Merge tag 'net-next-6.3' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-nextgrafted

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().
  ...

1 files changed, 868 insertions, 0 deletions

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/pm/base.c b/drivers/gpu/drm/nouveau/nvkm/engine/pm/base.c
new file mode 100644
index 000000000..8fe0444f7
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/pm/base.c
@@ -0,0 +1,868 @@
+/*
+ * Copyright 2013 Red Hat 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: Ben Skeggs
+ */
+#include "priv.h"
+
+#include <core/client.h>
+#include <core/option.h>
+
+#include <nvif/class.h>
+#include <nvif/if0002.h>
+#include <nvif/if0003.h>
+#include <nvif/ioctl.h>
+#include <nvif/unpack.h>
+
+static u8
+nvkm_pm_count_perfdom(struct nvkm_pm *pm)
+{
+	struct nvkm_perfdom *dom;
+	u8 domain_nr = 0;
+
+	list_for_each_entry(dom, &pm->domains, head)
+		domain_nr++;
+	return domain_nr;
+}
+
+static u16
+nvkm_perfdom_count_perfsig(struct nvkm_perfdom *dom)
+{
+	u16 signal_nr = 0;
+	int i;
+
+	if (dom) {
+		for (i = 0; i < dom->signal_nr; i++) {
+			if (dom->signal[i].name)
+				signal_nr++;
+		}
+	}
+	return signal_nr;
+}
+
+static struct nvkm_perfdom *
+nvkm_perfdom_find(struct nvkm_pm *pm, int di)
+{
+	struct nvkm_perfdom *dom;
+	int tmp = 0;
+
+	list_for_each_entry(dom, &pm->domains, head) {
+		if (tmp++ == di)
+			return dom;
+	}
+	return NULL;
+}
+
+static struct nvkm_perfsig *
+nvkm_perfsig_find(struct nvkm_pm *pm, u8 di, u8 si, struct nvkm_perfdom **pdom)
+{
+	struct nvkm_perfdom *dom = *pdom;
+
+	if (dom == NULL) {
+		dom = nvkm_perfdom_find(pm, di);
+		if (dom == NULL)
+			return NULL;
+		*pdom = dom;
+	}
+
+	if (!dom->signal[si].name)
+		return NULL;
+	return &dom->signal[si];
+}
+
+static u8
+nvkm_perfsig_count_perfsrc(struct nvkm_perfsig *sig)
+{
+	u8 source_nr = 0, i;
+
+	for (i = 0; i < ARRAY_SIZE(sig->source); i++) {
+		if (sig->source[i])
+			source_nr++;
+	}
+	return source_nr;
+}
+
+static struct nvkm_perfsrc *
+nvkm_perfsrc_find(struct nvkm_pm *pm, struct nvkm_perfsig *sig, int si)
+{
+	struct nvkm_perfsrc *src;
+	bool found = false;
+	int tmp = 1; /* Sources ID start from 1 */
+	u8 i;
+
+	for (i = 0; i < ARRAY_SIZE(sig->source) && sig->source[i]; i++) {
+		if (sig->source[i] == si) {
+			found = true;
+			break;
+		}
+	}
+
+	if (found) {
+		list_for_each_entry(src, &pm->sources, head) {
+			if (tmp++ == si)
+				return src;
+		}
+	}
+
+	return NULL;
+}
+
+static int
+nvkm_perfsrc_enable(struct nvkm_pm *pm, struct nvkm_perfctr *ctr)
+{
+	struct nvkm_subdev *subdev = &pm->engine.subdev;
+	struct nvkm_device *device = subdev->device;
+	struct nvkm_perfdom *dom = NULL;
+	struct nvkm_perfsig *sig;
+	struct nvkm_perfsrc *src;
+	u32 mask, value;
+	int i, j;
+
+	for (i = 0; i < 4; i++) {
+		for (j = 0; j < 8 && ctr->source[i][j]; j++) {
+			sig = nvkm_perfsig_find(pm, ctr->domain,
+						ctr->signal[i], &dom);
+			if (!sig)
+				return -EINVAL;
+
+			src = nvkm_perfsrc_find(pm, sig, ctr->source[i][j]);
+			if (!src)
+				return -EINVAL;
+
+			/* set enable bit if needed */
+			mask = value = 0x00000000;
+			if (src->enable)
+				mask = value = 0x80000000;
+			mask  |= (src->mask << src->shift);
+			value |= ((ctr->source[i][j] >> 32) << src->shift);
+
+			/* enable the source */
+			nvkm_mask(device, src->addr, mask, value);
+			nvkm_debug(subdev,
+				   "enabled source %08x %08x %08x\n",
+				   src->addr, mask, value);
+		}
+	}
+	return 0;
+}
+
+static int
+nvkm_perfsrc_disable(struct nvkm_pm *pm, struct nvkm_perfctr *ctr)
+{
+	struct nvkm_subdev *subdev = &pm->engine.subdev;
+	struct nvkm_device *device = subdev->device;
+	struct nvkm_perfdom *dom = NULL;
+	struct nvkm_perfsig *sig;
+	struct nvkm_perfsrc *src;
+	u32 mask;
+	int i, j;
+
+	for (i = 0; i < 4; i++) {
+		for (j = 0; j < 8 && ctr->source[i][j]; j++) {
+			sig = nvkm_perfsig_find(pm, ctr->domain,
+						ctr->signal[i], &dom);
+			if (!sig)
+				return -EINVAL;
+
+			src = nvkm_perfsrc_find(pm, sig, ctr->source[i][j]);
+			if (!src)
+				return -EINVAL;
+
+			/* unset enable bit if needed */
+			mask = 0x00000000;
+			if (src->enable)
+				mask = 0x80000000;
+			mask |= (src->mask << src->shift);
+
+			/* disable the source */
+			nvkm_mask(device, src->addr, mask, 0);
+			nvkm_debug(subdev, "disabled source %08x %08x\n",
+				   src->addr, mask);
+		}
+	}
+	return 0;
+}
+
+/*******************************************************************************
+ * Perfdom object classes
+ ******************************************************************************/
+static int
+nvkm_perfdom_init(struct nvkm_perfdom *dom, void *data, u32 size)
+{
+	union {
+		struct nvif_perfdom_init none;
+	} *args = data;
+	struct nvkm_object *object = &dom->object;
+	struct nvkm_pm *pm = dom->perfmon->pm;
+	int ret = -ENOSYS, i;
+
+	nvif_ioctl(object, "perfdom init size %d\n", size);
+	if (!(ret = nvif_unvers(ret, &data, &size, args->none))) {
+		nvif_ioctl(object, "perfdom init\n");
+	} else
+		return ret;
+
+	for (i = 0; i < 4; i++) {
+		if (dom->ctr[i]) {
+			dom->func->init(pm, dom, dom->ctr[i]);
+
+			/* enable sources */
+			nvkm_perfsrc_enable(pm, dom->ctr[i]);
+		}
+	}
+
+	/* start next batch of counters for sampling */
+	dom->func->next(pm, dom);
+	return 0;
+}
+
+static int
+nvkm_perfdom_sample(struct nvkm_perfdom *dom, void *data, u32 size)
+{
+	union {
+		struct nvif_perfdom_sample none;
+	} *args = data;
+	struct nvkm_object *object = &dom->object;
+	struct nvkm_pm *pm = dom->perfmon->pm;
+	int ret = -ENOSYS;
+
+	nvif_ioctl(object, "perfdom sample size %d\n", size);
+	if (!(ret = nvif_unvers(ret, &data, &size, args->none))) {
+		nvif_ioctl(object, "perfdom sample\n");
+	} else
+		return ret;
+	pm->sequence++;
+
+	/* sample previous batch of counters */
+	list_for_each_entry(dom, &pm->domains, head)
+		dom->func->next(pm, dom);
+
+	return 0;
+}
+
+static int
+nvkm_perfdom_read(struct nvkm_perfdom *dom, void *data, u32 size)
+{
+	union {
+		struct nvif_perfdom_read_v0 v0;
+	} *args = data;
+	struct nvkm_object *object = &dom->object;
+	struct nvkm_pm *pm = dom->perfmon->pm;
+	int ret = -ENOSYS, i;
+
+	nvif_ioctl(object, "perfdom read size %d\n", size);
+	if (!(ret = nvif_unpack(ret, &data, &size, args->v0, 0, 0, false))) {
+		nvif_ioctl(object, "perfdom read vers %d\n", args->v0.version);
+	} else
+		return ret;
+
+	for (i = 0; i < 4; i++) {
+		if (dom->ctr[i])
+			dom->func->read(pm, dom, dom->ctr[i]);
+	}
+
+	if (!dom->clk)
+		return -EAGAIN;
+
+	for (i = 0; i < 4; i++)
+		if (dom->ctr[i])
+			args->v0.ctr[i] = dom->ctr[i]->ctr;
+	args->v0.clk = dom->clk;
+	return 0;
+}
+
+static int
+nvkm_perfdom_mthd(struct nvkm_object *object, u32 mthd, void *data, u32 size)
+{
+	struct nvkm_perfdom *dom = nvkm_perfdom(object);
+	switch (mthd) {
+	case NVIF_PERFDOM_V0_INIT:
+		return nvkm_perfdom_init(dom, data, size);
+	case NVIF_PERFDOM_V0_SAMPLE:
+		return nvkm_perfdom_sample(dom, data, size);
+	case NVIF_PERFDOM_V0_READ:
+		return nvkm_perfdom_read(dom, data, size);
+	default:
+		break;
+	}
+	return -EINVAL;
+}
+
+static void *
+nvkm_perfdom_dtor(struct nvkm_object *object)
+{
+	struct nvkm_perfdom *dom = nvkm_perfdom(object);
+	struct nvkm_pm *pm = dom->perfmon->pm;
+	int i;
+
+	for (i = 0; i < 4; i++) {
+		struct nvkm_perfctr *ctr = dom->ctr[i];
+		if (ctr) {
+			nvkm_perfsrc_disable(pm, ctr);
+			if (ctr->head.next)
+				list_del(&ctr->head);
+		}
+		kfree(ctr);
+	}
+
+	return dom;
+}
+
+static int
+nvkm_perfctr_new(struct nvkm_perfdom *dom, int slot, u8 domain,
+		 struct nvkm_perfsig *signal[4], u64 source[4][8],
+		 u16 logic_op, struct nvkm_perfctr **pctr)
+{
+	struct nvkm_perfctr *ctr;
+	int i, j;
+
+	if (!dom)
+		return -EINVAL;
+
+	ctr = *pctr = kzalloc(sizeof(*ctr), GFP_KERNEL);
+	if (!ctr)
+		return -ENOMEM;
+
+	ctr->domain   = domain;
+	ctr->logic_op = logic_op;
+	ctr->slot     = slot;
+	for (i = 0; i < 4; i++) {
+		if (signal[i]) {
+			ctr->signal[i] = signal[i] - dom->signal;
+			for (j = 0; j < 8; j++)
+				ctr->source[i][j] = source[i][j];
+		}
+	}
+	list_add_tail(&ctr->head, &dom->list);
+
+	return 0;
+}
+
+static const struct nvkm_object_func
+nvkm_perfdom = {
+	.dtor = nvkm_perfdom_dtor,
+	.mthd = nvkm_perfdom_mthd,
+};
+
+static int
+nvkm_perfdom_new_(struct nvkm_perfmon *perfmon,
+		  const struct nvkm_oclass *oclass, void *data, u32 size,
+		  struct nvkm_object **pobject)
+{
+	union {
+		struct nvif_perfdom_v0 v0;
+	} *args = data;
+	struct nvkm_pm *pm = perfmon->pm;
+	struct nvkm_object *parent = oclass->parent;
+	struct nvkm_perfdom *sdom = NULL;
+	struct nvkm_perfctr *ctr[4] = {};
+	struct nvkm_perfdom *dom;
+	int c, s, m;
+	int ret = -ENOSYS;
+
+	nvif_ioctl(parent, "create perfdom size %d\n", size);
+	if (!(ret = nvif_unpack(ret, &data, &size, args->v0, 0, 0, false))) {
+		nvif_ioctl(parent, "create perfdom vers %d dom %d mode %02x\n",
+			   args->v0.version, args->v0.domain, args->v0.mode);
+	} else
+		return ret;
+
+	for (c = 0; c < ARRAY_SIZE(args->v0.ctr); c++) {
+		struct nvkm_perfsig *sig[4] = {};
+		u64 src[4][8] = {};
+
+		for (s = 0; s < ARRAY_SIZE(args->v0.ctr[c].signal); s++) {
+			sig[s] = nvkm_perfsig_find(pm, args->v0.domain,
+						   args->v0.ctr[c].signal[s],
+						   &sdom);
+			if (args->v0.ctr[c].signal[s] && !sig[s])
+				return -EINVAL;
+
+			for (m = 0; m < 8; m++) {
+				src[s][m] = args->v0.ctr[c].source[s][m];
+				if (src[s][m] && !nvkm_perfsrc_find(pm, sig[s],
+							            src[s][m]))
+					return -EINVAL;
+			}
+		}
+
+		ret = nvkm_perfctr_new(sdom, c, args->v0.domain, sig, src,
+				       args->v0.ctr[c].logic_op, &ctr[c]);
+		if (ret)
+			return ret;
+	}
+
+	if (!sdom)
+		return -EINVAL;
+
+	if (!(dom = kzalloc(sizeof(*dom), GFP_KERNEL)))
+		return -ENOMEM;
+	nvkm_object_ctor(&nvkm_perfdom, oclass, &dom->object);
+	dom->perfmon = perfmon;
+	*pobject = &dom->object;
+
+	dom->func = sdom->func;
+	dom->addr = sdom->addr;
+	dom->mode = args->v0.mode;
+	for (c = 0; c < ARRAY_SIZE(ctr); c++)
+		dom->ctr[c] = ctr[c];
+	return 0;
+}
+
+/*******************************************************************************
+ * Perfmon object classes
+ ******************************************************************************/
+static int
+nvkm_perfmon_mthd_query_domain(struct nvkm_perfmon *perfmon,
+			       void *data, u32 size)
+{
+	union {
+		struct nvif_perfmon_query_domain_v0 v0;
+	} *args = data;
+	struct nvkm_object *object = &perfmon->object;
+	struct nvkm_pm *pm = perfmon->pm;
+	struct nvkm_perfdom *dom;
+	u8 domain_nr;
+	int di, ret = -ENOSYS;
+
+	nvif_ioctl(object, "perfmon query domain size %d\n", size);
+	if (!(ret = nvif_unpack(ret, &data, &size, args->v0, 0, 0, false))) {
+		nvif_ioctl(object, "perfmon domain vers %d iter %02x\n",
+			   args->v0.version, args->v0.iter);
+		di = (args->v0.iter & 0xff) - 1;
+	} else
+		return ret;
+
+	domain_nr = nvkm_pm_count_perfdom(pm);
+	if (di >= (int)domain_nr)
+		return -EINVAL;
+
+	if (di >= 0) {
+		dom = nvkm_perfdom_find(pm, di);
+		if (dom == NULL)
+			return -EINVAL;
+
+		args->v0.id         = di;
+		args->v0.signal_nr  = nvkm_perfdom_count_perfsig(dom);
+		strncpy(args->v0.name, dom->name, sizeof(args->v0.name) - 1);
+
+		/* Currently only global counters (PCOUNTER) are implemented
+		 * but this will be different for local counters (MP). */
+		args->v0.counter_nr = 4;
+	}
+
+	if (++di < domain_nr) {
+		args->v0.iter = ++di;
+		return 0;
+	}
+
+	args->v0.iter = 0xff;
+	return 0;
+}
+
+static int
+nvkm_perfmon_mthd_query_signal(struct nvkm_perfmon *perfmon,
+			       void *data, u32 size)
+{
+	union {
+		struct nvif_perfmon_query_signal_v0 v0;
+	} *args = data;
+	struct nvkm_object *object = &perfmon->object;
+	struct nvkm_pm *pm = perfmon->pm;
+	struct nvkm_device *device = pm->engine.subdev.device;
+	struct nvkm_perfdom *dom;
+	struct nvkm_perfsig *sig;
+	const bool all = nvkm_boolopt(device->cfgopt, "NvPmShowAll", false);
+	const bool raw = nvkm_boolopt(device->cfgopt, "NvPmUnnamed", all);
+	int ret = -ENOSYS, si;
+
+	nvif_ioctl(object, "perfmon query signal size %d\n", size);
+	if (!(ret = nvif_unpack(ret, &data, &size, args->v0, 0, 0, false))) {
+		nvif_ioctl(object,
+			   "perfmon query signal vers %d dom %d iter %04x\n",
+			   args->v0.version, args->v0.domain, args->v0.iter);
+		si = (args->v0.iter & 0xffff) - 1;
+	} else
+		return ret;
+
+	dom = nvkm_perfdom_find(pm, args->v0.domain);
+	if (dom == NULL || si >= (int)dom->signal_nr)
+		return -EINVAL;
+
+	if (si >= 0) {
+		sig = &dom->signal[si];
+		if (raw || !sig->name) {
+			snprintf(args->v0.name, sizeof(args->v0.name),
+				 "/%s/%02x", dom->name, si);
+		} else {
+			strncpy(args->v0.name, sig->name,
+				sizeof(args->v0.name) - 1);
+		}
+
+		args->v0.signal = si;
+		args->v0.source_nr = nvkm_perfsig_count_perfsrc(sig);
+	}
+
+	while (++si < dom->signal_nr) {
+		if (all || dom->signal[si].name) {
+			args->v0.iter = ++si;
+			return 0;
+		}
+	}
+
+	args->v0.iter = 0xffff;
+	return 0;
+}
+
+static int
+nvkm_perfmon_mthd_query_source(struct nvkm_perfmon *perfmon,
+			       void *data, u32 size)
+{
+	union {
+		struct nvif_perfmon_query_source_v0 v0;
+	} *args = data;
+	struct nvkm_object *object = &perfmon->object;
+	struct nvkm_pm *pm = perfmon->pm;
+	struct nvkm_perfdom *dom = NULL;
+	struct nvkm_perfsig *sig;
+	struct nvkm_perfsrc *src;
+	u8 source_nr = 0;
+	int si, ret = -ENOSYS;
+
+	nvif_ioctl(object, "perfmon query source size %d\n", size);
+	if (!(ret = nvif_unpack(ret, &data, &size, args->v0, 0, 0, false))) {
+		nvif_ioctl(object,
+			   "perfmon source vers %d dom %d sig %02x iter %02x\n",
+			   args->v0.version, args->v0.domain, args->v0.signal,
+			   args->v0.iter);
+		si = (args->v0.iter & 0xff) - 1;
+	} else
+		return ret;
+
+	sig = nvkm_perfsig_find(pm, args->v0.domain, args->v0.signal, &dom);
+	if (!sig)
+		return -EINVAL;
+
+	source_nr = nvkm_perfsig_count_perfsrc(sig);
+	if (si >= (int)source_nr)
+		return -EINVAL;
+
+	if (si >= 0) {
+		src = nvkm_perfsrc_find(pm, sig, sig->source[si]);
+		if (!src)
+			return -EINVAL;
+
+		args->v0.source = sig->source[si];
+		args->v0.mask   = src->mask;
+		strncpy(args->v0.name, src->name, sizeof(args->v0.name) - 1);
+	}
+
+	if (++si < source_nr) {
+		args->v0.iter = ++si;
+		return 0;
+	}
+
+	args->v0.iter = 0xff;
+	return 0;
+}
+
+static int
+nvkm_perfmon_mthd(struct nvkm_object *object, u32 mthd, void *data, u32 size)
+{
+	struct nvkm_perfmon *perfmon = nvkm_perfmon(object);
+	switch (mthd) {
+	case NVIF_PERFMON_V0_QUERY_DOMAIN:
+		return nvkm_perfmon_mthd_query_domain(perfmon, data, size);
+	case NVIF_PERFMON_V0_QUERY_SIGNAL:
+		return nvkm_perfmon_mthd_query_signal(perfmon, data, size);
+	case NVIF_PERFMON_V0_QUERY_SOURCE:
+		return nvkm_perfmon_mthd_query_source(perfmon, data, size);
+	default:
+		break;
+	}
+	return -EINVAL;
+}
+
+static int
+nvkm_perfmon_child_new(const struct nvkm_oclass *oclass, void *data, u32 size,
+		       struct nvkm_object **pobject)
+{
+	struct nvkm_perfmon *perfmon = nvkm_perfmon(oclass->parent);
+	return nvkm_perfdom_new_(perfmon, oclass, data, size, pobject);
+}
+
+static int
+nvkm_perfmon_child_get(struct nvkm_object *object, int index,
+		       struct nvkm_oclass *oclass)
+{
+	if (index == 0) {
+		oclass->base.oclass = NVIF_CLASS_PERFDOM;
+		oclass->base.minver = 0;
+		oclass->base.maxver = 0;
+		oclass->ctor = nvkm_perfmon_child_new;
+		return 0;
+	}
+	return -EINVAL;
+}
+
+static void *
+nvkm_perfmon_dtor(struct nvkm_object *object)
+{
+	struct nvkm_perfmon *perfmon = nvkm_perfmon(object);
+	struct nvkm_pm *pm = perfmon->pm;
+	spin_lock(&pm->client.lock);
+	if (pm->client.object == &perfmon->object)
+		pm->client.object = NULL;
+	spin_unlock(&pm->client.lock);
+	return perfmon;
+}
+
+static const struct nvkm_object_func
+nvkm_perfmon = {
+	.dtor = nvkm_perfmon_dtor,
+	.mthd = nvkm_perfmon_mthd,
+	.sclass = nvkm_perfmon_child_get,
+};
+
+static int
+nvkm_perfmon_new(struct nvkm_pm *pm, const struct nvkm_oclass *oclass,
+		 void *data, u32 size, struct nvkm_object **pobject)
+{
+	struct nvkm_perfmon *perfmon;
+
+	if (!(perfmon = kzalloc(sizeof(*perfmon), GFP_KERNEL)))
+		return -ENOMEM;
+	nvkm_object_ctor(&nvkm_perfmon, oclass, &perfmon->object);
+	perfmon->pm = pm;
+	*pobject = &perfmon->object;
+	return 0;
+}
+
+/*******************************************************************************
+ * PPM engine/subdev functions
+ ******************************************************************************/
+
+static int
+nvkm_pm_oclass_new(struct nvkm_device *device, const struct nvkm_oclass *oclass,
+		   void *data, u32 size, struct nvkm_object **pobject)
+{
+	struct nvkm_pm *pm = nvkm_pm(oclass->engine);
+	int ret;
+
+	ret = nvkm_perfmon_new(pm, oclass, data, size, pobject);
+	if (ret)
+		return ret;
+
+	spin_lock(&pm->client.lock);
+	if (pm->client.object == NULL)
+		pm->client.object = *pobject;
+	ret = (pm->client.object == *pobject) ? 0 : -EBUSY;
+	spin_unlock(&pm->client.lock);
+	return ret;
+}
+
+static const struct nvkm_device_oclass
+nvkm_pm_oclass = {
+	.base.oclass = NVIF_CLASS_PERFMON,
+	.base.minver = -1,
+	.base.maxver = -1,
+	.ctor = nvkm_pm_oclass_new,
+};
+
+static int
+nvkm_pm_oclass_get(struct nvkm_oclass *oclass, int index,
+		   const struct nvkm_device_oclass **class)
+{
+	if (index == 0) {
+		oclass->base = nvkm_pm_oclass.base;
+		*class = &nvkm_pm_oclass;
+		return index;
+	}
+	return 1;
+}
+
+static int
+nvkm_perfsrc_new(struct nvkm_pm *pm, struct nvkm_perfsig *sig,
+		 const struct nvkm_specsrc *spec)
+{
+	const struct nvkm_specsrc *ssrc;
+	const struct nvkm_specmux *smux;
+	struct nvkm_perfsrc *src;
+	u8 source_nr = 0;
+
+	if (!spec) {
+		/* No sources are defined for this signal. */
+		return 0;
+	}
+
+	ssrc = spec;
+	while (ssrc->name) {
+		smux = ssrc->mux;
+		while (smux->name) {
+			bool found = false;
+			u8 source_id = 0;
+			u32 len;
+
+			list_for_each_entry(src, &pm->sources, head) {
+				if (src->addr == ssrc->addr &&
+				    src->shift == smux->shift) {
+					found = true;
+					break;
+				}
+				source_id++;
+			}
+
+			if (!found) {
+				src = kzalloc(sizeof(*src), GFP_KERNEL);
+				if (!src)
+					return -ENOMEM;
+
+				src->addr   = ssrc->addr;
+				src->mask   = smux->mask;
+				src->shift  = smux->shift;
+				src->enable = smux->enable;
+
+				len = strlen(ssrc->name) +
+				      strlen(smux->name) + 2;
+				src->name = kzalloc(len, GFP_KERNEL);
+				if (!src->name) {
+					kfree(src);
+					return -ENOMEM;
+				}
+				snprintf(src->name, len, "%s_%s", ssrc->name,
+					 smux->name);
+
+				list_add_tail(&src->head, &pm->sources);
+			}
+
+			sig->source[source_nr++] = source_id + 1;
+			smux++;
+		}
+		ssrc++;
+	}
+
+	return 0;
+}
+
+int
+nvkm_perfdom_new(struct nvkm_pm *pm, const char *name, u32 mask,
+		 u32 base, u32 size_unit, u32 size_domain,
+		 const struct nvkm_specdom *spec)
+{
+	const struct nvkm_specdom *sdom;
+	const struct nvkm_specsig *ssig;
+	struct nvkm_perfdom *dom;
+	int ret, i;
+
+	for (i = 0; i == 0 || mask; i++) {
+		u32 addr = base + (i * size_unit);
+		if (i && !(mask & (1 << i)))
+			continue;
+
+		sdom = spec;
+		while (sdom->signal_nr) {
+			dom = kzalloc(struct_size(dom, signal, sdom->signal_nr),
+				      GFP_KERNEL);
+			if (!dom)
+				return -ENOMEM;
+
+			if (mask) {
+				snprintf(dom->name, sizeof(dom->name),
+					 "%s/%02x/%02x", name, i,
+					 (int)(sdom - spec));
+			} else {
+				snprintf(dom->name, sizeof(dom->name),
+					 "%s/%02x", name, (int)(sdom - spec));
+			}
+
+			list_add_tail(&dom->head, &pm->domains);
+			INIT_LIST_HEAD(&dom->list);
+			dom->func = sdom->func;
+			dom->addr = addr;
+			dom->signal_nr = sdom->signal_nr;
+
+			ssig = (sdom++)->signal;
+			while (ssig->name) {
+				struct nvkm_perfsig *sig =
+					&dom->signal[ssig->signal];
+				sig->name = ssig->name;
+				ret = nvkm_perfsrc_new(pm, sig, ssig->source);
+				if (ret)
+					return ret;
+				ssig++;
+			}
+
+			addr += size_domain;
+		}
+
+		mask &= ~(1 << i);
+	}
+
+	return 0;
+}
+
+static int
+nvkm_pm_fini(struct nvkm_engine *engine, bool suspend)
+{
+	struct nvkm_pm *pm = nvkm_pm(engine);
+	if (pm->func->fini)
+		pm->func->fini(pm);
+	return 0;
+}
+
+static void *
+nvkm_pm_dtor(struct nvkm_engine *engine)
+{
+	struct nvkm_pm *pm = nvkm_pm(engine);
+	struct nvkm_perfdom *dom, *next_dom;
+	struct nvkm_perfsrc *src, *next_src;
+
+	list_for_each_entry_safe(dom, next_dom, &pm->domains, head) {
+		list_del(&dom->head);
+		kfree(dom);
+	}
+
+	list_for_each_entry_safe(src, next_src, &pm->sources, head) {
+		list_del(&src->head);
+		kfree(src->name);
+		kfree(src);
+	}
+
+	return pm;
+}
+
+static const struct nvkm_engine_func
+nvkm_pm = {
+	.dtor = nvkm_pm_dtor,
+	.fini = nvkm_pm_fini,
+	.base.sclass = nvkm_pm_oclass_get,
+};
+
+int
+nvkm_pm_ctor(const struct nvkm_pm_func *func, struct nvkm_device *device,
+	     enum nvkm_subdev_type type, int inst, struct nvkm_pm *pm)
+{
+	pm->func = func;
+	INIT_LIST_HEAD(&pm->domains);
+	INIT_LIST_HEAD(&pm->sources);
+	spin_lock_init(&pm->client.lock);
+	return nvkm_engine_ctor(&nvkm_pm, device, type, inst, true, &pm->engine);
+}