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, 1426 insertions, 0 deletions

diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/nv04.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/nv04.c
new file mode 100644
index 000000000..81bd682c2
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/nv04.c
@@ -0,0 +1,1426 @@
+/*
+ * Copyright 2007 Stephane Marchesin
+ * All Rights Reserved.
+ *
+ * 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
+ * paragr) 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
+ * PRECISION INSIGHT AND/OR ITS SUPPLIERS 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.
+ */
+#include "priv.h"
+#include "regs.h"
+
+#include <core/client.h>
+#include <core/gpuobj.h>
+#include <engine/fifo.h>
+#include <engine/fifo/chan.h>
+#include <subdev/instmem.h>
+#include <subdev/timer.h>
+
+static u32
+nv04_gr_ctx_regs[] = {
+	0x0040053c,
+	0x00400544,
+	0x00400540,
+	0x00400548,
+	NV04_PGRAPH_CTX_SWITCH1,
+	NV04_PGRAPH_CTX_SWITCH2,
+	NV04_PGRAPH_CTX_SWITCH3,
+	NV04_PGRAPH_CTX_SWITCH4,
+	NV04_PGRAPH_CTX_CACHE1,
+	NV04_PGRAPH_CTX_CACHE2,
+	NV04_PGRAPH_CTX_CACHE3,
+	NV04_PGRAPH_CTX_CACHE4,
+	0x00400184,
+	0x004001a4,
+	0x004001c4,
+	0x004001e4,
+	0x00400188,
+	0x004001a8,
+	0x004001c8,
+	0x004001e8,
+	0x0040018c,
+	0x004001ac,
+	0x004001cc,
+	0x004001ec,
+	0x00400190,
+	0x004001b0,
+	0x004001d0,
+	0x004001f0,
+	0x00400194,
+	0x004001b4,
+	0x004001d4,
+	0x004001f4,
+	0x00400198,
+	0x004001b8,
+	0x004001d8,
+	0x004001f8,
+	0x0040019c,
+	0x004001bc,
+	0x004001dc,
+	0x004001fc,
+	0x00400174,
+	NV04_PGRAPH_DMA_START_0,
+	NV04_PGRAPH_DMA_START_1,
+	NV04_PGRAPH_DMA_LENGTH,
+	NV04_PGRAPH_DMA_MISC,
+	NV04_PGRAPH_DMA_PITCH,
+	NV04_PGRAPH_BOFFSET0,
+	NV04_PGRAPH_BBASE0,
+	NV04_PGRAPH_BLIMIT0,
+	NV04_PGRAPH_BOFFSET1,
+	NV04_PGRAPH_BBASE1,
+	NV04_PGRAPH_BLIMIT1,
+	NV04_PGRAPH_BOFFSET2,
+	NV04_PGRAPH_BBASE2,
+	NV04_PGRAPH_BLIMIT2,
+	NV04_PGRAPH_BOFFSET3,
+	NV04_PGRAPH_BBASE3,
+	NV04_PGRAPH_BLIMIT3,
+	NV04_PGRAPH_BOFFSET4,
+	NV04_PGRAPH_BBASE4,
+	NV04_PGRAPH_BLIMIT4,
+	NV04_PGRAPH_BOFFSET5,
+	NV04_PGRAPH_BBASE5,
+	NV04_PGRAPH_BLIMIT5,
+	NV04_PGRAPH_BPITCH0,
+	NV04_PGRAPH_BPITCH1,
+	NV04_PGRAPH_BPITCH2,
+	NV04_PGRAPH_BPITCH3,
+	NV04_PGRAPH_BPITCH4,
+	NV04_PGRAPH_SURFACE,
+	NV04_PGRAPH_STATE,
+	NV04_PGRAPH_BSWIZZLE2,
+	NV04_PGRAPH_BSWIZZLE5,
+	NV04_PGRAPH_BPIXEL,
+	NV04_PGRAPH_NOTIFY,
+	NV04_PGRAPH_PATT_COLOR0,
+	NV04_PGRAPH_PATT_COLOR1,
+	NV04_PGRAPH_PATT_COLORRAM+0x00,
+	NV04_PGRAPH_PATT_COLORRAM+0x04,
+	NV04_PGRAPH_PATT_COLORRAM+0x08,
+	NV04_PGRAPH_PATT_COLORRAM+0x0c,
+	NV04_PGRAPH_PATT_COLORRAM+0x10,
+	NV04_PGRAPH_PATT_COLORRAM+0x14,
+	NV04_PGRAPH_PATT_COLORRAM+0x18,
+	NV04_PGRAPH_PATT_COLORRAM+0x1c,
+	NV04_PGRAPH_PATT_COLORRAM+0x20,
+	NV04_PGRAPH_PATT_COLORRAM+0x24,
+	NV04_PGRAPH_PATT_COLORRAM+0x28,
+	NV04_PGRAPH_PATT_COLORRAM+0x2c,
+	NV04_PGRAPH_PATT_COLORRAM+0x30,
+	NV04_PGRAPH_PATT_COLORRAM+0x34,
+	NV04_PGRAPH_PATT_COLORRAM+0x38,
+	NV04_PGRAPH_PATT_COLORRAM+0x3c,
+	NV04_PGRAPH_PATT_COLORRAM+0x40,
+	NV04_PGRAPH_PATT_COLORRAM+0x44,
+	NV04_PGRAPH_PATT_COLORRAM+0x48,
+	NV04_PGRAPH_PATT_COLORRAM+0x4c,
+	NV04_PGRAPH_PATT_COLORRAM+0x50,
+	NV04_PGRAPH_PATT_COLORRAM+0x54,
+	NV04_PGRAPH_PATT_COLORRAM+0x58,
+	NV04_PGRAPH_PATT_COLORRAM+0x5c,
+	NV04_PGRAPH_PATT_COLORRAM+0x60,
+	NV04_PGRAPH_PATT_COLORRAM+0x64,
+	NV04_PGRAPH_PATT_COLORRAM+0x68,
+	NV04_PGRAPH_PATT_COLORRAM+0x6c,
+	NV04_PGRAPH_PATT_COLORRAM+0x70,
+	NV04_PGRAPH_PATT_COLORRAM+0x74,
+	NV04_PGRAPH_PATT_COLORRAM+0x78,
+	NV04_PGRAPH_PATT_COLORRAM+0x7c,
+	NV04_PGRAPH_PATT_COLORRAM+0x80,
+	NV04_PGRAPH_PATT_COLORRAM+0x84,
+	NV04_PGRAPH_PATT_COLORRAM+0x88,
+	NV04_PGRAPH_PATT_COLORRAM+0x8c,
+	NV04_PGRAPH_PATT_COLORRAM+0x90,
+	NV04_PGRAPH_PATT_COLORRAM+0x94,
+	NV04_PGRAPH_PATT_COLORRAM+0x98,
+	NV04_PGRAPH_PATT_COLORRAM+0x9c,
+	NV04_PGRAPH_PATT_COLORRAM+0xa0,
+	NV04_PGRAPH_PATT_COLORRAM+0xa4,
+	NV04_PGRAPH_PATT_COLORRAM+0xa8,
+	NV04_PGRAPH_PATT_COLORRAM+0xac,
+	NV04_PGRAPH_PATT_COLORRAM+0xb0,
+	NV04_PGRAPH_PATT_COLORRAM+0xb4,
+	NV04_PGRAPH_PATT_COLORRAM+0xb8,
+	NV04_PGRAPH_PATT_COLORRAM+0xbc,
+	NV04_PGRAPH_PATT_COLORRAM+0xc0,
+	NV04_PGRAPH_PATT_COLORRAM+0xc4,
+	NV04_PGRAPH_PATT_COLORRAM+0xc8,
+	NV04_PGRAPH_PATT_COLORRAM+0xcc,
+	NV04_PGRAPH_PATT_COLORRAM+0xd0,
+	NV04_PGRAPH_PATT_COLORRAM+0xd4,
+	NV04_PGRAPH_PATT_COLORRAM+0xd8,
+	NV04_PGRAPH_PATT_COLORRAM+0xdc,
+	NV04_PGRAPH_PATT_COLORRAM+0xe0,
+	NV04_PGRAPH_PATT_COLORRAM+0xe4,
+	NV04_PGRAPH_PATT_COLORRAM+0xe8,
+	NV04_PGRAPH_PATT_COLORRAM+0xec,
+	NV04_PGRAPH_PATT_COLORRAM+0xf0,
+	NV04_PGRAPH_PATT_COLORRAM+0xf4,
+	NV04_PGRAPH_PATT_COLORRAM+0xf8,
+	NV04_PGRAPH_PATT_COLORRAM+0xfc,
+	NV04_PGRAPH_PATTERN,
+	0x0040080c,
+	NV04_PGRAPH_PATTERN_SHAPE,
+	0x00400600,
+	NV04_PGRAPH_ROP3,
+	NV04_PGRAPH_CHROMA,
+	NV04_PGRAPH_BETA_AND,
+	NV04_PGRAPH_BETA_PREMULT,
+	NV04_PGRAPH_CONTROL0,
+	NV04_PGRAPH_CONTROL1,
+	NV04_PGRAPH_CONTROL2,
+	NV04_PGRAPH_BLEND,
+	NV04_PGRAPH_STORED_FMT,
+	NV04_PGRAPH_SOURCE_COLOR,
+	0x00400560,
+	0x00400568,
+	0x00400564,
+	0x0040056c,
+	0x00400400,
+	0x00400480,
+	0x00400404,
+	0x00400484,
+	0x00400408,
+	0x00400488,
+	0x0040040c,
+	0x0040048c,
+	0x00400410,
+	0x00400490,
+	0x00400414,
+	0x00400494,
+	0x00400418,
+	0x00400498,
+	0x0040041c,
+	0x0040049c,
+	0x00400420,
+	0x004004a0,
+	0x00400424,
+	0x004004a4,
+	0x00400428,
+	0x004004a8,
+	0x0040042c,
+	0x004004ac,
+	0x00400430,
+	0x004004b0,
+	0x00400434,
+	0x004004b4,
+	0x00400438,
+	0x004004b8,
+	0x0040043c,
+	0x004004bc,
+	0x00400440,
+	0x004004c0,
+	0x00400444,
+	0x004004c4,
+	0x00400448,
+	0x004004c8,
+	0x0040044c,
+	0x004004cc,
+	0x00400450,
+	0x004004d0,
+	0x00400454,
+	0x004004d4,
+	0x00400458,
+	0x004004d8,
+	0x0040045c,
+	0x004004dc,
+	0x00400460,
+	0x004004e0,
+	0x00400464,
+	0x004004e4,
+	0x00400468,
+	0x004004e8,
+	0x0040046c,
+	0x004004ec,
+	0x00400470,
+	0x004004f0,
+	0x00400474,
+	0x004004f4,
+	0x00400478,
+	0x004004f8,
+	0x0040047c,
+	0x004004fc,
+	0x00400534,
+	0x00400538,
+	0x00400514,
+	0x00400518,
+	0x0040051c,
+	0x00400520,
+	0x00400524,
+	0x00400528,
+	0x0040052c,
+	0x00400530,
+	0x00400d00,
+	0x00400d40,
+	0x00400d80,
+	0x00400d04,
+	0x00400d44,
+	0x00400d84,
+	0x00400d08,
+	0x00400d48,
+	0x00400d88,
+	0x00400d0c,
+	0x00400d4c,
+	0x00400d8c,
+	0x00400d10,
+	0x00400d50,
+	0x00400d90,
+	0x00400d14,
+	0x00400d54,
+	0x00400d94,
+	0x00400d18,
+	0x00400d58,
+	0x00400d98,
+	0x00400d1c,
+	0x00400d5c,
+	0x00400d9c,
+	0x00400d20,
+	0x00400d60,
+	0x00400da0,
+	0x00400d24,
+	0x00400d64,
+	0x00400da4,
+	0x00400d28,
+	0x00400d68,
+	0x00400da8,
+	0x00400d2c,
+	0x00400d6c,
+	0x00400dac,
+	0x00400d30,
+	0x00400d70,
+	0x00400db0,
+	0x00400d34,
+	0x00400d74,
+	0x00400db4,
+	0x00400d38,
+	0x00400d78,
+	0x00400db8,
+	0x00400d3c,
+	0x00400d7c,
+	0x00400dbc,
+	0x00400590,
+	0x00400594,
+	0x00400598,
+	0x0040059c,
+	0x004005a8,
+	0x004005ac,
+	0x004005b0,
+	0x004005b4,
+	0x004005c0,
+	0x004005c4,
+	0x004005c8,
+	0x004005cc,
+	0x004005d0,
+	0x004005d4,
+	0x004005d8,
+	0x004005dc,
+	0x004005e0,
+	NV04_PGRAPH_PASSTHRU_0,
+	NV04_PGRAPH_PASSTHRU_1,
+	NV04_PGRAPH_PASSTHRU_2,
+	NV04_PGRAPH_DVD_COLORFMT,
+	NV04_PGRAPH_SCALED_FORMAT,
+	NV04_PGRAPH_MISC24_0,
+	NV04_PGRAPH_MISC24_1,
+	NV04_PGRAPH_MISC24_2,
+	0x00400500,
+	0x00400504,
+	NV04_PGRAPH_VALID1,
+	NV04_PGRAPH_VALID2,
+	NV04_PGRAPH_DEBUG_3
+};
+
+#define nv04_gr(p) container_of((p), struct nv04_gr, base)
+
+struct nv04_gr {
+	struct nvkm_gr base;
+	struct nv04_gr_chan *chan[16];
+	spinlock_t lock;
+};
+
+#define nv04_gr_chan(p) container_of((p), struct nv04_gr_chan, object)
+
+struct nv04_gr_chan {
+	struct nvkm_object object;
+	struct nv04_gr *gr;
+	int chid;
+	u32 nv04[ARRAY_SIZE(nv04_gr_ctx_regs)];
+};
+
+/*******************************************************************************
+ * Graphics object classes
+ ******************************************************************************/
+
+/*
+ * Software methods, why they are needed, and how they all work:
+ *
+ * NV04 and NV05 keep most of the state in PGRAPH context itself, but some
+ * 2d engine settings are kept inside the grobjs themselves. The grobjs are
+ * 3 words long on both. grobj format on NV04 is:
+ *
+ * word 0:
+ *  - bits 0-7: class
+ *  - bit 12: color key active
+ *  - bit 13: clip rect active
+ *  - bit 14: if set, destination surface is swizzled and taken from buffer 5
+ *            [set by NV04_SWIZZLED_SURFACE], otherwise it's linear and taken
+ *            from buffer 0 [set by NV04_CONTEXT_SURFACES_2D or
+ *            NV03_CONTEXT_SURFACE_DST].
+ *  - bits 15-17: 2d operation [aka patch config]
+ *  - bit 24: patch valid [enables rendering using this object]
+ *  - bit 25: surf3d valid [for tex_tri and multitex_tri only]
+ * word 1:
+ *  - bits 0-1: mono format
+ *  - bits 8-13: color format
+ *  - bits 16-31: DMA_NOTIFY instance
+ * word 2:
+ *  - bits 0-15: DMA_A instance
+ *  - bits 16-31: DMA_B instance
+ *
+ * On NV05 it's:
+ *
+ * word 0:
+ *  - bits 0-7: class
+ *  - bit 12: color key active
+ *  - bit 13: clip rect active
+ *  - bit 14: if set, destination surface is swizzled and taken from buffer 5
+ *            [set by NV04_SWIZZLED_SURFACE], otherwise it's linear and taken
+ *            from buffer 0 [set by NV04_CONTEXT_SURFACES_2D or
+ *            NV03_CONTEXT_SURFACE_DST].
+ *  - bits 15-17: 2d operation [aka patch config]
+ *  - bits 20-22: dither mode
+ *  - bit 24: patch valid [enables rendering using this object]
+ *  - bit 25: surface_dst/surface_color/surf2d/surf3d valid
+ *  - bit 26: surface_src/surface_zeta valid
+ *  - bit 27: pattern valid
+ *  - bit 28: rop valid
+ *  - bit 29: beta1 valid
+ *  - bit 30: beta4 valid
+ * word 1:
+ *  - bits 0-1: mono format
+ *  - bits 8-13: color format
+ *  - bits 16-31: DMA_NOTIFY instance
+ * word 2:
+ *  - bits 0-15: DMA_A instance
+ *  - bits 16-31: DMA_B instance
+ *
+ * NV05 will set/unset the relevant valid bits when you poke the relevant
+ * object-binding methods with object of the proper type, or with the NULL
+ * type. It'll only allow rendering using the grobj if all needed objects
+ * are bound. The needed set of objects depends on selected operation: for
+ * example rop object is needed by ROP_AND, but not by SRCCOPY_AND.
+ *
+ * NV04 doesn't have these methods implemented at all, and doesn't have the
+ * relevant bits in grobj. Instead, it'll allow rendering whenever bit 24
+ * is set. So we have to emulate them in software, internally keeping the
+ * same bits as NV05 does. Since grobjs are aligned to 16 bytes on nv04,
+ * but the last word isn't actually used for anything, we abuse it for this
+ * purpose.
+ *
+ * Actually, NV05 can optionally check bit 24 too, but we disable this since
+ * there's no use for it.
+ *
+ * For unknown reasons, NV04 implements surf3d binding in hardware as an
+ * exception. Also for unknown reasons, NV04 doesn't implement the clipping
+ * methods on the surf3d object, so we have to emulate them too.
+ */
+
+static void
+nv04_gr_set_ctx1(struct nvkm_device *device, u32 inst, u32 mask, u32 value)
+{
+	int subc = (nvkm_rd32(device, NV04_PGRAPH_TRAPPED_ADDR) >> 13) & 0x7;
+	u32 tmp;
+
+	tmp  = nvkm_rd32(device, 0x700000 + inst);
+	tmp &= ~mask;
+	tmp |= value;
+	nvkm_wr32(device, 0x700000 + inst, tmp);
+
+	nvkm_wr32(device, NV04_PGRAPH_CTX_SWITCH1, tmp);
+	nvkm_wr32(device, NV04_PGRAPH_CTX_CACHE1 + (subc << 2), tmp);
+}
+
+static void
+nv04_gr_set_ctx_val(struct nvkm_device *device, u32 inst, u32 mask, u32 value)
+{
+	int class, op, valid = 1;
+	u32 tmp, ctx1;
+
+	ctx1 = nvkm_rd32(device, 0x700000 + inst);
+	class = ctx1 & 0xff;
+	op = (ctx1 >> 15) & 7;
+
+	tmp = nvkm_rd32(device, 0x70000c + inst);
+	tmp &= ~mask;
+	tmp |= value;
+	nvkm_wr32(device, 0x70000c + inst, tmp);
+
+	/* check for valid surf2d/surf_dst/surf_color */
+	if (!(tmp & 0x02000000))
+		valid = 0;
+	/* check for valid surf_src/surf_zeta */
+	if ((class == 0x1f || class == 0x48) && !(tmp & 0x04000000))
+		valid = 0;
+
+	switch (op) {
+	/* SRCCOPY_AND, SRCCOPY: no extra objects required */
+	case 0:
+	case 3:
+		break;
+	/* ROP_AND: requires pattern and rop */
+	case 1:
+		if (!(tmp & 0x18000000))
+			valid = 0;
+		break;
+	/* BLEND_AND: requires beta1 */
+	case 2:
+		if (!(tmp & 0x20000000))
+			valid = 0;
+		break;
+	/* SRCCOPY_PREMULT, BLEND_PREMULT: beta4 required */
+	case 4:
+	case 5:
+		if (!(tmp & 0x40000000))
+			valid = 0;
+		break;
+	}
+
+	nv04_gr_set_ctx1(device, inst, 0x01000000, valid << 24);
+}
+
+static bool
+nv04_gr_mthd_set_operation(struct nvkm_device *device, u32 inst, u32 data)
+{
+	u8 class = nvkm_rd32(device, 0x700000) & 0x000000ff;
+	if (data > 5)
+		return false;
+	/* Old versions of the objects only accept first three operations. */
+	if (data > 2 && class < 0x40)
+		return false;
+	nv04_gr_set_ctx1(device, inst, 0x00038000, data << 15);
+	/* changing operation changes set of objects needed for validation */
+	nv04_gr_set_ctx_val(device, inst, 0, 0);
+	return true;
+}
+
+static bool
+nv04_gr_mthd_surf3d_clip_h(struct nvkm_device *device, u32 inst, u32 data)
+{
+	u32 min = data & 0xffff, max;
+	u32 w = data >> 16;
+	if (min & 0x8000)
+		/* too large */
+		return false;
+	if (w & 0x8000)
+		/* yes, it accepts negative for some reason. */
+		w |= 0xffff0000;
+	max = min + w;
+	max &= 0x3ffff;
+	nvkm_wr32(device, 0x40053c, min);
+	nvkm_wr32(device, 0x400544, max);
+	return true;
+}
+
+static bool
+nv04_gr_mthd_surf3d_clip_v(struct nvkm_device *device, u32 inst, u32 data)
+{
+	u32 min = data & 0xffff, max;
+	u32 w = data >> 16;
+	if (min & 0x8000)
+		/* too large */
+		return false;
+	if (w & 0x8000)
+		/* yes, it accepts negative for some reason. */
+		w |= 0xffff0000;
+	max = min + w;
+	max &= 0x3ffff;
+	nvkm_wr32(device, 0x400540, min);
+	nvkm_wr32(device, 0x400548, max);
+	return true;
+}
+
+static u8
+nv04_gr_mthd_bind_class(struct nvkm_device *device, u32 inst)
+{
+	return nvkm_rd32(device, 0x700000 + (inst << 4));
+}
+
+static bool
+nv04_gr_mthd_bind_surf2d(struct nvkm_device *device, u32 inst, u32 data)
+{
+	switch (nv04_gr_mthd_bind_class(device, data)) {
+	case 0x30:
+		nv04_gr_set_ctx1(device, inst, 0x00004000, 0);
+		nv04_gr_set_ctx_val(device, inst, 0x02000000, 0);
+		return true;
+	case 0x42:
+		nv04_gr_set_ctx1(device, inst, 0x00004000, 0);
+		nv04_gr_set_ctx_val(device, inst, 0x02000000, 0x02000000);
+		return true;
+	}
+	return false;
+}
+
+static bool
+nv04_gr_mthd_bind_surf2d_swzsurf(struct nvkm_device *device, u32 inst, u32 data)
+{
+	switch (nv04_gr_mthd_bind_class(device, data)) {
+	case 0x30:
+		nv04_gr_set_ctx1(device, inst, 0x00004000, 0);
+		nv04_gr_set_ctx_val(device, inst, 0x02000000, 0);
+		return true;
+	case 0x42:
+		nv04_gr_set_ctx1(device, inst, 0x00004000, 0);
+		nv04_gr_set_ctx_val(device, inst, 0x02000000, 0x02000000);
+		return true;
+	case 0x52:
+		nv04_gr_set_ctx1(device, inst, 0x00004000, 0x00004000);
+		nv04_gr_set_ctx_val(device, inst, 0x02000000, 0x02000000);
+		return true;
+	}
+	return false;
+}
+
+static bool
+nv01_gr_mthd_bind_patt(struct nvkm_device *device, u32 inst, u32 data)
+{
+	switch (nv04_gr_mthd_bind_class(device, data)) {
+	case 0x30:
+		nv04_gr_set_ctx_val(device, inst, 0x08000000, 0);
+		return true;
+	case 0x18:
+		nv04_gr_set_ctx_val(device, inst, 0x08000000, 0x08000000);
+		return true;
+	}
+	return false;
+}
+
+static bool
+nv04_gr_mthd_bind_patt(struct nvkm_device *device, u32 inst, u32 data)
+{
+	switch (nv04_gr_mthd_bind_class(device, data)) {
+	case 0x30:
+		nv04_gr_set_ctx_val(device, inst, 0x08000000, 0);
+		return true;
+	case 0x44:
+		nv04_gr_set_ctx_val(device, inst, 0x08000000, 0x08000000);
+		return true;
+	}
+	return false;
+}
+
+static bool
+nv04_gr_mthd_bind_rop(struct nvkm_device *device, u32 inst, u32 data)
+{
+	switch (nv04_gr_mthd_bind_class(device, data)) {
+	case 0x30:
+		nv04_gr_set_ctx_val(device, inst, 0x10000000, 0);
+		return true;
+	case 0x43:
+		nv04_gr_set_ctx_val(device, inst, 0x10000000, 0x10000000);
+		return true;
+	}
+	return false;
+}
+
+static bool
+nv04_gr_mthd_bind_beta1(struct nvkm_device *device, u32 inst, u32 data)
+{
+	switch (nv04_gr_mthd_bind_class(device, data)) {
+	case 0x30:
+		nv04_gr_set_ctx_val(device, inst, 0x20000000, 0);
+		return true;
+	case 0x12:
+		nv04_gr_set_ctx_val(device, inst, 0x20000000, 0x20000000);
+		return true;
+	}
+	return false;
+}
+
+static bool
+nv04_gr_mthd_bind_beta4(struct nvkm_device *device, u32 inst, u32 data)
+{
+	switch (nv04_gr_mthd_bind_class(device, data)) {
+	case 0x30:
+		nv04_gr_set_ctx_val(device, inst, 0x40000000, 0);
+		return true;
+	case 0x72:
+		nv04_gr_set_ctx_val(device, inst, 0x40000000, 0x40000000);
+		return true;
+	}
+	return false;
+}
+
+static bool
+nv04_gr_mthd_bind_surf_dst(struct nvkm_device *device, u32 inst, u32 data)
+{
+	switch (nv04_gr_mthd_bind_class(device, data)) {
+	case 0x30:
+		nv04_gr_set_ctx_val(device, inst, 0x02000000, 0);
+		return true;
+	case 0x58:
+		nv04_gr_set_ctx_val(device, inst, 0x02000000, 0x02000000);
+		return true;
+	}
+	return false;
+}
+
+static bool
+nv04_gr_mthd_bind_surf_src(struct nvkm_device *device, u32 inst, u32 data)
+{
+	switch (nv04_gr_mthd_bind_class(device, data)) {
+	case 0x30:
+		nv04_gr_set_ctx_val(device, inst, 0x04000000, 0);
+		return true;
+	case 0x59:
+		nv04_gr_set_ctx_val(device, inst, 0x04000000, 0x04000000);
+		return true;
+	}
+	return false;
+}
+
+static bool
+nv04_gr_mthd_bind_surf_color(struct nvkm_device *device, u32 inst, u32 data)
+{
+	switch (nv04_gr_mthd_bind_class(device, data)) {
+	case 0x30:
+		nv04_gr_set_ctx_val(device, inst, 0x02000000, 0);
+		return true;
+	case 0x5a:
+		nv04_gr_set_ctx_val(device, inst, 0x02000000, 0x02000000);
+		return true;
+	}
+	return false;
+}
+
+static bool
+nv04_gr_mthd_bind_surf_zeta(struct nvkm_device *device, u32 inst, u32 data)
+{
+	switch (nv04_gr_mthd_bind_class(device, data)) {
+	case 0x30:
+		nv04_gr_set_ctx_val(device, inst, 0x04000000, 0);
+		return true;
+	case 0x5b:
+		nv04_gr_set_ctx_val(device, inst, 0x04000000, 0x04000000);
+		return true;
+	}
+	return false;
+}
+
+static bool
+nv01_gr_mthd_bind_clip(struct nvkm_device *device, u32 inst, u32 data)
+{
+	switch (nv04_gr_mthd_bind_class(device, data)) {
+	case 0x30:
+		nv04_gr_set_ctx1(device, inst, 0x2000, 0);
+		return true;
+	case 0x19:
+		nv04_gr_set_ctx1(device, inst, 0x2000, 0x2000);
+		return true;
+	}
+	return false;
+}
+
+static bool
+nv01_gr_mthd_bind_chroma(struct nvkm_device *device, u32 inst, u32 data)
+{
+	switch (nv04_gr_mthd_bind_class(device, data)) {
+	case 0x30:
+		nv04_gr_set_ctx1(device, inst, 0x1000, 0);
+		return true;
+	/* Yes, for some reason even the old versions of objects
+	 * accept 0x57 and not 0x17. Consistency be damned.
+	 */
+	case 0x57:
+		nv04_gr_set_ctx1(device, inst, 0x1000, 0x1000);
+		return true;
+	}
+	return false;
+}
+
+static bool
+nv03_gr_mthd_gdi(struct nvkm_device *device, u32 inst, u32 mthd, u32 data)
+{
+	bool (*func)(struct nvkm_device *, u32, u32);
+	switch (mthd) {
+	case 0x0184: func = nv01_gr_mthd_bind_patt; break;
+	case 0x0188: func = nv04_gr_mthd_bind_rop; break;
+	case 0x018c: func = nv04_gr_mthd_bind_beta1; break;
+	case 0x0190: func = nv04_gr_mthd_bind_surf_dst; break;
+	case 0x02fc: func = nv04_gr_mthd_set_operation; break;
+	default:
+		return false;
+	}
+	return func(device, inst, data);
+}
+
+static bool
+nv04_gr_mthd_gdi(struct nvkm_device *device, u32 inst, u32 mthd, u32 data)
+{
+	bool (*func)(struct nvkm_device *, u32, u32);
+	switch (mthd) {
+	case 0x0188: func = nv04_gr_mthd_bind_patt; break;
+	case 0x018c: func = nv04_gr_mthd_bind_rop; break;
+	case 0x0190: func = nv04_gr_mthd_bind_beta1; break;
+	case 0x0194: func = nv04_gr_mthd_bind_beta4; break;
+	case 0x0198: func = nv04_gr_mthd_bind_surf2d; break;
+	case 0x02fc: func = nv04_gr_mthd_set_operation; break;
+	default:
+		return false;
+	}
+	return func(device, inst, data);
+}
+
+static bool
+nv01_gr_mthd_blit(struct nvkm_device *device, u32 inst, u32 mthd, u32 data)
+{
+	bool (*func)(struct nvkm_device *, u32, u32);
+	switch (mthd) {
+	case 0x0184: func = nv01_gr_mthd_bind_chroma; break;
+	case 0x0188: func = nv01_gr_mthd_bind_clip; break;
+	case 0x018c: func = nv01_gr_mthd_bind_patt; break;
+	case 0x0190: func = nv04_gr_mthd_bind_rop; break;
+	case 0x0194: func = nv04_gr_mthd_bind_beta1; break;
+	case 0x0198: func = nv04_gr_mthd_bind_surf_dst; break;
+	case 0x019c: func = nv04_gr_mthd_bind_surf_src; break;
+	case 0x02fc: func = nv04_gr_mthd_set_operation; break;
+	default:
+		return false;
+	}
+	return func(device, inst, data);
+}
+
+static bool
+nv04_gr_mthd_blit(struct nvkm_device *device, u32 inst, u32 mthd, u32 data)
+{
+	bool (*func)(struct nvkm_device *, u32, u32);
+	switch (mthd) {
+	case 0x0184: func = nv01_gr_mthd_bind_chroma; break;
+	case 0x0188: func = nv01_gr_mthd_bind_clip; break;
+	case 0x018c: func = nv04_gr_mthd_bind_patt; break;
+	case 0x0190: func = nv04_gr_mthd_bind_rop; break;
+	case 0x0194: func = nv04_gr_mthd_bind_beta1; break;
+	case 0x0198: func = nv04_gr_mthd_bind_beta4; break;
+	case 0x019c: func = nv04_gr_mthd_bind_surf2d; break;
+	case 0x02fc: func = nv04_gr_mthd_set_operation; break;
+	default:
+		return false;
+	}
+	return func(device, inst, data);
+}
+
+static bool
+nv04_gr_mthd_iifc(struct nvkm_device *device, u32 inst, u32 mthd, u32 data)
+{
+	bool (*func)(struct nvkm_device *, u32, u32);
+	switch (mthd) {
+	case 0x0188: func = nv01_gr_mthd_bind_chroma; break;
+	case 0x018c: func = nv01_gr_mthd_bind_clip; break;
+	case 0x0190: func = nv04_gr_mthd_bind_patt; break;
+	case 0x0194: func = nv04_gr_mthd_bind_rop; break;
+	case 0x0198: func = nv04_gr_mthd_bind_beta1; break;
+	case 0x019c: func = nv04_gr_mthd_bind_beta4; break;
+	case 0x01a0: func = nv04_gr_mthd_bind_surf2d_swzsurf; break;
+	case 0x03e4: func = nv04_gr_mthd_set_operation; break;
+	default:
+		return false;
+	}
+	return func(device, inst, data);
+}
+
+static bool
+nv01_gr_mthd_ifc(struct nvkm_device *device, u32 inst, u32 mthd, u32 data)
+{
+	bool (*func)(struct nvkm_device *, u32, u32);
+	switch (mthd) {
+	case 0x0184: func = nv01_gr_mthd_bind_chroma; break;
+	case 0x0188: func = nv01_gr_mthd_bind_clip; break;
+	case 0x018c: func = nv01_gr_mthd_bind_patt; break;
+	case 0x0190: func = nv04_gr_mthd_bind_rop; break;
+	case 0x0194: func = nv04_gr_mthd_bind_beta1; break;
+	case 0x0198: func = nv04_gr_mthd_bind_surf_dst; break;
+	case 0x02fc: func = nv04_gr_mthd_set_operation; break;
+	default:
+		return false;
+	}
+	return func(device, inst, data);
+}
+
+static bool
+nv04_gr_mthd_ifc(struct nvkm_device *device, u32 inst, u32 mthd, u32 data)
+{
+	bool (*func)(struct nvkm_device *, u32, u32);
+	switch (mthd) {
+	case 0x0184: func = nv01_gr_mthd_bind_chroma; break;
+	case 0x0188: func = nv01_gr_mthd_bind_clip; break;
+	case 0x018c: func = nv04_gr_mthd_bind_patt; break;
+	case 0x0190: func = nv04_gr_mthd_bind_rop; break;
+	case 0x0194: func = nv04_gr_mthd_bind_beta1; break;
+	case 0x0198: func = nv04_gr_mthd_bind_beta4; break;
+	case 0x019c: func = nv04_gr_mthd_bind_surf2d; break;
+	case 0x02fc: func = nv04_gr_mthd_set_operation; break;
+	default:
+		return false;
+	}
+	return func(device, inst, data);
+}
+
+static bool
+nv03_gr_mthd_sifc(struct nvkm_device *device, u32 inst, u32 mthd, u32 data)
+{
+	bool (*func)(struct nvkm_device *, u32, u32);
+	switch (mthd) {
+	case 0x0184: func = nv01_gr_mthd_bind_chroma; break;
+	case 0x0188: func = nv01_gr_mthd_bind_patt; break;
+	case 0x018c: func = nv04_gr_mthd_bind_rop; break;
+	case 0x0190: func = nv04_gr_mthd_bind_beta1; break;
+	case 0x0194: func = nv04_gr_mthd_bind_surf_dst; break;
+	case 0x02fc: func = nv04_gr_mthd_set_operation; break;
+	default:
+		return false;
+	}
+	return func(device, inst, data);
+}
+
+static bool
+nv04_gr_mthd_sifc(struct nvkm_device *device, u32 inst, u32 mthd, u32 data)
+{
+	bool (*func)(struct nvkm_device *, u32, u32);
+	switch (mthd) {
+	case 0x0184: func = nv01_gr_mthd_bind_chroma; break;
+	case 0x0188: func = nv04_gr_mthd_bind_patt; break;
+	case 0x018c: func = nv04_gr_mthd_bind_rop; break;
+	case 0x0190: func = nv04_gr_mthd_bind_beta1; break;
+	case 0x0194: func = nv04_gr_mthd_bind_beta4; break;
+	case 0x0198: func = nv04_gr_mthd_bind_surf2d; break;
+	case 0x02fc: func = nv04_gr_mthd_set_operation; break;
+	default:
+		return false;
+	}
+	return func(device, inst, data);
+}
+
+static bool
+nv03_gr_mthd_sifm(struct nvkm_device *device, u32 inst, u32 mthd, u32 data)
+{
+	bool (*func)(struct nvkm_device *, u32, u32);
+	switch (mthd) {
+	case 0x0188: func = nv01_gr_mthd_bind_patt; break;
+	case 0x018c: func = nv04_gr_mthd_bind_rop; break;
+	case 0x0190: func = nv04_gr_mthd_bind_beta1; break;
+	case 0x0194: func = nv04_gr_mthd_bind_surf_dst; break;
+	case 0x0304: func = nv04_gr_mthd_set_operation; break;
+	default:
+		return false;
+	}
+	return func(device, inst, data);
+}
+
+static bool
+nv04_gr_mthd_sifm(struct nvkm_device *device, u32 inst, u32 mthd, u32 data)
+{
+	bool (*func)(struct nvkm_device *, u32, u32);
+	switch (mthd) {
+	case 0x0188: func = nv04_gr_mthd_bind_patt; break;
+	case 0x018c: func = nv04_gr_mthd_bind_rop; break;
+	case 0x0190: func = nv04_gr_mthd_bind_beta1; break;
+	case 0x0194: func = nv04_gr_mthd_bind_beta4; break;
+	case 0x0198: func = nv04_gr_mthd_bind_surf2d; break;
+	case 0x0304: func = nv04_gr_mthd_set_operation; break;
+	default:
+		return false;
+	}
+	return func(device, inst, data);
+}
+
+static bool
+nv04_gr_mthd_surf3d(struct nvkm_device *device, u32 inst, u32 mthd, u32 data)
+{
+	bool (*func)(struct nvkm_device *, u32, u32);
+	switch (mthd) {
+	case 0x02f8: func = nv04_gr_mthd_surf3d_clip_h; break;
+	case 0x02fc: func = nv04_gr_mthd_surf3d_clip_v; break;
+	default:
+		return false;
+	}
+	return func(device, inst, data);
+}
+
+static bool
+nv03_gr_mthd_ttri(struct nvkm_device *device, u32 inst, u32 mthd, u32 data)
+{
+	bool (*func)(struct nvkm_device *, u32, u32);
+	switch (mthd) {
+	case 0x0188: func = nv01_gr_mthd_bind_clip; break;
+	case 0x018c: func = nv04_gr_mthd_bind_surf_color; break;
+	case 0x0190: func = nv04_gr_mthd_bind_surf_zeta; break;
+	default:
+		return false;
+	}
+	return func(device, inst, data);
+}
+
+static bool
+nv01_gr_mthd_prim(struct nvkm_device *device, u32 inst, u32 mthd, u32 data)
+{
+	bool (*func)(struct nvkm_device *, u32, u32);
+	switch (mthd) {
+	case 0x0184: func = nv01_gr_mthd_bind_clip; break;
+	case 0x0188: func = nv01_gr_mthd_bind_patt; break;
+	case 0x018c: func = nv04_gr_mthd_bind_rop; break;
+	case 0x0190: func = nv04_gr_mthd_bind_beta1; break;
+	case 0x0194: func = nv04_gr_mthd_bind_surf_dst; break;
+	case 0x02fc: func = nv04_gr_mthd_set_operation; break;
+	default:
+		return false;
+	}
+	return func(device, inst, data);
+}
+
+static bool
+nv04_gr_mthd_prim(struct nvkm_device *device, u32 inst, u32 mthd, u32 data)
+{
+	bool (*func)(struct nvkm_device *, u32, u32);
+	switch (mthd) {
+	case 0x0184: func = nv01_gr_mthd_bind_clip; break;
+	case 0x0188: func = nv04_gr_mthd_bind_patt; break;
+	case 0x018c: func = nv04_gr_mthd_bind_rop; break;
+	case 0x0190: func = nv04_gr_mthd_bind_beta1; break;
+	case 0x0194: func = nv04_gr_mthd_bind_beta4; break;
+	case 0x0198: func = nv04_gr_mthd_bind_surf2d; break;
+	case 0x02fc: func = nv04_gr_mthd_set_operation; break;
+	default:
+		return false;
+	}
+	return func(device, inst, data);
+}
+
+static bool
+nv04_gr_mthd(struct nvkm_device *device, u32 inst, u32 mthd, u32 data)
+{
+	bool (*func)(struct nvkm_device *, u32, u32, u32);
+	switch (nvkm_rd32(device, 0x700000 + inst) & 0x000000ff) {
+	case 0x1c ... 0x1e:
+		   func = nv01_gr_mthd_prim; break;
+	case 0x1f: func = nv01_gr_mthd_blit; break;
+	case 0x21: func = nv01_gr_mthd_ifc; break;
+	case 0x36: func = nv03_gr_mthd_sifc; break;
+	case 0x37: func = nv03_gr_mthd_sifm; break;
+	case 0x48: func = nv03_gr_mthd_ttri; break;
+	case 0x4a: func = nv04_gr_mthd_gdi; break;
+	case 0x4b: func = nv03_gr_mthd_gdi; break;
+	case 0x53: func = nv04_gr_mthd_surf3d; break;
+	case 0x5c ... 0x5e:
+		   func = nv04_gr_mthd_prim; break;
+	case 0x5f: func = nv04_gr_mthd_blit; break;
+	case 0x60: func = nv04_gr_mthd_iifc; break;
+	case 0x61: func = nv04_gr_mthd_ifc; break;
+	case 0x76: func = nv04_gr_mthd_sifc; break;
+	case 0x77: func = nv04_gr_mthd_sifm; break;
+	default:
+		return false;
+	}
+	return func(device, inst, mthd, data);
+}
+
+static int
+nv04_gr_object_bind(struct nvkm_object *object, struct nvkm_gpuobj *parent,
+		    int align, struct nvkm_gpuobj **pgpuobj)
+{
+	int ret = nvkm_gpuobj_new(object->engine->subdev.device, 16, align,
+				  false, parent, pgpuobj);
+	if (ret == 0) {
+		nvkm_kmap(*pgpuobj);
+		nvkm_wo32(*pgpuobj, 0x00, object->oclass);
+#ifdef __BIG_ENDIAN
+		nvkm_mo32(*pgpuobj, 0x00, 0x00080000, 0x00080000);
+#endif
+		nvkm_wo32(*pgpuobj, 0x04, 0x00000000);
+		nvkm_wo32(*pgpuobj, 0x08, 0x00000000);
+		nvkm_wo32(*pgpuobj, 0x0c, 0x00000000);
+		nvkm_done(*pgpuobj);
+	}
+	return ret;
+}
+
+const struct nvkm_object_func
+nv04_gr_object = {
+	.bind = nv04_gr_object_bind,
+};
+
+/*******************************************************************************
+ * PGRAPH context
+ ******************************************************************************/
+
+static struct nv04_gr_chan *
+nv04_gr_channel(struct nv04_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	struct nv04_gr_chan *chan = NULL;
+	if (nvkm_rd32(device, NV04_PGRAPH_CTX_CONTROL) & 0x00010000) {
+		int chid = nvkm_rd32(device, NV04_PGRAPH_CTX_USER) >> 24;
+		if (chid < ARRAY_SIZE(gr->chan))
+			chan = gr->chan[chid];
+	}
+	return chan;
+}
+
+static int
+nv04_gr_load_context(struct nv04_gr_chan *chan, int chid)
+{
+	struct nvkm_device *device = chan->gr->base.engine.subdev.device;
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(nv04_gr_ctx_regs); i++)
+		nvkm_wr32(device, nv04_gr_ctx_regs[i], chan->nv04[i]);
+
+	nvkm_wr32(device, NV04_PGRAPH_CTX_CONTROL, 0x10010100);
+	nvkm_mask(device, NV04_PGRAPH_CTX_USER, 0xff000000, chid << 24);
+	nvkm_mask(device, NV04_PGRAPH_FFINTFC_ST2, 0xfff00000, 0x00000000);
+	return 0;
+}
+
+static int
+nv04_gr_unload_context(struct nv04_gr_chan *chan)
+{
+	struct nvkm_device *device = chan->gr->base.engine.subdev.device;
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(nv04_gr_ctx_regs); i++)
+		chan->nv04[i] = nvkm_rd32(device, nv04_gr_ctx_regs[i]);
+
+	nvkm_wr32(device, NV04_PGRAPH_CTX_CONTROL, 0x10000000);
+	nvkm_mask(device, NV04_PGRAPH_CTX_USER, 0xff000000, 0x0f000000);
+	return 0;
+}
+
+static void
+nv04_gr_context_switch(struct nv04_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	struct nv04_gr_chan *prev = NULL;
+	struct nv04_gr_chan *next = NULL;
+	int chid;
+
+	nv04_gr_idle(&gr->base);
+
+	/* If previous context is valid, we need to save it */
+	prev = nv04_gr_channel(gr);
+	if (prev)
+		nv04_gr_unload_context(prev);
+
+	/* load context for next channel */
+	chid = (nvkm_rd32(device, NV04_PGRAPH_TRAPPED_ADDR) >> 24) & 0x0f;
+	next = gr->chan[chid];
+	if (next)
+		nv04_gr_load_context(next, chid);
+}
+
+static u32 *ctx_reg(struct nv04_gr_chan *chan, u32 reg)
+{
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(nv04_gr_ctx_regs); i++) {
+		if (nv04_gr_ctx_regs[i] == reg)
+			return &chan->nv04[i];
+	}
+
+	return NULL;
+}
+
+static void *
+nv04_gr_chan_dtor(struct nvkm_object *object)
+{
+	struct nv04_gr_chan *chan = nv04_gr_chan(object);
+	struct nv04_gr *gr = chan->gr;
+	unsigned long flags;
+
+	spin_lock_irqsave(&gr->lock, flags);
+	gr->chan[chan->chid] = NULL;
+	spin_unlock_irqrestore(&gr->lock, flags);
+	return chan;
+}
+
+static int
+nv04_gr_chan_fini(struct nvkm_object *object, bool suspend)
+{
+	struct nv04_gr_chan *chan = nv04_gr_chan(object);
+	struct nv04_gr *gr = chan->gr;
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	unsigned long flags;
+
+	spin_lock_irqsave(&gr->lock, flags);
+	nvkm_mask(device, NV04_PGRAPH_FIFO, 0x00000001, 0x00000000);
+	if (nv04_gr_channel(gr) == chan)
+		nv04_gr_unload_context(chan);
+	nvkm_mask(device, NV04_PGRAPH_FIFO, 0x00000001, 0x00000001);
+	spin_unlock_irqrestore(&gr->lock, flags);
+	return 0;
+}
+
+static const struct nvkm_object_func
+nv04_gr_chan = {
+	.dtor = nv04_gr_chan_dtor,
+	.fini = nv04_gr_chan_fini,
+};
+
+static int
+nv04_gr_chan_new(struct nvkm_gr *base, struct nvkm_fifo_chan *fifoch,
+		 const struct nvkm_oclass *oclass, struct nvkm_object **pobject)
+{
+	struct nv04_gr *gr = nv04_gr(base);
+	struct nv04_gr_chan *chan;
+	unsigned long flags;
+
+	if (!(chan = kzalloc(sizeof(*chan), GFP_KERNEL)))
+		return -ENOMEM;
+	nvkm_object_ctor(&nv04_gr_chan, oclass, &chan->object);
+	chan->gr = gr;
+	chan->chid = fifoch->id;
+	*pobject = &chan->object;
+
+	*ctx_reg(chan, NV04_PGRAPH_DEBUG_3) = 0xfad4ff31;
+
+	spin_lock_irqsave(&gr->lock, flags);
+	gr->chan[chan->chid] = chan;
+	spin_unlock_irqrestore(&gr->lock, flags);
+	return 0;
+}
+
+/*******************************************************************************
+ * PGRAPH engine/subdev functions
+ ******************************************************************************/
+
+bool
+nv04_gr_idle(struct nvkm_gr *gr)
+{
+	struct nvkm_subdev *subdev = &gr->engine.subdev;
+	struct nvkm_device *device = subdev->device;
+	u32 mask = 0xffffffff;
+
+	if (device->card_type == NV_40)
+		mask &= ~NV40_PGRAPH_STATUS_SYNC_STALL;
+
+	if (nvkm_msec(device, 2000,
+		if (!(nvkm_rd32(device, NV04_PGRAPH_STATUS) & mask))
+			break;
+	) < 0) {
+		nvkm_error(subdev, "idle timed out with status %08x\n",
+			   nvkm_rd32(device, NV04_PGRAPH_STATUS));
+		return false;
+	}
+
+	return true;
+}
+
+static const struct nvkm_bitfield
+nv04_gr_intr_name[] = {
+	{ NV_PGRAPH_INTR_NOTIFY, "NOTIFY" },
+	{}
+};
+
+static const struct nvkm_bitfield
+nv04_gr_nstatus[] = {
+	{ NV04_PGRAPH_NSTATUS_STATE_IN_USE,       "STATE_IN_USE" },
+	{ NV04_PGRAPH_NSTATUS_INVALID_STATE,      "INVALID_STATE" },
+	{ NV04_PGRAPH_NSTATUS_BAD_ARGUMENT,       "BAD_ARGUMENT" },
+	{ NV04_PGRAPH_NSTATUS_PROTECTION_FAULT,   "PROTECTION_FAULT" },
+	{}
+};
+
+const struct nvkm_bitfield
+nv04_gr_nsource[] = {
+	{ NV03_PGRAPH_NSOURCE_NOTIFICATION,       "NOTIFICATION" },
+	{ NV03_PGRAPH_NSOURCE_DATA_ERROR,         "DATA_ERROR" },
+	{ NV03_PGRAPH_NSOURCE_PROTECTION_ERROR,   "PROTECTION_ERROR" },
+	{ NV03_PGRAPH_NSOURCE_RANGE_EXCEPTION,    "RANGE_EXCEPTION" },
+	{ NV03_PGRAPH_NSOURCE_LIMIT_COLOR,        "LIMIT_COLOR" },
+	{ NV03_PGRAPH_NSOURCE_LIMIT_ZETA,         "LIMIT_ZETA" },
+	{ NV03_PGRAPH_NSOURCE_ILLEGAL_MTHD,       "ILLEGAL_MTHD" },
+	{ NV03_PGRAPH_NSOURCE_DMA_R_PROTECTION,   "DMA_R_PROTECTION" },
+	{ NV03_PGRAPH_NSOURCE_DMA_W_PROTECTION,   "DMA_W_PROTECTION" },
+	{ NV03_PGRAPH_NSOURCE_FORMAT_EXCEPTION,   "FORMAT_EXCEPTION" },
+	{ NV03_PGRAPH_NSOURCE_PATCH_EXCEPTION,    "PATCH_EXCEPTION" },
+	{ NV03_PGRAPH_NSOURCE_STATE_INVALID,      "STATE_INVALID" },
+	{ NV03_PGRAPH_NSOURCE_DOUBLE_NOTIFY,      "DOUBLE_NOTIFY" },
+	{ NV03_PGRAPH_NSOURCE_NOTIFY_IN_USE,      "NOTIFY_IN_USE" },
+	{ NV03_PGRAPH_NSOURCE_METHOD_CNT,         "METHOD_CNT" },
+	{ NV03_PGRAPH_NSOURCE_BFR_NOTIFICATION,   "BFR_NOTIFICATION" },
+	{ NV03_PGRAPH_NSOURCE_DMA_VTX_PROTECTION, "DMA_VTX_PROTECTION" },
+	{ NV03_PGRAPH_NSOURCE_DMA_WIDTH_A,        "DMA_WIDTH_A" },
+	{ NV03_PGRAPH_NSOURCE_DMA_WIDTH_B,        "DMA_WIDTH_B" },
+	{}
+};
+
+static void
+nv04_gr_intr(struct nvkm_gr *base)
+{
+	struct nv04_gr *gr = nv04_gr(base);
+	struct nvkm_subdev *subdev = &gr->base.engine.subdev;
+	struct nvkm_device *device = subdev->device;
+	u32 stat = nvkm_rd32(device, NV03_PGRAPH_INTR);
+	u32 nsource = nvkm_rd32(device, NV03_PGRAPH_NSOURCE);
+	u32 nstatus = nvkm_rd32(device, NV03_PGRAPH_NSTATUS);
+	u32 addr = nvkm_rd32(device, NV04_PGRAPH_TRAPPED_ADDR);
+	u32 chid = (addr & 0x0f000000) >> 24;
+	u32 subc = (addr & 0x0000e000) >> 13;
+	u32 mthd = (addr & 0x00001ffc);
+	u32 data = nvkm_rd32(device, NV04_PGRAPH_TRAPPED_DATA);
+	u32 class = nvkm_rd32(device, 0x400180 + subc * 4) & 0xff;
+	u32 inst = (nvkm_rd32(device, 0x40016c) & 0xffff) << 4;
+	u32 show = stat;
+	char msg[128], src[128], sta[128];
+	struct nv04_gr_chan *chan;
+	unsigned long flags;
+
+	spin_lock_irqsave(&gr->lock, flags);
+	chan = gr->chan[chid];
+
+	if (stat & NV_PGRAPH_INTR_NOTIFY) {
+		if (chan && (nsource & NV03_PGRAPH_NSOURCE_ILLEGAL_MTHD)) {
+			if (!nv04_gr_mthd(device, inst, mthd, data))
+				show &= ~NV_PGRAPH_INTR_NOTIFY;
+		}
+	}
+
+	if (stat & NV_PGRAPH_INTR_CONTEXT_SWITCH) {
+		nvkm_wr32(device, NV03_PGRAPH_INTR, NV_PGRAPH_INTR_CONTEXT_SWITCH);
+		stat &= ~NV_PGRAPH_INTR_CONTEXT_SWITCH;
+		show &= ~NV_PGRAPH_INTR_CONTEXT_SWITCH;
+		nv04_gr_context_switch(gr);
+	}
+
+	nvkm_wr32(device, NV03_PGRAPH_INTR, stat);
+	nvkm_wr32(device, NV04_PGRAPH_FIFO, 0x00000001);
+
+	if (show) {
+		nvkm_snprintbf(msg, sizeof(msg), nv04_gr_intr_name, show);
+		nvkm_snprintbf(src, sizeof(src), nv04_gr_nsource, nsource);
+		nvkm_snprintbf(sta, sizeof(sta), nv04_gr_nstatus, nstatus);
+		nvkm_error(subdev, "intr %08x [%s] nsource %08x [%s] "
+				   "nstatus %08x [%s] ch %d [%s] subc %d "
+				   "class %04x mthd %04x data %08x\n",
+			   show, msg, nsource, src, nstatus, sta, chid,
+			   chan ? chan->object.client->name : "unknown",
+			   subc, class, mthd, data);
+	}
+
+	spin_unlock_irqrestore(&gr->lock, flags);
+}
+
+static int
+nv04_gr_init(struct nvkm_gr *base)
+{
+	struct nv04_gr *gr = nv04_gr(base);
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+
+	/* Enable PGRAPH interrupts */
+	nvkm_wr32(device, NV03_PGRAPH_INTR, 0xFFFFFFFF);
+	nvkm_wr32(device, NV03_PGRAPH_INTR_EN, 0xFFFFFFFF);
+
+	nvkm_wr32(device, NV04_PGRAPH_VALID1, 0);
+	nvkm_wr32(device, NV04_PGRAPH_VALID2, 0);
+	/*nvkm_wr32(device, NV04_PGRAPH_DEBUG_0, 0x000001FF);
+	nvkm_wr32(device, NV04_PGRAPH_DEBUG_0, 0x001FFFFF);*/
+	nvkm_wr32(device, NV04_PGRAPH_DEBUG_0, 0x1231c000);
+	/*1231C000 blob, 001 haiku*/
+	/*V_WRITE(NV04_PGRAPH_DEBUG_1, 0xf2d91100);*/
+	nvkm_wr32(device, NV04_PGRAPH_DEBUG_1, 0x72111100);
+	/*0x72111100 blob , 01 haiku*/
+	/*nvkm_wr32(device, NV04_PGRAPH_DEBUG_2, 0x11d5f870);*/
+	nvkm_wr32(device, NV04_PGRAPH_DEBUG_2, 0x11d5f071);
+	/*haiku same*/
+
+	/*nvkm_wr32(device, NV04_PGRAPH_DEBUG_3, 0xfad4ff31);*/
+	nvkm_wr32(device, NV04_PGRAPH_DEBUG_3, 0xf0d4ff31);
+	/*haiku and blob 10d4*/
+
+	nvkm_wr32(device, NV04_PGRAPH_STATE        , 0xFFFFFFFF);
+	nvkm_wr32(device, NV04_PGRAPH_CTX_CONTROL  , 0x10000100);
+	nvkm_mask(device, NV04_PGRAPH_CTX_USER, 0xff000000, 0x0f000000);
+
+	/* These don't belong here, they're part of a per-channel context */
+	nvkm_wr32(device, NV04_PGRAPH_PATTERN_SHAPE, 0x00000000);
+	nvkm_wr32(device, NV04_PGRAPH_BETA_AND     , 0xFFFFFFFF);
+	return 0;
+}
+
+static const struct nvkm_gr_func
+nv04_gr = {
+	.init = nv04_gr_init,
+	.intr = nv04_gr_intr,
+	.chan_new = nv04_gr_chan_new,
+	.sclass = {
+		{ -1, -1, 0x0012, &nv04_gr_object }, /* beta1 */
+		{ -1, -1, 0x0017, &nv04_gr_object }, /* chroma */
+		{ -1, -1, 0x0018, &nv04_gr_object }, /* pattern (nv01) */
+		{ -1, -1, 0x0019, &nv04_gr_object }, /* clip */
+		{ -1, -1, 0x001c, &nv04_gr_object }, /* line */
+		{ -1, -1, 0x001d, &nv04_gr_object }, /* tri */
+		{ -1, -1, 0x001e, &nv04_gr_object }, /* rect */
+		{ -1, -1, 0x001f, &nv04_gr_object },
+		{ -1, -1, 0x0021, &nv04_gr_object },
+		{ -1, -1, 0x0030, &nv04_gr_object }, /* null */
+		{ -1, -1, 0x0036, &nv04_gr_object },
+		{ -1, -1, 0x0037, &nv04_gr_object },
+		{ -1, -1, 0x0038, &nv04_gr_object }, /* dvd subpicture */
+		{ -1, -1, 0x0039, &nv04_gr_object }, /* m2mf */
+		{ -1, -1, 0x0042, &nv04_gr_object }, /* surf2d */
+		{ -1, -1, 0x0043, &nv04_gr_object }, /* rop */
+		{ -1, -1, 0x0044, &nv04_gr_object }, /* pattern */
+		{ -1, -1, 0x0048, &nv04_gr_object },
+		{ -1, -1, 0x004a, &nv04_gr_object },
+		{ -1, -1, 0x004b, &nv04_gr_object },
+		{ -1, -1, 0x0052, &nv04_gr_object }, /* swzsurf */
+		{ -1, -1, 0x0053, &nv04_gr_object },
+		{ -1, -1, 0x0054, &nv04_gr_object }, /* ttri */
+		{ -1, -1, 0x0055, &nv04_gr_object }, /* mtri */
+		{ -1, -1, 0x0057, &nv04_gr_object }, /* chroma */
+		{ -1, -1, 0x0058, &nv04_gr_object }, /* surf_dst */
+		{ -1, -1, 0x0059, &nv04_gr_object }, /* surf_src */
+		{ -1, -1, 0x005a, &nv04_gr_object }, /* surf_color */
+		{ -1, -1, 0x005b, &nv04_gr_object }, /* surf_zeta */
+		{ -1, -1, 0x005c, &nv04_gr_object }, /* line */
+		{ -1, -1, 0x005d, &nv04_gr_object }, /* tri */
+		{ -1, -1, 0x005e, &nv04_gr_object }, /* rect */
+		{ -1, -1, 0x005f, &nv04_gr_object },
+		{ -1, -1, 0x0060, &nv04_gr_object },
+		{ -1, -1, 0x0061, &nv04_gr_object },
+		{ -1, -1, 0x0064, &nv04_gr_object }, /* iifc (nv05) */
+		{ -1, -1, 0x0065, &nv04_gr_object }, /* ifc (nv05) */
+		{ -1, -1, 0x0066, &nv04_gr_object }, /* sifc (nv05) */
+		{ -1, -1, 0x0072, &nv04_gr_object }, /* beta4 */
+		{ -1, -1, 0x0076, &nv04_gr_object },
+		{ -1, -1, 0x0077, &nv04_gr_object },
+		{}
+	}
+};
+
+int
+nv04_gr_new(struct nvkm_device *device, enum nvkm_subdev_type type, int inst, struct nvkm_gr **pgr)
+{
+	struct nv04_gr *gr;
+
+	if (!(gr = kzalloc(sizeof(*gr), GFP_KERNEL)))
+		return -ENOMEM;
+	spin_lock_init(&gr->lock);
+	*pgr = &gr->base;
+
+	return nvkm_gr_ctor(&nv04_gr, device, type, inst, true, &gr->base);
+}