diff options
author | Thomas Petazzoni <thomas.petazzoni@free-electrons.com> | 2014-05-19 11:04:39 +0100 |
---|---|---|
committer | Russell King <rmk+kernel@arm.linux.org.uk> | 2014-06-01 01:15:04 +0100 |
commit | 1c8c3cf0b5239388e712508a85821f4718f4d889 (patch) | |
tree | 8454ad999882d20118d1af924c0cf6d2d42c97bd /arch/x86/Makefile | |
parent | 437b680a222098f67ecb1690f8eb91f550bfaac8 (diff) | |
download | linux-1c8c3cf0b5239388e712508a85821f4718f4d889.tar.gz linux-1c8c3cf0b5239388e712508a85821f4718f4d889.tar.bz2 linux-1c8c3cf0b5239388e712508a85821f4718f4d889.zip |
ARM: 8060/1: mm: allow sub-architectures to override PCI I/O memory type
Due to a design incompatibility between the PCIe Marvell controller
and the Cortex-A9, stressing PCIe devices with a lot of traffic
quickly causes a deadlock.
One part of the workaround for this is to have all PCIe regions mapped
as strongly-ordered (MT_UNCACHED) instead of the default
MT_DEVICE. While the arch_ioremap_caller() mechanism allows
sub-architecture code to override ioremap(), used to map PCIe memory
regions, there isn't such a mechanism to override the behavior of
pci_ioremap_io().
This commit adds the arch_pci_ioremap_mem_type variable, initialized
to MT_DEVICE by default, and that sub-architecture code can
override. We have chosen to expose a single variable rather than
offering the possibility of overriding the entire pci_ioremap_io(),
because implementing pci_ioremap_io() requires calling functions
(get_mem_type()) that are private to the arch/arm/mm/ code.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Acked-by: Catalin Marinas <catalin.marinas@arm.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Diffstat (limited to 'arch/x86/Makefile')
0 files changed, 0 insertions, 0 deletions