drm/nvc0-/gr: bring in some macros to abstract falcon isa differences

Need. A. Compiler...

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

drivers/gpu/drm/nouveau/core/engine/graph/fuc/com.fuc

@@ -38,7 +38,7 @@ queue_put:
 	cmpu b32 $r8 $r9
 	bra ne #queue_put_next
 		mov $r15 E_CMD_OVERFLOW
-		call #error
+		call(error)
 		ret
 
 	// store cmd/data on queue
@@ -92,18 +92,16 @@ queue_get_done:
 // Out: $r15 value
 //
 nv_rd32:
-	mov $r11 0x728
-	shl b32 $r11 6
 	mov b32 $r12 $r14
 	bset $r12 31			// MMIO_CTRL_PENDING
-	iowr I[$r11 + 0x000] $r12	// MMIO_CTRL
+	nv_iowr(NV_PGRAPH_FECS_MMIO_CTRL, 0, $r12)
 	nv_rd32_wait:
-		iord $r12 I[$r11 + 0x000]
+		nv_iord($r12, NV_PGRAPH_FECS_MMIO_CTRL, 0)
 		xbit $r12 $r12 31
 		bra ne #nv_rd32_wait
 	mov $r10 6			// DONE_MMIO_RD
-	call #wait_doneo
-	iord $r15 I[$r11 + 0x100]	// MMIO_RDVAL
+	call(wait_doneo)
+	nv_iord($r15, NV_PGRAPH_FECS_MMIO_RDVAL, 0)
 	ret
 
 // nv_wr32 - write 32-bit value to nv register
@@ -112,37 +110,17 @@ nv_rd32:
 //      $r15 value
 //
 nv_wr32:
-	mov $r11 0x728
-	shl b32 $r11 6
-	iowr I[$r11 + 0x200] $r15	// MMIO_WRVAL
+	nv_iowr(NV_PGRAPH_FECS_MMIO_WRVAL, 0, $r15)
 	mov b32 $r12 $r14
 	bset $r12 31			// MMIO_CTRL_PENDING
 	bset $r12 30			// MMIO_CTRL_WRITE
-	iowr I[$r11 + 0x000] $r12	// MMIO_CTRL
+	nv_iowr(NV_PGRAPH_FECS_MMIO_CTRL, 0, $r12)
 	nv_wr32_wait:
-		iord $r12 I[$r11 + 0x000]
+		nv_iord($r12, NV_PGRAPH_FECS_MMIO_CTRL, 0)
 		xbit $r12 $r12 31
 		bra ne #nv_wr32_wait
 	ret
 
-// (re)set watchdog timer
-//
-// In : $r15 timeout
-//
-watchdog_reset:
-	mov $r8 0x430
-	shl b32 $r8 6
-	bset $r15 31
-	iowr I[$r8 + 0x000] $r15
-	ret
-
-// clear watchdog timer
-watchdog_clear:
-	mov $r8 0x430
-	shl b32 $r8 6
-	iowr I[$r8 + 0x000] $r0
-	ret
-
 // wait_donez - wait on FUC_DONE bit to become clear
 //
 // In : $r10 bit to wait on
@@ -163,13 +141,9 @@ wait_donez:
 //
 wait_doneo:
 	trace_set(T_WAIT);
-	mov $r8 0x818
-	shl b32 $r8 6
-	iowr I[$r8 + 0x000] $r10
+	nv_iowr(NV_PGRAPH_FECS_CC_SCRATCH_VAL(6), 0, $r10)
 	wait_doneo_e:
-		mov $r8 0x400
-		shl b32 $r8 6
-		iord $r8 I[$r8 + 0x000]
+		nv_iord($r8, NV_PGRAPH_FECS_SIGNAL, 0)
 		xbit $r8 $r8 $r10
 		bra e #wait_doneo_e
 	trace_clr(T_WAIT)
@@ -209,21 +183,18 @@ mmctx_size:
 //
 mmctx_xfer:
 	trace_set(T_MMCTX)
-	mov $r8 0x710
-	shl b32 $r8 6
 	clear b32 $r9
 	or $r11 $r11
 	bra e #mmctx_base_disabled
-		iowr I[$r8 + 0x000] $r11	// MMCTX_BASE
+		nv_iowr(NV_PGRAPH_FECS_MMCTX_BASE, 0, $r11)
 		bset $r9 0			// BASE_EN
 	mmctx_base_disabled:
 	or $r14 $r14
 	bra e #mmctx_multi_disabled
-		iowr I[$r8 + 0x200] $r14 	// MMCTX_MULTI_STRIDE
-		iowr I[$r8 + 0x300] $r15 	// MMCTX_MULTI_MASK
+		nv_iowr(NV_PGRAPH_FECS_MMCTX_MULTI_STRIDE, 0, $r14)
+		nv_iowr(NV_PGRAPH_FECS_MMCTX_MULTI_MASK, 0, $r15)
 		bset $r9 1			// MULTI_EN
 	mmctx_multi_disabled:
-	add b32 $r8 0x100
 
 	xbit $r11 $r10 0
 	shl b32 $r11 16			// DIR
@@ -231,20 +202,20 @@ mmctx_xfer:
 	xbit $r14 $r10 1
 	shl b32 $r14 17
 	or $r11 $r14			// START_TRIGGER
-	iowr I[$r8 + 0x000] $r11	// MMCTX_CTRL
+	nv_iowr(NV_PGRAPH_FECS_MMCTX_CTRL, 0, $r11)
 
 	// loop over the mmio list, and send requests to the hw
 	mmctx_exec_loop:
 		// wait for space in mmctx queue
 		mmctx_wait_free:
-			iord $r14 I[$r8 + 0x000] // MMCTX_CTRL
+			nv_iord($r14, NV_PGRAPH_FECS_MMCTX_CTRL, 0)
 			and $r14 0x1f
 			bra e #mmctx_wait_free
 
 		// queue up an entry
 		ld b32 $r14 D[$r12]
 		or $r14 $r9
-		iowr I[$r8 + 0x300] $r14
+		nv_iowr(NV_PGRAPH_FECS_MMCTX_QUEUE, 0, $r14)
 		add b32 $r12 4
 		cmpu b32 $r12 $r13
 		bra ne #mmctx_exec_loop
@@ -253,22 +224,22 @@ mmctx_xfer:
 	bra ne #mmctx_stop
 		// wait for queue to empty
 		mmctx_fini_wait:
-			iord $r11 I[$r8 + 0x000]	// MMCTX_CTRL
+			nv_iord($r11, NV_PGRAPH_FECS_MMCTX_CTRL, 0)
 			and $r11 0x1f
 			cmpu b32 $r11 0x10
 			bra ne #mmctx_fini_wait
 		mov $r10 2				// DONE_MMCTX
-		call #wait_donez
+		call(wait_donez)
 		bra #mmctx_done
 	mmctx_stop:
 		xbit $r11 $r10 0
 		shl b32 $r11 16			// DIR
 		bset $r11 12			// QLIMIT = 0x10
 		bset $r11 18			// STOP_TRIGGER
-		iowr I[$r8 + 0x000] $r11	// MMCTX_CTRL
+		nv_iowr(NV_PGRAPH_FECS_MMCTX_CTRL, 0, $r11)
 		mmctx_stop_wait:
 			// wait for STOP_TRIGGER to clear
-			iord $r11 I[$r8 + 0x000] // MMCTX_CTRL
+			nv_iord($r11, NV_PGRAPH_FECS_MMCTX_CTRL, 0)
 			xbit $r11 $r11 18
 			bra ne #mmctx_stop_wait
 	mmctx_done:
@@ -280,28 +251,24 @@ mmctx_xfer:
 strand_wait:
 	push $r10
 	mov $r10 2
-	call #wait_donez
+	call(wait_donez)
 	pop $r10
 	ret
 
 // unknown - call before issuing strand commands
 //
 strand_pre:
-	mov $r8 0x4afc
-	sethi $r8 0x20000
-	mov $r9 0xc
-	iowr I[$r8] $r9
-	call #strand_wait
+	mov $r9 NV_PGRAPH_FECS_STRAND_CMD_ENABLE
+	nv_iowr(NV_PGRAPH_FECS_STRAND_CMD, 0x3f, $r9)
+	call(strand_wait)
 	ret
 
 // unknown - call after issuing strand commands
 //
 strand_post:
-	mov $r8 0x4afc
-	sethi $r8 0x20000
-	mov $r9 0xd
-	iowr I[$r8] $r9
-	call #strand_wait
+	mov $r9 NV_PGRAPH_FECS_STRAND_CMD_DISABLE
+	nv_iowr(NV_PGRAPH_FECS_STRAND_CMD, 0x3f, $r9)
+	call(strand_wait)
 	ret
 
 // Selects strand set?!
@@ -309,18 +276,14 @@ strand_post:
 // In: $r14 id
 //
 strand_set:
-	mov $r10 0x4ffc
-	sethi $r10 0x20000
-	sub b32 $r11 $r10 0x500
 	mov $r12 0xf
-	iowr I[$r10 + 0x000] $r12		// 0x93c = 0xf
-	mov $r12 0xb
-	iowr I[$r11 + 0x000] $r12		// 0x928 = 0xb
-	call #strand_wait
-	iowr I[$r10 + 0x000] $r14		// 0x93c = <id>
-	mov $r12 0xa
-	iowr I[$r11 + 0x000] $r12		// 0x928 = 0xa
-	call #strand_wait
+	nv_iowr(NV_PGRAPH_FECS_STRAND_FILTER, 0x3f, $r12)
+	mov $r12 NV_PGRAPH_FECS_STRAND_CMD_DEACTIVATE_FILTER
+	nv_iowr(NV_PGRAPH_FECS_STRAND_CMD, 0x3f, $r12)
+	nv_iowr(NV_PGRAPH_FECS_STRAND_FILTER, 0x3f, $r14)
+	mov $r12 NV_PGRAPH_FECS_STRAND_CMD_ACTIVATE_FILTER
+	nv_iowr(NV_PGRAPH_FECS_STRAND_CMD, 0x3f, $r12)
+	call(strand_wait)
 	ret
 
 // Initialise strand context data
@@ -332,30 +295,27 @@ strand_set:
 //
 strand_ctx_init:
 	trace_set(T_STRINIT)
-	call #strand_pre
+	call(strand_pre)
 	mov $r14 3
-	call #strand_set
-	mov $r10 0x46fc
-	sethi $r10 0x20000
-	add b32 $r11 $r10 0x400
-	iowr I[$r10 + 0x100] $r0	// STRAND_FIRST_GENE = 0
-	mov $r12 1
-	iowr I[$r11 + 0x000] $r12	// STRAND_CMD = LATCH_FIRST_GENE
-	call #strand_wait
+	call(strand_set)
+
+	clear b32 $r12
+	nv_iowr(NV_PGRAPH_FECS_STRAND_SELECT, 0x3f, $r12)
+	mov $r12 NV_PGRAPH_FECS_STRAND_CMD_SEEK
+	nv_iowr(NV_PGRAPH_FECS_STRAND_CMD, 0x3f, $r12)
+	call(strand_wait)
 	sub b32 $r12 $r0 1
-	iowr I[$r10 + 0x000] $r12	// STRAND_GENE_CNT = 0xffffffff
-	mov $r12 2
-	iowr I[$r11 + 0x000] $r12	// STRAND_CMD = LATCH_GENE_CNT
-	call #strand_wait
-	call #strand_post
+	nv_iowr(NV_PGRAPH_FECS_STRAND_DATA, 0x3f, $r12)
+	mov $r12 NV_PGRAPH_FECS_STRAND_CMD_GET_INFO
+	nv_iowr(NV_PGRAPH_FECS_STRAND_CMD, 0x3f, $r12)
+	call(strand_wait)
+	call(strand_post)
 
 	// read the size of each strand, poke the context offset of
 	// each into STRAND_{SAVE,LOAD}_SWBASE now, no need to worry
 	// about it later then.
-	mov $r8 0x880
-	shl b32 $r8 6
-	iord $r9 I[$r8 + 0x000]		// STRANDS
-	add b32 $r8 0x2200
+	nv_mkio($r8, NV_PGRAPH_FECS_STRAND_SAVE_SWBASE, 0x00)
+	nv_iord($r9, NV_PGRAPH_FECS_STRANDS_CNT, 0x00)
 	shr b32 $r14 $r15 8
 	ctx_init_strand_loop:
 		iowr I[$r8 + 0x000] $r14	// STRAND_SAVE_SWBASE

drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpc.fuc

@@ -58,12 +58,9 @@ mmio_list_base:
 //
 error:
 	push $r14
-	mov $r14 -0x67ec 	// 0x9814
-	sethi $r14 0x400000
-	call #nv_wr32		// HUB_CTXCTL_CC_SCRATCH[5] = error code
-	add b32 $r14 0x41c
+	nv_wr32(NV_PGRAPH_FECS_CC_SCRATCH_VAL(5), $r15)
 	mov $r15 1
-	call #nv_wr32		// HUB_CTXCTL_INTR_UP_SET
+	nv_wr32(NV_PGRAPH_FECS_INTR_UP_SET, $r15)
 	pop $r14
 	ret
 
@@ -84,46 +81,40 @@ init:
 	mov $sp $r0
 
 	// enable fifo access
-	mov $r1 0x1200
-	mov $r2 2
-	iowr I[$r1 + 0x000] $r2		// FIFO_ENABLE
+	mov $r2 NV_PGRAPH_GPCX_GPCCS_ACCESS_FIFO
+	nv_iowr(NV_PGRAPH_GPCX_GPCCS_ACCESS, 0, $r2)
 
 	// setup i0 handler, and route all interrupts to it
 	mov $r1 #ih
 	mov $iv0 $r1
-	mov $r1 0x400
-	iowr I[$r1 + 0x300] $r0		// INTR_DISPATCH
+	nv_iowr(NV_PGRAPH_GPCX_GPCCS_INTR_ROUTE, 0, $r0)
 
 	// enable fifo interrupt
-	mov $r2 4
-	iowr I[$r1 + 0x000] $r2		// INTR_EN_SET
+	mov $r2 NV_PGRAPH_GPCX_GPCCS_INTR_EN_SET_FIFO
+	nv_iowr(NV_PGRAPH_GPCX_GPCCS_INTR_EN_SET, 0, $r2)
 
 	// enable interrupts
 	bset $flags ie0
 
 	// figure out which GPC we are, and how many TPCs we have
-	mov $r1 0x608
-	shl b32 $r1 6
-	iord $r2 I[$r1 + 0x000]		// UNITS
+	nv_iord($r2, NV_PGRAPH_GPCX_GPCCS_UNITS, 0)
 	mov $r3 1
 	and $r2 0x1f
 	shl b32 $r3 $r2
 	sub b32 $r3 1
 	st b32 D[$r0 + #tpc_count] $r2
 	st b32 D[$r0 + #tpc_mask] $r3
-	add b32 $r1 0x400
-	iord $r2 I[$r1 + 0x000]		// MYINDEX
+	nv_iord($r2, NV_PGRAPH_GPCX_GPCCS_MYINDEX, 0)
 	st b32 D[$r0 + #gpc_id] $r2
 
 #if NV_PGRAPH_GPCX_UNK__SIZE > 0
 	// figure out which, and how many, UNKs are actually present
-	mov $r14 0x0c30
-	sethi $r14 0x500000
+	imm32($r14, 0x500c30)
 	clear b32 $r2
 	clear b32 $r3
 	clear b32 $r4
 	init_unk_loop:
-		call #nv_rd32
+		call(nv_rd32)
 		cmp b32 $r15 0
 		bra z #init_unk_next
 			mov $r15 1
@@ -146,23 +137,21 @@ init:
 
 	// set mmctx base addresses now so we don't have to do it later,
 	// they don't currently ever change
-	mov $r4 0x700
-	shl b32 $r4 6
 	shr b32 $r5 $r2 8
-	iowr I[$r4 + 0x000] $r5		// MMCTX_SAVE_SWBASE
-	iowr I[$r4 + 0x100] $r5		// MMCTX_LOAD_SWBASE
+	nv_iowr(NV_PGRAPH_GPCX_GPCCS_MMCTX_SAVE_SWBASE, 0, $r5)
+	nv_iowr(NV_PGRAPH_GPCX_GPCCS_MMCTX_LOAD_SWBASE, 0, $r5)
 
 	// calculate GPC mmio context size
 	ld b32 $r14 D[$r0 + #gpc_mmio_list_head]
 	ld b32 $r15 D[$r0 + #gpc_mmio_list_tail]
-	call #mmctx_size
+	call(mmctx_size)
 	add b32 $r2 $r15
 	add b32 $r3 $r15
 
 	// calculate per-TPC mmio context size
 	ld b32 $r14 D[$r0 + #tpc_mmio_list_head]
 	ld b32 $r15 D[$r0 + #tpc_mmio_list_tail]
-	call #mmctx_size
+	call(mmctx_size)
 	ld b32 $r14 D[$r0 + #tpc_count]
 	mulu $r14 $r15
 	add b32 $r2 $r14
@@ -172,7 +161,7 @@ init:
 	// calculate per-UNK mmio context size
 	ld b32 $r14 D[$r0 + #unk_mmio_list_head]
 	ld b32 $r15 D[$r0 + #unk_mmio_list_tail]
-	call #mmctx_size
+	call(mmctx_size)
 	ld b32 $r14 D[$r0 + #unk_count]
 	mulu $r14 $r15
 	add b32 $r2 $r14
@@ -180,9 +169,8 @@ init:
 #endif
 
 	// round up base/size to 256 byte boundary (for strand SWBASE)
-	add b32 $r4 0x1300
 	shr b32 $r3 2
-	iowr I[$r4 + 0x000] $r3		// MMCTX_LOAD_COUNT, wtf for?!?
+	nv_iowr(NV_PGRAPH_GPCX_GPCCS_MMCTX_LOAD_COUNT, 0, $r3) // wtf for?!
 	shr b32 $r2 8
 	shr b32 $r3 6
 	add b32 $r2 1
@@ -192,7 +180,7 @@ init:
 
 	// calculate size of strand context data
 	mov b32 $r15 $r2
-	call #strand_ctx_init
+	call(strand_ctx_init)
 	add b32 $r3 $r15
 
 	// save context size, and tell HUB we're done
@@ -208,7 +196,7 @@ main:
 	bset $flags $p0
 	sleep $p0
 	mov $r13 #cmd_queue
-	call #queue_get
+	call(queue_get)
 	bra $p1 #main
 
 	// 0x0000-0x0003 are all context transfers
@@ -224,13 +212,13 @@ main:
 		or $r1 $r14
 		mov $flags $r1
 		// transfer context data
-		call #ctx_xfer
+		call(ctx_xfer)
 		bra #main
 
 	main_not_ctx_xfer:
 	shl b32 $r15 $r14 16
 	or $r15 E_BAD_COMMAND
-	call #error
+	call(error)
 	bra #main
 
 // interrupt handler
@@ -247,22 +235,20 @@ ih:
 	clear b32 $r0
 
 	// incoming fifo command?
-	iord $r10 I[$r0 + 0x200]	// INTR
-	and $r11 $r10 0x00000004
+	nv_iord($r10, NV_PGRAPH_GPCX_GPCCS_INTR, 0)
+	and $r11 $r10 NV_PGRAPH_GPCX_GPCCS_INTR_FIFO
 	bra e #ih_no_fifo
 		// queue incoming fifo command for later processing
-		mov $r11 0x1900
 		mov $r13 #cmd_queue
-		iord $r14 I[$r11 + 0x100]	// FIFO_CMD
-		iord $r15 I[$r11 + 0x000]	// FIFO_DATA
-		call #queue_put
-		add b32 $r11 0x400
+		nv_iord($r14, NV_PGRAPH_GPCX_GPCCS_FIFO_CMD, 0)
+		nv_iord($r15, NV_PGRAPH_GPCX_GPCCS_FIFO_DATA, 0)
+		call(queue_put)
 		mov $r14 1
-		iowr I[$r11 + 0x000] $r14	// FIFO_ACK
+		nv_iowr(NV_PGRAPH_GPCX_GPCCS_FIFO_ACK, 0, $r14)
 
 	// ack, and wake up main()
 	ih_no_fifo:
-	iowr I[$r0 + 0x100] $r10	// INTR_ACK
+	nv_iowr(NV_PGRAPH_GPCX_GPCCS_INTR_ACK, 0, $r10)
 
 	pop $r15
 	pop $r14
@@ -283,9 +269,7 @@ hub_barrier_done:
 	mov $r15 1
 	ld b32 $r14 D[$r0 + #gpc_id]
 	shl b32 $r15 $r14
-	mov $r14 -0x6be8 	// 0x409418 - HUB_BAR_SET
-	sethi $r14 0x400000
-	call #nv_wr32
+	nv_wr32(0x409418, $r15)	// 0x409418 - HUB_BAR_SET
 	ret
 
 // Disables various things, waits a bit, and re-enables them..
@@ -295,16 +279,15 @@ hub_barrier_done:
 // funny things happen.
 //
 ctx_redswitch:
-	mov $r14 0x614
-	shl b32 $r14 6
-	mov $r15 0x020
-	iowr I[$r14] $r15	// GPC_RED_SWITCH = POWER
-	mov $r15 8
+	mov $r15 NV_PGRAPH_GPCX_GPCCS_RED_SWITCH_POWER
+	nv_iowr(NV_PGRAPH_GPCX_GPCCS_RED_SWITCH, 0, $r15)
+	mov $r14 8
 	ctx_redswitch_delay:
-		sub b32 $r15 1
+		sub b32 $r14 1
 		bra ne #ctx_redswitch_delay
-	mov $r15 0xa20
-	iowr I[$r14] $r15	// GPC_RED_SWITCH = UNK11, ENABLE, POWER
+	or $r15 NV_PGRAPH_GPCX_GPCCS_RED_SWITCH_UNK11
+	or $r15 NV_PGRAPH_GPCX_GPCCS_RED_SWITCH_ENABLE
+	nv_iowr(NV_PGRAPH_GPCX_GPCCS_RED_SWITCH, 0, $r15)
 	ret
 
 // Transfer GPC context data between GPU and storage area
@@ -317,46 +300,37 @@ ctx_redswitch:
 //
 ctx_xfer:
 	// set context base address
-	mov $r1 0xa04
-	shl b32 $r1 6
-	iowr I[$r1 + 0x000] $r15// MEM_BASE
+	nv_iowr(NV_PGRAPH_GPCX_GPCCS_MEM_BASE, 0, $r15)
 	bra not $p1 #ctx_xfer_not_load
-		call #ctx_redswitch
+		call(ctx_redswitch)
 	ctx_xfer_not_load:
 
 	// strands
-	mov $r1 0x4afc
-	sethi $r1 0x20000
-	mov $r2 0xc
-	iowr I[$r1] $r2		// STRAND_CMD(0x3f) = 0x0c
-	call #strand_wait
-	mov $r2 0x47fc
-	sethi $r2 0x20000
-	iowr I[$r2] $r0		// STRAND_FIRST_GENE(0x3f) = 0x00
-	xbit $r2 $flags $p1
-	add b32 $r2 3
-	iowr I[$r1] $r2		// STRAND_CMD(0x3f) = 0x03/0x04 (SAVE/LOAD)
+	call(strand_pre)
+	clear b32 $r2
+	nv_iowr(NV_PGRAPH_GPCX_GPCCS_STRAND_SELECT, 0x3f, $r2)
+	xbit $r2 $flags $p1	// SAVE/LOAD
+	add b32 $r2 NV_PGRAPH_GPCX_GPCCS_STRAND_CMD_SAVE
+	nv_iowr(NV_PGRAPH_GPCX_GPCCS_STRAND_CMD, 0x3f, $r2)
 
 	// mmio context
 	xbit $r10 $flags $p1	// direction
 	or $r10 2		// first
-	mov $r11 0x0000
-	sethi $r11 0x500000
+	imm32($r11,0x500000)
 	ld b32 $r12 D[$r0 + #gpc_id]
 	shl b32 $r12 15
 	add b32 $r11 $r12	// base = NV_PGRAPH_GPCn
 	ld b32 $r12 D[$r0 + #gpc_mmio_list_head]
 	ld b32 $r13 D[$r0 + #gpc_mmio_list_tail]
 	mov $r14 0		// not multi
-	call #mmctx_xfer
+	call(mmctx_xfer)
 
 	// per-TPC mmio context
 	xbit $r10 $flags $p1	// direction
 #if !NV_PGRAPH_GPCX_UNK__SIZE
 	or $r10 4		// last
 #endif
-	mov $r11 0x4000
-	sethi $r11 0x500000	// base = NV_PGRAPH_GPC0_TPC0
+	imm32($r11, 0x504000)
 	ld b32 $r12 D[$r0 + #gpc_id]
 	shl b32 $r12 15
 	add b32 $r11 $r12	// base = NV_PGRAPH_GPCn_TPC0
@@ -364,14 +338,13 @@ ctx_xfer:
 	ld b32 $r13 D[$r0 + #tpc_mmio_list_tail]
 	ld b32 $r15 D[$r0 + #tpc_mask]
 	mov $r14 0x800		// stride = 0x800
-	call #mmctx_xfer
+	call(mmctx_xfer)
 
 #if NV_PGRAPH_GPCX_UNK__SIZE > 0
 	// per-UNK mmio context
 	xbit $r10 $flags $p1	// direction
 	or $r10 4		// last
-	mov $r11 0x3000
-	sethi $r11 0x500000	// base = NV_PGRAPH_GPC0_UNK0
+	imm32($r11, 0x503000)
 	ld b32 $r12 D[$r0 + #gpc_id]
 	shl b32 $r12 15
 	add b32 $r11 $r12	// base = NV_PGRAPH_GPCn_UNK0
@@ -379,11 +352,11 @@ ctx_xfer:
 	ld b32 $r13 D[$r0 + #unk_mmio_list_tail]
 	ld b32 $r15 D[$r0 + #unk_mask]
 	mov $r14 0x200		// stride = 0x200
-	call #mmctx_xfer
+	call(mmctx_xfer)
 #endif
 
 	// wait for strands to finish
-	call #strand_wait
+	call(strand_wait)
 
 	// if load, or a save without a load following, do some
 	// unknown stuff that's done after finishing a block of
@@ -391,14 +364,10 @@ ctx_xfer:
 	bra $p1 #ctx_xfer_post
 	bra not $p2 #ctx_xfer_done
 	ctx_xfer_post:
-		mov $r1 0x4afc
-		sethi $r1 0x20000
-		mov $r2 0xd
-		iowr I[$r1] $r2		// STRAND_CMD(0x3f) = 0x0d
-		call #strand_wait
+		call(strand_post)
 
 	// mark completion in HUB's barrier
 	ctx_xfer_done:
-	call #hub_barrier_done
+	call(hub_barrier_done)
 	ret
 #endif

drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcnvc0.fuc.h

@@ -37,14 +37,14 @@ uint32_t nvc0_grgpc_data[] = {
 };
 
 uint32_t nvc0_grgpc_code[] = {
-	0x03180ef5,
+	0x03a10ef5,
 /* 0x0004: queue_put */
 	0x9800d898,
 	0x86f001d9,
 	0x0489b808,
 	0xf00c1bf4,
 	0x21f502f7,
-	0x00f802fe,
+	0x00f8037e,
 /* 0x001c: queue_put_next */
 	0xb60798c4,
 	0x8dbb0384,
@@ -68,184 +68,214 @@ uint32_t nvc0_grgpc_code[] = {
 /* 0x0066: queue_get_done */
 	0x00f80132,
 /* 0x0068: nv_rd32 */
-	0x0728b7f1,
-	0xb906b4b6,
-	0xc9f002ec,
-	0x00bcd01f,
-/* 0x0078: nv_rd32_wait */
-	0xc800bccf,
-	0x1bf41fcc,
-	0x06a7f0fa,
-	0x010921f5,
-	0xf840bfcf,
-/* 0x008d: nv_wr32 */
-	0x28b7f100,
-	0x06b4b607,
-	0xb980bfd0,
-	0xc9f002ec,
-	0x1ec9f01f,
-/* 0x00a3: nv_wr32_wait */
-	0xcf00bcd0,
-	0xccc800bc,
-	0xfa1bf41f,
-/* 0x00ae: watchdog_reset */
-	0x87f100f8,
-	0x84b60430,
-	0x1ff9f006,
-	0xf8008fd0,
-/* 0x00bd: watchdog_clear */
-	0x3087f100,
-	0x0684b604,
-	0xf80080d0,
-/* 0x00c9: wait_donez */
-	0xf094bd00,
-	0x07f10099,
-	0x03f00f00,
-	0x0009d002,
-	0x07f104bd,
-	0x03f00600,
-	0x000ad002,
-/* 0x00e6: wait_donez_ne */
-	0x87f104bd,
-	0x83f00000,
-	0x0088cf01,
-	0xf4888aff,
-	0x94bdf31b,
-	0xf10099f0,
-	0xf0170007,
-	0x09d00203,
-	0xf804bd00,
-/* 0x0109: wait_doneo */
-	0xf094bd00,
-	0x07f10099,
-	0x03f00f00,
-	0x0009d002,
-	0x87f104bd,
-	0x84b60818,
-	0x008ad006,
-/* 0x0124: wait_doneo_e */
-	0x040087f1,
-	0xcf0684b6,
-	0x8aff0088,
-	0xf30bf488,
+	0xf002ecb9,
+	0x07f11fc9,
+	0x03f0ca00,
+	0x000cd001,
+/* 0x007a: nv_rd32_wait */
+	0xc7f104bd,
+	0xc3f0ca00,
+	0x00cccf01,
+	0xf41fccc8,
+	0xa7f0f31b,
+	0x1021f506,
+	0x00f7f101,
+	0x01f3f0cb,
+	0xf800ffcf,
+/* 0x009d: nv_wr32 */
+	0x0007f100,
+	0x0103f0cc,
+	0xbd000fd0,
+	0x02ecb904,
+	0xf01fc9f0,
+	0x07f11ec9,
+	0x03f0ca00,
+	0x000cd001,
+/* 0x00be: nv_wr32_wait */
+	0xc7f104bd,
+	0xc3f0ca00,
+	0x00cccf01,
+	0xf41fccc8,
+	0x00f8f31b,
+/* 0x00d0: wait_donez */
 	0x99f094bd,
 	0x0007f100,
-	0x0203f017,
+	0x0203f00f,
 	0xbd0009d0,
-/* 0x0147: mmctx_size */
-	0xbd00f804,
-/* 0x0149: nv_mmctx_size_loop */
-	0x00e89894,
-	0xb61a85b6,
-	0x84b60180,
-	0x0098bb02,
-	0xb804e0b6,
-	0x1bf404ef,
-	0x029fb9eb,
-/* 0x0166: mmctx_xfer */
-	0x94bd00f8,
-	0xf10199f0,
-	0xf00f0007,
-	0x09d00203,
-	0xf104bd00,
-	0xb6071087,
-	0x94bd0684,
-	0xf405bbfd,
-	0x8bd0090b,
-	0x0099f000,
-/* 0x018c: mmctx_base_disabled */
-	0xf405eefd,
-	0x8ed00c0b,
-	0xc08fd080,
-/* 0x019b: mmctx_multi_disabled */
-	0xb70199f0,
-	0xc8010080,
+	0x0007f104,
+	0x0203f006,
+	0xbd000ad0,
+/* 0x00ed: wait_donez_ne */
+	0x0087f104,
+	0x0183f000,
+	0xff0088cf,
+	0x1bf4888a,
+	0xf094bdf3,
+	0x07f10099,
+	0x03f01700,
+	0x0009d002,
+	0x00f804bd,
+/* 0x0110: wait_doneo */
+	0x99f094bd,
+	0x0007f100,
+	0x0203f00f,
+	0xbd0009d0,
+	0x0007f104,
+	0x0203f006,
+	0xbd000ad0,
+/* 0x012d: wait_doneo_e */
+	0x0087f104,
+	0x0183f000,
+	0xff0088cf,
+	0x0bf4888a,
+	0xf094bdf3,
+	0x07f10099,
+	0x03f01700,
+	0x0009d002,
+	0x00f804bd,
+/* 0x0150: mmctx_size */
+/* 0x0152: nv_mmctx_size_loop */
+	0xe89894bd,
+	0x1a85b600,
+	0xb60180b6,
+	0x98bb0284,
+	0x04e0b600,
+	0xf404efb8,
+	0x9fb9eb1b,
+/* 0x016f: mmctx_xfer */
+	0xbd00f802,
+	0x0199f094,
+	0x0f0007f1,
+	0xd00203f0,
+	0x04bd0009,
+	0xbbfd94bd,
+	0x120bf405,
+	0xc40007f1,
+	0xd00103f0,
+	0x04bd000b,
+/* 0x0197: mmctx_base_disabled */
+	0xfd0099f0,
+	0x0bf405ee,
+	0x0007f11e,
+	0x0103f0c6,
+	0xbd000ed0,
+	0x0007f104,
+	0x0103f0c7,
+	0xbd000fd0,
+	0x0199f004,
+/* 0x01b8: mmctx_multi_disabled */
+	0xb600abc8,
+	0xb9f010b4,
+	0x01aec80c,
+	0xfd11e4b6,
+	0x07f105be,
+	0x03f0c500,
+	0x000bd001,
+/* 0x01d6: mmctx_exec_loop */
+/* 0x01d6: mmctx_wait_free */
+	0xe7f104bd,
+	0xe3f0c500,
+	0x00eecf01,
+	0xf41fe4f0,
+	0xce98f30b,
+	0x05e9fd00,
+	0xc80007f1,
+	0xd00103f0,
+	0x04bd000e,
+	0xb804c0b6,
+	0x1bf404cd,
+	0x02abc8d8,
+/* 0x0207: mmctx_fini_wait */
+	0xf11f1bf4,
+	0xf0c500b7,
+	0xbbcf01b3,
+	0x1fb4f000,
+	0xf410b4b0,
+	0xa7f0f01b,
+	0xd021f402,
+/* 0x0223: mmctx_stop */
+	0xc82b0ef4,
 	0xb4b600ab,
 	0x0cb9f010,
-	0xb601aec8,
-	0xbefd11e4,
-	0x008bd005,
-/* 0x01b4: mmctx_exec_loop */
-/* 0x01b4: mmctx_wait_free */
-	0xf0008ecf,
-	0x0bf41fe4,
-	0x00ce98fa,
-	0xd005e9fd,
-	0xc0b6c08e,
-	0x04cdb804,
-	0xc8e81bf4,
-	0x1bf402ab,
-/* 0x01d5: mmctx_fini_wait */
-	0x008bcf18,
-	0xb01fb4f0,
-	0x1bf410b4,
-	0x02a7f0f7,
-	0xf4c921f4,
-/* 0x01ea: mmctx_stop */
-	0xabc81b0e,
-	0x10b4b600,
-	0xf00cb9f0,
-	0x8bd012b9,
-/* 0x01f9: mmctx_stop_wait */
-	0x008bcf00,
-	0xf412bbc8,
-/* 0x0202: mmctx_done */
-	0x94bdfa1b,
-	0xf10199f0,
-	0xf0170007,
-	0x09d00203,
-	0xf804bd00,
-/* 0x0215: strand_wait */
-	0xf0a0f900,
-	0x21f402a7,
-	0xf8a0fcc9,
-/* 0x0221: strand_pre */
-	0xfc87f100,
-	0x0283f04a,
-	0xd00c97f0,
-	0x21f50089,
-	0x00f80215,
-/* 0x0234: strand_post */
-	0x4afc87f1,
-	0xf00283f0,
-	0x89d00d97,
-	0x1521f500,
-/* 0x0247: strand_set */
-	0xf100f802,
-	0xf04ffca7,
-	0xaba202a3,
-	0xc7f00500,
-	0x00acd00f,
-	0xd00bc7f0,
-	0x21f500bc,
-	0xaed00215,
-	0x0ac7f000,
-	0xf500bcd0,
-	0xf8021521,
-/* 0x0271: strand_ctx_init */
-	0xf094bd00,
-	0x07f10399,
-	0x03f00f00,
+	0xf112b9f0,
+	0xf0c50007,
+	0x0bd00103,
+/* 0x023b: mmctx_stop_wait */
+	0xf104bd00,
+	0xf0c500b7,
+	0xbbcf01b3,
+	0x12bbc800,
+/* 0x024b: mmctx_done */
+	0xbdf31bf4,
+	0x0199f094,
+	0x170007f1,
+	0xd00203f0,
+	0x04bd0009,
+/* 0x025e: strand_wait */
+	0xa0f900f8,
+	0xf402a7f0,
+	0xa0fcd021,
+/* 0x026a: strand_pre */
+	0x97f000f8,
+	0xfc07f10c,
+	0x0203f04a,
+	0xbd0009d0,
+	0x5e21f504,
+/* 0x027f: strand_post */
+	0xf000f802,
+	0x07f10d97,
+	0x03f04afc,
 	0x0009d002,
 	0x21f504bd,
-	0xe7f00221,
-	0x4721f503,
-	0xfca7f102,
-	0x02a3f046,
-	0x0400aba0,
-	0xf040a0d0,
-	0xbcd001c7,
-	0x1521f500,
-	0x010c9202,
-	0xf000acd0,
-	0xbcd002c7,
-	0x1521f500,
-	0x3421f502,
-	0x8087f102,
-	0x0684b608,
-	0xb70089cf,
-	0x95220080,
-/* 0x02ca: ctx_init_strand_loop */
+	0x00f8025e,
+/* 0x0294: strand_set */
+	0xf10fc7f0,
+	0xf04ffc07,
+	0x0cd00203,
+	0xf004bd00,
+	0x07f10bc7,
+	0x03f04afc,
+	0x000cd002,
+	0x07f104bd,
+	0x03f04ffc,
+	0x000ed002,
+	0xc7f004bd,
+	0xfc07f10a,
+	0x0203f04a,
+	0xbd000cd0,
+	0x5e21f504,
+/* 0x02d3: strand_ctx_init */
+	0xbd00f802,
+	0x0399f094,
+	0x0f0007f1,
+	0xd00203f0,
+	0x04bd0009,
+	0x026a21f5,
+	0xf503e7f0,
+	0xbd029421,
+	0xfc07f1c4,
+	0x0203f047,
+	0xbd000cd0,
+	0x01c7f004,
+	0x4afc07f1,
+	0xd00203f0,
+	0x04bd000c,
+	0x025e21f5,
+	0xf1010c92,
+	0xf046fc07,
+	0x0cd00203,
+	0xf004bd00,
+	0x07f102c7,
+	0x03f04afc,
+	0x000cd002,
+	0x21f504bd,
+	0x21f5025e,
+	0x87f1027f,
+	0x83f04200,
+	0x0097f102,
+	0x0293f020,
+	0x950099cf,
+/* 0x034a: ctx_init_strand_loop */
 	0x8ed008fe,
 	0x408ed000,
 	0xb6808acf,
@@ -259,167 +289,199 @@ uint32_t nvc0_grgpc_code[] = {
 	0x170007f1,
 	0xd00203f0,
 	0x04bd0009,
-/* 0x02fe: error */
+/* 0x037e: error */
 	0xe0f900f8,
-	0x9814e7f1,
-	0xf440e3f0,
-	0xe0b78d21,
-	0xf7f0041c,
-	0x8d21f401,
-	0x00f8e0fc,
-/* 0x0318: init */
-	0x04fe04bd,
-	0x0017f100,
-	0x0227f012,
-	0xf10012d0,
-	0xfe042617,
-	0x17f10010,
-	0x10d00400,
-	0x0427f0c0,
-	0xf40012d0,
-	0x17f11031,
-	0x14b60608,
-	0x0012cf06,
+	0xf102ffb9,
+	0xf09814e7,
+	0x21f440e3,
+	0x01f7f09d,
+	0xf102ffb9,
+	0xf09c1ce7,
+	0x21f440e3,
+	0xf8e0fc9d,
+/* 0x03a1: init */
+	0xfe04bd00,
+	0x27f00004,
+	0x0007f102,
+	0x0003f012,
+	0xbd0002d0,
+	0xd517f104,
+	0x0010fe04,
+	0x070007f1,
+	0xd00003f0,
+	0x04bd0000,
+	0xf10427f0,
+	0xf0040007,
+	0x02d00003,
+	0xf404bd00,
+	0x27f11031,
+	0x23f08200,
+	0x0022cf01,
 	0xf00137f0,
 	0x32bb1f24,
 	0x0132b604,
 	0x80050280,
-	0x10b70603,
-	0x12cf0400,
-	0x04028000,
-	0x010027f1,
-	0xcf0223f0,
-	0x34bd0022,
-	0x070047f1,
-	0x950644b6,
-	0x45d00825,
-	0x4045d000,
-	0x98000e98,
-	0x21f5010f,
-	0x2fbb0147,
-	0x003fbb00,
-	0x98010e98,
-	0x21f5020f,
-	0x0e980147,
-	0x00effd05,
-	0xbb002ebb,
-	0x40b7003e,
-	0x35b61300,
-	0x0043d002,
-	0xb60825b6,
-	0x20b60635,
-	0x0130b601,
-	0xb60824b6,
-	0x2fb90834,
-	0x7121f502,
-	0x003fbb02,
-	0x010007f1,
+	0x27f10603,
+	0x23f08600,
+	0x0022cf01,
+	0xf1040280,
+	0xf0010027,
+	0x22cf0223,
+	0x9534bd00,
+	0x07f10825,
+	0x03f0c000,
+	0x0005d001,
+	0x07f104bd,
+	0x03f0c100,
+	0x0005d001,
+	0x0e9804bd,
+	0x010f9800,
+	0x015021f5,
+	0xbb002fbb,
+	0x0e98003f,
+	0x020f9801,
+	0x015021f5,
+	0xfd050e98,
+	0x2ebb00ef,
+	0x003ebb00,
+	0xf10235b6,
+	0xf0d30007,
+	0x03d00103,
+	0xb604bd00,
+	0x35b60825,
+	0x0120b606,
+	0xb60130b6,
+	0x34b60824,
+	0x022fb908,
+	0x02d321f5,
+	0xf1003fbb,
+	0xf0010007,
+	0x03d00203,
+	0xbd04bd00,
+	0x1f29f024,
+	0x080007f1,
 	0xd00203f0,
-	0x04bd0003,
-	0x29f024bd,
-	0x0007f11f,
-	0x0203f008,
-	0xbd0002d0,
-/* 0x03e9: main */
-	0x0031f404,
-	0xf00028f4,
-	0x21f41cd7,
-	0xf401f439,
-	0xf404e4b0,
-	0x81fe1e18,
-	0x0627f001,
-	0x12fd20bd,
-	0x01e4b604,
-	0xfe051efd,
-	0x21f50018,
-	0x0ef404ad,
-/* 0x0419: main_not_ctx_xfer */
-	0x10ef94d3,
-	0xf501f5f0,
-	0xf402fe21,
-/* 0x0426: ih */
-	0x80f9c60e,
-	0xf90188fe,
-	0xf990f980,
-	0xf9b0f9a0,
-	0xf9e0f9d0,
-	0xcf04bdf0,
-	0xabc4800a,
-	0x1d0bf404,
-	0x1900b7f1,
-	0xcf1cd7f0,
-	0xbfcf40be,
+	0x04bd0002,
+/* 0x0498: main */
+	0xf40031f4,
+	0xd7f00028,
+	0x3921f41c,
+	0xb0f401f4,
+	0x18f404e4,
+	0x0181fe1e,
+	0xbd0627f0,
+	0x0412fd20,
+	0xfd01e4b6,
+	0x18fe051e,
+	0x8d21f500,
+	0xd30ef405,
+/* 0x04c8: main_not_ctx_xfer */
+	0xf010ef94,
+	0x21f501f5,
+	0x0ef4037e,
+/* 0x04d5: ih */
+	0xfe80f9c6,
+	0x80f90188,
+	0xa0f990f9,
+	0xd0f9b0f9,
+	0xf0f9e0f9,
+	0xa7f104bd,
+	0xa3f00200,
+	0x00aacf00,
+	0xf404abc4,
+	0xd7f02c0b,
+	0x00e7f11c,
+	0x00e3f01a,
+	0xf100eecf,
+	0xf01900f7,
+	0xffcf00f3,
 	0x0421f400,
-	0x0400b0b7,
-	0xd001e7f0,
-/* 0x045e: ih_no_fifo */
-	0x0ad000be,
-	0xfcf0fc40,
-	0xfcd0fce0,
-	0xfca0fcb0,
-	0xfe80fc90,
-	0x80fc0088,
-	0xf80032f4,
-/* 0x0479: hub_barrier_done */
-	0x01f7f001,
-	0xbb040e98,
-	0xe7f104fe,
-	0xe3f09418,
-	0x8d21f440,
-/* 0x048e: ctx_redswitch */
-	0xe7f100f8,
-	0xe4b60614,
-	0x20f7f006,
-	0xf000efd0,
-/* 0x049e: ctx_redswitch_delay */
-	0xf2b608f7,
-	0xfd1bf401,
-	0x0a20f7f1,
-	0xf800efd0,
-/* 0x04ad: ctx_xfer */
-	0x0417f100,
-	0x0614b60a,
-	0xf4001fd0,
-	0x21f50711,
-/* 0x04be: ctx_xfer_not_load */
-	0x17f1048e,
-	0x13f04afc,
-	0x0c27f002,
-	0xf50012d0,
-	0xf1021521,
-	0xf047fc27,
-	0x20d00223,
-	0x012cf000,
-	0xd00320b6,
-	0xacf00012,
-	0x02a5f001,
-	0xf000b7f0,
-	0x0c9850b3,
-	0x0fc4b604,
-	0x9800bcbb,
-	0x0d98000c,
-	0x00e7f001,
-	0x016621f5,
+	0xf101e7f0,
+	0xf01d0007,
+	0x0ed00003,
+/* 0x0523: ih_no_fifo */
+	0xf104bd00,
+	0xf0010007,
+	0x0ad00003,
+	0xfc04bd00,
+	0xfce0fcf0,
+	0xfcb0fcd0,
+	0xfc90fca0,
+	0x0088fe80,
+	0x32f480fc,
+/* 0x0547: hub_barrier_done */
+	0xf001f800,
+	0x0e9801f7,
+	0x04febb04,
+	0xf102ffb9,
+	0xf09418e7,
+	0x21f440e3,
+/* 0x055f: ctx_redswitch */
+	0xf000f89d,
+	0x07f120f7,
+	0x03f08500,
+	0x000fd001,
+	0xe7f004bd,
+/* 0x0571: ctx_redswitch_delay */
+	0x01e2b608,
+	0xf1fd1bf4,
+	0xf10800f5,
+	0xf10200f5,
+	0xf0850007,
+	0x0fd00103,
+	0xf804bd00,
+/* 0x058d: ctx_xfer */
+	0x0007f100,
+	0x0203f081,
+	0xbd000fd0,
+	0x0711f404,
+	0x055f21f5,
+/* 0x05a0: ctx_xfer_not_load */
+	0x026a21f5,
+	0x07f124bd,
+	0x03f047fc,
+	0x0002d002,
+	0x2cf004bd,
+	0x0320b601,
+	0x4afc07f1,
+	0xd00203f0,
+	0x04bd0002,
 	0xf001acf0,
-	0xb7f104a5,
-	0xb3f04000,
+	0xb7f102a5,
+	0xb3f00000,
 	0x040c9850,
 	0xbb0fc4b6,
 	0x0c9800bc,
-	0x020d9801,
-	0xf1060f98,
-	0xf50800e7,
-	0xf5016621,
-	0xf4021521,
-	0x12f40601,
-/* 0x0535: ctx_xfer_post */
-	0xfc17f114,
-	0x0213f04a,
-	0xd00d27f0,
-	0x21f50012,
-/* 0x0546: ctx_xfer_done */
-	0x21f50215,
-	0x00f80479,
+	0x010d9800,
+	0xf500e7f0,
+	0xf0016f21,
+	0xa5f001ac,
+	0x00b7f104,
+	0x50b3f040,
+	0xb6040c98,
+	0xbcbb0fc4,
+	0x010c9800,
+	0x98020d98,
+	0xe7f1060f,
+	0x21f50800,
+	0x21f5016f,
+	0x01f4025e,
+	0x0712f406,
+/* 0x0618: ctx_xfer_post */
+	0x027f21f5,
+/* 0x061c: ctx_xfer_done */
+	0x054721f5,
+	0x000000f8,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
 	0x00000000,
 	0x00000000,
 	0x00000000,

drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcnvd7.fuc.h

@@ -41,14 +41,14 @@ uint32_t nvd7_grgpc_data[] = {
 };
 
 uint32_t nvd7_grgpc_code[] = {
-	0x03180ef5,
+	0x03a10ef5,
 /* 0x0004: queue_put */
 	0x9800d898,
 	0x86f001d9,
 	0x0489b808,
 	0xf00c1bf4,
 	0x21f502f7,
-	0x00f802fe,
+	0x00f8037e,
 /* 0x001c: queue_put_next */
 	0xb60798c4,
 	0x8dbb0384,
@@ -72,184 +72,214 @@ uint32_t nvd7_grgpc_code[] = {
 /* 0x0066: queue_get_done */
 	0x00f80132,
 /* 0x0068: nv_rd32 */
-	0x0728b7f1,
-	0xb906b4b6,
-	0xc9f002ec,
-	0x00bcd01f,
-/* 0x0078: nv_rd32_wait */
-	0xc800bccf,
-	0x1bf41fcc,
-	0x06a7f0fa,
-	0x010921f5,
-	0xf840bfcf,
-/* 0x008d: nv_wr32 */
-	0x28b7f100,
-	0x06b4b607,
-	0xb980bfd0,
-	0xc9f002ec,
-	0x1ec9f01f,
-/* 0x00a3: nv_wr32_wait */
-	0xcf00bcd0,
-	0xccc800bc,
-	0xfa1bf41f,
-/* 0x00ae: watchdog_reset */
-	0x87f100f8,
-	0x84b60430,
-	0x1ff9f006,
-	0xf8008fd0,
-/* 0x00bd: watchdog_clear */
-	0x3087f100,
-	0x0684b604,
-	0xf80080d0,
-/* 0x00c9: wait_donez */
-	0xf094bd00,
-	0x07f10099,
-	0x03f00f00,
-	0x0009d002,
-	0x07f104bd,
-	0x03f00600,
-	0x000ad002,
-/* 0x00e6: wait_donez_ne */
-	0x87f104bd,
-	0x83f00000,
-	0x0088cf01,
-	0xf4888aff,
-	0x94bdf31b,
-	0xf10099f0,
-	0xf0170007,
-	0x09d00203,
-	0xf804bd00,
-/* 0x0109: wait_doneo */
-	0xf094bd00,
-	0x07f10099,
-	0x03f00f00,
-	0x0009d002,
-	0x87f104bd,
-	0x84b60818,
-	0x008ad006,
-/* 0x0124: wait_doneo_e */
-	0x040087f1,
-	0xcf0684b6,
-	0x8aff0088,
-	0xf30bf488,
+	0xf002ecb9,
+	0x07f11fc9,
+	0x03f0ca00,
+	0x000cd001,
+/* 0x007a: nv_rd32_wait */
+	0xc7f104bd,
+	0xc3f0ca00,
+	0x00cccf01,
+	0xf41fccc8,
+	0xa7f0f31b,
+	0x1021f506,
+	0x00f7f101,
+	0x01f3f0cb,
+	0xf800ffcf,
+/* 0x009d: nv_wr32 */
+	0x0007f100,
+	0x0103f0cc,
+	0xbd000fd0,
+	0x02ecb904,
+	0xf01fc9f0,
+	0x07f11ec9,
+	0x03f0ca00,
+	0x000cd001,
+/* 0x00be: nv_wr32_wait */
+	0xc7f104bd,
+	0xc3f0ca00,
+	0x00cccf01,
+	0xf41fccc8,
+	0x00f8f31b,
+/* 0x00d0: wait_donez */
 	0x99f094bd,
 	0x0007f100,
-	0x0203f017,
+	0x0203f00f,
 	0xbd0009d0,
-/* 0x0147: mmctx_size */
-	0xbd00f804,
-/* 0x0149: nv_mmctx_size_loop */
-	0x00e89894,
-	0xb61a85b6,
-	0x84b60180,
-	0x0098bb02,
-	0xb804e0b6,
-	0x1bf404ef,
-	0x029fb9eb,
-/* 0x0166: mmctx_xfer */
-	0x94bd00f8,
-	0xf10199f0,
-	0xf00f0007,
-	0x09d00203,
-	0xf104bd00,
-	0xb6071087,
-	0x94bd0684,
-	0xf405bbfd,
-	0x8bd0090b,
-	0x0099f000,
-/* 0x018c: mmctx_base_disabled */
-	0xf405eefd,
-	0x8ed00c0b,
-	0xc08fd080,
-/* 0x019b: mmctx_multi_disabled */
-	0xb70199f0,
-	0xc8010080,
+	0x0007f104,
+	0x0203f006,
+	0xbd000ad0,
+/* 0x00ed: wait_donez_ne */
+	0x0087f104,
+	0x0183f000,
+	0xff0088cf,
+	0x1bf4888a,
+	0xf094bdf3,
+	0x07f10099,
+	0x03f01700,
+	0x0009d002,
+	0x00f804bd,
+/* 0x0110: wait_doneo */
+	0x99f094bd,
+	0x0007f100,
+	0x0203f00f,
+	0xbd0009d0,
+	0x0007f104,
+	0x0203f006,
+	0xbd000ad0,
+/* 0x012d: wait_doneo_e */
+	0x0087f104,
+	0x0183f000,
+	0xff0088cf,
+	0x0bf4888a,
+	0xf094bdf3,
+	0x07f10099,
+	0x03f01700,
+	0x0009d002,
+	0x00f804bd,
+/* 0x0150: mmctx_size */
+/* 0x0152: nv_mmctx_size_loop */
+	0xe89894bd,
+	0x1a85b600,
+	0xb60180b6,
+	0x98bb0284,
+	0x04e0b600,
+	0xf404efb8,
+	0x9fb9eb1b,
+/* 0x016f: mmctx_xfer */
+	0xbd00f802,
+	0x0199f094,
+	0x0f0007f1,
+	0xd00203f0,
+	0x04bd0009,
+	0xbbfd94bd,
+	0x120bf405,
+	0xc40007f1,
+	0xd00103f0,
+	0x04bd000b,
+/* 0x0197: mmctx_base_disabled */
+	0xfd0099f0,
+	0x0bf405ee,
+	0x0007f11e,
+	0x0103f0c6,
+	0xbd000ed0,
+	0x0007f104,
+	0x0103f0c7,
+	0xbd000fd0,
+	0x0199f004,
+/* 0x01b8: mmctx_multi_disabled */
+	0xb600abc8,
+	0xb9f010b4,
+	0x01aec80c,
+	0xfd11e4b6,
+	0x07f105be,
+	0x03f0c500,
+	0x000bd001,
+/* 0x01d6: mmctx_exec_loop */
+/* 0x01d6: mmctx_wait_free */
+	0xe7f104bd,
+	0xe3f0c500,
+	0x00eecf01,
+	0xf41fe4f0,
+	0xce98f30b,
+	0x05e9fd00,
+	0xc80007f1,
+	0xd00103f0,
+	0x04bd000e,
+	0xb804c0b6,
+	0x1bf404cd,
+	0x02abc8d8,
+/* 0x0207: mmctx_fini_wait */
+	0xf11f1bf4,
+	0xf0c500b7,
+	0xbbcf01b3,
+	0x1fb4f000,
+	0xf410b4b0,
+	0xa7f0f01b,
+	0xd021f402,
+/* 0x0223: mmctx_stop */
+	0xc82b0ef4,
 	0xb4b600ab,
 	0x0cb9f010,
-	0xb601aec8,
-	0xbefd11e4,
-	0x008bd005,
-/* 0x01b4: mmctx_exec_loop */
-/* 0x01b4: mmctx_wait_free */
-	0xf0008ecf,
-	0x0bf41fe4,
-	0x00ce98fa,
-	0xd005e9fd,
-	0xc0b6c08e,
-	0x04cdb804,
-	0xc8e81bf4,
-	0x1bf402ab,
-/* 0x01d5: mmctx_fini_wait */
-	0x008bcf18,
-	0xb01fb4f0,
-	0x1bf410b4,
-	0x02a7f0f7,
-	0xf4c921f4,
-/* 0x01ea: mmctx_stop */
-	0xabc81b0e,
-	0x10b4b600,
-	0xf00cb9f0,
-	0x8bd012b9,
-/* 0x01f9: mmctx_stop_wait */
-	0x008bcf00,
-	0xf412bbc8,
-/* 0x0202: mmctx_done */
-	0x94bdfa1b,
-	0xf10199f0,
-	0xf0170007,
-	0x09d00203,
-	0xf804bd00,
-/* 0x0215: strand_wait */
-	0xf0a0f900,
-	0x21f402a7,
-	0xf8a0fcc9,
-/* 0x0221: strand_pre */
-	0xfc87f100,
-	0x0283f04a,
-	0xd00c97f0,
-	0x21f50089,
-	0x00f80215,
-/* 0x0234: strand_post */
-	0x4afc87f1,
-	0xf00283f0,
-	0x89d00d97,
-	0x1521f500,
-/* 0x0247: strand_set */
-	0xf100f802,
-	0xf04ffca7,
-	0xaba202a3,
-	0xc7f00500,
-	0x00acd00f,
-	0xd00bc7f0,
-	0x21f500bc,
-	0xaed00215,
-	0x0ac7f000,
-	0xf500bcd0,
-	0xf8021521,
-/* 0x0271: strand_ctx_init */
-	0xf094bd00,
-	0x07f10399,
-	0x03f00f00,
+	0xf112b9f0,
+	0xf0c50007,
+	0x0bd00103,
+/* 0x023b: mmctx_stop_wait */
+	0xf104bd00,
+	0xf0c500b7,
+	0xbbcf01b3,
+	0x12bbc800,
+/* 0x024b: mmctx_done */
+	0xbdf31bf4,
+	0x0199f094,
+	0x170007f1,
+	0xd00203f0,
+	0x04bd0009,
+/* 0x025e: strand_wait */
+	0xa0f900f8,
+	0xf402a7f0,
+	0xa0fcd021,
+/* 0x026a: strand_pre */
+	0x97f000f8,
+	0xfc07f10c,
+	0x0203f04a,
+	0xbd0009d0,
+	0x5e21f504,
+/* 0x027f: strand_post */
+	0xf000f802,
+	0x07f10d97,
+	0x03f04afc,
 	0x0009d002,
 	0x21f504bd,
-	0xe7f00221,
-	0x4721f503,
-	0xfca7f102,
-	0x02a3f046,
-	0x0400aba0,
-	0xf040a0d0,
-	0xbcd001c7,
-	0x1521f500,
-	0x010c9202,
-	0xf000acd0,
-	0xbcd002c7,
-	0x1521f500,
-	0x3421f502,
-	0x8087f102,
-	0x0684b608,
-	0xb70089cf,
-	0x95220080,
-/* 0x02ca: ctx_init_strand_loop */
+	0x00f8025e,
+/* 0x0294: strand_set */
+	0xf10fc7f0,
+	0xf04ffc07,
+	0x0cd00203,
+	0xf004bd00,
+	0x07f10bc7,
+	0x03f04afc,
+	0x000cd002,
+	0x07f104bd,
+	0x03f04ffc,
+	0x000ed002,
+	0xc7f004bd,
+	0xfc07f10a,
+	0x0203f04a,
+	0xbd000cd0,
+	0x5e21f504,
+/* 0x02d3: strand_ctx_init */
+	0xbd00f802,
+	0x0399f094,
+	0x0f0007f1,
+	0xd00203f0,
+	0x04bd0009,
+	0x026a21f5,
+	0xf503e7f0,
+	0xbd029421,
+	0xfc07f1c4,
+	0x0203f047,
+	0xbd000cd0,
+	0x01c7f004,
+	0x4afc07f1,
+	0xd00203f0,
+	0x04bd000c,
+	0x025e21f5,
+	0xf1010c92,
+	0xf046fc07,
+	0x0cd00203,
+	0xf004bd00,
+	0x07f102c7,
+	0x03f04afc,
+	0x000cd002,
+	0x21f504bd,
+	0x21f5025e,
+	0x87f1027f,
+	0x83f04200,
+	0x0097f102,
+	0x0293f020,
+	0x950099cf,
+/* 0x034a: ctx_init_strand_loop */
 	0x8ed008fe,
 	0x408ed000,
 	0xb6808acf,
@@ -263,198 +293,230 @@ uint32_t nvd7_grgpc_code[] = {
 	0x170007f1,
 	0xd00203f0,
 	0x04bd0009,
-/* 0x02fe: error */
+/* 0x037e: error */
 	0xe0f900f8,
-	0x9814e7f1,
-	0xf440e3f0,
-	0xe0b78d21,
-	0xf7f0041c,
-	0x8d21f401,
-	0x00f8e0fc,
-/* 0x0318: init */
-	0x04fe04bd,
-	0x0017f100,
-	0x0227f012,
-	0xf10012d0,
-	0xfe047017,
-	0x17f10010,
-	0x10d00400,
-	0x0427f0c0,
-	0xf40012d0,
-	0x17f11031,
-	0x14b60608,
-	0x0012cf06,
+	0xf102ffb9,
+	0xf09814e7,
+	0x21f440e3,
+	0x01f7f09d,
+	0xf102ffb9,
+	0xf09c1ce7,
+	0x21f440e3,
+	0xf8e0fc9d,
+/* 0x03a1: init */
+	0xfe04bd00,
+	0x27f00004,
+	0x0007f102,
+	0x0003f012,
+	0xbd0002d0,
+	0x1f17f104,
+	0x0010fe05,
+	0x070007f1,
+	0xd00003f0,
+	0x04bd0000,
+	0xf10427f0,
+	0xf0040007,
+	0x02d00003,
+	0xf404bd00,
+	0x27f11031,
+	0x23f08200,
+	0x0022cf01,
 	0xf00137f0,
 	0x32bb1f24,
 	0x0132b604,
 	0x80050280,
-	0x10b70603,
-	0x12cf0400,
-	0x04028000,
-	0x0c30e7f1,
-	0xbd50e3f0,
-	0xbd34bd24,
-/* 0x0371: init_unk_loop */
-	0x6821f444,
-	0xf400f6b0,
-	0xf7f00f0b,
-	0x04f2bb01,
-	0xb6054ffd,
-/* 0x0386: init_unk_next */
-	0x20b60130,
-	0x04e0b601,
-	0xf40126b0,
-/* 0x0392: init_unk_done */
-	0x0380e21b,
-	0x08048007,
-	0x010027f1,
-	0xcf0223f0,
-	0x34bd0022,
-	0x070047f1,
-	0x950644b6,
-	0x45d00825,
-	0x4045d000,
-	0x98000e98,
-	0x21f5010f,
-	0x2fbb0147,
-	0x003fbb00,
-	0x98010e98,
-	0x21f5020f,
-	0x0e980147,
-	0x00effd05,
-	0xbb002ebb,
-	0x0e98003e,
-	0x030f9802,
-	0x014721f5,
-	0xfd070e98,
+	0x27f10603,
+	0x23f08600,
+	0x0022cf01,
+	0xf1040280,
+	0xf00c30e7,
+	0x24bd50e3,
+	0x44bd34bd,
+/* 0x0410: init_unk_loop */
+	0xb06821f4,
+	0x0bf400f6,
+	0x01f7f00f,
+	0xfd04f2bb,
+	0x30b6054f,
+/* 0x0425: init_unk_next */
+	0x0120b601,
+	0xb004e0b6,
+	0x1bf40126,
+/* 0x0431: init_unk_done */
+	0x070380e2,
+	0xf1080480,
+	0xf0010027,
+	0x22cf0223,
+	0x9534bd00,
+	0x07f10825,
+	0x03f0c000,
+	0x0005d001,
+	0x07f104bd,
+	0x03f0c100,
+	0x0005d001,
+	0x0e9804bd,
+	0x010f9800,
+	0x015021f5,
+	0xbb002fbb,
+	0x0e98003f,
+	0x020f9801,
+	0x015021f5,
+	0xfd050e98,
 	0x2ebb00ef,
 	0x003ebb00,
-	0x130040b7,
-	0xd00235b6,
-	0x25b60043,
-	0x0635b608,
-	0xb60120b6,
-	0x24b60130,
-	0x0834b608,
-	0xf5022fb9,
-	0xbb027121,
-	0x07f1003f,
-	0x03f00100,
-	0x0003d002,
-	0x24bd04bd,
-	0xf11f29f0,
-	0xf0080007,
-	0x02d00203,
-/* 0x0433: main */
+	0x98020e98,
+	0x21f5030f,
+	0x0e980150,
+	0x00effd07,
+	0xbb002ebb,
+	0x35b6003e,
+	0x0007f102,
+	0x0103f0d3,
+	0xbd0003d0,
+	0x0825b604,
+	0xb60635b6,
+	0x30b60120,
+	0x0824b601,
+	0xb90834b6,
+	0x21f5022f,
+	0x3fbb02d3,
+	0x0007f100,
+	0x0203f001,
+	0xbd0003d0,
+	0xf024bd04,
+	0x07f11f29,
+	0x03f00800,
+	0x0002d002,
+/* 0x04e2: main */
+	0x31f404bd,
+	0x0028f400,
+	0xf424d7f0,
+	0x01f43921,
+	0x04e4b0f4,
+	0xfe1e18f4,
+	0x27f00181,
+	0xfd20bd06,
+	0xe4b60412,
+	0x051efd01,
+	0xf50018fe,
+	0xf405d721,
+/* 0x0512: main_not_ctx_xfer */
+	0xef94d30e,
+	0x01f5f010,
+	0x037e21f5,
+/* 0x051f: ih */
+	0xf9c60ef4,
+	0x0188fe80,
+	0x90f980f9,
+	0xb0f9a0f9,
+	0xe0f9d0f9,
+	0x04bdf0f9,
+	0x0200a7f1,
+	0xcf00a3f0,
+	0xabc400aa,
+	0x2c0bf404,
+	0xf124d7f0,
+	0xf01a00e7,
+	0xeecf00e3,
+	0x00f7f100,
+	0x00f3f019,
+	0xf400ffcf,
+	0xe7f00421,
+	0x0007f101,
+	0x0003f01d,
+	0xbd000ed0,
+/* 0x056d: ih_no_fifo */
+	0x0007f104,
+	0x0003f001,
+	0xbd000ad0,
+	0xfcf0fc04,
+	0xfcd0fce0,
+	0xfca0fcb0,
+	0xfe80fc90,
+	0x80fc0088,
+	0xf80032f4,
+/* 0x0591: hub_barrier_done */
+	0x01f7f001,
+	0xbb040e98,
+	0xffb904fe,
+	0x18e7f102,
+	0x40e3f094,
+	0xf89d21f4,
+/* 0x05a9: ctx_redswitch */
+	0x20f7f000,
+	0x850007f1,
+	0xd00103f0,
+	0x04bd000f,
+/* 0x05bb: ctx_redswitch_delay */
+	0xb608e7f0,
+	0x1bf401e2,
+	0x00f5f1fd,
+	0x00f5f108,
+	0x0007f102,
+	0x0103f085,
+	0xbd000fd0,
+/* 0x05d7: ctx_xfer */
+	0xf100f804,
+	0xf0810007,
+	0x0fd00203,
 	0xf404bd00,
-	0x28f40031,
-	0x24d7f000,
-	0xf43921f4,
-	0xe4b0f401,
-	0x1e18f404,
-	0xf00181fe,
-	0x20bd0627,
-	0xb60412fd,
-	0x1efd01e4,
-	0x0018fe05,
-	0x04f721f5,
-/* 0x0463: main_not_ctx_xfer */
-	0x94d30ef4,
-	0xf5f010ef,
-	0xfe21f501,
-	0xc60ef402,
-/* 0x0470: ih */
-	0x88fe80f9,
-	0xf980f901,
-	0xf9a0f990,
-	0xf9d0f9b0,
-	0xbdf0f9e0,
-	0x800acf04,
-	0xf404abc4,
-	0xb7f11d0b,
-	0xd7f01900,
-	0x40becf24,
-	0xf400bfcf,
-	0xb0b70421,
-	0xe7f00400,
-	0x00bed001,
-/* 0x04a8: ih_no_fifo */
-	0xfc400ad0,
-	0xfce0fcf0,
-	0xfcb0fcd0,
-	0xfc90fca0,
-	0x0088fe80,
-	0x32f480fc,
-/* 0x04c3: hub_barrier_done */
-	0xf001f800,
-	0x0e9801f7,
-	0x04febb04,
-	0x9418e7f1,
-	0xf440e3f0,
-	0x00f88d21,
-/* 0x04d8: ctx_redswitch */
-	0x0614e7f1,
-	0xf006e4b6,
-	0xefd020f7,
-	0x08f7f000,
-/* 0x04e8: ctx_redswitch_delay */
-	0xf401f2b6,
-	0xf7f1fd1b,
-	0xefd00a20,
-/* 0x04f7: ctx_xfer */
-	0xf100f800,
-	0xb60a0417,
-	0x1fd00614,
-	0x0711f400,
-	0x04d821f5,
-/* 0x0508: ctx_xfer_not_load */
-	0x4afc17f1,
-	0xf00213f0,
-	0x12d00c27,
-	0x1521f500,
-	0xfc27f102,
-	0x0223f047,
-	0xf00020d0,
-	0x20b6012c,
-	0x0012d003,
-	0xf001acf0,
-	0xb7f002a5,
-	0x50b3f000,
-	0xb6040c98,
-	0xbcbb0fc4,
-	0x000c9800,
-	0xf0010d98,
-	0x21f500e7,
-	0xacf00166,
-	0x00b7f101,
-	0x50b3f040,
-	0xb6040c98,
-	0xbcbb0fc4,
-	0x010c9800,
-	0x98020d98,
-	0xe7f1060f,
-	0x21f50800,
-	0xacf00166,
-	0x04a5f001,
-	0x3000b7f1,
+	0x21f50711,
+/* 0x05ea: ctx_xfer_not_load */
+	0x21f505a9,
+	0x24bd026a,
+	0x47fc07f1,
+	0xd00203f0,
+	0x04bd0002,
+	0xb6012cf0,
+	0x07f10320,
+	0x03f04afc,
+	0x0002d002,
+	0xacf004bd,
+	0x02a5f001,
+	0x0000b7f1,
 	0x9850b3f0,
 	0xc4b6040c,
 	0x00bcbb0f,
-	0x98020c98,
-	0x0f98030d,
-	0x00e7f108,
-	0x6621f502,
-	0x1521f501,
-	0x0601f402,
-/* 0x05a3: ctx_xfer_post */
-	0xf11412f4,
-	0xf04afc17,
-	0x27f00213,
-	0x0012d00d,
-	0x021521f5,
-/* 0x05b4: ctx_xfer_done */
-	0x04c321f5,
-	0x000000f8,
+	0x98000c98,
+	0xe7f0010d,
+	0x6f21f500,
+	0x01acf001,
+	0x4000b7f1,
+	0x9850b3f0,
+	0xc4b6040c,
+	0x00bcbb0f,
+	0x98010c98,
+	0x0f98020d,
+	0x00e7f106,
+	0x6f21f508,
+	0x01acf001,
+	0xf104a5f0,
+	0xf03000b7,
+	0x0c9850b3,
+	0x0fc4b604,
+	0x9800bcbb,
+	0x0d98020c,
+	0x080f9803,
+	0x0200e7f1,
+	0x016f21f5,
+	0x025e21f5,
+	0xf40601f4,
+/* 0x0686: ctx_xfer_post */
+	0x21f50712,
+/* 0x068a: ctx_xfer_done */
+	0x21f5027f,
+	0x00f80591,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
 	0x00000000,
 	0x00000000,
 	0x00000000,

drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcnve0.fuc.h

@@ -41,14 +41,14 @@ uint32_t nve0_grgpc_data[] = {
 };
 
 uint32_t nve0_grgpc_code[] = {
-	0x03180ef5,
+	0x03a10ef5,
 /* 0x0004: queue_put */
 	0x9800d898,
 	0x86f001d9,
 	0x0489b808,
 	0xf00c1bf4,
 	0x21f502f7,
-	0x00f802fe,
+	0x00f8037e,
 /* 0x001c: queue_put_next */
 	0xb60798c4,
 	0x8dbb0384,
@@ -72,184 +72,214 @@ uint32_t nve0_grgpc_code[] = {
 /* 0x0066: queue_get_done */
 	0x00f80132,
 /* 0x0068: nv_rd32 */
-	0x0728b7f1,
-	0xb906b4b6,
-	0xc9f002ec,
-	0x00bcd01f,
-/* 0x0078: nv_rd32_wait */
-	0xc800bccf,
-	0x1bf41fcc,
-	0x06a7f0fa,
-	0x010921f5,
-	0xf840bfcf,
-/* 0x008d: nv_wr32 */
-	0x28b7f100,
-	0x06b4b607,
-	0xb980bfd0,
-	0xc9f002ec,
-	0x1ec9f01f,
-/* 0x00a3: nv_wr32_wait */
-	0xcf00bcd0,
-	0xccc800bc,
-	0xfa1bf41f,
-/* 0x00ae: watchdog_reset */
-	0x87f100f8,
-	0x84b60430,
-	0x1ff9f006,
-	0xf8008fd0,
-/* 0x00bd: watchdog_clear */
-	0x3087f100,
-	0x0684b604,
-	0xf80080d0,
-/* 0x00c9: wait_donez */
-	0xf094bd00,
-	0x07f10099,
-	0x03f00f00,
-	0x0009d002,
-	0x07f104bd,
-	0x03f00600,
-	0x000ad002,
-/* 0x00e6: wait_donez_ne */
-	0x87f104bd,
-	0x83f00000,
-	0x0088cf01,
-	0xf4888aff,
-	0x94bdf31b,
-	0xf10099f0,
-	0xf0170007,
-	0x09d00203,
-	0xf804bd00,
-/* 0x0109: wait_doneo */
-	0xf094bd00,
-	0x07f10099,
-	0x03f00f00,
-	0x0009d002,
-	0x87f104bd,
-	0x84b60818,
-	0x008ad006,
-/* 0x0124: wait_doneo_e */
-	0x040087f1,
-	0xcf0684b6,
-	0x8aff0088,
-	0xf30bf488,
+	0xf002ecb9,
+	0x07f11fc9,
+	0x03f0ca00,
+	0x000cd001,
+/* 0x007a: nv_rd32_wait */
+	0xc7f104bd,
+	0xc3f0ca00,
+	0x00cccf01,
+	0xf41fccc8,
+	0xa7f0f31b,
+	0x1021f506,
+	0x00f7f101,
+	0x01f3f0cb,
+	0xf800ffcf,
+/* 0x009d: nv_wr32 */
+	0x0007f100,
+	0x0103f0cc,
+	0xbd000fd0,
+	0x02ecb904,
+	0xf01fc9f0,
+	0x07f11ec9,
+	0x03f0ca00,
+	0x000cd001,
+/* 0x00be: nv_wr32_wait */
+	0xc7f104bd,
+	0xc3f0ca00,
+	0x00cccf01,
+	0xf41fccc8,
+	0x00f8f31b,
+/* 0x00d0: wait_donez */
 	0x99f094bd,
 	0x0007f100,
-	0x0203f017,
+	0x0203f00f,
 	0xbd0009d0,
-/* 0x0147: mmctx_size */
-	0xbd00f804,
-/* 0x0149: nv_mmctx_size_loop */
-	0x00e89894,
-	0xb61a85b6,
-	0x84b60180,
-	0x0098bb02,
-	0xb804e0b6,
-	0x1bf404ef,
-	0x029fb9eb,
-/* 0x0166: mmctx_xfer */
-	0x94bd00f8,
-	0xf10199f0,
-	0xf00f0007,
-	0x09d00203,
-	0xf104bd00,
-	0xb6071087,
-	0x94bd0684,
-	0xf405bbfd,
-	0x8bd0090b,
-	0x0099f000,
-/* 0x018c: mmctx_base_disabled */
-	0xf405eefd,
-	0x8ed00c0b,
-	0xc08fd080,
-/* 0x019b: mmctx_multi_disabled */
-	0xb70199f0,
-	0xc8010080,
+	0x0007f104,
+	0x0203f006,
+	0xbd000ad0,
+/* 0x00ed: wait_donez_ne */
+	0x0087f104,
+	0x0183f000,
+	0xff0088cf,
+	0x1bf4888a,
+	0xf094bdf3,
+	0x07f10099,
+	0x03f01700,
+	0x0009d002,
+	0x00f804bd,
+/* 0x0110: wait_doneo */
+	0x99f094bd,
+	0x0007f100,
+	0x0203f00f,
+	0xbd0009d0,
+	0x0007f104,
+	0x0203f006,
+	0xbd000ad0,
+/* 0x012d: wait_doneo_e */
+	0x0087f104,
+	0x0183f000,
+	0xff0088cf,
+	0x0bf4888a,
+	0xf094bdf3,
+	0x07f10099,
+	0x03f01700,
+	0x0009d002,
+	0x00f804bd,
+/* 0x0150: mmctx_size */
+/* 0x0152: nv_mmctx_size_loop */
+	0xe89894bd,
+	0x1a85b600,
+	0xb60180b6,
+	0x98bb0284,
+	0x04e0b600,
+	0xf404efb8,
+	0x9fb9eb1b,
+/* 0x016f: mmctx_xfer */
+	0xbd00f802,
+	0x0199f094,
+	0x0f0007f1,
+	0xd00203f0,
+	0x04bd0009,
+	0xbbfd94bd,
+	0x120bf405,
+	0xc40007f1,
+	0xd00103f0,
+	0x04bd000b,
+/* 0x0197: mmctx_base_disabled */
+	0xfd0099f0,
+	0x0bf405ee,
+	0x0007f11e,
+	0x0103f0c6,
+	0xbd000ed0,
+	0x0007f104,
+	0x0103f0c7,
+	0xbd000fd0,
+	0x0199f004,
+/* 0x01b8: mmctx_multi_disabled */
+	0xb600abc8,
+	0xb9f010b4,
+	0x01aec80c,
+	0xfd11e4b6,
+	0x07f105be,
+	0x03f0c500,
+	0x000bd001,
+/* 0x01d6: mmctx_exec_loop */
+/* 0x01d6: mmctx_wait_free */
+	0xe7f104bd,
+	0xe3f0c500,
+	0x00eecf01,
+	0xf41fe4f0,
+	0xce98f30b,
+	0x05e9fd00,
+	0xc80007f1,
+	0xd00103f0,
+	0x04bd000e,
+	0xb804c0b6,
+	0x1bf404cd,
+	0x02abc8d8,
+/* 0x0207: mmctx_fini_wait */
+	0xf11f1bf4,
+	0xf0c500b7,
+	0xbbcf01b3,
+	0x1fb4f000,
+	0xf410b4b0,
+	0xa7f0f01b,
+	0xd021f402,
+/* 0x0223: mmctx_stop */
+	0xc82b0ef4,
 	0xb4b600ab,
 	0x0cb9f010,
-	0xb601aec8,
-	0xbefd11e4,
-	0x008bd005,
-/* 0x01b4: mmctx_exec_loop */
-/* 0x01b4: mmctx_wait_free */
-	0xf0008ecf,
-	0x0bf41fe4,
-	0x00ce98fa,
-	0xd005e9fd,
-	0xc0b6c08e,
-	0x04cdb804,
-	0xc8e81bf4,
-	0x1bf402ab,
-/* 0x01d5: mmctx_fini_wait */
-	0x008bcf18,
-	0xb01fb4f0,
-	0x1bf410b4,
-	0x02a7f0f7,
-	0xf4c921f4,
-/* 0x01ea: mmctx_stop */
-	0xabc81b0e,
-	0x10b4b600,
-	0xf00cb9f0,
-	0x8bd012b9,
-/* 0x01f9: mmctx_stop_wait */
-	0x008bcf00,
-	0xf412bbc8,
-/* 0x0202: mmctx_done */
-	0x94bdfa1b,
-	0xf10199f0,
-	0xf0170007,
-	0x09d00203,
-	0xf804bd00,
-/* 0x0215: strand_wait */
-	0xf0a0f900,
-	0x21f402a7,
-	0xf8a0fcc9,
-/* 0x0221: strand_pre */
-	0xfc87f100,
-	0x0283f04a,
-	0xd00c97f0,
-	0x21f50089,
-	0x00f80215,
-/* 0x0234: strand_post */
-	0x4afc87f1,
-	0xf00283f0,
-	0x89d00d97,
-	0x1521f500,
-/* 0x0247: strand_set */
-	0xf100f802,
-	0xf04ffca7,
-	0xaba202a3,
-	0xc7f00500,
-	0x00acd00f,
-	0xd00bc7f0,
-	0x21f500bc,
-	0xaed00215,
-	0x0ac7f000,
-	0xf500bcd0,
-	0xf8021521,
-/* 0x0271: strand_ctx_init */
-	0xf094bd00,
-	0x07f10399,
-	0x03f00f00,
+	0xf112b9f0,
+	0xf0c50007,
+	0x0bd00103,
+/* 0x023b: mmctx_stop_wait */
+	0xf104bd00,
+	0xf0c500b7,
+	0xbbcf01b3,
+	0x12bbc800,
+/* 0x024b: mmctx_done */
+	0xbdf31bf4,
+	0x0199f094,
+	0x170007f1,
+	0xd00203f0,
+	0x04bd0009,
+/* 0x025e: strand_wait */
+	0xa0f900f8,
+	0xf402a7f0,
+	0xa0fcd021,
+/* 0x026a: strand_pre */
+	0x97f000f8,
+	0xfc07f10c,
+	0x0203f04a,
+	0xbd0009d0,
+	0x5e21f504,
+/* 0x027f: strand_post */
+	0xf000f802,
+	0x07f10d97,
+	0x03f04afc,
 	0x0009d002,
 	0x21f504bd,
-	0xe7f00221,
-	0x4721f503,
-	0xfca7f102,
-	0x02a3f046,
-	0x0400aba0,
-	0xf040a0d0,
-	0xbcd001c7,
-	0x1521f500,
-	0x010c9202,
-	0xf000acd0,
-	0xbcd002c7,
-	0x1521f500,
-	0x3421f502,
-	0x8087f102,
-	0x0684b608,
-	0xb70089cf,
-	0x95220080,
-/* 0x02ca: ctx_init_strand_loop */
+	0x00f8025e,
+/* 0x0294: strand_set */
+	0xf10fc7f0,
+	0xf04ffc07,
+	0x0cd00203,
+	0xf004bd00,
+	0x07f10bc7,
+	0x03f04afc,
+	0x000cd002,
+	0x07f104bd,
+	0x03f04ffc,
+	0x000ed002,
+	0xc7f004bd,
+	0xfc07f10a,
+	0x0203f04a,
+	0xbd000cd0,
+	0x5e21f504,
+/* 0x02d3: strand_ctx_init */
+	0xbd00f802,
+	0x0399f094,
+	0x0f0007f1,
+	0xd00203f0,
+	0x04bd0009,
+	0x026a21f5,
+	0xf503e7f0,
+	0xbd029421,
+	0xfc07f1c4,
+	0x0203f047,
+	0xbd000cd0,
+	0x01c7f004,
+	0x4afc07f1,
+	0xd00203f0,
+	0x04bd000c,
+	0x025e21f5,
+	0xf1010c92,
+	0xf046fc07,
+	0x0cd00203,
+	0xf004bd00,
+	0x07f102c7,
+	0x03f04afc,
+	0x000cd002,
+	0x21f504bd,
+	0x21f5025e,
+	0x87f1027f,
+	0x83f04200,
+	0x0097f102,
+	0x0293f020,
+	0x950099cf,
+/* 0x034a: ctx_init_strand_loop */
 	0x8ed008fe,
 	0x408ed000,
 	0xb6808acf,
@@ -263,198 +293,230 @@ uint32_t nve0_grgpc_code[] = {
 	0x170007f1,
 	0xd00203f0,
 	0x04bd0009,
-/* 0x02fe: error */
+/* 0x037e: error */
 	0xe0f900f8,
-	0x9814e7f1,
-	0xf440e3f0,
-	0xe0b78d21,
-	0xf7f0041c,
-	0x8d21f401,
-	0x00f8e0fc,
-/* 0x0318: init */
-	0x04fe04bd,
-	0x0017f100,
-	0x0227f012,
-	0xf10012d0,
-	0xfe047017,
-	0x17f10010,
-	0x10d00400,
-	0x0427f0c0,
-	0xf40012d0,
-	0x17f11031,
-	0x14b60608,
-	0x0012cf06,
+	0xf102ffb9,
+	0xf09814e7,
+	0x21f440e3,
+	0x01f7f09d,
+	0xf102ffb9,
+	0xf09c1ce7,
+	0x21f440e3,
+	0xf8e0fc9d,
+/* 0x03a1: init */
+	0xfe04bd00,
+	0x27f00004,
+	0x0007f102,
+	0x0003f012,
+	0xbd0002d0,
+	0x1f17f104,
+	0x0010fe05,
+	0x070007f1,
+	0xd00003f0,
+	0x04bd0000,
+	0xf10427f0,
+	0xf0040007,
+	0x02d00003,
+	0xf404bd00,
+	0x27f11031,
+	0x23f08200,
+	0x0022cf01,
 	0xf00137f0,
 	0x32bb1f24,
 	0x0132b604,
 	0x80050280,
-	0x10b70603,
-	0x12cf0400,
-	0x04028000,
-	0x0c30e7f1,
-	0xbd50e3f0,
-	0xbd34bd24,
-/* 0x0371: init_unk_loop */
-	0x6821f444,
-	0xf400f6b0,
-	0xf7f00f0b,
-	0x04f2bb01,
-	0xb6054ffd,
-/* 0x0386: init_unk_next */
-	0x20b60130,
-	0x04e0b601,
-	0xf40126b0,
-/* 0x0392: init_unk_done */
-	0x0380e21b,
-	0x08048007,
-	0x010027f1,
-	0xcf0223f0,
-	0x34bd0022,
-	0x070047f1,
-	0x950644b6,
-	0x45d00825,
-	0x4045d000,
-	0x98000e98,
-	0x21f5010f,
-	0x2fbb0147,
-	0x003fbb00,
-	0x98010e98,
-	0x21f5020f,
-	0x0e980147,
-	0x00effd05,
-	0xbb002ebb,
-	0x0e98003e,
-	0x030f9802,
-	0x014721f5,
-	0xfd070e98,
+	0x27f10603,
+	0x23f08600,
+	0x0022cf01,
+	0xf1040280,
+	0xf00c30e7,
+	0x24bd50e3,
+	0x44bd34bd,
+/* 0x0410: init_unk_loop */
+	0xb06821f4,
+	0x0bf400f6,
+	0x01f7f00f,
+	0xfd04f2bb,
+	0x30b6054f,
+/* 0x0425: init_unk_next */
+	0x0120b601,
+	0xb004e0b6,
+	0x1bf40126,
+/* 0x0431: init_unk_done */
+	0x070380e2,
+	0xf1080480,
+	0xf0010027,
+	0x22cf0223,
+	0x9534bd00,
+	0x07f10825,
+	0x03f0c000,
+	0x0005d001,
+	0x07f104bd,
+	0x03f0c100,
+	0x0005d001,
+	0x0e9804bd,
+	0x010f9800,
+	0x015021f5,
+	0xbb002fbb,
+	0x0e98003f,
+	0x020f9801,
+	0x015021f5,
+	0xfd050e98,
 	0x2ebb00ef,
 	0x003ebb00,
-	0x130040b7,
-	0xd00235b6,
-	0x25b60043,
-	0x0635b608,
-	0xb60120b6,
-	0x24b60130,
-	0x0834b608,
-	0xf5022fb9,
-	0xbb027121,
-	0x07f1003f,
-	0x03f00100,
-	0x0003d002,
-	0x24bd04bd,
-	0xf11f29f0,
-	0xf0080007,
-	0x02d00203,
-/* 0x0433: main */
+	0x98020e98,
+	0x21f5030f,
+	0x0e980150,
+	0x00effd07,
+	0xbb002ebb,
+	0x35b6003e,
+	0x0007f102,
+	0x0103f0d3,
+	0xbd0003d0,
+	0x0825b604,
+	0xb60635b6,
+	0x30b60120,
+	0x0824b601,
+	0xb90834b6,
+	0x21f5022f,
+	0x3fbb02d3,
+	0x0007f100,
+	0x0203f001,
+	0xbd0003d0,
+	0xf024bd04,
+	0x07f11f29,
+	0x03f00800,
+	0x0002d002,
+/* 0x04e2: main */
+	0x31f404bd,
+	0x0028f400,
+	0xf424d7f0,
+	0x01f43921,
+	0x04e4b0f4,
+	0xfe1e18f4,
+	0x27f00181,
+	0xfd20bd06,
+	0xe4b60412,
+	0x051efd01,
+	0xf50018fe,
+	0xf405d721,
+/* 0x0512: main_not_ctx_xfer */
+	0xef94d30e,
+	0x01f5f010,
+	0x037e21f5,
+/* 0x051f: ih */
+	0xf9c60ef4,
+	0x0188fe80,
+	0x90f980f9,
+	0xb0f9a0f9,
+	0xe0f9d0f9,
+	0x04bdf0f9,
+	0x0200a7f1,
+	0xcf00a3f0,
+	0xabc400aa,
+	0x2c0bf404,
+	0xf124d7f0,
+	0xf01a00e7,
+	0xeecf00e3,
+	0x00f7f100,
+	0x00f3f019,
+	0xf400ffcf,
+	0xe7f00421,
+	0x0007f101,
+	0x0003f01d,
+	0xbd000ed0,
+/* 0x056d: ih_no_fifo */
+	0x0007f104,
+	0x0003f001,
+	0xbd000ad0,
+	0xfcf0fc04,
+	0xfcd0fce0,
+	0xfca0fcb0,
+	0xfe80fc90,
+	0x80fc0088,
+	0xf80032f4,
+/* 0x0591: hub_barrier_done */
+	0x01f7f001,
+	0xbb040e98,
+	0xffb904fe,
+	0x18e7f102,
+	0x40e3f094,
+	0xf89d21f4,
+/* 0x05a9: ctx_redswitch */
+	0x20f7f000,
+	0x850007f1,
+	0xd00103f0,
+	0x04bd000f,
+/* 0x05bb: ctx_redswitch_delay */
+	0xb608e7f0,
+	0x1bf401e2,
+	0x00f5f1fd,
+	0x00f5f108,
+	0x0007f102,
+	0x0103f085,
+	0xbd000fd0,
+/* 0x05d7: ctx_xfer */
+	0xf100f804,
+	0xf0810007,
+	0x0fd00203,
 	0xf404bd00,
-	0x28f40031,
-	0x24d7f000,
-	0xf43921f4,
-	0xe4b0f401,
-	0x1e18f404,
-	0xf00181fe,
-	0x20bd0627,
-	0xb60412fd,
-	0x1efd01e4,
-	0x0018fe05,
-	0x04f721f5,
-/* 0x0463: main_not_ctx_xfer */
-	0x94d30ef4,
-	0xf5f010ef,
-	0xfe21f501,
-	0xc60ef402,
-/* 0x0470: ih */
-	0x88fe80f9,
-	0xf980f901,
-	0xf9a0f990,
-	0xf9d0f9b0,
-	0xbdf0f9e0,
-	0x800acf04,
-	0xf404abc4,
-	0xb7f11d0b,
-	0xd7f01900,
-	0x40becf24,
-	0xf400bfcf,
-	0xb0b70421,
-	0xe7f00400,
-	0x00bed001,
-/* 0x04a8: ih_no_fifo */
-	0xfc400ad0,
-	0xfce0fcf0,
-	0xfcb0fcd0,
-	0xfc90fca0,
-	0x0088fe80,
-	0x32f480fc,
-/* 0x04c3: hub_barrier_done */
-	0xf001f800,
-	0x0e9801f7,
-	0x04febb04,
-	0x9418e7f1,
-	0xf440e3f0,
-	0x00f88d21,
-/* 0x04d8: ctx_redswitch */
-	0x0614e7f1,
-	0xf006e4b6,
-	0xefd020f7,
-	0x08f7f000,
-/* 0x04e8: ctx_redswitch_delay */
-	0xf401f2b6,
-	0xf7f1fd1b,
-	0xefd00a20,
-/* 0x04f7: ctx_xfer */
-	0xf100f800,
-	0xb60a0417,
-	0x1fd00614,
-	0x0711f400,
-	0x04d821f5,
-/* 0x0508: ctx_xfer_not_load */
-	0x4afc17f1,
-	0xf00213f0,
-	0x12d00c27,
-	0x1521f500,
-	0xfc27f102,
-	0x0223f047,
-	0xf00020d0,
-	0x20b6012c,
-	0x0012d003,
-	0xf001acf0,
-	0xb7f002a5,
-	0x50b3f000,
-	0xb6040c98,
-	0xbcbb0fc4,
-	0x000c9800,
-	0xf0010d98,
-	0x21f500e7,
-	0xacf00166,
-	0x00b7f101,
-	0x50b3f040,
-	0xb6040c98,
-	0xbcbb0fc4,
-	0x010c9800,
-	0x98020d98,
-	0xe7f1060f,
-	0x21f50800,
-	0xacf00166,
-	0x04a5f001,
-	0x3000b7f1,
+	0x21f50711,
+/* 0x05ea: ctx_xfer_not_load */
+	0x21f505a9,
+	0x24bd026a,
+	0x47fc07f1,
+	0xd00203f0,
+	0x04bd0002,
+	0xb6012cf0,
+	0x07f10320,
+	0x03f04afc,
+	0x0002d002,
+	0xacf004bd,
+	0x02a5f001,
+	0x0000b7f1,
 	0x9850b3f0,
 	0xc4b6040c,
 	0x00bcbb0f,
-	0x98020c98,
-	0x0f98030d,
-	0x00e7f108,
-	0x6621f502,
-	0x1521f501,
-	0x0601f402,
-/* 0x05a3: ctx_xfer_post */
-	0xf11412f4,
-	0xf04afc17,
-	0x27f00213,
-	0x0012d00d,
-	0x021521f5,
-/* 0x05b4: ctx_xfer_done */
-	0x04c321f5,
-	0x000000f8,
+	0x98000c98,
+	0xe7f0010d,
+	0x6f21f500,
+	0x01acf001,
+	0x4000b7f1,
+	0x9850b3f0,
+	0xc4b6040c,
+	0x00bcbb0f,
+	0x98010c98,
+	0x0f98020d,
+	0x00e7f106,
+	0x6f21f508,
+	0x01acf001,
+	0xf104a5f0,
+	0xf03000b7,
+	0x0c9850b3,
+	0x0fc4b604,
+	0x9800bcbb,
+	0x0d98020c,
+	0x080f9803,
+	0x0200e7f1,
+	0x016f21f5,
+	0x025e21f5,
+	0xf40601f4,
+/* 0x0686: ctx_xfer_post */
+	0x21f50712,
+/* 0x068a: ctx_xfer_done */
+	0x21f5027f,
+	0x00f80591,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
 	0x00000000,
 	0x00000000,
 	0x00000000,

drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcnvf0.fuc.h

@@ -41,14 +41,14 @@ uint32_t nvf0_grgpc_data[] = {
 };
 
 uint32_t nvf0_grgpc_code[] = {
-	0x03180ef5,
+	0x03a10ef5,
 /* 0x0004: queue_put */
 	0x9800d898,
 	0x86f001d9,
 	0x0489b808,
 	0xf00c1bf4,
 	0x21f502f7,
-	0x00f802fe,
+	0x00f8037e,
 /* 0x001c: queue_put_next */
 	0xb60798c4,
 	0x8dbb0384,
@@ -72,184 +72,214 @@ uint32_t nvf0_grgpc_code[] = {
 /* 0x0066: queue_get_done */
 	0x00f80132,
 /* 0x0068: nv_rd32 */
-	0x0728b7f1,
-	0xb906b4b6,
-	0xc9f002ec,
-	0x00bcd01f,
-/* 0x0078: nv_rd32_wait */
-	0xc800bccf,
-	0x1bf41fcc,
-	0x06a7f0fa,
-	0x010921f5,
-	0xf840bfcf,
-/* 0x008d: nv_wr32 */
-	0x28b7f100,
-	0x06b4b607,
-	0xb980bfd0,
-	0xc9f002ec,
-	0x1ec9f01f,
-/* 0x00a3: nv_wr32_wait */
-	0xcf00bcd0,
-	0xccc800bc,
-	0xfa1bf41f,
-/* 0x00ae: watchdog_reset */
-	0x87f100f8,
-	0x84b60430,
-	0x1ff9f006,
-	0xf8008fd0,
-/* 0x00bd: watchdog_clear */
-	0x3087f100,
-	0x0684b604,
-	0xf80080d0,
-/* 0x00c9: wait_donez */
-	0xf094bd00,
-	0x07f10099,
-	0x03f03700,
-	0x0009d002,
-	0x07f104bd,
-	0x03f00600,
-	0x000ad002,
-/* 0x00e6: wait_donez_ne */
-	0x87f104bd,
-	0x83f00000,
-	0x0088cf01,
-	0xf4888aff,
-	0x94bdf31b,
-	0xf10099f0,
-	0xf0170007,
-	0x09d00203,
-	0xf804bd00,
-/* 0x0109: wait_doneo */
-	0xf094bd00,
-	0x07f10099,
-	0x03f03700,
-	0x0009d002,
-	0x87f104bd,
-	0x84b60818,
-	0x008ad006,
-/* 0x0124: wait_doneo_e */
-	0x040087f1,
-	0xcf0684b6,
-	0x8aff0088,
-	0xf30bf488,
+	0xf002ecb9,
+	0x07f11fc9,
+	0x03f0ca00,
+	0x000cd001,
+/* 0x007a: nv_rd32_wait */
+	0xc7f104bd,
+	0xc3f0ca00,
+	0x00cccf01,
+	0xf41fccc8,
+	0xa7f0f31b,
+	0x1021f506,
+	0x00f7f101,
+	0x01f3f0cb,
+	0xf800ffcf,
+/* 0x009d: nv_wr32 */
+	0x0007f100,
+	0x0103f0cc,
+	0xbd000fd0,
+	0x02ecb904,
+	0xf01fc9f0,
+	0x07f11ec9,
+	0x03f0ca00,
+	0x000cd001,
+/* 0x00be: nv_wr32_wait */
+	0xc7f104bd,
+	0xc3f0ca00,
+	0x00cccf01,
+	0xf41fccc8,
+	0x00f8f31b,
+/* 0x00d0: wait_donez */
 	0x99f094bd,
 	0x0007f100,
-	0x0203f017,
+	0x0203f037,
 	0xbd0009d0,
-/* 0x0147: mmctx_size */
-	0xbd00f804,
-/* 0x0149: nv_mmctx_size_loop */
-	0x00e89894,
-	0xb61a85b6,
-	0x84b60180,
-	0x0098bb02,
-	0xb804e0b6,
-	0x1bf404ef,
-	0x029fb9eb,
-/* 0x0166: mmctx_xfer */
-	0x94bd00f8,
-	0xf10199f0,
-	0xf0370007,
-	0x09d00203,
-	0xf104bd00,
-	0xb6071087,
-	0x94bd0684,
-	0xf405bbfd,
-	0x8bd0090b,
-	0x0099f000,
-/* 0x018c: mmctx_base_disabled */
-	0xf405eefd,
-	0x8ed00c0b,
-	0xc08fd080,
-/* 0x019b: mmctx_multi_disabled */
-	0xb70199f0,
-	0xc8010080,
+	0x0007f104,
+	0x0203f006,
+	0xbd000ad0,
+/* 0x00ed: wait_donez_ne */
+	0x0087f104,
+	0x0183f000,
+	0xff0088cf,
+	0x1bf4888a,
+	0xf094bdf3,
+	0x07f10099,
+	0x03f01700,
+	0x0009d002,
+	0x00f804bd,
+/* 0x0110: wait_doneo */
+	0x99f094bd,
+	0x0007f100,
+	0x0203f037,
+	0xbd0009d0,
+	0x0007f104,
+	0x0203f006,
+	0xbd000ad0,
+/* 0x012d: wait_doneo_e */
+	0x0087f104,
+	0x0183f000,
+	0xff0088cf,
+	0x0bf4888a,
+	0xf094bdf3,
+	0x07f10099,
+	0x03f01700,
+	0x0009d002,
+	0x00f804bd,
+/* 0x0150: mmctx_size */
+/* 0x0152: nv_mmctx_size_loop */
+	0xe89894bd,
+	0x1a85b600,
+	0xb60180b6,
+	0x98bb0284,
+	0x04e0b600,
+	0xf404efb8,
+	0x9fb9eb1b,
+/* 0x016f: mmctx_xfer */
+	0xbd00f802,
+	0x0199f094,
+	0x370007f1,
+	0xd00203f0,
+	0x04bd0009,
+	0xbbfd94bd,
+	0x120bf405,
+	0xc40007f1,
+	0xd00103f0,
+	0x04bd000b,
+/* 0x0197: mmctx_base_disabled */
+	0xfd0099f0,
+	0x0bf405ee,
+	0x0007f11e,
+	0x0103f0c6,
+	0xbd000ed0,
+	0x0007f104,
+	0x0103f0c7,
+	0xbd000fd0,
+	0x0199f004,
+/* 0x01b8: mmctx_multi_disabled */
+	0xb600abc8,
+	0xb9f010b4,
+	0x01aec80c,
+	0xfd11e4b6,
+	0x07f105be,
+	0x03f0c500,
+	0x000bd001,
+/* 0x01d6: mmctx_exec_loop */
+/* 0x01d6: mmctx_wait_free */
+	0xe7f104bd,
+	0xe3f0c500,
+	0x00eecf01,
+	0xf41fe4f0,
+	0xce98f30b,
+	0x05e9fd00,
+	0xc80007f1,
+	0xd00103f0,
+	0x04bd000e,
+	0xb804c0b6,
+	0x1bf404cd,
+	0x02abc8d8,
+/* 0x0207: mmctx_fini_wait */
+	0xf11f1bf4,
+	0xf0c500b7,
+	0xbbcf01b3,
+	0x1fb4f000,
+	0xf410b4b0,
+	0xa7f0f01b,
+	0xd021f402,
+/* 0x0223: mmctx_stop */
+	0xc82b0ef4,
 	0xb4b600ab,
 	0x0cb9f010,
-	0xb601aec8,
-	0xbefd11e4,
-	0x008bd005,
-/* 0x01b4: mmctx_exec_loop */
-/* 0x01b4: mmctx_wait_free */
-	0xf0008ecf,
-	0x0bf41fe4,
-	0x00ce98fa,
-	0xd005e9fd,
-	0xc0b6c08e,
-	0x04cdb804,
-	0xc8e81bf4,
-	0x1bf402ab,
-/* 0x01d5: mmctx_fini_wait */
-	0x008bcf18,
-	0xb01fb4f0,
-	0x1bf410b4,
-	0x02a7f0f7,
-	0xf4c921f4,
-/* 0x01ea: mmctx_stop */
-	0xabc81b0e,
-	0x10b4b600,
-	0xf00cb9f0,
-	0x8bd012b9,
-/* 0x01f9: mmctx_stop_wait */
-	0x008bcf00,
-	0xf412bbc8,
-/* 0x0202: mmctx_done */
-	0x94bdfa1b,
-	0xf10199f0,
-	0xf0170007,
-	0x09d00203,
-	0xf804bd00,
-/* 0x0215: strand_wait */
-	0xf0a0f900,
-	0x21f402a7,
-	0xf8a0fcc9,
-/* 0x0221: strand_pre */
-	0xfc87f100,
-	0x0283f04a,
-	0xd00c97f0,
-	0x21f50089,
-	0x00f80215,
-/* 0x0234: strand_post */
-	0x4afc87f1,
-	0xf00283f0,
-	0x89d00d97,
-	0x1521f500,
-/* 0x0247: strand_set */
-	0xf100f802,
-	0xf04ffca7,
-	0xaba202a3,
-	0xc7f00500,
-	0x00acd00f,
-	0xd00bc7f0,
-	0x21f500bc,
-	0xaed00215,
-	0x0ac7f000,
-	0xf500bcd0,
-	0xf8021521,
-/* 0x0271: strand_ctx_init */
-	0xf094bd00,
-	0x07f10399,
-	0x03f03700,
+	0xf112b9f0,
+	0xf0c50007,
+	0x0bd00103,
+/* 0x023b: mmctx_stop_wait */
+	0xf104bd00,
+	0xf0c500b7,
+	0xbbcf01b3,
+	0x12bbc800,
+/* 0x024b: mmctx_done */
+	0xbdf31bf4,
+	0x0199f094,
+	0x170007f1,
+	0xd00203f0,
+	0x04bd0009,
+/* 0x025e: strand_wait */
+	0xa0f900f8,
+	0xf402a7f0,
+	0xa0fcd021,
+/* 0x026a: strand_pre */
+	0x97f000f8,
+	0xfc07f10c,
+	0x0203f04a,
+	0xbd0009d0,
+	0x5e21f504,
+/* 0x027f: strand_post */
+	0xf000f802,
+	0x07f10d97,
+	0x03f04afc,
 	0x0009d002,
 	0x21f504bd,
-	0xe7f00221,
-	0x4721f503,
-	0xfca7f102,
-	0x02a3f046,
-	0x0400aba0,
-	0xf040a0d0,
-	0xbcd001c7,
-	0x1521f500,
-	0x010c9202,
-	0xf000acd0,
-	0xbcd002c7,
-	0x1521f500,
-	0x3421f502,
-	0x8087f102,
-	0x0684b608,
-	0xb70089cf,
-	0x95220080,
-/* 0x02ca: ctx_init_strand_loop */
+	0x00f8025e,
+/* 0x0294: strand_set */
+	0xf10fc7f0,
+	0xf04ffc07,
+	0x0cd00203,
+	0xf004bd00,
+	0x07f10bc7,
+	0x03f04afc,
+	0x000cd002,
+	0x07f104bd,
+	0x03f04ffc,
+	0x000ed002,
+	0xc7f004bd,
+	0xfc07f10a,
+	0x0203f04a,
+	0xbd000cd0,
+	0x5e21f504,
+/* 0x02d3: strand_ctx_init */
+	0xbd00f802,
+	0x0399f094,
+	0x370007f1,
+	0xd00203f0,
+	0x04bd0009,
+	0x026a21f5,
+	0xf503e7f0,
+	0xbd029421,
+	0xfc07f1c4,
+	0x0203f047,
+	0xbd000cd0,
+	0x01c7f004,
+	0x4afc07f1,
+	0xd00203f0,
+	0x04bd000c,
+	0x025e21f5,
+	0xf1010c92,
+	0xf046fc07,
+	0x0cd00203,
+	0xf004bd00,
+	0x07f102c7,
+	0x03f04afc,
+	0x000cd002,
+	0x21f504bd,
+	0x21f5025e,
+	0x87f1027f,
+	0x83f04200,
+	0x0097f102,
+	0x0293f020,
+	0x950099cf,
+/* 0x034a: ctx_init_strand_loop */
 	0x8ed008fe,
 	0x408ed000,
 	0xb6808acf,
@@ -263,198 +293,230 @@ uint32_t nvf0_grgpc_code[] = {
 	0x170007f1,
 	0xd00203f0,
 	0x04bd0009,
-/* 0x02fe: error */
+/* 0x037e: error */
 	0xe0f900f8,
-	0x9814e7f1,
-	0xf440e3f0,
-	0xe0b78d21,
-	0xf7f0041c,
-	0x8d21f401,
-	0x00f8e0fc,
-/* 0x0318: init */
-	0x04fe04bd,
-	0x0017f100,
-	0x0227f012,
-	0xf10012d0,
-	0xfe047017,
-	0x17f10010,
-	0x10d00400,
-	0x0427f0c0,
-	0xf40012d0,
-	0x17f11031,
-	0x14b60608,
-	0x0012cf06,
+	0xf102ffb9,
+	0xf09814e7,
+	0x21f440e3,
+	0x01f7f09d,
+	0xf102ffb9,
+	0xf09c1ce7,
+	0x21f440e3,
+	0xf8e0fc9d,
+/* 0x03a1: init */
+	0xfe04bd00,
+	0x27f00004,
+	0x0007f102,
+	0x0003f012,
+	0xbd0002d0,
+	0x1f17f104,
+	0x0010fe05,
+	0x070007f1,
+	0xd00003f0,
+	0x04bd0000,
+	0xf10427f0,
+	0xf0040007,
+	0x02d00003,
+	0xf404bd00,
+	0x27f11031,
+	0x23f08200,
+	0x0022cf01,
 	0xf00137f0,
 	0x32bb1f24,
 	0x0132b604,
 	0x80050280,
-	0x10b70603,
-	0x12cf0400,
-	0x04028000,
-	0x0c30e7f1,
-	0xbd50e3f0,
-	0xbd34bd24,
-/* 0x0371: init_unk_loop */
-	0x6821f444,
-	0xf400f6b0,
-	0xf7f00f0b,
-	0x04f2bb01,
-	0xb6054ffd,
-/* 0x0386: init_unk_next */
-	0x20b60130,
-	0x04e0b601,
-	0xf40226b0,
-/* 0x0392: init_unk_done */
-	0x0380e21b,
-	0x08048007,
-	0x010027f1,
-	0xcf0223f0,
-	0x34bd0022,
-	0x070047f1,
-	0x950644b6,
-	0x45d00825,
-	0x4045d000,
-	0x98000e98,
-	0x21f5010f,
-	0x2fbb0147,
-	0x003fbb00,
-	0x98010e98,
-	0x21f5020f,
-	0x0e980147,
-	0x00effd05,
-	0xbb002ebb,
-	0x0e98003e,
-	0x030f9802,
-	0x014721f5,
-	0xfd070e98,
+	0x27f10603,
+	0x23f08600,
+	0x0022cf01,
+	0xf1040280,
+	0xf00c30e7,
+	0x24bd50e3,
+	0x44bd34bd,
+/* 0x0410: init_unk_loop */
+	0xb06821f4,
+	0x0bf400f6,
+	0x01f7f00f,
+	0xfd04f2bb,
+	0x30b6054f,
+/* 0x0425: init_unk_next */
+	0x0120b601,
+	0xb004e0b6,
+	0x1bf40226,
+/* 0x0431: init_unk_done */
+	0x070380e2,
+	0xf1080480,
+	0xf0010027,
+	0x22cf0223,
+	0x9534bd00,
+	0x07f10825,
+	0x03f0c000,
+	0x0005d001,
+	0x07f104bd,
+	0x03f0c100,
+	0x0005d001,
+	0x0e9804bd,
+	0x010f9800,
+	0x015021f5,
+	0xbb002fbb,
+	0x0e98003f,
+	0x020f9801,
+	0x015021f5,
+	0xfd050e98,
 	0x2ebb00ef,
 	0x003ebb00,
-	0x130040b7,
-	0xd00235b6,
-	0x25b60043,
-	0x0635b608,
-	0xb60120b6,
-	0x24b60130,
-	0x0834b608,
-	0xf5022fb9,
-	0xbb027121,
-	0x07f1003f,
-	0x03f00100,
-	0x0003d002,
-	0x24bd04bd,
-	0xf11f29f0,
-	0xf0300007,
-	0x02d00203,
-/* 0x0433: main */
+	0x98020e98,
+	0x21f5030f,
+	0x0e980150,
+	0x00effd07,
+	0xbb002ebb,
+	0x35b6003e,
+	0x0007f102,
+	0x0103f0d3,
+	0xbd0003d0,
+	0x0825b604,
+	0xb60635b6,
+	0x30b60120,
+	0x0824b601,
+	0xb90834b6,
+	0x21f5022f,
+	0x3fbb02d3,
+	0x0007f100,
+	0x0203f001,
+	0xbd0003d0,
+	0xf024bd04,
+	0x07f11f29,
+	0x03f03000,
+	0x0002d002,
+/* 0x04e2: main */
+	0x31f404bd,
+	0x0028f400,
+	0xf424d7f0,
+	0x01f43921,
+	0x04e4b0f4,
+	0xfe1e18f4,
+	0x27f00181,
+	0xfd20bd06,
+	0xe4b60412,
+	0x051efd01,
+	0xf50018fe,
+	0xf405d721,
+/* 0x0512: main_not_ctx_xfer */
+	0xef94d30e,
+	0x01f5f010,
+	0x037e21f5,
+/* 0x051f: ih */
+	0xf9c60ef4,
+	0x0188fe80,
+	0x90f980f9,
+	0xb0f9a0f9,
+	0xe0f9d0f9,
+	0x04bdf0f9,
+	0x0200a7f1,
+	0xcf00a3f0,
+	0xabc400aa,
+	0x2c0bf404,
+	0xf124d7f0,
+	0xf01a00e7,
+	0xeecf00e3,
+	0x00f7f100,
+	0x00f3f019,
+	0xf400ffcf,
+	0xe7f00421,
+	0x0007f101,
+	0x0003f01d,
+	0xbd000ed0,
+/* 0x056d: ih_no_fifo */
+	0x0007f104,
+	0x0003f001,
+	0xbd000ad0,
+	0xfcf0fc04,
+	0xfcd0fce0,
+	0xfca0fcb0,
+	0xfe80fc90,
+	0x80fc0088,
+	0xf80032f4,
+/* 0x0591: hub_barrier_done */
+	0x01f7f001,
+	0xbb040e98,
+	0xffb904fe,
+	0x18e7f102,
+	0x40e3f094,
+	0xf89d21f4,
+/* 0x05a9: ctx_redswitch */
+	0x20f7f000,
+	0x850007f1,
+	0xd00103f0,
+	0x04bd000f,
+/* 0x05bb: ctx_redswitch_delay */
+	0xb608e7f0,
+	0x1bf401e2,
+	0x00f5f1fd,
+	0x00f5f108,
+	0x0007f102,
+	0x0103f085,
+	0xbd000fd0,
+/* 0x05d7: ctx_xfer */
+	0xf100f804,
+	0xf0810007,
+	0x0fd00203,
 	0xf404bd00,
-	0x28f40031,
-	0x24d7f000,
-	0xf43921f4,
-	0xe4b0f401,
-	0x1e18f404,
-	0xf00181fe,
-	0x20bd0627,
-	0xb60412fd,
-	0x1efd01e4,
-	0x0018fe05,
-	0x04f721f5,
-/* 0x0463: main_not_ctx_xfer */
-	0x94d30ef4,
-	0xf5f010ef,
-	0xfe21f501,
-	0xc60ef402,
-/* 0x0470: ih */
-	0x88fe80f9,
-	0xf980f901,
-	0xf9a0f990,
-	0xf9d0f9b0,
-	0xbdf0f9e0,
-	0x800acf04,
-	0xf404abc4,
-	0xb7f11d0b,
-	0xd7f01900,
-	0x40becf24,
-	0xf400bfcf,
-	0xb0b70421,
-	0xe7f00400,
-	0x00bed001,
-/* 0x04a8: ih_no_fifo */
-	0xfc400ad0,
-	0xfce0fcf0,
-	0xfcb0fcd0,
-	0xfc90fca0,
-	0x0088fe80,
-	0x32f480fc,
-/* 0x04c3: hub_barrier_done */
-	0xf001f800,
-	0x0e9801f7,
-	0x04febb04,
-	0x9418e7f1,
-	0xf440e3f0,
-	0x00f88d21,
-/* 0x04d8: ctx_redswitch */
-	0x0614e7f1,
-	0xf006e4b6,
-	0xefd020f7,
-	0x08f7f000,
-/* 0x04e8: ctx_redswitch_delay */
-	0xf401f2b6,
-	0xf7f1fd1b,
-	0xefd00a20,
-/* 0x04f7: ctx_xfer */
-	0xf100f800,
-	0xb60a0417,
-	0x1fd00614,
-	0x0711f400,
-	0x04d821f5,
-/* 0x0508: ctx_xfer_not_load */
-	0x4afc17f1,
-	0xf00213f0,
-	0x12d00c27,
-	0x1521f500,
-	0xfc27f102,
-	0x0223f047,
-	0xf00020d0,
-	0x20b6012c,
-	0x0012d003,
-	0xf001acf0,
-	0xb7f002a5,
-	0x50b3f000,
-	0xb6040c98,
-	0xbcbb0fc4,
-	0x000c9800,
-	0xf0010d98,
-	0x21f500e7,
-	0xacf00166,
-	0x00b7f101,
-	0x50b3f040,
-	0xb6040c98,
-	0xbcbb0fc4,
-	0x010c9800,
-	0x98020d98,
-	0xe7f1060f,
-	0x21f50800,
-	0xacf00166,
-	0x04a5f001,
-	0x3000b7f1,
+	0x21f50711,
+/* 0x05ea: ctx_xfer_not_load */
+	0x21f505a9,
+	0x24bd026a,
+	0x47fc07f1,
+	0xd00203f0,
+	0x04bd0002,
+	0xb6012cf0,
+	0x07f10320,
+	0x03f04afc,
+	0x0002d002,
+	0xacf004bd,
+	0x02a5f001,
+	0x0000b7f1,
 	0x9850b3f0,
 	0xc4b6040c,
 	0x00bcbb0f,
-	0x98020c98,
-	0x0f98030d,
-	0x00e7f108,
-	0x6621f502,
-	0x1521f501,
-	0x0601f402,
-/* 0x05a3: ctx_xfer_post */
-	0xf11412f4,
-	0xf04afc17,
-	0x27f00213,
-	0x0012d00d,
-	0x021521f5,
-/* 0x05b4: ctx_xfer_done */
-	0x04c321f5,
-	0x000000f8,
+	0x98000c98,
+	0xe7f0010d,
+	0x6f21f500,
+	0x01acf001,
+	0x4000b7f1,
+	0x9850b3f0,
+	0xc4b6040c,
+	0x00bcbb0f,
+	0x98010c98,
+	0x0f98020d,
+	0x00e7f106,
+	0x6f21f508,
+	0x01acf001,
+	0xf104a5f0,
+	0xf03000b7,
+	0x0c9850b3,
+	0x0fc4b604,
+	0x9800bcbb,
+	0x0d98020c,
+	0x080f9803,
+	0x0200e7f1,
+	0x016f21f5,
+	0x025e21f5,
+	0xf40601f4,
+/* 0x0686: ctx_xfer_post */
+	0x21f50712,
+/* 0x068a: ctx_xfer_done */
+	0x21f5027f,
+	0x00f80591,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
 	0x00000000,
 	0x00000000,
 	0x00000000,

drivers/gpu/drm/nouveau/core/engine/graph/fuc/hub.fuc

@@ -68,60 +68,57 @@ error:
 //
 init:
 	clear b32 $r0
-	mov $sp $r0
 	mov $xdbase $r0
 
+	// setup stack
+	nv_iord($r1, NV_PGRAPH_FECS_CAPS, 0)
+	extr $r1 $r1 9:17
+	shl b32 $r1 8
+	mov $sp $r1
+
 	// enable fifo access
-	mov $r1 0x1200
-	mov $r2 2
-	iowr I[$r1 + 0x000] $r2	// FIFO_ENABLE
+	mov $r2 NV_PGRAPH_FECS_ACCESS_FIFO
+	nv_iowr(NV_PGRAPH_FECS_ACCESS, 0, $r2)
 
 	// setup i0 handler, and route all interrupts to it
 	mov $r1 #ih
 	mov $iv0 $r1
-	mov $r1 0x400
-	iowr I[$r1 + 0x300] $r0	// INTR_DISPATCH
+
+	clear b32 $r2
+	nv_iowr(NV_PGRAPH_FECS_INTR_ROUTE, 0, $r2)
 
 	// route HUB_CHANNEL_SWITCH to fuc interrupt 8
-	mov $r3 0x404
-	shl b32 $r3 6
 	mov $r2 0x2003		// { HUB_CHANNEL_SWITCH, ZERO } -> intr 8
-	iowr I[$r3 + 0x000] $r2
+	nv_iowr(NV_PGRAPH_FECS_IROUTE, 0, $r2)
 
 	// not sure what these are, route them because NVIDIA does, and
 	// the IRQ handler will signal the host if we ever get one.. we
 	// may find out if/why we need to handle these if so..
 	//
-	mov $r2 0x2004
-	iowr I[$r3 + 0x004] $r2 // { 0x04, ZERO } -> intr 9
-	mov $r2 0x200b
-	iowr I[$r3 + 0x008] $r2 // { 0x0b, ZERO } -> intr 10
-	mov $r2 0x200c
-	iowr I[$r3 + 0x01c] $r2 // { 0x0c, ZERO } -> intr 15
+	mov $r2 0x2004		// { 0x04, ZERO } -> intr 9
+	nv_iowr(NV_PGRAPH_FECS_IROUTE, 1, $r2)
+	mov $r2 0x200b		// { 0x0b, ZERO } -> intr 10
+	nv_iowr(NV_PGRAPH_FECS_IROUTE, 2, $r2)
+	mov $r2 0x200c		// { 0x0c, ZERO } -> intr 15
+	nv_iowr(NV_PGRAPH_FECS_IROUTE, 7, $r2)
 
 	// enable all INTR_UP interrupts
-	mov $r2 0xc24
-	shl b32 $r2 6
-	not b32 $r3 $r0
-	iowr I[$r2] $r3
+	sub b32 $r3 $r0 1
+	nv_iowr(NV_PGRAPH_FECS_INTR_UP_EN, 0, $r3)
 
 	// enable fifo, ctxsw, 9, 10, 15 interrupts
-	mov $r2 -0x78fc		// 0x8704
-	sethi $r2 0
-	iowr I[$r1 + 0x000] $r2	// INTR_EN_SET
+	imm32($r2, 0x8704)
+	nv_iowr(NV_PGRAPH_FECS_INTR_EN_SET, 0, $r2)
 
 	// fifo level triggered, rest edge
-	sub b32 $r1 0x100
-	mov $r2 4
-	iowr I[$r1] $r2
+	mov $r2 NV_PGRAPH_FECS_INTR_MODE_FIFO_LEVEL
+	nv_iowr(NV_PGRAPH_FECS_INTR_MODE, 0, $r2)
 
 	// enable interrupts
 	bset $flags ie0
 
 	// fetch enabled GPC/ROP counts
-	mov $r14 -0x69fc	// 0x409604
-	sethi $r14 0x400000
-	call #nv_rd32
+	nv_rd32($r14, 0x409604)
 	extr $r1 $r15 16:20
 	st b32 D[$r0 + #rop_count] $r1
 	and $r15 0x1f
@@ -131,10 +128,8 @@ init:
 	mov $r1 1
 	shl b32 $r1 $r15
 	sub b32 $r1 1
-	mov $r2 0x40c
-	shl b32 $r2 6
-	iowr I[$r2 + 0x000] $r1
-	iowr I[$r2 + 0x100] $r1
+	nv_iowr(NV_PGRAPH_FECS_BAR_MASK0, 0, $r1)
+	nv_iowr(NV_PGRAPH_FECS_BAR_MASK1, 0, $r1)
 
 	// context size calculation, reserve first 256 bytes for use by fuc
 	mov $r1 256
@@ -142,26 +137,24 @@ init:
 	// calculate size of mmio context data
 	ld b32 $r14 D[$r0 + #hub_mmio_list_head]
 	ld b32 $r15 D[$r0 + #hub_mmio_list_tail]
-	call #mmctx_size
+	call(mmctx_size)
 
 	// set mmctx base addresses now so we don't have to do it later,
 	// they don't (currently) ever change
-	mov $r3 0x700
-	shl b32 $r3 6
 	shr b32 $r4 $r1 8
-	iowr I[$r3 + 0x000] $r4		// MMCTX_SAVE_SWBASE
-	iowr I[$r3 + 0x100] $r4		// MMCTX_LOAD_SWBASE
+	nv_iowr(NV_PGRAPH_FECS_MMCTX_SAVE_SWBASE, 0, $r4)
+	nv_iowr(NV_PGRAPH_FECS_MMCTX_LOAD_SWBASE, 0, $r4)
 	add b32 $r3 0x1300
 	add b32 $r1 $r15
 	shr b32 $r15 2
-	iowr I[$r3 + 0x000] $r15	// MMCTX_LOAD_COUNT, wtf for?!?
+	nv_iowr(NV_PGRAPH_FECS_MMCTX_LOAD_COUNT, 0, $r15) // wtf??
 
 	// strands, base offset needs to be aligned to 256 bytes
 	shr b32 $r1 8
 	add b32 $r1 1
 	shl b32 $r1 8
 	mov b32 $r15 $r1
-	call #strand_ctx_init
+	call(strand_ctx_init)
 	add b32 $r1 $r15
 
 	// initialise each GPC in sequence by passing in the offset of its
@@ -173,30 +166,29 @@ init:
 	// in GPCn_CC_SCRATCH[1]
 	//
 	ld b32 $r3 D[$r0 + #gpc_count]
-	mov $r4 0x2000
-	sethi $r4 0x500000
+	imm32($r4, 0x502000)
 	init_gpc:
 		// setup, and start GPC ucode running
 		add b32 $r14 $r4 0x804
 		mov b32 $r15 $r1
-		call #nv_wr32			// CC_SCRATCH[1] = ctx offset
+		call(nv_wr32)			// CC_SCRATCH[1] = ctx offset
 		add b32 $r14 $r4 0x10c
 		clear b32 $r15
-		call #nv_wr32
+		call(nv_wr32)
 		add b32 $r14 $r4 0x104
-		call #nv_wr32			// ENTRY
+		call(nv_wr32)			// ENTRY
 		add b32 $r14 $r4 0x100
 		mov $r15 2			// CTRL_START_TRIGGER
-		call #nv_wr32			// CTRL
+		call(nv_wr32)			// CTRL
 
 		// wait for it to complete, and adjust context size
 		add b32 $r14 $r4 0x800
 		init_gpc_wait:
-			call #nv_rd32
+			call(nv_rd32)
 			xbit $r15 $r15 31
 			bra e #init_gpc_wait
 		add b32 $r14 $r4 0x804
-		call #nv_rd32
+		call(nv_rd32)
 		add b32 $r1 $r15
 
 		// next!
@@ -218,17 +210,15 @@ main:
 	bset $flags $p0
 	sleep $p0
 	mov $r13 #cmd_queue
-	call #queue_get
+	call(queue_get)
 	bra $p1 #main
 
 	// context switch, requested by GPU?
 	cmpu b32 $r14 0x4001
 	bra ne #main_not_ctx_switch
 		trace_set(T_AUTO)
-		mov $r1 0xb00
-		shl b32 $r1 6
-		iord $r2 I[$r1 + 0x100]		// CHAN_NEXT
-		iord $r1 I[$r1 + 0x000]		// CHAN_CUR
+		nv_iord($r1, NV_PGRAPH_FECS_CHAN_ADDR, 0)
+		nv_iord($r2, NV_PGRAPH_FECS_CHAN_NEXT, 0)
 
 		xbit $r3 $r1 31
 		bra e #chsw_no_prev
@@ -239,12 +229,12 @@ main:
 				trace_set(T_SAVE)
 				bclr $flags $p1
 				bset $flags $p2
-				call #ctx_xfer
+				call(ctx_xfer)
 				trace_clr(T_SAVE);
 				pop $r2
 				trace_set(T_LOAD);
 				bset $flags $p1
-				call #ctx_xfer
+				call(ctx_xfer)
 				trace_clr(T_LOAD);
 				bra #chsw_done
 			chsw_prev_no_next:
@@ -252,25 +242,21 @@ main:
 				mov b32 $r2 $r1
 				bclr $flags $p1
 				bclr $flags $p2
-				call #ctx_xfer
+				call(ctx_xfer)
 				pop $r2
-				mov $r1 0xb00
-				shl b32 $r1 6
-				iowr I[$r1] $r2
+				nv_iowr(NV_PGRAPH_FECS_CHAN_ADDR, 0, $r2)
 				bra #chsw_done
 		chsw_no_prev:
 			xbit $r3 $r2 31
 			bra e #chsw_done
 				bset $flags $p1
 				bclr $flags $p2
-				call #ctx_xfer
+				call(ctx_xfer)
 
 		// ack the context switch request
 		chsw_done:
-		mov $r1 0xb0c
-		shl b32 $r1 6
-		mov $r2 1
-		iowr I[$r1 + 0x000] $r2		// 0x409b0c
+		mov $r2 NV_PGRAPH_FECS_CHSW_ACK
+		nv_iowr(NV_PGRAPH_FECS_CHSW, 0, $r2)
 		trace_clr(T_AUTO)
 		bra #main
 
@@ -279,7 +265,7 @@ main:
 	cmpu b32 $r14 0x0001
 	bra ne #main_not_ctx_chan
 		mov b32 $r2 $r15
-		call #ctx_chan
+		call(ctx_chan)
 		bra #main_done
 
 	// request to store current channel context?
@@ -289,14 +275,14 @@ main:
 		trace_set(T_SAVE)
 		bclr $flags $p1
 		bclr $flags $p2
-		call #ctx_xfer
+		call(ctx_xfer)
 		trace_clr(T_SAVE)
 		bra #main_done
 
 	main_not_ctx_save:
 		shl b32 $r15 $r14 16
 		or $r15 E_BAD_COMMAND
-		call #error
+		call(error)
 		bra #main
 
 	main_done:
@@ -319,41 +305,38 @@ ih:
 	clear b32 $r0
 
 	// incoming fifo command?
-	iord $r10 I[$r0 + 0x200]	// INTR
-	and $r11 $r10 0x00000004
+	nv_iord($r10, NV_PGRAPH_FECS_INTR, 0)
+	and $r11 $r10 NV_PGRAPH_FECS_INTR_FIFO
 	bra e #ih_no_fifo
 		// queue incoming fifo command for later processing
-		mov $r11 0x1900
 		mov $r13 #cmd_queue
-		iord $r14 I[$r11 + 0x100]	// FIFO_CMD
-		iord $r15 I[$r11 + 0x000]	// FIFO_DATA
-		call #queue_put
+		nv_iord($r14, NV_PGRAPH_FECS_FIFO_CMD, 0)
+		nv_iord($r15, NV_PGRAPH_FECS_FIFO_DATA, 0)
+		call(queue_put)
 		add b32 $r11 0x400
 		mov $r14 1
-		iowr I[$r11 + 0x000] $r14	// FIFO_ACK
+		nv_iowr(NV_PGRAPH_FECS_FIFO_ACK, 0, $r14)
 
 	// context switch request?
 	ih_no_fifo:
-	and $r11 $r10 0x00000100
+	and $r11 $r10 NV_PGRAPH_FECS_INTR_CHSW
 	bra e #ih_no_ctxsw
 		// enqueue a context switch for later processing
 		mov $r13 #cmd_queue
 		mov $r14 0x4001
-		call #queue_put
+		call(queue_put)
 
 	// anything we didn't handle, bring it to the host's attention
 	ih_no_ctxsw:
-	mov $r11 0x104
+	mov $r11 0x104 // FIFO | CHSW
 	not b32 $r11
 	and $r11 $r10 $r11
 	bra e #ih_no_other
-		mov $r10 0xc1c
-		shl b32 $r10 6
-		iowr I[$r10] $r11	// INTR_UP_SET
+		nv_iowr(NV_PGRAPH_FECS_INTR_UP_SET, 0, $r11)
 
 	// ack, and wake up main()
 	ih_no_other:
-	iowr I[$r0 + 0x100] $r10	// INTR_ACK
+	nv_iowr(NV_PGRAPH_FECS_INTR_ACK, 0, $r10)
 
 	pop $r15
 	pop $r14
@@ -370,12 +353,10 @@ ih:
 #if CHIPSET < GK100
 // Not real sure, but, MEM_CMD 7 will hang forever if this isn't done
 ctx_4160s:
-	mov $r14 0x4160
-	sethi $r14 0x400000
 	mov $r15 1
-	call #nv_wr32
+	nv_wr32(0x404160, $r15)
 	ctx_4160s_wait:
-		call #nv_rd32
+		nv_rd32($r15, 0x404160)
 		xbit $r15 $r15 4
 		bra e #ctx_4160s_wait
 	ret
@@ -384,10 +365,8 @@ ctx_4160s:
 // to hang with STATUS=0x00000007 until it's cleared.. fbcon can
 // still function with it set however...
 ctx_4160c:
-	mov $r14 0x4160
-	sethi $r14 0x400000
 	clear b32 $r15
-	call #nv_wr32
+	nv_wr32(0x404160, $r15)
 	ret
 #endif
 
@@ -396,18 +375,14 @@ ctx_4160c:
 // In: $r15 value to set 0x404170 to
 //
 ctx_4170s:
-	mov $r14 0x4170
-	sethi $r14 0x400000
 	or $r15 0x10
-	call #nv_wr32
+	nv_wr32(0x404170, $r15)
 	ret
 
 // Waits for a ctx_4170s() call to complete
 //
 ctx_4170w:
-	mov $r14 0x4170
-	sethi $r14 0x400000
-	call #nv_rd32
+	nv_rd32($r15, 0x404170)
 	and $r15 0x10
 	bra ne #ctx_4170w
 	ret
@@ -419,16 +394,18 @@ ctx_4170w:
 // funny things happen.
 //
 ctx_redswitch:
-	mov $r14 0x614
-	shl b32 $r14 6
-	mov $r15 0x270
-	iowr I[$r14] $r15	// HUB_RED_SWITCH = ENABLE_GPC, POWER_ALL
+	mov $r14 NV_PGRAPH_FECS_RED_SWITCH_ENABLE_GPC
+	or  $r14 NV_PGRAPH_FECS_RED_SWITCH_POWER_ROP
+	or  $r14 NV_PGRAPH_FECS_RED_SWITCH_POWER_GPC
+	or  $r14 NV_PGRAPH_FECS_RED_SWITCH_POWER_MAIN
+	nv_iowr(NV_PGRAPH_FECS_RED_SWITCH, 0, $r14)
 	mov $r15 8
 	ctx_redswitch_delay:
 		sub b32 $r15 1
 		bra ne #ctx_redswitch_delay
-	mov $r15 0x770
-	iowr I[$r14] $r15	// HUB_RED_SWITCH = ENABLE_ALL, POWER_ALL
+	or  $r14 NV_PGRAPH_FECS_RED_SWITCH_ENABLE_ROP
+	or  $r14 NV_PGRAPH_FECS_RED_SWITCH_ENABLE_MAIN
+	nv_iowr(NV_PGRAPH_FECS_RED_SWITCH, 0, $r14)
 	ret
 
 // Not a clue what this is for, except that unless the value is 0x10, the
@@ -437,15 +414,18 @@ ctx_redswitch:
 // In: $r15 value to set to (0x00/0x10 are used)
 //
 ctx_86c:
-	mov $r14 0x86c
-	shl b32 $r14 6
-	iowr I[$r14] $r15	// HUB(0x86c) = val
-	mov $r14 -0x75ec
-	sethi $r14 0x400000
-	call #nv_wr32		// ROP(0xa14) = val
-	mov $r14 -0x5794
-	sethi $r14 0x410000
-	call #nv_wr32		// GPC(0x86c) = val
+	nv_iowr(0x40986c, 0, $r15)
+	nv_wr32(0x408a14, $r15)
+	nv_wr32(0x41a86c, $r15)
+	ret
+
+// In: $r15 NV_PGRAPH_FECS_MEM_CMD_*
+ctx_mem:
+	nv_iowr(NV_PGRAPH_FECS_MEM_CMD, 0, $r15)
+	ctx_mem_wait:
+		nv_iord($r15, NV_PGRAPH_FECS_MEM_CMD, 0)
+		or $r15 $r15
+		bra ne #ctx_mem_wait
 	ret
 
 // ctx_load - load's a channel's ctxctl data, and selects its vm
@@ -457,23 +437,14 @@ ctx_load:
 
 	// switch to channel, somewhat magic in parts..
 	mov $r10 12		// DONE_UNK12
-	call #wait_donez
-	mov $r1 0xa24
-	shl b32 $r1 6
-	iowr I[$r1 + 0x000] $r0	// 0x409a24
-	mov $r3 0xb00
-	shl b32 $r3 6
-	iowr I[$r3 + 0x100] $r2	// CHAN_NEXT
-	mov $r1 0xa0c
-	shl b32 $r1 6
-	mov $r4 7
-	iowr I[$r1 + 0x000] $r2 // MEM_CHAN
-	iowr I[$r1 + 0x100] $r4	// MEM_CMD
-	ctx_chan_wait_0:
-		iord $r4 I[$r1 + 0x100]
-		and $r4 0x1f
-		bra ne #ctx_chan_wait_0
-	iowr I[$r3 + 0x000] $r2	// CHAN_CUR
+	call(wait_donez)
+	clear b32 $r15
+	nv_iowr(0x409a24, 0, $r15)
+	nv_iowr(NV_PGRAPH_FECS_CHAN_NEXT, 0, $r2)
+	nv_iowr(NV_PGRAPH_FECS_MEM_CHAN, 0, $r2)
+	mov $r15 NV_PGRAPH_FECS_MEM_CMD_LOAD_CHAN
+	call(ctx_mem)
+	nv_iowr(NV_PGRAPH_FECS_CHAN_ADDR, 0, $r2)
 
 	// load channel header, fetch PGRAPH context pointer
 	mov $xtargets $r0
@@ -482,14 +453,10 @@ ctx_load:
 	add b32 $r2 2
 
 	trace_set(T_LCHAN)
-	mov $r1 0xa04
-	shl b32 $r1 6
-	iowr I[$r1 + 0x000] $r2		// MEM_BASE
-	mov $r1 0xa20
-	shl b32 $r1 6
-	mov $r2 0x0002
-	sethi $r2 0x80000000
-	iowr I[$r1 + 0x000] $r2		// MEM_TARGET = vram
+	nv_iowr(NV_PGRAPH_FECS_MEM_BASE, 0, $r2)
+	imm32($r2, NV_PGRAPH_FECS_MEM_TARGET_UNK31)
+	or  $r2 NV_PGRAPH_FECS_MEM_TARGET_AS_VRAM
+	nv_iowr(NV_PGRAPH_FECS_MEM_TARGET, 0, $r2)
 	mov $r1 0x10			// chan + 0x0210
 	mov $r2 #xfer_data
 	sethi $r2 0x00020000		// 16 bytes
@@ -507,13 +474,9 @@ ctx_load:
 
 	// set transfer base to start of context, and fetch context header
 	trace_set(T_LCTXH)
-	mov $r2 0xa04
-	shl b32 $r2 6
-	iowr I[$r2 + 0x000] $r1		// MEM_BASE
-	mov $r2 1
-	mov $r1 0xa20
-	shl b32 $r1 6
-	iowr I[$r1 + 0x000] $r2		// MEM_TARGET = vm
+	nv_iowr(NV_PGRAPH_FECS_MEM_BASE, 0, $r1)
+	mov $r2 NV_PGRAPH_FECS_MEM_TARGET_AS_VM
+	nv_iowr(NV_PGRAPH_FECS_MEM_TARGET, 0, $r2)
 	mov $r1 #chan_data
 	sethi $r1 0x00060000		// 256 bytes
 	xdld $r0 $r1
@@ -532,21 +495,15 @@ ctx_load:
 //
 ctx_chan:
 #if CHIPSET < GK100
-	call #ctx_4160s
+	call(ctx_4160s)
 #endif
-	call #ctx_load
+	call(ctx_load)
 	mov $r10 12			// DONE_UNK12
-	call #wait_donez
-	mov $r1 0xa10
-	shl b32 $r1 6
-	mov $r2 5
-	iowr I[$r1 + 0x000] $r2		// MEM_CMD = 5 (???)
-	ctx_chan_wait:
-		iord $r2 I[$r1 + 0x000]
-		or $r2 $r2
-		bra ne #ctx_chan_wait
+	call(wait_donez)
+	mov $r15 5 // MEM_CMD 5 ???
+	call(ctx_mem)
 #if CHIPSET < GK100
-	call #ctx_4160c
+	call(ctx_4160c)
 #endif
 	ret
 
@@ -562,9 +519,7 @@ ctx_chan:
 ctx_mmio_exec:
 	// set transfer base to be the mmio list
 	ld b32 $r3 D[$r0 + #chan_mmio_address]
-	mov $r2 0xa04
-	shl b32 $r2 6
-	iowr I[$r2 + 0x000] $r3		// MEM_BASE
+	nv_iowr(NV_PGRAPH_FECS_MEM_BASE, 0, $r3)
 
 	clear b32 $r3
 	ctx_mmio_loop:
@@ -580,7 +535,7 @@ ctx_mmio_exec:
 		ctx_mmio_pull:
 		ld b32 $r14 D[$r4 + #xfer_data + 0x00]
 		ld b32 $r15 D[$r4 + #xfer_data + 0x04]
-		call #nv_wr32
+		call(nv_wr32)
 
 		// next!
 		add b32 $r3 8
@@ -590,7 +545,7 @@ ctx_mmio_exec:
 	// set transfer base back to the current context
 	ctx_mmio_done:
 	ld b32 $r3 D[$r0 + #ctx_current]
-	iowr I[$r2 + 0x000] $r3		// MEM_BASE
+	nv_iowr(NV_PGRAPH_FECS_MEM_BASE, 0, $r3)
 
 	// disable the mmio list now, we don't need/want to execute it again
 	st b32 D[$r0 + #chan_mmio_count] $r0
@@ -610,12 +565,10 @@ ctx_mmio_exec:
 //
 ctx_xfer:
 	// according to mwk, some kind of wait for idle
-	mov $r15 0xc00
-	shl b32 $r15 6
 	mov $r14 4
-	iowr I[$r15 + 0x200] $r14
+	nv_iowr(0x409c08, 0, $r14)
 	ctx_xfer_idle:
-		iord $r14 I[$r15 + 0x000]
+		nv_iord($r14, 0x409c00, 0)
 		and $r14 0x2000
 		bra ne #ctx_xfer_idle
 
@@ -623,50 +576,42 @@ ctx_xfer:
 	bra $p2 #ctx_xfer_pre_load
 	ctx_xfer_pre:
 		mov $r15 0x10
-		call #ctx_86c
+		call(ctx_86c)
 #if CHIPSET < GK100
-		call #ctx_4160s
+		call(ctx_4160s)
 #endif
 		bra not $p1 #ctx_xfer_exec
 
 	ctx_xfer_pre_load:
 		mov $r15 2
-		call #ctx_4170s
-		call #ctx_4170w
-		call #ctx_redswitch
+		call(ctx_4170s)
+		call(ctx_4170w)
+		call(ctx_redswitch)
 		clear b32 $r15
-		call #ctx_4170s
-		call #ctx_load
+		call(ctx_4170s)
+		call(ctx_load)
 
 	// fetch context pointer, and initiate xfer on all GPCs
 	ctx_xfer_exec:
 	ld b32 $r1 D[$r0 + #ctx_current]
-	mov $r2 0x414
-	shl b32 $r2 6
-	iowr I[$r2 + 0x000] $r0	// BAR_STATUS = reset
-	mov $r14 -0x5b00
-	sethi $r14 0x410000
-	mov b32 $r15 $r1
-	call #nv_wr32		// GPC_BCAST_WRCMD_DATA = ctx pointer
-	add b32 $r14 4
+
+	clear b32 $r2
+	nv_iowr(NV_PGRAPH_FECS_BAR, 0, $r2)
+
+	nv_wr32(0x41a500, $r1)	// GPC_BCAST_WRCMD_DATA = ctx pointer
 	xbit $r15 $flags $p1
 	xbit $r2 $flags $p2
 	shl b32 $r2 1
 	or $r15 $r2
-	call #nv_wr32		// GPC_BCAST_WRCMD_CMD = GPC_XFER(type)
+	nv_wr32(0x41a504, $r15)	// GPC_BCAST_WRCMD_CMD = GPC_XFER(type)
 
 	// strands
-	mov $r1 0x4afc
-	sethi $r1 0x20000
-	mov $r2 0xc
-	iowr I[$r1] $r2		// STRAND_CMD(0x3f) = 0x0c
-	call #strand_wait
-	mov $r2 0x47fc
-	sethi $r2 0x20000
-	iowr I[$r2] $r0		// STRAND_FIRST_GENE(0x3f) = 0x00
-	xbit $r2 $flags $p1
-	add b32 $r2 3
-	iowr I[$r1] $r2		// STRAND_CMD(0x3f) = 0x03/0x04 (SAVE/LOAD)
+	call(strand_pre)
+	clear b32 $r2
+	nv_iowr(NV_PGRAPH_FECS_STRAND_SELECT, 0x3f, $r2)
+	xbit $r2 $flags $p1	// SAVE/LOAD
+	add b32 $r2 NV_PGRAPH_FECS_STRAND_CMD_SAVE
+	nv_iowr(NV_PGRAPH_FECS_STRAND_CMD, 0x3f, $r2)
 
 	// mmio context
 	xbit $r10 $flags $p1	// direction
@@ -675,48 +620,42 @@ ctx_xfer:
 	ld b32 $r12 D[$r0 + #hub_mmio_list_head]
 	ld b32 $r13 D[$r0 + #hub_mmio_list_tail]
 	mov $r14 0		// not multi
-	call #mmctx_xfer
+	call(mmctx_xfer)
 
 	// wait for GPCs to all complete
 	mov $r10 8		// DONE_BAR
-	call #wait_doneo
+	call(wait_doneo)
 
 	// wait for strand xfer to complete
-	call #strand_wait
+	call(strand_wait)
 
 	// post-op
 	bra $p1 #ctx_xfer_post
 		mov $r10 12		// DONE_UNK12
-		call #wait_donez
-		mov $r1 0xa10
-		shl b32 $r1 6
-		mov $r2 5
-		iowr I[$r1] $r2		// MEM_CMD
-		ctx_xfer_post_save_wait:
-			iord $r2 I[$r1]
-			or $r2 $r2
-			bra ne #ctx_xfer_post_save_wait
+		call(wait_donez)
+		mov $r15 5 // MEM_CMD 5 ???
+		call(ctx_mem)
 
 	bra $p2 #ctx_xfer_done
 	ctx_xfer_post:
 		mov $r15 2
-		call #ctx_4170s
+		call(ctx_4170s)
 		clear b32 $r15
-		call #ctx_86c
-		call #strand_post
-		call #ctx_4170w
+		call(ctx_86c)
+		call(strand_post)
+		call(ctx_4170w)
 		clear b32 $r15
-		call #ctx_4170s
+		call(ctx_4170s)
 
 		bra not $p1 #ctx_xfer_no_post_mmio
 		ld b32 $r1 D[$r0 + #chan_mmio_count]
 		or $r1 $r1
 		bra e #ctx_xfer_no_post_mmio
-			call #ctx_mmio_exec
+			call(ctx_mmio_exec)
 
 		ctx_xfer_no_post_mmio:
 #if CHIPSET < GK100
-		call #ctx_4160c
+		call(ctx_4160c)
 #endif
 
 	ctx_xfer_done:

drivers/gpu/drm/nouveau/core/engine/graph/fuc/macros.fuc

@@ -28,8 +28,54 @@
 #define GF117 0xd7
 #define GK100 0xe0
 #define GK110 0xf0
+#define GK208 0x108
 
+#define NV_PGRAPH_FECS_INTR_ACK                                        0x409004
+#define NV_PGRAPH_FECS_INTR                                            0x409008
+#define NV_PGRAPH_FECS_INTR_CHSW                                     0x00000100
+#define NV_PGRAPH_FECS_INTR_FIFO                                     0x00000004
+#define NV_PGRAPH_FECS_INTR_MODE                                       0x40900c
+#define NV_PGRAPH_FECS_INTR_MODE_FIFO                                0x00000004
+#define NV_PGRAPH_FECS_INTR_MODE_FIFO_LEVEL                          0x00000004
+#define NV_PGRAPH_FECS_INTR_MODE_FIFO_EDGE                           0x00000000
+#define NV_PGRAPH_FECS_INTR_EN_SET                                     0x409010
+#define NV_PGRAPH_FECS_INTR_EN_SET_FIFO                              0x00000004
+#define NV_PGRAPH_FECS_INTR_ROUTE                                      0x40901c
+#define NV_PGRAPH_FECS_ACCESS                                          0x409048
+#define NV_PGRAPH_FECS_ACCESS_FIFO                                   0x00000002
+#define NV_PGRAPH_FECS_FIFO_DATA                                       0x409064
+#define NV_PGRAPH_FECS_FIFO_CMD                                        0x409068
+#define NV_PGRAPH_FECS_FIFO_ACK                                        0x409074
+#define NV_PGRAPH_FECS_CAPS                                            0x409108
 #define NV_PGRAPH_FECS_SIGNAL                                          0x409400
+#define NV_PGRAPH_FECS_IROUTE                                          0x409404
+#define NV_PGRAPH_FECS_BAR_MASK0                                       0x40940c
+#define NV_PGRAPH_FECS_BAR_MASK1                                       0x409410
+#define NV_PGRAPH_FECS_BAR                                             0x409414
+#define NV_PGRAPH_FECS_BAR_SET                                         0x409418
+#define NV_PGRAPH_FECS_RED_SWITCH                                      0x409614
+#define NV_PGRAPH_FECS_RED_SWITCH_ENABLE_ROP                         0x00000400
+#define NV_PGRAPH_FECS_RED_SWITCH_ENABLE_GPC                         0x00000200
+#define NV_PGRAPH_FECS_RED_SWITCH_ENABLE_MAIN                        0x00000100
+#define NV_PGRAPH_FECS_RED_SWITCH_POWER_ROP                          0x00000040
+#define NV_PGRAPH_FECS_RED_SWITCH_POWER_GPC                          0x00000020
+#define NV_PGRAPH_FECS_RED_SWITCH_POWER_MAIN                         0x00000010
+#define NV_PGRAPH_FECS_RED_SWITCH_PAUSE_GPC                          0x00000002
+#define NV_PGRAPH_FECS_RED_SWITCH_PAUSE_MAIN                         0x00000001
+#define NV_PGRAPH_FECS_MMCTX_SAVE_SWBASE                               0x409700
+#define NV_PGRAPH_FECS_MMCTX_LOAD_SWBASE                               0x409704
+#define NV_PGRAPH_FECS_MMCTX_LOAD_COUNT                                0x40974c
+#define NV_PGRAPH_FECS_MMCTX_SAVE_SWBASE                               0x409700
+#define NV_PGRAPH_FECS_MMCTX_LOAD_SWBASE                               0x409704
+#define NV_PGRAPH_FECS_MMCTX_BASE                                      0x409710
+#define NV_PGRAPH_FECS_MMCTX_CTRL                                      0x409714
+#define NV_PGRAPH_FECS_MMCTX_MULTI_STRIDE                              0x409718
+#define NV_PGRAPH_FECS_MMCTX_MULTI_MASK                                0x40971c
+#define NV_PGRAPH_FECS_MMCTX_QUEUE                                     0x409720
+#define NV_PGRAPH_FECS_MMIO_CTRL                                       0x409728
+#define NV_PGRAPH_FECS_MMIO_RDVAL                                      0x40972c
+#define NV_PGRAPH_FECS_MMIO_WRVAL                                      0x409730
+#define NV_PGRAPH_FECS_MMCTX_LOAD_COUNT                                0x40974c
 #if CHIPSET < GK110
 #define NV_PGRAPH_FECS_CC_SCRATCH_VAL(n)                    ((n) * 4 + 0x409800)
 #define NV_PGRAPH_FECS_CC_SCRATCH_SET(n)                    ((n) * 4 + 0x409820)
@@ -39,8 +85,59 @@
 #define NV_PGRAPH_FECS_CC_SCRATCH_CLR(n)                    ((n) * 4 + 0x409840)
 #define NV_PGRAPH_FECS_CC_SCRATCH_SET(n)                    ((n) * 4 + 0x4098c0)
 #endif
+#define NV_PGRAPH_FECS_STRANDS_CNT                                     0x409880
+#define NV_PGRAPH_FECS_STRAND_SAVE_SWBASE                              0x409908
+#define NV_PGRAPH_FECS_STRAND_LOAD_SWBASE                              0x40990c
+#define NV_PGRAPH_FECS_STRAND_WORDS                                    0x409910
+#define NV_PGRAPH_FECS_STRAND_DATA                                     0x409918
+#define NV_PGRAPH_FECS_STRAND_SELECT                                   0x40991c
+#define NV_PGRAPH_FECS_STRAND_CMD                                      0x409928
+#define NV_PGRAPH_FECS_STRAND_CMD_SEEK                               0x00000001
+#define NV_PGRAPH_FECS_STRAND_CMD_GET_INFO                           0x00000002
+#define NV_PGRAPH_FECS_STRAND_CMD_SAVE                               0x00000003
+#define NV_PGRAPH_FECS_STRAND_CMD_LOAD                               0x00000004
+#define NV_PGRAPH_FECS_STRAND_CMD_ACTIVATE_FILTER                    0x0000000a
+#define NV_PGRAPH_FECS_STRAND_CMD_DEACTIVATE_FILTER                  0x0000000b
+#define NV_PGRAPH_FECS_STRAND_CMD_ENABLE                             0x0000000c
+#define NV_PGRAPH_FECS_STRAND_CMD_DISABLE                            0x0000000d
+#define NV_PGRAPH_FECS_STRAND_FILTER                                   0x40993c
+#define NV_PGRAPH_FECS_MEM_BASE                                        0x409a04
+#define NV_PGRAPH_FECS_MEM_CHAN                                        0x409a0c
+#define NV_PGRAPH_FECS_MEM_CMD                                         0x409a10
+#define NV_PGRAPH_FECS_MEM_CMD_LOAD_CHAN                             0x00000007
+#define NV_PGRAPH_FECS_MEM_TARGET                                      0x409a20
+#define NV_PGRAPH_FECS_MEM_TARGET_UNK31                              0x80000000
+#define NV_PGRAPH_FECS_MEM_TARGET_AS                                 0x0000001f
+#define NV_PGRAPH_FECS_MEM_TARGET_AS_VM                              0x00000001
+#define NV_PGRAPH_FECS_MEM_TARGET_AS_VRAM                            0x00000002
+#define NV_PGRAPH_FECS_CHAN_ADDR                                       0x409b00
+#define NV_PGRAPH_FECS_CHAN_NEXT                                       0x409b04
+#define NV_PGRAPH_FECS_CHSW                                            0x409b0c
+#define NV_PGRAPH_FECS_CHSW_ACK                                      0x00000001
 #define NV_PGRAPH_FECS_INTR_UP_SET                                     0x409c1c
+#define NV_PGRAPH_FECS_INTR_UP_EN                                      0x409c24
 
+#define NV_PGRAPH_GPCX_GPCCS_INTR_ACK                                  0x41a004
+#define NV_PGRAPH_GPCX_GPCCS_INTR                                      0x41a008
+#define NV_PGRAPH_GPCX_GPCCS_INTR_FIFO                               0x00000004
+#define NV_PGRAPH_GPCX_GPCCS_INTR_EN_SET                               0x41a010
+#define NV_PGRAPH_GPCX_GPCCS_INTR_EN_SET_FIFO                        0x00000004
+#define NV_PGRAPH_GPCX_GPCCS_INTR_ROUTE                                0x41a01c
+#define NV_PGRAPH_GPCX_GPCCS_ACCESS                                    0x41a048
+#define NV_PGRAPH_GPCX_GPCCS_ACCESS_FIFO                             0x00000002
+#define NV_PGRAPH_GPCX_GPCCS_FIFO_DATA                                 0x41a064
+#define NV_PGRAPH_GPCX_GPCCS_FIFO_CMD                                  0x41a068
+#define NV_PGRAPH_GPCX_GPCCS_FIFO_ACK                                  0x41a074
+#define NV_PGRAPH_GPCX_GPCCS_UNITS                                     0x41a608
+#define NV_PGRAPH_GPCX_GPCCS_RED_SWITCH                                0x41a614
+#define NV_PGRAPH_GPCX_GPCCS_RED_SWITCH_UNK11                        0x00000800
+#define NV_PGRAPH_GPCX_GPCCS_RED_SWITCH_ENABLE                       0x00000200
+#define NV_PGRAPH_GPCX_GPCCS_RED_SWITCH_POWER                        0x00000020
+#define NV_PGRAPH_GPCX_GPCCS_RED_SWITCH_PAUSE                        0x00000002
+#define NV_PGRAPH_GPCX_GPCCS_MYINDEX                                   0x41a618
+#define NV_PGRAPH_GPCX_GPCCS_MMCTX_SAVE_SWBASE                         0x41a700
+#define NV_PGRAPH_GPCX_GPCCS_MMCTX_LOAD_SWBASE                         0x41a704
+#define NV_PGRAPH_GPCX_GPCCS_MMCTX_LOAD_COUNT                          0x41a74c
 #if CHIPSET < GK110
 #define NV_PGRAPH_GPCX_GPCCS_CC_SCRATCH_VAL(n)              ((n) * 4 + 0x41a800)
 #define NV_PGRAPH_GPCX_GPCCS_CC_SCRATCH_SET(n)              ((n) * 4 + 0x41a820)
@@ -50,6 +147,11 @@
 #define NV_PGRAPH_GPCX_GPCCS_CC_SCRATCH_CLR(n)              ((n) * 4 + 0x41a840)
 #define NV_PGRAPH_GPCX_GPCCS_CC_SCRATCH_SET(n)              ((n) * 4 + 0x41a8c0)
 #endif
+#define NV_PGRAPH_GPCX_GPCCS_STRAND_SELECT                             0x41a91c
+#define NV_PGRAPH_GPCX_GPCCS_STRAND_CMD                                0x41a928
+#define NV_PGRAPH_GPCX_GPCCS_STRAND_CMD_SAVE                         0x00000003
+#define NV_PGRAPH_GPCX_GPCCS_STRAND_CMD_LOAD                         0x00000004
+#define NV_PGRAPH_GPCX_GPCCS_MEM_BASE                                  0x41aa04
 
 #define mmctx_data(r,c) .b32 (((c - 1) << 26) | r)
 #define queue_init      .skip 72 // (2 * 4) + ((8 * 4) * 2)
@@ -65,24 +167,50 @@
 #define T_LCHAN   8
 #define T_LCTXH   9
 
-#define nv_mkmm(rv,r) /*
-*/	movw rv  ((r) & 0x0000fffc) /*
-*/	sethi rv ((r) & 0x00ff0000)
+#if CHIPSET < GK208
+#define imm32(reg,val) /*
+*/	movw reg  ((val) & 0x0000ffff) /*
+*/	sethi reg ((val) & 0xffff0000)
+#else
+#define imm32(reg,val) /*
+*/	mov reg (val)
+#endif
+
 #define nv_mkio(rv,r,i) /*
-*/	nv_mkmm(rv, (((r) & 0xffc) << 6) | ((i) << 2))
+*/	imm32(rv, (((r) & 0xffc) << 6) | ((i) << 2))
+
+#define hash #
+#define fn(a) a
+#if CHIPSET < GK208
+#define call(a) call fn(hash)a
+#else
+#define call(a) lcall fn(hash)a
+#endif
 
 #define nv_iord(rv,r,i) /*
 */	nv_mkio(rv,r,i) /*
 */	iord rv I[rv]
+
 #define nv_iowr(r,i,rv) /*
 */	nv_mkio($r0,r,i) /*
 */	iowr I[$r0] rv /*
 */	clear b32 $r0
 
+#define nv_rd32(reg,addr) /*
+*/	imm32($r14, addr) /*
+*/	call(nv_rd32) /*
+*/	mov b32 reg $r15
+
+#define nv_wr32(addr,reg) /*
+*/	mov b32 $r15 reg /*
+*/	imm32($r14, addr) /*
+*/	call(nv_wr32)
+
 #define trace_set(bit) /*
 */	clear b32 $r9 /*
 */	bset $r9 bit /*
 */	nv_iowr(NV_PGRAPH_FECS_CC_SCRATCH_SET(7), 0, $r9)
+
 #define trace_clr(bit) /*
 */	clear b32 $r9 /*
 */	bset $r9 bit /*