Hi,
> diff --git a/hw/display/virtio-gpu-udmabuf.c b/hw/display/virtio-gpu-udmabuf.c
> index 2c0e7b2455..8bbce08550 100644
> --- a/hw/display/virtio-gpu-udmabuf.c
> +++ b/hw/display/virtio-gpu-udmabuf.c
> @@ -72,7 +72,10 @@ static void virtio_gpu_create_udmabuf(struct
> virtio_gpu_simple_resource *res)
>
> static void virtio_gpu_remap_udmabuf(struct virtio_gpu_simple_resource *res)
> {
> - res->remapsz = res->width * res->height * 4;
> + if (res->blob_size) {
> + res->remapsz = res->blob_size;
> + }
> +
> res->remapsz = QEMU_ALIGN_UP(res->remapsz, qemu_real_host_page_size);
>
> res->remapped = mmap(NULL, res->remapsz, PROT_READ,
> @@ -152,7 +155,9 @@ void virtio_gpu_init_udmabuf(struct
> virtio_gpu_simple_resource *res)
> pdata = res->remapped;
> }
>
> - (void) pdata;
> + if (pdata) {
> + res->blob = pdata;
> + }
> }
That is confusing. Patch #2 introduces code which is removed here
(probably leftovers from udmabuf prototype which didn't use blob
resources). I think it would be useful to merge patch #2 with this one.
Also we might cleanup things a bit. Do we need separate remapsz +
blob_size fields in the first place? Probably not ...
take care,
Gerd
PS: Can you explicitly cc me on the next version of the patch series?
Some patches (this one for example) are on the list only and not in
my inbox. Thanks.
