OA_TAKEN(stream->oa_buffer.tail - gtt_offset, stream->oa_buffer.head - gtt_offset) >= report_size; spin_unlock_irqrestore(&stream->oa_buffer.ptr_lock, flags); return pollin; } /** * append_oa_status - Appends a status record to a userspace read() buffer. * @stream: An i915-perf stream opened for OA metrics * @buf: destination buffer given by userspace * @count: the number of bytes userspace wants to read * @offset: (inout): the current position for writing into @buf * @type: The kind of status to report to userspace * * Writes a status record (such as `DRM_I915_PERF_RECORD_OA_REPORT_LOST`) * into the userspace read() buffer. * * The @buf @offset will only be updated on success. * * Returns: 0 on success, negative error code on failure. */ static int append_oa_status(struct i915_perf_stream *stream, char __user *buf, size_t count, size_t *offset, enum drm_i915_perf_record_type type) { struct drm_i915_perf_record_header header = { type, 0, sizeof(header) }; if ((count - *offset) < header.size) return -ENOSPC; if (copy_to_user(buf + *offset, &header, sizeof(header))) return -EFAULT; (*offset) += header.size; return 0; } /** * append_oa_sample - Copies single OA report into userspace read() buffer. * @stream: An i915-perf stream opened for OA metrics * @buf: destination buffer given by userspace * @count: the number of bytes userspace wants to read * @offset: (inout): the current position for writing into @buf * @report: A single OA report to (optionally) include as part of the sample * * The contents of a sample are configured through `DRM_I915_PERF_PROP_SAMPLE_*` * properties when opening a stream, tracked as `stream->sample_flags`. This * function copies the requested components of a single sample to the given * read() @buf. * * The @buf @offset will only be updated on success. * * Returns: 0 on success, negative error code on failure. */ static int append_oa_sample(struct i915_perf_stream *stream, char __user *buf, size_t count, size_t *offset, const u8 *report) { int report_size = stream->oa_buffer.format_size; struct drm_i915_perf_record_header header; header.type = DRM_I915_PERF_RECORD_SAMPLE; header.pad = 0; header.size = stream->sample_size; if ((count - *offset) < header.size) return -ENOSPC; buf += *offset; if (copy_to_user(buf, &header, sizeof(header))) return -EFAULT; buf += sizeof(header); if (copy_to_user(buf, report, report_size)) return -EFAULT; (*offset) += header.size; return 0; } /** * gen8_append_oa_reports - Copies all buffered OA reports into * userspace read() buffer. * @stream: An i915-perf stream opened for OA metrics * @buf: destination buffer given by userspace * @count: the number of bytes userspace wants to read * @offset: (inout): the current position for writing into @buf * * Notably any error condition resulting in a short read (-%ENOSPC or * -%EFAULT) will be returned even though one or more records may * have been successfully copied. In this case it's up to the caller * to decide if the error should be squashed before returning to * userspace. * * Note: reports are consumed from the head, and appended to the * tail, so the tail chases the head?... If you think that's mad * and back-to-front you're not alone, but this follows the * Gen PRM naming convention. * * Returns: 0 on success, negative error code on failure. */ static int gen8_append_oa_reports(struct i915_perf_stream *stream, char __user *buf, size_t count, size_t *offset) { struct intel_uncore *uncore = stream->uncore; int report_size = stream->oa_buffer.format_size; u8 *oa_buf_base = stream->oa_buffer.vaddr; u32 gtt_offset = i915_ggtt_offset(stream->oa_buffer.vma); u32 mask = (OA_BUFFER_SIZE - 1); size_t start_offset = *offset; unsigned long flags; u32 head, tail; u32 taken; int ret = 0; if (drm_WARN_ON(&uncore->i915->drm, !stream->enabled)) return -EIO; spin_lock_irqsave(&stream->oa_buffer.ptr_lock, flags); head = stream->oa_buffer.head; tail = stream->oa_buffer.tail; spin_unlock_irqrestore(&stream->oa_buffer.ptr_lock, flags); /* * NB: oa_buffer.head/tail include the gtt_offset which we don't want * while indexing relative to oa_buf_base. */ head -= gtt_offset; tail -= gtt_offset; /* * An out of bounds or misaligned head or tail pointer implies a driver * bug since we validate + align the tail pointers we read from the * hardware and we are in full control of the head pointer which should * only be incremented by multiples of the report size (notably also * all a power of two). */ if (drm_WARN_ONCE(&uncore->i915->drm, head > OA_BUFFER_SIZE || head % report_size || tail > OA_BUFFER_SIZE || tail % report_size, "Inconsistent OA buffer pointers: head = %u, tail = %u\n", head, tail)) return -EIO; for (/* none */; (taken = OA_TAKEN(tail, head)); head = (head + report_size) & mask) { u8 *report = oa_buf_base + head; u32 *report32 = (void *)report; u32 ctx_id; u32 reason; /* * All the report sizes factor neatly into the buffer * size so we never expect to see a report split * between the beginning and end of the buffer. * * Given the initial alignment check a misalignment * here would imply a driver bug that would result * in an overrun. */ if (drm_WARN_ON(&uncore->i915->drm, (OA_BUFFER_SIZE - head) < report_size)) { drm_err(&uncore->i915->drm, "Spurious OA head ptr: non-integral report offset\n"); break; } /* * The reason field includes flags identifying what * triggered this specific report (mostly timer * triggered or e.g. due to a context switch). * * This field is never expected to be zero so we can * check that the report isn't invalid before copying * it to userspace... */ reason = ((report32[0] >> OAREPORT_REASON_SHIFT) & (GRAPHICS_VER(stream->perf->i915) == 12 ? OAREPORT_REASON_MASK_EXTENDED : OAREPORT_REASON_MASK)); ctx_id = report32[2] & stream->specific_ctx_id_mask; /* * Squash whatever is in the CTX_ID field if it's marked as * invalid to be sure we avoid false-positive, single-context * filtering below... * * Note: that we don't clear the valid_ctx_bit so userspace can * understand that the ID has been squashed by the kernel. */ if (!(report32[0] & stream->perf->gen8_valid_ctx_bit) && GRAPHICS_VER(stream->perf->i915) <= 11) ctx_id = report32[2] = INVALID_CTX_ID; /* * NB: For Gen 8 the OA unit no longer supports clock gating * off for a specific context and the kernel can't securely * stop the counters from updating as system-wide / global * values. * * Automatic reports now include a context ID so reports can be * filtered on the cpu but it's not worth trying to * automatically subtract/hide counter progress for other * contexts while filtering since we can't stop userspace * issuing MI_REPORT_PERF_COUNT commands which would still * provide a side-band view of the real values. * * To allow userspace (such as Mesa/GL_INTEL_performance_query) * to normalize counters for a single filtered context then it * needs be forwarded bookend context-switch reports so that it * can track switches in between MI_REPORT_PERF_COUNT commands * and can itself subtract/ignore the progress of counters * associated with other contexts. Note that the hardware * automatically triggers reports when switching to a new * context which are tagged with the ID of the newly active * context. To avoid the complexity (and likely fragility) of * reading ahead while parsing reports to try and minimize * forwarding redundant context switch reports (i.e. between * other, unrelated contexts) we simply elect to forward them * all. * * We don't rely solely on the reason field to identify context * switches since it's not-uncommon for periodic samples to * identify a switch before any 'context switch' report. */ if (!stream->perf->exclusive_stream->ctx || stream->specific_ctx_id == ctx_id || stream->oa_buffer.last_ctx_id == stream->specific_ctx_id || reason & OAREPORT_REASON_CTX_SWITCH) { /* * While filtering for a single context we avoid * leaking the IDs of other contexts. */ if (stream->perf->exclusive_stream->ctx && stream->specific_ctx_id != ctx_id) { report32[2] = INVALID_CTX_ID; } ret = append_oa_sample(stream, buf, count, offset, report); if (ret) break; stream->oa_buffer.last_ctx_id = ctx_id; } /* * Clear out the first 2 dword as a mean to detect unlanded * reports. */ report32[0] = 0; report32[1] = 0; } if (start_offset != *offset) { i915_reg_t oaheadptr; oaheadptr = GRAPHICS_VER(stream->perf->i915) == 12 ? GEN12_OAG_OAHEADPTR : GEN8_OAHEADPTR; spin_lock_irqsave(&stream->oa_buffer.ptr_lock, flags); /* * We removed the gtt_offset for the copy loop above, indexing * relative to oa_buf_base so put back here... */ head += gtt_offset; intel_uncore_write(uncore, oaheadptr, head & GEN12_OAG_OAHEADPTR_MASK); stream->oa_buffer.head = head; spin_unlock_irqrestore(&stream->oa_buffer.ptr_lock, flags); } return ret; } /** * gen8_oa_read - copy status records then buffered OA reports * @stream: An i915-perf stream opened for OA metrics * @buf: destination buffer given by userspace * @count: the number of bytes userspace wants to read * @offset: (inout): the current position for writing into @buf * * Checks OA unit status registers and if necessary appends corresponding * status records for userspace (such as for a buffer full condition) and then * initiate appending any buffered OA reports. * * Updates @offset according to the number of bytes successfully copied into * the userspace buffer. * * NB: some data may be successfully copied to the userspace buffer * even if an error is returned, and this is reflected in the * updated @offset. * * Returns: zero on success or a negative error code */ static int gen8_oa_read(struct i915_perf_stream *stream, char __user *buf, size_t count, size_t *offset) { struct intel_uncore *uncore = stream->uncore; u32 oastatus; i915_reg_t oastatus_reg; int ret; if (drm_WARN_ON(&uncore->i915->drm, !stream->oa_buffer.vaddr)) return -EIO; oastatus_reg = GRAPHICS_VER(stream->perf->i915) == 12 ? GEN12_OAG_OASTATUS : GEN8_OASTATUS; oastatus = intel_uncore_read(uncore, oastatus_reg); /* * We treat OABUFFER_OVERFLOW as a significant error: * * Although theoretically we could handle this more gracefully * sometimes, some Gens don't correctly suppress certain * automatically triggered reports in this condition and so we * have to assume that old reports are now being trampled * over. * * Considering how we don't currently give userspace control * over the OA buffer size and always configure a large 16MB * buffer, then a buffer overflow does anyway likely indicate * that something has gone quite badly wrong. */ if (oastatus & GEN8_OASTATUS_OABUFFER_OVERFLOW) { ret = append_oa_status(stream, buf, count, offset, DRM_I915_PERF_RECORD_OA_BUFFER_LOST); if (ret) return ret; drm_dbg(&stream->perf->i915->drm, "OA buffer overflow (exponent = %d): force restart\n", stream->period_exponent); stream->perf->ops.oa_disable(stream); stream->perf->ops.oa_enable(stream); /* * Note: .oa_enable() is expected to re-init the oabuffer and * reset GEN8_OASTATUS for us */ oastatus = intel_uncore_read(uncore, oastatus_reg); } if (oastatus & GEN8_OASTATUS_REPORT_LOST) { ret = append_oa_status(stream, buf, count, offset, DRM_I915_PERF_RECORD_OA_REPORT_LOST); if (ret) return ret; intel_uncore_rmw(uncore, oastatus_reg, GEN8_OASTATUS_COUNTER_OVERFLOW | GEN8_OASTATUS_REPORT_LOST, IS_GRAPHICS_VER(uncore->i915, 8, 11) ? (GEN8_OASTATUS_HEAD_POINTER_WRAP | GEN8_OASTATUS_TAIL_POINTER_WRAP) : 0); } return gen8_append_oa_reports(stream, buf, count, offset); } /** * gen7_append_oa_reports - Copies all buffered OA reports into * userspace read() buffer. * @stream: An i915-perf stream opened for OA metrics * @buf: destination buffer given by userspace * @count: the number of bytes userspace wants to read * @offset: (inout): the current position for writing into @buf * * Notably any error condition resulting in a short read (-%ENOSPC or * -%EFAULT) will be returned even though one or more records may * have been successfully copied. In this case it's up to the caller * to decide if the error should be squashed before returning to * userspace. * * Note: reports are consumed from the head, and appended to the * tail, so the tail chases the head?... If you think that's mad * and back-to-front you're not alone, but this follows the * Gen PRM naming convention. * * Returns: 0 on success, negative error code on failure. */ static int gen7_append_oa_reports(struct i915_perf_stream *stream, char __user *buf, size_t count, size_t *offset) { struct intel_uncore *uncore = stream->uncore; int report_size = stream->oa_buffer.format_size; u8 *oa_buf_base = stream->oa_buffer.vaddr; u32 gtt_offset = i915_ggtt_offset(stream->oa_buffer.vma); u32 mask = (OA_BUFFER_SIZE - 1); size_t start_offset = *offset; unsigned long flags; u32 head, tail; u32 taken; int ret = 0; if (drm_WARN_ON(&uncore->i915->drm, !stream->enabled)) return -EIO; spin_lock_irqsave(&stream->oa_buffer.ptr_lock, flags); head = stream->oa_buffer.head; tail = stream->oa_buffer.tail; spin_unlock_irqrestore(&stream->oa_buffer.ptr_lock, flags); /* NB: oa_buffer.head/tail include the gtt_offset which we don't want * while indexing relative to oa_buf_base. */ head -= gtt_offset; tail -= gtt_offset; /* An out of bounds or misaligned head or tail pointer implies a driver * bug since we validate + align the tail pointers we read from the * hardware and we are in full control of the head pointer which should * only be incremented by multiples of the report size (notably also * all a power of two). */ if (drm_WARN_ONCE(&uncore->i915->drm, head > OA_BUFFER_SIZE || head % report_size || tail > OA_BUFFER_SIZE || tail % report_size, "Inconsistent OA buffer pointers: head = %u, tail = %u\n", head, tail)) return -EIO; for (/* none */; (taken = OA_TAKEN(tail, head)); head = (head + report_size) & mask) { u8 *report = oa_buf_base + head; u32 *report32 = (void *)report; /* All the report sizes factor neatly into the buffer * size so we never expect to see a report split * between the beginning and end of the buffer. * * Given the initial alignment check a misalignment * here would imply a driver bug that would result * in an overrun. */ if (drm_WARN_ON(&uncore->i915->drm, (OA_BUFFER_SIZE - head) < report_size)) { drm_err(&uncore->i915->drm, "Spurious OA head ptr: non-integral report offset\n"); break; } /* The report-ID field for periodic samples includes * some undocumented flags related to what triggered * the report and is never expected to be zero so we * can check that the report isn't invalid before * copying it to userspace... */ if (report32[0] == 0) { if (__ratelimit(&stream->perf->spurious_report_rs)) DRM_NOTE("Skipping spurious, invalid OA report\n"); continue; } ret = append_oa_sample(stream, buf, count, offset, report); if (ret) break; /* Clear out the first 2 dwords as a mean to detect unlanded * reports. */ report32[0] = 0; report32[1] = 0; } if (start_offset != *offset) { spin_lock_irqsave(&stream->oa_buffer.ptr_lock, flags); /* We removed the gtt_offset for the copy loop above, indexing * relative to oa_buf_base so put back here... */ head += gtt_offset; intel_uncore_write(uncore, GEN7_OASTATUS2, (head & GEN7_OASTATUS2_HEAD_MASK) | GEN7_OASTATUS2_MEM_SELECT_GGTT); stream->oa_buffer.head = head; spin_unlock_irqrestore(&stream->oa_buffer.ptr_lock, flags); } return ret; } /** * gen7_oa_read - copy status records then buffered OA reports * @stream: An i915-perf stream opened for OA metrics * @buf: destination buffer given by userspace * @count: the number of bytes userspace wants to read * @offset: (inout): the current position for writing into @buf * * Checks Gen 7 specific OA unit status registers and if necessary appends * corresponding status records for userspace (such as for a buffer full * condition) and then initiate appending any buffered OA reports. * * Updates @offset according to the number of bytes successfully copied into * the userspace buffer. * * Returns: zero on success or a negative error code */ static int gen7_oa_read(struct i915_perf_stream *stream, char __user *buf, size_t count, size_t *offset) { struct intel_uncore *uncore = stream->uncore; u32 oastatus1; int ret; if (drm_WARN_ON(&uncore->i915->drm, !stream->oa_buffer.vaddr)) return -EIO; oastatus1 = intel_uncore_read(uncore, GEN7_OASTATUS1); /* XXX: On Haswell we don't have a safe way to clear oastatus1 * bits while the OA unit is enabled (while the tail pointer * may be updated asynchronously) so we ignore status bits * that have already been reported to userspace. */ oastatus1 &= ~stream->perf->gen7_latched_oastatus1; /* We treat OABUFFER_OVERFLOW as a significant error: * * - The status can be interpreted to mean that the buffer is * currently full (with a higher precedence than OA_TAKEN() * which will start to report a near-empty buffer after an * overflow) but it's awkward that we can't clear the status * on Haswell, so without a reset we won't be able to catch * the state again. * * - Since it also implies the HW has started overwriting old * reports it may also affect our sanity checks for invalid * reports when copying to userspace that assume new reports * are being written to cleared memory. * * - In the future we may want to introduce a flight recorder * mode where the driver will automatically maintain a safe * guard band between head/tail, avoiding this overflow * condition, but we avoid the added driver complexity for * now. */ if (unlikely(oastatus1 & GEN7_OASTATUS1_OABUFFER_OVERFLOW)) { ret = append_oa_status(stream, buf, count, offset, DRM_I915_PERF_RECORD_OA_BUFFER_LOST); if (ret) return ret; drm_dbg(&stream->perf->i915->drm, "OA buffer overflow (exponent = %d): force restart\n", stream->period_exponent); stream->perf->ops.oa_disable(stream); stream->perf->ops.oa_enable(stream); oastatus1 = intel_uncore_read(uncore, GEN7_OASTATUS1); } if (unlikely(oastatus1 & GEN7_OASTATUS1_REPORT_LOST)) { ret = append_oa_status(stream, buf, count, offset, DRM_I915_PERF_RECORD_OA_REPORT_LOST); if (ret) return ret; stream->perf->gen7_latched_oastatus1 |= GEN7_OASTATUS1_REPORT_LOST; } return gen7_append_oa_reports(stream, buf, count, offset); } /** * i915_oa_wait_unlocked - handles blocking IO until OA data available * @stream: An i915-perf stream opened for OA metrics * * Called when userspace tries to read() from a blocking stream FD opened * for OA metrics. It waits until the hrtimer callback finds a non-empty * OA buffer and wakes us. * * Note: it's acceptable to have this return with some false positives * since any subsequent read handling will return -EAGAIN if there isn't * really data ready for userspace yet. * * Returns: zero on success or a negative error code */ static int i915_oa_wait_unlocked(struct i915_perf_stream *stream) { /* We would wait indefinitely if periodic sampling is not enabled */ if (!stream->periodic) return -EIO; return wait_event_interruptible(stream->poll_wq, oa_buffer_check_unlocked(stream)); } /** * i915_oa_poll_wait - call poll_wait() for an OA stream poll() * @stream: An i915-perf stream opened for OA metrics * @file: An i915 perf stream file * @wait: poll() state table * * For handling userspace polling on an i915 perf stream opened for OA metrics, * this starts a poll_wait with the wait queue that our hrtimer callback wakes * when it sees data ready to read in the circular OA buffer. */ static void i915_oa_poll_wait(struct i915_perf_stream *stream, struct file *file, poll_table *wait) { poll_wait(file, &stream->poll_wq, wait); } /** * i915_oa_read - just calls through to &i915_oa_ops->read * @stream: An i915-perf stream opened for OA metrics * @buf: destination buffer given by userspace * @count: the number of bytes userspace wants to read * @offset: (inout): the current position for writing into @buf * * Updates @offset according to the number of bytes successfully copied into * the userspace buffer. * * Returns: zero on success or a negative error code */ static int i915_oa_read(struct i915_perf_stream *stream, char __user *buf, size_t count, size_t *offset) { return stream->perf->ops.read(stream, buf, count, offset); } static struct intel_context *oa_pin_context(struct i915_perf_stream *stream) { struct i915_gem_engines_iter it; struct i915_gem_context *ctx = stream->ctx; struct intel_context *ce; struct i915_gem_ww_ctx ww; int err = -ENODEV; for_each_gem_engine(ce, i915_gem_context_lock_engines(ctx), it) { if (ce->engine != stream->engine) /* first match! */ continue; err = 0; break; } i915_gem_context_unlock_engines(ctx); if (err) return ERR_PTR(err); i915_gem_ww_ctx_init(&ww, true); retry: /* * As the ID is the gtt offset of the context's vma we * pin the vma to ensure the ID remains fixed. */ err = intel_context_pin_ww(ce, &ww); if (err == -EDEADLK) { err = i915_gem_ww_ctx_backoff(&ww); if (!err) goto retry; } i915_gem_ww_ctx_fini(&ww); if (err) return ERR_PTR(err); stream->pinned_ctx = ce; return stream->pinned_ctx; } /** * oa_get_render_ctx_id - determine and hold ctx hw id * @stream: An i915-perf stream opened for OA metrics * * Determine the render context hw id, and ensure it remains fixed for the * lifetime of the stream. This ensures that we don't have to worry about * updating the context ID in OACONTROL on the fly. * * Returns: zero on success or a negative error code */ static int oa_get_render_ctx_id(struct i915_perf_stream *stream) { struct intel_context *ce; ce = oa_pin_context(stream); if (IS_ERR(ce)) return PTR_ERR(ce); switch (GRAPHICS_VER(ce->engine->i915)) { case 7: { /* * On Haswell we don't do any post processing of the reports * and don't need to use the mask. */ stream->specific_ctx_id = i915_ggtt_offset(ce->state); stream->specific_ctx_id_mask = 0; break; } case 8: case 9: if (intel_engine_uses_guc(ce->engine)) { /* * When using GuC, the context descriptor we write in * i915 is read by GuC and rewritten before it's * actually written into the hardware. The LRCA is * what is put into the context id field of the * context descriptor by GuC. Because it's aligned to * a page, the lower 12bits are always at 0 and * dropped by GuC. They won't be part of the context * ID in the OA reports, so squash those lower bits. */ stream->specific_ctx_id = ce->lrc.lrca >> 12; /* * GuC uses the top bit to signal proxy submission, so * ignore that bit. */ stream->specific_ctx_id_mask = (1U << (GEN8_CTX_ID_WIDTH - 1)) - 1; } else { stream->specific_ctx_id_mask = (1U << GEN8_CTX_ID_WIDTH) - 1; stream->specific_ctx_id = stream->specific_ctx_id_mask; } break; case 11: case 12: if (GRAPHICS_VER_FULL(ce->engine->i915) >= IP_VER(12, 50)) { stream->specific_ctx_id_mask = ((1U << XEHP_SW_CTX_ID_WIDTH) - 1) << (XEHP_SW_CTX_ID_SHIFT - 32); stream->specific_ctx_id = (XEHP_MAX_CONTEXT_HW_ID - 1) << (XEHP_SW_CTX_ID_SHIFT - 32); } else { stream->specific_ctx_id_mask = ((1U << GEN11_SW_CTX_ID_WIDTH) - 1) << (GEN11_SW_CTX_ID_SHIFT - 32); /* * Pick an unused context id * 0 - BITS_PER_LONG are used by other contexts * GEN12_MAX_CONTEXT_HW_ID (0x7ff) is used by idle context */ stream->specific_ctx_id = (GEN12_MAX_CONTEXT_HW_ID - 1) << (GEN11_SW_CTX_ID_SHIFT - 32); } break; default: MISSING_CASE(GRAPHICS_VER(ce->engine->i915)); } ce->tag = stream->specific_ctx_id; drm_dbg(&stream->perf->i915->drm, "filtering on ctx_id=0x%x ctx_id_mask=0x%x\n", stream->specific_ctx_id, stream->specific_ctx_id_mask); return 0; } /** * oa_put_render_ctx_id - counterpart to oa_get_render_ctx_id releases hold * @stream: An i915-perf stream opened for OA metrics * * In case anything needed doing to ensure the context HW ID would remain valid * for the lifetime of the stream, then that can be undone here. */ static void oa_put_render_ctx_id(struct i915_perf_stream *stream) { struct intel_context *ce; ce = fetch_and_zero(&stream->pinned_ctx); if (ce) { ce->tag = 0; /* recomputed on next submission after parking */ intel_context_unpin(ce); } stream->specific_ctx_id = INVALID_CTX_ID; stream->specific_ctx_id_mask = 0; } static void free_oa_buffer(struct i915_perf_stream *stream) { i915_vma_unpin_and_release(&stream->oa_buffer.vma, I915_VMA_RELEASE_MAP); stream->oa_buffer.vaddr = NULL; } static void free_oa_configs(struct i915_perf_stream *stream) { struct i915_oa_config_bo *oa_bo, *tmp; i915_oa_config_put(stream->oa_config); llist_for_each_entry_safe(oa_bo, tmp, stream->oa_config_bos.first, node) free_oa_config_bo(oa_bo); } static void free_noa_wait(struct i915_perf_stream *stream) { i915_vma_unpin_and_release(&stream->noa_wait, 0); } static void i915_oa_stream_destroy(struct i915_perf_stream *stream) { struct i915_perf *perf = stream->perf; if (WARN_ON(stream != perf->exclusive_stream)) return; /* * Unset exclusive_stream first, it will be checked while disabling * the metric set on gen8+. * * See i915_oa_init_reg_state() and lrc_configure_all_contexts() */ WRITE_ONCE(perf->exclusive_stream, NULL); perf->ops.disable_metric_set(stream); free_oa_buffer(stream); intel_uncore_forcewake_put(stream->uncore, FORCEWAKE_ALL); intel_engine_pm_put(stream->engine); if (stream->ctx) oa_put_render_ctx_id(stream); free_oa_configs(stream); free_noa_wait(stream); if (perf->spurious_report_rs.missed) { DRM_NOTE("%d spurious OA report notices suppressed due to ratelimiting\n", perf->spurious_report_rs.missed); } } static void gen7_init_oa_buffer(struct i915_perf_stream *stream) { struct intel_uncore *uncore = stream->uncore; u32 gtt_offset = i915_ggtt_offset(stream->oa_buffer.vma); unsigned long flags; spin_lock_irqsave(&stream->oa_buffer.ptr_lock, flags); /* Pre-DevBDW: OABUFFER must be set with counters off, * before OASTATUS1, but after OASTATUS2 */ intel_uncore_write(uncore, GEN7_OASTATUS2, /* head */ gtt_offset | GEN7_OASTATUS2_MEM_SELECT_GGTT); stream->oa_buffer.head = gtt_offset; intel_uncore_write(uncore, GEN7_OABUFFER, gtt_offset); intel_uncore_write(uncore, GEN7_OASTATUS1, /* tail */ gtt_offset | OABUFFER_SIZE_16M); /* Mark that we need updated tail pointers to read from... */ stream->oa_buffer.aging_tail = INVALID_TAIL_PTR; stream->oa_buffer.tail = gtt_offset; spin_unlock_irqrestore(&stream->oa_buffer.ptr_lock, flags); /* On Haswell we have to track which OASTATUS1 flags we've * already seen since they can't be cleared while periodic * sampling is enabled. */ stream->perf->gen7_latched_oastatus1 = 0; /* NB: although the OA buffer will initially be allocated * zeroed via shmfs (and so this memset is redundant when * first allocating), we may re-init the OA buffer, either * when re-enabling a stream or in error/reset paths. * * The reason we clear the buffer for each re-init is for the * sanity check in gen7_append_oa_reports() that looks at the * report-id field to make sure it's non-zero which relies on * the assumption that new reports are being written to zeroed * memory... */ memset(stream->oa_buffer.vaddr, 0, OA_BUFFER_SIZE); } static void gen8_init_oa_buffer(struct i915_perf_stream *stream) { struct intel_uncore *uncore = stream->uncore; u32 gtt_offset = i915_ggtt_offset(stream->oa_buffer.vma); unsigned long flags; spin_lock_irqsave(&stream->oa_buffer.ptr_lock, flags); intel_uncore_write(uncore, GEN8_OASTATUS, 0); intel_uncore_write(uncore, GEN8_OAHEADPTR, gtt_offset); stream->oa_buffer.head = gtt_offset; intel_uncore_write(uncore, GEN8_OABUFFER_UDW, 0); /* * PRM says: * * "This MMIO must be set before the OATAILPTR * register and after the OAHEADPTR register. This is * to enable proper functionality of the overflow * bit." */ intel_uncore_write(uncore, GEN8_OABUFFER, gtt_offset | OABUFFER_SIZE_16M | GEN8_OABUFFER_MEM_SELECT_GGTT); intel_uncore_write(uncore, GEN8_OATAILPTR, gtt_offset & GEN8_OATAILPTR_MASK); /* Mark that we need updated tail pointers to read from... */ stream->oa_buffer.aging_tail = INVALID_TAIL_PTR; stream->oa_buffer.tail = gtt_offset; /* * Reset state used to recognise context switches, affecting which * reports we will forward to userspace while filtering for a single * context. */ stream->oa_buffer.last_ctx_id = INVALID_CTX_ID; spin_unlock_irqrestore(&stream->oa_buffer.ptr_lock, flags); /* * NB: although the OA buffer will initially be allocated * zeroed via shmfs (and so this memset is redundant when * first allocating), we may re-init the OA buffer, either * when re-enabling a stream or in error/reset paths. * * The reason we clear the buffer for each re-init is for the * sanity check in gen8_append_oa_reports() that looks at the * reason field to make sure it's non-zero which relies on * the assumption that new reports are being written to zeroed * memory... */ memset(stream->oa_buffer.vaddr, 0, OA_BUFFER_SIZE); } static void gen12_init_oa_buffer(struct i915_perf_stream *stream) { struct intel_uncore *uncore = stream->uncore; u32 gtt_offset = i915_ggtt_offset(stream->oa_buffer.vma); unsigned long flags; spin_lock_irqsave(&stream->oa_buffer.ptr_lock, flags); intel_uncore_write(uncore, GEN12_OAG_OASTATUS, 0); intel_uncore_write(uncore, GEN12_OAG_OAHEADPTR, gtt_offset & GEN12_OAG_OAHEADPTR_MASK); stream->oa_buffer.head = gtt_offset; /* * PRM says: * * "This MMIO must be set before the OATAILPTR * register and after the OAHEADPTR register. This is * to enable proper functionality of the overflow * bit." */ intel_uncore_write(uncore, GEN12_OAG_OABUFFER, gtt_offset | OABUFFER_SIZE_16M | GEN8_OABUFFER_MEM_SELECT_GGTT); intel_uncore_write(uncore, GEN12_OAG_OATAILPTR, gtt_offset & GEN12_OAG_OATAILPTR_MASK); /* Mark that we need updated tail pointers to read from... */ stream->oa_buffer.aging_tail = INVALID_TAIL_PTR; stream->oa_buffer.tail = gtt_offset; /* * Reset state used to recognise context switches, affecting which * reports we will forward to userspace while filtering for a single * context. */ stream->oa_buffer.last_ctx_id = INVALID_CTX_ID; spin_unlock_irqrestore(&stream->oa_buffer.ptr_lock, flags); /* * NB: although the OA buffer will initially be allocated * zeroed via shmfs (and so this memset is redundant when * first allocating), we may re-init the OA buffer, either * when re-enabling a stream or in error/reset paths. * * The reason we clear the buffer for each re-init is for the * sanity check in gen8_append_oa_reports() that looks at the * reason field to make sure it's non-zero which relies on * the assumption that new reports are being written to zeroed * memory... */ memset(stream->oa_buffer.vaddr, 0, stream->oa_buffer.vma->size); } static int alloc_oa_buffer(struct i915_perf_stream *stream) { struct drm_i915_private *i915 = stream->perf->i915; struct drm_i915_gem_object *bo; struct i915_vma *vma; int ret; if (drm_WARN_ON(&i915->drm, stream->oa_buffer.vma)) return -ENODEV; BUILD_BUG_ON_NOT_POWER_OF_2(OA_BUFFER_SIZE); BUILD_BUG_ON(OA_BUFFER_SIZE < SZ_128K || OA_BUFFER_SIZE > SZ_16M); bo = i915_gem_object_create_shmem(stream->perf->i915, OA_BUFFER_SIZE); if (IS_ERR(bo)) { drm_err(&i915->drm, "Failed to allocate OA buffer\n"); return PTR_ERR(bo); } i915_gem_object_set_cache_coherency(bo, I915_CACHE_LLC); /* PreHSW required 512K alignment, HSW requires 16M */ vma = i915_gem_object_ggtt_pin(bo, NULL, 0, SZ_16M, 0); if (IS_ERR(vma)) { ret = PTR_ERR(vma); goto err_unref; } stream->oa_buffer.vma = vma; stream->oa_buffer.vaddr = i915_gem_object_pin_map_unlocked(bo, I915_MAP_WB); if (IS_ERR(stream->oa_buffer.vaddr)) { ret = PTR_ERR(stream->oa_buffer.vaddr); goto err_unpin; } return 0; err_unpin: __i915_vma_unpin(vma); err_unref: i915_gem_object_put(bo); stream->oa_buffer.vaddr = NULL; stream->oa_buffer.vma = NULL; return ret; } static u32 *save_restore_register(struct i915_perf_stream *stream, u32 *cs, bool save, i915_reg_t reg, u32 offset, u32 dword_count) { u32 cmd; u32 d; cmd = save ? MI_STORE_REGISTER_MEM : MI_LOAD_REGISTER_MEM; cmd |= MI_SRM_LRM_GLOBAL_GTT; if (GRAPHICS_VER(stream->perf->i915) >= 8) cmd++; for (d = 0; d < dword_count; d++) { *cs++ = cmd; *cs++ = i915_mmio_reg_offset(reg) + 4 * d; *cs++ = intel_gt_scratch_offset(stream->engine->gt, offset) + 4 * d; *cs++ = 0; } return cs; } static int alloc_noa_wait(struct i915_perf_stream *stream) { struct drm_i915_private *i915 = stream->perf->i915; struct drm_i915_gem_object *bo; struct i915_vma *vma; const u64 delay_ticks = 0xffffffffffffffff - intel_gt_ns_to_clock_interval(to_gt(stream->perf->i915), atomic64_read(&stream->perf->noa_programming_delay)); const u32 base = stream->engine->mmio_base; #define CS_GPR(x) GEN8_RING_CS_GPR(base, x) u32 *batch, *ts0, *cs, *jump; struct i915_gem_ww_ctx ww; int ret, i; enum { START_TS, NOW_TS, DELTA_TS, JUMP_PREDICATE, DELTA_TARGET, N_CS_GPR }; bo = i915_gem_object_create_internal(i915, 4096); if (IS_ERR(bo)) { drm_err(&i915->drm, "Failed to allocate NOA wait batchbuffer\n"); return PTR_ERR(bo); } i915_gem_ww_ctx_init(&ww, true); retry: ret = i915_gem_object_lock(bo, &ww); if (ret) goto out_ww; /* * We pin in GGTT because we jump into this buffer now because * multiple OA config BOs will have a jump to this address and it * needs to be fixed during the lifetime of the i915/perf stream. */ vma = i915_gem_object_ggtt_pin_ww(bo, &ww, NULL, 0, 0, PIN_HIGH); if (IS_ERR(vma)) { ret = PTR_ERR(vma); goto out_ww; } batch = cs = i915_gem_object_pin_map(bo, I915_MAP_WB); if (IS_ERR(batch)) { ret = PTR_ERR(batch); goto err_unpin; } /* Save registers. */ for (i = 0; i < N_CS_GPR; i++) cs = save_restore_register( stream, cs, true /* save */, CS_GPR(i), INTEL_GT_SCRATCH_FIELD_PERF_CS_GPR + 8 * i, 2); cs = save_restore_register( stream, cs, true /* save */, MI_PREDICATE_RESULT_1(RENDER_RING_BASE), INTEL_GT_SCRATCH_FIELD_PERF_PREDICATE_RESULT_1, 1); /* First timestamp snapshot location. */ ts0 = cs; /* * Initial snapshot of the timestamp register to implement the wait. * We work with 32b values, so clear out the top 32b bits of the * register because the ALU works 64bits. */ *cs++ = MI_LOAD_REGISTER_IMM(1); *cs++ = i915_mmio_reg_offset(CS_GPR(START_TS)) + 4; *cs++ = 0; *cs++ = MI_LOAD_REGISTER_REG | (3 - 2); *cs++ = i915_mmio_reg_offset(RING_TIMESTAMP(base)); *cs++ = i915_mmio_reg_offset(CS_GPR(START_TS)); /* * This is the location we're going to jump back into until the * required amount of time has passed. */ jump = cs; /* * Take another snapshot of the timestamp register. Take care to clear * up the top 32bits of CS_GPR(1) as we're using it for other * operations below. */ *cs++ = MI_LOAD_REGISTER_IMM(1); *cs++ = i915_mmio_reg_offset(CS_GPR(NOW_TS)) + 4; *cs++ = 0; *cs++ = MI_LOAD_REGISTER_REG | (3 - 2); *cs++ = i915_mmio_reg_offset(RING_TIMESTAMP(base)); *cs++ = i915_mmio_reg_offset(CS_GPR(NOW_TS)); /* * Do a diff between the 2 timestamps and store the result back into * CS_GPR(1). */ *cs++ = MI_MATH(5); *cs++ = MI_MATH_LOAD(MI_MATH_REG_SRCA, MI_MATH_REG(NOW_TS)); *cs++ = MI_MATH_LOAD(MI_MATH_REG_SRCB, MI_MATH_REG(START_TS)); *cs++ = MI_MATH_SUB; *cs++ = MI_MATH_STORE(MI_MATH_REG(DELTA_TS), MI_MATH_REG_ACCU); *cs++ = MI_MATH_STORE(MI_MATH_REG(JUMP_PREDICATE), MI_MATH_REG_CF); /* * Transfer the carry flag (set to 1 if ts1 < ts0, meaning the * timestamp have rolled over the 32bits) into the predicate register * to be used for the predicated jump. */ *cs++ = MI_LOAD_REGISTER_REG | (3 - 2); *cs++ = i915_mmio_reg_offset(CS_GPR(JUMP_PREDICATE)); *cs++ = i915_mmio_reg_offset(MI_PREDICATE_RESULT_1(RENDER_RING_BASE)); /* Restart from the beginning if we had timestamps roll over. */ *cs++ = (GRAPHICS_VER(i915) < 8 ? MI_BATCH_BUFFER_START : MI_BATCH_BUFFER_START_GEN8) | MI_BATCH_PREDICATE; *cs++ = i915_ggtt_offset(vma) + (ts0 - batch) * 4; *cs++ = 0; /* * Now add the diff between to previous timestamps and add it to : * (((1 * << 64) - 1) - delay_ns) * * When the Carry Flag contains 1 this means the elapsed time is * longer than the expected delay, and we can exit the wait loop. */ *cs++ = MI_LOAD_REGISTER_IMM(2); *cs++ = i915_mmio_reg_offset(CS_GPR(DELTA_TARGET)); *cs++ = lower_32_bits(delay_ticks); *cs++ = i915_mmio_reg_offset(CS_GPR(DELTA_TARGET)) + 4; *cs++ = upper_32_bits(delay_ticks); *cs++ = MI_MATH(4); *cs++ = MI_MATH_LOAD(MI_MATH_REG_SRCA, MI_MATH_REG(DELTA_TS)); *cs++ = MI_MATH_LOAD(MI_MATH_REG_SRCB, MI_MATH_REG(DELTA_TARGET)); *cs++ = MI_MATH_ADD; *cs++ = MI_MATH_STOREINV(MI_MATH_REG(JUMP_PREDICATE), MI_MATH_REG_CF); *cs++ = MI_ARB_CHECK; /* * Transfer the result into the predicate register to be used for the * predicated jump. */ *cs++ = MI_LOAD_REGISTER_REG | (3 - 2); *cs++ = i915_mmio_reg_offset(CS_GPR(JUMP_PREDICATE)); *cs++ = i915_mmio_reg_offset(MI_PREDICATE_RESULT_1(RENDER_RING_BASE)); /* Predicate the jump. */ *cs++ = (GRAPHICS_VER(i915) < 8 ? MI_BATCH_BUFFER_START : MI_BATCH_BUFFER_START_GEN8) | MI_BATCH_PREDICATE; *cs++ = i915_ggtt_offset(vma) + (jump - batch) * 4; *cs++ = 0; /* Restore registers. */ for (i = 0; i < N_CS_GPR; i++) cs = save_restore_register( stream, cs, false /* restore */, CS_GPR(i), INTEL_GT_SCRATCH_FIELD_PERF_CS_GPR + 8 * i, 2); cs = save_restore_register( stream, cs, false /* restore */, MI_PREDICATE_RESULT_1(RENDER_RING_BASE), INTEL_GT_SCRATCH_FIELD_PERF_PREDICATE_RESULT_1, 1); /* And return to the ring. */ *cs++ = MI_BATCH_BUFFER_END; GEM_BUG_ON(cs - batch > PAGE_SIZE / sizeof(*batch)); i915_gem_object_flush_map(bo); __i915_gem_object_release_map(bo); stream->noa_wait = vma; goto out_ww; err_unpin: i915_vma_unpin_and_release(&vma, 0); out_ww: if (ret == -EDEADLK) { ret = i915_gem_ww_ctx_backoff(&ww); if (!ret) goto retry; } i915_gem_ww_ctx_fini(&ww); if (ret) i915_gem_object_put(bo); return ret; } static u32 *write_cs_mi_lri(u32 *cs, const struct i915_oa_reg *reg_data, u32 n_regs) { u32 i; for (i = 0; i < n_regs; i++) { if ((i % MI_LOAD_REGISTER_IMM_MAX_REGS) == 0) { u32 n_lri = min_t(u32, n_regs - i, MI_LOAD_REGISTER_IMM_MAX_REGS); *cs++ = MI_LOAD_REGISTER_IMM(n_lri); } *cs++ = i915_mmio_reg_offset(reg_data[i].addr); *cs++ = reg_data[i].value; } return cs; } static int num_lri_dwords(int num_regs) { int count = 0; if (num_regs > 0) { count += DIV_ROUND_UP(num_regs, MI_LOAD_REGISTER_IMM_MAX_REGS); count += num_regs * 2; } return count; } static struct i915_oa_config_bo * alloc_oa_config_buffer(struct i915_perf_stream *stream, struct i915_oa_config *oa_config) { struct drm_i915_gem_object *obj; struct i915_oa_config_bo *oa_bo; struct i915_gem_ww_ctx ww; size_t config_length = 0; u32 *cs; int err; oa_bo = kzalloc(sizeof(*oa_bo), GFP_KERNEL); if (!oa_bo) return ERR_PTR(-ENOMEM); config_length += num_lri_dwords(oa_config->mux_regs_len); config_length += num_lri_dwords(oa_config->b_counter_regs_len); config_length += num_lri_dwords(oa_config->flex_regs_len); config_length += 3; /* MI_BATCH_BUFFER_START */ config_length = ALIGN(sizeof(u32) * config_length, I915_GTT_PAGE_SIZE); obj = i915_gem_object_create_shmem(stream->perf->i915, config_length); if (IS_ERR(obj)) { err = PTR_ERR(obj); goto err_free; } i915_gem_ww_ctx_init(&ww, true); retry: err = i915_gem_object_lock(obj, &ww); if (err) goto out_ww; cs = i915_gem_object_pin_map(obj, I915_MAP_WB); if (IS_ERR(cs)) { err = PTR_ERR(cs); goto out_ww; } cs = write_cs_mi_lri(cs, oa_config->mux_regs, oa_config->mux_regs_len); cs = write_cs_mi_lri(cs, oa_config->b_counter_regs, oa_config->b_counter_regs_len); cs = write_cs_mi_lri(cs, oa_config->flex_regs, oa_config->flex_regs_len); /* Jump into the active wait. */ *cs++ = (GRAPHICS_VER(stream->perf->i915) < 8 ? MI_BATCH_BUFFER_START : MI_BATCH_BUFFER_START_GEN8); *cs++ = i915_ggtt_offset(stream->noa_wait); *cs++ = 0; i915_gem_object_flush_map(obj); __i915_gem_object_release_map(obj); oa_bo->vma = i915_vma_instance(obj, &stream->engine->gt->ggtt->vm, NULL); if (IS_ERR(oa_bo->vma)) { err = PTR_ERR(oa_bo->vma); goto out_ww; } oa_bo->oa_config = i915_oa_config_get(oa_config); llist_add(&oa_bo->node, &stream->oa_config_bos); out_ww: if (err == -EDEADLK) { err = i915_gem_ww_ctx_backoff(&ww); if (!err) goto retry; } i915_gem_ww_ctx_fini(&ww); if (err) i915_gem_object_put(obj); err_free: if (err) { kfree(oa_bo); return ERR_PTR(err); } return oa_bo; } static struct i915_vma * get_oa_vma(struct i915_perf_stream *stream, struct i915_oa_config *oa_config) { struct i915_oa_config_bo *oa_bo; /* * Look for the buffer in the already allocated BOs attached * to the stream. */ llist_for_each_entry(oa_bo, stream->oa_config_bos.first, node) { if (oa_bo->oa_config == oa_config && memcmp(oa_bo->oa_config->uuid, oa_config->uuid, sizeof(oa_config->uuid)) == 0) goto out; } oa_bo = alloc_oa_config_buffer(stream, oa_config); if (IS_ERR(oa_bo)) return ERR_CAST(oa_bo); out: return i915_vma_get(oa_bo->vma); } static int emit_oa_config(struct i915_perf_stream *stream, struct i915_oa_config *oa_config, struct intel_context *ce, struct i915_active *active) { struct i915_request *rq; struct i915_vma *vma; struct i915_gem_ww_ctx ww; int err; vma = get_oa_vma(stream, oa_config); if (IS_ERR(vma)) return PTR_ERR(vma); i915_gem_ww_ctx_init(&ww, true); retry: err = i915_gem_object_lock(vma->obj, &ww); if (err) goto err; err = i915_vma_pin_ww(vma, &ww, 0, 0, PIN_GLOBAL | PIN_HIGH); if (err) goto err; intel_engine_pm_get(ce->engine); rq = i915_request_create(ce); intel_engine_pm_put(ce->engine); if (IS_ERR(rq)) { err = PTR_ERR(rq); goto err_vma_unpin; } if (!IS_ERR_OR_NULL(active)) { /* After all individual context modifications */ err = i915_request_await_active(rq, active, I915_ACTIVE_AWAIT_ACTIVE); if (err) goto err_add_request; err = i915_active_add_request(active, rq); if (err) goto err_add_request; } err = i915_request_await_object(rq, vma->obj, 0); if (!err) err = i915_vma_move_to_active(vma, rq, 0); if (err) goto err_add_request; err = rq->engine->emit_bb_start(rq, vma->node.start, 0, I915_DISPATCH_SECURE); if (err) goto err_add_request; err_add_request: i915_request_add(rq); err_vma_unpin: i915_vma_unpin(vma); err: if (err == -EDEADLK) { err = i915_gem_ww_ctx_backoff(&ww); if (!err) goto retry; } i915_gem_ww_ctx_fini(&ww); i915_vma_put(vma); return err; } static struct intel_context *oa_context(struct i915_perf_stream *stream) { return stream->pinned_ctx ?: stream->engine->kernel_context; } static int hsw_enable_metric_set(struct i915_perf_stream *stream, struct i915_active *active) { struct intel_uncore *uncore = stream->uncore; /* * PRM: * * OA unit is using “crclk” for its functionality. When trunk * level clock gating takes place, OA clock would be gated, * unable to count the events from non-render clock domain. * Render clock gating must be disabled when OA is enabled to * count the events from non-render domain. Unit level clock * gating for RCS should also be disabled. */ intel_uncore_rmw(uncore, GEN7_MISCCPCTL, GEN7_DOP_CLOCK_GATE_ENABLE, 0); intel_uncore_rmw(uncore, GEN6_UCGCTL1, 0, GEN6_CSUNIT_CLOCK_GATE_DISABLE); return emit_oa_config(stream, stream->oa_config, oa_context(stream), active); } static void hsw_disable_metric_set(struct i915_perf_stream *stream) { struct intel_uncore *uncore = stream->uncore; intel_uncore_rmw(uncore, GEN6_UCGCTL1, GEN6_CSUNIT_CLOCK_GATE_DISABLE, 0); intel_uncore_rmw(uncore, GEN7_MISCCPCTL, 0, GEN7_DOP_CLOCK_GATE_ENABLE); intel_uncore_rmw(uncore, GDT_CHICKEN_BITS, GT_NOA_ENABLE, 0); } static u32 oa_config_flex_reg(const struct i915_oa_config *oa_config, i915_reg_t reg) { u32 mmio = i915_mmio_reg_offset(reg); int i; /* * This arbitrary default will select the 'EU FPU0 Pipeline * Active' event. In the future it's anticipated that there * will be an explicit 'No Event' we can select, but not yet... */ if (!oa_config) return 0; for (i = 0; i < oa_config->flex_regs_len; i++) { if (i915_mmio_reg_offset(oa_config->flex_regs[i].addr) == mmio) return oa_config->flex_regs[i].value; } return 0; } /* * NB: It must always remain pointer safe to run this even if the OA unit * has been disabled. * * It's fine to put out-of-date values into these per-context registers * in the case that the OA unit has been disabled. */ static void gen8_update_reg_state_unlocked(const struct intel_context *ce, const struct i915_perf_stream *stream) { u32 ctx_oactxctrl = stream->perf->ctx_oactxctrl_offset; u32 ctx_flexeu0 = stream->perf->ctx_flexeu0_offset; /* The MMIO offsets for Flex EU registers aren't contiguous */ static const i915_reg_t flex_regs[] = { EU_PERF_CNTL0, EU_PERF_CNTL1, EU_PERF_CNTL2, EU_PERF_CNTL3, EU_PERF_CNTL4, EU_PERF_CNTL5, EU_PERF_CNTL6, }; u32 *reg_state = ce->lrc_reg_state; int i; reg_state[ctx_oactxctrl + 1] = (stream->period_exponent << GEN8_OA_TIMER_PERIOD_SHIFT) | (stream->periodic ? GEN8_OA_TIMER_ENABLE : 0) | GEN8_OA_COUNTER_RESUME; for (i = 0; i < ARRAY_SIZE(flex_regs); i++) reg_state[ctx_flexeu0 + i * 2 + 1] = oa_config_flex_reg(stream->oa_config, flex_regs[i]); } struct flex { i915_reg_t reg; u32 offset; u32 value; }; static int gen8_store_flex(struct i915_request *rq, struct intel_context *ce, const struct flex *flex, unsigned int count) { u32 offset; u32 *cs; cs = intel_ring_begin(rq, 4 * count); if (IS_ERR(cs)) return PTR_ERR(cs); offset = i915_ggtt_offset(ce->state) + LRC_STATE_OFFSET; do { *cs++ = MI_STORE_DWORD_IMM_GEN4 | MI_USE_GGTT; *cs++ = offset + flex->offset * sizeof(u32); *cs++ = 0; *cs++ = flex->value; } while (flex++, --count); intel_ring_advance(rq, cs); return 0; } static int gen8_load_flex(struct i915_request *rq, struct intel_context *ce, const struct flex *flex, unsigned int count) { u32 *cs; GEM_BUG_ON(!count || count > 63); cs = intel_ring_begin(rq, 2 * count + 2); if (IS_ERR(cs)) return PTR_ERR(cs); *cs++ = MI_LOAD_REGISTER_IMM(count); do { *cs++ = i915_mmio_reg_offset(flex->reg); *cs++ = flex->value; } while (flex++, --count); *cs++ = MI_NOOP; intel_ring_advance(rq, cs); return 0; } static int gen8_modify_context(struct intel_context *ce, const struct flex *flex, unsigned int count) { struct i915_request *rq; int err; rq = intel_engine_create_kernel_request(ce->engine); if (IS_ERR(rq)) return PTR_ERR(rq); /* Serialise with the remote context */ err = intel_context_prepare_remote_request(ce, rq); if (err == 0) err = gen8_store_flex(rq, ce, flex, count); i915_request_add(rq); return err; } static int gen8_modify_self(struct intel_context *ce, const struct flex *flex, unsigned int count, struct i915_active *active) { struct i915_request *rq; int err; intel_engine_pm_get(ce->engine); rq = i915_request_create(ce); intel_engine_pm_put(ce->engine); if (IS_ERR(rq)) return PTR_ERR(rq); if (!IS_ERR_OR_NULL(active)) { err = i915_active_add_request(active, rq); if (err) goto err_add_request; } err = gen8_load_flex(rq, ce, flex, count); if (err) goto err_add_request; err_add_request: i915_request_add(rq); return err; } static int gen8_configure_context(struct i915_gem_context *ctx, struct flex *flex, unsigned int count) { struct i915_gem_engines_iter it; struct intel_context *ce; int err = 0; for_each_gem_engine(ce, i915_gem_context_lock_engines(ctx), it) { GEM_BUG_ON(ce == ce->engine->kernel_context); if (ce->engine->class != RENDER_CLASS) continue; /* Otherwise OA settings will be set upon first use */ if (!intel_context_pin_if_active(ce)) continue; flex->value = intel_sseu_make_rpcs(ce->engine->gt, &ce->sseu); err = gen8_modify_context(ce, flex, count); intel_context_unpin(ce); if (err) break; } i915_gem_context_unlock_engines(ctx); return err; } static int gen12_configure_oar_context(struct i915_perf_stream *stream, struct i915_active *active) { int err; struct intel_context *ce = stream->pinned_ctx; u32 format = stream->oa_buffer.format; struct flex regs_context[] = { { GEN8_OACTXCONTROL, stream->perf->ctx_oactxctrl_offset + 1, active ? GEN8_OA_COUNTER_RESUME : 0, }, }; /* Offsets in regs_lri are not used since this configuration is only * applied using LRI. Initialize the correct offsets for posterity. */ #define GEN12_OAR_OACONTROL_OFFSET 0x5B0 struct flex regs_lri[] = { { GEN12_OAR_OACONTROL, GEN12_OAR_OACONTROL_OFFSET + 1, (format << GEN12_OAR_OACONTROL_COUNTER_FORMAT_SHIFT) | (active ? GEN12_OAR_OACONTROL_COUNTER_ENABLE : 0) }, { RING_CONTEXT_CONTROL(ce->engine->mmio_base), CTX_CONTEXT_CONTROL, _MASKED_FIELD(GEN12_CTX_CTRL_OAR_CONTEXT_ENABLE, active ? GEN12_CTX_CTRL_OAR_CONTEXT_ENABLE : 0) }, }; /* Modify the context image of pinned context with regs_context*/ err = intel_context_lock_pinned(ce); if (err) return err; err = gen8_modify_context(ce, regs_context, ARRAY_SIZE(regs_context)); intel_context_unlock_pinned(ce); if (err) return err; /* Apply regs_lri using LRI with pinned context */ return gen8_modify_self(ce, regs_lri, ARRAY_SIZE(regs_lri), active); } /* * Manages updating the per-context aspects of the OA stream * configuration across all contexts. * * The awkward consideration here is that OACTXCONTROL controls the * exponent for periodic sampling which is primarily used for system * wide profiling where we'd like a consistent sampling period even in * the face of context switches. * * Our approach of updating the register state context (as opposed to * say using a workaround batch buffer) ensures that the hardware * won't automatically reload an out-of-date timer exponent even * transiently before a WA BB could be parsed. * * This function needs to: * - Ensure the currently running context's per-context OA state is * updated * - Ensure that all existing contexts will have the correct per-context * OA state if they are scheduled for use. * - Ensure any new contexts will be initialized with the correct * per-context OA state. * * Note: it's only the RCS/Render context that has any OA state. * Note: the first flex register passed must always be R_PWR_CLK_STATE */ static int oa_configure_all_contexts(struct i915_perf_stream *stream, struct flex *regs, size_t num_regs, struct i915_active *active) { struct drm_i915_private *i915 = stream->perf->i915; struct intel_engine_cs *engine; struct i915_gem_context *ctx, *cn; int err; lockdep_assert_held(&stream->perf->lock); /* * The OA register config is setup through the context image. This image * might be written to by the GPU on context switch (in particular on * lite-restore). This means we can't safely update a context's image, * if this context is scheduled/submitted to run on the GPU. * * We could emit the OA register config through the batch buffer but * this might leave small interval of time where the OA unit is * configured at an invalid sampling period. * * Note that since we emit all requests from a single ring, there * is still an implicit global barrier here that may cause a high * priority context to wait for an otherwise independent low priority * context. Contexts idle at the time of reconfiguration are not * trapped behind the barrier. */ spin_lock(&i915->gem.contexts.lock); list_for_each_entry_safe(ctx, cn, &i915->gem.contexts.list, link) { if (!kref_get_unless_zero(&ctx->ref)) continue; spin_unlock(&i915->gem.contexts.lock); err = gen8_configure_context(ctx, regs, num_regs); if (err) { i915_gem_context_put(ctx); return err; } spin_lock(&i915->gem.contexts.lock); list_safe_reset_next(ctx, cn, link); i915_gem_context_put(ctx); } spin_unlock(&i915->gem.contexts.lock); /* * After updating all other contexts, we need to modify ourselves. * If we don't modify the kernel_context, we do not get events while * idle. */ for_each_uabi_engine(engine, i915) { struct intel_context *ce = engine->kernel_context; if (engine->class != RENDER_CLASS) continue; regs[0].value = intel_sseu_make_rpcs(engine->gt, &ce->sseu); err = gen8_modify_self(ce, regs, num_regs, active); if (err) return err; } return 0; } static int gen12_configure_all_contexts(struct i915_perf_stream *stream, const struct i915_oa_config *oa_config, struct i915_active *active) { struct flex regs[] = { { GEN8_R_PWR_CLK_STATE(RENDER_RING_BASE), CTX_R_PWR_CLK_STATE, }, }; return oa_configure_all_contexts(stream, regs, ARRAY_SIZE(regs), active); } static int lrc_configure_all_contexts(struct i915_perf_stream *stream, const struct i915_oa_config *oa_config, struct i915_active *active) { /* The MMIO offsets for Flex EU registers aren't contiguous */ const u32 ctx_flexeu0 = stream->perf->ctx_flexeu0_offset; #define ctx_flexeuN(N) (ctx_flexeu0 + 2 * (N) + 1) struct flex regs[] = { { GEN8_R_PWR_CLK_STATE(RENDER_RING_BASE), CTX_R_PWR_CLK_STATE, }, { GEN8_OACTXCONTROL, stream->perf->ctx_oactxctrl_offset + 1, }, { EU_PERF_CNTL0, ctx_flexeuN(0) }, { EU_PERF_CNTL1, ctx_flexeuN(1) }, { EU_PERF_CNTL2, ctx_flexeuN(2) }, { EU_PERF_CNTL3, ctx_flexeuN(3) }, { EU_PERF_CNTL4, ctx_flexeuN(4) }, { EU_PERF_CNTL5, ctx_flexeuN(5) }, { EU_PERF_CNTL6, ctx_flexeuN(6) }, }; #undef ctx_flexeuN int i; regs[1].value = (stream->period_exponent << GEN8_OA_TIMER_PERIOD_SHIFT) | (stream->periodic ? GEN8_OA_TIMER_ENABLE : 0) | GEN8_OA_COUNTER_RESUME; for (i = 2; i < ARRAY_SIZE(regs); i++) regs[i].value = oa_config_flex_reg(oa_config, regs[i].reg); return oa_configure_all_contexts(stream, regs, ARRAY_SIZE(regs), active); } static int gen8_enable_metric_set(struct i915_perf_stream *stream, struct i915_active *active) { struct intel_uncore *uncore = stream->uncore; struct i915_oa_config *oa_config = stream->oa_config; int ret; /* * We disable slice/unslice clock ratio change reports on SKL since * they are too noisy. The HW generates a lot of redundant reports * where the ratio hasn't really changed causing a lot of redundant * work to processes and increasing the chances we'll hit buffer * overruns. * * Although we don't currently use the 'disable overrun' OABUFFER * feature it's worth noting that clock ratio reports have to be * disabled before considering to use that feature since the HW doesn't * correctly block these reports. * * Currently none of the high-level metrics we have depend on knowing * this ratio to normalize. * * Note: This register is not power context saved and restored, but * that's OK considering that we disable RC6 while the OA unit is * enabled. * * The _INCLUDE_CLK_RATIO bit allows the slice/unslice frequency to * be read back from automatically triggered reports, as part of the * RPT_ID field. */ if (IS_GRAPHICS_VER(stream->perf->i915, 9, 11)) { intel_uncore_write(uncore, GEN8_OA_DEBUG, _MASKED_BIT_ENABLE(GEN9_OA_DEBUG_DISABLE_CLK_RATIO_REPORTS | GEN9_OA_DEBUG_INCLUDE_CLK_RATIO)); } /* * Update all contexts prior writing the mux configurations as we need * to make sure all slices/subslices are ON before writing to NOA * registers. */ ret = lrc_configure_all_contexts(stream, oa_config, active); if (ret) return ret; return emit_oa_config(stream, stream->oa_config, oa_context(stream), active); } static u32 oag_report_ctx_switches(const struct i915_perf_stream *stream) { return _MASKED_FIELD(GEN12_OAG_OA_DEBUG_DISABLE_CTX_SWITCH_REPORTS, (stream->sample_flags & SAMPLE_OA_REPORT) ? 0 : GEN12_OAG_OA_DEBUG_DISABLE_CTX_SWITCH_REPORTS); } static int gen12_enable_metric_set(struct i915_perf_stream *stream, struct i915_active *active) { struct intel_uncore *uncore = stream->uncore; struct i915_oa_config *oa_config = stream->oa_config; bool periodic = stream->periodic; u32 period_exponent = stream->period_exponent; int ret; intel_uncore_write(uncore, GEN12_OAG_OA_DEBUG, /* Disable clk ratio reports, like previous Gens. */ _MASKED_BIT_ENABLE(GEN12_OAG_OA_DEBUG_DISABLE_CLK_RATIO_REPORTS | GEN12_OAG_OA_DEBUG_INCLUDE_CLK_RATIO) | /* * If the user didn't require OA reports, instruct * the hardware not to emit ctx switch reports. */ oag_report_ctx_switches(stream)); intel_uncore_write(uncore, GEN12_OAG_OAGLBCTXCTRL, periodic ? (GEN12_OAG_OAGLBCTXCTRL_COUNTER_RESUME | GEN12_OAG_OAGLBCTXCTRL_TIMER_ENABLE | (period_exponent << GEN12_OAG_OAGLBCTXCTRL_TIMER_PERIOD_SHIFT)) : 0); /* * Update all contexts prior writing the mux configurations as we need * to make sure all slices/subslices are ON before writing to NOA * registers. */ ret = gen12_configure_all_contexts(stream, oa_config, active); if (ret) return ret; /* * For Gen12, performance counters are context * saved/restored. Only enable it for the context that * requested this. */ if (stream->ctx) { ret = gen12_configure_oar_context(stream, active); if (ret) return ret; } return emit_oa_config(stream, stream->oa_config, oa_context(stream), active); } static void gen8_disable_metric_set(struct i915_perf_stream *stream) { struct intel_uncore *uncore = stream->uncore; /* Reset all contexts' slices/subslices configurations. */ lrc_configure_all_contexts(stream, NULL, NULL); intel_uncore_rmw(uncore, GDT_CHICKEN_BITS, GT_NOA_ENABLE, 0); } static void gen11_disable_metric_set(struct i915_perf_stream *stream) { struct intel_uncore *uncore = stream->uncore; /* Reset all contexts' slices/subslices configurations. */ lrc_configure_all_contexts(stream, NULL, NULL); /* Make sure we disable noa to save power. */ intel_uncore_rmw(uncore, RPM_CONFIG1, GEN10_GT_NOA_ENABLE, 0); } static void gen12_disable_metric_set(struct i915_perf_stream *stream) { struct intel_uncore *uncore = stream->uncore; /* Reset all contexts' slices/subslices configurations. */ gen12_configure_all_contexts(stream, NULL, NULL); /* disable the context save/restore or OAR counters */ if (stream->ctx) gen12_configure_oar_context(stream, NULL); /* Make sure we disable noa to save power. */ intel_uncore_rmw(uncore, RPM_CONFIG1, GEN10_GT_NOA_ENABLE, 0); } static void gen7_oa_enable(struct i915_perf_stream *stream) { struct intel_uncore *uncore = stream->uncore; struct i915_gem_context *ctx = stream->ctx; u32 ctx_id = stream->specific_ctx_id; bool periodic = stream->periodic; u32 period_exponent = stream->period_exponent; u32 report_format = stream->oa_buffer.format; /* * Reset buf pointers so we don't forward reports from before now. * * Think carefully if considering trying to avoid this, since it * also ensures status flags and the buffer itself are cleared * in error paths, and we have checks for invalid reports based * on the assumption that certain fields are written to zeroed * memory which this helps maintains. */ gen7_init_oa_buffer(stream); intel_uncore_write(uncore, GEN7_OACONTROL, (ctx_id & GEN7_OACONTROL_CTX_MASK) | (period_exponent << GEN7_OACONTROL_TIMER_PERIOD_SHIFT) | (periodic ? GEN7_OACONTROL_TIMER_ENABLE : 0) | (report_format << GEN7_OACONTROL_FORMAT_SHIFT) | (ctx ? GEN7_OACONTROL_PER_CTX_ENABLE : 0) | GEN7_OACONTROL_ENABLE); } static void gen8_oa_enable(struct i915_perf_stream *stream) { struct intel_uncore *uncore = stream->uncore; u32 report_format = stream->oa_buffer.format; /* * Reset buf pointers so we don't forward reports from before now. * * Think carefully if considering trying to avoid this, since it * also ensures status flags and the buffer itself are cleared * in error paths, and we have checks for invalid reports based * on the assumption that certain fields are written to zeroed * memory which this helps maintains. */ gen8_init_oa_buffer(stream); /* * Note: we don't rely on the hardware to perform single context * filtering and instead filter on the cpu based on the context-id * field of reports */ intel_uncore_write(uncore, GEN8_OACONTROL, (report_format << GEN8_OA_REPORT_FORMAT_SHIFT) | GEN8_OA_COUNTER_ENABLE); } static void gen12_oa_enable(struct i915_perf_stream *stream) { struct intel_uncore *uncore = stream->uncore; u32 report_format = stream->oa_buffer.format; /* * If we don't want OA reports from the OA buffer, then we don't even * need to program the OAG unit. */ if (!(stream->sample_flags & SAMPLE_OA_REPORT)) return; gen12_init_oa_buffer(stream); intel_uncore_write(uncore, GEN12_OAG_OACONTROL, (report_format << GEN12_OAG_OACONTROL_OA_COUNTER_FORMAT_SHIFT) | GEN12_OAG_OACONTROL_OA_COUNTER_ENABLE); } /** * i915_oa_stream_enable - handle `I915_PERF_IOCTL_ENABLE` for OA stream * @stream: An i915 perf stream opened for OA metrics * * [Re]enables hardware periodic sampling according to the period configured * when opening the stream. This also starts a hrtimer that will periodically * check for data in the circular OA buffer for notifying userspace (e.g. * during a read() or poll()). */ static void i915_oa_stream_enable(struct i915_perf_stream *stream) { stream->pollin = false; stream->perf->ops.oa_enable(stream); if (stream->sample_flags & SAMPLE_OA_REPORT) hrtimer_start(&stream->poll_check_timer, ns_to_ktime(stream->poll_oa_period), HRTIMER_MODE_REL_PINNED); } static void gen7_oa_disable(struct i915_perf_stream *stream) { struct intel_uncore *uncore = stream->uncore; intel_uncore_write(uncore, GEN7_OACONTROL, 0); if (intel_wait_for_register(uncore, GEN7_OACONTROL, GEN7_OACONTROL_ENABLE, 0, 50)) drm_err(&stream->perf->i915->drm, "wait for OA to be disabled timed out\n"); } static void gen8_oa_disable(struct i915_perf_stream *stream) { struct intel_uncore *uncore = stream->uncore; intel_uncore_write(uncore, GEN8_OACONTROL, 0); if (intel_wait_for_register(uncore, GEN8_OACONTROL, GEN8_OA_COUNTER_ENABLE, 0, 50)) drm_err(&stream->perf->i915->drm, "wait for OA to be disabled timed out\n"); } static void gen12_oa_disable(struct i915_perf_stream *stream) { struct intel_uncore *uncore = stream->uncore; intel_uncore_write(uncore, GEN12_OAG_OACONTROL, 0); if (intel_wait_for_register(uncore, GEN12_OAG_OACONTROL, GEN12_OAG_OACONTROL_OA_COUNTER_ENABLE, 0, 50)) drm_err(&stream->perf->i915->drm, "wait for OA to be disabled timed out\n"); intel_uncore_write(uncore, GEN12_OA_TLB_INV_CR, 1); if (intel_wait_for_register(uncore, GEN12_OA_TLB_INV_CR, 1, 0, 50)) drm_err(&stream->perf->i915->drm, "wait for OA tlb invalidate timed out\n"); } /** * i915_oa_stream_disable - handle `I915_PERF_IOCTL_DISABLE` for OA stream * @stream: An i915 perf stream opened for OA metrics * * Stops the OA unit from periodically writing counter reports into the * circular OA buffer. This also stops the hrtimer that periodically checks for * data in the circular OA buffer, for notifying userspace. */ static void i915_oa_stream_disable(struct i915_perf_stream *stream) { stream->perf->ops.oa_disable(stream); if (stream->sample_flags & SAMPLE_OA_REPORT) hrtimer_cancel(&stream->poll_check_timer); } static const struct i915_perf_stream_ops i915_oa_stream_ops = { .destroy = i915_oa_stream_destroy, .enable = i915_oa_stream_enable, .disable = i915_oa_stream_disable, .wait_unlocked = i915_oa_wait_unlocked, .poll_wait = i915_oa_poll_wait, .read = i915_oa_read, }; static int i915_perf_stream_enable_sync(struct i915_perf_stream *stream) { struct i915_active *active; int err; active = i915_active_create(); if (!active) return -ENOMEM; err = stream->perf->ops.enable_metric_set(stream, active); if (err == 0) __i915_active_wait(active, TASK_UNINTERRUPTIBLE); i915_active_put(active); return err; } static void get_default_sseu_config(struct intel_sseu *out_sseu, struct intel_engine_cs *engine) { const struct sseu_dev_info *devinfo_sseu = &engine->gt->info.sseu; *out_sseu = intel_sseu_from_device_info(devinfo_sseu); if (GRAPHICS_VER(engine->i915) == 11) { /* * We only need subslice count so it doesn't matter which ones * we select - just turn off low bits in the amount of half of * all available subslices per slice. */ out_sseu->subslice_mask = ~(~0 << (hweight8(out_sseu->subslice_mask) / 2)); out_sseu->slice_mask = 0x1; } } static int get_sseu_config(struct intel_sseu *out_sseu, struct intel_engine_cs *engine, const struct drm_i915_gem_context_param_sseu *drm_sseu) { if (drm_sseu->engine.engine_class != engine->uabi_class || drm_sseu->engine.engine_instance != engine->uabi_instance) return -EINVAL; return i915_gem_user_to_context_sseu(engine->gt, drm_sseu, out_sseu); } /** * i915_oa_stream_init - validate combined props for OA stream and init * @stream: An i915 perf stream * @param: The open parameters passed to `DRM_I915_PERF_OPEN` * @props: The property state that configures stream (individually validated) * * While read_properties_unlocked() validates properties in isolation it * doesn't ensure that the combination necessarily makes sense. * * At this point it has been determined that userspace wants a stream of * OA metrics, but still we need to further validate the combined * properties are OK. * * If the configuration makes sense then we can allocate memory for * a circular OA buffer and apply the requested metric set configuration. * * Returns: zero on success or a negative error code. */ static int i915_oa_stream_init(struct i915_perf_stream *stream, struct drm_i915_perf_open_param *param, struct perf_open_properties *props) { struct drm_i915_private *i915 = stream->perf->i915; struct i915_perf *perf = stream->perf; int format_size; int ret; if (!props->engine) { drm_dbg(&stream->perf->i915->drm, "OA engine not specified\n"); return -EINVAL; } /* * If the sysfs metrics/ directory wasn't registered for some * reason then don't let userspace try their luck with config * IDs */ if (!perf->metrics_kobj) { drm_dbg(&stream->perf->i915->drm, "OA metrics weren't advertised via sysfs\n"); return -EINVAL; } if (!(props->sample_flags & SAMPLE_OA_REPORT) && (GRAPHICS_VER(perf->i915) < 12 || !stream->ctx)) { drm_dbg(&stream->perf->i915->drm, "Only OA report sampling supported\n"); return -EINVAL; } if (!perf->ops.enable_metric_set) { drm_dbg(&stream->perf->i915->drm, "OA unit not supported\n"); return -ENODEV; } /* * To avoid the complexity of having to accurately filter * counter reports and marshal to the appropriate client * we currently only allow exclusive access */ if (perf->exclusive_stream) { drm_dbg(&stream->perf->i915->drm, "OA unit already in use\n"); return -EBUSY; } if (!props->oa_format) { drm_dbg(&stream->perf->i915->drm, "OA report format not specified\n"); return -EINVAL; } stream->engine = props->engine; stream->uncore = stream->engine->gt->uncore; stream->sample_size = sizeof(struct drm_i915_perf_record_header); format_size = perf->oa_formats[props->oa_format].size; stream->sample_flags = props->sample_flags; stream->sample_size += format_size; stream->oa_buffer.format_size = format_size; if (drm_WARN_ON(&i915->drm, stream->oa_buffer.format_size == 0)) return -EINVAL; stream->hold_preemption = props->hold_preemption; stream->oa_buffer.format = perf->oa_formats[props->oa_format].format; stream->periodic = props->oa_periodic; if (stream->periodic) stream->period_exponent = props->oa_period_exponent; if (stream->ctx) { ret = oa_get_render_ctx_id(stream); if (ret) { drm_dbg(&stream->perf->i915->drm, "Invalid context id to filter with\n"); return ret; } } ret = alloc_noa_wait(stream); if (ret) { drm_dbg(&stream->perf->i915->drm, "Unable to allocate NOA wait batch buffer\n"); goto err_noa_wait_alloc; } stream->oa_config = i915_perf_get_oa_config(perf, props->metrics_set); if (!stream->oa_config) { drm_dbg(&stream->perf->i915->drm, "Invalid OA config id=%i\n", props->metrics_set); ret = -EINVAL; goto err_config; } /* PRM - observability performance counters: * * OACONTROL, performance counter enable, note: * * "When this bit is set, in order to have coherent counts, * RC6 power state and trunk clock gating must be disabled. * This can be achieved by programming MMIO registers as * 0xA094=0 and 0xA090[31]=1" * * In our case we are expecting that taking pm + FORCEWAKE * references will effectively disable RC6. */ intel_engine_pm_get(stream->engine); intel_uncore_forcewake_get(stream->uncore, FORCEWAKE_ALL); ret = alloc_oa_buffer(stream); if (ret) goto err_oa_buf_alloc; stream->ops = &i915_oa_stream_ops; perf->sseu = props->sseu; WRITE_ONCE(perf->exclusive_stream, stream); ret = i915_perf_stream_enable_sync(stream); if (ret) { drm_dbg(&stream->perf->i915->drm, "Unable to enable metric set\n"); goto err_enable; } drm_dbg(&stream->perf->i915->drm, "opening stream oa config uuid=%s\n", stream->oa_config->uuid); hrtimer_init(&stream->poll_check_timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL); stream->poll_check_timer.function = oa_poll_check_timer_cb; init_waitqueue_head(&stream->poll_wq); spin_lock_init(&stream->oa_buffer.ptr_lock); return 0; err_enable: WRITE_ONCE(perf->exclusive_stream, NULL); perf->ops.disable_metric_set(stream); free_oa_buffer(stream); err_oa_buf_alloc: free_oa_configs(stream); intel_uncore_forcewake_put(stream->uncore, FORCEWAKE_ALL); intel_engine_pm_put(stream->engine); err_config: free_noa_wait(stream); err_noa_wait_alloc: if (stream->ctx) oa_put_render_ctx_id(stream); return ret; } void i915_oa_init_reg_state(const struct intel_context *ce, const struct intel_engine_cs *engine) { struct i915_perf_stream *stream; if (engine->class != RENDER_CLASS) return; /* perf.exclusive_stream serialised by lrc_configure_all_contexts() */ stream = READ_ONCE(engine->i915->perf.exclusive_stream); if (stream && GRAPHICS_VER(stream->perf->i915) < 12) gen8_update_reg_state_unlocked(ce, stream); } /** * i915_perf_read - handles read() FOP for i915 perf stream FDs * @file: An i915 perf stream file * @buf: destination buffer given by userspace * @count: the number of bytes userspace wants to read * @ppos: (inout) file seek position (unused) * * The entry point for handling a read() on a stream file descriptor from * userspace. Most of the work is left to the i915_perf_read_locked() and * &i915_perf_stream_ops->read but to save having stream implementations (of * which we might have multiple later) we handle blocking read here. * * We can also consistently treat trying to read from a disabled stream * as an IO error so implementations can assume the stream is enabled * while reading. * * Returns: The number of bytes copied or a negative error code on failure. */ static ssize_t i915_perf_read(struct file *file, char __user *buf, size_t count, loff_t *ppos) { struct i915_perf_stream *stream = file->private_data; struct i915_perf *perf = stream->perf; size_t offset = 0; int ret; /* To ensure it's handled consistently we simply treat all reads of a * disabled stream as an error. In particular it might otherwise lead * to a deadlock for blocking file descriptors... */ if (!stream->enabled || !(stream->sample_flags & SAMPLE_OA_REPORT)) return -EIO; if (!(file->f_flags & O_NONBLOCK)) { /* There's the small chance of false positives from * stream->ops->wait_unlocked. * * E.g. with single context filtering since we only wait until * oabuffer has >= 1 report we don't immediately know whether * any reports really belong to the current context */ do { ret = stream->ops->wait_unlocked(stream); if (ret) return ret; mutex_lock(&perf->lock); ret = stream->ops->read(stream, buf, count, &offset); mutex_unlock(&perf->lock); } while (!offset && !ret); } else { mutex_lock(&perf->lock); ret = stream->ops->read(stream, buf, count, &offset); mutex_unlock(&perf->lock); } /* We allow the poll checking to sometimes report false positive EPOLLIN * events where we might actually report EAGAIN on read() if there's * not really any data available. In this situation though we don't * want to enter a busy loop between poll() reporting a EPOLLIN event * and read() returning -EAGAIN. Clearing the oa.pollin state here * effectively ensures we back off until the next hrtimer callback * before reporting another EPOLLIN event. * The exception to this is if ops->read() returned -ENOSPC which means * that more OA data is available than could fit in the user provided * buffer. In this case we want the next poll() call to not block. */ if (ret != -ENOSPC) stream->pollin = false; /* Possible values for ret are 0, -EFAULT, -ENOSPC, -EIO, ... */ return offset ?: (ret ?: -EAGAIN); } static enum hrtimer_restart oa_poll_check_timer_cb(struct hrtimer *hrtimer) { struct i915_perf_stream *stream = container_of(hrtimer, typeof(*stream), poll_check_timer); if (oa_buffer_check_unlocked(stream)) { stream->pollin = true; wake_up(&stream->poll_wq); } hrtimer_forward_now(hrtimer, ns_to_ktime(stream->poll_oa_period)); return HRTIMER_RESTART; } /** * i915_perf_poll_locked - poll_wait() with a suitable wait queue for stream * @stream: An i915 perf stream * @file: An i915 perf stream file * @wait: poll() state table * * For handling userspace polling on an i915 perf stream, this calls through to * &i915_perf_stream_ops->poll_wait to call poll_wait() with a wait queue that * will be woken for new stream data. * * Note: The &perf->lock mutex has been taken to serialize * with any non-file-operation driver hooks. * * Returns: any poll events that are ready without sleeping */ static __poll_t i915_perf_poll_locked(struct i915_perf_stream *stream, struct file *file, poll_table *wait) { __poll_t events = 0; stream->ops->poll_wait(stream, file, wait); /* Note: we don't explicitly check whether there's something to read * here since this path may be very hot depending on what else * userspace is polling, or on the timeout in use. We rely solely on * the hrtimer/oa_poll_check_timer_cb to notify us when there are * samples to read. */ if (stream->pollin) events |= EPOLLIN; return events; } /** * i915_perf_poll - call poll_wait() with a suitable wait queue for stream * @file: An i915 perf stream file * @wait: poll() state table * * For handling userspace polling on an i915 perf stream, this ensures * poll_wait() gets called with a wait queue that will be woken for new stream * data. * * Note: Implementation deferred to i915_perf_poll_locked() * * Returns: any poll events that are ready without sleeping */ static __poll_t i915_perf_poll(struct file *file, poll_table *wait) { struct i915_perf_stream *stream = file->private_data; struct i915_perf *perf = stream->perf; __poll_t ret; mutex_lock(&perf->lock); ret = i915_perf_poll_locked(stream, file, wait); mutex_unlock(&perf->lock); return ret; } /** * i915_perf_enable_locked - handle `I915_PERF_IOCTL_ENABLE` ioctl * @stream: A disabled i915 perf stream * * [Re]enables the associated capture of data for this stream. * * If a stream was previously enabled then there's currently no intention * to provide userspace any guarantee about the preservation of previously * buffered data. */ static void i915_perf_enable_locked(struct i915_perf_stream *stream) { if (stream->enabled) return; /* Allow stream->ops->enable() to refer to this */ stream->enabled = true; if (stream->ops->enable) stream->ops->enable(stream); if (stream->hold_preemption) intel_context_set_nopreempt(stream->pinned_ctx); } /** * i915_perf_disable_locked - handle `I915_PERF_IOCTL_DISABLE` ioctl * @stream: An enabled i915 perf stream * * Disables the associated capture of data for this stream. * * The intention is that disabling an re-enabling a stream will ideally be * cheaper than destroying and re-opening a stream with the same configuration, * though there are no formal guarantees about what state or buffered data * must be retained between disabling and re-enabling a stream. * * Note: while a stream is disabled it's considered an error for userspace * to attempt to read from the stream (-EIO). */ static void i915_perf_disable_locked(struct i915_perf_stream *stream) { if (!stream->enabled) return; /* Allow stream->ops->disable() to refer to this */ stream->enabled = false; if (stream->hold_preemption) intel_context_clear_nopreempt(stream->pinned_ctx); if (stream->ops->disable) stream->ops->disable(stream); } static long i915_perf_config_locked(struct i915_perf_stream *stream, unsigned long metrics_set) { struct i915_oa_config *config; long ret = stream->oa_config->id; config = i915_perf_get_oa_config(stream->perf, metrics_set); if (!config) return -EINVAL; if (config != stream->oa_config) { int err; /* * If OA is bound to a specific context, emit the * reconfiguration inline from that context. The update * will then be ordered with respect to submission on that * context. * * When set globally, we use a low priority kernel context, * so it will effectively take effect when idle. */ err = emit_oa_config(stream, config, oa_context(stream), NULL); if (!err) config = xchg(&stream->oa_config, config); else ret = err; } i915_oa_config_put(config); return ret; } /** * i915_perf_ioctl_locked - support ioctl() usage with i915 perf stream FDs * @stream: An i915 perf stream * @cmd: the ioctl request * @arg: the ioctl data * * Note: The &perf->lock mutex has been taken to serialize * with any non-file-operation driver hooks. * * Returns: zero on success or a negative error code. Returns -EINVAL for * an unknown ioctl request. */ static long i915_perf_ioctl_locked(struct i915_perf_stream *stream, unsigned int cmd, unsigned long arg) { switch (cmd) { case I915_PERF_IOCTL_ENABLE: i915_perf_enable_locked(stream); return 0; case I915_PERF_IOCTL_DISABLE: i915_perf_disable_locked(stream); return 0; case I915_PERF_IOCTL_CONFIG: return i915_perf_config_locked(stream, arg); } return -EINVAL; } /** * i915_perf_ioctl - support ioctl() usage with i915 perf stream FDs * @file: An i915 perf stream file * @cmd: the ioctl request * @arg: the ioctl data * * Implementation deferred to i915_perf_ioctl_locked(). * * Returns: zero on success or a negative error code. Returns -EINVAL for * an unknown ioctl request. */ static long i915_perf_ioctl(struct file *file, unsigned int cmd, unsigned long arg) { struct i915_perf_stream *stream = file->private_data; struct i915_perf *perf = stream->perf; long ret; mutex_lock(&perf->lock); ret = i915_perf_ioctl_locked(stream, cmd, arg); mutex_unlock(&perf->lock); return ret; } /** * i915_perf_destroy_locked - destroy an i915 perf stream * @stream: An i915 perf stream * * Frees all resources associated with the given i915 perf @stream, disabling * any associated data capture in the process. * * Note: The &perf->lock mutex has been taken to serialize * with any non-file-operation driver hooks. */ static void i915_perf_destroy_locked(struct i915_perf_stream *stream) { if (stream->enabled) i915_perf_disable_locked(stream); if (stream->ops->destroy) stream->ops->destroy(stream); if (stream->ctx) i915_gem_context_put(stream->ctx); kfree(stream); } /** * i915_perf_release - handles userspace close() of a stream file * @inode: anonymous inode associated with file * @file: An i915 perf stream file * * Cleans up any resources associated with an open i915 perf stream file. * * NB: close() can't really fail from the userspace point of view. * * Returns: zero on success or a negative error code. */ static int i915_perf_release(struct inode *inode, struct file *file) { struct i915_perf_stream *stream = file->private_data; struct i915_perf *perf = stream->perf; mutex_lock(&perf->lock); i915_perf_destroy_locked(stream); mutex_unlock(&perf->lock); /* Release the reference the perf stream kept on the driver. */ drm_dev_put(&perf->i915->drm); return 0; } static const struct file_operations fops = { .owner = THIS_MODULE, .llseek = no_llseek, .release = i915_perf_release, .poll = i915_perf_poll, .read = i915_perf_read, .unlocked_ioctl = i915_perf_ioctl, /* Our ioctl have no arguments, so it's safe to use the same function * to handle 32bits compatibility. */ .compat_ioctl = i915_perf_ioctl, }; /** * i915_perf_open_ioctl_locked - DRM ioctl() for userspace to open a stream FD * @perf: i915 perf instance * @param: The open parameters passed to 'DRM_I915_PERF_OPEN` * @props: individually validated u64 property value pairs * @file: drm file * * See i915_perf_ioctl_open() for interface details. * * Implements further stream config validation and stream initialization on * behalf of i915_perf_open_ioctl() with the &perf->lock mutex * taken to serialize with any non-file-operation driver hooks. * * Note: at this point the @props have only been validated in isolation and * it's still necessary to validate that the combination of properties makes * sense. * * In the case where userspace is interested in OA unit metrics then further * config validation and stream initialization details will be handled by * i915_oa_stream_init(). The code here should only validate config state that * will be relevant to all stream types / backends. * * Returns: zero on success or a negative error code. */ static int i915_perf_open_ioctl_locked(struct i915_perf *perf, struct drm_i915_perf_open_param *param, struct perf_open_properties *props, struct drm_file *file) { struct i915_gem_context *specific_ctx = NULL; struct i915_perf_stream *stream = NULL; unsigned long f_flags = 0; bool privileged_op = true; int stream_fd; int ret; if (props->single_context) { u32 ctx_handle = props->ctx_handle; struct drm_i915_file_private *file_priv = file->driver_priv; specific_ctx = i915_gem_context_lookup(file_priv, ctx_handle); if (IS_ERR(specific_ctx)) { drm_dbg(&perf->i915->drm, "Failed to look up context with ID %u for opening perf stream\n", ctx_handle); ret = PTR_ERR(specific_ctx); goto err; } } /* * On Haswell the OA unit supports clock gating off for a specific * context and in this mode there's no visibility of metrics for the * rest of the system, which we consider acceptable for a * non-privileged client. * * For Gen8->11 the OA unit no longer supports clock gating off for a * specific context and the kernel can't securely stop the counters * from updating as system-wide / global values. Even though we can * filter reports based on the included context ID we can't block * clients from seeing the raw / global counter values via * MI_REPORT_PERF_COUNT commands and so consider it a privileged op to * enable the OA unit by default. * * For Gen12+ we gain a new OAR unit that only monitors the RCS on a * per context basis. So we can relax requirements there if the user * doesn't request global stream access (i.e. query based sampling * using MI_RECORD_PERF_COUNT. */ if (IS_HASWELL(perf->i915) && specific_ctx) privileged_op = false; else if (GRAPHICS_VER(perf->i915) == 12 && specific_ctx && (props->sample_flags & SAMPLE_OA_REPORT) == 0) privileged_op = false; if (props->hold_preemption) { if (!props->single_context) { drm_dbg(&perf->i915->drm, "preemption disable with no context\n"); ret = -EINVAL; goto err; } privileged_op = true; } /* * Asking for SSEU configuration is a priviliged operation. */ if (props->has_sseu) privileged_op = true; else get_default_sseu_config(&props->sseu, props->engine); /* Similar to perf's kernel.perf_paranoid_cpu sysctl option * we check a dev.i915.perf_stream_paranoid sysctl option * to determine if it's ok to access system wide OA counters * without CAP_PERFMON or CAP_SYS_ADMIN privileges. */ if (privileged_op && i915_perf_stream_paranoid && !perfmon_capable()) { drm_dbg(&perf->i915->drm, "Insufficient privileges to open i915 perf stream\n"); ret = -EACCES; goto err_ctx; } stream = kzalloc(sizeof(*stream), GFP_KERNEL); if (!stream) { ret = -ENOMEM; goto err_ctx; } stream->perf = perf; stream->ctx = specific_ctx; stream->poll_oa_period = props->poll_oa_period; ret = i915_oa_stream_init(stream, param, props); if (ret) goto err_alloc; /* we avoid simply assigning stream->sample_flags = props->sample_flags * to have _stream_init check the combination of sample flags more * thoroughly, but still this is the expected result at this point. */ if (WARN_ON(stream->sample_flags != props->sample_flags)) { ret = -ENODEV; goto err_flags; } if (param->flags & I915_PERF_FLAG_FD_CLOEXEC) f_flags |= O_CLOEXEC; if (param->flags & I915_PERF_FLAG_FD_NONBLOCK) f_flags |= O_NONBLOCK; stream_fd = anon_inode_getfd("[i915_perf]", &fops, stream, f_flags); if (stream_fd < 0) { ret = stream_fd; goto err_flags; } if (!(param->flags & I915_PERF_FLAG_DISABLED)) i915_perf_enable_locked(stream); /* Take a reference on the driver that will be kept with stream_fd * until its release. */ drm_dev_get(&perf->i915->drm); return stream_fd; err_flags: if (stream->ops->destroy) stream->ops->destroy(stream); err_alloc: kfree(stream); err_ctx: if (specific_ctx) i915_gem_context_put(specific_ctx); err: return ret; } static u64 oa_exponent_to_ns(struct i915_perf *perf, int exponent) { return intel_gt_clock_interval_to_ns(to_gt(perf->i915), 2ULL << exponent); } static __always_inline bool oa_format_valid(struct i915_perf *perf, enum drm_i915_oa_format format) { return test_bit(format, perf->format_mask); } static __always_inline void oa_format_add(struct i915_perf *perf, enum drm_i915_oa_format format) { __set_bit(format, perf->format_mask); } /** * read_properties_unlocked - validate + copy userspace stream open properties * @perf: i915 perf instance * @uprops: The array of u64 key value pairs given by userspace * @n_props: The number of key value pairs expected in @uprops * @props: The stream configuration built up while validating properties * * Note this function only validates properties in isolation it doesn't * validate that the combination of properties makes sense or that all * properties necessary for a particular kind of stream have been set. * * Note that there currently aren't any ordering requirements for properties so * we shouldn't validate or assume anything about ordering here. This doesn't * rule out defining new properties with ordering requirements in the future. */ static int read_properties_unlocked(struct i915_perf *perf, u64 __user *uprops, u32 n_props, struct perf_open_properties *props) { u64 __user *uprop = uprops; u32 i; int ret; memset(props, 0, sizeof(struct perf_open_properties)); props->poll_oa_period = DEFAULT_POLL_PERIOD_NS; if (!n_props) { drm_dbg(&perf->i915->drm, "No i915 perf properties given\n"); return -EINVAL; } /* At the moment we only support using i915-perf on the RCS. */ props->engine = intel_engine_lookup_user(perf->i915, I915_ENGINE_CLASS_RENDER, 0); if (!props->engine) { drm_dbg(&perf->i915->drm, "No RENDER-capable engines\n"); return -EINVAL; } /* Considering that ID = 0 is reserved and assuming that we don't * (currently) expect any configurations to ever specify duplicate * values for a particular property ID then the last _PROP_MAX value is * one greater than the maximum number of properties we expect to get * from userspace. */ if (n_props >= DRM_I915_PERF_PROP_MAX) { drm_dbg(&perf->i915->drm, "More i915 perf properties specified than exist\n"); return -EINVAL; } for (i = 0; i < n_props; i++) { u64 oa_period, oa_freq_hz; u64 id, value; ret = get_user(id, uprop); if (ret) return ret; ret = get_user(value, uprop + 1); if (ret) return ret; if (id == 0 || id >= DRM_I915_PERF_PROP_MAX) { drm_dbg(&perf->i915->drm, "Unknown i915 perf property ID\n"); return -EINVAL; } switch ((enum drm_i915_perf_property_id)id) { case DRM_I915_PERF_PROP_CTX_HANDLE: props->single_context = 1; props->ctx_handle = value; break; case DRM_I915_PERF_PROP_SAMPLE_OA: if (value) props->sample_flags |= SAMPLE_OA_REPORT; break; case DRM_I915_PERF_PROP_OA_METRICS_SET: if (value == 0) { drm_dbg(&perf->i915->drm, "Unknown OA metric set ID\n"); return -EINVAL; } props->metrics_set = value; break; case DRM_I915_PERF_PROP_OA_FORMAT: if (value == 0 || value >= I915_OA_FORMAT_MAX) { drm_dbg(&perf->i915->drm, "Out-of-range OA report format %llu\n", value); return -EINVAL; } if (!oa_format_valid(perf, value)) { drm_dbg(&perf->i915->drm, "Unsupported OA report format %llu\n", value); return -EINVAL; } props->oa_format = value; break; case DRM_I915_PERF_PROP_OA_EXPONENT: if (value > OA_EXPONENT_MAX) { drm_dbg(&perf->i915->drm, "OA timer exponent too high (> %u)\n", OA_EXPONENT_MAX); return -EINVAL; } /* Theoretically we can program the OA unit to sample * e.g. every 160ns for HSW, 167ns for BDW/SKL or 104ns * for BXT. We don't allow such high sampling * frequencies by default unless root. */ BUILD_BUG_ON(sizeof(oa_period) != 8); oa_period = oa_exponent_to_ns(perf, value); /* This check is primarily to ensure that oa_period <= * UINT32_MAX (before passing to do_div which only * accepts a u32 denominator), but we can also skip * checking anything < 1Hz which implicitly can't be * limited via an integer oa_max_sample_rate. */ if (oa_period <= NSEC_PER_SEC) { u64 tmp = NSEC_PER_SEC; do_div(tmp, oa_period); oa_freq_hz = tmp; } else oa_freq_hz = 0; if (oa_freq_hz > i915_oa_max_sample_rate && !perfmon_capable()) { drm_dbg(&perf->i915->drm, "OA exponent would exceed the max sampling frequency (sysctl dev.i915.oa_max_sample_rate) %uHz without CAP_PERFMON or CAP_SYS_ADMIN privileges\n", i915_oa_max_sample_rate); return -EACCES; } props->oa_periodic = true; props->oa_period_exponent = value; break; case DRM_I915_PERF_PROP_HOLD_PREEMPTION: props->hold_preemption = !!value; break; case DRM_I915_PERF_PROP_GLOBAL_SSEU: { struct drm_i915_gem_context_param_sseu user_sseu; if (GRAPHICS_VER_FULL(perf->i915) >= IP_VER(12, 50)) { drm_dbg(&perf->i915->drm, "SSEU config not supported on gfx %x\n", GRAPHICS_VER_FULL(perf->i915)); return -ENODEV; } if (copy_from_user(&user_sseu, u64_to_user_ptr(value), sizeof(user_sseu))) { drm_dbg(&perf->i915->drm, "Unable to copy global sseu parameter\n"); return -EFAULT; } ret = get_sseu_config(&props->sseu, props->engine, &user_sseu); if (ret) { drm_dbg(&perf->i915->drm, "Invalid SSEU configuration\n"); return ret; } props->has_sseu = true; break; } case DRM_I915_PERF_PROP_POLL_OA_PERIOD: if (value < 100000 /* 100us */) { drm_dbg(&perf->i915->drm, "OA availability timer too small (%lluns < 100us)\n", value); return -EINVAL; } props->poll_oa_period = value; break; case DRM_I915_PERF_PROP_MAX: MISSING_CASE(id); return -EINVAL; } uprop += 2; } return 0; } /** * i915_perf_open_ioctl - DRM ioctl() for userspace to open a stream FD * @dev: drm device * @data: ioctl data copied from userspace (unvalidated) * @file: drm file * * Validates the stream open parameters given by userspace including flags * and an array of u64 key, value pair properties. * * Very little is assumed up front about the nature of the stream being * opened (for instance we don't assume it's for periodic OA unit metrics). An * i915-perf stream is expected to be a suitable interface for other forms of * buffered data written by the GPU besides periodic OA metrics. * * Note we copy the properties from userspace outside of the i915 perf * mutex to avoid an awkward lockdep with mmap_lock. * * Most of the implementation details are handled by * i915_perf_open_ioctl_locked() after taking the &perf->lock * mutex for serializing with any non-file-operation driver hooks. * * Return: A newly opened i915 Perf stream file descriptor or negative * error code on failure. */ int i915_perf_open_ioctl(struct drm_device *dev, void *data, struct drm_file *file) { struct i915_perf *perf = &to_i915(dev)->perf; struct drm_i915_perf_open_param *param = data; struct perf_open_properties props; u32 known_open_flags; int ret; if (!perf->i915) { drm_dbg(&perf->i915->drm, "i915 perf interface not available for this system\n"); return -ENOTSUPP; } known_open_flags = I915_PERF_FLAG_FD_CLOEXEC | I915_PERF_FLAG_FD_NONBLOCK | I915_PERF_FLAG_DISABLED; if (param->flags & ~known_open_flags) { drm_dbg(&perf->i915->drm, "Unknown drm_i915_perf_open_param flag\n"); return -EINVAL; } ret = read_properties_unlocked(perf, u64_to_user_ptr(param->properties_ptr), param->num_properties, &props); if (ret) return ret; mutex_lock(&perf->lock); ret = i915_perf_open_ioctl_locked(perf, param, &props, file); mutex_unlock(&perf->lock); return ret; } /** * i915_perf_register - exposes i915-perf to userspace * @i915: i915 device instance * * In particular OA metric sets are advertised under a sysfs metrics/ * directory allowing userspace to enumerate valid IDs that can be * used to open an i915-perf stream. */ void i915_perf_register(struct drm_i915_private *i915) { struct i915_perf *perf = &i915->perf; if (!perf->i915) return; /* To be sure we're synchronized with an attempted * i915_perf_open_ioctl(); considering that we register after * being exposed to userspace. */ mutex_lock(&perf->lock); perf->metrics_kobj = kobject_create_and_add("metrics", &i915->drm.primary->kdev->kobj); mutex_unlock(&perf->lock); } /** * i915_perf_unregister - hide i915-perf from userspace * @i915: i915 device instance * * i915-perf state cleanup is split up into an 'unregister' and * 'deinit' phase where the interface is first hidden from * userspace by i915_perf_unregister() before cleaning up * remaining state in i915_perf_fini(). */ void i915_perf_unregister(struct drm_i915_private *i915) { struct i915_perf *perf = &i915->perf; if (!perf->metrics_kobj) return; kobject_put(perf->metrics_kobj); perf->metrics_kobj = NULL; } static bool gen8_is_valid_flex_addr(struct i915_perf *perf, u32 addr) { static const i915_reg_t flex_eu_regs[] = { EU_PERF_CNTL0, EU_PERF_CNTL1, EU_PERF_CNTL2, EU_PERF_CNTL3, EU_PERF_CNTL4, EU_PERF_CNTL5, EU_PERF_CNTL6, }; int i; for (i = 0; i < ARRAY_SIZE(flex_eu_regs); i++) { if (i915_mmio_reg_offset(flex_eu_regs[i]) == addr) return true; } return false; } static bool reg_in_range_table(u32 addr, const struct i915_range *table) { while (table->start || table->end) { if (addr >= table->start && addr <= table->end) return true; table++; } return false; } #define REG_EQUAL(addr, mmio) \ ((addr) == i915_mmio_reg_offset(mmio)) static const struct i915_range gen7_oa_b_counters[] = { { .start = 0x2710, .end = 0x272c }, /* OASTARTTRIG[1-8] */ { .start = 0x2740, .end = 0x275c }, /* OAREPORTTRIG[1-8] */ { .start = 0x2770, .end = 0x27ac }, /* OACEC[0-7][0-1] */ {} }; static const struct i915_range gen12_oa_b_counters[] = { { .start = 0x2b2c, .end = 0x2b2c }, /* GEN12_OAG_OA_PESS */ { .start = 0xd900, .end = 0xd91c }, /* GEN12_OAG_OASTARTTRIG[1-8] */ { .start = 0xd920, .end = 0xd93c }, /* GEN12_OAG_OAREPORTTRIG1[1-8] */ { .start = 0xd940, .end = 0xd97c }, /* GEN12_OAG_CEC[0-7][0-1] */ { .start = 0xdc00, .end = 0xdc3c }, /* GEN12_OAG_SCEC[0-7][0-1] */ { .start = 0xdc40, .end = 0xdc40 }, /* GEN12_OAG_SPCTR_CNF */ { .start = 0xdc44, .end = 0xdc44 }, /* GEN12_OAA_DBG_REG */ {} }; static const struct i915_range gen7_oa_mux_regs[] = { { .start = 0x91b8, .end = 0x91cc }, /* OA_PERFCNT[1-2], OA_PERFMATRIX */ { .start = 0x9800, .end = 0x9888 }, /* MICRO_BP0_0 - NOA_WRITE */ { .start = 0xe180, .end = 0xe180 }, /* HALF_SLICE_CHICKEN2 */ {} }; static const struct i915_range hsw_oa_mux_regs[] = { { .start = 0x09e80, .end = 0x09ea4 }, /* HSW_MBVID2_NOA[0-9] */ { .start = 0x09ec0, .end = 0x09ec0 }, /* HSW_MBVID2_MISR0 */ { .start = 0x25100, .end = 0x2ff90 }, {} }; static const struct i915_range chv_oa_mux_regs[] = { { .start = 0x182300, .end = 0x1823a4 }, {} }; static const struct i915_range gen8_oa_mux_regs[] = { { .start = 0x0d00, .end = 0x0d2c }, /* RPM_CONFIG[0-1], NOA_CONFIG[0-8] */ { .start = 0x20cc, .end = 0x20cc }, /* WAIT_FOR_RC6_EXIT */ {} }; static const struct i915_range gen11_oa_mux_regs[] = { { .start = 0x91c8, .end = 0x91dc }, /* OA_PERFCNT[3-4] */ {} }; static const struct i915_range gen12_oa_mux_regs[] = { { .start = 0x0d00, .end = 0x0d04 }, /* RPM_CONFIG[0-1] */ { .start = 0x0d0c, .end = 0x0d2c }, /* NOA_CONFIG[0-8] */ { .start = 0x9840, .end = 0x9840 }, /* GDT_CHICKEN_BITS */ { .start = 0x9884, .end = 0x9888 }, /* NOA_WRITE */ { .start = 0x20cc, .end = 0x20cc }, /* WAIT_FOR_RC6_EXIT */ {} }; static bool gen7_is_valid_b_counter_addr(struct i915_perf *perf, u32 addr) { return reg_in_range_table(addr, gen7_oa_b_counters); } static bool gen8_is_valid_mux_addr(struct i915_perf *perf, u32 addr) { return reg_in_range_table(addr, gen7_oa_mux_regs) || reg_in_range_table(addr, gen8_oa_mux_regs); } static bool gen11_is_valid_mux_addr(struct i915_perf *perf, u32 addr) { return reg_in_range_table(addr, gen7_oa_mux_regs) || reg_in_range_table(addr, gen8_oa_mux_regs) || reg_in_range_table(addr, gen11_oa_mux_regs); } static bool hsw_is_valid_mux_addr(struct i915_perf *perf, u32 addr) { return reg_in_range_table(addr, gen7_oa_mux_regs) || reg_in_range_table(addr, hsw_oa_mux_regs); } static bool chv_is_valid_mux_addr(struct i915_perf *perf, u32 addr) { return reg_in_range_table(addr, gen7_oa_mux_regs) || reg_in_range_table(addr, chv_oa_mux_regs); } static bool gen12_is_valid_b_counter_addr(struct i915_perf *perf, u32 addr) { return reg_in_range_table(addr, gen12_oa_b_counters); } static bool gen12_is_valid_mux_addr(struct i915_perf *perf, u32 addr) { return reg_in_range_table(addr, gen12_oa_mux_regs); } static u32 mask_reg_value(u32 reg, u32 val) { /* HALF_SLICE_CHICKEN2 is programmed with a the * WaDisableSTUnitPowerOptimization workaround. Make sure the value * programmed by userspace doesn't change this. */ if (REG_EQUAL(reg, HALF_SLICE_CHICKEN2)) val = val & ~_MASKED_BIT_ENABLE(GEN8_ST_PO_DISABLE); /* WAIT_FOR_RC6_EXIT has only one bit fullfilling the function * indicated by its name and a bunch of selection fields used by OA * configs. */ if (REG_EQUAL(reg, WAIT_FOR_RC6_EXIT)) val = val & ~_MASKED_BIT_ENABLE(HSW_WAIT_FOR_RC6_EXIT_ENABLE); return val; } static struct i915_oa_reg *alloc_oa_regs(struct i915_perf *perf, bool (*is_valid)(struct i915_perf *perf, u32 addr), u32 __user *regs, u32 n_regs) { struct i915_oa_reg *oa_regs; int err; u32 i; if (!n_regs) return NULL; /* No is_valid function means we're not allowing any register to be programmed. */ GEM_BUG_ON(!is_valid); if (!is_valid) return ERR_PTR(-EINVAL); oa_regs = kmalloc_array(n_regs, sizeof(*oa_regs), GFP_KERNEL); if (!oa_regs) return ERR_PTR(-ENOMEM); for (i = 0; i < n_regs; i++) { u32 addr, value; err = get_user(addr, regs); if (err) goto addr_err; if (!is_valid(perf, addr)) { drm_dbg(&perf->i915->drm, "Invalid oa_reg address: %X\n", addr); err = -EINVAL; goto addr_err; } err = get_user(value, regs + 1); if (err) goto addr_err; oa_regs[i].addr = _MMIO(addr); oa_regs[i].value = mask_reg_value(addr, value); regs += 2; } return oa_regs; addr_err: kfree(oa_regs); return ERR_PTR(err); } static ssize_t show_dynamic_id(struct kobject *kobj, struct kobj_attribute *attr, char *buf) { struct i915_oa_config *oa_config = container_of(attr, typeof(*oa_config), sysfs_metric_id); return sprintf(buf, "%d\n", oa_config->id); } static int create_dynamic_oa_sysfs_entry(struct i915_perf *perf, struct i915_oa_config *oa_config) { sysfs_attr_init(&oa_config->sysfs_metric_id.attr); oa_config->sysfs_metric_id.attr.name = "id"; oa_config->sysfs_metric_id.attr.mode = S_IRUGO; oa_config->sysfs_metric_id.show = show_dynamic_id; oa_config->sysfs_metric_id.store = NULL; oa_config->attrs[0] = &oa_config->sysfs_metric_id.attr; oa_config->attrs[1] = NULL; oa_config->sysfs_metric.name = oa_config->uuid; oa_config->sysfs_metric.attrs = oa_config->attrs; return sysfs_create_group(perf->metrics_kobj, &oa_config->sysfs_metric); } /** * i915_perf_add_config_ioctl - DRM ioctl() for userspace to add a new OA config * @dev: drm device * @data: ioctl data (pointer to struct drm_i915_perf_oa_config) copied from * userspace (unvalidated) * @file: drm file * * Validates the submitted OA register to be saved into a new OA config that * can then be used for programming the OA unit and its NOA network. * * Returns: A new allocated config number to be used with the perf open ioctl * or a negative error code on failure. */ int i915_perf_add_config_ioctl(struct drm_device *dev, void *data, struct drm_file *file) { struct i915_perf *perf = &to_i915(dev)->perf; struct drm_i915_perf_oa_config *args = data; struct i915_oa_config *oa_config, *tmp; struct i915_oa_reg *regs; int err, id; if (!perf->i915) { drm_dbg(&perf->i915->drm, "i915 perf interface not available for this system\n"); return -ENOTSUPP; } if (!perf->metrics_kobj) { drm_dbg(&perf->i915->drm, "OA metrics weren't advertised via sysfs\n"); return -EINVAL; } if (i915_perf_stream_paranoid && !perfmon_capable()) { drm_dbg(&perf->i915->drm, "Insufficient privileges to add i915 OA config\n"); return -EACCES; } if ((!args->mux_regs_ptr || !args->n_mux_regs) && (!args->boolean_regs_ptr || !args->n_boolean_regs) && (!args->flex_regs_ptr || !args->n_flex_regs)) { drm_dbg(&perf->i915->drm, "No OA registers given\n"); return -EINVAL; } oa_config = kzalloc(sizeof(*oa_config), GFP_KERNEL); if (!oa_config) { drm_dbg(&perf->i915->drm, "Failed to allocate memory for the OA config\n"); return -ENOMEM; } oa_config->perf = perf; kref_init(&oa_config->ref); if (!uuid_is_valid(args->uuid)) { drm_dbg(&perf->i915->drm, "Invalid uuid format for OA config\n"); err = -EINVAL; goto reg_err; } /* Last character in oa_config->uuid will be 0 because oa_config is * kzalloc. */ memcpy(oa_config->uuid, args->uuid, sizeof(args->uuid)); oa_config->mux_regs_len = args->n_mux_regs; regs = alloc_oa_regs(perf, perf->ops.is_valid_mux_reg, u64_to_user_ptr(args->mux_regs_ptr), args->n_mux_regs); if (IS_ERR(regs)) { drm_dbg(&perf->i915->drm, "Failed to create OA config for mux_regs\n"); err = PTR_ERR(regs); goto reg_err; } oa_config->mux_regs = regs; oa_config->b_counter_regs_len = args->n_boolean_regs; regs = alloc_oa_regs(perf, perf->ops.is_valid_b_counter_reg, u64_to_user_ptr(args->boolean_regs_ptr), args->n_boolean_regs); if (IS_ERR(regs)) { drm_dbg(&perf->i915->drm, "Failed to create OA config for b_counter_regs\n"); err = PTR_ERR(regs); goto reg_err; } oa_config->b_counter_regs = regs; if (GRAPHICS_VER(perf->i915) < 8) { if (args->n_flex_regs != 0) { err = -EINVAL; goto reg_err; } } else { oa_config->flex_regs_len = args->n_flex_regs; regs = alloc_oa_regs(perf, perf->ops.is_valid_flex_reg, u64_to_user_ptr(args->flex_regs_ptr), args->n_flex_regs); if (IS_ERR(regs)) { drm_dbg(&perf->i915->drm, "Failed to create OA config for flex_regs\n"); err = PTR_ERR(regs); goto reg_err; } oa_config->flex_regs = regs; } err = mutex_lock_interruptible(&perf->metrics_lock); if (err) goto reg_err; /* We shouldn't have too many configs, so this iteration shouldn't be * too costly. */ idr_for_each_entry(&perf->metrics_idr, tmp, id) { if (!strcmp(tmp->uuid, oa_config->uuid)) { drm_dbg(&perf->i915->drm, "OA config already exists with this uuid\n"); err = -EADDRINUSE; goto sysfs_err; } } err = create_dynamic_oa_sysfs_entry(perf, oa_config); if (err) { drm_dbg(&perf->i915->drm, "Failed to create sysfs entry for OA config\n"); goto sysfs_err; } /* Config id 0 is invalid, id 1 for kernel stored test config. */ oa_config->id = idr_alloc(&perf->metrics_idr, oa_config, 2, 0, GFP_KERNEL); if (oa_config->id < 0) { drm_dbg(&perf->i915->drm, "Failed to create sysfs entry for OA config\n"); err = oa_config->id; goto sysfs_err; } id = oa_config->id; drm_dbg(&perf->i915->drm, "Added config %s id=%i\n", oa_config->uuid, oa_config->id); mutex_unlock(&perf->metrics_lock); return id; sysfs_err: mutex_unlock(&perf->metrics_lock); reg_err: i915_oa_config_put(oa_config); drm_dbg(&perf->i915->drm, "Failed to add new OA config\n"); return err; } /** * i915_perf_remove_config_ioctl - DRM ioctl() for userspace to remove an OA config * @dev: drm device * @data: ioctl data (pointer to u64 integer) copied from userspace * @file: drm file * * Configs can be removed while being used, the will stop appearing in sysfs * and their content will be freed when the stream using the config is closed. * * Returns: 0 on success or a negative error code on failure. */ int i915_perf_remove_config_ioctl(struct drm_device *dev, void *data, struct drm_file *file) { struct i915_perf *perf = &to_i915(dev)->perf; u64 *arg = data; struct i915_oa_config *oa_config; int ret; if (!perf->i915) { drm_dbg(&perf->i915->drm, "i915 perf interface not available for this system\n"); return -ENOTSUPP; } if (i915_perf_stream_paranoid && !perfmon_capable()) { drm_dbg(&perf->i915->drm, "Insufficient privileges to remove i915 OA config\n"); return -EACCES; } ret = mutex_lock_interruptible(&perf->metrics_lock); if (ret) return ret; oa_config = idr_find(&perf->metrics_idr, *arg); if (!oa_config) { drm_dbg(&perf->i915->drm, "Failed to remove unknown OA config\n"); ret = -ENOENT; goto err_unlock; } GEM_BUG_ON(*arg != oa_config->id); sysfs_remove_group(perf->metrics_kobj, &oa_config->sysfs_metric); idr_remove(&perf->metrics_idr, *arg); mutex_unlock(&perf->metrics_lock); drm_dbg(&perf->i915->drm, "Removed config %s id=%i\n", oa_config->uuid, oa_config->id); i915_oa_config_put(oa_config); return 0; err_unlock: mutex_unlock(&perf->metrics_lock); return ret; } static struct ctl_table oa_table[] = { { .procname = "perf_stream_paranoid", .data = &i915_perf_stream_paranoid, .maxlen = sizeof(i915_perf_stream_paranoid), .mode = 0644, .proc_handler = proc_dointvec_minmax, .extra1 = SYSCTL_ZERO, .extra2 = SYSCTL_ONE, }, { .procname = "oa_max_sample_rate", .data = &i915_oa_max_sample_rate, .maxlen = sizeof(i915_oa_max_sample_rate), .mode = 0644, .proc_handler = proc_dointvec_minmax, .extra1 = SYSCTL_ZERO, .extra2 = &oa_sample_rate_hard_limit, }, {} }; static void oa_init_supported_formats(struct i915_perf *perf) { struct drm_i915_private *i915 = perf->i915; enum intel_platform platform = INTEL_INFO(i915)->platform; switch (platform) { case INTEL_HASWELL: oa_format_add(perf, I915_OA_FORMAT_A13); oa_format_add(perf, I915_OA_FORMAT_A13); oa_format_add(perf, I915_OA_FORMAT_A29); oa_format_add(perf, I915_OA_FORMAT_A13_B8_C8); oa_format_add(perf, I915_OA_FORMAT_B4_C8); oa_format_add(perf, I915_OA_FORMAT_A45_B8_C8); oa_format_add(perf, I915_OA_FORMAT_B4_C8_A16); oa_format_add(perf, I915_OA_FORMAT_C4_B8); break; case INTEL_BROADWELL: case INTEL_CHERRYVIEW: case INTEL_SKYLAKE: case INTEL_BROXTON: case INTEL_KABYLAKE: case INTEL_GEMINILAKE: case INTEL_COFFEELAKE: case INTEL_COMETLAKE: case INTEL_ICELAKE: case INTEL_ELKHARTLAKE: case INTEL_JASPERLAKE: case INTEL_TIGERLAKE: case INTEL_ROCKETLAKE: case INTEL_DG1: case INTEL_ALDERLAKE_S: case INTEL_ALDERLAKE_P: oa_format_add(perf, I915_OA_FORMAT_A12); oa_format_add(perf, I915_OA_FORMAT_A12_B8_C8); oa_format_add(perf, I915_OA_FORMAT_A32u40_A4u32_B8_C8); oa_format_add(perf, I915_OA_FORMAT_C4_B8); break; default: MISSING_CASE(platform); } } /** * i915_perf_init - initialize i915-perf state on module bind * @i915: i915 device instance * * Initializes i915-perf state without exposing anything to userspace. * * Note: i915-perf initialization is split into an 'init' and 'register' * phase with the i915_perf_register() exposing state to userspace. */ void i915_perf_init(struct drm_i915_private *i915) { struct i915_perf *perf = &i915->perf; /* XXX const struct i915_perf_ops! */ /* i915_perf is not enabled for DG2 yet */ if (IS_DG2(i915)) return; perf->oa_formats = oa_formats; if (IS_HASWELL(i915)) { perf->ops.is_valid_b_counter_reg = gen7_is_valid_b_counter_addr; perf->ops.is_valid_mux_reg = hsw_is_valid_mux_addr; perf->ops.is_valid_flex_reg = NULL; perf->ops.enable_metric_set = hsw_enable_metric_set; perf->ops.disable_metric_set = hsw_disable_metric_set; perf->ops.oa_enable = gen7_oa_enable; perf->ops.oa_disable = gen7_oa_disable; perf->ops.read = gen7_oa_read; perf->ops.oa_hw_tail_read = gen7_oa_hw_tail_read; } else if (HAS_LOGICAL_RING_CONTEXTS(i915)) { /* Note: that although we could theoretically also support the * legacy ringbuffer mode on BDW (and earlier iterations of * this driver, before upstreaming did this) it didn't seem * worth the complexity to maintain now that BDW+ enable * execlist mode by default. */ perf->ops.read = gen8_oa_read; if (IS_GRAPHICS_VER(i915, 8, 9)) { perf->ops.is_valid_b_counter_reg = gen7_is_valid_b_counter_addr; perf->ops.is_valid_mux_reg = gen8_is_valid_mux_addr; perf->ops.is_valid_flex_reg = gen8_is_valid_flex_addr; if (IS_CHERRYVIEW(i915)) { perf->ops.is_valid_mux_reg = chv_is_valid_mux_addr; } perf->ops.oa_enable = gen8_oa_enable; perf->ops.oa_disable = gen8_oa_disable; perf->ops.enable_metric_set = gen8_enable_metric_set; perf->ops.disable_metric_set = gen8_disable_metric_set; perf->ops.oa_hw_tail_read = gen8_oa_hw_tail_read; if (GRAPHICS_VER(i915) == 8) { perf->ctx_oactxctrl_offset = 0x120; perf->ctx_flexeu0_offset = 0x2ce; perf->gen8_valid_ctx_bit = BIT(25); } else { perf->ctx_oactxctrl_offset = 0x128; perf->ctx_flexeu0_offset = 0x3de; perf->gen8_valid_ctx_bit = BIT(16); } } else if (GRAPHICS_VER(i915) == 11) { perf->ops.is_valid_b_counter_reg = gen7_is_valid_b_counter_addr; perf->ops.is_valid_mux_reg = gen11_is_valid_mux_addr; perf->ops.is_valid_flex_reg = gen8_is_valid_flex_addr; perf->ops.oa_enable = gen8_oa_enable; perf->ops.oa_disable = gen8_oa_disable; perf->ops.enable_metric_set = gen8_enable_metric_set; perf->ops.disable_metric_set = gen11_disable_metric_set; perf->ops.oa_hw_tail_read = gen8_oa_hw_tail_read; perf->ctx_oactxctrl_offset = 0x124; perf->ctx_flexeu0_offset = 0x78e; perf->gen8_valid_ctx_bit = BIT(16); } else if (GRAPHICS_VER(i915) == 12) { perf->ops.is_valid_b_counter_reg = gen12_is_valid_b_counter_addr; perf->ops.is_valid_mux_reg = gen12_is_valid_mux_addr; perf->ops.is_valid_flex_reg = gen8_is_valid_flex_addr; perf->ops.oa_enable = gen12_oa_enable; perf->ops.oa_disable = gen12_oa_disable; perf->ops.enable_metric_set = gen12_enable_metric_set; perf->ops.disable_metric_set = gen12_disable_metric_set; perf->ops.oa_hw_tail_read = gen12_oa_hw_tail_read; perf->ctx_flexeu0_offset = 0; perf->ctx_oactxctrl_offset = 0x144; } } if (perf->ops.enable_metric_set) { mutex_init(&perf->lock); /* Choose a representative limit */ oa_sample_rate_hard_limit = to_gt(i915)->clock_frequency / 2; mutex_init(&perf->metrics_lock); idr_init_base(&perf->metrics_idr, 1); /* We set up some ratelimit state to potentially throttle any * _NOTES about spurious, invalid OA reports which we don't * forward to userspace. * * We print a _NOTE about any throttling when closing the * stream instead of waiting until driver _fini which no one * would ever see. * * Using the same limiting factors as printk_ratelimit() */ ratelimit_state_init(&perf->spurious_report_rs, 5 * HZ, 10); /* Since we use a DRM_NOTE for spurious reports it would be * inconsistent to let __ratelimit() automatically print a * warning for throttling. */ ratelimit_set_flags(&perf->spurious_report_rs, RATELIMIT_MSG_ON_RELEASE); ratelimit_state_init(&perf->tail_pointer_race, 5 * HZ, 10); ratelimit_set_flags(&perf->tail_pointer_race, RATELIMIT_MSG_ON_RELEASE); atomic64_set(&perf->noa_programming_delay, 500 * 1000 /* 500us */); perf->i915 = i915; oa_init_supported_formats(perf); } } static int destroy_config(int id, void *p, void *data) { i915_oa_config_put(p); return 0; } int i915_perf_sysctl_register(void) { sysctl_header = register_sysctl("dev/i915", oa_table); return 0; } void i915_perf_sysctl_unregister(void) { unregister_sysctl_table(sysctl_header); } /** * i915_perf_fini - Counter part to i915_perf_init() * @i915: i915 device instance */ void i915_perf_fini(struct drm_i915_private *i915) { struct i915_perf *perf = &i915->perf; if (!perf->i915) return; idr_for_each(&perf->metrics_idr, destroy_config, perf); idr_destroy(&perf->metrics_idr); memset(&perf->ops, 0, sizeof(perf->ops)); perf->i915 = NULL; } /** * i915_perf_ioctl_version - Version of the i915-perf subsystem * * This version number is used by userspace to detect available features. */ int i915_perf_ioctl_version(void) { /* * 1: Initial version * I915_PERF_IOCTL_ENABLE * I915_PERF_IOCTL_DISABLE * * 2: Added runtime modification of OA config. * I915_PERF_IOCTL_CONFIG * * 3: Add DRM_I915_PERF_PROP_HOLD_PREEMPTION parameter to hold * preemption on a particular context so that performance data is * accessible from a delta of MI_RPC reports without looking at the * OA buffer. * * 4: Add DRM_I915_PERF_PROP_ALLOWED_SSEU to limit what contexts can * be run for the duration of the performance recording based on * their SSEU configuration. * * 5: Add DRM_I915_PERF_PROP_POLL_OA_PERIOD parameter that controls the * interval for the hrtimer used to check for OA data. */ return 5; } #if IS_ENABLED(CONFIG_DRM_I915_SELFTEST) #include "selftests/i915_perf.c" #endif