drm/vmwgfx: Add basic support for SVGA3

SVGA3 is the next version of our PCI device. Some of the changes
include using MMIO for register accesses instead of ioports,
deprecating the FIFO MMIO and removing a lot of the old and
legacy functionality. SVGA3 doesn't support guest backed
objects right now so everything except 3D is working.

v2: Fixes all the static analyzer warnings

Signed-off-by: Zack Rusin <zackr@vmware.com>
Cc: Martin Krastev <krastevm@vmware.com>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210505191007.305872-1-zackr@vmware.com

1 files changed, 96 insertions, 24 deletions

diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h
index 2fb6898ceca9..dcdbc59c2b5f 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h
@@ -66,6 +66,7 @@
 #define VMWGFX_ENABLE_SCREEN_TARGET_OTABLE 1
 
 #define VMWGFX_PCI_ID_SVGA2              0x0405
+#define VMWGFX_PCI_ID_SVGA3              0x0406
 
 /*
  * Perhaps we should have sysfs entries for these.
@@ -284,7 +285,6 @@ struct vmw_fifo_state {
 	uint32_t capabilities;
 	struct mutex fifo_mutex;
 	struct rw_semaphore rwsem;
-	bool dx;
 };
 
 /**
@@ -485,14 +485,14 @@ struct vmw_private {
 	struct drm_device drm;
 	struct ttm_device bdev;
 
-	struct vmw_fifo_state fifo;
-
 	struct drm_vma_offset_manager vma_manager;
+	unsigned long pci_id;
 	u32 vmw_chipset;
 	resource_size_t io_start;
 	resource_size_t vram_start;
 	resource_size_t vram_size;
 	resource_size_t prim_bb_mem;
+	void __iomem *rmmio;
 	u32 *fifo_mem;
 	resource_size_t fifo_mem_size;
 	uint32_t fb_max_width;
@@ -623,6 +623,7 @@ struct vmw_private {
 	 */
 	struct vmw_otable_batch otable_batch;
 
+	struct vmw_fifo_state *fifo;
 	struct vmw_cmdbuf_man *cman;
 	DECLARE_BITMAP(irqthread_pending, VMW_IRQTHREAD_MAX);
 
@@ -646,6 +647,14 @@ static inline struct vmw_fpriv *vmw_fpriv(struct drm_file *file_priv)
 }
 
 /*
+ * SVGA v3 has mmio register access and lacks fifo cmds
+ */
+static inline bool vmw_is_svga_v3(const struct vmw_private *dev)
+{
+	return dev->pci_id == VMWGFX_PCI_ID_SVGA3;
+}
+
+/*
  * The locking here is fine-grained, so that it is performed once
  * for every read- and write operation. This is of course costly, but we
  * don't perform much register access in the timing critical paths anyway.
@@ -655,10 +664,14 @@ static inline struct vmw_fpriv *vmw_fpriv(struct drm_file *file_priv)
 static inline void vmw_write(struct vmw_private *dev_priv,
 			     unsigned int offset, uint32_t value)
 {
-	spin_lock(&dev_priv->hw_lock);
-	outl(offset, dev_priv->io_start + VMWGFX_INDEX_PORT);
-	outl(value, dev_priv->io_start + VMWGFX_VALUE_PORT);
-	spin_unlock(&dev_priv->hw_lock);
+	if (vmw_is_svga_v3(dev_priv)) {
+		iowrite32(value, dev_priv->rmmio + offset);
+	} else {
+		spin_lock(&dev_priv->hw_lock);
+		outl(offset, dev_priv->io_start + SVGA_INDEX_PORT);
+		outl(value, dev_priv->io_start + SVGA_VALUE_PORT);
+		spin_unlock(&dev_priv->hw_lock);
+	}
 }
 
 static inline uint32_t vmw_read(struct vmw_private *dev_priv,
@@ -666,10 +679,14 @@ static inline uint32_t vmw_read(struct vmw_private *dev_priv,
 {
 	u32 val;
 
-	spin_lock(&dev_priv->hw_lock);
-	outl(offset, dev_priv->io_start + VMWGFX_INDEX_PORT);
-	val = inl(dev_priv->io_start + VMWGFX_VALUE_PORT);
-	spin_unlock(&dev_priv->hw_lock);
+	if (vmw_is_svga_v3(dev_priv)) {
+		val = ioread32(dev_priv->rmmio + offset);
+	} else {
+		spin_lock(&dev_priv->hw_lock);
+		outl(offset, dev_priv->io_start + SVGA_INDEX_PORT);
+		val = inl(dev_priv->io_start + SVGA_VALUE_PORT);
+		spin_unlock(&dev_priv->hw_lock);
+	}
 
 	return val;
 }
@@ -932,19 +949,14 @@ extern int vmw_present_ioctl(struct drm_device *dev, void *data,
 			     struct drm_file *file_priv);
 extern int vmw_present_readback_ioctl(struct drm_device *dev, void *data,
 				      struct drm_file *file_priv);
-extern __poll_t vmw_fops_poll(struct file *filp,
-				  struct poll_table_struct *wait);
-extern ssize_t vmw_fops_read(struct file *filp, char __user *buffer,
-			     size_t count, loff_t *offset);
 
 /**
  * Fifo utilities - vmwgfx_fifo.c
  */
 
-extern int vmw_fifo_init(struct vmw_private *dev_priv,
-			 struct vmw_fifo_state *fifo);
-extern void vmw_fifo_release(struct vmw_private *dev_priv,
-			     struct vmw_fifo_state *fifo);
+extern struct vmw_fifo_state *vmw_fifo_create(struct vmw_private *dev_priv);
+extern void vmw_fifo_destroy(struct vmw_private *dev_priv);
+extern bool vmw_cmd_supported(struct vmw_private *vmw);
 extern void *
 vmw_cmd_ctx_reserve(struct vmw_private *dev_priv, uint32_t bytes, int ctx_id);
 extern void vmw_cmd_commit(struct vmw_private *dev_priv, uint32_t bytes);
@@ -970,6 +982,31 @@ extern int vmw_cmd_flush(struct vmw_private *dev_priv,
 #define VMW_CMD_RESERVE(__priv, __bytes)                                     \
 	VMW_CMD_CTX_RESERVE(__priv, __bytes, SVGA3D_INVALID_ID)
 
+
+/**
+ * vmw_fifo_caps - Returns the capabilities of the FIFO command
+ * queue or 0 if fifo memory isn't present.
+ * @dev_priv: The device private context
+ */
+static inline uint32_t vmw_fifo_caps(const struct vmw_private *dev_priv)
+{
+	if (!dev_priv->fifo_mem || !dev_priv->fifo)
+		return 0;
+	return dev_priv->fifo->capabilities;
+}
+
+
+/**
+ * vmw_is_cursor_bypass3_enabled - Returns TRUE iff Cursor Bypass 3
+ * is enabled in the FIFO.
+ * @dev_priv: The device private context
+ */
+static inline bool
+vmw_is_cursor_bypass3_enabled(const struct vmw_private *dev_priv)
+{
+	return (vmw_fifo_caps(dev_priv) & SVGA_FIFO_CAP_CURSOR_BYPASS_3) != 0;
+}
+
 /**
  * TTM glue - vmwgfx_ttm_glue.c
  */
@@ -1079,9 +1116,6 @@ bool vmw_cmd_describe(const void *buf, u32 *size, char const **cmd);
  * IRQs and wating - vmwgfx_irq.c
  */
 
-extern int vmw_wait_seqno(struct vmw_private *dev_priv, bool lazy,
-			  uint32_t seqno, bool interruptible,
-			  unsigned long timeout);
 extern int vmw_irq_install(struct drm_device *dev, int irq);
 extern void vmw_irq_uninstall(struct drm_device *dev);
 extern bool vmw_seqno_passed(struct vmw_private *dev_priv,
@@ -1092,8 +1126,7 @@ extern int vmw_fallback_wait(struct vmw_private *dev_priv,
 			     uint32_t seqno,
 			     bool interruptible,
 			     unsigned long timeout);
-extern void vmw_update_seqno(struct vmw_private *dev_priv,
-				struct vmw_fifo_state *fifo_state);
+extern void vmw_update_seqno(struct vmw_private *dev_priv);
 extern void vmw_seqno_waiter_add(struct vmw_private *dev_priv);
 extern void vmw_seqno_waiter_remove(struct vmw_private *dev_priv);
 extern void vmw_goal_waiter_add(struct vmw_private *dev_priv);
@@ -1572,6 +1605,7 @@ static inline void vmw_fifo_resource_dec(struct vmw_private *dev_priv)
  */
 static inline u32 vmw_fifo_mem_read(struct vmw_private *vmw, uint32 fifo_reg)
 {
+	BUG_ON(vmw_is_svga_v3(vmw));
 	return READ_ONCE(*(vmw->fifo_mem + fifo_reg));
 }
 
@@ -1586,6 +1620,44 @@ static inline u32 vmw_fifo_mem_read(struct vmw_private *vmw, uint32 fifo_reg)
 static inline void vmw_fifo_mem_write(struct vmw_private *vmw, u32 fifo_reg,
 				      u32 value)
 {
+	BUG_ON(vmw_is_svga_v3(vmw));
 	WRITE_ONCE(*(vmw->fifo_mem + fifo_reg), value);
 }
+
+static inline u32 vmw_fence_read(struct vmw_private *dev_priv)
+{
+	u32 fence;
+	if (vmw_is_svga_v3(dev_priv))
+		fence = vmw_read(dev_priv, SVGA_REG_FENCE);
+	else
+		fence = vmw_fifo_mem_read(dev_priv, SVGA_FIFO_FENCE);
+	return fence;
+}
+
+static inline void vmw_fence_write(struct vmw_private *dev_priv,
+				  u32 fence)
+{
+	BUG_ON(vmw_is_svga_v3(dev_priv));
+	vmw_fifo_mem_write(dev_priv, SVGA_FIFO_FENCE, fence);
+}
+
+static inline u32 vmw_irq_status_read(struct vmw_private *vmw)
+{
+	u32 status;
+	if (vmw_is_svga_v3(vmw))
+		status = vmw_read(vmw, SVGA_REG_IRQ_STATUS);
+	else
+		status = inl(vmw->io_start + SVGA_IRQSTATUS_PORT);
+	return status;
+}
+
+static inline void vmw_irq_status_write(struct vmw_private *vmw,
+					uint32 status)
+{
+	if (vmw_is_svga_v3(vmw))
+		vmw_write(vmw, SVGA_REG_IRQ_STATUS, status);
+	else
+		outl(status, vmw->io_start + SVGA_IRQSTATUS_PORT);
+}
+
 #endif