On Fri, 30 Sep 2011 20:59:39 -0700, Ben Widawsky <[email protected]> wrote: > Clean the code up, and always use a BO when creating a new buffer. I've > not seen any regressions but haven't yet tried this on < Gen6.
You have a bunch of different patches squashed into this here:
Replacing GL_TRUE/GL_FALSE with true/false.
Using a BO even for range-flushes
Rearranging logic.
Using a BO for range flushes is a definite behavior change that should
be separate from others and not hidden in a patch titled "i965: Cleanup
MapRangeBuffer".
> + if (intel_obj->buffer == NULL)
> + goto error_out;
> +
> /* If the mapping is synchronized with other GL operations, flush
> * the batchbuffer so that GEM knows about the buffer access for later
> * syncing.
> @@ -322,11 +325,6 @@ intel_bufferobj_map_range(struct gl_context * ctx,
> drm_intel_bo_references(intel->batch.bo, intel_obj->buffer))
> intel_flush(ctx);
>
> - if (intel_obj->buffer == NULL) {
> - obj->Pointer = NULL;
> - return NULL;
> - }
> -
I don't object to moving the code up, but making a goto for error case
used only once seems wrong.
pgpmKOKwqGpxc.pgp
Description: PGP signature
_______________________________________________ mesa-dev mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/mesa-dev
