drm/nouveau: Add DRM driver for NVIDIA GPUs

This adds a drm/kms staging non-API stable driver for GPUs from NVIDIA.

This driver is a KMS-based driver and requires a compatible nouveau
userspace libdrm and nouveau X.org driver.

This driver requires firmware files not available in this kernel tree,
interested parties can find them via the nouveau project git archive.

This driver is reverse engineered, and is in no way supported by nVidia.

Support for nearly the complete range of nvidia hw from nv04->g80 (nv50)
is available, and the kms driver should support driving nearly all
output types (displayport is under development still) along with supporting
suspend/resume.

This work is all from the upstream nouveau project found at
nouveau.freedesktop.org.

The original authors list from nouveau git tree is:
Anssi Hannula <anssi.hannula@iki.fi>
Ben Skeggs <bskeggs@redhat.com>
Francisco Jerez <currojerez@riseup.net>
Maarten Maathuis <madman2003@gmail.com>
Marcin Kościelnicki <koriakin@0x04.net>
Matthew Garrett <mjg@redhat.com>
Matt Parnell <mparnell@gmail.com>
Patrice Mandin <patmandin@gmail.com>
Pekka Paalanen <pq@iki.fi>
Xavier Chantry <shiningxc@gmail.com>
along with project founder Stephane Marchesin <marchesin@icps.u-strasbg.fr>

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

1 files changed, 157 insertions, 0 deletions

diff --git a/drivers/gpu/drm/nouveau/nouveau_dma.h b/drivers/gpu/drm/nouveau/nouveau_dma.h
new file mode 100644
index 000000000000..04e85d8f757e
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nouveau_dma.h
@@ -0,0 +1,157 @@
+/*
+ * Copyright (C) 2007 Ben Skeggs.
+ * All Rights Reserved.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining
+ * a copy of this software and associated documentation files (the
+ * "Software"), to deal in the Software without restriction, including
+ * without limitation the rights to use, copy, modify, merge, publish,
+ * distribute, sublicense, and/or sell copies of the Software, and to
+ * permit persons to whom the Software is furnished to do so, subject to
+ * the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the
+ * next paragraph) shall be included in all copies or substantial
+ * portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
+ * IN NO EVENT SHALL THE COPYRIGHT OWNER(S) AND/OR ITS SUPPLIERS BE
+ * LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION
+ * OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION
+ * WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#ifndef __NOUVEAU_DMA_H__
+#define __NOUVEAU_DMA_H__
+
+#ifndef NOUVEAU_DMA_DEBUG
+#define NOUVEAU_DMA_DEBUG 0
+#endif
+
+/*
+ * There's a hw race condition where you can't jump to your PUT offset,
+ * to avoid this we jump to offset + SKIPS and fill the difference with
+ * NOPs.
+ *
+ * xf86-video-nv configures the DMA fetch size to 32 bytes, and uses
+ * a SKIPS value of 8.  Lets assume that the race condition is to do
+ * with writing into the fetch area, we configure a fetch size of 128
+ * bytes so we need a larger SKIPS value.
+ */
+#define NOUVEAU_DMA_SKIPS (128 / 4)
+
+/* Hardcoded object assignments to subchannels (subchannel id). */
+enum {
+	NvSubM2MF	= 0,
+	NvSub2D		= 1,
+	NvSubCtxSurf2D  = 1,
+	NvSubGdiRect    = 2,
+	NvSubImageBlit  = 3
+};
+
+/* Object handles. */
+enum {
+	NvM2MF		= 0x80000001,
+	NvDmaFB		= 0x80000002,
+	NvDmaTT		= 0x80000003,
+	NvDmaVRAM	= 0x80000004,
+	NvDmaGART	= 0x80000005,
+	NvNotify0       = 0x80000006,
+	Nv2D		= 0x80000007,
+	NvCtxSurf2D	= 0x80000008,
+	NvRop		= 0x80000009,
+	NvImagePatt	= 0x8000000a,
+	NvClipRect	= 0x8000000b,
+	NvGdiRect	= 0x8000000c,
+	NvImageBlit	= 0x8000000d,
+
+	/* G80+ display objects */
+	NvEvoVRAM	= 0x01000000,
+	NvEvoFB16	= 0x01000001,
+	NvEvoFB32	= 0x01000002
+};
+
+#define NV_MEMORY_TO_MEMORY_FORMAT                                    0x00000039
+#define NV_MEMORY_TO_MEMORY_FORMAT_NAME                               0x00000000
+#define NV_MEMORY_TO_MEMORY_FORMAT_SET_REF                            0x00000050
+#define NV_MEMORY_TO_MEMORY_FORMAT_NOP                                0x00000100
+#define NV_MEMORY_TO_MEMORY_FORMAT_NOTIFY                             0x00000104
+#define NV_MEMORY_TO_MEMORY_FORMAT_NOTIFY_STYLE_WRITE                 0x00000000
+#define NV_MEMORY_TO_MEMORY_FORMAT_NOTIFY_STYLE_WRITE_LE_AWAKEN       0x00000001
+#define NV_MEMORY_TO_MEMORY_FORMAT_DMA_NOTIFY                         0x00000180
+#define NV_MEMORY_TO_MEMORY_FORMAT_DMA_SOURCE                         0x00000184
+#define NV_MEMORY_TO_MEMORY_FORMAT_OFFSET_IN                          0x0000030c
+
+#define NV50_MEMORY_TO_MEMORY_FORMAT                                  0x00005039
+#define NV50_MEMORY_TO_MEMORY_FORMAT_UNK200                           0x00000200
+#define NV50_MEMORY_TO_MEMORY_FORMAT_UNK21C                           0x0000021c
+#define NV50_MEMORY_TO_MEMORY_FORMAT_OFFSET_IN_HIGH                   0x00000238
+#define NV50_MEMORY_TO_MEMORY_FORMAT_OFFSET_OUT_HIGH                  0x0000023c
+
+static __must_check inline int
+RING_SPACE(struct nouveau_channel *chan, int size)
+{
+	if (chan->dma.free < size) {
+		int ret;
+
+		ret = nouveau_dma_wait(chan, size);
+		if (ret)
+			return ret;
+	}
+
+	chan->dma.free -= size;
+	return 0;
+}
+
+static inline void
+OUT_RING(struct nouveau_channel *chan, int data)
+{
+	if (NOUVEAU_DMA_DEBUG) {
+		NV_INFO(chan->dev, "Ch%d/0x%08x: 0x%08x\n",
+			chan->id, chan->dma.cur << 2, data);
+	}
+
+	nouveau_bo_wr32(chan->pushbuf_bo, chan->dma.cur++, data);
+}
+
+extern void
+OUT_RINGp(struct nouveau_channel *chan, const void *data, unsigned nr_dwords);
+
+static inline void
+BEGIN_RING(struct nouveau_channel *chan, int subc, int mthd, int size)
+{
+	OUT_RING(chan, (subc << 13) | (size << 18) | mthd);
+}
+
+#define WRITE_PUT(val) do {                                                    \
+	DRM_MEMORYBARRIER();                                                   \
+	nouveau_bo_rd32(chan->pushbuf_bo, 0);                                  \
+	nvchan_wr32(chan, chan->user_put, ((val) << 2) + chan->pushbuf_base);  \
+} while (0)
+
+static inline void
+FIRE_RING(struct nouveau_channel *chan)
+{
+	if (NOUVEAU_DMA_DEBUG) {
+		NV_INFO(chan->dev, "Ch%d/0x%08x: PUSH!\n",
+			chan->id, chan->dma.cur << 2);
+	}
+
+	if (chan->dma.cur == chan->dma.put)
+		return;
+	chan->accel_done = true;
+
+	WRITE_PUT(chan->dma.cur);
+	chan->dma.put = chan->dma.cur;
+}
+
+static inline void
+WIND_RING(struct nouveau_channel *chan)
+{
+	chan->dma.cur = chan->dma.put;
+}
+
+#endif