From patchwork Wed Sep 5 23:38:57 2018 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Deepak Singh Rawat X-Patchwork-Id: 10589637 Return-Path: Received: from mail.wl.linuxfoundation.org (pdx-wl-mail.web.codeaurora.org [172.30.200.125]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id A611613BB for ; Wed, 5 Sep 2018 23:39:50 +0000 (UTC) Received: from mail.wl.linuxfoundation.org (localhost [127.0.0.1]) by mail.wl.linuxfoundation.org (Postfix) with ESMTP id 95AE12AA6B for ; Wed, 5 Sep 2018 23:39:50 +0000 (UTC) Received: by mail.wl.linuxfoundation.org (Postfix, from userid 486) id 898412AA78; Wed, 5 Sep 2018 23:39:50 +0000 (UTC) X-Spam-Checker-Version: SpamAssassin 3.3.1 (2010-03-16) on pdx-wl-mail.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-5.2 required=2.0 tests=BAYES_00,MAILING_LIST_MULTI, RCVD_IN_DNSWL_MED autolearn=ham version=3.3.1 Received: from gabe.freedesktop.org (gabe.freedesktop.org [131.252.210.177]) (using TLSv1.2 with cipher DHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.wl.linuxfoundation.org (Postfix) with ESMTPS id 0B9732AA6B for ; Wed, 5 Sep 2018 23:39:50 +0000 (UTC) Received: from gabe.freedesktop.org (localhost [127.0.0.1]) by gabe.freedesktop.org (Postfix) with ESMTP id 636A46E577; Wed, 5 Sep 2018 23:39:40 +0000 (UTC) X-Original-To: dri-devel@lists.freedesktop.org Delivered-To: dri-devel@lists.freedesktop.org Received: from EX13-EDG-OU-002.vmware.com (ex13-edg-ou-002.vmware.com [208.91.0.190]) by gabe.freedesktop.org (Postfix) with ESMTPS id 946D56E574 for ; Wed, 5 Sep 2018 23:39:35 +0000 (UTC) Received: from sc9-mailhost2.vmware.com (10.113.161.72) by EX13-EDG-OU-002.vmware.com (10.113.208.156) with Microsoft SMTP Server id 15.0.1156.6; Wed, 5 Sep 2018 16:39:28 -0700 Received: from ubuntu.localdomain (promb-2n-dhcp79.eng.vmware.com [10.20.88.79]) by sc9-mailhost2.vmware.com (Postfix) with ESMTP id 462AFB0749; Wed, 5 Sep 2018 19:39:35 -0400 (EDT) From: Deepak Rawat To: , , , Subject: [PATCH 10/14] drm/vmwgfx: implement SOU plane update for surface backed fb Date: Wed, 5 Sep 2018 16:38:57 -0700 Message-ID: <20180905233901.2321-11-drawat@vmware.com> X-Mailer: git-send-email 2.17.1 In-Reply-To: <20180905233901.2321-1-drawat@vmware.com> References: <20180905233901.2321-1-drawat@vmware.com> MIME-Version: 1.0 Received-SPF: None (EX13-EDG-OU-002.vmware.com: drawat@vmware.com does not designate permitted sender hosts) X-BeenThere: dri-devel@lists.freedesktop.org X-Mailman-Version: 2.1.23 Precedence: list List-Id: Direct Rendering Infrastructure - Development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Cc: Deepak Rawat , lukasz.spintzyk@displaylink.com Errors-To: dri-devel-bounces@lists.freedesktop.org Sender: "dri-devel" X-Virus-Scanned: ClamAV using ClamSMTP Using the new interface implement SOU plane update for surface backed fb. Signed-off-by: Deepak Rawat --- drivers/gpu/drm/vmwgfx/vmwgfx_kms.h | 3 + drivers/gpu/drm/vmwgfx/vmwgfx_scrn.c | 187 +++++++++++++++++++++++++++ 2 files changed, 190 insertions(+) diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_kms.h b/drivers/gpu/drm/vmwgfx/vmwgfx_kms.h index 1c3dd0e14414..45b12448e138 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_kms.h +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_kms.h @@ -165,10 +165,13 @@ struct vmw_du_update_plane { * * @base: base closure structure. * @ctx: validation context. + * @cmd_start: (optional) FIFO command start address (used by SOU only). */ struct vmw_du_update_plane_surface { struct vmw_du_update_plane base; struct vmw_validation_ctx *ctx; + /* This member is to handle special case SOU surface update */ + void *cmd_start; }; /** diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_scrn.c b/drivers/gpu/drm/vmwgfx/vmwgfx_scrn.c index ad0de7f0cd60..8427a51cde3f 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_scrn.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_scrn.c @@ -499,6 +499,193 @@ vmw_sou_primary_plane_prepare_fb(struct drm_plane *plane, return vmw_bo_pin_in_vram(dev_priv, vps->bo, true); } +static int vmw_sou_surface_prepare(struct vmw_du_update_plane *update) +{ + struct vmw_du_update_plane_surface *srf_update; + struct vmw_framebuffer_surface *vfbs; + + srf_update = container_of(update, typeof(*srf_update), base); + vfbs = container_of(update->vfb, typeof(*vfbs), base); + + return vmw_kms_helper_resource_prepare(&vfbs->surface->res, true, + srf_update->ctx); +} + +static uint32_t vmw_sou_surface_fifo_size(struct vmw_du_update_plane *update, + uint32_t num_hits) +{ + return sizeof(struct vmw_kms_sou_dirty_cmd) + sizeof(SVGASignedRect) * + num_hits; +} + +static uint32_t vmw_sou_surface_post_prepare(struct vmw_du_update_plane *update, + void *cmd) +{ + struct vmw_du_update_plane_surface *srf_update; + + srf_update = container_of(update, typeof(*srf_update), base); + + /* + * SOU SVGA_3D_CMD_BLIT_SURFACE_TO_SCREEN is special in the sense that + * it bounding box is filled before iterating over all the clips. So + * store the FIFO start address and revisit to fill the details. + */ + srf_update->cmd_start = cmd; + + return 0; +} + +static uint32_t vmw_sou_surface_pre_clip(struct vmw_du_update_plane *update, + void *cmd, uint32_t num_hits) +{ + struct vmw_kms_sou_dirty_cmd *blit = cmd; + struct vmw_framebuffer_surface *vfbs; + + vfbs = container_of(update->vfb, typeof(*vfbs), base); + + blit->header.id = SVGA_3D_CMD_BLIT_SURFACE_TO_SCREEN; + blit->header.size = sizeof(blit->body) + sizeof(SVGASignedRect) * + num_hits; + + blit->body.srcImage.sid = vfbs->surface->res.id; + blit->body.destScreenId = update->du->unit; + + /* Update the source and destination bounding box later in post_clip */ + blit->body.srcRect.left = 0; + blit->body.srcRect.top = 0; + blit->body.srcRect.right = 0; + blit->body.srcRect.bottom = 0; + + blit->body.destRect.left = 0; + blit->body.destRect.top = 0; + blit->body.destRect.right = 0; + blit->body.destRect.bottom = 0; + + return sizeof(*blit); +} + +static uint32_t vmw_sou_surface_clip_rect(struct vmw_du_update_plane *update, + void *cmd, struct drm_rect *clip, + uint32_t src_x, uint32_t src_y) +{ + SVGASignedRect *rect = cmd; + + /* + * rects are relative to dest bounding box rect on screen object, so + * translate to it later in post_clip + */ + rect->left = clip->x1; + rect->top = clip->y1; + rect->right = clip->x2; + rect->bottom = clip->y2; + + return sizeof(*rect); +} + +static uint32_t vmw_sou_surface_post_clip(struct vmw_du_update_plane *update, + void *cmd, struct drm_rect *bb) +{ + struct vmw_du_update_plane_surface *srf_update; + struct drm_plane_state *state = update->plane->state; + struct drm_rect src_bb; + struct vmw_kms_sou_dirty_cmd *blit; + SVGASignedRect *rect; + uint32_t num_hits; + int translate_src_x; + int translate_src_y; + int i; + + srf_update = container_of(update, typeof(*srf_update), base); + + blit = srf_update->cmd_start; + rect = (SVGASignedRect *)&blit[1]; + + num_hits = (blit->header.size - sizeof(blit->body))/ + sizeof(SVGASignedRect); + + src_bb = *bb; + + /* To translate bb back to fb src coord */ + translate_src_x = (state->src_x >> 16) - state->crtc_x; + translate_src_y = (state->src_y >> 16) - state->crtc_y; + + drm_rect_translate(&src_bb, translate_src_x, translate_src_y); + + blit->body.srcRect.left = src_bb.x1; + blit->body.srcRect.top = src_bb.y1; + blit->body.srcRect.right = src_bb.x2; + blit->body.srcRect.bottom = src_bb.y2; + + blit->body.destRect.left = bb->x1; + blit->body.destRect.top = bb->y1; + blit->body.destRect.right = bb->x2; + blit->body.destRect.bottom = bb->y2; + + /* rects are relative to dest bb rect */ + for (i = 0; i < num_hits; i++) { + rect->left -= bb->x1; + rect->top -= bb->y1; + rect->right -= bb->x1; + rect->bottom -= bb->y1; + rect++; + } + + return 0; +} + +static void vmw_sou_surface_finish(struct vmw_du_update_plane *update) +{ + struct vmw_du_update_plane_surface *srf_update = + container_of(update, typeof(*srf_update), base); + + vmw_kms_helper_resource_finish(srf_update->ctx, update->out_fence); +} + +/** + * vmw_sou_plane_update_surface - update display unit for surface backed fb + * @dev_priv: device private + * @plane: plane state + * @old_state: old plane state + * @vfb: framebuffer which is blitted to display unit + * @out_fence: (optional) If non-NULL, will return a ref-counted pointer to a + * struct vmw_fence_obj. The returned fence pointer may be NULL in which case + * the device has already synchronized. + * + * RETURNS: + * + * 0 on success or a negative error code on failure. + */ +static int vmw_sou_plane_update_surface(struct vmw_private *dev_priv, + struct drm_plane *plane, + struct drm_plane_state *old_state, + struct vmw_framebuffer *vfb, + struct vmw_fence_obj **out_fence) +{ + struct vmw_du_update_plane_surface srf_update; + struct vmw_validation_ctx ctx; + + memset(&srf_update, 0, sizeof(struct vmw_du_update_plane_surface)); + srf_update.base.plane = plane; + srf_update.base.old_state = old_state; + srf_update.base.dev_priv = dev_priv; + srf_update.base.du = vmw_crtc_to_du(plane->state->crtc); + srf_update.base.vfb = vfb; + srf_update.base.out_fence = out_fence; + + srf_update.ctx = &ctx; + + srf_update.base.prepare = vmw_sou_surface_prepare; + srf_update.base.calc_fifo_size = vmw_sou_surface_fifo_size; + srf_update.base.post_prepare = vmw_sou_surface_post_prepare; + srf_update.base.pre_clip = vmw_sou_surface_pre_clip; + srf_update.base.clip = vmw_sou_surface_clip_rect; + srf_update.base.post_clip = vmw_sou_surface_post_clip; + srf_update.base.finish = vmw_sou_surface_finish; + /* Nothing special to do for revert */ + srf_update.base.revert = vmw_sou_surface_finish; + + return vmw_du_helper_plane_update(&srf_update.base); +} static void vmw_sou_primary_plane_atomic_update(struct drm_plane *plane,