diff options
| author | Gary King <gking@nvidia.com> | 2010-08-02 15:55:16 -0700 |
|---|---|---|
| committer | Dan Willemsen <dwillemsen@nvidia.com> | 2012-03-21 22:11:42 -0700 |
| commit | 07797aa61561ee4ca10f016091b74bae4060dcea (patch) | |
| tree | 0b312a4ac9932d0441adb970c5c1a255249aa99f /arch/arm/mm/Kconfig | |
| parent | 19dd7a33996a7fb41c534208cdb5dc17dd956cae (diff) | |
[ARM] mm: add page allocator for modifying cache attributes
ARM CPUs with speculative prefetching have undefined behaviors when the
same physical page is mapped to two different virtual addresses with
conflicting cache attributes.
since many recent systems include IOMMU functionality (i.e., remapping
of discontiguous physical pages into a virtually-contiguous address
range for I/O devices), it is desirable to support allocating any
available OS memory for use by the I/O devices. however, since many
systems do not support cache coherency between the CPU and DMA devices,
these devices are left with using DMA-coherent allocations from the OS
(which severely limits the benefit of an IOMMU) or performing cache
maintenance (which can be a severe performance loss, particularly on
systems with outer caches, compared to using DMA-coherent memory).
this change adds an API for allocating pages from the OS with specific
cache maintenance properties and ensures that the kernel's mapping
of the page reflects the desired cache attributes, in line with the
ARMv7 architectural requirements
Change-Id: If0bd3cfe339b9a9b10fd6d45a748cd5e65931cf0
Signed-off-by: Gary King <gking@nvidia.com>
Rebase-Id: R0a86c40e1144d8da937993ebd942a8cf6ffbed11
Diffstat (limited to 'arch/arm/mm/Kconfig')
| -rw-r--r-- | arch/arm/mm/Kconfig | 23 |
1 files changed, 23 insertions, 0 deletions
diff --git a/arch/arm/mm/Kconfig b/arch/arm/mm/Kconfig index 7edef9121632..8b9f13156fc1 100644 --- a/arch/arm/mm/Kconfig +++ b/arch/arm/mm/Kconfig @@ -892,6 +892,29 @@ config ARM_L1_CACHE_SHIFT default 6 if ARM_L1_CACHE_SHIFT_6 default 5 +config ARM_ATTRIB_ALLOCATOR + bool "Support custom cache attribute allocations in low memory" + select ARCH_LOWMEM_IN_PTES if (CPU_V7) + depends on MMU && !CPU_CACHE_VIVT + help + Historically, the kernel has only reserved a small region + of physical memory for uncached access, and relied on + explicit cache maintenance for ensuring coherency between + the CPU and DMA. + + However, many recent systems support mapping discontiguous + physical pages into contiguous DMA addresses (so-called + system MMUs). For some DMA clients (notably graphics and + multimedia engines), performing explict cache maintenance + between CPU and DMA mappings can be prohibitively expensive, + and since ARMv7, mapping the same physical page with different + cache attributes is disallowed and has unpredictable behavior. + + Say 'Y' here to include page allocation support with explicit + cache attributes; on ARMv7 systems this will also force the + kernel's page tables to be mapped using page tables rather + than sections. + config ARM_DMA_MEM_BUFFERABLE bool "Use non-cacheable memory for DMA" if (CPU_V6 || CPU_V6K) && !CPU_V7 depends on !(MACH_REALVIEW_PB1176 || REALVIEW_EB_ARM11MP || \ |