Suspend will disable pcie device. Thus, resume should do full hw initialization again. Add some APIs to ast_drm_thaw() before ast_post_gpu() to fix the issue. Fixes: 5b71707dd13 ("drm/ast: Enable and unlock device access early during init") Signed-off-by: Jammy Huang <jammy_huang@xxxxxxxxxxxxxx> --- drivers/gpu/drm/ast/ast_drv.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/drivers/gpu/drm/ast/ast_drv.c b/drivers/gpu/drm/ast/ast_drv.c index f8c49ba68e78..45a9c7bf49c8 100644 --- a/drivers/gpu/drm/ast/ast_drv.c +++ b/drivers/gpu/drm/ast/ast_drv.c @@ -391,6 +391,11 @@ static int ast_drm_freeze(struct drm_device *dev) static int ast_drm_thaw(struct drm_device *dev) { + struct ast_device *ast = to_ast_device(dev); + + ast_enable_vga(ioregs); + ast_open_key(ioregs); + ast_enable_mmio(dev, ioregs); ast_post_gpu(dev); return drm_mode_config_helper_resume(dev); base-commit: e9d22f7a6655941fc8b2b942ed354ec780936b3e prerequisite-patch-id: a03a33902f33bcc38091e2cdce9d79f630464c30 prerequisite-patch-id: 505779a1e2094f2ee2f2a55ad44aac5cd1d5342f -- 2.25.1