1 files changed, 328 insertions, 226 deletions

diff --git a/drivers/gpu/host1x/job.c b/drivers/gpu/host1x/job.c
index cc807667d8f1..3ed49e1fd933 100644
--- a/drivers/gpu/host1x/job.c
+++ b/drivers/gpu/host1x/job.c
@@ -1,23 +1,14 @@
+// SPDX-License-Identifier: GPL-2.0-only
 /*
  * Tegra host1x Job
  *
- * Copyright (c) 2010-2013, NVIDIA Corporation.
- *
- * 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/>.
+ * Copyright (c) 2010-2015, NVIDIA Corporation.
  */
 
 #include <linux/dma-mapping.h>
 #include <linux/err.h>
+#include <linux/host1x.h>
+#include <linux/iommu.h>
 #include <linux/kref.h>
 #include <linux/module.h>
 #include <linux/scatterlist.h>
@@ -27,27 +18,33 @@
 
 #include "channel.h"
 #include "dev.h"
-#include "host1x_bo.h"
 #include "job.h"
 #include "syncpt.h"
 
+#define HOST1X_WAIT_SYNCPT_OFFSET 0x8
+
 struct host1x_job *host1x_job_alloc(struct host1x_channel *ch,
 				    u32 num_cmdbufs, u32 num_relocs,
-				    u32 num_waitchks)
+				    bool skip_firewall)
 {
 	struct host1x_job *job = NULL;
-	unsigned int num_unpins = num_cmdbufs + num_relocs;
+	unsigned int num_unpins = num_relocs;
+	bool enable_firewall;
 	u64 total;
 	void *mem;
 
+	enable_firewall = IS_ENABLED(CONFIG_TEGRA_HOST1X_FIREWALL) && !skip_firewall;
+
+	if (!enable_firewall)
+		num_unpins += num_cmdbufs;
+
 	/* Check that we're not going to overflow */
 	total = sizeof(struct host1x_job) +
-		num_relocs * sizeof(struct host1x_reloc) +
-		num_unpins * sizeof(struct host1x_job_unpin_data) +
-		num_waitchks * sizeof(struct host1x_waitchk) +
-		num_cmdbufs * sizeof(struct host1x_job_gather) +
-		num_unpins * sizeof(dma_addr_t) +
-		num_unpins * sizeof(u32 *);
+		(u64)num_relocs * sizeof(struct host1x_reloc) +
+		(u64)num_unpins * sizeof(struct host1x_job_unpin_data) +
+		(u64)num_cmdbufs * sizeof(struct host1x_job_cmd) +
+		(u64)num_unpins * sizeof(dma_addr_t) +
+		(u64)num_unpins * sizeof(u32 *);
 	if (total > ULONG_MAX)
 		return NULL;
 
@@ -55,19 +52,19 @@ struct host1x_job *host1x_job_alloc(struct host1x_channel *ch,
 	if (!job)
 		return NULL;
 
+	job->enable_firewall = enable_firewall;
+
 	kref_init(&job->ref);
 	job->channel = ch;
 
 	/* Redistribute memory to the structs  */
 	mem += sizeof(struct host1x_job);
-	job->relocarray = num_relocs ? mem : NULL;
+	job->relocs = num_relocs ? mem : NULL;
 	mem += num_relocs * sizeof(struct host1x_reloc);
 	job->unpins = num_unpins ? mem : NULL;
 	mem += num_unpins * sizeof(struct host1x_job_unpin_data);
-	job->waitchk = num_waitchks ? mem : NULL;
-	mem += num_waitchks * sizeof(struct host1x_waitchk);
-	job->gathers = num_cmdbufs ? mem : NULL;
-	mem += num_cmdbufs * sizeof(struct host1x_job_gather);
+	job->cmds = num_cmdbufs ? mem : NULL;
+	mem += num_cmdbufs * sizeof(struct host1x_job_cmd);
 	job->addr_phys = num_unpins ? mem : NULL;
 
 	job->reloc_addr_phys = job->addr_phys;
@@ -75,17 +72,35 @@ struct host1x_job *host1x_job_alloc(struct host1x_channel *ch,
 
 	return job;
 }
+EXPORT_SYMBOL(host1x_job_alloc);
 
 struct host1x_job *host1x_job_get(struct host1x_job *job)
 {
 	kref_get(&job->ref);
 	return job;
 }
+EXPORT_SYMBOL(host1x_job_get);
 
 static void job_free(struct kref *ref)
 {
 	struct host1x_job *job = container_of(ref, struct host1x_job, ref);
 
+	if (job->release)
+		job->release(job);
+
+	if (job->fence) {
+		/*
+		 * remove_callback is atomic w.r.t. fence signaling, so
+		 * after the call returns, we know that the callback is not
+		 * in execution, and the fence can be safely freed.
+		 */
+		dma_fence_remove_callback(job->fence, &job->fence_cb);
+		dma_fence_put(job->fence);
+	}
+
+	if (job->syncpt)
+		host1x_syncpt_put(job->syncpt);
+
 	kfree(job);
 }
 
@@ -93,182 +108,228 @@ void host1x_job_put(struct host1x_job *job)
 {
 	kref_put(&job->ref, job_free);
 }
+EXPORT_SYMBOL(host1x_job_put);
 
 void host1x_job_add_gather(struct host1x_job *job, struct host1x_bo *bo,
-			   u32 words, u32 offset)
+			   unsigned int words, unsigned int offset)
 {
-	struct host1x_job_gather *cur_gather = &job->gathers[job->num_gathers];
+	struct host1x_job_gather *gather = &job->cmds[job->num_cmds].gather;
 
-	cur_gather->words = words;
-	cur_gather->bo = bo;
-	cur_gather->offset = offset;
-	job->num_gathers++;
-}
+	gather->words = words;
+	gather->bo = bo;
+	gather->offset = offset;
 
-/*
- * NULL an already satisfied WAIT_SYNCPT host method, by patching its
- * args in the command stream. The method data is changed to reference
- * a reserved (never given out or incr) HOST1X_SYNCPT_RESERVED syncpt
- * with a matching threshold value of 0, so is guaranteed to be popped
- * by the host HW.
- */
-static void host1x_syncpt_patch_offset(struct host1x_syncpt *sp,
-				       struct host1x_bo *h, u32 offset)
-{
-	void *patch_addr = NULL;
-
-	/* patch the wait */
-	patch_addr = host1x_bo_kmap(h, offset >> PAGE_SHIFT);
-	if (patch_addr) {
-		host1x_syncpt_patch_wait(sp,
-					 patch_addr + (offset & ~PAGE_MASK));
-		host1x_bo_kunmap(h, offset >> PAGE_SHIFT, patch_addr);
-	} else
-		pr_err("Could not map cmdbuf for wait check\n");
+	job->num_cmds++;
 }
+EXPORT_SYMBOL(host1x_job_add_gather);
 
-/*
- * Check driver supplied waitchk structs for syncpt thresholds
- * that have already been satisfied and NULL the comparison (to
- * avoid a wrap condition in the HW).
- */
-static int do_waitchks(struct host1x_job *job, struct host1x *host,
-		       struct host1x_bo *patch)
+void host1x_job_add_wait(struct host1x_job *job, u32 id, u32 thresh,
+			 bool relative, u32 next_class)
 {
-	int i;
+	struct host1x_job_cmd *cmd = &job->cmds[job->num_cmds];
 
-	/* compare syncpt vs wait threshold */
-	for (i = 0; i < job->num_waitchk; i++) {
-		struct host1x_waitchk *wait = &job->waitchk[i];
-		struct host1x_syncpt *sp =
-			host1x_syncpt_get(host, wait->syncpt_id);
+	cmd->is_wait = true;
+	cmd->wait.id = id;
+	cmd->wait.threshold = thresh;
+	cmd->wait.next_class = next_class;
+	cmd->wait.relative = relative;
 
-		/* validate syncpt id */
-		if (wait->syncpt_id > host1x_syncpt_nb_pts(host))
-			continue;
+	job->num_cmds++;
+}
+EXPORT_SYMBOL(host1x_job_add_wait);
 
-		/* skip all other gathers */
-		if (patch != wait->bo)
-			continue;
+static unsigned int pin_job(struct host1x *host, struct host1x_job *job)
+{
+	unsigned long mask = HOST1X_RELOC_READ | HOST1X_RELOC_WRITE;
+	struct host1x_client *client = job->client;
+	struct device *dev = client->dev;
+	struct host1x_job_gather *g;
+	unsigned int i;
+	int err;
 
-		trace_host1x_syncpt_wait_check(wait->bo, wait->offset,
-					       wait->syncpt_id, wait->thresh,
-					       host1x_syncpt_read_min(sp));
+	job->num_unpins = 0;
 
-		if (host1x_syncpt_is_expired(sp, wait->thresh)) {
-			dev_dbg(host->dev,
-				"drop WAIT id %d (%s) thresh 0x%x, min 0x%x\n",
-				wait->syncpt_id, sp->name, wait->thresh,
-				host1x_syncpt_read_min(sp));
+	for (i = 0; i < job->num_relocs; i++) {
+		struct host1x_reloc *reloc = &job->relocs[i];
+		enum dma_data_direction direction;
+		struct host1x_bo_mapping *map;
+		struct host1x_bo *bo;
 
-			host1x_syncpt_patch_offset(sp, patch, wait->offset);
+		reloc->target.bo = host1x_bo_get(reloc->target.bo);
+		if (!reloc->target.bo) {
+			err = -EINVAL;
+			goto unpin;
 		}
 
-		wait->bo = NULL;
-	}
+		bo = reloc->target.bo;
 
-	return 0;
-}
+		switch (reloc->flags & mask) {
+		case HOST1X_RELOC_READ:
+			direction = DMA_TO_DEVICE;
+			break;
 
-static unsigned int pin_job(struct host1x_job *job)
-{
-	unsigned int i;
+		case HOST1X_RELOC_WRITE:
+			direction = DMA_FROM_DEVICE;
+			break;
 
-	job->num_unpins = 0;
+		case HOST1X_RELOC_READ | HOST1X_RELOC_WRITE:
+			direction = DMA_BIDIRECTIONAL;
+			break;
 
-	for (i = 0; i < job->num_relocs; i++) {
-		struct host1x_reloc *reloc = &job->relocarray[i];
-		struct sg_table *sgt;
-		dma_addr_t phys_addr;
+		default:
+			err = -EINVAL;
+			goto unpin;
+		}
 
-		reloc->target = host1x_bo_get(reloc->target);
-		if (!reloc->target)
+		map = host1x_bo_pin(dev, bo, direction, NULL);
+		if (IS_ERR(map)) {
+			err = PTR_ERR(map);
 			goto unpin;
+		}
 
-		phys_addr = host1x_bo_pin(reloc->target, &sgt);
-		if (!phys_addr)
+		/*
+		 * host1x clients are generally not able to do scatter-gather themselves, so fail
+		 * if the buffer is discontiguous and we fail to map its SG table to a single
+		 * contiguous chunk of I/O virtual memory.
+		 */
+		if (map->chunks > 1) {
+			err = -EINVAL;
 			goto unpin;
+		}
 
-		job->addr_phys[job->num_unpins] = phys_addr;
-		job->unpins[job->num_unpins].bo = reloc->target;
-		job->unpins[job->num_unpins].sgt = sgt;
+		job->addr_phys[job->num_unpins] = map->phys;
+		job->unpins[job->num_unpins].map = map;
 		job->num_unpins++;
 	}
 
-	for (i = 0; i < job->num_gathers; i++) {
-		struct host1x_job_gather *g = &job->gathers[i];
-		struct sg_table *sgt;
-		dma_addr_t phys_addr;
+	/*
+	 * We will copy gathers BO content later, so there is no need to
+	 * hold and pin them.
+	 */
+	if (job->enable_firewall)
+		return 0;
+
+	for (i = 0; i < job->num_cmds; i++) {
+		struct host1x_bo_mapping *map;
+		size_t gather_size = 0;
+		struct scatterlist *sg;
+		unsigned long shift;
+		struct iova *alloc;
+		unsigned int j;
+
+		if (job->cmds[i].is_wait)
+			continue;
+
+		g = &job->cmds[i].gather;
 
 		g->bo = host1x_bo_get(g->bo);
-		if (!g->bo)
+		if (!g->bo) {
+			err = -EINVAL;
 			goto unpin;
+		}
 
-		phys_addr = host1x_bo_pin(g->bo, &sgt);
-		if (!phys_addr)
+		map = host1x_bo_pin(host->dev, g->bo, DMA_TO_DEVICE, NULL);
+		if (IS_ERR(map)) {
+			err = PTR_ERR(map);
 			goto unpin;
+		}
+
+		if (host->domain) {
+			for_each_sgtable_sg(map->sgt, sg, j)
+				gather_size += sg->length;
+
+			gather_size = iova_align(&host->iova, gather_size);
+
+			shift = iova_shift(&host->iova);
+			alloc = alloc_iova(&host->iova, gather_size >> shift,
+					   host->iova_end >> shift, true);
+			if (!alloc) {
+				err = -ENOMEM;
+				goto put;
+			}
+
+			err = iommu_map_sgtable(host->domain, iova_dma_addr(&host->iova, alloc),
+						map->sgt, IOMMU_READ);
+			if (err == 0) {
+				__free_iova(&host->iova, alloc);
+				err = -EINVAL;
+				goto put;
+			}
 
-		job->addr_phys[job->num_unpins] = phys_addr;
-		job->unpins[job->num_unpins].bo = g->bo;
-		job->unpins[job->num_unpins].sgt = sgt;
+			map->phys = iova_dma_addr(&host->iova, alloc);
+			map->size = gather_size;
+		}
+
+		job->addr_phys[job->num_unpins] = map->phys;
+		job->unpins[job->num_unpins].map = map;
 		job->num_unpins++;
+
+		job->gather_addr_phys[i] = map->phys;
 	}
 
-	return job->num_unpins;
+	return 0;
 
+put:
+	host1x_bo_put(g->bo);
 unpin:
 	host1x_job_unpin(job);
-	return 0;
+	return err;
 }
 
-static unsigned int do_relocs(struct host1x_job *job, struct host1x_bo *cmdbuf)
+static int do_relocs(struct host1x_job *job, struct host1x_job_gather *g)
 {
-	int i = 0;
-	u32 last_page = ~0;
-	void *cmdbuf_page_addr = NULL;
+	void *cmdbuf_addr = NULL;
+	struct host1x_bo *cmdbuf = g->bo;
+	unsigned int i;
 
 	/* pin & patch the relocs for one gather */
 	for (i = 0; i < job->num_relocs; i++) {
-		struct host1x_reloc *reloc = &job->relocarray[i];
+		struct host1x_reloc *reloc = &job->relocs[i];
 		u32 reloc_addr = (job->reloc_addr_phys[i] +
-			reloc->target_offset) >> reloc->shift;
+				  reloc->target.offset) >> reloc->shift;
 		u32 *target;
 
 		/* skip all other gathers */
-		if (cmdbuf != reloc->cmdbuf)
+		if (cmdbuf != reloc->cmdbuf.bo)
 			continue;
 
-		if (last_page != reloc->cmdbuf_offset >> PAGE_SHIFT) {
-			if (cmdbuf_page_addr)
-				host1x_bo_kunmap(cmdbuf, last_page,
-						 cmdbuf_page_addr);
+		if (job->enable_firewall) {
+			target = (u32 *)job->gather_copy_mapped +
+					reloc->cmdbuf.offset / sizeof(u32) +
+						g->offset / sizeof(u32);
+			goto patch_reloc;
+		}
 
-			cmdbuf_page_addr = host1x_bo_kmap(cmdbuf,
-					reloc->cmdbuf_offset >> PAGE_SHIFT);
-			last_page = reloc->cmdbuf_offset >> PAGE_SHIFT;
+		if (!cmdbuf_addr) {
+			cmdbuf_addr = host1x_bo_mmap(cmdbuf);
 
-			if (unlikely(!cmdbuf_page_addr)) {
+			if (unlikely(!cmdbuf_addr)) {
 				pr_err("Could not map cmdbuf for relocation\n");
 				return -ENOMEM;
 			}
 		}
 
-		target = cmdbuf_page_addr + (reloc->cmdbuf_offset & ~PAGE_MASK);
+		target = cmdbuf_addr + reloc->cmdbuf.offset;
+patch_reloc:
 		*target = reloc_addr;
 	}
 
-	if (cmdbuf_page_addr)
-		host1x_bo_kunmap(cmdbuf, last_page, cmdbuf_page_addr);
+	if (cmdbuf_addr)
+		host1x_bo_munmap(cmdbuf, cmdbuf_addr);
 
 	return 0;
 }
 
 static bool check_reloc(struct host1x_reloc *reloc, struct host1x_bo *cmdbuf,
-		       unsigned int offset)
+			unsigned int offset)
 {
 	offset *= sizeof(u32);
 
-	if (reloc->cmdbuf != cmdbuf || reloc->cmdbuf_offset != offset)
+	if (reloc->cmdbuf.bo != cmdbuf || reloc->cmdbuf.offset != offset)
+		return false;
+
+	/* relocation shift value validation isn't implemented yet */
+	if (reloc->shift)
 		return false;
 
 	return true;
@@ -281,7 +342,7 @@ struct host1x_firewall {
 	unsigned int num_relocs;
 	struct host1x_reloc *reloc;
 
-	struct host1x_bo *cmdbuf_id;
+	struct host1x_bo *cmdbuf;
 	unsigned int offset;
 
 	u32 words;
@@ -291,25 +352,53 @@ struct host1x_firewall {
 	u32 count;
 };
 
+static int check_register(struct host1x_firewall *fw, unsigned long offset)
+{
+	if (!fw->job->is_addr_reg)
+		return 0;
+
+	if (fw->job->is_addr_reg(fw->dev, fw->class, offset)) {
+		if (!fw->num_relocs)
+			return -EINVAL;
+
+		if (!check_reloc(fw->reloc, fw->cmdbuf, fw->offset))
+			return -EINVAL;
+
+		fw->num_relocs--;
+		fw->reloc++;
+	}
+
+	return 0;
+}
+
+static int check_class(struct host1x_firewall *fw, u32 class)
+{
+	if (!fw->job->is_valid_class) {
+		if (fw->class != class)
+			return -EINVAL;
+	} else {
+		if (!fw->job->is_valid_class(fw->class))
+			return -EINVAL;
+	}
+
+	return 0;
+}
+
 static int check_mask(struct host1x_firewall *fw)
 {
 	u32 mask = fw->mask;
 	u32 reg = fw->reg;
+	int ret;
 
 	while (mask) {
 		if (fw->words == 0)
 			return -EINVAL;
 
 		if (mask & 1) {
-			if (fw->job->is_addr_reg(fw->dev, fw->class, reg)) {
-				if (!fw->num_relocs)
-					return -EINVAL;
-				if (!check_reloc(fw->reloc, fw->cmdbuf_id,
-						 fw->offset))
-					return -EINVAL;
-				fw->reloc++;
-				fw->num_relocs--;
-			}
+			ret = check_register(fw, reg);
+			if (ret < 0)
+				return ret;
+
 			fw->words--;
 			fw->offset++;
 		}
@@ -324,19 +413,16 @@ static int check_incr(struct host1x_firewall *fw)
 {
 	u32 count = fw->count;
 	u32 reg = fw->reg;
+	int ret;
 
 	while (count) {
 		if (fw->words == 0)
 			return -EINVAL;
 
-		if (fw->job->is_addr_reg(fw->dev, fw->class, reg)) {
-			if (!fw->num_relocs)
-				return -EINVAL;
-			if (!check_reloc(fw->reloc, fw->cmdbuf_id, fw->offset))
-				return -EINVAL;
-			fw->reloc++;
-			fw->num_relocs--;
-		}
+		ret = check_register(fw, reg);
+		if (ret < 0)
+			return ret;
+
 		reg++;
 		fw->words--;
 		fw->offset++;
@@ -348,21 +434,17 @@ static int check_incr(struct host1x_firewall *fw)
 
 static int check_nonincr(struct host1x_firewall *fw)
 {
-	int is_addr_reg = fw->job->is_addr_reg(fw->dev, fw->class, fw->reg);
 	u32 count = fw->count;
+	int ret;
 
 	while (count) {
 		if (fw->words == 0)
 			return -EINVAL;
 
-		if (is_addr_reg) {
-			if (!fw->num_relocs)
-				return -EINVAL;
-			if (!check_reloc(fw->reloc, fw->cmdbuf_id, fw->offset))
-				return -EINVAL;
-			fw->reloc++;
-			fw->num_relocs--;
-		}
+		ret = check_register(fw, fw->reg);
+		if (ret < 0)
+			return ret;
+
 		fw->words--;
 		fw->offset++;
 		count--;
@@ -375,13 +457,11 @@ static int validate(struct host1x_firewall *fw, struct host1x_job_gather *g)
 {
 	u32 *cmdbuf_base = (u32 *)fw->job->gather_copy_mapped +
 		(g->offset / sizeof(u32));
+	u32 job_class = fw->class;
 	int err = 0;
 
-	if (!fw->job->is_addr_reg)
-		return 0;
-
 	fw->words = g->words;
-	fw->cmdbuf_id = g->bo;
+	fw->cmdbuf = g->bo;
 	fw->offset = 0;
 
 	while (fw->words && !err) {
@@ -399,7 +479,9 @@ static int validate(struct host1x_firewall *fw, struct host1x_job_gather *g)
 			fw->class = word >> 6 & 0x3ff;
 			fw->mask = word & 0x3f;
 			fw->reg = word >> 16 & 0xfff;
-			err = check_mask(fw);
+			err = check_class(fw, job_class);
+			if (!err)
+				err = check_mask(fw);
 			if (err)
 				goto out;
 			break;
@@ -427,7 +509,6 @@ static int validate(struct host1x_firewall *fw, struct host1x_job_gather *g)
 				goto out;
 			break;
 		case 4:
-		case 5:
 		case 14:
 			break;
 		default:
@@ -436,47 +517,60 @@ static int validate(struct host1x_firewall *fw, struct host1x_job_gather *g)
 		}
 	}
 
-	/* No relocs should remain at this point */
-	if (fw->num_relocs)
-		err = -EINVAL;
-
 out:
 	return err;
 }
 
-static inline int copy_gathers(struct host1x_job *job, struct device *dev)
+static inline int copy_gathers(struct device *host, struct host1x_job *job,
+			       struct device *dev)
 {
 	struct host1x_firewall fw;
 	size_t size = 0;
 	size_t offset = 0;
-	int i;
+	unsigned int i;
 
 	fw.job = job;
 	fw.dev = dev;
-	fw.reloc = job->relocarray;
+	fw.reloc = job->relocs;
 	fw.num_relocs = job->num_relocs;
-	fw.class = 0;
+	fw.class = job->class;
+
+	for (i = 0; i < job->num_cmds; i++) {
+		struct host1x_job_gather *g;
+
+		if (job->cmds[i].is_wait)
+			continue;
+
+		g = &job->cmds[i].gather;
 
-	for (i = 0; i < job->num_gathers; i++) {
-		struct host1x_job_gather *g = &job->gathers[i];
 		size += g->words * sizeof(u32);
 	}
 
-	job->gather_copy_mapped = dma_alloc_writecombine(dev, size,
-							 &job->gather_copy,
-							 GFP_KERNEL);
-	if (!job->gather_copy_mapped) {
-		int err = PTR_ERR(job->gather_copy_mapped);
-		job->gather_copy_mapped = NULL;
-		return err;
-	}
+	/*
+	 * Try a non-blocking allocation from a higher priority pools first,
+	 * as awaiting for the allocation here is a major performance hit.
+	 */
+	job->gather_copy_mapped = dma_alloc_wc(host, size, &job->gather_copy,
+					       GFP_NOWAIT);
+
+	/* the higher priority allocation failed, try the generic-blocking */
+	if (!job->gather_copy_mapped)
+		job->gather_copy_mapped = dma_alloc_wc(host, size,
+						       &job->gather_copy,
+						       GFP_KERNEL);
+	if (!job->gather_copy_mapped)
+		return -ENOMEM;
 
 	job->gather_copy_size = size;
 
-	for (i = 0; i < job->num_gathers; i++) {
-		struct host1x_job_gather *g = &job->gathers[i];
+	for (i = 0; i < job->num_cmds; i++) {
+		struct host1x_job_gather *g;
 		void *gather;
 
+		if (job->cmds[i].is_wait)
+			continue;
+		g = &job->cmds[i].gather;
+
 		/* Copy the gather */
 		gather = host1x_bo_mmap(g->bo);
 		memcpy(job->gather_copy_mapped + offset, gather + g->offset,
@@ -494,6 +588,10 @@ static inline int copy_gathers(struct host1x_job *job, struct device *dev)
 		offset += g->words * sizeof(u32);
 	}
 
+	/* No relocs should remain at this point */
+	if (fw.num_relocs)
+		return -EINVAL;
+
 	return 0;
 }
 
@@ -502,84 +600,88 @@ int host1x_job_pin(struct host1x_job *job, struct device *dev)
 	int err;
 	unsigned int i, j;
 	struct host1x *host = dev_get_drvdata(dev->parent);
-	DECLARE_BITMAP(waitchk_mask, host1x_syncpt_nb_pts(host));
-
-	bitmap_zero(waitchk_mask, host1x_syncpt_nb_pts(host));
-	for (i = 0; i < job->num_waitchk; i++) {
-		u32 syncpt_id = job->waitchk[i].syncpt_id;
-		if (syncpt_id < host1x_syncpt_nb_pts(host))
-			set_bit(syncpt_id, waitchk_mask);
-	}
-
-	/* get current syncpt values for waitchk */
-	for_each_set_bit(i, waitchk_mask, host1x_syncpt_nb_pts(host))
-		host1x_syncpt_load(host->syncpt + i);
 
 	/* pin memory */
-	err = pin_job(job);
-	if (!err)
+	err = pin_job(host, job);
+	if (err)
 		goto out;
 
+	if (job->enable_firewall) {
+		err = copy_gathers(host->dev, job, dev);
+		if (err)
+			goto out;
+	}
+
 	/* patch gathers */
-	for (i = 0; i < job->num_gathers; i++) {
-		struct host1x_job_gather *g = &job->gathers[i];
+	for (i = 0; i < job->num_cmds; i++) {
+		struct host1x_job_gather *g;
+
+		if (job->cmds[i].is_wait)
+			continue;
+		g = &job->cmds[i].gather;
 
 		/* process each gather mem only once */
 		if (g->handled)
 			continue;
 
-		g->base = job->gather_addr_phys[i];
+		/* copy_gathers() sets gathers base if firewall is enabled */
+		if (!job->enable_firewall)
+			g->base = job->gather_addr_phys[i];
 
-		for (j = 0; j < job->num_gathers; j++)
-			if (job->gathers[j].bo == g->bo)
-				job->gathers[j].handled = true;
-
-		err = do_relocs(job, g->bo);
-		if (err)
-			break;
+		for (j = i + 1; j < job->num_cmds; j++) {
+			if (!job->cmds[j].is_wait &&
+			    job->cmds[j].gather.bo == g->bo) {
+				job->cmds[j].gather.handled = true;
+				job->cmds[j].gather.base = g->base;
+			}
+		}
 
-		err = do_waitchks(job, host, g->bo);
+		err = do_relocs(job, g);
 		if (err)
 			break;
 	}
 
-	if (IS_ENABLED(CONFIG_TEGRA_HOST1X_FIREWALL) && !err) {
-		err = copy_gathers(job, dev);
-		if (err) {
-			host1x_job_unpin(job);
-			return err;
-		}
-	}
-
 out:
+	if (err)
+		host1x_job_unpin(job);
 	wmb();
 
 	return err;
 }
+EXPORT_SYMBOL(host1x_job_pin);
 
 void host1x_job_unpin(struct host1x_job *job)
 {
+	struct host1x *host = dev_get_drvdata(job->channel->dev->parent);
 	unsigned int i;
 
 	for (i = 0; i < job->num_unpins; i++) {
-		struct host1x_job_unpin_data *unpin = &job->unpins[i];
-		host1x_bo_unpin(unpin->bo, unpin->sgt);
-		host1x_bo_put(unpin->bo);
+		struct host1x_bo_mapping *map = job->unpins[i].map;
+		struct host1x_bo *bo = map->bo;
+
+		if (!job->enable_firewall && map->size && host->domain) {
+			iommu_unmap(host->domain, job->addr_phys[i], map->size);
+			free_iova(&host->iova, iova_pfn(&host->iova, job->addr_phys[i]));
+		}
+
+		host1x_bo_unpin(map);
+		host1x_bo_put(bo);
 	}
+
 	job->num_unpins = 0;
 
 	if (job->gather_copy_size)
-		dma_free_writecombine(job->channel->dev, job->gather_copy_size,
-				      job->gather_copy_mapped,
-				      job->gather_copy);
+		dma_free_wc(host->dev, job->gather_copy_size,
+			    job->gather_copy_mapped, job->gather_copy);
 }
+EXPORT_SYMBOL(host1x_job_unpin);
 
 /*
  * Debug routine used to dump job entries
  */
 void host1x_job_dump(struct device *dev, struct host1x_job *job)
 {
-	dev_dbg(dev, "    SYNCPT_ID   %d\n", job->syncpt_id);
+	dev_dbg(dev, "    SYNCPT_ID   %d\n", job->syncpt->id);
 	dev_dbg(dev, "    SYNCPT_VAL  %d\n", job->syncpt_end);
 	dev_dbg(dev, "    FIRST_GET   0x%x\n", job->first_get);
 	dev_dbg(dev, "    TIMEOUT     %d\n", job->timeout);