Replace the uncached memory mapping of the framebuffer with a write-
combining one. This improves performance, and avoids issues with
unaligned accesses and DC ZVA instructions performed by the accelerated
memcpy/memset routines.
Instead of manipulating the memory attributes directly, use the
SetMemorySpaceAttributes() DXE services, which validates the attributes
against the capabilities of the region before making the actual change.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Reviewed-by: Jeremy Linton <jeremy.linton@arm.com>
Tested-by: Ryan Harkin <ryan.harkin@linaro.org>
Reviewed-by: Leif Lindholm <leif.lindholm@linaro.org>
When we fail to modify the memory attributes for the VRAM allocation,
the allocation - which was made using AllocatePages() - is freed using
FreePool(). This is incorrect by itself, but it masks a second bug, i.e.,
that the address of the allocation is not in VramBaseAddress but in
*VramBaseAddress. So fix both issues.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Reviewed-by: Jeremy Linton <jeremy.linton@arm.com>
Tested-by: Ryan Harkin <ryan.harkin@linaro.org>
Reviewed-by: Leif Lindholm <leif.lindholm@linaro.org>
Ard Biesheuvel
Ronald Cron
oliviermartin