Merge tag 'drm-misc-fixes-2024-08-29' of https://gitlab.freedesktop.org/drm/misc/kernel into drm-fixes

A revert for a previous TTM commit causing stuttering, 3 fixes for
vmwgfx related to buffer operations, a fix for video/aperture with
non-VGA primary devices, and a preemption status fix for v3d

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

From: Maxime Ripard <mripard@redhat.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240829-efficient-swift-from-lemuria-f60c05@houat

12 files changed, 161 insertions, 37 deletions

diff --git a/drivers/firmware/sysfb.c b/drivers/firmware/sysfb.c
index 921f61507ae8..02a07d3d0d40 100644
--- a/drivers/firmware/sysfb.c
+++ b/drivers/firmware/sysfb.c
@@ -39,6 +39,8 @@ static struct platform_device *pd;
 static DEFINE_MUTEX(disable_lock);
 static bool disabled;
 
+static struct device *sysfb_parent_dev(const struct screen_info *si);
+
 static bool sysfb_unregister(void)
 {
 	if (IS_ERR_OR_NULL(pd))
@@ -52,6 +54,7 @@ static bool sysfb_unregister(void)
 
 /**
  * sysfb_disable() - disable the Generic System Framebuffers support
+ * @dev:	the device to check if non-NULL
  *
  * This disables the registration of system framebuffer devices that match the
  * generic drivers that make use of the system framebuffer set up by firmware.
@@ -61,17 +64,21 @@ static bool sysfb_unregister(void)
  * Context: The function can sleep. A @disable_lock mutex is acquired to serialize
  *          against sysfb_init(), that registers a system framebuffer device.
  */
-void sysfb_disable(void)
+void sysfb_disable(struct device *dev)
 {
+	struct screen_info *si = &screen_info;
+
 	mutex_lock(&disable_lock);
-	sysfb_unregister();
-	disabled = true;
+	if (!dev || dev == sysfb_parent_dev(si)) {
+		sysfb_unregister();
+		disabled = true;
+	}
 	mutex_unlock(&disable_lock);
 }
 EXPORT_SYMBOL_GPL(sysfb_disable);
 
 #if defined(CONFIG_PCI)
-static __init bool sysfb_pci_dev_is_enabled(struct pci_dev *pdev)
+static bool sysfb_pci_dev_is_enabled(struct pci_dev *pdev)
 {
 	/*
 	 * TODO: Try to integrate this code into the PCI subsystem
@@ -87,13 +94,13 @@ static __init bool sysfb_pci_dev_is_enabled(struct pci_dev *pdev)
 	return true;
 }
 #else
-static __init bool sysfb_pci_dev_is_enabled(struct pci_dev *pdev)
+static bool sysfb_pci_dev_is_enabled(struct pci_dev *pdev)
 {
 	return false;
 }
 #endif
 
-static __init struct device *sysfb_parent_dev(const struct screen_info *si)
+static struct device *sysfb_parent_dev(const struct screen_info *si)
 {
 	struct pci_dev *pdev;
 
diff --git a/drivers/gpu/drm/v3d/v3d_sched.c b/drivers/gpu/drm/v3d/v3d_sched.c
index b8682818bafa..ad1e6236ff6f 100644
--- a/drivers/gpu/drm/v3d/v3d_sched.c
+++ b/drivers/gpu/drm/v3d/v3d_sched.c
@@ -134,6 +134,8 @@ v3d_job_start_stats(struct v3d_job *job, enum v3d_queue queue)
 	struct v3d_stats *local_stats = &file->stats[queue];
 	u64 now = local_clock();
 
+	preempt_disable();
+
 	write_seqcount_begin(&local_stats->lock);
 	local_stats->start_ns = now;
 	write_seqcount_end(&local_stats->lock);
@@ -141,6 +143,8 @@ v3d_job_start_stats(struct v3d_job *job, enum v3d_queue queue)
 	write_seqcount_begin(&global_stats->lock);
 	global_stats->start_ns = now;
 	write_seqcount_end(&global_stats->lock);
+
+	preempt_enable();
 }
 
 static void
@@ -162,8 +166,10 @@ v3d_job_update_stats(struct v3d_job *job, enum v3d_queue queue)
 	struct v3d_stats *local_stats = &file->stats[queue];
 	u64 now = local_clock();
 
+	preempt_disable();
 	v3d_stats_update(local_stats, now);
 	v3d_stats_update(global_stats, now);
+	preempt_enable();
 }
 
 static struct dma_fence *v3d_bin_job_run(struct drm_sched_job *sched_job)
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_blit.c b/drivers/gpu/drm/vmwgfx/vmwgfx_blit.c
index 717d624e9a05..890a66a2361f 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_blit.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_blit.c
@@ -27,6 +27,8 @@
  **************************************************************************/
 
 #include "vmwgfx_drv.h"
+
+#include "vmwgfx_bo.h"
 #include <linux/highmem.h>
 
 /*
@@ -420,13 +422,105 @@ static int vmw_bo_cpu_blit_line(struct vmw_bo_blit_line_data *d,
 	return 0;
 }
 
+static void *map_external(struct vmw_bo *bo, struct iosys_map *map)
+{
+	struct vmw_private *vmw =
+		container_of(bo->tbo.bdev, struct vmw_private, bdev);
+	void *ptr = NULL;
+	int ret;
+
+	if (bo->tbo.base.import_attach) {
+		ret = dma_buf_vmap(bo->tbo.base.dma_buf, map);
+		if (ret) {
+			drm_dbg_driver(&vmw->drm,
+				       "Wasn't able to map external bo!\n");
+			goto out;
+		}
+		ptr = map->vaddr;
+	} else {
+		ptr = vmw_bo_map_and_cache(bo);
+	}
+
+out:
+	return ptr;
+}
+
+static void unmap_external(struct vmw_bo *bo, struct iosys_map *map)
+{
+	if (bo->tbo.base.import_attach)
+		dma_buf_vunmap(bo->tbo.base.dma_buf, map);
+	else
+		vmw_bo_unmap(bo);
+}
+
+static int vmw_external_bo_copy(struct vmw_bo *dst, u32 dst_offset,
+				u32 dst_stride, struct vmw_bo *src,
+				u32 src_offset, u32 src_stride,
+				u32 width_in_bytes, u32 height,
+				struct vmw_diff_cpy *diff)
+{
+	struct vmw_private *vmw =
+		container_of(dst->tbo.bdev, struct vmw_private, bdev);
+	size_t dst_size = dst->tbo.resource->size;
+	size_t src_size = src->tbo.resource->size;
+	struct iosys_map dst_map = {0};
+	struct iosys_map src_map = {0};
+	int ret, i;
+	int x_in_bytes;
+	u8 *vsrc;
+	u8 *vdst;
+
+	vsrc = map_external(src, &src_map);
+	if (!vsrc) {
+		drm_dbg_driver(&vmw->drm, "Wasn't able to map src\n");
+		ret = -ENOMEM;
+		goto out;
+	}
+
+	vdst = map_external(dst, &dst_map);
+	if (!vdst) {
+		drm_dbg_driver(&vmw->drm, "Wasn't able to map dst\n");
+		ret = -ENOMEM;
+		goto out;
+	}
+
+	vsrc += src_offset;
+	vdst += dst_offset;
+	if (src_stride == dst_stride) {
+		dst_size -= dst_offset;
+		src_size -= src_offset;
+		memcpy(vdst, vsrc,
+		       min(dst_stride * height, min(dst_size, src_size)));
+	} else {
+		WARN_ON(dst_stride < width_in_bytes);
+		for (i = 0; i < height; ++i) {
+			memcpy(vdst, vsrc, width_in_bytes);
+			vsrc += src_stride;
+			vdst += dst_stride;
+		}
+	}
+
+	x_in_bytes = (dst_offset % dst_stride);
+	diff->rect.x1 =  x_in_bytes / diff->cpp;
+	diff->rect.y1 = ((dst_offset - x_in_bytes) / dst_stride);
+	diff->rect.x2 = diff->rect.x1 + width_in_bytes / diff->cpp;
+	diff->rect.y2 = diff->rect.y1 + height;
+
+	ret = 0;
+out:
+	unmap_external(src, &src_map);
+	unmap_external(dst, &dst_map);
+
+	return ret;
+}
+
 /**
  * vmw_bo_cpu_blit - in-kernel cpu blit.
  *
- * @dst: Destination buffer object.
+ * @vmw_dst: Destination buffer object.
  * @dst_offset: Destination offset of blit start in bytes.
  * @dst_stride: Destination stride in bytes.
- * @src: Source buffer object.
+ * @vmw_src: Source buffer object.
  * @src_offset: Source offset of blit start in bytes.
  * @src_stride: Source stride in bytes.
  * @w: Width of blit.
@@ -444,13 +538,15 @@ static int vmw_bo_cpu_blit_line(struct vmw_bo_blit_line_data *d,
  * Neither of the buffer objects may be placed in PCI memory
  * (Fixed memory in TTM terminology) when using this function.
  */
-int vmw_bo_cpu_blit(struct ttm_buffer_object *dst,
+int vmw_bo_cpu_blit(struct vmw_bo *vmw_dst,
 		    u32 dst_offset, u32 dst_stride,
-		    struct ttm_buffer_object *src,
+		    struct vmw_bo *vmw_src,
 		    u32 src_offset, u32 src_stride,
 		    u32 w, u32 h,
 		    struct vmw_diff_cpy *diff)
 {
+	struct ttm_buffer_object *src = &vmw_src->tbo;
+	struct ttm_buffer_object *dst = &vmw_dst->tbo;
 	struct ttm_operation_ctx ctx = {
 		.interruptible = false,
 		.no_wait_gpu = false
@@ -460,6 +556,11 @@ int vmw_bo_cpu_blit(struct ttm_buffer_object *dst,
 	int ret = 0;
 	struct page **dst_pages = NULL;
 	struct page **src_pages = NULL;
+	bool src_external = (src->ttm->page_flags & TTM_TT_FLAG_EXTERNAL) != 0;
+	bool dst_external = (dst->ttm->page_flags & TTM_TT_FLAG_EXTERNAL) != 0;
+
+	if (WARN_ON(dst == src))
+		return -EINVAL;
 
 	/* Buffer objects need to be either pinned or reserved: */
 	if (!(dst->pin_count))
@@ -479,6 +580,11 @@ int vmw_bo_cpu_blit(struct ttm_buffer_object *dst,
 			return ret;
 	}
 
+	if (src_external || dst_external)
+		return vmw_external_bo_copy(vmw_dst, dst_offset, dst_stride,
+					    vmw_src, src_offset, src_stride,
+					    w, h, diff);
+
 	if (!src->ttm->pages && src->ttm->sg) {
 		src_pages = kvmalloc_array(src->ttm->num_pages,
 					   sizeof(struct page *), GFP_KERNEL);
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_bo.c b/drivers/gpu/drm/vmwgfx/vmwgfx_bo.c
index f42ebc4a7c22..a0e433fbcba6 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_bo.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_bo.c
@@ -360,6 +360,8 @@ void *vmw_bo_map_and_cache_size(struct vmw_bo *vbo, size_t size)
 	void *virtual;
 	int ret;
 
+	atomic_inc(&vbo->map_count);
+
 	virtual = ttm_kmap_obj_virtual(&vbo->map, &not_used);
 	if (virtual)
 		return virtual;
@@ -383,11 +385,17 @@ void *vmw_bo_map_and_cache_size(struct vmw_bo *vbo, size_t size)
  */
 void vmw_bo_unmap(struct vmw_bo *vbo)
 {
+	int map_count;
+
 	if (vbo->map.bo == NULL)
 		return;
 
-	ttm_bo_kunmap(&vbo->map);
-	vbo->map.bo = NULL;
+	map_count = atomic_dec_return(&vbo->map_count);
+
+	if (!map_count) {
+		ttm_bo_kunmap(&vbo->map);
+		vbo->map.bo = NULL;
+	}
 }
 
 
@@ -421,6 +429,7 @@ static int vmw_bo_init(struct vmw_private *dev_priv,
 	vmw_bo->tbo.priority = 3;
 	vmw_bo->res_tree = RB_ROOT;
 	xa_init(&vmw_bo->detached_resources);
+	atomic_set(&vmw_bo->map_count, 0);
 
 	params->size = ALIGN(params->size, PAGE_SIZE);
 	drm_gem_private_object_init(vdev, &vmw_bo->tbo.base, params->size);
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_bo.h b/drivers/gpu/drm/vmwgfx/vmwgfx_bo.h
index 62b4342d5f7c..43b5439ec9f7 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_bo.h
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_bo.h
@@ -71,6 +71,8 @@ struct vmw_bo_params {
  * @map: Kmap object for semi-persistent mappings
  * @res_tree: RB tree of resources using this buffer object as a backing MOB
  * @res_prios: Eviction priority counts for attached resources
+ * @map_count: The number of currently active maps. Will differ from the
+ * cpu_writers because it includes kernel maps.
  * @cpu_writers: Number of synccpu write grabs. Protected by reservation when
  * increased. May be decreased without reservation.
  * @dx_query_ctx: DX context if this buffer object is used as a DX query MOB
@@ -90,6 +92,7 @@ struct vmw_bo {
 	u32 res_prios[TTM_MAX_BO_PRIORITY];
 	struct xarray detached_resources;
 
+	atomic_t map_count;
 	atomic_t cpu_writers;
 	/* Not ref-counted.  Protected by binding_mutex */
 	struct vmw_resource *dx_query_ctx;
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h
index 32f50e595809..3f4719b3c268 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h
@@ -1353,9 +1353,9 @@ void vmw_diff_memcpy(struct vmw_diff_cpy *diff, u8 *dest, const u8 *src,
 
 void vmw_memcpy(struct vmw_diff_cpy *diff, u8 *dest, const u8 *src, size_t n);
 
-int vmw_bo_cpu_blit(struct ttm_buffer_object *dst,
+int vmw_bo_cpu_blit(struct vmw_bo *dst,
 		    u32 dst_offset, u32 dst_stride,
-		    struct ttm_buffer_object *src,
+		    struct vmw_bo *src,
 		    u32 src_offset, u32 src_stride,
 		    u32 w, u32 h,
 		    struct vmw_diff_cpy *diff);
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_stdu.c b/drivers/gpu/drm/vmwgfx/vmwgfx_stdu.c
index 5453f7cf0e2d..fab155a68054 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_stdu.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_stdu.c
@@ -502,7 +502,7 @@ static void vmw_stdu_bo_cpu_commit(struct vmw_kms_dirty *dirty)
 		container_of(dirty->unit, typeof(*stdu), base);
 	s32 width, height;
 	s32 src_pitch, dst_pitch;
-	struct ttm_buffer_object *src_bo, *dst_bo;
+	struct vmw_bo *src_bo, *dst_bo;
 	u32 src_offset, dst_offset;
 	struct vmw_diff_cpy diff = VMW_CPU_BLIT_DIFF_INITIALIZER(stdu->cpp);
 
@@ -517,11 +517,11 @@ static void vmw_stdu_bo_cpu_commit(struct vmw_kms_dirty *dirty)
 
 	/* Assume we are blitting from Guest (bo) to Host (display_srf) */
 	src_pitch = stdu->display_srf->metadata.base_size.width * stdu->cpp;
-	src_bo = &stdu->display_srf->res.guest_memory_bo->tbo;
+	src_bo = stdu->display_srf->res.guest_memory_bo;
 	src_offset = ddirty->top * src_pitch + ddirty->left * stdu->cpp;
 
 	dst_pitch = ddirty->pitch;
-	dst_bo = &ddirty->buf->tbo;
+	dst_bo = ddirty->buf;
 	dst_offset = ddirty->fb_top * dst_pitch + ddirty->fb_left * stdu->cpp;
 
 	(void) vmw_bo_cpu_blit(dst_bo, dst_offset, dst_pitch,
@@ -1170,7 +1170,7 @@ vmw_stdu_bo_populate_update_cpu(struct vmw_du_update_plane  *update, void *cmd,
 	struct vmw_diff_cpy diff = VMW_CPU_BLIT_DIFF_INITIALIZER(0);
 	struct vmw_stdu_update_gb_image *cmd_img = cmd;
 	struct vmw_stdu_update *cmd_update;
-	struct ttm_buffer_object *src_bo, *dst_bo;
+	struct vmw_bo *src_bo, *dst_bo;
 	u32 src_offset, dst_offset;
 	s32 src_pitch, dst_pitch;
 	s32 width, height;
@@ -1184,11 +1184,11 @@ vmw_stdu_bo_populate_update_cpu(struct vmw_du_update_plane  *update, void *cmd,
 
 	diff.cpp = stdu->cpp;
 
-	dst_bo = &stdu->display_srf->res.guest_memory_bo->tbo;
+	dst_bo = stdu->display_srf->res.guest_memory_bo;
 	dst_pitch = stdu->display_srf->metadata.base_size.width * stdu->cpp;
 	dst_offset = bb->y1 * dst_pitch + bb->x1 * stdu->cpp;
 
-	src_bo = &vfbbo->buffer->tbo;
+	src_bo = vfbbo->buffer;
 	src_pitch = update->vfb->base.pitches[0];
 	src_offset = bo_update->fb_top * src_pitch + bo_update->fb_left *
 		stdu->cpp;
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_surface.c b/drivers/gpu/drm/vmwgfx/vmwgfx_surface.c
index 8ae6a761c900..1625b30d9970 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_surface.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_surface.c
@@ -2283,9 +2283,11 @@ int vmw_dumb_create(struct drm_file *file_priv,
 	/*
 	 * Without mob support we're just going to use raw memory buffer
 	 * because we wouldn't be able to support full surface coherency
-	 * without mobs
+	 * without mobs. There also no reason to support surface coherency
+	 * without 3d (i.e. gpu usage on the host) because then all the
+	 * contents is going to be rendered guest side.
 	 */
-	if (!dev_priv->has_mob) {
+	if (!dev_priv->has_mob || !vmw_supports_3d(dev_priv)) {
 		int cpp = DIV_ROUND_UP(args->bpp, 8);
 
 		switch (cpp) {
diff --git a/drivers/of/platform.c b/drivers/of/platform.c
index 389d4ea6bfc1..ef622d41eb5b 100644
--- a/drivers/of/platform.c
+++ b/drivers/of/platform.c
@@ -592,7 +592,7 @@ static int __init of_platform_default_populate_init(void)
 			 * This can happen for example on DT systems that do EFI
 			 * booting and may provide a GOP handle to the EFI stub.
 			 */
-			sysfb_disable();
+			sysfb_disable(NULL);
 			of_platform_device_create(node, NULL, NULL);
 			of_node_put(node);
 		}
diff --git a/drivers/video/aperture.c b/drivers/video/aperture.c
index 561be8feca96..2b5a1e666e9b 100644
--- a/drivers/video/aperture.c
+++ b/drivers/video/aperture.c
@@ -293,7 +293,7 @@ int aperture_remove_conflicting_devices(resource_size_t base, resource_size_t si
 	 * ask for this, so let's assume that a real driver for the display
 	 * was already probed and prevent sysfb to register devices later.
 	 */
-	sysfb_disable();
+	sysfb_disable(NULL);
 
 	aperture_detach_devices(base, size);
 
@@ -346,15 +346,10 @@ EXPORT_SYMBOL(__aperture_remove_legacy_vga_devices);
  */
 int aperture_remove_conflicting_pci_devices(struct pci_dev *pdev, const char *name)
 {
-	bool primary = false;
 	resource_size_t base, size;
 	int bar, ret = 0;
 
-	if (pdev == vga_default_device())
-		primary = true;
-
-	if (primary)
-		sysfb_disable();
+	sysfb_disable(&pdev->dev);
 
 	for (bar = 0; bar < PCI_STD_NUM_BARS; ++bar) {
 		if (!(pci_resource_flags(pdev, bar) & IORESOURCE_MEM))
@@ -370,7 +365,7 @@ int aperture_remove_conflicting_pci_devices(struct pci_dev *pdev, const char *na
 	 * that consumes the VGA framebuffer I/O range. Remove this
 	 * device as well.
 	 */
-	if (primary)
+	if (pdev == vga_default_device())
 		ret = __aperture_remove_legacy_vga_devices(pdev);
 
 	return ret;
diff --git a/include/drm/ttm/ttm_bo.h b/include/drm/ttm/ttm_bo.h
index ef0f52f56ebc..6ccf96c91f3a 100644
--- a/include/drm/ttm/ttm_bo.h
+++ b/include/drm/ttm/ttm_bo.h
@@ -39,11 +39,7 @@
 #include "ttm_device.h"
 
 /* Default number of pre-faulted pages in the TTM fault handler */
-#if CONFIG_PGTABLE_LEVELS > 2
-#define TTM_BO_VM_NUM_PREFAULT (1 << (PMD_SHIFT - PAGE_SHIFT))
-#else
 #define TTM_BO_VM_NUM_PREFAULT 16
-#endif
 
 struct iosys_map;
 
diff --git a/include/linux/sysfb.h b/include/linux/sysfb.h
index c9cb657dad08..bef5f06a91de 100644
--- a/include/linux/sysfb.h
+++ b/include/linux/sysfb.h
@@ -58,11 +58,11 @@ struct efifb_dmi_info {
 
 #ifdef CONFIG_SYSFB
 
-void sysfb_disable(void);
+void sysfb_disable(struct device *dev);
 
 #else /* CONFIG_SYSFB */
 
-static inline void sysfb_disable(void)
+static inline void sysfb_disable(struct device *dev)
 {
 }