Re: [PATCH] drm/virtio: Add drm_panic support

From: Jocelyn Falempe
Date: Thu Oct 31 2024 - 06:35:06 EST

On 31/10/2024 04:20, Ryosuke Yasuoka wrote:
From: Jocelyn Falempe <jfalempe@xxxxxxxxxx>

Virtio gpu supports the drm_panic module, which displays a message to
the screen when a kernel panic occurs.

Thanks for this patch.
I think a few things can be simplified, see below.
Other than that, it looks good to me.


Signed-off-by: Ryosuke Yasuoka <ryasuoka@xxxxxxxxxx>
Signed-off-by: Jocelyn Falempe <jfalempe@xxxxxxxxxx>
---
drivers/gpu/drm/virtio/virtgpu_drv.h | 28 +++++
drivers/gpu/drm/virtio/virtgpu_plane.c | 161 +++++++++++++++++++++++++
drivers/gpu/drm/virtio/virtgpu_vq.c | 156 +++++++++++++++++++++++-
3 files changed, 339 insertions(+), 6 deletions(-)

diff --git a/drivers/gpu/drm/virtio/virtgpu_drv.h b/drivers/gpu/drm/virtio/virtgpu_drv.h
index 64c236169db8..765698a6dfa3 100644
--- a/drivers/gpu/drm/virtio/virtgpu_drv.h
+++ b/drivers/gpu/drm/virtio/virtgpu_drv.h
@@ -125,7 +125,14 @@ struct virtio_gpu_object_array {
struct drm_gem_object *objs[] __counted_by(total);
};
+#define MAX_INLINE_CMD_SIZE 96
+#define MAX_INLINE_RESP_SIZE 24
+#define VBUFFER_SIZE (sizeof(struct virtio_gpu_vbuffer) \
+ + MAX_INLINE_CMD_SIZE \
+ + MAX_INLINE_RESP_SIZE)
+
struct virtio_gpu_vbuffer;
+struct virtio_gpu_vbuffer_inline;

I think virtio_gpu_vbuffer_inline is not used anymore, so you can drop it.

struct virtio_gpu_device;
typedef void (*virtio_gpu_resp_cb)(struct virtio_gpu_device *vgdev,
@@ -329,12 +336,26 @@ void virtio_gpu_cmd_create_resource(struct virtio_gpu_device *vgdev,
struct virtio_gpu_fence *fence);
void virtio_gpu_cmd_unref_resource(struct virtio_gpu_device *vgdev,
struct virtio_gpu_object *bo);
+void virtio_gpu_panic_cmd_transfer_to_host_2d(struct virtio_gpu_device *vgdev,
+ uint64_t offset,
+ uint32_t width, uint32_t height,
+ uint32_t x, uint32_t y,
+ struct virtio_gpu_object_array *objs,
+ struct virtio_gpu_fence *fence,
+ struct virtio_gpu_vbuffer *vbuf);
void virtio_gpu_cmd_transfer_to_host_2d(struct virtio_gpu_device *vgdev,
uint64_t offset,
uint32_t width, uint32_t height,
uint32_t x, uint32_t y,
struct virtio_gpu_object_array *objs,
struct virtio_gpu_fence *fence);
+void virtio_gpu_panic_cmd_resource_flush(struct virtio_gpu_device *vgdev,
+ struct virtio_gpu_vbuffer *vbuf,
+ uint32_t resource_id,
+ uint32_t x, uint32_t y,
+ uint32_t width, uint32_t height,
+ struct virtio_gpu_object_array *objs,
+ struct virtio_gpu_fence *fence);
void virtio_gpu_cmd_resource_flush(struct virtio_gpu_device *vgdev,
uint32_t resource_id,
uint32_t x, uint32_t y,
@@ -399,6 +420,7 @@ void virtio_gpu_ctrl_ack(struct virtqueue *vq);
void virtio_gpu_cursor_ack(struct virtqueue *vq);
void virtio_gpu_dequeue_ctrl_func(struct work_struct *work);
void virtio_gpu_dequeue_cursor_func(struct work_struct *work);
+void virtio_gpu_panic_notify(struct virtio_gpu_device *vgdev);
void virtio_gpu_notify(struct virtio_gpu_device *vgdev);
int
@@ -435,6 +457,12 @@ struct drm_plane *virtio_gpu_plane_init(struct virtio_gpu_device *vgdev,
enum drm_plane_type type,
int index);
+struct virtio_gpu_vbuffer_inline {
+ struct virtio_gpu_vbuffer base;
+ u8 cmd_buf[MAX_INLINE_CMD_SIZE];
+ u8 resp_buf[MAX_INLINE_RESP_SIZE];
+};
This struct is no more needed, as the panic vbuffers are pre-allocated using VBUFFER_SIZE.

+
/* virtgpu_fence.c */
struct virtio_gpu_fence *virtio_gpu_fence_alloc(struct virtio_gpu_device *vgdev,
uint64_t base_fence_ctx,
diff --git a/drivers/gpu/drm/virtio/virtgpu_plane.c b/drivers/gpu/drm/virtio/virtgpu_plane.c
index a72a2dbda031..84b65f6b0906 100644
--- a/drivers/gpu/drm/virtio/virtgpu_plane.c
+++ b/drivers/gpu/drm/virtio/virtgpu_plane.c
@@ -26,6 +26,8 @@
#include <drm/drm_atomic_helper.h>
#include <drm/drm_damage_helper.h>
#include <drm/drm_fourcc.h>
+#include <drm/drm_managed.h>
+#include <drm/drm_panic.h>
#include "virtgpu_drv.h"
@@ -108,6 +110,28 @@ static int virtio_gpu_plane_atomic_check(struct drm_plane *plane,
return ret;
}
+/* For drm panic */
+static void virtio_gpu_panic_update_dumb_bo(struct virtio_gpu_device *vgdev,
+ struct drm_plane_state *state,
+ struct drm_rect *rect,
+ struct virtio_gpu_object_array *objs,
+ struct virtio_gpu_vbuffer *vbuf)
+{
+ struct virtio_gpu_object *bo =
+ gem_to_virtio_gpu_obj(state->fb->obj[0]);
+ uint32_t w = rect->x2 - rect->x1;
+ uint32_t h = rect->y2 - rect->y1;
+ uint32_t x = rect->x1;
+ uint32_t y = rect->y1;
+ uint32_t off = x * state->fb->format->cpp[0] +
+ y * state->fb->pitches[0];
+
+ virtio_gpu_array_add_obj(objs, &bo->base.base);
+
+ virtio_gpu_panic_cmd_transfer_to_host_2d(vgdev, off, w, h, x, y,
+ objs, NULL, vbuf);
+}
+
static void virtio_gpu_update_dumb_bo(struct virtio_gpu_device *vgdev,
struct drm_plane_state *state,
struct drm_rect *rect)
@@ -131,6 +155,24 @@ static void virtio_gpu_update_dumb_bo(struct virtio_gpu_device *vgdev,
objs, NULL);
}
+/* For drm_panic */
+static void virtio_gpu_panic_resource_flush(struct drm_plane *plane,
+ struct virtio_gpu_vbuffer *vbuf,
+ uint32_t x, uint32_t y,
+ uint32_t width, uint32_t height)
+{
+ struct drm_device *dev = plane->dev;
+ struct virtio_gpu_device *vgdev = dev->dev_private;
+ struct virtio_gpu_framebuffer *vgfb;
+ struct virtio_gpu_object *bo;
+
+ vgfb = to_virtio_gpu_framebuffer(plane->state->fb);
+ bo = gem_to_virtio_gpu_obj(vgfb->base.obj[0]);
+
+ virtio_gpu_panic_cmd_resource_flush(vgdev, vbuf, bo->hw_res_handle, x, y,
+ width, height, NULL, NULL);
+}
+
static void virtio_gpu_resource_flush(struct drm_plane *plane,
uint32_t x, uint32_t y,
uint32_t width, uint32_t height)
@@ -359,11 +401,123 @@ static void virtio_gpu_cursor_plane_update(struct drm_plane *plane,
virtio_gpu_cursor_ping(vgdev, output);
}
+static int virtio_drm_get_scanout_buffer(struct drm_plane *plane,
+ struct drm_scanout_buffer *sb)
+{
+ struct virtio_gpu_object *bo;
+
+ if (!plane->state || !plane->state->fb || !plane->state->visible)
+ return -ENODEV;
+
+ bo = gem_to_virtio_gpu_obj(plane->state->fb->obj[0]);
+
+ /* try to vmap it if possible */
+ if (!bo->base.vaddr) {
+ int ret;
+
+ ret = drm_gem_shmem_vmap(&bo->base, &sb->map[0]);
+ if (ret)
+ return ret;
+ } else {
+ iosys_map_set_vaddr(&sb->map[0], bo->base.vaddr);
+ }
+
+ sb->format = plane->state->fb->format;
+ sb->height = plane->state->fb->height;
+ sb->width = plane->state->fb->width;
+ sb->pitch[0] = plane->state->fb->pitches[0];
+ return 0;
+}
+
+struct virtio_gpu_panic_object_array {
+ struct ww_acquire_ctx ticket;
+ struct list_head next;
+ u32 nents, total;
+ struct drm_gem_object *objs;
+};
+
+static void *virtio_panic_buffer;
+
+static void virtio_gpu_panic_get_vbuf(struct virtqueue *vq,
+ struct virtio_gpu_vbuffer *vbuf,
+ struct list_head *reclaim_list)
+{
+ unsigned int len;
+
+ /* waiting vbuf to be used */
+ while (1) {
+ if (vbuf == virtqueue_get_buf(vq, &len))
+ break;
+ }
+ list_add_tail(&vbuf->list, reclaim_list);
+}
Instead of using a reclaim list, I think you can just call drm_gem_object_put() if there is an obj here, and that's all.

+
+static void virtio_panic_flush(struct drm_plane *plane)
+{
+ struct virtio_gpu_object *bo;
+ struct drm_device *dev = plane->dev;
+ struct virtio_gpu_device *vgdev = dev->dev_private;
+ struct drm_rect rect;
+ struct virtio_gpu_vbuffer_inline *vbuf_inline_dumb_bo;
+ struct virtio_gpu_vbuffer_inline *vbuf_inline_resource_flush;
I think they should be regular struct virtio_gpu_vbuffer *

+
+ vbuf_inline_dumb_bo = virtio_panic_buffer;
+ vbuf_inline_resource_flush = virtio_panic_buffer + VBUFFER_SIZE;
+
+ rect.x1 = 0;
+ rect.y1 = 0;
+ rect.x2 = plane->state->fb->width;
+ rect.y2 = plane->state->fb->height;
+
+ bo = gem_to_virtio_gpu_obj(plane->state->fb->obj[0]);
+
+ struct drm_gem_object obj;
+ struct virtio_gpu_panic_object_array objs = {
+ .next = { NULL, NULL },
+ .nents = 0,
+ .total = 1,
+ .objs = &obj
+ };
+ struct list_head reclaim_list;
+ struct virtio_gpu_vbuffer *entry, *tmp;
+
+ INIT_LIST_HEAD(&reclaim_list);
+
+ if (bo->dumb)
+ virtio_gpu_panic_update_dumb_bo(vgdev,
+ plane->state,
+ &rect,
+ (struct virtio_gpu_object_array *)&objs,
+ &vbuf_inline_dumb_bo->base);
+
+ virtio_gpu_panic_resource_flush(plane, &vbuf_inline_resource_flush->base,
+ plane->state->src_x >> 16,
+ plane->state->src_y >> 16,
+ plane->state->src_w >> 16,
+ plane->state->src_h >> 16);
+ virtio_gpu_panic_notify(vgdev);
+
+ virtio_gpu_panic_get_vbuf(vgdev->ctrlq.vq,
+ &vbuf_inline_dumb_bo->base,
+ &reclaim_list);
+ virtio_gpu_panic_get_vbuf(vgdev->ctrlq.vq,
+ &vbuf_inline_resource_flush->base,
+ &reclaim_list);
+
+ list_for_each_entry_safe(entry, tmp, &reclaim_list, list) {
+ if (entry->objs)
+ drm_gem_object_put(&objs.objs[0]);
+ list_del(&entry->list);
+ }
+}
+
static const struct drm_plane_helper_funcs virtio_gpu_primary_helper_funcs = {
.prepare_fb = virtio_gpu_plane_prepare_fb,
.cleanup_fb = virtio_gpu_plane_cleanup_fb,
.atomic_check = virtio_gpu_plane_atomic_check,
.atomic_update = virtio_gpu_primary_plane_update,
+ .get_scanout_buffer = virtio_drm_get_scanout_buffer,
+ .panic_flush = virtio_panic_flush,
};
static const struct drm_plane_helper_funcs virtio_gpu_cursor_helper_funcs = {
@@ -383,6 +537,13 @@ struct drm_plane *virtio_gpu_plane_init(struct virtio_gpu_device *vgdev,
const uint32_t *formats;
int nformats;
+ /* allocate panic buffers */
+ if (index == 0 && type == DRM_PLANE_TYPE_PRIMARY) {
+ virtio_panic_buffer = drmm_kzalloc(dev, 2 * VBUFFER_SIZE, GFP_KERNEL);
+ if (!virtio_panic_buffer)
+ return ERR_PTR(-ENOMEM);
+ }
+
if (type == DRM_PLANE_TYPE_CURSOR) {
formats = virtio_gpu_cursor_formats;
nformats = ARRAY_SIZE(virtio_gpu_cursor_formats);
diff --git a/drivers/gpu/drm/virtio/virtgpu_vq.c b/drivers/gpu/drm/virtio/virtgpu_vq.c
index 0d3d0d09f39b..b9b8e6a5c314 100644
--- a/drivers/gpu/drm/virtio/virtgpu_vq.c
+++ b/drivers/gpu/drm/virtio/virtgpu_vq.c
@@ -36,12 +36,6 @@
#include "virtgpu_drv.h"
#include "virtgpu_trace.h"
-#define MAX_INLINE_CMD_SIZE 96
-#define MAX_INLINE_RESP_SIZE 24
-#define VBUFFER_SIZE (sizeof(struct virtio_gpu_vbuffer) \
- + MAX_INLINE_CMD_SIZE \
- + MAX_INLINE_RESP_SIZE)
-
static void convert_to_hw_box(struct virtio_gpu_box *dst,
const struct drm_virtgpu_3d_box *src)
{
@@ -311,6 +305,40 @@ static struct sg_table *vmalloc_to_sgt(char *data, uint32_t size, int *sg_ents)
return sgt;
}
+/* For drm_panic */
+static int virtio_gpu_panic_queue_ctrl_sgs(struct virtio_gpu_device *vgdev,
+ struct virtio_gpu_vbuffer *vbuf,
+ struct virtio_gpu_fence *fence,
fence is always NULL, so you can remove it from the arguments.

+ int elemcnt,
+ struct scatterlist **sgs,
+ int outcnt,
+ int incnt)
+{
+ struct virtqueue *vq = vgdev->ctrlq.vq;
+ int ret;
+
+ if (vgdev->has_indirect)
+ elemcnt = 1;
+
+again:
+
+ if (vq->num_free < elemcnt) {
+ virtio_gpu_panic_notify(vgdev);
+ wait_event(vgdev->ctrlq.ack_queue, vq->num_free >= elemcnt);
In a panic handler, it can't wait for an event. If the virtio queue is full, maybe just return an error, and no panic message would be seen.

+ goto again;
+ }
+
+ ret = virtqueue_add_sgs(vq, sgs, outcnt, incnt, vbuf, GFP_ATOMIC);
+ WARN_ON(ret);
+
+ vbuf->seqno = ++vgdev->ctrlq.seqno;
+ trace_virtio_gpu_cmd_queue(vq, virtio_gpu_vbuf_ctrl_hdr(vbuf), vbuf->seqno);
+
+ atomic_inc(&vgdev->pending_commands);
+
+ return 0;
+}
+
static int virtio_gpu_queue_ctrl_sgs(struct virtio_gpu_device *vgdev,
struct virtio_gpu_vbuffer *vbuf,
struct virtio_gpu_fence *fence,
@@ -368,6 +396,36 @@ static int virtio_gpu_queue_ctrl_sgs(struct virtio_gpu_device *vgdev,
return 0;
}
+/* For drm_panic */
+static int virtio_gpu_panic_queue_fenced_ctrl_buffer(struct virtio_gpu_device *vgdev,
+ struct virtio_gpu_vbuffer *vbuf,
+ struct virtio_gpu_fence *fence)
fence is always NULL when calling this, so you can remove it from the arguments, and remove the "fenced" from the function name.
+{
+ struct scatterlist *sgs[3], vcmd, vresp;
+ int elemcnt = 0, outcnt = 0, incnt = 0, ret;
+
+ /* set up vcmd */
+ sg_init_one(&vcmd, vbuf->buf, vbuf->size);
+ elemcnt++;
+ sgs[outcnt] = &vcmd;
+ outcnt++;
+
+ /* set up vresp */
+ if (vbuf->resp_size) {
+ sg_init_one(&vresp, vbuf->resp_buf, vbuf->resp_size);
+ elemcnt++;
+ sgs[outcnt + incnt] = &vresp;
+ incnt++;
+ }
+
+ ret = virtio_gpu_panic_queue_ctrl_sgs(vgdev, vbuf,
+ fence, elemcnt,
+ sgs, outcnt,
+ incnt);
+
+ return ret;
+}
+
static int virtio_gpu_queue_fenced_ctrl_buffer(struct virtio_gpu_device *vgdev,
struct virtio_gpu_vbuffer *vbuf,
struct virtio_gpu_fence *fence)
@@ -422,6 +480,21 @@ static int virtio_gpu_queue_fenced_ctrl_buffer(struct virtio_gpu_device *vgdev,
return ret;
}
+/* For drm_panic */
+void virtio_gpu_panic_notify(struct virtio_gpu_device *vgdev)
+{
+ bool notify;
+
+ if (!atomic_read(&vgdev->pending_commands))
+ return;
+
+ atomic_set(&vgdev->pending_commands, 0);
+ notify = virtqueue_kick_prepare(vgdev->ctrlq.vq);
+
+ if (notify)
+ virtqueue_notify(vgdev->ctrlq.vq);
+}
+
void virtio_gpu_notify(struct virtio_gpu_device *vgdev)
{
bool notify;
@@ -567,6 +640,44 @@ void virtio_gpu_cmd_set_scanout(struct virtio_gpu_device *vgdev,
virtio_gpu_queue_ctrl_buffer(vgdev, vbuf);
}
+/* For drm_panic */
+static void virtio_gpu_panic_alloc_cmd(struct virtio_gpu_device *vgdev,
+ struct virtio_gpu_vbuffer *vbuf,
+ int cmd_size)
+{
+ vbuf->buf = (void *)vbuf + sizeof(*vbuf);
+ vbuf->size = cmd_size;
+ vbuf->resp_cb = NULL;
+ vbuf->resp_size = sizeof(struct virtio_gpu_ctrl_hdr);
+ vbuf->resp_buf = (void *)vbuf->buf + cmd_size;
+}
I think you can rename this to virtio_gpu_panic_init_cmd, because we're not allocating a buffer, just initializing a pre-allocated buffer.

+
+/* For drm_panic */
+void virtio_gpu_panic_cmd_resource_flush(struct virtio_gpu_device *vgdev,
+ struct virtio_gpu_vbuffer *vbuf,
+ uint32_t resource_id,
+ uint32_t x, uint32_t y,
+ uint32_t width, uint32_t height,
+ struct virtio_gpu_object_array *objs,
+ struct virtio_gpu_fence *fence)
objs and fence are always NULL, so you can remove them from the arguments.

+{
+ struct virtio_gpu_resource_flush *cmd_p;
+
+ virtio_gpu_panic_alloc_cmd(vgdev, vbuf,
+ sizeof(struct virtio_gpu_resource_flush));
+ cmd_p = (void *)vbuf->buf;
+ vbuf->objs = objs;
+
+ cmd_p->hdr.type = cpu_to_le32(VIRTIO_GPU_CMD_RESOURCE_FLUSH);
+ cmd_p->resource_id = cpu_to_le32(resource_id);
+ cmd_p->r.width = cpu_to_le32(width);
+ cmd_p->r.height = cpu_to_le32(height);
+ cmd_p->r.x = cpu_to_le32(x);
+ cmd_p->r.y = cpu_to_le32(y);
+
+ virtio_gpu_queue_fenced_ctrl_buffer(vgdev, vbuf, fence);
Use virtio_gpu_panic_queue_fenced_ctrl_buffer() there ?
+}
+
void virtio_gpu_cmd_resource_flush(struct virtio_gpu_device *vgdev,
uint32_t resource_id,
uint32_t x, uint32_t y,
@@ -591,6 +702,39 @@ void virtio_gpu_cmd_resource_flush(struct virtio_gpu_device *vgdev,
virtio_gpu_queue_fenced_ctrl_buffer(vgdev, vbuf, fence);
}
+/* For drm_panic */
+void virtio_gpu_panic_cmd_transfer_to_host_2d(struct virtio_gpu_device *vgdev,
+ uint64_t offset,
+ uint32_t width, uint32_t height,
+ uint32_t x, uint32_t y,
+ struct virtio_gpu_object_array *objs,
+ struct virtio_gpu_fence *fence,
+ struct virtio_gpu_vbuffer *vbuf)
fence is always NULL when calling this function, so you can remove it from the arguments.

+{
+ struct virtio_gpu_object *bo = gem_to_virtio_gpu_obj(objs->objs[0]);
+ struct virtio_gpu_transfer_to_host_2d *cmd_p;
+ bool use_dma_api = !virtio_has_dma_quirk(vgdev->vdev);
+
+ if (virtio_gpu_is_shmem(bo) && use_dma_api)
+ dma_sync_sgtable_for_device(vgdev->vdev->dev.parent,
+ bo->base.sgt, DMA_TO_DEVICE);
+
+ virtio_gpu_panic_alloc_cmd(vgdev, vbuf,
+ sizeof(struct virtio_gpu_transfer_to_host_2d));
+ cmd_p = (void *)vbuf->buf;
+ vbuf->objs = objs;
+
+ cmd_p->hdr.type = cpu_to_le32(VIRTIO_GPU_CMD_TRANSFER_TO_HOST_2D);
+ cmd_p->resource_id = cpu_to_le32(bo->hw_res_handle);
+ cmd_p->offset = cpu_to_le64(offset);
+ cmd_p->r.width = cpu_to_le32(width);
+ cmd_p->r.height = cpu_to_le32(height);
+ cmd_p->r.x = cpu_to_le32(x);
+ cmd_p->r.y = cpu_to_le32(y);
+
+ virtio_gpu_panic_queue_fenced_ctrl_buffer(vgdev, vbuf, fence);
+}
+
void virtio_gpu_cmd_transfer_to_host_2d(struct virtio_gpu_device *vgdev,
uint64_t offset,
uint32_t width, uint32_t height,