diff options
author | Keith Whitwell <[email protected]> | 2008-05-01 11:28:47 +0100 |
---|---|---|
committer | Keith Whitwell <[email protected]> | 2008-05-01 12:00:45 +0100 |
commit | 27e46611f04108765fa99890822a474820d5c563 (patch) | |
tree | 1c4a98fc2895da0543e54c2f3a1989114e958ae1 /src/gallium/drivers/softpipe/sp_texture.c | |
parent | c9ed86a96483063f3d6789ed16645a3dca77d726 (diff) |
softpipe: use CPU flags for mapping
But when creating surfaces, adjust incoming flags from GPU->CPU usage.
Diffstat (limited to 'src/gallium/drivers/softpipe/sp_texture.c')
-rw-r--r-- | src/gallium/drivers/softpipe/sp_texture.c | 34 |
1 files changed, 31 insertions, 3 deletions
diff --git a/src/gallium/drivers/softpipe/sp_texture.c b/src/gallium/drivers/softpipe/sp_texture.c index ee3fa994f94..2b31cd4f251 100644 --- a/src/gallium/drivers/softpipe/sp_texture.c +++ b/src/gallium/drivers/softpipe/sp_texture.c @@ -151,6 +151,23 @@ softpipe_get_tex_surface(struct pipe_screen *screen, ps->pitch = ps->width; ps->offset = spt->level_offset[level]; ps->usage = usage; + + /* Because we are softpipe, anything that the state tracker + * thought was going to be done with the GPU will actually get + * done with the CPU. Let's adjust the flags to take that into + * account. + */ + if (ps->usage & PIPE_BUFFER_USAGE_GPU_WRITE) + ps->usage |= PIPE_BUFFER_USAGE_CPU_WRITE; + + if (ps->usage & PIPE_BUFFER_USAGE_GPU_READ) + ps->usage |= PIPE_BUFFER_USAGE_CPU_READ; + + + pipe_texture_reference(&ps->texture, pt); + ps->face = face; + ps->level = level; + ps->zslice = zslice; if (pt->target == PIPE_TEXTURE_CUBE || pt->target == PIPE_TEXTURE_3D) { ps->offset += ((pt->target == PIPE_TEXTURE_CUBE) ? face : zslice) * @@ -164,8 +181,18 @@ softpipe_get_tex_surface(struct pipe_screen *screen, if (usage & (PIPE_BUFFER_USAGE_CPU_WRITE | PIPE_BUFFER_USAGE_GPU_WRITE)) { - /* XXX if writing to the texture, invalidate the texcache entries!!! */ - assert(0); + /* XXX if writing to the texture, invalidate the texcache entries!!! + * + * Actually, no. Flushing dependent contexts is still done + * explicitly and separately. Hardware drivers won't insert + * FLUSH commands into a command stream at this point, + * neither should softpipe try to flush caches. + * + * Those contexts could be living in separate threads & doing + * all sorts of unrelated stuff... Context<->texture + * dependency tracking needs to happen elsewhere. + */ + /* assert(0); */ } } return ps; @@ -181,6 +208,7 @@ softpipe_tex_surface_release(struct pipe_screen *screen, * where it would happen. For softpipe, nothing to do. */ assert ((*s)->texture); + pipe_texture_reference(&(*s)->texture, NULL); screen->winsys->surface_release(screen->winsys, s); } @@ -206,7 +234,7 @@ softpipe_surface_map( struct pipe_screen *screen, * of the map: */ if (surface->texture && - (flags & PIPE_BUFFER_USAGE_GPU_WRITE)) + (flags & PIPE_BUFFER_USAGE_CPU_WRITE)) { /* Do something to notify sharing contexts of a texture change. * In softpipe, that would mean flushing the texture cache. |