tegrakernel/kernel/nvidia/drivers/video/tegra/host/iommu_context_dev.c

/*
 * Host1x Application Specific Virtual Memory
 *
 * Copyright (c) 2015-2018, NVIDIA Corporation.  All rights reserved.
 *
 * This program is free software; you can redistribute it and/or modify it
 * under the terms and conditions of the GNU General Public License,
 * version 2, as published by the Free Software Foundation.
 *
 * This program is distributed in the hope it will be useful, but WITHOUT
 * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
 * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
 * more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with this program.  If not, see <http://www.gnu.org/licenses/>.
 */

#include <linux/dma-mapping.h>
#include <linux/of_platform.h>
#include <linux/of_device.h>
#include <linux/module.h>
#include <linux/iommu.h>
#include <linux/mutex.h>
#include <linux/slab.h>
#include <linux/list.h>
#include <linux/of.h>
#include <linux/version.h>
#include <linux/dma-buf.h>
#include <linux/nvhost.h>

#include <iommu_context_dev.h>

#include "chip_support.h"

static struct of_device_id tegra_iommu_context_dev_of_match[] = {
	{ .compatible = "nvidia,tegra186-iommu-context" },
	{ },
};

struct iommu_static_mapping {
	struct list_head list;
	dma_addr_t paddr;
	void *vaddr;
	size_t size;
};

struct iommu_ctx {
	struct nvhost_device_data pdata;
	struct platform_device *pdev;
	struct list_head list;
	struct device_dma_parameters dma_parms;
	bool allocated;
	void *prev_identifier;
};

static LIST_HEAD(iommu_ctx_list);
static LIST_HEAD(iommu_static_mappings_list);
static DEFINE_MUTEX(iommu_ctx_list_mutex);

struct platform_device *iommu_context_dev_allocate(void *identifier)
{
	struct iommu_ctx *ctx, *ctx_new = NULL;
	bool dirty = false;

	mutex_lock(&iommu_ctx_list_mutex);
	/*
	 * First check if we have same identifier stashed into
	 * some context device
	 * If yes, use that context device since it will have all
	 * the mappings stashed too
	 */
	list_for_each_entry(ctx, &iommu_ctx_list, list) {
		if (!ctx->allocated && identifier == ctx->prev_identifier) {
			ctx->allocated = true;
			mutex_unlock(&iommu_ctx_list_mutex);
			return ctx->pdev;
		}
	}

	/*
	 * Otherwise, find a device which does not have any identifier stashed
	 * If there is no device left without identifier stashed, use any of
	 * the free device and explicitly remove all the stashings from it
	 */
	list_for_each_entry(ctx, &iommu_ctx_list, list) {
		if (!ctx->allocated && !ctx_new) {
			ctx_new = ctx;
			dirty = true;
		}
		if (!ctx->allocated && !ctx->prev_identifier) {
			ctx_new = ctx;
			dirty = false;
			break;
		}
	}

	if (ctx_new) {
		if (dirty) {
			/*
			 * Ensure that all stashed mappings are removed from this context device
			 * before this context device gets reassigned to some other process
			 */
			dma_buf_release_stash(&ctx_new->pdev->dev);
		}

		ctx_new->prev_identifier = identifier;
		ctx_new->allocated = true;
		mutex_unlock(&iommu_ctx_list_mutex);
		return ctx_new->pdev;
	}

	mutex_unlock(&iommu_ctx_list_mutex);

	return NULL;
}

void iommu_context_dev_release(struct platform_device *pdev)
{
	struct iommu_ctx *ctx = platform_get_drvdata(pdev);

	mutex_lock(&iommu_ctx_list_mutex);
	ctx->allocated = false;
	mutex_unlock(&iommu_ctx_list_mutex);
}

static int __iommu_context_dev_map_static(struct platform_device *pdev,
					  struct iommu_static_mapping *mapping)
{
#if LINUX_VERSION_CODE < KERNEL_VERSION(4, 9, 0)
	DEFINE_DMA_ATTRS(attrs);
#endif
	const struct dma_map_ops *ops = get_dma_ops(&pdev->dev);
	int num_pages = DIV_ROUND_UP(mapping->size, PAGE_SIZE);
#if LINUX_VERSION_CODE < KERNEL_VERSION(4, 4, 0)
	dma_addr_t base_iova = mapping->paddr;
#endif
	int i, err;

#if LINUX_VERSION_CODE < KERNEL_VERSION(4, 9, 0)
	dma_set_attr(DMA_ATTR_SKIP_IOVA_GAP, &attrs);
#endif

#if LINUX_VERSION_CODE < KERNEL_VERSION(4, 4, 0)
	/* allocate the area first */
	base_iova = ops->iova_alloc_at(&pdev->dev, &mapping->paddr,
				       num_pages * PAGE_SIZE, &attrs);
	err = dma_mapping_error(&pdev->dev, base_iova);
	if (err) {
		dev_warn(&pdev->dev, "failed to allocate space\n");
		return err;
	}
#endif

	/* map each page of the buffer to this ctx dev */
	for (i = 0; i < num_pages; i++) {
		dma_addr_t iova = mapping->paddr + PAGE_SIZE * i;
		void *va = mapping->vaddr + PAGE_SIZE * i;
		struct page *page;

		/* hack: the memory may be allocated from vmalloc
		 * pool. this happens if dma_alloc_attrs() is called
		 * with a device that is attached to smmu domain. */

		page = virt_addr_valid(va) ? virt_to_page(va) :
			vmalloc_to_page(va);

#if LINUX_VERSION_CODE < KERNEL_VERSION(4, 4, 0)
		iova = ops->map_page_at(&pdev->dev, page, iova,
			     (unsigned long)va & ~PAGE_MASK, PAGE_SIZE,
			     DMA_BIDIRECTIONAL, NULL);
#elif LINUX_VERSION_CODE < KERNEL_VERSION(4, 9, 0)
		iova = ops->map_at(&pdev->dev, iova, page_to_phys(page),
					PAGE_SIZE, DMA_BIDIRECTIONAL, &attrs);
#else
		iova = ops->map_at(&pdev->dev, iova, page_to_phys(page),
					PAGE_SIZE, DMA_BIDIRECTIONAL,
					DMA_ATTR_SKIP_IOVA_GAP);
#endif
		err = dma_mapping_error(&pdev->dev, iova);
		if (err) {
			dev_warn(&pdev->dev, "failed to map page\n");
			return err;
		}
	}

	return 0;
}

int iommu_context_dev_map_static(void *vaddr, dma_addr_t paddr, size_t size)
{
	struct iommu_static_mapping *mapping;
	struct iommu_ctx *ctx;

	mapping = kzalloc(sizeof(*mapping), GFP_KERNEL);
	if (!mapping) {
		dev_err(NULL, "%s: could not allocate mapping struct\n",
			   __func__);
		return -ENOMEM;
	}

	INIT_LIST_HEAD(&mapping->list);
	mapping->vaddr = vaddr;
	mapping->paddr = paddr;
	mapping->size = size;

	mutex_lock(&iommu_ctx_list_mutex);

	list_add_tail(&mapping->list, &iommu_static_mappings_list);

	/* go through context devices */
	list_for_each_entry(ctx, &iommu_ctx_list, list)
		__iommu_context_dev_map_static(ctx->pdev, mapping);

	mutex_unlock(&iommu_ctx_list_mutex);

	return 0;
}

static int iommu_context_dev_probe(struct platform_device *pdev)
{
	struct iommu_static_mapping *mapping;
	struct iommu_ctx *ctx;

	if (!pdev->dev.archdata.iommu) {
		dev_err(&pdev->dev, "iommu is not enabled for context device. aborting.");
		return -ENOSYS;
	}

	dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(40));

	ctx = devm_kzalloc(&pdev->dev, sizeof(*ctx), GFP_KERNEL);
	if (!ctx) {
		dev_err(&pdev->dev,
			   "%s: could not allocate iommu ctx\n", __func__);
		return -ENOMEM;
	}

	INIT_LIST_HEAD(&ctx->list);
	ctx->pdev = pdev;

	mutex_lock(&iommu_ctx_list_mutex);
	list_add_tail(&ctx->list, &iommu_ctx_list);

	/* map all static buffers to this address space */
	list_for_each_entry(mapping, &iommu_static_mappings_list, list)
		__iommu_context_dev_map_static(pdev, mapping);

	mutex_unlock(&iommu_ctx_list_mutex);

	platform_set_drvdata(pdev, ctx);

	pdev->dev.dma_parms = &ctx->dma_parms;
	dma_set_max_seg_size(&pdev->dev, UINT_MAX);

	/* flag required to handle stashings in context devices */
	pdev->dev.context_dev = true;

	dev_info(&pdev->dev, "initialized (streamid=%d)",
		 iommu_get_hwid(pdev->dev.archdata.iommu, &pdev->dev, 0));

	if (vm_op().init_syncpt_interface)
		vm_op().init_syncpt_interface(pdev);

	return 0;
}

static int __exit iommu_context_dev_remove(struct platform_device *pdev)
{
	struct iommu_ctx *ctx = platform_get_drvdata(pdev);

	mutex_lock(&iommu_ctx_list_mutex);
	list_del(&ctx->list);
	mutex_unlock(&iommu_ctx_list_mutex);

	return 0;
}

static struct platform_driver iommu_context_dev_driver = {
	.probe = iommu_context_dev_probe,
	.remove = __exit_p(iommu_context_dev_remove),
	.driver = {
		.owner = THIS_MODULE,
		.name = "iommu_context_dev",
#ifdef CONFIG_OF
		.of_match_table = tegra_iommu_context_dev_of_match,
#endif
	},
};

static int __init iommu_context_dev_init(void)
{
	return platform_driver_register(&iommu_context_dev_driver);
}

static void __exit iommu_context_dev_exit(void)
{
	platform_driver_unregister(&iommu_context_dev_driver);
}

module_init(iommu_context_dev_init);
module_exit(iommu_context_dev_exit);