From 5b7c4cabbb65f5c469464da6c5f614cbd7f730f2 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Tue, 21 Feb 2023 18:24:12 -0800 Subject: Merge tag 'net-next-6.3' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-next Pull networking updates from Jakub Kicinski: "Core: - Add dedicated kmem_cache for typical/small skb->head, avoid having to access struct page at kfree time, and improve memory use. - Introduce sysctl to set default RPS configuration for new netdevs. - Define Netlink protocol specification format which can be used to describe messages used by each family and auto-generate parsers. Add tools for generating kernel data structures and uAPI headers. - Expose all net/core sysctls inside netns. - Remove 4s sleep in netpoll if carrier is instantly detected on boot. - Add configurable limit of MDB entries per port, and port-vlan. - Continue populating drop reasons throughout the stack. - Retire a handful of legacy Qdiscs and classifiers. Protocols: - Support IPv4 big TCP (TSO frames larger than 64kB). - Add IP_LOCAL_PORT_RANGE socket option, to control local port range on socket by socket basis. - Track and report in procfs number of MPTCP sockets used. - Support mixing IPv4 and IPv6 flows in the in-kernel MPTCP path manager. - IPv6: don't check net.ipv6.route.max_size and rely on garbage collection to free memory (similarly to IPv4). - Support Penultimate Segment Pop (PSP) flavor in SRv6 (RFC8986). - ICMP: add per-rate limit counters. - Add support for user scanning requests in ieee802154. - Remove static WEP support. - Support minimal Wi-Fi 7 Extremely High Throughput (EHT) rate reporting. - WiFi 7 EHT channel puncturing support (client & AP). BPF: - Add a rbtree data structure following the "next-gen data structure" precedent set by recently added linked list, that is, by using kfunc + kptr instead of adding a new BPF map type. - Expose XDP hints via kfuncs with initial support for RX hash and timestamp metadata. - Add BPF_F_NO_TUNNEL_KEY extension to bpf_skb_set_tunnel_key to better support decap on GRE tunnel devices not operating in collect metadata. - Improve x86 JIT's codegen for PROBE_MEM runtime error checks. - Remove the need for trace_printk_lock for bpf_trace_printk and bpf_trace_vprintk helpers. - Extend libbpf's bpf_tracing.h support for tracing arguments of kprobes/uprobes and syscall as a special case. - Significantly reduce the search time for module symbols by livepatch and BPF. - Enable cpumasks to be used as kptrs, which is useful for tracing programs tracking which tasks end up running on which CPUs in different time intervals. - Add support for BPF trampoline on s390x and riscv64. - Add capability to export the XDP features supported by the NIC. - Add __bpf_kfunc tag for marking kernel functions as kfuncs. - Add cgroup.memory=nobpf kernel parameter option to disable BPF memory accounting for container environments. Netfilter: - Remove the CLUSTERIP target. It has been marked as obsolete for years, and we still have WARN splats wrt races of the out-of-band /proc interface installed by this target. - Add 'destroy' commands to nf_tables. They are identical to the existing 'delete' commands, but do not return an error if the referenced object (set, chain, rule...) did not exist. Driver API: - Improve cpumask_local_spread() locality to help NICs set the right IRQ affinity on AMD platforms. - Separate C22 and C45 MDIO bus transactions more clearly. - Introduce new DCB table to control DSCP rewrite on egress. - Support configuration of Physical Layer Collision Avoidance (PLCA) Reconciliation Sublayer (RS) (802.3cg-2019). Modern version of shared medium Ethernet. - Support for MAC Merge layer (IEEE 802.3-2018 clause 99). Allowing preemption of low priority frames by high priority frames. - Add support for controlling MACSec offload using netlink SET. - Rework devlink instance refcounts to allow registration and de-registration under the instance lock. Split the code into multiple files, drop some of the unnecessarily granular locks and factor out common parts of netlink operation handling. - Add TX frame aggregation parameters (for USB drivers). - Add a new attr TCA_EXT_WARN_MSG to report TC (offload) warning messages with notifications for debug. - Allow offloading of UDP NEW connections via act_ct. - Add support for per action HW stats in TC. - Support hardware miss to TC action (continue processing in SW from a specific point in the action chain). - Warn if old Wireless Extension user space interface is used with modern cfg80211/mac80211 drivers. Do not support Wireless Extensions for Wi-Fi 7 devices at all. Everyone should switch to using nl80211 interface instead. - Improve the CAN bit timing configuration. Use extack to return error messages directly to user space, update the SJW handling, including the definition of a new default value that will benefit CAN-FD controllers, by increasing their oscillator tolerance. New hardware / drivers: - Ethernet: - nVidia BlueField-3 support (control traffic driver) - Ethernet support for imx93 SoCs - Motorcomm yt8531 gigabit Ethernet PHY - onsemi NCN26000 10BASE-T1S PHY (with support for PLCA) - Microchip LAN8841 PHY (incl. cable diagnostics and PTP) - Amlogic gxl MDIO mux - WiFi: - RealTek RTL8188EU (rtl8xxxu) - Qualcomm Wi-Fi 7 devices (ath12k) - CAN: - Renesas R-Car V4H Drivers: - Bluetooth: - Set Per Platform Antenna Gain (PPAG) for Intel controllers. - Ethernet NICs: - Intel (1G, igc): - support TSN / Qbv / packet scheduling features of i226 model - Intel (100G, ice): - use GNSS subsystem instead of TTY - multi-buffer XDP support - extend support for GPIO pins to E823 devices - nVidia/Mellanox: - update the shared buffer configuration on PFC commands - implement PTP adjphase function for HW offset control - TC support for Geneve and GRE with VF tunnel offload - more efficient crypto key management method - multi-port eswitch support - Netronome/Corigine: - add DCB IEEE support - support IPsec offloading for NFP3800 - Freescale/NXP (enetc): - support XDP_REDIRECT for XDP non-linear buffers - improve reconfig, avoid link flap and waiting for idle - support MAC Merge layer - Other NICs: - sfc/ef100: add basic devlink support for ef100 - ionic: rx_push mode operation (writing descriptors via MMIO) - bnxt: use the auxiliary bus abstraction for RDMA - r8169: disable ASPM and reset bus in case of tx timeout - cpsw: support QSGMII mode for J721e CPSW9G - cpts: support pulse-per-second output - ngbe: add an mdio bus driver - usbnet: optimize usbnet_bh() by avoiding unnecessary queuing - r8152: handle devices with FW with NCM support - amd-xgbe: support 10Mbps, 2.5GbE speeds and rx-adaptation - virtio-net: support multi buffer XDP - virtio/vsock: replace virtio_vsock_pkt with sk_buff - tsnep: XDP support - Ethernet high-speed switches: - nVidia/Mellanox (mlxsw): - add support for latency TLV (in FW control messages) - Microchip (sparx5): - separate explicit and implicit traffic forwarding rules, make the implicit rules always active - add support for egress DSCP rewrite - IS0 VCAP support (Ingress Classification) - IS2 VCAP filters (protos, L3 addrs, L4 ports, flags, ToS etc.) - ES2 VCAP support (Egress Access Control) - support for Per-Stream Filtering and Policing (802.1Q, 8.6.5.1) - Ethernet embedded switches: - Marvell (mv88e6xxx): - add MAB (port auth) offload support - enable PTP receive for mv88e6390 - NXP (ocelot): - support MAC Merge layer - support for the the vsc7512 internal copper phys - Microchip: - lan9303: convert to PHYLINK - lan966x: support TC flower filter statistics - lan937x: PTP support for KSZ9563/KSZ8563 and LAN937x - lan937x: support Credit Based Shaper configuration - ksz9477: support Energy Efficient Ethernet - other: - qca8k: convert to regmap read/write API, use bulk operations - rswitch: Improve TX timestamp accuracy - Intel WiFi (iwlwifi): - EHT (Wi-Fi 7) rate reporting - STEP equalizer support: transfer some STEP (connection to radio on platforms with integrated wifi) related parameters from the BIOS to the firmware. - Qualcomm 802.11ax WiFi (ath11k): - IPQ5018 support - Fine Timing Measurement (FTM) responder role support - channel 177 support - MediaTek WiFi (mt76): - per-PHY LED support - mt7996: EHT (Wi-Fi 7) support - Wireless Ethernet Dispatch (WED) reset support - switch to using page pool allocator - RealTek WiFi (rtw89): - support new version of Bluetooth co-existance - Mobile: - rmnet: support TX aggregation" * tag 'net-next-6.3' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-next: (1872 commits) page_pool: add a comment explaining the fragment counter usage net: ethtool: fix __ethtool_dev_mm_supported() implementation ethtool: pse-pd: Fix double word in comments xsk: add linux/vmalloc.h to xsk.c sefltests: netdevsim: wait for devlink instance after netns removal selftest: fib_tests: Always cleanup before exit net/mlx5e: Align IPsec ASO result memory to be as required by hardware net/mlx5e: TC, Set CT miss to the specific ct action instance net/mlx5e: Rename CHAIN_TO_REG to MAPPED_OBJ_TO_REG net/mlx5: Refactor tc miss handling to a single function net/mlx5: Kconfig: Make tc offload depend on tc skb extension net/sched: flower: Support hardware miss to tc action net/sched: flower: Move filter handle initialization earlier net/sched: cls_api: Support hardware miss to tc action net/sched: Rename user cookie and act cookie sfc: fix builds without CONFIG_RTC_LIB sfc: clean up some inconsistent indentings net/mlx4_en: Introduce flexible array to silence overflow warning net: lan966x: Fix possible deadlock inside PTP net/ulp: Remove redundant ->clone() test in inet_clone_ulp(). ... --- drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c | 965 ++++++++++++++++++++++++++++++++ 1 file changed, 965 insertions(+) create mode 100644 drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c (limited to 'drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c new file mode 100644 index 000000000..d2139ac12 --- /dev/null +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c @@ -0,0 +1,965 @@ +/* + * Copyright 2015 Advanced Micro Devices, Inc. + * + * 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 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 COPYRIGHT HOLDER(S) OR AUTHOR(S) 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: monk liu + */ + +#include +#include +#include "amdgpu.h" +#include "amdgpu_sched.h" +#include "amdgpu_ras.h" +#include + +#define to_amdgpu_ctx_entity(e) \ + container_of((e), struct amdgpu_ctx_entity, entity) + +const unsigned int amdgpu_ctx_num_entities[AMDGPU_HW_IP_NUM] = { + [AMDGPU_HW_IP_GFX] = 1, + [AMDGPU_HW_IP_COMPUTE] = 4, + [AMDGPU_HW_IP_DMA] = 2, + [AMDGPU_HW_IP_UVD] = 1, + [AMDGPU_HW_IP_VCE] = 1, + [AMDGPU_HW_IP_UVD_ENC] = 1, + [AMDGPU_HW_IP_VCN_DEC] = 1, + [AMDGPU_HW_IP_VCN_ENC] = 1, + [AMDGPU_HW_IP_VCN_JPEG] = 1, +}; + +bool amdgpu_ctx_priority_is_valid(int32_t ctx_prio) +{ + switch (ctx_prio) { + case AMDGPU_CTX_PRIORITY_UNSET: + case AMDGPU_CTX_PRIORITY_VERY_LOW: + case AMDGPU_CTX_PRIORITY_LOW: + case AMDGPU_CTX_PRIORITY_NORMAL: + case AMDGPU_CTX_PRIORITY_HIGH: + case AMDGPU_CTX_PRIORITY_VERY_HIGH: + return true; + default: + return false; + } +} + +static enum drm_sched_priority +amdgpu_ctx_to_drm_sched_prio(int32_t ctx_prio) +{ + switch (ctx_prio) { + case AMDGPU_CTX_PRIORITY_UNSET: + return DRM_SCHED_PRIORITY_UNSET; + + case AMDGPU_CTX_PRIORITY_VERY_LOW: + return DRM_SCHED_PRIORITY_MIN; + + case AMDGPU_CTX_PRIORITY_LOW: + return DRM_SCHED_PRIORITY_MIN; + + case AMDGPU_CTX_PRIORITY_NORMAL: + return DRM_SCHED_PRIORITY_NORMAL; + + case AMDGPU_CTX_PRIORITY_HIGH: + return DRM_SCHED_PRIORITY_HIGH; + + case AMDGPU_CTX_PRIORITY_VERY_HIGH: + return DRM_SCHED_PRIORITY_HIGH; + + /* This should not happen as we sanitized userspace provided priority + * already, WARN if this happens. + */ + default: + WARN(1, "Invalid context priority %d\n", ctx_prio); + return DRM_SCHED_PRIORITY_NORMAL; + } + +} + +static int amdgpu_ctx_priority_permit(struct drm_file *filp, + int32_t priority) +{ + if (!amdgpu_ctx_priority_is_valid(priority)) + return -EINVAL; + + /* NORMAL and below are accessible by everyone */ + if (priority <= AMDGPU_CTX_PRIORITY_NORMAL) + return 0; + + if (capable(CAP_SYS_NICE)) + return 0; + + if (drm_is_current_master(filp)) + return 0; + + return -EACCES; +} + +static enum amdgpu_gfx_pipe_priority amdgpu_ctx_prio_to_gfx_pipe_prio(int32_t prio) +{ + switch (prio) { + case AMDGPU_CTX_PRIORITY_HIGH: + case AMDGPU_CTX_PRIORITY_VERY_HIGH: + return AMDGPU_GFX_PIPE_PRIO_HIGH; + default: + return AMDGPU_GFX_PIPE_PRIO_NORMAL; + } +} + +static enum amdgpu_ring_priority_level amdgpu_ctx_sched_prio_to_ring_prio(int32_t prio) +{ + switch (prio) { + case AMDGPU_CTX_PRIORITY_HIGH: + return AMDGPU_RING_PRIO_1; + case AMDGPU_CTX_PRIORITY_VERY_HIGH: + return AMDGPU_RING_PRIO_2; + default: + return AMDGPU_RING_PRIO_0; + } +} + +static unsigned int amdgpu_ctx_get_hw_prio(struct amdgpu_ctx *ctx, u32 hw_ip) +{ + struct amdgpu_device *adev = ctx->mgr->adev; + unsigned int hw_prio; + int32_t ctx_prio; + + ctx_prio = (ctx->override_priority == AMDGPU_CTX_PRIORITY_UNSET) ? + ctx->init_priority : ctx->override_priority; + + switch (hw_ip) { + case AMDGPU_HW_IP_GFX: + case AMDGPU_HW_IP_COMPUTE: + hw_prio = amdgpu_ctx_prio_to_gfx_pipe_prio(ctx_prio); + break; + case AMDGPU_HW_IP_VCE: + case AMDGPU_HW_IP_VCN_ENC: + hw_prio = amdgpu_ctx_sched_prio_to_ring_prio(ctx_prio); + break; + default: + hw_prio = AMDGPU_RING_PRIO_DEFAULT; + break; + } + + hw_ip = array_index_nospec(hw_ip, AMDGPU_HW_IP_NUM); + if (adev->gpu_sched[hw_ip][hw_prio].num_scheds == 0) + hw_prio = AMDGPU_RING_PRIO_DEFAULT; + + return hw_prio; +} + +/* Calculate the time spend on the hw */ +static ktime_t amdgpu_ctx_fence_time(struct dma_fence *fence) +{ + struct drm_sched_fence *s_fence; + + if (!fence) + return ns_to_ktime(0); + + /* When the fence is not even scheduled it can't have spend time */ + s_fence = to_drm_sched_fence(fence); + if (!test_bit(DMA_FENCE_FLAG_TIMESTAMP_BIT, &s_fence->scheduled.flags)) + return ns_to_ktime(0); + + /* When it is still running account how much already spend */ + if (!test_bit(DMA_FENCE_FLAG_TIMESTAMP_BIT, &s_fence->finished.flags)) + return ktime_sub(ktime_get(), s_fence->scheduled.timestamp); + + return ktime_sub(s_fence->finished.timestamp, + s_fence->scheduled.timestamp); +} + +static ktime_t amdgpu_ctx_entity_time(struct amdgpu_ctx *ctx, + struct amdgpu_ctx_entity *centity) +{ + ktime_t res = ns_to_ktime(0); + uint32_t i; + + spin_lock(&ctx->ring_lock); + for (i = 0; i < amdgpu_sched_jobs; i++) { + res = ktime_add(res, amdgpu_ctx_fence_time(centity->fences[i])); + } + spin_unlock(&ctx->ring_lock); + return res; +} + +static int amdgpu_ctx_init_entity(struct amdgpu_ctx *ctx, u32 hw_ip, + const u32 ring) +{ + struct drm_gpu_scheduler **scheds = NULL, *sched = NULL; + struct amdgpu_device *adev = ctx->mgr->adev; + struct amdgpu_ctx_entity *entity; + enum drm_sched_priority drm_prio; + unsigned int hw_prio, num_scheds; + int32_t ctx_prio; + int r; + + entity = kzalloc(struct_size(entity, fences, amdgpu_sched_jobs), + GFP_KERNEL); + if (!entity) + return -ENOMEM; + + ctx_prio = (ctx->override_priority == AMDGPU_CTX_PRIORITY_UNSET) ? + ctx->init_priority : ctx->override_priority; + entity->hw_ip = hw_ip; + entity->sequence = 1; + hw_prio = amdgpu_ctx_get_hw_prio(ctx, hw_ip); + drm_prio = amdgpu_ctx_to_drm_sched_prio(ctx_prio); + + hw_ip = array_index_nospec(hw_ip, AMDGPU_HW_IP_NUM); + scheds = adev->gpu_sched[hw_ip][hw_prio].sched; + num_scheds = adev->gpu_sched[hw_ip][hw_prio].num_scheds; + + /* disable load balance if the hw engine retains context among dependent jobs */ + if (hw_ip == AMDGPU_HW_IP_VCN_ENC || + hw_ip == AMDGPU_HW_IP_VCN_DEC || + hw_ip == AMDGPU_HW_IP_UVD_ENC || + hw_ip == AMDGPU_HW_IP_UVD) { + sched = drm_sched_pick_best(scheds, num_scheds); + scheds = &sched; + num_scheds = 1; + } + + r = drm_sched_entity_init(&entity->entity, drm_prio, scheds, num_scheds, + &ctx->guilty); + if (r) + goto error_free_entity; + + /* It's not an error if we fail to install the new entity */ + if (cmpxchg(&ctx->entities[hw_ip][ring], NULL, entity)) + goto cleanup_entity; + + return 0; + +cleanup_entity: + drm_sched_entity_fini(&entity->entity); + +error_free_entity: + kfree(entity); + + return r; +} + +static ktime_t amdgpu_ctx_fini_entity(struct amdgpu_ctx_entity *entity) +{ + ktime_t res = ns_to_ktime(0); + int i; + + if (!entity) + return res; + + for (i = 0; i < amdgpu_sched_jobs; ++i) { + res = ktime_add(res, amdgpu_ctx_fence_time(entity->fences[i])); + dma_fence_put(entity->fences[i]); + } + + kfree(entity); + return res; +} + +static int amdgpu_ctx_get_stable_pstate(struct amdgpu_ctx *ctx, + u32 *stable_pstate) +{ + struct amdgpu_device *adev = ctx->mgr->adev; + enum amd_dpm_forced_level current_level; + + current_level = amdgpu_dpm_get_performance_level(adev); + + switch (current_level) { + case AMD_DPM_FORCED_LEVEL_PROFILE_STANDARD: + *stable_pstate = AMDGPU_CTX_STABLE_PSTATE_STANDARD; + break; + case AMD_DPM_FORCED_LEVEL_PROFILE_MIN_SCLK: + *stable_pstate = AMDGPU_CTX_STABLE_PSTATE_MIN_SCLK; + break; + case AMD_DPM_FORCED_LEVEL_PROFILE_MIN_MCLK: + *stable_pstate = AMDGPU_CTX_STABLE_PSTATE_MIN_MCLK; + break; + case AMD_DPM_FORCED_LEVEL_PROFILE_PEAK: + *stable_pstate = AMDGPU_CTX_STABLE_PSTATE_PEAK; + break; + default: + *stable_pstate = AMDGPU_CTX_STABLE_PSTATE_NONE; + break; + } + return 0; +} + +static int amdgpu_ctx_init(struct amdgpu_ctx_mgr *mgr, int32_t priority, + struct drm_file *filp, struct amdgpu_ctx *ctx) +{ + u32 current_stable_pstate; + int r; + + r = amdgpu_ctx_priority_permit(filp, priority); + if (r) + return r; + + memset(ctx, 0, sizeof(*ctx)); + + kref_init(&ctx->refcount); + ctx->mgr = mgr; + spin_lock_init(&ctx->ring_lock); + + ctx->reset_counter = atomic_read(&mgr->adev->gpu_reset_counter); + ctx->reset_counter_query = ctx->reset_counter; + ctx->vram_lost_counter = atomic_read(&mgr->adev->vram_lost_counter); + ctx->init_priority = priority; + ctx->override_priority = AMDGPU_CTX_PRIORITY_UNSET; + + r = amdgpu_ctx_get_stable_pstate(ctx, ¤t_stable_pstate); + if (r) + return r; + + if (mgr->adev->pm.stable_pstate_ctx) + ctx->stable_pstate = mgr->adev->pm.stable_pstate_ctx->stable_pstate; + else + ctx->stable_pstate = current_stable_pstate; + + return 0; +} + +static int amdgpu_ctx_set_stable_pstate(struct amdgpu_ctx *ctx, + u32 stable_pstate) +{ + struct amdgpu_device *adev = ctx->mgr->adev; + enum amd_dpm_forced_level level; + u32 current_stable_pstate; + int r; + + mutex_lock(&adev->pm.stable_pstate_ctx_lock); + if (adev->pm.stable_pstate_ctx && adev->pm.stable_pstate_ctx != ctx) { + r = -EBUSY; + goto done; + } + + r = amdgpu_ctx_get_stable_pstate(ctx, ¤t_stable_pstate); + if (r || (stable_pstate == current_stable_pstate)) + goto done; + + switch (stable_pstate) { + case AMDGPU_CTX_STABLE_PSTATE_NONE: + level = AMD_DPM_FORCED_LEVEL_AUTO; + break; + case AMDGPU_CTX_STABLE_PSTATE_STANDARD: + level = AMD_DPM_FORCED_LEVEL_PROFILE_STANDARD; + break; + case AMDGPU_CTX_STABLE_PSTATE_MIN_SCLK: + level = AMD_DPM_FORCED_LEVEL_PROFILE_MIN_SCLK; + break; + case AMDGPU_CTX_STABLE_PSTATE_MIN_MCLK: + level = AMD_DPM_FORCED_LEVEL_PROFILE_MIN_MCLK; + break; + case AMDGPU_CTX_STABLE_PSTATE_PEAK: + level = AMD_DPM_FORCED_LEVEL_PROFILE_PEAK; + break; + default: + r = -EINVAL; + goto done; + } + + r = amdgpu_dpm_force_performance_level(adev, level); + + if (level == AMD_DPM_FORCED_LEVEL_AUTO) + adev->pm.stable_pstate_ctx = NULL; + else + adev->pm.stable_pstate_ctx = ctx; +done: + mutex_unlock(&adev->pm.stable_pstate_ctx_lock); + + return r; +} + +static void amdgpu_ctx_fini(struct kref *ref) +{ + struct amdgpu_ctx *ctx = container_of(ref, struct amdgpu_ctx, refcount); + struct amdgpu_ctx_mgr *mgr = ctx->mgr; + struct amdgpu_device *adev = mgr->adev; + unsigned i, j, idx; + + if (!adev) + return; + + for (i = 0; i < AMDGPU_HW_IP_NUM; ++i) { + for (j = 0; j < AMDGPU_MAX_ENTITY_NUM; ++j) { + ktime_t spend; + + spend = amdgpu_ctx_fini_entity(ctx->entities[i][j]); + atomic64_add(ktime_to_ns(spend), &mgr->time_spend[i]); + } + } + + if (drm_dev_enter(adev_to_drm(adev), &idx)) { + amdgpu_ctx_set_stable_pstate(ctx, ctx->stable_pstate); + drm_dev_exit(idx); + } + + kfree(ctx); +} + +int amdgpu_ctx_get_entity(struct amdgpu_ctx *ctx, u32 hw_ip, u32 instance, + u32 ring, struct drm_sched_entity **entity) +{ + int r; + + if (hw_ip >= AMDGPU_HW_IP_NUM) { + DRM_ERROR("unknown HW IP type: %d\n", hw_ip); + return -EINVAL; + } + + /* Right now all IPs have only one instance - multiple rings. */ + if (instance != 0) { + DRM_DEBUG("invalid ip instance: %d\n", instance); + return -EINVAL; + } + + if (ring >= amdgpu_ctx_num_entities[hw_ip]) { + DRM_DEBUG("invalid ring: %d %d\n", hw_ip, ring); + return -EINVAL; + } + + if (ctx->entities[hw_ip][ring] == NULL) { + r = amdgpu_ctx_init_entity(ctx, hw_ip, ring); + if (r) + return r; + } + + *entity = &ctx->entities[hw_ip][ring]->entity; + return 0; +} + +static int amdgpu_ctx_alloc(struct amdgpu_device *adev, + struct amdgpu_fpriv *fpriv, + struct drm_file *filp, + int32_t priority, + uint32_t *id) +{ + struct amdgpu_ctx_mgr *mgr = &fpriv->ctx_mgr; + struct amdgpu_ctx *ctx; + int r; + + ctx = kmalloc(sizeof(*ctx), GFP_KERNEL); + if (!ctx) + return -ENOMEM; + + mutex_lock(&mgr->lock); + r = idr_alloc(&mgr->ctx_handles, ctx, 1, AMDGPU_VM_MAX_NUM_CTX, GFP_KERNEL); + if (r < 0) { + mutex_unlock(&mgr->lock); + kfree(ctx); + return r; + } + + *id = (uint32_t)r; + r = amdgpu_ctx_init(mgr, priority, filp, ctx); + if (r) { + idr_remove(&mgr->ctx_handles, *id); + *id = 0; + kfree(ctx); + } + mutex_unlock(&mgr->lock); + return r; +} + +static void amdgpu_ctx_do_release(struct kref *ref) +{ + struct amdgpu_ctx *ctx; + u32 i, j; + + ctx = container_of(ref, struct amdgpu_ctx, refcount); + for (i = 0; i < AMDGPU_HW_IP_NUM; ++i) { + for (j = 0; j < amdgpu_ctx_num_entities[i]; ++j) { + if (!ctx->entities[i][j]) + continue; + + drm_sched_entity_destroy(&ctx->entities[i][j]->entity); + } + } + + amdgpu_ctx_fini(ref); +} + +static int amdgpu_ctx_free(struct amdgpu_fpriv *fpriv, uint32_t id) +{ + struct amdgpu_ctx_mgr *mgr = &fpriv->ctx_mgr; + struct amdgpu_ctx *ctx; + + mutex_lock(&mgr->lock); + ctx = idr_remove(&mgr->ctx_handles, id); + if (ctx) + kref_put(&ctx->refcount, amdgpu_ctx_do_release); + mutex_unlock(&mgr->lock); + return ctx ? 0 : -EINVAL; +} + +static int amdgpu_ctx_query(struct amdgpu_device *adev, + struct amdgpu_fpriv *fpriv, uint32_t id, + union drm_amdgpu_ctx_out *out) +{ + struct amdgpu_ctx *ctx; + struct amdgpu_ctx_mgr *mgr; + unsigned reset_counter; + + if (!fpriv) + return -EINVAL; + + mgr = &fpriv->ctx_mgr; + mutex_lock(&mgr->lock); + ctx = idr_find(&mgr->ctx_handles, id); + if (!ctx) { + mutex_unlock(&mgr->lock); + return -EINVAL; + } + + /* TODO: these two are always zero */ + out->state.flags = 0x0; + out->state.hangs = 0x0; + + /* determine if a GPU reset has occured since the last call */ + reset_counter = atomic_read(&adev->gpu_reset_counter); + /* TODO: this should ideally return NO, GUILTY, or INNOCENT. */ + if (ctx->reset_counter_query == reset_counter) + out->state.reset_status = AMDGPU_CTX_NO_RESET; + else + out->state.reset_status = AMDGPU_CTX_UNKNOWN_RESET; + ctx->reset_counter_query = reset_counter; + + mutex_unlock(&mgr->lock); + return 0; +} + +#define AMDGPU_RAS_COUNTE_DELAY_MS 3000 + +static int amdgpu_ctx_query2(struct amdgpu_device *adev, + struct amdgpu_fpriv *fpriv, uint32_t id, + union drm_amdgpu_ctx_out *out) +{ + struct amdgpu_ras *con = amdgpu_ras_get_context(adev); + struct amdgpu_ctx *ctx; + struct amdgpu_ctx_mgr *mgr; + + if (!fpriv) + return -EINVAL; + + mgr = &fpriv->ctx_mgr; + mutex_lock(&mgr->lock); + ctx = idr_find(&mgr->ctx_handles, id); + if (!ctx) { + mutex_unlock(&mgr->lock); + return -EINVAL; + } + + out->state.flags = 0x0; + out->state.hangs = 0x0; + + if (ctx->reset_counter != atomic_read(&adev->gpu_reset_counter)) + out->state.flags |= AMDGPU_CTX_QUERY2_FLAGS_RESET; + + if (ctx->vram_lost_counter != atomic_read(&adev->vram_lost_counter)) + out->state.flags |= AMDGPU_CTX_QUERY2_FLAGS_VRAMLOST; + + if (atomic_read(&ctx->guilty)) + out->state.flags |= AMDGPU_CTX_QUERY2_FLAGS_GUILTY; + + if (adev->ras_enabled && con) { + /* Return the cached values in O(1), + * and schedule delayed work to cache + * new vaues. + */ + int ce_count, ue_count; + + ce_count = atomic_read(&con->ras_ce_count); + ue_count = atomic_read(&con->ras_ue_count); + + if (ce_count != ctx->ras_counter_ce) { + ctx->ras_counter_ce = ce_count; + out->state.flags |= AMDGPU_CTX_QUERY2_FLAGS_RAS_CE; + } + + if (ue_count != ctx->ras_counter_ue) { + ctx->ras_counter_ue = ue_count; + out->state.flags |= AMDGPU_CTX_QUERY2_FLAGS_RAS_UE; + } + + schedule_delayed_work(&con->ras_counte_delay_work, + msecs_to_jiffies(AMDGPU_RAS_COUNTE_DELAY_MS)); + } + + mutex_unlock(&mgr->lock); + return 0; +} + + + +static int amdgpu_ctx_stable_pstate(struct amdgpu_device *adev, + struct amdgpu_fpriv *fpriv, uint32_t id, + bool set, u32 *stable_pstate) +{ + struct amdgpu_ctx *ctx; + struct amdgpu_ctx_mgr *mgr; + int r; + + if (!fpriv) + return -EINVAL; + + mgr = &fpriv->ctx_mgr; + mutex_lock(&mgr->lock); + ctx = idr_find(&mgr->ctx_handles, id); + if (!ctx) { + mutex_unlock(&mgr->lock); + return -EINVAL; + } + + if (set) + r = amdgpu_ctx_set_stable_pstate(ctx, *stable_pstate); + else + r = amdgpu_ctx_get_stable_pstate(ctx, stable_pstate); + + mutex_unlock(&mgr->lock); + return r; +} + +int amdgpu_ctx_ioctl(struct drm_device *dev, void *data, + struct drm_file *filp) +{ + int r; + uint32_t id, stable_pstate; + int32_t priority; + + union drm_amdgpu_ctx *args = data; + struct amdgpu_device *adev = drm_to_adev(dev); + struct amdgpu_fpriv *fpriv = filp->driver_priv; + + id = args->in.ctx_id; + priority = args->in.priority; + + /* For backwards compatibility reasons, we need to accept + * ioctls with garbage in the priority field */ + if (!amdgpu_ctx_priority_is_valid(priority)) + priority = AMDGPU_CTX_PRIORITY_NORMAL; + + switch (args->in.op) { + case AMDGPU_CTX_OP_ALLOC_CTX: + r = amdgpu_ctx_alloc(adev, fpriv, filp, priority, &id); + args->out.alloc.ctx_id = id; + break; + case AMDGPU_CTX_OP_FREE_CTX: + r = amdgpu_ctx_free(fpriv, id); + break; + case AMDGPU_CTX_OP_QUERY_STATE: + r = amdgpu_ctx_query(adev, fpriv, id, &args->out); + break; + case AMDGPU_CTX_OP_QUERY_STATE2: + r = amdgpu_ctx_query2(adev, fpriv, id, &args->out); + break; + case AMDGPU_CTX_OP_GET_STABLE_PSTATE: + if (args->in.flags) + return -EINVAL; + r = amdgpu_ctx_stable_pstate(adev, fpriv, id, false, &stable_pstate); + if (!r) + args->out.pstate.flags = stable_pstate; + break; + case AMDGPU_CTX_OP_SET_STABLE_PSTATE: + if (args->in.flags & ~AMDGPU_CTX_STABLE_PSTATE_FLAGS_MASK) + return -EINVAL; + stable_pstate = args->in.flags & AMDGPU_CTX_STABLE_PSTATE_FLAGS_MASK; + if (stable_pstate > AMDGPU_CTX_STABLE_PSTATE_PEAK) + return -EINVAL; + r = amdgpu_ctx_stable_pstate(adev, fpriv, id, true, &stable_pstate); + break; + default: + return -EINVAL; + } + + return r; +} + +struct amdgpu_ctx *amdgpu_ctx_get(struct amdgpu_fpriv *fpriv, uint32_t id) +{ + struct amdgpu_ctx *ctx; + struct amdgpu_ctx_mgr *mgr; + + if (!fpriv) + return NULL; + + mgr = &fpriv->ctx_mgr; + + mutex_lock(&mgr->lock); + ctx = idr_find(&mgr->ctx_handles, id); + if (ctx) + kref_get(&ctx->refcount); + mutex_unlock(&mgr->lock); + return ctx; +} + +int amdgpu_ctx_put(struct amdgpu_ctx *ctx) +{ + if (ctx == NULL) + return -EINVAL; + + kref_put(&ctx->refcount, amdgpu_ctx_do_release); + return 0; +} + +uint64_t amdgpu_ctx_add_fence(struct amdgpu_ctx *ctx, + struct drm_sched_entity *entity, + struct dma_fence *fence) +{ + struct amdgpu_ctx_entity *centity = to_amdgpu_ctx_entity(entity); + uint64_t seq = centity->sequence; + struct dma_fence *other = NULL; + unsigned idx = 0; + + idx = seq & (amdgpu_sched_jobs - 1); + other = centity->fences[idx]; + WARN_ON(other && !dma_fence_is_signaled(other)); + + dma_fence_get(fence); + + spin_lock(&ctx->ring_lock); + centity->fences[idx] = fence; + centity->sequence++; + spin_unlock(&ctx->ring_lock); + + atomic64_add(ktime_to_ns(amdgpu_ctx_fence_time(other)), + &ctx->mgr->time_spend[centity->hw_ip]); + + dma_fence_put(other); + return seq; +} + +struct dma_fence *amdgpu_ctx_get_fence(struct amdgpu_ctx *ctx, + struct drm_sched_entity *entity, + uint64_t seq) +{ + struct amdgpu_ctx_entity *centity = to_amdgpu_ctx_entity(entity); + struct dma_fence *fence; + + spin_lock(&ctx->ring_lock); + + if (seq == ~0ull) + seq = centity->sequence - 1; + + if (seq >= centity->sequence) { + spin_unlock(&ctx->ring_lock); + return ERR_PTR(-EINVAL); + } + + + if (seq + amdgpu_sched_jobs < centity->sequence) { + spin_unlock(&ctx->ring_lock); + return NULL; + } + + fence = dma_fence_get(centity->fences[seq & (amdgpu_sched_jobs - 1)]); + spin_unlock(&ctx->ring_lock); + + return fence; +} + +static void amdgpu_ctx_set_entity_priority(struct amdgpu_ctx *ctx, + struct amdgpu_ctx_entity *aentity, + int hw_ip, + int32_t priority) +{ + struct amdgpu_device *adev = ctx->mgr->adev; + unsigned int hw_prio; + struct drm_gpu_scheduler **scheds = NULL; + unsigned num_scheds; + + /* set sw priority */ + drm_sched_entity_set_priority(&aentity->entity, + amdgpu_ctx_to_drm_sched_prio(priority)); + + /* set hw priority */ + if (hw_ip == AMDGPU_HW_IP_COMPUTE || hw_ip == AMDGPU_HW_IP_GFX) { + hw_prio = amdgpu_ctx_get_hw_prio(ctx, hw_ip); + hw_prio = array_index_nospec(hw_prio, AMDGPU_RING_PRIO_MAX); + scheds = adev->gpu_sched[hw_ip][hw_prio].sched; + num_scheds = adev->gpu_sched[hw_ip][hw_prio].num_scheds; + drm_sched_entity_modify_sched(&aentity->entity, scheds, + num_scheds); + } +} + +void amdgpu_ctx_priority_override(struct amdgpu_ctx *ctx, + int32_t priority) +{ + int32_t ctx_prio; + unsigned i, j; + + ctx->override_priority = priority; + + ctx_prio = (ctx->override_priority == AMDGPU_CTX_PRIORITY_UNSET) ? + ctx->init_priority : ctx->override_priority; + for (i = 0; i < AMDGPU_HW_IP_NUM; ++i) { + for (j = 0; j < amdgpu_ctx_num_entities[i]; ++j) { + if (!ctx->entities[i][j]) + continue; + + amdgpu_ctx_set_entity_priority(ctx, ctx->entities[i][j], + i, ctx_prio); + } + } +} + +int amdgpu_ctx_wait_prev_fence(struct amdgpu_ctx *ctx, + struct drm_sched_entity *entity) +{ + struct amdgpu_ctx_entity *centity = to_amdgpu_ctx_entity(entity); + struct dma_fence *other; + unsigned idx; + long r; + + spin_lock(&ctx->ring_lock); + idx = centity->sequence & (amdgpu_sched_jobs - 1); + other = dma_fence_get(centity->fences[idx]); + spin_unlock(&ctx->ring_lock); + + if (!other) + return 0; + + r = dma_fence_wait(other, true); + if (r < 0 && r != -ERESTARTSYS) + DRM_ERROR("Error (%ld) waiting for fence!\n", r); + + dma_fence_put(other); + return r; +} + +void amdgpu_ctx_mgr_init(struct amdgpu_ctx_mgr *mgr, + struct amdgpu_device *adev) +{ + unsigned int i; + + mgr->adev = adev; + mutex_init(&mgr->lock); + idr_init_base(&mgr->ctx_handles, 1); + + for (i = 0; i < AMDGPU_HW_IP_NUM; ++i) + atomic64_set(&mgr->time_spend[i], 0); +} + +long amdgpu_ctx_mgr_entity_flush(struct amdgpu_ctx_mgr *mgr, long timeout) +{ + struct amdgpu_ctx *ctx; + struct idr *idp; + uint32_t id, i, j; + + idp = &mgr->ctx_handles; + + mutex_lock(&mgr->lock); + idr_for_each_entry(idp, ctx, id) { + for (i = 0; i < AMDGPU_HW_IP_NUM; ++i) { + for (j = 0; j < amdgpu_ctx_num_entities[i]; ++j) { + struct drm_sched_entity *entity; + + if (!ctx->entities[i][j]) + continue; + + entity = &ctx->entities[i][j]->entity; + timeout = drm_sched_entity_flush(entity, timeout); + } + } + } + mutex_unlock(&mgr->lock); + return timeout; +} + +void amdgpu_ctx_mgr_entity_fini(struct amdgpu_ctx_mgr *mgr) +{ + struct amdgpu_ctx *ctx; + struct idr *idp; + uint32_t id, i, j; + + idp = &mgr->ctx_handles; + + idr_for_each_entry(idp, ctx, id) { + if (kref_read(&ctx->refcount) != 1) { + DRM_ERROR("ctx %p is still alive\n", ctx); + continue; + } + + for (i = 0; i < AMDGPU_HW_IP_NUM; ++i) { + for (j = 0; j < amdgpu_ctx_num_entities[i]; ++j) { + struct drm_sched_entity *entity; + + if (!ctx->entities[i][j]) + continue; + + entity = &ctx->entities[i][j]->entity; + drm_sched_entity_fini(entity); + } + } + } +} + +void amdgpu_ctx_mgr_fini(struct amdgpu_ctx_mgr *mgr) +{ + struct amdgpu_ctx *ctx; + struct idr *idp; + uint32_t id; + + amdgpu_ctx_mgr_entity_fini(mgr); + + idp = &mgr->ctx_handles; + + idr_for_each_entry(idp, ctx, id) { + if (kref_put(&ctx->refcount, amdgpu_ctx_fini) != 1) + DRM_ERROR("ctx %p is still alive\n", ctx); + } + + idr_destroy(&mgr->ctx_handles); + mutex_destroy(&mgr->lock); +} + +void amdgpu_ctx_mgr_usage(struct amdgpu_ctx_mgr *mgr, + ktime_t usage[AMDGPU_HW_IP_NUM]) +{ + struct amdgpu_ctx *ctx; + unsigned int hw_ip, i; + uint32_t id; + + /* + * This is a little bit racy because it can be that a ctx or a fence are + * destroyed just in the moment we try to account them. But that is ok + * since exactly that case is explicitely allowed by the interface. + */ + mutex_lock(&mgr->lock); + for (hw_ip = 0; hw_ip < AMDGPU_HW_IP_NUM; ++hw_ip) { + uint64_t ns = atomic64_read(&mgr->time_spend[hw_ip]); + + usage[hw_ip] = ns_to_ktime(ns); + } + + idr_for_each_entry(&mgr->ctx_handles, ctx, id) { + for (hw_ip = 0; hw_ip < AMDGPU_HW_IP_NUM; ++hw_ip) { + for (i = 0; i < amdgpu_ctx_num_entities[hw_ip]; ++i) { + struct amdgpu_ctx_entity *centity; + ktime_t spend; + + centity = ctx->entities[hw_ip][i]; + if (!centity) + continue; + spend = amdgpu_ctx_entity_time(ctx, centity); + usage[hw_ip] = ktime_add(usage[hw_ip], spend); + } + } + } + mutex_unlock(&mgr->lock); +} -- cgit v1.2.3