video: tegra: host: Simplify robustness

nvhost attempts to purge all buffers from timed out contexts. This is
very error prone, and not necessary. Change behavior so that only the
hung job, and the immediately following jobs from same context are
purged. This simplifies code, and corrects bug where the push buffer
modifications caused panics.

Bug 982946

Change-Id: Ifb26484cf02ef40f8d5b20338eebc0a731f453cf
Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com>
Reviewed-on: http://git-master/r/102234
Reviewed-by: Rohan Somvanshi <rsomvanshi@nvidia.com>
Tested-by: Rohan Somvanshi <rsomvanshi@nvidia.com>

1 files changed, 2 insertions, 141 deletions

diff --git a/drivers/video/tegra/host/host1x/host1x_cdma.c b/drivers/video/tegra/host/host1x/host1x_cdma.c
index 813b35476f5b..fcb1f05f0025 100644
--- a/drivers/video/tegra/host/host1x/host1x_cdma.c
+++ b/drivers/video/tegra/host/host1x/host1x_cdma.c
@@ -193,96 +193,25 @@ static u32 push_buffer_putptr(struct push_buffer *pb)
  */
 
 /**
- * Init timeout and syncpt incr buffer resources
+ * Init timeout resources
  */
 static int cdma_timeout_init(struct nvhost_cdma *cdma,
 				 u32 syncpt_id)
 {
-	struct nvhost_master *dev = cdma_to_dev(cdma);
-	struct nvmap_client *nvmap = cdma_to_nvmap(cdma);
-	struct syncpt_buffer *sb = &cdma->syncpt_buffer;
-	struct nvhost_channel *ch = cdma_to_channel(cdma);
-	u32 i = 0;
-
 	if (syncpt_id == NVSYNCPT_INVALID)
 		return -EINVAL;
 
-	/* allocate and map syncpt incr memory */
-	sb->mem = nvmap_alloc(nvmap,
-			(SYNCPT_INCR_BUFFER_SIZE_WORDS * sizeof(u32)), 32,
-			NVMAP_HANDLE_WRITE_COMBINE, 0);
-	if (IS_ERR_OR_NULL(sb->mem)) {
-		sb->mem = NULL;
-		goto fail;
-	}
-	sb->mapped = nvmap_mmap(sb->mem);
-	if (sb->mapped == NULL)
-		goto fail;
-
-	/* pin syncpt buffer and get physical address */
-	sb->phys = nvmap_pin(nvmap, sb->mem);
-	if (sb->phys >= 0xfffff000) {
-		sb->phys = 0;
-		goto fail;
-	}
-
-	dev_dbg(&dev->dev->dev, "%s: SYNCPT_INCR buffer at 0x%x\n",
-		 __func__, sb->phys);
-
-	sb->words_per_incr = (syncpt_id == NVSYNCPT_3D) ? 5 : 3;
-	sb->incr_per_buffer = (SYNCPT_INCR_BUFFER_SIZE_WORDS /
-				sb->words_per_incr);
-
-	/* init buffer with SETCL and INCR_SYNCPT methods */
-	while (i < sb->incr_per_buffer) {
-		sb->mapped[i++] = nvhost_opcode_setclass(NV_HOST1X_CLASS_ID,
-						0, 0);
-		sb->mapped[i++] = nvhost_opcode_imm_incr_syncpt(
-						NV_SYNCPT_IMMEDIATE,
-						syncpt_id);
-		if (syncpt_id == NVSYNCPT_3D) {
-			/* also contains base increments */
-			sb->mapped[i++] = nvhost_opcode_nonincr(
-						NV_CLASS_HOST_INCR_SYNCPT_BASE,
-						1);
-			sb->mapped[i++] = nvhost_class_host_incr_syncpt_base(
-						NVWAITBASE_3D, 1);
-		}
-		sb->mapped[i++] = nvhost_opcode_setclass(ch->dev->class,
-						0, 0);
-	}
-	wmb();
-
 	INIT_DELAYED_WORK(&cdma->timeout.wq, cdma_timeout_handler);
 	cdma->timeout.initialized = true;
 
 	return 0;
-fail:
-	cdma_op().timeout_destroy(cdma);
-	return -ENOMEM;
 }
 
 /**
- * Clean up timeout syncpt buffer resources
+ * Clean up timeout resources
  */
 static void cdma_timeout_destroy(struct nvhost_cdma *cdma)
 {
-	struct nvmap_client *nvmap = cdma_to_nvmap(cdma);
-	struct syncpt_buffer *sb = &cdma->syncpt_buffer;
-
-	if (sb->mapped)
-		nvmap_munmap(sb->mem, sb->mapped);
-
-	if (sb->phys != 0)
-		nvmap_unpin(nvmap, sb->mem);
-
-	if (sb->mem)
-		nvmap_free(nvmap, sb->mem);
-
-	sb->mem = NULL;
-	sb->mapped = NULL;
-	sb->phys = 0;
-
 	if (cdma->timeout.initialized)
 		cancel_delayed_work(&cdma->timeout.wq);
 	cdma->timeout.initialized = false;
@@ -326,73 +255,6 @@ static void cdma_timeout_cpu_incr(struct nvhost_cdma *cdma, u32 getptr,
 }
 
 /**
- * This routine is called at the point we transition back into a timed
- * ctx. The syncpts are incremented via pushbuffer with a flag indicating
- * whether there's a CTXSAVE that should be still executed (for the
- * preceding HW ctx).
- */
-static void cdma_timeout_pb_incr(struct nvhost_cdma *cdma, u32 getptr,
-				u32 syncpt_incrs, u32 nr_slots,
-				bool exec_ctxsave)
-{
-	struct nvhost_master *dev = cdma_to_dev(cdma);
-	struct syncpt_buffer *sb = &cdma->syncpt_buffer;
-	struct push_buffer *pb = &cdma->push_buffer;
-	struct host1x_hwctx *hwctx = to_host1x_hwctx(cdma->timeout.ctx);
-	u32 getidx, *p;
-
-	/* should have enough slots to incr to desired count */
-	BUG_ON(syncpt_incrs > (nr_slots * sb->incr_per_buffer));
-
-	getidx = getptr - pb->phys;
-	if (exec_ctxsave) {
-		/* don't disrupt the CTXSAVE of a good/non-timed out ctx */
-		nr_slots -= hwctx->save_slots;
-		syncpt_incrs -= hwctx->save_incrs;
-
-		getidx += (hwctx->save_slots * 8);
-		getidx &= (PUSH_BUFFER_SIZE - 1);
-
-		dev_dbg(&dev->dev->dev,
-			"%s: exec CTXSAVE of prev ctx (slots %d, incrs %d)\n",
-			__func__, nr_slots, syncpt_incrs);
-	}
-
-	while (syncpt_incrs) {
-		u32 incrs, count;
-
-		/* GATHER count are incrs * number of DWORDs per incr */
-		incrs = min(syncpt_incrs, sb->incr_per_buffer);
-		count = incrs * sb->words_per_incr;
-
-		p = (u32 *)((u32)pb->mapped + getidx);
-		*(p++) = nvhost_opcode_gather(count);
-		*(p++) = sb->phys;
-
-		dev_dbg(&dev->dev->dev,
-			"%s: GATHER at 0x%x, from 0x%x, dcount = %d\n",
-			__func__,
-			pb->phys + getidx, sb->phys,
-			(incrs * sb->words_per_incr));
-
-		syncpt_incrs -= incrs;
-		getidx = (getidx + 8) & (PUSH_BUFFER_SIZE - 1);
-		nr_slots--;
-	}
-
-	/* NOP remaining slots */
-	while (nr_slots--) {
-		p = (u32 *)((u32)pb->mapped + getidx);
-		*(p++) = NVHOST_OPCODE_NOOP;
-		*(p++) = NVHOST_OPCODE_NOOP;
-		dev_dbg(&dev->dev->dev, "%s: NOP at 0x%x\n",
-			__func__, pb->phys + getidx);
-		getidx = (getidx + 8) & (PUSH_BUFFER_SIZE - 1);
-	}
-	wmb();
-}
-
-/**
  * Start channel DMA
  */
 static void cdma_start(struct nvhost_cdma *cdma)
@@ -653,7 +515,6 @@ int host1x_init_cdma_support(struct nvhost_chip_support *op)
 	op->cdma.timeout_teardown_begin = cdma_timeout_teardown_begin;
 	op->cdma.timeout_teardown_end = cdma_timeout_teardown_end;
 	op->cdma.timeout_cpu_incr = cdma_timeout_cpu_incr;
-	op->cdma.timeout_pb_incr = cdma_timeout_pb_incr;
 
 	op->push_buffer.reset = push_buffer_reset;
 	op->push_buffer.init = push_buffer_init;