ARM: tegra: speed up framebuffer copy
Jon Mayo [Tue, 30 Aug 2011 01:09:16 +0000 (18:09 -0700)]
Use a memcpy with less overhead in tegra_move_framebuffer, this makes
this function about 30 times faster.

Bug 843089

Original-Change-Id: I4ae9127db6d5ff5d9680e3ff2c3d28463395e39b
Reviewed-on: http://git-master/r/49735
Reviewed-by: Jonathan Mayo <jmayo@nvidia.com>
Tested-by: Jonathan Mayo <jmayo@nvidia.com>
Reviewed-by: Krishna Reddy <vdumpa@nvidia.com>
Reviewed-by: Yu-Huan Hsu <yhsu@nvidia.com>
Reviewed-by: Kevin Huang <kevinh@nvidia.com>

Rebase-Id: R0906917433643ac4ce9ac97284007527ef2d67df

arch/arm/mach-tegra/common.c

index 1d51838..7d3d76b 100644 (file)
@@ -444,7 +444,7 @@ void tegra_move_framebuffer(unsigned long to, unsigned long from,
                for (i = 0 ; i < size; i += PAGE_SIZE) {
                        page = phys_to_page(from + i);
                        from_virt = kmap(page);
-                       memcpy_toio(to_io + i, from_virt, PAGE_SIZE);
+                       memcpy(to_io + i, from_virt, PAGE_SIZE);
                        kunmap(page);
                }
        } else {