linux/drivers/gpu/drm/xe/xe_mmio.c

// SPDX-License-Identifier: MIT
/*
 * Copyright © 2021-2023 Intel Corporation
 */

#include "xe_mmio.h"

#include <linux/delay.h>
#include <linux/io-64-nonatomic-lo-hi.h>
#include <linux/minmax.h>
#include <linux/pci.h>

#include <drm/drm_managed.h>
#include <drm/drm_print.h>

#include "regs/xe_bars.h"
#include "regs/xe_regs.h"
#include "xe_device.h"
#include "xe_gt.h"
#include "xe_gt_printk.h"
#include "xe_gt_sriov_vf.h"
#include "xe_macros.h"
#include "xe_sriov.h"
#include "xe_trace.h"
#include "xe_wa.h"

#include "generated/xe_device_wa_oob.h"

static void tiles_fini(void *arg)
{
	struct xe_device *xe = arg;
	struct xe_tile *tile;
	int id;

	for_each_remote_tile(tile, xe, id)
		tile->mmio.regs = NULL;
}

/*
 * On multi-tile devices, partition the BAR space for MMIO on each tile,
 * possibly accounting for register override on the number of tiles available.
 * tile_mmio_size contains both the tile's 4MB register space, as well as
 * additional space for the GTT and other (possibly unused) regions).
 * Resulting memory layout is like below:
 *
 * .----------------------. <- tile_count * tile_mmio_size
 * |         ....         |
 * |----------------------| <- 2 * tile_mmio_size
 * |   tile1 GTT + other  |
 * |----------------------| <- 1 * tile_mmio_size + 4MB
 * |   tile1->mmio.regs   |
 * |----------------------| <- 1 * tile_mmio_size
 * |   tile0 GTT + other  |
 * |----------------------| <- 4MB
 * |   tile0->mmio.regs   |
 * '----------------------' <- 0MB
 */
static void mmio_multi_tile_setup(struct xe_device *xe, size_t tile_mmio_size)
{
	struct xe_tile *tile;
	struct xe_gt *gt;
	u8 id;

	/*
	 * Nothing to be done as tile 0 has already been setup earlier with the
	 * entire BAR mapped - see xe_mmio_probe_early()
	 */
	if (xe->info.tile_count == 1)
		return;

	/* Possibly override number of tile based on configuration register */
	if (!xe->info.skip_mtcfg) {
		struct xe_mmio *mmio = xe_root_tile_mmio(xe);
		u8 tile_count, gt_count;
		u32 mtcfg;

		/*
		 * Although the per-tile mmio regs are not yet initialized, this
		 * is fine as it's going to the root tile's mmio, that's
		 * guaranteed to be initialized earlier in xe_mmio_probe_early()
		 */
		mtcfg = xe_mmio_read32(mmio, XEHP_MTCFG_ADDR);
		tile_count = REG_FIELD_GET(TILE_COUNT, mtcfg) + 1;

		if (tile_count < xe->info.tile_count) {
			drm_info(&xe->drm, "tile_count: %d, reduced_tile_count %d\n",
				 xe->info.tile_count, tile_count);
			xe->info.tile_count = tile_count;

			/*
			 * We've already setup gt_count according to the full
			 * tile count.  Re-calculate it to only include the GTs
			 * that belong to the remaining tile(s).
			 */
			gt_count = 0;
			for_each_gt(gt, xe, id)
				if (gt->info.id < tile_count * xe->info.max_gt_per_tile)
					gt_count++;
			xe->info.gt_count = gt_count;
		}
	}

	for_each_remote_tile(tile, xe, id)
		xe_mmio_init(&tile->mmio, tile, xe->mmio.regs + id * tile_mmio_size, SZ_4M);
}

int xe_mmio_probe_tiles(struct xe_device *xe)
{
	size_t tile_mmio_size = SZ_16M;

	mmio_multi_tile_setup(xe, tile_mmio_size);

	return devm_add_action_or_reset(xe->drm.dev, tiles_fini, xe);
}

static void mmio_fini(void *arg)
{
	struct xe_device *xe = arg;
	struct xe_tile *root_tile = xe_device_get_root_tile(xe);

	pci_iounmap(to_pci_dev(xe->drm.dev), xe->mmio.regs);
	xe->mmio.regs = NULL;
	root_tile->mmio.regs = NULL;
}

int xe_mmio_probe_early(struct xe_device *xe)
{
	struct xe_tile *root_tile = xe_device_get_root_tile(xe);
	struct pci_dev *pdev = to_pci_dev(xe->drm.dev);

	/*
	 * Map the entire BAR.
	 * The first 16MB of the BAR, belong to the root tile, and include:
	 * registers (0-4MB), reserved space (4MB-8MB) and GGTT (8MB-16MB).
	 */
	xe->mmio.size = pci_resource_len(pdev, GTTMMADR_BAR);
	xe->mmio.regs = pci_iomap(pdev, GTTMMADR_BAR, 0);
	if (!xe->mmio.regs) {
		drm_err(&xe->drm, "failed to map registers\n");
		return -EIO;
	}

	/* Setup first tile; other tiles (if present) will be setup later. */
	xe_mmio_init(&root_tile->mmio, root_tile, xe->mmio.regs, SZ_4M);

	return devm_add_action_or_reset(xe->drm.dev, mmio_fini, xe);
}
ALLOW_ERROR_INJECTION(xe_mmio_probe_early, ERRNO); /* See xe_pci_probe() */

/**
 * xe_mmio_init() - Initialize an MMIO instance
 * @mmio: Pointer to the MMIO instance to initialize
 * @tile: The tile to which the MMIO region belongs
 * @ptr: Pointer to the start of the MMIO region
 * @size: The size of the MMIO region in bytes
 *
 * This is a convenience function for minimal initialization of struct xe_mmio.
 */
void xe_mmio_init(struct xe_mmio *mmio, struct xe_tile *tile, void __iomem *ptr, u32 size)
{
	xe_tile_assert(tile, size <= XE_REG_ADDR_MAX);

	mmio->regs = ptr;
	mmio->regs_size = size;
	mmio->tile = tile;
}

static void mmio_flush_pending_writes(struct xe_mmio *mmio)
{
#define DUMMY_REG_OFFSET	0x130030
	int i;

	if (!XE_DEVICE_WA(mmio->tile->xe, 15015404425))
		return;

	/* 4 dummy writes */
	for (i = 0; i < 4; i++)
		writel(0, mmio->regs + DUMMY_REG_OFFSET);
}

u8 xe_mmio_read8(struct xe_mmio *mmio, struct xe_reg reg)
{
	u32 addr = xe_mmio_adjusted_addr(mmio, reg.addr);
	u8 val;

	mmio_flush_pending_writes(mmio);

	val = readb(mmio->regs + addr);
	trace_xe_reg_rw(mmio, false, addr, val, sizeof(val));

	return val;
}

u16 xe_mmio_read16(struct xe_mmio *mmio, struct xe_reg reg)
{
	u32 addr = xe_mmio_adjusted_addr(mmio, reg.addr);
	u16 val;

	mmio_flush_pending_writes(mmio);

	val = readw(mmio->regs + addr);
	trace_xe_reg_rw(mmio, false, addr, val, sizeof(val));

	return val;
}

void xe_mmio_write32(struct xe_mmio *mmio, struct xe_reg reg, u32 val)
{
	u32 addr = xe_mmio_adjusted_addr(mmio, reg.addr);

	trace_xe_reg_rw(mmio, true, addr, val, sizeof(val));

	if (!reg.vf && IS_SRIOV_VF(mmio->tile->xe))
		xe_gt_sriov_vf_write32(mmio->sriov_vf_gt ?:
				       mmio->tile->primary_gt, reg, val);
	else
		writel(val, mmio->regs + addr);
}

u32 xe_mmio_read32(struct xe_mmio *mmio, struct xe_reg reg)
{
	u32 addr = xe_mmio_adjusted_addr(mmio, reg.addr);
	u32 val;

	mmio_flush_pending_writes(mmio);

	if (!reg.vf && IS_SRIOV_VF(mmio->tile->xe))
		val = xe_gt_sriov_vf_read32(mmio->sriov_vf_gt ?:
					    mmio->tile->primary_gt, reg);
	else
		val = readl(mmio->regs + addr);

	trace_xe_reg_rw(mmio, false, addr, val, sizeof(val));

	return val;
}

u32 xe_mmio_rmw32(struct xe_mmio *mmio, struct xe_reg reg, u32 clr, u32 set)
{
	u32 old, reg_val;

	old = xe_mmio_read32(mmio, reg);
	reg_val = (old & ~clr) | set;
	xe_mmio_write32(mmio, reg, reg_val);

	return old;
}

int xe_mmio_write32_and_verify(struct xe_mmio *mmio,
			       struct xe_reg reg, u32 val, u32 mask, u32 eval)
{
	u32 reg_val;

	xe_mmio_write32(mmio, reg, val);
	reg_val = xe_mmio_read32(mmio, reg);

	return (reg_val & mask) != eval ? -EINVAL : 0;
}

bool xe_mmio_in_range(const struct xe_mmio *mmio,
		      const struct xe_mmio_range *range,
		      struct xe_reg reg)
{
	u32 addr = xe_mmio_adjusted_addr(mmio, reg.addr);

	return range && addr >= range->start && addr <= range->end;
}

/**
 * xe_mmio_read64_2x32() - Read a 64-bit register as two 32-bit reads
 * @mmio: MMIO target
 * @reg: register to read value from
 *
 * Although Intel GPUs have some 64-bit registers, the hardware officially
 * only supports GTTMMADR register reads of 32 bits or smaller.  Even if
 * a readq operation may return a reasonable value, that violation of the
 * spec shouldn't be relied upon and all 64-bit register reads should be
 * performed as two 32-bit reads of the upper and lower dwords.
 *
 * When reading registers that may be changing (such as
 * counters), a rollover of the lower dword between the two 32-bit reads
 * can be problematic.  This function attempts to ensure the upper dword has
 * stabilized before returning the 64-bit value.
 *
 * Note that because this function may re-read the register multiple times
 * while waiting for the value to stabilize it should not be used to read
 * any registers where read operations have side effects.
 *
 * Returns the value of the 64-bit register.
 */
u64 xe_mmio_read64_2x32(struct xe_mmio *mmio, struct xe_reg reg)
{
	struct xe_reg reg_udw = { .addr = reg.addr + 0x4 };
	u32 ldw, udw, oldudw, retries;

	reg.addr = xe_mmio_adjusted_addr(mmio, reg.addr);
	reg_udw.addr = xe_mmio_adjusted_addr(mmio, reg_udw.addr);

	/* we shouldn't adjust just one register address */
	xe_tile_assert(mmio->tile, reg_udw.addr == reg.addr + 0x4);

	oldudw = xe_mmio_read32(mmio, reg_udw);
	for (retries = 5; retries; --retries) {
		ldw = xe_mmio_read32(mmio, reg);
		udw = xe_mmio_read32(mmio, reg_udw);

		if (udw == oldudw)
			break;

		oldudw = udw;
	}

	drm_WARN(&mmio->tile->xe->drm, retries == 0,
		 "64-bit read of %#x did not stabilize\n", reg.addr);

	return (u64)udw << 32 | ldw;
}

static int __xe_mmio_wait32(struct xe_mmio *mmio, struct xe_reg reg, u32 mask, u32 val,
			    u32 timeout_us, u32 *out_val, bool atomic, bool expect_match)
{
	ktime_t cur = ktime_get_raw();
	const ktime_t end = ktime_add_us(cur, timeout_us);
	int ret = -ETIMEDOUT;
	s64 wait = 10;
	u32 read;
	bool check;

	for (;;) {
		read = xe_mmio_read32(mmio, reg);

		check = (read & mask) == val;
		if (!expect_match)
			check = !check;

		if (check) {
			ret = 0;
			break;
		}

		cur = ktime_get_raw();
		if (!ktime_before(cur, end))
			break;

		if (ktime_after(ktime_add_us(cur, wait), end))
			wait = ktime_us_delta(end, cur);

		if (atomic)
			udelay(wait);
		else
			usleep_range(wait, wait << 1);
		wait <<= 1;
	}

	if (ret != 0) {
		read = xe_mmio_read32(mmio, reg);

		check = (read & mask) == val;
		if (!expect_match)
			check = !check;

		if (check)
			ret = 0;
	}

	if (out_val)
		*out_val = read;

	return ret;
}

/**
 * xe_mmio_wait32() - Wait for a register to match the desired masked value
 * @mmio: MMIO target
 * @reg: register to read value from
 * @mask: mask to be applied to the value read from the register
 * @val: desired value after applying the mask
 * @timeout_us: time out after this period of time. Wait logic tries to be
 * smart, applying an exponential backoff until @timeout_us is reached.
 * @out_val: if not NULL, points where to store the last unmasked value
 * @atomic: needs to be true if calling from an atomic context
 *
 * This function polls for the desired masked value and returns zero on success
 * or -ETIMEDOUT if timed out.
 *
 * Note that @timeout_us represents the minimum amount of time to wait before
 * giving up. The actual time taken by this function can be a little more than
 * @timeout_us for different reasons, specially in non-atomic contexts. Thus,
 * it is possible that this function succeeds even after @timeout_us has passed.
 */
int xe_mmio_wait32(struct xe_mmio *mmio, struct xe_reg reg, u32 mask, u32 val, u32 timeout_us,
		   u32 *out_val, bool atomic)
{
	return __xe_mmio_wait32(mmio, reg, mask, val, timeout_us, out_val, atomic, true);
}

/**
 * xe_mmio_wait32_not() - Wait for a register to return anything other than the given masked value
 * @mmio: MMIO target
 * @reg: register to read value from
 * @mask: mask to be applied to the value read from the register
 * @val: value not to be matched after applying the mask
 * @timeout_us: time out after this period of time
 * @out_val: if not NULL, points where to store the last unmasked value
 * @atomic: needs to be true if calling from an atomic context
 *
 * This function works exactly like xe_mmio_wait32() with the exception that
 * @val is expected not to be matched.
 */
int xe_mmio_wait32_not(struct xe_mmio *mmio, struct xe_reg reg, u32 mask, u32 val, u32 timeout_us,
		       u32 *out_val, bool atomic)
{
	return __xe_mmio_wait32(mmio, reg, mask, val, timeout_us, out_val, atomic, false);
}