These patches add support to the Qualcomm msm8660 target. It currently
provides uniprocessor support only, though SMP support is forthcoming.
Note that this code is based on top of Gregory Bean's gpiomux code, and
won't build until the mach-types are updated.
Abhijeet Dharmapurikar (3):
GIC: Dont disable INT in ack callback
msm: irqs-8x60: interrupt map
msm: 8x60: setup correct handlers for private interrupts
Daniel Walker (3):
arm: mm: add proc info for ScorpionMP
msm: add hotplug stub functions
msm: allow uart to be conditionally disabled
Gregory Bean (1):
msm: add MSM8x60 FFA support
Jeff Ohlstein (5):
msm: timer: support 8x60 timers
msm: clock: add dummy clock driver
dma: add stub functions for dma features not yet present on 8x60
msm: add build support for msm8x60 target
msm: physical offset for MSM8X60
Stepan Moskovchenko (3):
msm: Add MSM IOMMU support
msm: Platform initialization for the IOMMU driver
msm: Platform data for msm8x60 IOMMUs
Steve Muckle (9):
msm: create config option for proc-comm
arm: Kconfig option for ARCH_MSM_SCORPIONMP
msm: io: MSM8X60 io support
msm: initial irq definitions for MSM8X60
msm: MSM8X60 RUMI3 board support
msm: irq: rename existing entry-macro to entry-macro-vic
msm: 8x60: gic initialization fixup for RUMI
msm: add msm8x60_surf machine
msm: MSM8X60 simulator board support
arch/arm/Kconfig | 3 +-
arch/arm/common/gic.c | 14 -
arch/arm/mach-msm/Kconfig | 52 +-
arch/arm/mach-msm/Makefile | 20 +-
arch/arm/mach-msm/board-msm8x60.c | 102 ++
arch/arm/mach-msm/clock-dummy.c | 54 +
arch/arm/mach-msm/devices-msm8x60-iommu.c | 883 ++++++++++
arch/arm/mach-msm/hotplug.c | 22 +
arch/arm/mach-msm/include/mach/debug-macro.S | 2 +-
arch/arm/mach-msm/include/mach/dma.h | 10 +-
arch/arm/mach-msm/include/mach/entry-macro-qgic.S | 96 ++
arch/arm/mach-msm/include/mach/entry-macro-vic.S | 37 +
arch/arm/mach-msm/include/mach/entry-macro.S | 43 +-
arch/arm/mach-msm/include/mach/io.h | 1 +
arch/arm/mach-msm/include/mach/iommu.h | 103 ++
arch/arm/mach-msm/include/mach/iommu_hw-8xxx.h | 1871 +++++++++++++++++++++
arch/arm/mach-msm/include/mach/irqs-8x60.h | 267 +++
arch/arm/mach-msm/include/mach/irqs.h | 2 +
arch/arm/mach-msm/include/mach/memory.h | 2 +
arch/arm/mach-msm/include/mach/msm_iomap-8x60.h | 59 +
arch/arm/mach-msm/include/mach/smp.h | 39 +
arch/arm/mach-msm/io.c | 15 +
arch/arm/mach-msm/iommu.c | 597 +++++++
arch/arm/mach-msm/iommu_dev.c | 364 ++++
arch/arm/mach-msm/timer.c | 226 ++-
arch/arm/mm/proc-v7.S | 24 +
26 files changed, 4798 insertions(+), 110 deletions(-)
create mode 100644 arch/arm/mach-msm/board-msm8x60.c
create mode 100644 arch/arm/mach-msm/clock-dummy.c
create mode 100644 arch/arm/mach-msm/devices-msm8x60-iommu.c
create mode 100644 arch/arm/mach-msm/hotplug.c
create mode 100644 arch/arm/mach-msm/include/mach/entry-macro-qgic.S
create mode 100644 arch/arm/mach-msm/include/mach/entry-macro-vic.S
create mode 100644 arch/arm/mach-msm/include/mach/iommu.h
create mode 100644 arch/arm/mach-msm/include/mach/iommu_hw-8xxx.h
create mode 100644 arch/arm/mach-msm/include/mach/irqs-8x60.h
create mode 100644 arch/arm/mach-msm/include/mach/smp.h
create mode 100644 arch/arm/mach-msm/iommu.c
create mode 100644 arch/arm/mach-msm/iommu_dev.c
--
1.7.2.1
Sent by an employee of the Qualcomm Innovation Center, Inc.
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum.
Need to add this until real clock support for 8x60 goes in, or else some
drivers won't compile.
Signed-off-by: Jeff Ohlstein <[email protected]>
---
arch/arm/mach-msm/Makefile | 1 +
arch/arm/mach-msm/clock-dummy.c | 54 +++++++++++++++++++++++++++++++++++++++
2 files changed, 55 insertions(+), 0 deletions(-)
create mode 100644 arch/arm/mach-msm/clock-dummy.c
diff --git a/arch/arm/mach-msm/Makefile b/arch/arm/mach-msm/Makefile
index 64a95ae..ff19e01 100644
--- a/arch/arm/mach-msm/Makefile
+++ b/arch/arm/mach-msm/Makefile
@@ -7,6 +7,7 @@ else
obj-y += irq.o
endif
+obj-$(CONFIG_ARCH_MSM8X60) += clock-dummy.o
obj-$(CONFIG_MSM_PROC_COMM) += proc_comm.o clock-pcom.o vreg.o
obj-$(CONFIG_MSM_PROC_COMM) += clock.o gpio.o
obj-$(CONFIG_ARCH_QSD8X50) += sirc.o
diff --git a/arch/arm/mach-msm/clock-dummy.c b/arch/arm/mach-msm/clock-dummy.c
new file mode 100644
index 0000000..1250d22
--- /dev/null
+++ b/arch/arm/mach-msm/clock-dummy.c
@@ -0,0 +1,54 @@
+/* Copyright (c) 2010, Code Aurora Forum. All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 and
+ * only version 2 as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
+ * 02110-1301, USA.
+ *
+ */
+#include <linux/clk.h>
+#include <linux/err.h>
+#include <linux/module.h>
+
+struct clk *clk_get(struct device *dev, const char *id)
+{
+ return ERR_PTR(-ENOENT);
+}
+EXPORT_SYMBOL(clk_get);
+
+int clk_enable(struct clk *clk)
+{
+ return -ENOENT;
+}
+EXPORT_SYMBOL(clk_enable);
+
+void clk_disable(struct clk *clk)
+{
+}
+EXPORT_SYMBOL(clk_disable);
+
+unsigned long clk_get_rate(struct clk *clk)
+{
+ return 0;
+}
+EXPORT_SYMBOL(clk_get_rate);
+
+int clk_set_rate(struct clk *clk, unsigned long rate)
+{
+ return -ENOENT;
+}
+EXPORT_SYMBOL(clk_set_rate);
+
+void clk_put(struct clk *clk)
+{
+}
+EXPORT_SYMBOL(clk_put);
--
1.7.2.1
Sent by an employee of the Qualcomm Innovation Center, Inc.
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum.
From: Steve Muckle <[email protected]>
Some builds may not support the proc-comm interface with
the baseband processor.
Signed-off-by: Steve Muckle <[email protected]>
---
arch/arm/mach-msm/Kconfig | 7 +++++++
arch/arm/mach-msm/Makefile | 6 ++----
2 files changed, 9 insertions(+), 4 deletions(-)
diff --git a/arch/arm/mach-msm/Kconfig b/arch/arm/mach-msm/Kconfig
index 4e79580..08ed76c 100644
--- a/arch/arm/mach-msm/Kconfig
+++ b/arch/arm/mach-msm/Kconfig
@@ -10,6 +10,7 @@ config ARCH_MSM7X00A
select MSM_SMD
select MSM_SMD_PKG3
select CPU_V6
+ select MSM_PROC_COMM
config ARCH_MSM7X30
bool "MSM7x30"
@@ -18,6 +19,7 @@ config ARCH_MSM7X30
select MSM_VIC
select CPU_V7
select MSM_REMOTE_SPINLOCK_DEKKERS
+ select MSM_PROC_COMM
config ARCH_QSD8X50
bool "QSD8X50"
@@ -26,6 +28,8 @@ config ARCH_QSD8X50
select MSM_VIC
select CPU_V7
select MSM_REMOTE_SPINLOCK_LDREX
+ select MSM_PROC_COMM
+
endchoice
config MSM_SOC_REV_A
@@ -103,6 +107,9 @@ endchoice
config MSM_SMD_PKG3
bool
+config MSM_PROC_COMM
+ bool
+
config MSM_SMD
bool
diff --git a/arch/arm/mach-msm/Makefile b/arch/arm/mach-msm/Makefile
index faeb1f9..78fb536 100644
--- a/arch/arm/mach-msm/Makefile
+++ b/arch/arm/mach-msm/Makefile
@@ -1,8 +1,5 @@
-obj-y += proc_comm.o
obj-y += io.o idle.o timer.o dma.o
-obj-y += vreg.o
obj-y += acpuclock-arm11.o
-obj-y += clock.o clock-pcom.o
ifdef CONFIG_MSM_VIC
obj-y += irq-vic.o
@@ -10,10 +7,11 @@ else
obj-y += irq.o
endif
+obj-$(CONFIG_MSM_PROC_COMM) += proc_comm.o clock-pcom.o vreg.o
+obj-$(CONFIG_MSM_PROC_COMM) += clock.o gpio.o
obj-$(CONFIG_ARCH_QSD8X50) += sirc.o
obj-$(CONFIG_MSM_SMD) += smd.o smd_debug.o
obj-$(CONFIG_MSM_SMD) += last_radio_log.o
-
obj-$(CONFIG_MACH_TROUT) += board-trout.o board-trout-gpio.o board-trout-mmc.o devices-msm7x00.o
obj-$(CONFIG_MACH_HALIBUT) += board-halibut.o devices-msm7x00.o
obj-$(CONFIG_ARCH_MSM7X30) += board-msm7x30.o devices-msm7x30.o
--
1.7.2.1
Sent by an employee of the Qualcomm Innovation Center, Inc.
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum.
From: Steve Muckle <[email protected]>
Signed-off-by: Steve Muckle <[email protected]>
---
arch/arm/mach-msm/Kconfig | 7 +++++++
arch/arm/mach-msm/Makefile | 1 +
arch/arm/mach-msm/board-msm8x60.c | 7 +++++++
3 files changed, 15 insertions(+), 0 deletions(-)
diff --git a/arch/arm/mach-msm/Kconfig b/arch/arm/mach-msm/Kconfig
index dfa67d8..3017e12 100644
--- a/arch/arm/mach-msm/Kconfig
+++ b/arch/arm/mach-msm/Kconfig
@@ -98,6 +98,13 @@ config MACH_MSM8X60_RUMI3
help
Support for the Qualcomm MSM8x60 RUMI3 emulator.
+config MACH_MSM8X60_SURF
+ depends on ARCH_MSM8X60
+ default n
+ bool "MSM8x60 SURF"
+ help
+ Support for the Qualcomm MSM8x60 SURF eval board.
+
endmenu
config MSM_DEBUG_UART
diff --git a/arch/arm/mach-msm/Makefile b/arch/arm/mach-msm/Makefile
index 17c03fa..1d62025 100644
--- a/arch/arm/mach-msm/Makefile
+++ b/arch/arm/mach-msm/Makefile
@@ -28,3 +28,4 @@ obj-$(CONFIG_ARCH_QSD8X50) += gpiomux-8x50.o gpiomux-7x00.o
obj-$(CONFIG_ARCH_MSM8X60) += gpiomux-8x60.o
obj-$(CONFIG_HOTPLUG_CPU) += hotplug.o
obj-$(CONFIG_MACH_MSM8X60_RUMI3) += board-msm8x60.o
+obj-$(CONFIG_MACH_MSM8X60_SURF) += board-msm8x60.o
diff --git a/arch/arm/mach-msm/board-msm8x60.c b/arch/arm/mach-msm/board-msm8x60.c
index e2e5a5c..fd953c2 100644
--- a/arch/arm/mach-msm/board-msm8x60.c
+++ b/arch/arm/mach-msm/board-msm8x60.c
@@ -78,3 +78,10 @@ MACHINE_START(MSM8X60_RUMI3, "QCT MSM8X60 RUMI3")
.init_machine = msm8x60_init,
.timer = &msm_timer,
MACHINE_END
+
+MACHINE_START(MSM8X60_SURF, "QCT MSM8X60 SURF")
+ .map_io = msm8x60_map_io,
+ .init_irq = msm8x60_init_irq,
+ .init_machine = msm8x60_init,
+ .timer = &msm_timer,
+MACHINE_END
--
1.7.2.1
Sent by an employee of the Qualcomm Innovation Center, Inc.
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum.
The MSM8x60 has a different physical memory offset than other targets.
Signed-off-by: Jeff Ohlstein <[email protected]>
---
arch/arm/mach-msm/include/mach/memory.h | 2 ++
1 files changed, 2 insertions(+), 0 deletions(-)
diff --git a/arch/arm/mach-msm/include/mach/memory.h b/arch/arm/mach-msm/include/mach/memory.h
index 50c7847..070e17d 100644
--- a/arch/arm/mach-msm/include/mach/memory.h
+++ b/arch/arm/mach-msm/include/mach/memory.h
@@ -23,6 +23,8 @@
#define PHYS_OFFSET UL(0x20000000)
#elif defined(CONFIG_ARCH_MSM7X30)
#define PHYS_OFFSET UL(0x00200000)
+#elif defined(CONFIG_ARCH_MSM8X60)
+#define PHYS_OFFSET UL(0x40200000)
#else
#define PHYS_OFFSET UL(0x10000000)
#endif
--
1.7.2.1
Sent by an employee of the Qualcomm Innovation Center, Inc.
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum.
Signed-off-by: Jeff Ohlstein <[email protected]>
---
arch/arm/mach-msm/Kconfig | 7 +++++++
arch/arm/mach-msm/Makefile | 7 ++++++-
2 files changed, 13 insertions(+), 1 deletions(-)
diff --git a/arch/arm/mach-msm/Kconfig b/arch/arm/mach-msm/Kconfig
index 27fc183..dfa67d8 100644
--- a/arch/arm/mach-msm/Kconfig
+++ b/arch/arm/mach-msm/Kconfig
@@ -33,6 +33,13 @@ config ARCH_QSD8X50
select MSM_PROC_COMM
select HAS_MSM_DEBUG_UART_PHYS
+config ARCH_MSM8X60
+ bool "MSM8X60"
+ select ARCH_MSM_SCORPIONMP
+ select ARM_GIC
+ select CPU_V7
+ select MSM_V2_TLMM
+
endchoice
config MSM_SOC_REV_A
diff --git a/arch/arm/mach-msm/Makefile b/arch/arm/mach-msm/Makefile
index 8a01f8a..17c03fa 100644
--- a/arch/arm/mach-msm/Makefile
+++ b/arch/arm/mach-msm/Makefile
@@ -1,11 +1,16 @@
-obj-y += io.o idle.o timer.o dma.o
+obj-y += io.o idle.o timer.o
+ifndef CONFIG_ARCH_MSM8X60
obj-y += acpuclock-arm11.o
+obj-y += gpio.o dma.o
+endif
ifdef CONFIG_MSM_VIC
obj-y += irq-vic.o
else
+ifndef CONFIG_ARCH_MSM8X60
obj-y += irq.o
endif
+endif
obj-$(CONFIG_ARCH_MSM8X60) += clock-dummy.o
obj-$(CONFIG_MSM_PROC_COMM) += proc_comm.o clock-pcom.o vreg.o
--
1.7.2.1
Sent by an employee of the Qualcomm Innovation Center, Inc.
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum.
From: Gregory Bean <[email protected]>
The MSM8X60 FFA contains different components than the MSM8X60 SURF,
and therefore requires a different ARCH type and machine ID.
Signed-off-by: Gregory Bean <[email protected]>
---
arch/arm/mach-msm/Kconfig | 7 +++++++
arch/arm/mach-msm/Makefile | 1 +
arch/arm/mach-msm/board-msm8x60.c | 7 +++++++
3 files changed, 15 insertions(+), 0 deletions(-)
diff --git a/arch/arm/mach-msm/Kconfig b/arch/arm/mach-msm/Kconfig
index d2610a3..d6b83de 100644
--- a/arch/arm/mach-msm/Kconfig
+++ b/arch/arm/mach-msm/Kconfig
@@ -112,6 +112,13 @@ config MACH_MSM8X60_SIM
help
Support for the Qualcomm MSM8x60 simulator.
+config MACH_MSM8X60_FFA
+ depends on ARCH_MSM8X60
+ default n
+ bool "MSM8x60 FFA"
+ help
+ Support for the Qualcomm MSM8x60 FFA eval board.
+
endmenu
config MSM_DEBUG_UART
diff --git a/arch/arm/mach-msm/Makefile b/arch/arm/mach-msm/Makefile
index 9fb73ab..c984f4d 100644
--- a/arch/arm/mach-msm/Makefile
+++ b/arch/arm/mach-msm/Makefile
@@ -30,3 +30,4 @@ obj-$(CONFIG_HOTPLUG_CPU) += hotplug.o
obj-$(CONFIG_MACH_MSM8X60_RUMI3) += board-msm8x60.o
obj-$(CONFIG_MACH_MSM8X60_SURF) += board-msm8x60.o
obj-$(CONFIG_MACH_MSM8X60_SIM) += board-msm8x60.o
+obj-$(CONFIG_MACH_MSM8X60_FFA) += board-msm8x60.o
diff --git a/arch/arm/mach-msm/board-msm8x60.c b/arch/arm/mach-msm/board-msm8x60.c
index 7a2ed7f..02d22ea 100644
--- a/arch/arm/mach-msm/board-msm8x60.c
+++ b/arch/arm/mach-msm/board-msm8x60.c
@@ -93,3 +93,10 @@ MACHINE_START(MSM8X60_SIM, "QCT MSM8X60 SIMULATOR")
.init_machine = msm8x60_init,
.timer = &msm_timer,
MACHINE_END
+
+MACHINE_START(MSM8X60_FFA, "QCT MSM8X60 FFA")
+ .map_io = msm8x60_map_io,
+ .init_irq = msm8x60_init_irq,
+ .init_machine = msm8x60_init,
+ .timer = &msm_timer,
+MACHINE_END
--
1.7.2.1
Sent by an employee of the Qualcomm Innovation Center, Inc.
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum.
From: Steve Muckle <[email protected]>
Board configuration for MSM8X60 simulation.
Signed-off-by: Steve Muckle <[email protected]>
---
arch/arm/mach-msm/Kconfig | 7 +++++++
arch/arm/mach-msm/Makefile | 1 +
arch/arm/mach-msm/board-msm8x60.c | 10 +++++++++-
3 files changed, 17 insertions(+), 1 deletions(-)
diff --git a/arch/arm/mach-msm/Kconfig b/arch/arm/mach-msm/Kconfig
index 3017e12..d2610a3 100644
--- a/arch/arm/mach-msm/Kconfig
+++ b/arch/arm/mach-msm/Kconfig
@@ -105,6 +105,13 @@ config MACH_MSM8X60_SURF
help
Support for the Qualcomm MSM8x60 SURF eval board.
+config MACH_MSM8X60_SIM
+ depends on ARCH_MSM8X60
+ default n
+ bool "MSM8x60 Simulator"
+ help
+ Support for the Qualcomm MSM8x60 simulator.
+
endmenu
config MSM_DEBUG_UART
diff --git a/arch/arm/mach-msm/Makefile b/arch/arm/mach-msm/Makefile
index 1d62025..9fb73ab 100644
--- a/arch/arm/mach-msm/Makefile
+++ b/arch/arm/mach-msm/Makefile
@@ -29,3 +29,4 @@ obj-$(CONFIG_ARCH_MSM8X60) += gpiomux-8x60.o
obj-$(CONFIG_HOTPLUG_CPU) += hotplug.o
obj-$(CONFIG_MACH_MSM8X60_RUMI3) += board-msm8x60.o
obj-$(CONFIG_MACH_MSM8X60_SURF) += board-msm8x60.o
+obj-$(CONFIG_MACH_MSM8X60_SIM) += board-msm8x60.o
diff --git a/arch/arm/mach-msm/board-msm8x60.c b/arch/arm/mach-msm/board-msm8x60.c
index fd953c2..7a2ed7f 100644
--- a/arch/arm/mach-msm/board-msm8x60.c
+++ b/arch/arm/mach-msm/board-msm8x60.c
@@ -56,7 +56,8 @@ static void __init msm8x60_init_irq(void)
/* RUMI does not adhere to GIC spec by enabling STIs by default.
* Enable/clear is supposed to be RO for STIs, but is RW on RUMI.
*/
- writel(0x0000FFFF, MSM_QGIC_DIST_BASE + GIC_DIST_ENABLE_SET);
+ if (!machine_is_msm8x60_sim())
+ writel(0x0000FFFF, MSM_QGIC_DIST_BASE + GIC_DIST_ENABLE_SET);
/* FIXME: Not installing AVS_SVICINT and AVS_SVICINTSWDONE yet
* as they are configured as level, which does not play nice with
@@ -85,3 +86,10 @@ MACHINE_START(MSM8X60_SURF, "QCT MSM8X60 SURF")
.init_machine = msm8x60_init,
.timer = &msm_timer,
MACHINE_END
+
+MACHINE_START(MSM8X60_SIM, "QCT MSM8X60 SIMULATOR")
+ .map_io = msm8x60_map_io,
+ .init_irq = msm8x60_init_irq,
+ .init_machine = msm8x60_init,
+ .timer = &msm_timer,
+MACHINE_END
--
1.7.2.1
Sent by an employee of the Qualcomm Innovation Center, Inc.
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum.
From: Stepan Moskovchenko <[email protected]>
Register a driver for the MSM IOMMU devices and a driver
for the translation context devices. Set up the global
IOMMU registers and initialize the context banks.
Signed-off-by: Stepan Moskovchenko <[email protected]>
---
arch/arm/mach-msm/Makefile | 2 +-
arch/arm/mach-msm/iommu_dev.c | 364 +++++++++++++++++++++++++++++++++++++++++
2 files changed, 365 insertions(+), 1 deletions(-)
create mode 100644 arch/arm/mach-msm/iommu_dev.c
diff --git a/arch/arm/mach-msm/Makefile b/arch/arm/mach-msm/Makefile
index c984f4d..0d16232 100644
--- a/arch/arm/mach-msm/Makefile
+++ b/arch/arm/mach-msm/Makefile
@@ -12,7 +12,7 @@ obj-y += irq.o
endif
endif
-obj-$(CONFIG_ARCH_MSM8X60) += clock-dummy.o
+obj-$(CONFIG_ARCH_MSM8X60) += clock-dummy.o iommu.o iommu_dev.o
obj-$(CONFIG_MSM_PROC_COMM) += proc_comm.o clock-pcom.o vreg.o
obj-$(CONFIG_MSM_PROC_COMM) += clock.o gpio.o
obj-$(CONFIG_ARCH_QSD8X50) += sirc.o
diff --git a/arch/arm/mach-msm/iommu_dev.c b/arch/arm/mach-msm/iommu_dev.c
new file mode 100644
index 0000000..59ac7e5
--- /dev/null
+++ b/arch/arm/mach-msm/iommu_dev.c
@@ -0,0 +1,364 @@
+/* Copyright (c) 2010, Code Aurora Forum. All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 and
+ * only version 2 as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
+ * 02110-1301, USA.
+ */
+
+#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/io.h>
+#include <linux/clk.h>
+#include <linux/iommu.h>
+#include <linux/interrupt.h>
+#include <linux/err.h>
+#include <linux/slab.h>
+
+#include <mach/iommu_hw-8xxx.h>
+#include <mach/iommu.h>
+
+struct iommu_ctx_iter_data {
+ /* input */
+ const char *name;
+
+ /* output */
+ struct device *dev;
+};
+
+static struct platform_device *msm_iommu_root_dev;
+
+static int each_iommu_ctx(struct device *dev, void *data)
+{
+ struct iommu_ctx_iter_data *res = data;
+ struct msm_iommu_ctx_dev *c = dev->platform_data;
+
+ if (!res || !c || !c->name || !res->name)
+ return -EINVAL;
+
+ if (!strcmp(res->name, c->name)) {
+ res->dev = dev;
+ return 1;
+ }
+ return 0;
+}
+
+static int each_iommu(struct device *dev, void *data)
+{
+ return device_for_each_child(dev, data, each_iommu_ctx);
+}
+
+struct device *msm_iommu_get_ctx(const char *ctx_name)
+{
+ struct iommu_ctx_iter_data r;
+ int found;
+
+ if (!msm_iommu_root_dev) {
+ pr_err("No root IOMMU device.\n");
+ goto fail;
+ }
+
+ r.name = ctx_name;
+ found = device_for_each_child(&msm_iommu_root_dev->dev, &r, each_iommu);
+
+ if (!found) {
+ pr_err("Could not find context <%s>\n", ctx_name);
+ goto fail;
+ }
+
+ return r.dev;
+fail:
+ return NULL;
+}
+EXPORT_SYMBOL(msm_iommu_get_ctx);
+
+static void msm_iommu_reset(void __iomem *base)
+{
+ int ctx, ncb;
+
+ SET_RPUE(base, 0);
+ SET_RPUEIE(base, 0);
+ SET_ESRRESTORE(base, 0);
+ SET_TBE(base, 0);
+ SET_CR(base, 0);
+ SET_SPDMBE(base, 0);
+ SET_TESTBUSCR(base, 0);
+ SET_TLBRSW(base, 0);
+ SET_GLOBAL_TLBIALL(base, 0);
+ SET_RPU_ACR(base, 0);
+ SET_TLBLKCRWE(base, 1);
+ ncb = GET_NCB(base)+1;
+
+ for (ctx = 0; ctx < ncb; ctx++) {
+ SET_BPRCOSH(base, ctx, 0);
+ SET_BPRCISH(base, ctx, 0);
+ SET_BPRCNSH(base, ctx, 0);
+ SET_BPSHCFG(base, ctx, 0);
+ SET_BPMTCFG(base, ctx, 0);
+ SET_ACTLR(base, ctx, 0);
+ SET_SCTLR(base, ctx, 0);
+ SET_FSRRESTORE(base, ctx, 0);
+ SET_TTBR0(base, ctx, 0);
+ SET_TTBR1(base, ctx, 0);
+ SET_TTBCR(base, ctx, 0);
+ SET_BFBCR(base, ctx, 0);
+ SET_PAR(base, ctx, 0);
+ SET_FAR(base, ctx, 0);
+ SET_CTX_TLBIALL(base, ctx, 0);
+ SET_TLBFLPTER(base, ctx, 0);
+ SET_TLBSLPTER(base, ctx, 0);
+ SET_TLBLKCR(base, ctx, 0);
+ SET_PRRR(base, ctx, 0);
+ SET_NMRR(base, ctx, 0);
+ SET_CONTEXTIDR(base, ctx, 0);
+ }
+}
+
+static int msm_iommu_probe(struct platform_device *pdev)
+{
+ struct resource *r;
+ struct clk *iommu_clk;
+ struct msm_iommu_drvdata *drvdata;
+ struct msm_iommu_dev *iommu_dev = pdev->dev.platform_data;
+ void __iomem *regs_base;
+ resource_size_t len;
+ int ret = 0, ncb, nm2v, irq;
+
+ if (pdev->id != -1) {
+ drvdata = kzalloc(sizeof(*drvdata), GFP_KERNEL);
+
+ if (!drvdata) {
+ ret = -ENOMEM;
+ goto fail;
+ }
+
+ if (!iommu_dev) {
+ ret = -ENODEV;
+ goto fail;
+ }
+
+ if (iommu_dev->clk_rate != 0) {
+ iommu_clk = clk_get(&pdev->dev, "iommu_clk");
+
+ if (IS_ERR(iommu_clk)) {
+ ret = -ENODEV;
+ goto fail;
+ }
+
+ if (iommu_dev->clk_rate > 0)
+ clk_set_rate(iommu_clk, iommu_dev->clk_rate);
+
+ clk_enable(iommu_clk);
+ clk_put(iommu_clk);
+ }
+
+ r = platform_get_resource_byname(pdev, IORESOURCE_MEM,
+ "physbase");
+ if (!r) {
+ ret = -ENODEV;
+ goto fail;
+ }
+
+ len = r->end - r->start + 1;
+
+ r = request_mem_region(r->start, len, r->name);
+ if (!r) {
+ pr_err("Could not request memory region: "
+ "start=%p, len=%d\n", (void *) r->start, len);
+ ret = -EBUSY;
+ goto fail;
+ }
+
+ regs_base = ioremap(r->start, len);
+
+ if (!regs_base) {
+ pr_err("Could not ioremap: start=%p, len=%d\n",
+ (void *) r->start, len);
+ ret = -EBUSY;
+ goto fail;
+ }
+
+ irq = platform_get_irq_byname(pdev, "secure_irq");
+ if (irq < 0) {
+ ret = -ENODEV;
+ goto fail;
+ }
+
+ mb();
+
+ if (GET_IDR(regs_base) == 0) {
+ pr_err("Invalid IDR value detected\n");
+ ret = -ENODEV;
+ goto fail;
+ }
+
+ ret = request_irq(irq, msm_iommu_fault_handler, 0,
+ "msm_iommu_secure_irpt_handler", drvdata);
+ if (ret) {
+ pr_err("Request IRQ %d failed with ret=%d\n", irq, ret);
+ goto fail;
+ }
+
+ msm_iommu_reset(regs_base);
+ drvdata->base = regs_base;
+ drvdata->irq = irq;
+
+ nm2v = GET_NM2VCBMT((unsigned long) regs_base);
+ ncb = GET_NCB((unsigned long) regs_base);
+
+ pr_info("device %s mapped at %p, irq %d with %d ctx banks\n",
+ iommu_dev->name, regs_base, irq, ncb+1);
+
+ platform_set_drvdata(pdev, drvdata);
+ } else
+ msm_iommu_root_dev = pdev;
+
+ return 0;
+
+fail:
+ kfree(drvdata);
+ return ret;
+}
+
+static int msm_iommu_remove(struct platform_device *pdev)
+{
+ struct msm_iommu_drvdata *drv = NULL;
+
+ drv = platform_get_drvdata(pdev);
+ if (drv) {
+ memset(drv, 0, sizeof(struct msm_iommu_drvdata));
+ kfree(drv);
+ platform_set_drvdata(pdev, NULL);
+ }
+ return 0;
+}
+
+static int msm_iommu_ctx_probe(struct platform_device *pdev)
+{
+ struct msm_iommu_ctx_dev *c = pdev->dev.platform_data;
+ struct msm_iommu_drvdata *drvdata;
+ struct msm_iommu_ctx_drvdata *ctx_drvdata = NULL;
+ int i, ret = 0;
+ if (!c || !pdev->dev.parent) {
+ ret = -EINVAL;
+ goto fail;
+ }
+
+ drvdata = dev_get_drvdata(pdev->dev.parent);
+
+ if (!drvdata) {
+ ret = -ENODEV;
+ goto fail;
+ }
+
+ ctx_drvdata = kzalloc(sizeof(*ctx_drvdata), GFP_KERNEL);
+ if (!ctx_drvdata) {
+ ret = -ENOMEM;
+ goto fail;
+ }
+ ctx_drvdata->num = c->num;
+ ctx_drvdata->pdev = pdev;
+
+ INIT_LIST_HEAD(&ctx_drvdata->attached_elm);
+ platform_set_drvdata(pdev, ctx_drvdata);
+
+ /* Program the M2V tables for this context */
+ for (i = 0; i < MAX_NUM_MIDS; i++) {
+ int mid = c->mids[i];
+ if (mid == -1)
+ break;
+
+ SET_M2VCBR_N(drvdata->base, mid, 0);
+ SET_CBACR_N(drvdata->base, c->num, 0);
+
+ /* Set VMID = MID */
+ SET_VMID(drvdata->base, mid, mid);
+
+ /* Set the context number for that MID to this context */
+ SET_CBNDX(drvdata->base, mid, c->num);
+
+ /* Set MID associated with this context bank */
+ SET_CBVMID(drvdata->base, c->num, mid);
+
+ /* Set security bit override to be Non-secure */
+ SET_NSCFG(drvdata->base, mid, 3);
+ }
+
+ pr_info("context device %s with bank index %d\n", c->name, c->num);
+
+ return 0;
+fail:
+ kfree(ctx_drvdata);
+ return ret;
+}
+
+static int msm_iommu_ctx_remove(struct platform_device *pdev)
+{
+ struct msm_iommu_ctx_drvdata *drv = NULL;
+ drv = platform_get_drvdata(pdev);
+ if (drv) {
+ memset(drv, 0, sizeof(struct msm_iommu_ctx_drvdata));
+ kfree(drv);
+ platform_set_drvdata(pdev, NULL);
+ }
+ return 0;
+}
+
+static struct platform_driver msm_iommu_driver = {
+ .driver = {
+ .name = "msm_iommu",
+ },
+ .probe = msm_iommu_probe,
+ .remove = msm_iommu_remove,
+};
+
+static struct platform_driver msm_iommu_ctx_driver = {
+ .driver = {
+ .name = "msm_iommu_ctx",
+ },
+ .probe = msm_iommu_ctx_probe,
+ .remove = msm_iommu_ctx_remove,
+};
+
+static int msm_iommu_driver_init(void)
+{
+ int ret;
+ ret = platform_driver_register(&msm_iommu_driver);
+ if (ret != 0) {
+ pr_err("Failed to register IOMMU driver\n");
+ goto error;
+ }
+
+ ret = platform_driver_register(&msm_iommu_ctx_driver);
+ if (ret != 0) {
+ pr_err("Failed to register IOMMU context driver\n");
+ goto error;
+ }
+
+error:
+ return ret;
+}
+
+static void msm_iommu_driver_exit(void)
+{
+ platform_driver_unregister(&msm_iommu_ctx_driver);
+ platform_driver_unregister(&msm_iommu_driver);
+}
+
+subsys_initcall(msm_iommu_driver_init);
+module_exit(msm_iommu_driver_exit);
+
+MODULE_LICENSE("GPL v2");
+MODULE_AUTHOR("Stepan Moskovchenko <[email protected]>");
--
1.7.2.1
Sent by an employee of the Qualcomm Innovation Center, Inc.
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum.
From: Stepan Moskovchenko <[email protected]>
Add the platform data for the IOMMUs found on the Qualcomm
msm8x60 SoC.
Signed-off-by: Stepan Moskovchenko <[email protected]>
---
arch/arm/mach-msm/Makefile | 2 +-
arch/arm/mach-msm/devices-msm8x60-iommu.c | 883 +++++++++++++++++++++++
arch/arm/mach-msm/include/mach/msm_iomap-8x60.h | 32 +
3 files changed, 916 insertions(+), 1 deletions(-)
create mode 100644 arch/arm/mach-msm/devices-msm8x60-iommu.c
diff --git a/arch/arm/mach-msm/Makefile b/arch/arm/mach-msm/Makefile
index 0d16232..bf9eb38 100644
--- a/arch/arm/mach-msm/Makefile
+++ b/arch/arm/mach-msm/Makefile
@@ -12,7 +12,7 @@ obj-y += irq.o
endif
endif
-obj-$(CONFIG_ARCH_MSM8X60) += clock-dummy.o iommu.o iommu_dev.o
+obj-$(CONFIG_ARCH_MSM8X60) += clock-dummy.o iommu.o iommu_dev.o devices-msm8x60-iommu.o
obj-$(CONFIG_MSM_PROC_COMM) += proc_comm.o clock-pcom.o vreg.o
obj-$(CONFIG_MSM_PROC_COMM) += clock.o gpio.o
obj-$(CONFIG_ARCH_QSD8X50) += sirc.o
diff --git a/arch/arm/mach-msm/devices-msm8x60-iommu.c b/arch/arm/mach-msm/devices-msm8x60-iommu.c
new file mode 100644
index 0000000..89b9d44
--- /dev/null
+++ b/arch/arm/mach-msm/devices-msm8x60-iommu.c
@@ -0,0 +1,883 @@
+/* Copyright (c) 2010, Code Aurora Forum. All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 and
+ * only version 2 as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
+ * 02110-1301, USA.
+ */
+
+#include <linux/kernel.h>
+#include <linux/platform_device.h>
+#include <linux/bootmem.h>
+
+#include <mach/msm_iomap-8x60.h>
+#include <mach/irqs-8x60.h>
+#include <mach/iommu.h>
+
+static struct resource msm_iommu_jpegd_resources[] = {
+ {
+ .start = MSM_IOMMU_JPEGD_PHYS,
+ .end = MSM_IOMMU_JPEGD_PHYS + MSM_IOMMU_JPEGD_SIZE - 1,
+ .name = "physbase",
+ .flags = IORESOURCE_MEM,
+ },
+ {
+ .name = "nonsecure_irq",
+ .start = SMMU_JPEGD_CB_SC_NON_SECURE_IRQ,
+ .end = SMMU_JPEGD_CB_SC_NON_SECURE_IRQ,
+ .flags = IORESOURCE_IRQ,
+ },
+ {
+ .name = "secure_irq",
+ .start = SMMU_JPEGD_CB_SC_SECURE_IRQ,
+ .end = SMMU_JPEGD_CB_SC_SECURE_IRQ,
+ .flags = IORESOURCE_IRQ,
+ },
+};
+
+static struct resource msm_iommu_vpe_resources[] = {
+ {
+ .start = MSM_IOMMU_VPE_PHYS,
+ .end = MSM_IOMMU_VPE_PHYS + MSM_IOMMU_VPE_SIZE - 1,
+ .name = "physbase",
+ .flags = IORESOURCE_MEM,
+ },
+ {
+ .name = "nonsecure_irq",
+ .start = SMMU_VPE_CB_SC_NON_SECURE_IRQ,
+ .end = SMMU_VPE_CB_SC_NON_SECURE_IRQ,
+ .flags = IORESOURCE_IRQ,
+ },
+ {
+ .name = "secure_irq",
+ .start = SMMU_VPE_CB_SC_SECURE_IRQ,
+ .end = SMMU_VPE_CB_SC_SECURE_IRQ,
+ .flags = IORESOURCE_IRQ,
+ },
+};
+
+static struct resource msm_iommu_mdp0_resources[] = {
+ {
+ .start = MSM_IOMMU_MDP0_PHYS,
+ .end = MSM_IOMMU_MDP0_PHYS + MSM_IOMMU_MDP0_SIZE - 1,
+ .name = "physbase",
+ .flags = IORESOURCE_MEM,
+ },
+ {
+ .name = "nonsecure_irq",
+ .start = SMMU_MDP0_CB_SC_NON_SECURE_IRQ,
+ .end = SMMU_MDP0_CB_SC_NON_SECURE_IRQ,
+ .flags = IORESOURCE_IRQ,
+ },
+ {
+ .name = "secure_irq",
+ .start = SMMU_MDP0_CB_SC_SECURE_IRQ,
+ .end = SMMU_MDP0_CB_SC_SECURE_IRQ,
+ .flags = IORESOURCE_IRQ,
+ },
+};
+
+static struct resource msm_iommu_mdp1_resources[] = {
+ {
+ .start = MSM_IOMMU_MDP1_PHYS,
+ .end = MSM_IOMMU_MDP1_PHYS + MSM_IOMMU_MDP1_SIZE - 1,
+ .name = "physbase",
+ .flags = IORESOURCE_MEM,
+ },
+ {
+ .name = "nonsecure_irq",
+ .start = SMMU_MDP1_CB_SC_NON_SECURE_IRQ,
+ .end = SMMU_MDP1_CB_SC_NON_SECURE_IRQ,
+ .flags = IORESOURCE_IRQ,
+ },
+ {
+ .name = "secure_irq",
+ .start = SMMU_MDP1_CB_SC_SECURE_IRQ,
+ .end = SMMU_MDP1_CB_SC_SECURE_IRQ,
+ .flags = IORESOURCE_IRQ,
+ },
+};
+
+static struct resource msm_iommu_rot_resources[] = {
+ {
+ .start = MSM_IOMMU_ROT_PHYS,
+ .end = MSM_IOMMU_ROT_PHYS + MSM_IOMMU_ROT_SIZE - 1,
+ .name = "physbase",
+ .flags = IORESOURCE_MEM,
+ },
+ {
+ .name = "nonsecure_irq",
+ .start = SMMU_ROT_CB_SC_NON_SECURE_IRQ,
+ .end = SMMU_ROT_CB_SC_NON_SECURE_IRQ,
+ .flags = IORESOURCE_IRQ,
+ },
+ {
+ .name = "secure_irq",
+ .start = SMMU_ROT_CB_SC_SECURE_IRQ,
+ .end = SMMU_ROT_CB_SC_SECURE_IRQ,
+ .flags = IORESOURCE_IRQ,
+ },
+};
+
+static struct resource msm_iommu_ijpeg_resources[] = {
+ {
+ .start = MSM_IOMMU_IJPEG_PHYS,
+ .end = MSM_IOMMU_IJPEG_PHYS + MSM_IOMMU_IJPEG_SIZE - 1,
+ .name = "physbase",
+ .flags = IORESOURCE_MEM,
+ },
+ {
+ .name = "nonsecure_irq",
+ .start = SMMU_IJPEG_CB_SC_NON_SECURE_IRQ,
+ .end = SMMU_IJPEG_CB_SC_NON_SECURE_IRQ,
+ .flags = IORESOURCE_IRQ,
+ },
+ {
+ .name = "secure_irq",
+ .start = SMMU_IJPEG_CB_SC_SECURE_IRQ,
+ .end = SMMU_IJPEG_CB_SC_SECURE_IRQ,
+ .flags = IORESOURCE_IRQ,
+ },
+};
+
+static struct resource msm_iommu_vfe_resources[] = {
+ {
+ .start = MSM_IOMMU_VFE_PHYS,
+ .end = MSM_IOMMU_VFE_PHYS + MSM_IOMMU_VFE_SIZE - 1,
+ .name = "physbase",
+ .flags = IORESOURCE_MEM,
+ },
+ {
+ .name = "nonsecure_irq",
+ .start = SMMU_VFE_CB_SC_NON_SECURE_IRQ,
+ .end = SMMU_VFE_CB_SC_NON_SECURE_IRQ,
+ .flags = IORESOURCE_IRQ,
+ },
+ {
+ .name = "secure_irq",
+ .start = SMMU_VFE_CB_SC_SECURE_IRQ,
+ .end = SMMU_VFE_CB_SC_SECURE_IRQ,
+ .flags = IORESOURCE_IRQ,
+ },
+};
+
+static struct resource msm_iommu_vcodec_a_resources[] = {
+ {
+ .start = MSM_IOMMU_VCODEC_A_PHYS,
+ .end = MSM_IOMMU_VCODEC_A_PHYS + MSM_IOMMU_VCODEC_A_SIZE - 1,
+ .name = "physbase",
+ .flags = IORESOURCE_MEM,
+ },
+ {
+ .name = "nonsecure_irq",
+ .start = SMMU_VCODEC_A_CB_SC_NON_SECURE_IRQ,
+ .end = SMMU_VCODEC_A_CB_SC_NON_SECURE_IRQ,
+ .flags = IORESOURCE_IRQ,
+ },
+ {
+ .name = "secure_irq",
+ .start = SMMU_VCODEC_A_CB_SC_SECURE_IRQ,
+ .end = SMMU_VCODEC_A_CB_SC_SECURE_IRQ,
+ .flags = IORESOURCE_IRQ,
+ },
+};
+
+static struct resource msm_iommu_vcodec_b_resources[] = {
+ {
+ .start = MSM_IOMMU_VCODEC_B_PHYS,
+ .end = MSM_IOMMU_VCODEC_B_PHYS + MSM_IOMMU_VCODEC_B_SIZE - 1,
+ .name = "physbase",
+ .flags = IORESOURCE_MEM,
+ },
+ {
+ .name = "nonsecure_irq",
+ .start = SMMU_VCODEC_B_CB_SC_NON_SECURE_IRQ,
+ .end = SMMU_VCODEC_B_CB_SC_NON_SECURE_IRQ,
+ .flags = IORESOURCE_IRQ,
+ },
+ {
+ .name = "secure_irq",
+ .start = SMMU_VCODEC_B_CB_SC_SECURE_IRQ,
+ .end = SMMU_VCODEC_B_CB_SC_SECURE_IRQ,
+ .flags = IORESOURCE_IRQ,
+ },
+};
+
+static struct resource msm_iommu_gfx3d_resources[] = {
+ {
+ .start = MSM_IOMMU_GFX3D_PHYS,
+ .end = MSM_IOMMU_GFX3D_PHYS + MSM_IOMMU_GFX3D_SIZE - 1,
+ .name = "physbase",
+ .flags = IORESOURCE_MEM,
+ },
+ {
+ .name = "nonsecure_irq",
+ .start = SMMU_GFX3D_CB_SC_NON_SECURE_IRQ,
+ .end = SMMU_GFX3D_CB_SC_NON_SECURE_IRQ,
+ .flags = IORESOURCE_IRQ,
+ },
+ {
+ .name = "secure_irq",
+ .start = SMMU_GFX3D_CB_SC_SECURE_IRQ,
+ .end = SMMU_GFX3D_CB_SC_SECURE_IRQ,
+ .flags = IORESOURCE_IRQ,
+ },
+};
+
+static struct resource msm_iommu_gfx2d0_resources[] = {
+ {
+ .start = MSM_IOMMU_GFX2D0_PHYS,
+ .end = MSM_IOMMU_GFX2D0_PHYS + MSM_IOMMU_GFX2D0_SIZE - 1,
+ .name = "physbase",
+ .flags = IORESOURCE_MEM,
+ },
+ {
+ .name = "nonsecure_irq",
+ .start = SMMU_GFX2D0_CB_SC_NON_SECURE_IRQ,
+ .end = SMMU_GFX2D0_CB_SC_NON_SECURE_IRQ,
+ .flags = IORESOURCE_IRQ,
+ },
+ {
+ .name = "secure_irq",
+ .start = SMMU_GFX2D0_CB_SC_SECURE_IRQ,
+ .end = SMMU_GFX2D0_CB_SC_SECURE_IRQ,
+ .flags = IORESOURCE_IRQ,
+ },
+};
+
+static struct platform_device msm_root_iommu_dev = {
+ .name = "msm_iommu",
+ .id = -1,
+};
+
+static struct msm_iommu_dev jpegd_smmu = {
+ .name = "jpegd",
+ .clk_rate = -1
+};
+
+static struct msm_iommu_dev vpe_smmu = {
+ .name = "vpe"
+};
+
+static struct msm_iommu_dev mdp0_smmu = {
+ .name = "mdp0"
+};
+
+static struct msm_iommu_dev mdp1_smmu = {
+ .name = "mdp1"
+};
+
+static struct msm_iommu_dev rot_smmu = {
+ .name = "rot"
+};
+
+static struct msm_iommu_dev ijpeg_smmu = {
+ .name = "ijpeg"
+};
+
+static struct msm_iommu_dev vfe_smmu = {
+ .name = "vfe",
+ .clk_rate = -1
+};
+
+static struct msm_iommu_dev vcodec_a_smmu = {
+ .name = "vcodec_a"
+};
+
+static struct msm_iommu_dev vcodec_b_smmu = {
+ .name = "vcodec_b"
+};
+
+static struct msm_iommu_dev gfx3d_smmu = {
+ .name = "gfx3d",
+ .clk_rate = 27000000
+};
+
+static struct msm_iommu_dev gfx2d0_smmu = {
+ .name = "gfx2d0",
+ .clk_rate = 27000000
+};
+
+static struct platform_device msm_device_smmu_jpegd = {
+ .name = "msm_iommu",
+ .id = 0,
+ .dev = {
+ .parent = &msm_root_iommu_dev.dev,
+ },
+ .num_resources = ARRAY_SIZE(msm_iommu_jpegd_resources),
+ .resource = msm_iommu_jpegd_resources,
+};
+
+static struct platform_device msm_device_smmu_vpe = {
+ .name = "msm_iommu",
+ .id = 1,
+ .dev = {
+ .parent = &msm_root_iommu_dev.dev,
+ },
+ .num_resources = ARRAY_SIZE(msm_iommu_vpe_resources),
+ .resource = msm_iommu_vpe_resources,
+};
+
+static struct platform_device msm_device_smmu_mdp0 = {
+ .name = "msm_iommu",
+ .id = 2,
+ .dev = {
+ .parent = &msm_root_iommu_dev.dev,
+ },
+ .num_resources = ARRAY_SIZE(msm_iommu_mdp0_resources),
+ .resource = msm_iommu_mdp0_resources,
+};
+
+static struct platform_device msm_device_smmu_mdp1 = {
+ .name = "msm_iommu",
+ .id = 3,
+ .dev = {
+ .parent = &msm_root_iommu_dev.dev,
+ },
+ .num_resources = ARRAY_SIZE(msm_iommu_mdp1_resources),
+ .resource = msm_iommu_mdp1_resources,
+};
+
+static struct platform_device msm_device_smmu_rot = {
+ .name = "msm_iommu",
+ .id = 4,
+ .dev = {
+ .parent = &msm_root_iommu_dev.dev,
+ },
+ .num_resources = ARRAY_SIZE(msm_iommu_rot_resources),
+ .resource = msm_iommu_rot_resources,
+};
+
+static struct platform_device msm_device_smmu_ijpeg = {
+ .name = "msm_iommu",
+ .id = 5,
+ .dev = {
+ .parent = &msm_root_iommu_dev.dev,
+ },
+ .num_resources = ARRAY_SIZE(msm_iommu_ijpeg_resources),
+ .resource = msm_iommu_ijpeg_resources,
+};
+
+static struct platform_device msm_device_smmu_vfe = {
+ .name = "msm_iommu",
+ .id = 6,
+ .dev = {
+ .parent = &msm_root_iommu_dev.dev,
+ },
+ .num_resources = ARRAY_SIZE(msm_iommu_vfe_resources),
+ .resource = msm_iommu_vfe_resources,
+};
+
+static struct platform_device msm_device_smmu_vcodec_a = {
+ .name = "msm_iommu",
+ .id = 7,
+ .dev = {
+ .parent = &msm_root_iommu_dev.dev,
+ },
+ .num_resources = ARRAY_SIZE(msm_iommu_vcodec_a_resources),
+ .resource = msm_iommu_vcodec_a_resources,
+};
+
+static struct platform_device msm_device_smmu_vcodec_b = {
+ .name = "msm_iommu",
+ .id = 8,
+ .dev = {
+ .parent = &msm_root_iommu_dev.dev,
+ },
+ .num_resources = ARRAY_SIZE(msm_iommu_vcodec_b_resources),
+ .resource = msm_iommu_vcodec_b_resources,
+};
+
+static struct platform_device msm_device_smmu_gfx3d = {
+ .name = "msm_iommu",
+ .id = 9,
+ .dev = {
+ .parent = &msm_root_iommu_dev.dev,
+ },
+ .num_resources = ARRAY_SIZE(msm_iommu_gfx3d_resources),
+ .resource = msm_iommu_gfx3d_resources,
+};
+
+static struct platform_device msm_device_smmu_gfx2d0 = {
+ .name = "msm_iommu",
+ .id = 10,
+ .dev = {
+ .parent = &msm_root_iommu_dev.dev,
+ },
+ .num_resources = ARRAY_SIZE(msm_iommu_gfx2d0_resources),
+ .resource = msm_iommu_gfx2d0_resources,
+};
+
+static struct msm_iommu_ctx_dev jpegd_src_ctx = {
+ .name = "jpegd_src",
+ .num = 0,
+ .mids = {0, -1}
+};
+
+static struct msm_iommu_ctx_dev jpegd_dst_ctx = {
+ .name = "jpegd_dst",
+ .num = 1,
+ .mids = {1, -1}
+};
+
+static struct msm_iommu_ctx_dev vpe_src_ctx = {
+ .name = "vpe_src",
+ .num = 0,
+ .mids = {0, -1}
+};
+
+static struct msm_iommu_ctx_dev vpe_dst_ctx = {
+ .name = "vpe_dst",
+ .num = 1,
+ .mids = {1, -1}
+};
+
+static struct msm_iommu_ctx_dev mdp_vg1_ctx = {
+ .name = "mdp_vg1",
+ .num = 0,
+ .mids = {0, 2, -1}
+};
+
+static struct msm_iommu_ctx_dev mdp_rgb1_ctx = {
+ .name = "mdp_rgb1",
+ .num = 1,
+ .mids = {1, 3, 4, 5, 6, 7, 8, 9, 10, -1}
+};
+
+static struct msm_iommu_ctx_dev mdp_vg2_ctx = {
+ .name = "mdp_vg2",
+ .num = 0,
+ .mids = {0, 2, -1}
+};
+
+static struct msm_iommu_ctx_dev mdp_rgb2_ctx = {
+ .name = "mdp_rgb2",
+ .num = 1,
+ .mids = {1, 3, 4, 5, 6, 7, 8, 9, 10, -1}
+};
+
+static struct msm_iommu_ctx_dev rot_src_ctx = {
+ .name = "rot_src",
+ .num = 0,
+ .mids = {0, -1}
+};
+
+static struct msm_iommu_ctx_dev rot_dst_ctx = {
+ .name = "rot_dst",
+ .num = 1,
+ .mids = {1, -1}
+};
+
+static struct msm_iommu_ctx_dev ijpeg_src_ctx = {
+ .name = "ijpeg_src",
+ .num = 0,
+ .mids = {0, -1}
+};
+
+static struct msm_iommu_ctx_dev ijpeg_dst_ctx = {
+ .name = "ijpeg_dst",
+ .num = 1,
+ .mids = {1, -1}
+};
+
+static struct msm_iommu_ctx_dev vfe_imgwr_ctx = {
+ .name = "vfe_imgwr",
+ .num = 0,
+ .mids = {2, 3, 4, 5, 6, 7, 8, -1}
+};
+
+static struct msm_iommu_ctx_dev vfe_misc_ctx = {
+ .name = "vfe_misc",
+ .num = 1,
+ .mids = {0, 1, 9, -1}
+};
+
+static struct msm_iommu_ctx_dev vcodec_a_stream_ctx = {
+ .name = "vcodec_a_stream",
+ .num = 0,
+ .mids = {2, 5, -1}
+};
+
+static struct msm_iommu_ctx_dev vcodec_a_mm1_ctx = {
+ .name = "vcodec_a_mm1",
+ .num = 1,
+ .mids = {0, 1, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, -1}
+};
+
+static struct msm_iommu_ctx_dev vcodec_b_mm2_ctx = {
+ .name = "vcodec_b_mm2",
+ .num = 0,
+ .mids = {0, 1, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, -1}
+};
+
+static struct msm_iommu_ctx_dev gfx3d_rbpa_ctx = {
+ .name = "gfx3d_rbpa",
+ .num = 0,
+ .mids = {-1}
+};
+
+static struct msm_iommu_ctx_dev gfx3d_cpvgttc_ctx = {
+ .name = "gfx3d_cpvgttc",
+ .num = 1,
+ .mids = {0, 1, 2, 3, 4, 5, 6, 7, -1}
+};
+
+static struct msm_iommu_ctx_dev gfx3d_smmu_ctx = {
+ .name = "gfx3d_smmu",
+ .num = 2,
+ .mids = {8, 9, 10, 11, 12, -1}
+};
+
+static struct msm_iommu_ctx_dev gfx2d0_pixv1_ctx = {
+ .name = "gfx2d0_pixv1_smmu",
+ .num = 0,
+ .mids = {0, 3, 4, -1}
+};
+
+static struct msm_iommu_ctx_dev gfx2d0_texv3_ctx = {
+ .name = "gfx2d0_texv3_smmu",
+ .num = 1,
+ .mids = {1, 6, 7, -1}
+};
+
+static struct platform_device msm_device_jpegd_src_ctx = {
+ .name = "msm_iommu_ctx",
+ .id = 0,
+ .dev = {
+ .parent = &msm_device_smmu_jpegd.dev,
+ },
+};
+
+static struct platform_device msm_device_jpegd_dst_ctx = {
+ .name = "msm_iommu_ctx",
+ .id = 1,
+ .dev = {
+ .parent = &msm_device_smmu_jpegd.dev,
+ },
+};
+
+static struct platform_device msm_device_vpe_src_ctx = {
+ .name = "msm_iommu_ctx",
+ .id = 2,
+ .dev = {
+ .parent = &msm_device_smmu_vpe.dev,
+ },
+};
+
+static struct platform_device msm_device_vpe_dst_ctx = {
+ .name = "msm_iommu_ctx",
+ .id = 3,
+ .dev = {
+ .parent = &msm_device_smmu_vpe.dev,
+ },
+};
+
+static struct platform_device msm_device_mdp_vg1_ctx = {
+ .name = "msm_iommu_ctx",
+ .id = 4,
+ .dev = {
+ .parent = &msm_device_smmu_mdp0.dev,
+ },
+};
+
+static struct platform_device msm_device_mdp_rgb1_ctx = {
+ .name = "msm_iommu_ctx",
+ .id = 5,
+ .dev = {
+ .parent = &msm_device_smmu_mdp0.dev,
+ },
+};
+
+static struct platform_device msm_device_mdp_vg2_ctx = {
+ .name = "msm_iommu_ctx",
+ .id = 6,
+ .dev = {
+ .parent = &msm_device_smmu_mdp1.dev,
+ },
+};
+
+static struct platform_device msm_device_mdp_rgb2_ctx = {
+ .name = "msm_iommu_ctx",
+ .id = 7,
+ .dev = {
+ .parent = &msm_device_smmu_mdp1.dev,
+ },
+};
+
+static struct platform_device msm_device_rot_src_ctx = {
+ .name = "msm_iommu_ctx",
+ .id = 8,
+ .dev = {
+ .parent = &msm_device_smmu_rot.dev,
+ },
+};
+
+static struct platform_device msm_device_rot_dst_ctx = {
+ .name = "msm_iommu_ctx",
+ .id = 9,
+ .dev = {
+ .parent = &msm_device_smmu_rot.dev,
+ },
+};
+
+static struct platform_device msm_device_ijpeg_src_ctx = {
+ .name = "msm_iommu_ctx",
+ .id = 10,
+ .dev = {
+ .parent = &msm_device_smmu_ijpeg.dev,
+ },
+};
+
+static struct platform_device msm_device_ijpeg_dst_ctx = {
+ .name = "msm_iommu_ctx",
+ .id = 11,
+ .dev = {
+ .parent = &msm_device_smmu_ijpeg.dev,
+ },
+};
+
+static struct platform_device msm_device_vfe_imgwr_ctx = {
+ .name = "msm_iommu_ctx",
+ .id = 12,
+ .dev = {
+ .parent = &msm_device_smmu_vfe.dev,
+ },
+};
+
+static struct platform_device msm_device_vfe_misc_ctx = {
+ .name = "msm_iommu_ctx",
+ .id = 13,
+ .dev = {
+ .parent = &msm_device_smmu_vfe.dev,
+ },
+};
+
+static struct platform_device msm_device_vcodec_a_stream_ctx = {
+ .name = "msm_iommu_ctx",
+ .id = 14,
+ .dev = {
+ .parent = &msm_device_smmu_vcodec_a.dev,
+ },
+};
+
+static struct platform_device msm_device_vcodec_a_mm1_ctx = {
+ .name = "msm_iommu_ctx",
+ .id = 15,
+ .dev = {
+ .parent = &msm_device_smmu_vcodec_a.dev,
+ },
+};
+
+static struct platform_device msm_device_vcodec_b_mm2_ctx = {
+ .name = "msm_iommu_ctx",
+ .id = 16,
+ .dev = {
+ .parent = &msm_device_smmu_vcodec_b.dev,
+ },
+};
+
+static struct platform_device msm_device_gfx3d_rbpa_ctx = {
+ .name = "msm_iommu_ctx",
+ .id = 17,
+ .dev = {
+ .parent = &msm_device_smmu_gfx3d.dev,
+ },
+};
+
+static struct platform_device msm_device_gfx3d_cpvgttc_ctx = {
+ .name = "msm_iommu_ctx",
+ .id = 18,
+ .dev = {
+ .parent = &msm_device_smmu_gfx3d.dev,
+ },
+};
+
+static struct platform_device msm_device_gfx3d_smmu_ctx = {
+ .name = "msm_iommu_ctx",
+ .id = 19,
+ .dev = {
+ .parent = &msm_device_smmu_gfx3d.dev,
+ },
+};
+
+static struct platform_device msm_device_gfx2d0_pixv1_ctx = {
+ .name = "msm_iommu_ctx",
+ .id = 20,
+ .dev = {
+ .parent = &msm_device_smmu_gfx2d0.dev,
+ },
+};
+
+static struct platform_device msm_device_gfx2d0_texv3_ctx = {
+ .name = "msm_iommu_ctx",
+ .id = 21,
+ .dev = {
+ .parent = &msm_device_smmu_gfx2d0.dev,
+ },
+};
+
+static struct platform_device *msm_iommu_devs[] = {
+ &msm_device_smmu_jpegd,
+ &msm_device_smmu_vpe,
+ &msm_device_smmu_mdp0,
+ &msm_device_smmu_mdp1,
+ &msm_device_smmu_rot,
+ &msm_device_smmu_ijpeg,
+ &msm_device_smmu_vfe,
+ &msm_device_smmu_vcodec_a,
+ &msm_device_smmu_vcodec_b,
+ &msm_device_smmu_gfx3d,
+ &msm_device_smmu_gfx2d0,
+};
+
+static struct msm_iommu_dev *msm_iommu_data[] = {
+ &jpegd_smmu,
+ &vpe_smmu,
+ &mdp0_smmu,
+ &mdp1_smmu,
+ &rot_smmu,
+ &ijpeg_smmu,
+ &vfe_smmu,
+ &vcodec_a_smmu,
+ &vcodec_b_smmu,
+ &gfx3d_smmu,
+ &gfx2d0_smmu,
+};
+
+static struct platform_device *msm_iommu_ctx_devs[] = {
+ &msm_device_jpegd_src_ctx,
+ &msm_device_jpegd_dst_ctx,
+ &msm_device_vpe_src_ctx,
+ &msm_device_vpe_dst_ctx,
+ &msm_device_mdp_vg1_ctx,
+ &msm_device_mdp_rgb1_ctx,
+ &msm_device_mdp_vg2_ctx,
+ &msm_device_mdp_rgb2_ctx,
+ &msm_device_rot_src_ctx,
+ &msm_device_rot_dst_ctx,
+ &msm_device_ijpeg_src_ctx,
+ &msm_device_ijpeg_dst_ctx,
+ &msm_device_vfe_imgwr_ctx,
+ &msm_device_vfe_misc_ctx,
+ &msm_device_vcodec_a_stream_ctx,
+ &msm_device_vcodec_a_mm1_ctx,
+ &msm_device_vcodec_b_mm2_ctx,
+ &msm_device_gfx3d_rbpa_ctx,
+ &msm_device_gfx3d_cpvgttc_ctx,
+ &msm_device_gfx3d_smmu_ctx,
+ &msm_device_gfx2d0_pixv1_ctx,
+ &msm_device_gfx2d0_texv3_ctx,
+};
+
+static struct msm_iommu_ctx_dev *msm_iommu_ctx_data[] = {
+ &jpegd_src_ctx,
+ &jpegd_dst_ctx,
+ &vpe_src_ctx,
+ &vpe_dst_ctx,
+ &mdp_vg1_ctx,
+ &mdp_rgb1_ctx,
+ &mdp_vg2_ctx,
+ &mdp_rgb2_ctx,
+ &rot_src_ctx,
+ &rot_dst_ctx,
+ &ijpeg_src_ctx,
+ &ijpeg_dst_ctx,
+ &vfe_imgwr_ctx,
+ &vfe_misc_ctx,
+ &vcodec_a_stream_ctx,
+ &vcodec_a_mm1_ctx,
+ &vcodec_b_mm2_ctx,
+ &gfx3d_rbpa_ctx,
+ &gfx3d_cpvgttc_ctx,
+ &gfx3d_smmu_ctx,
+ &gfx2d0_pixv1_ctx,
+ &gfx2d0_texv3_ctx,
+};
+
+static int msm8x60_iommu_init(void)
+{
+ int ret, i;
+
+ ret = platform_device_register(&msm_root_iommu_dev);
+ if (ret != 0) {
+ pr_err("Failed to register root IOMMU device!\n");
+ goto failure;
+ }
+
+ for (i = 0; i < ARRAY_SIZE(msm_iommu_devs); i++) {
+ ret = platform_device_add_data(msm_iommu_devs[i],
+ msm_iommu_data[i],
+ sizeof(struct msm_iommu_dev));
+ if (ret != 0) {
+ pr_err("platform_device_add_data failed, "
+ "i = %d\n", i);
+ goto failure_unwind;
+ }
+
+ ret = platform_device_register(msm_iommu_devs[i]);
+
+ if (ret != 0) {
+ pr_err("platform_device_register smmu failed, "
+ "i = %d\n", i);
+ goto failure_unwind;
+ }
+ }
+
+ for (i = 0; i < ARRAY_SIZE(msm_iommu_ctx_devs); i++) {
+ ret = platform_device_add_data(msm_iommu_ctx_devs[i],
+ msm_iommu_ctx_data[i],
+ sizeof(*msm_iommu_ctx_devs[i]));
+ if (ret != 0) {
+ pr_err("platform_device_add_data smmu failed, "
+ "i = %d\n", i);
+ goto failure_unwind2;
+ }
+
+ ret = platform_device_register(msm_iommu_ctx_devs[i]);
+ if (ret != 0) {
+ pr_err("platform_device_register ctx failed, "
+ "i = %d\n", i);
+ goto failure_unwind2;
+ }
+ }
+ return 0;
+
+failure_unwind2:
+ while (--i >= 0)
+ platform_device_unregister(msm_iommu_ctx_devs[i]);
+failure_unwind:
+ while (--i >= 0)
+ platform_device_unregister(msm_iommu_devs[i]);
+
+ platform_device_unregister(&msm_root_iommu_dev);
+failure:
+ return ret;
+}
+
+static void msm8x60_iommu_exit(void)
+{
+ int i;
+
+ for (i = 0; i < ARRAY_SIZE(msm_iommu_ctx_devs); i++)
+ platform_device_unregister(msm_iommu_ctx_devs[i]);
+
+ for (i = 0; i < ARRAY_SIZE(msm_iommu_devs); ++i)
+ platform_device_unregister(msm_iommu_devs[i]);
+
+ platform_device_unregister(&msm_root_iommu_dev);
+}
+
+subsys_initcall(msm8x60_iommu_init);
+module_exit(msm8x60_iommu_exit);
+
+MODULE_LICENSE("GPL v2");
+MODULE_AUTHOR("Stepan Moskovchenko <[email protected]>");
diff --git a/arch/arm/mach-msm/include/mach/msm_iomap-8x60.h b/arch/arm/mach-msm/include/mach/msm_iomap-8x60.h
index 1f15bbb..c0f6ad3 100644
--- a/arch/arm/mach-msm/include/mach/msm_iomap-8x60.h
+++ b/arch/arm/mach-msm/include/mach/msm_iomap-8x60.h
@@ -65,5 +65,37 @@
#define MSM_SHARED_RAM_BASE IOMEM(0xF0200000)
#define MSM_SHARED_RAM_SIZE SZ_1M
+#define MSM_IOMMU_JPEGD_PHYS 0x07300000
+#define MSM_IOMMU_JPEGD_SIZE SZ_1M
+
+#define MSM_IOMMU_VPE_PHYS 0x07400000
+#define MSM_IOMMU_VPE_SIZE SZ_1M
+
+#define MSM_IOMMU_MDP0_PHYS 0x07500000
+#define MSM_IOMMU_MDP0_SIZE SZ_1M
+
+#define MSM_IOMMU_MDP1_PHYS 0x07600000
+#define MSM_IOMMU_MDP1_SIZE SZ_1M
+
+#define MSM_IOMMU_ROT_PHYS 0x07700000
+#define MSM_IOMMU_ROT_SIZE SZ_1M
+
+#define MSM_IOMMU_IJPEG_PHYS 0x07800000
+#define MSM_IOMMU_IJPEG_SIZE SZ_1M
+
+#define MSM_IOMMU_VFE_PHYS 0x07900000
+#define MSM_IOMMU_VFE_SIZE SZ_1M
+
+#define MSM_IOMMU_VCODEC_A_PHYS 0x07A00000
+#define MSM_IOMMU_VCODEC_A_SIZE SZ_1M
+
+#define MSM_IOMMU_VCODEC_B_PHYS 0x07B00000
+#define MSM_IOMMU_VCODEC_B_SIZE SZ_1M
+
+#define MSM_IOMMU_GFX3D_PHYS 0x07C00000
+#define MSM_IOMMU_GFX3D_SIZE SZ_1M
+
+#define MSM_IOMMU_GFX2D0_PHYS 0x07D00000
+#define MSM_IOMMU_GFX2D0_SIZE SZ_1M
#endif
--
1.7.2.1
Sent by an employee of the Qualcomm Innovation Center, Inc.
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum.
From: Stepan Moskovchenko <[email protected]>
Add support for the IOMMUs found on the upcoming Qualcomm
MSM8x60 chips. These IOMMUs allow virtualization of the
address space used by most of the multimedia cores on these
chips.
Signed-off-by: Stepan Moskovchenko <[email protected]>
---
arch/arm/mach-msm/include/mach/iommu.h | 103 ++
arch/arm/mach-msm/include/mach/iommu_hw-8xxx.h | 1871 ++++++++++++++++++++++++
arch/arm/mach-msm/iommu.c | 597 ++++++++
3 files changed, 2571 insertions(+), 0 deletions(-)
create mode 100644 arch/arm/mach-msm/include/mach/iommu.h
create mode 100644 arch/arm/mach-msm/include/mach/iommu_hw-8xxx.h
create mode 100644 arch/arm/mach-msm/iommu.c
diff --git a/arch/arm/mach-msm/include/mach/iommu.h b/arch/arm/mach-msm/include/mach/iommu.h
new file mode 100644
index 0000000..218ef57
--- /dev/null
+++ b/arch/arm/mach-msm/include/mach/iommu.h
@@ -0,0 +1,103 @@
+/* Copyright (c) 2010, Code Aurora Forum. All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 and
+ * only version 2 as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
+ * 02110-1301, USA.
+ */
+
+#ifndef MSM_IOMMU_H
+#define MSM_IOMMU_H
+
+#include <linux/interrupt.h>
+
+/* Maximum number of Machine IDs that we are allowing to be mapped to the same
+ * context bank. The number of MIDs mapped to the same CB does not affect
+ * performance, but there is a practical limit on how many distinct MIDs may
+ * be present. These mappings are typically determined at design time and are
+ * not expected to change at run time.
+ */
+#define MAX_NUM_MIDS 16
+
+/**
+ * struct msm_iommu_dev - a single IOMMU hardware instance
+ * name Human-readable name given to this IOMMU HW instance
+ * clk_rate Rate to set for this IOMMU's clock, if applicable to this
+ * particular IOMMU. 0 means don't set a rate.
+ * -1 means it is an AXI clock with no valid rate
+ *
+ */
+struct msm_iommu_dev {
+ const char *name;
+ int clk_rate;
+};
+
+/**
+ * struct msm_iommu_ctx_dev - an IOMMU context bank instance
+ * name Human-readable name given to this context bank
+ * num Index of this context bank within the hardware
+ * mids List of Machine IDs that are to be mapped into this context
+ * bank, terminated by -1. The MID is a set of signals on the
+ * AXI bus that identifies the function associated with a specific
+ * memory request. (See ARM spec).
+ */
+struct msm_iommu_ctx_dev {
+ const char *name;
+ int num;
+ int mids[MAX_NUM_MIDS];
+};
+
+
+/**
+ * struct msm_iommu_drvdata - A single IOMMU hardware instance
+ * @base: IOMMU config port base address (VA)
+ * @irq: Interrupt number
+ *
+ * A msm_iommu_drvdata holds the global driver data about a single piece
+ * of an IOMMU hardware instance.
+ */
+struct msm_iommu_drvdata {
+ void __iomem *base;
+ int irq;
+};
+
+/**
+ * struct msm_iommu_ctx_drvdata - an IOMMU context bank instance
+ * @num: Hardware context number of this context
+ * @pdev: Platform device associated wit this HW instance
+ * @attached_elm: List element for domains to track which devices are
+ * attached to them
+ *
+ * A msm_iommu_ctx_drvdata holds the driver data for a single context bank
+ * within each IOMMU hardware instance
+ */
+struct msm_iommu_ctx_drvdata {
+ int num;
+ struct platform_device *pdev;
+ struct list_head attached_elm;
+};
+
+/*
+ * Look up an IOMMU context device by its context name. NULL if none found.
+ * Useful for testing and drivers that do not yet fully have IOMMU stuff in
+ * their platform devices.
+ */
+struct device *msm_iommu_get_ctx(const char *ctx_name);
+
+/*
+ * Interrupt handler for the IOMMU context fault interrupt. Hooking the
+ * interrupt is not supported in the API yet, but this will print an error
+ * message and dump useful IOMMU registers.
+ */
+irqreturn_t msm_iommu_fault_handler(int irq, void *dev_id);
+
+#endif
diff --git a/arch/arm/mach-msm/include/mach/iommu_hw-8xxx.h b/arch/arm/mach-msm/include/mach/iommu_hw-8xxx.h
new file mode 100644
index 0000000..5565aee
--- /dev/null
+++ b/arch/arm/mach-msm/include/mach/iommu_hw-8xxx.h
@@ -0,0 +1,1871 @@
+/* Copyright (c) 2010, Code Aurora Forum. All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 and
+ * only version 2 as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
+ * 02110-1301, USA.
+ */
+
+#ifndef __ARCH_ARM_MACH_MSM_IOMMU_HW_8XXX_H
+#define __ARCH_ARM_MACH_MSM_IOMMU_HW_8XXX_H
+
+#define CTX_SHIFT 12
+
+#define GET_GLOBAL_REG(reg, base) (readl((base) + (reg)))
+#define GET_CTX_REG(reg, base, ctx) \
+ (readl((base) + (reg) + ((ctx) << CTX_SHIFT)))
+
+#define SET_GLOBAL_REG(reg, base, val) writel((val), ((base) + (reg)))
+
+#define SET_CTX_REG(reg, base, ctx, val) \
+ writel((val), ((base) + (reg) + ((ctx) << CTX_SHIFT)))
+
+/* Wrappers for numbered registers */
+#define SET_GLOBAL_REG_N(b, n, r, v) SET_GLOBAL_REG(b, ((r) + (n << 2)), (v))
+#define GET_GLOBAL_REG_N(b, n, r) GET_GLOBAL_REG(b, ((r) + (n << 2)))
+
+/* Field wrappers */
+#define GET_GLOBAL_FIELD(b, r, F) GET_FIELD(((b) + (r)), F##_MASK, F##_SHIFT)
+#define GET_CONTEXT_FIELD(b, c, r, F) \
+ GET_FIELD(((b) + (r) + ((c) << CTX_SHIFT)), F##_MASK, F##_SHIFT)
+
+#define SET_GLOBAL_FIELD(b, r, F, v) \
+ SET_FIELD(((b) + (r)), F##_MASK, F##_SHIFT, (v))
+#define SET_CONTEXT_FIELD(b, c, r, F, v) \
+ SET_FIELD(((b) + (r) + ((c) << CTX_SHIFT)), F##_MASK, F##_SHIFT, (v))
+
+#define GET_FIELD(addr, mask, shift) ((readl(addr) >> (shift)) & (mask))
+
+#define SET_FIELD(addr, mask, shift, v) \
+do { \
+ int t = readl(addr); \
+ writel((t & ~((mask) << (shift))) + (((v) & (mask)) << (shift)), addr);\
+} while (0)
+
+
+#define NUM_FL_PTE 4096
+#define NUM_SL_PTE 256
+
+/* First-level page table bits */
+#define FL_BASE_MASK 0xFFFFFC00
+#define FL_TYPE_TABLE (1 << 0)
+#define FL_TYPE_SECT (2 << 0)
+#define FL_SUPERSECTION (1 << 18)
+#define FL_AP_WRITE (1 << 10)
+#define FL_AP_READ (1 << 11)
+#define FL_SHARED (1 << 16)
+#define FL_OFFSET(va) (((va) & 0xFFF00000) >> 20)
+
+/* Second-level page table bits */
+#define SL_BASE_MASK_LARGE 0xFFFF0000
+#define SL_BASE_MASK_SMALL 0xFFFFF000
+#define SL_TYPE_LARGE (1 << 0)
+#define SL_TYPE_SMALL (2 << 0)
+#define SL_AP0 (1 << 4)
+#define SL_AP1 (2 << 4)
+#define SL_SHARED (1 << 10)
+#define SL_OFFSET(va) (((va) & 0xFF000) >> 12)
+
+/* Global register setters / getters */
+#define SET_M2VCBR_N(b, N, v) SET_GLOBAL_REG_N(M2VCBR_N, N, (b), (v))
+#define SET_CBACR_N(b, N, v) SET_GLOBAL_REG_N(CBACR_N, N, (b), (v))
+#define SET_TLBRSW(b, v) SET_GLOBAL_REG(TLBRSW, (b), (v))
+#define SET_TLBTR0(b, v) SET_GLOBAL_REG(TLBTR0, (b), (v))
+#define SET_TLBTR1(b, v) SET_GLOBAL_REG(TLBTR1, (b), (v))
+#define SET_TLBTR2(b, v) SET_GLOBAL_REG(TLBTR2, (b), (v))
+#define SET_TESTBUSCR(b, v) SET_GLOBAL_REG(TESTBUSCR, (b), (v))
+#define SET_GLOBAL_TLBIALL(b, v) SET_GLOBAL_REG(GLOBAL_TLBIALL, (b), (v))
+#define SET_TLBIVMID(b, v) SET_GLOBAL_REG(TLBIVMID, (b), (v))
+#define SET_CR(b, v) SET_GLOBAL_REG(CR, (b), (v))
+#define SET_EAR(b, v) SET_GLOBAL_REG(EAR, (b), (v))
+#define SET_ESR(b, v) SET_GLOBAL_REG(ESR, (b), (v))
+#define SET_ESRRESTORE(b, v) SET_GLOBAL_REG(ESRRESTORE, (b), (v))
+#define SET_ESYNR0(b, v) SET_GLOBAL_REG(ESYNR0, (b), (v))
+#define SET_ESYNR1(b, v) SET_GLOBAL_REG(ESYNR1, (b), (v))
+#define SET_RPU_ACR(b, v) SET_GLOBAL_REG(RPU_ACR, (b), (v))
+
+#define GET_M2VCBR_N(b, N) GET_GLOBAL_REG_N(M2VCBR_N, N, (b))
+#define GET_CBACR_N(b, N) GET_GLOBAL_REG_N(CBACR_N, N, (b))
+#define GET_TLBTR0(b) GET_GLOBAL_REG(TLBTR0, (b))
+#define GET_TLBTR1(b) GET_GLOBAL_REG(TLBTR1, (b))
+#define GET_TLBTR2(b) GET_GLOBAL_REG(TLBTR2, (b))
+#define GET_TESTBUSCR(b) GET_GLOBAL_REG(TESTBUSCR, (b))
+#define GET_GLOBAL_TLBIALL(b) GET_GLOBAL_REG(GLOBAL_TLBIALL, (b))
+#define GET_TLBIVMID(b) GET_GLOBAL_REG(TLBIVMID, (b))
+#define GET_CR(b) GET_GLOBAL_REG(CR, (b))
+#define GET_EAR(b) GET_GLOBAL_REG(EAR, (b))
+#define GET_ESR(b) GET_GLOBAL_REG(ESR, (b))
+#define GET_ESRRESTORE(b) GET_GLOBAL_REG(ESRRESTORE, (b))
+#define GET_ESYNR0(b) GET_GLOBAL_REG(ESYNR0, (b))
+#define GET_ESYNR1(b) GET_GLOBAL_REG(ESYNR1, (b))
+#define GET_REV(b) GET_GLOBAL_REG(REV, (b))
+#define GET_IDR(b) GET_GLOBAL_REG(IDR, (b))
+#define GET_RPU_ACR(b) GET_GLOBAL_REG(RPU_ACR, (b))
+
+
+/* Context register setters/getters */
+#define SET_SCTLR(b, c, v) SET_CTX_REG(SCTLR, (b), (c), (v))
+#define SET_ACTLR(b, c, v) SET_CTX_REG(ACTLR, (b), (c), (v))
+#define SET_CONTEXTIDR(b, c, v) SET_CTX_REG(CONTEXTIDR, (b), (c), (v))
+#define SET_TTBR0(b, c, v) SET_CTX_REG(TTBR0, (b), (c), (v))
+#define SET_TTBR1(b, c, v) SET_CTX_REG(TTBR1, (b), (c), (v))
+#define SET_TTBCR(b, c, v) SET_CTX_REG(TTBCR, (b), (c), (v))
+#define SET_PAR(b, c, v) SET_CTX_REG(PAR, (b), (c), (v))
+#define SET_FSR(b, c, v) SET_CTX_REG(FSR, (b), (c), (v))
+#define SET_FSRRESTORE(b, c, v) SET_CTX_REG(FSRRESTORE, (b), (c), (v))
+#define SET_FAR(b, c, v) SET_CTX_REG(FAR, (b), (c), (v))
+#define SET_FSYNR0(b, c, v) SET_CTX_REG(FSYNR0, (b), (c), (v))
+#define SET_FSYNR1(b, c, v) SET_CTX_REG(FSYNR1, (b), (c), (v))
+#define SET_PRRR(b, c, v) SET_CTX_REG(PRRR, (b), (c), (v))
+#define SET_NMRR(b, c, v) SET_CTX_REG(NMRR, (b), (c), (v))
+#define SET_TLBLKCR(b, c, v) SET_CTX_REG(TLBLCKR, (b), (c), (v))
+#define SET_V2PSR(b, c, v) SET_CTX_REG(V2PSR, (b), (c), (v))
+#define SET_TLBFLPTER(b, c, v) SET_CTX_REG(TLBFLPTER, (b), (c), (v))
+#define SET_TLBSLPTER(b, c, v) SET_CTX_REG(TLBSLPTER, (b), (c), (v))
+#define SET_BFBCR(b, c, v) SET_CTX_REG(BFBCR, (b), (c), (v))
+#define SET_CTX_TLBIALL(b, c, v) SET_CTX_REG(CTX_TLBIALL, (b), (c), (v))
+#define SET_TLBIASID(b, c, v) SET_CTX_REG(TLBIASID, (b), (c), (v))
+#define SET_TLBIVA(b, c, v) SET_CTX_REG(TLBIVA, (b), (c), (v))
+#define SET_TLBIVAA(b, c, v) SET_CTX_REG(TLBIVAA, (b), (c), (v))
+#define SET_V2PPR(b, c, v) SET_CTX_REG(V2PPR, (b), (c), (v))
+#define SET_V2PPW(b, c, v) SET_CTX_REG(V2PPW, (b), (c), (v))
+#define SET_V2PUR(b, c, v) SET_CTX_REG(V2PUR, (b), (c), (v))
+#define SET_V2PUW(b, c, v) SET_CTX_REG(V2PUW, (b), (c), (v))
+#define SET_RESUME(b, c, v) SET_CTX_REG(RESUME, (b), (c), (v))
+
+#define GET_SCTLR(b, c) GET_CTX_REG(SCTLR, (b), (c))
+#define GET_ACTLR(b, c) GET_CTX_REG(ACTLR, (b), (c))
+#define GET_CONTEXTIDR(b, c) GET_CTX_REG(CONTEXTIDR, (b), (c))
+#define GET_TTBR0(b, c) GET_CTX_REG(TTBR0, (b), (c))
+#define GET_TTBR1(b, c) GET_CTX_REG(TTBR1, (b), (c))
+#define GET_TTBCR(b, c) GET_CTX_REG(TTBCR, (b), (c))
+#define GET_PAR(b, c) GET_CTX_REG(PAR, (b), (c))
+#define GET_FSR(b, c) GET_CTX_REG(FSR, (b), (c))
+#define GET_FSRRESTORE(b, c) GET_CTX_REG(FSRRESTORE, (b), (c))
+#define GET_FAR(b, c) GET_CTX_REG(FAR, (b), (c))
+#define GET_FSYNR0(b, c) GET_CTX_REG(FSYNR0, (b), (c))
+#define GET_FSYNR1(b, c) GET_CTX_REG(FSYNR1, (b), (c))
+#define GET_PRRR(b, c) GET_CTX_REG(PRRR, (b), (c))
+#define GET_NMRR(b, c) GET_CTX_REG(NMRR, (b), (c))
+#define GET_TLBLCKR(b, c) GET_CTX_REG(TLBLCKR, (b), (c))
+#define GET_V2PSR(b, c) GET_CTX_REG(V2PSR, (b), (c))
+#define GET_TLBFLPTER(b, c) GET_CTX_REG(TLBFLPTER, (b), (c))
+#define GET_TLBSLPTER(b, c) GET_CTX_REG(TLBSLPTER, (b), (c))
+#define GET_BFBCR(b, c) GET_CTX_REG(BFBCR, (b), (c))
+#define GET_CTX_TLBIALL(b, c) GET_CTX_REG(CTX_TLBIALL, (b), (c))
+#define GET_TLBIASID(b, c) GET_CTX_REG(TLBIASID, (b), (c))
+#define GET_TLBIVA(b, c) GET_CTX_REG(TLBIVA, (b), (c))
+#define GET_TLBIVAA(b, c) GET_CTX_REG(TLBIVAA, (b), (c))
+#define GET_V2PPR(b, c) GET_CTX_REG(V2PPR, (b), (c))
+#define GET_V2PPW(b, c) GET_CTX_REG(V2PPW, (b), (c))
+#define GET_V2PUR(b, c) GET_CTX_REG(V2PUR, (b), (c))
+#define GET_V2PUW(b, c) GET_CTX_REG(V2PUW, (b), (c))
+#define GET_RESUME(b, c) GET_CTX_REG(RESUME, (b), (c))
+
+
+/* Global field setters / getters */
+/* Global Field Setters: */
+/* CBACR_N */
+#define SET_RWVMID(b, n, v) SET_GLOBAL_FIELD(b, (n<<2)|(CBACR_N), RWVMID, v)
+#define SET_RWE(b, n, v) SET_GLOBAL_FIELD(b, (n<<2)|(CBACR_N), RWE, v)
+#define SET_RWGE(b, n, v) SET_GLOBAL_FIELD(b, (n<<2)|(CBACR_N), RWGE, v)
+#define SET_CBVMID(b, n, v) SET_GLOBAL_FIELD(b, (n<<2)|(CBACR_N), CBVMID, v)
+#define SET_IRPTNDX(b, n, v) SET_GLOBAL_FIELD(b, (n<<2)|(CBACR_N), IRPTNDX, v)
+
+
+/* M2VCBR_N */
+#define SET_VMID(b, n, v) SET_GLOBAL_FIELD(b, (n<<2)|(M2VCBR_N), VMID, v)
+#define SET_CBNDX(b, n, v) SET_GLOBAL_FIELD(b, (n<<2)|(M2VCBR_N), CBNDX, v)
+#define SET_BYPASSD(b, n, v) SET_GLOBAL_FIELD(b, (n<<2)|(M2VCBR_N), BYPASSD, v)
+#define SET_BPRCOSH(b, n, v) SET_GLOBAL_FIELD(b, (n<<2)|(M2VCBR_N), BPRCOSH, v)
+#define SET_BPRCISH(b, n, v) SET_GLOBAL_FIELD(b, (n<<2)|(M2VCBR_N), BPRCISH, v)
+#define SET_BPRCNSH(b, n, v) SET_GLOBAL_FIELD(b, (n<<2)|(M2VCBR_N), BPRCNSH, v)
+#define SET_BPSHCFG(b, n, v) SET_GLOBAL_FIELD(b, (n<<2)|(M2VCBR_N), BPSHCFG, v)
+#define SET_NSCFG(b, n, v) SET_GLOBAL_FIELD(b, (n<<2)|(M2VCBR_N), NSCFG, v)
+#define SET_BPMTCFG(b, n, v) SET_GLOBAL_FIELD(b, (n<<2)|(M2VCBR_N), BPMTCFG, v)
+#define SET_BPMEMTYPE(b, n, v) \
+ SET_GLOBAL_FIELD(b, (n<<2)|(M2VCBR_N), BPMEMTYPE, v)
+
+
+/* CR */
+#define SET_RPUE(b, v) SET_GLOBAL_FIELD(b, CR, RPUE, v)
+#define SET_RPUERE(b, v) SET_GLOBAL_FIELD(b, CR, RPUERE, v)
+#define SET_RPUEIE(b, v) SET_GLOBAL_FIELD(b, CR, RPUEIE, v)
+#define SET_DCDEE(b, v) SET_GLOBAL_FIELD(b, CR, DCDEE, v)
+#define SET_CLIENTPD(b, v) SET_GLOBAL_FIELD(b, CR, CLIENTPD, v)
+#define SET_STALLD(b, v) SET_GLOBAL_FIELD(b, CR, STALLD, v)
+#define SET_TLBLKCRWE(b, v) SET_GLOBAL_FIELD(b, CR, TLBLKCRWE, v)
+#define SET_CR_TLBIALLCFG(b, v) SET_GLOBAL_FIELD(b, CR, CR_TLBIALLCFG, v)
+#define SET_TLBIVMIDCFG(b, v) SET_GLOBAL_FIELD(b, CR, TLBIVMIDCFG, v)
+#define SET_CR_HUME(b, v) SET_GLOBAL_FIELD(b, CR, CR_HUME, v)
+
+
+/* ESR */
+#define SET_CFG(b, v) SET_GLOBAL_FIELD(b, ESR, CFG, v)
+#define SET_BYPASS(b, v) SET_GLOBAL_FIELD(b, ESR, BYPASS, v)
+#define SET_ESR_MULTI(b, v) SET_GLOBAL_FIELD(b, ESR, ESR_MULTI, v)
+
+
+/* ESYNR0 */
+#define SET_ESYNR0_AMID(b, v) SET_GLOBAL_FIELD(b, ESYNR0, ESYNR0_AMID, v)
+#define SET_ESYNR0_APID(b, v) SET_GLOBAL_FIELD(b, ESYNR0, ESYNR0_APID, v)
+#define SET_ESYNR0_ABID(b, v) SET_GLOBAL_FIELD(b, ESYNR0, ESYNR0_ABID, v)
+#define SET_ESYNR0_AVMID(b, v) SET_GLOBAL_FIELD(b, ESYNR0, ESYNR0_AVMID, v)
+#define SET_ESYNR0_ATID(b, v) SET_GLOBAL_FIELD(b, ESYNR0, ESYNR0_ATID, v)
+
+
+/* ESYNR1 */
+#define SET_ESYNR1_AMEMTYPE(b, v) \
+ SET_GLOBAL_FIELD(b, ESYNR1, ESYNR1_AMEMTYPE, v)
+#define SET_ESYNR1_ASHARED(b, v) SET_GLOBAL_FIELD(b, ESYNR1, ESYNR1_ASHARED, v)
+#define SET_ESYNR1_AINNERSHARED(b, v) \
+ SET_GLOBAL_FIELD(b, ESYNR1, ESYNR1_AINNERSHARED, v)
+#define SET_ESYNR1_APRIV(b, v) SET_GLOBAL_FIELD(b, ESYNR1, ESYNR1_APRIV, v)
+#define SET_ESYNR1_APROTNS(b, v) SET_GLOBAL_FIELD(b, ESYNR1, ESYNR1_APROTNS, v)
+#define SET_ESYNR1_AINST(b, v) SET_GLOBAL_FIELD(b, ESYNR1, ESYNR1_AINST, v)
+#define SET_ESYNR1_AWRITE(b, v) SET_GLOBAL_FIELD(b, ESYNR1, ESYNR1_AWRITE, v)
+#define SET_ESYNR1_ABURST(b, v) SET_GLOBAL_FIELD(b, ESYNR1, ESYNR1_ABURST, v)
+#define SET_ESYNR1_ALEN(b, v) SET_GLOBAL_FIELD(b, ESYNR1, ESYNR1_ALEN, v)
+#define SET_ESYNR1_ASIZE(b, v) SET_GLOBAL_FIELD(b, ESYNR1, ESYNR1_ASIZE, v)
+#define SET_ESYNR1_ALOCK(b, v) SET_GLOBAL_FIELD(b, ESYNR1, ESYNR1_ALOCK, v)
+#define SET_ESYNR1_AOOO(b, v) SET_GLOBAL_FIELD(b, ESYNR1, ESYNR1_AOOO, v)
+#define SET_ESYNR1_AFULL(b, v) SET_GLOBAL_FIELD(b, ESYNR1, ESYNR1_AFULL, v)
+#define SET_ESYNR1_AC(b, v) SET_GLOBAL_FIELD(b, ESYNR1, ESYNR1_AC, v)
+#define SET_ESYNR1_DCD(b, v) SET_GLOBAL_FIELD(b, ESYNR1, ESYNR1_DCD, v)
+
+
+/* TESTBUSCR */
+#define SET_TBE(b, v) SET_GLOBAL_FIELD(b, TESTBUSCR, TBE, v)
+#define SET_SPDMBE(b, v) SET_GLOBAL_FIELD(b, TESTBUSCR, SPDMBE, v)
+#define SET_WGSEL(b, v) SET_GLOBAL_FIELD(b, TESTBUSCR, WGSEL, v)
+#define SET_TBLSEL(b, v) SET_GLOBAL_FIELD(b, TESTBUSCR, TBLSEL, v)
+#define SET_TBHSEL(b, v) SET_GLOBAL_FIELD(b, TESTBUSCR, TBHSEL, v)
+#define SET_SPDM0SEL(b, v) SET_GLOBAL_FIELD(b, TESTBUSCR, SPDM0SEL, v)
+#define SET_SPDM1SEL(b, v) SET_GLOBAL_FIELD(b, TESTBUSCR, SPDM1SEL, v)
+#define SET_SPDM2SEL(b, v) SET_GLOBAL_FIELD(b, TESTBUSCR, SPDM2SEL, v)
+#define SET_SPDM3SEL(b, v) SET_GLOBAL_FIELD(b, TESTBUSCR, SPDM3SEL, v)
+
+
+/* TLBIVMID */
+#define SET_TLBIVMID_VMID(b, v) SET_GLOBAL_FIELD(b, TLBIVMID, TLBIVMID_VMID, v)
+
+
+/* TLBRSW */
+#define SET_TLBRSW_INDEX(b, v) SET_GLOBAL_FIELD(b, TLBRSW, TLBRSW_INDEX, v)
+#define SET_TLBBFBS(b, v) SET_GLOBAL_FIELD(b, TLBRSW, TLBBFBS, v)
+
+
+/* TLBTR0 */
+#define SET_PR(b, v) SET_GLOBAL_FIELD(b, TLBTR0, PR, v)
+#define SET_PW(b, v) SET_GLOBAL_FIELD(b, TLBTR0, PW, v)
+#define SET_UR(b, v) SET_GLOBAL_FIELD(b, TLBTR0, UR, v)
+#define SET_UW(b, v) SET_GLOBAL_FIELD(b, TLBTR0, UW, v)
+#define SET_XN(b, v) SET_GLOBAL_FIELD(b, TLBTR0, XN, v)
+#define SET_NSDESC(b, v) SET_GLOBAL_FIELD(b, TLBTR0, NSDESC, v)
+#define SET_ISH(b, v) SET_GLOBAL_FIELD(b, TLBTR0, ISH, v)
+#define SET_SH(b, v) SET_GLOBAL_FIELD(b, TLBTR0, SH, v)
+#define SET_MT(b, v) SET_GLOBAL_FIELD(b, TLBTR0, MT, v)
+#define SET_DPSIZR(b, v) SET_GLOBAL_FIELD(b, TLBTR0, DPSIZR, v)
+#define SET_DPSIZC(b, v) SET_GLOBAL_FIELD(b, TLBTR0, DPSIZC, v)
+
+
+/* TLBTR1 */
+#define SET_TLBTR1_VMID(b, v) SET_GLOBAL_FIELD(b, TLBTR1, TLBTR1_VMID, v)
+#define SET_TLBTR1_PA(b, v) SET_GLOBAL_FIELD(b, TLBTR1, TLBTR1_PA, v)
+
+
+/* TLBTR2 */
+#define SET_TLBTR2_ASID(b, v) SET_GLOBAL_FIELD(b, TLBTR2, TLBTR2_ASID, v)
+#define SET_TLBTR2_V(b, v) SET_GLOBAL_FIELD(b, TLBTR2, TLBTR2_V, v)
+#define SET_TLBTR2_NSTID(b, v) SET_GLOBAL_FIELD(b, TLBTR2, TLBTR2_NSTID, v)
+#define SET_TLBTR2_NV(b, v) SET_GLOBAL_FIELD(b, TLBTR2, TLBTR2_NV, v)
+#define SET_TLBTR2_VA(b, v) SET_GLOBAL_FIELD(b, TLBTR2, TLBTR2_VA, v)
+
+
+/* Global Field Getters */
+/* CBACR_N */
+#define GET_RWVMID(b, n) GET_GLOBAL_FIELD(b, (n<<2)|(CBACR_N), RWVMID)
+#define GET_RWE(b, n) GET_GLOBAL_FIELD(b, (n<<2)|(CBACR_N), RWE)
+#define GET_RWGE(b, n) GET_GLOBAL_FIELD(b, (n<<2)|(CBACR_N), RWGE)
+#define GET_CBVMID(b, n) GET_GLOBAL_FIELD(b, (n<<2)|(CBACR_N), CBVMID)
+#define GET_IRPTNDX(b, n) GET_GLOBAL_FIELD(b, (n<<2)|(CBACR_N), IRPTNDX)
+
+
+/* M2VCBR_N */
+#define GET_VMID(b, n) GET_GLOBAL_FIELD(b, (n<<2)|(M2VCBR_N), VMID)
+#define GET_CBNDX(b, n) GET_GLOBAL_FIELD(b, (n<<2)|(M2VCBR_N), CBNDX)
+#define GET_BYPASSD(b, n) GET_GLOBAL_FIELD(b, (n<<2)|(M2VCBR_N), BYPASSD)
+#define GET_BPRCOSH(b, n) GET_GLOBAL_FIELD(b, (n<<2)|(M2VCBR_N), BPRCOSH)
+#define GET_BPRCISH(b, n) GET_GLOBAL_FIELD(b, (n<<2)|(M2VCBR_N), BPRCISH)
+#define GET_BPRCNSH(b, n) GET_GLOBAL_FIELD(b, (n<<2)|(M2VCBR_N), BPRCNSH)
+#define GET_BPSHCFG(b, n) GET_GLOBAL_FIELD(b, (n<<2)|(M2VCBR_N), BPSHCFG)
+#define GET_NSCFG(b, n) GET_GLOBAL_FIELD(b, (n<<2)|(M2VCBR_N), NSCFG)
+#define GET_BPMTCFG(b, n) GET_GLOBAL_FIELD(b, (n<<2)|(M2VCBR_N), BPMTCFG)
+#define GET_BPMEMTYPE(b, n) GET_GLOBAL_FIELD(b, (n<<2)|(M2VCBR_N), BPMEMTYPE)
+
+
+/* CR */
+#define GET_RPUE(b) GET_GLOBAL_FIELD(b, CR, RPUE)
+#define GET_RPUERE(b) GET_GLOBAL_FIELD(b, CR, RPUERE)
+#define GET_RPUEIE(b) GET_GLOBAL_FIELD(b, CR, RPUEIE)
+#define GET_DCDEE(b) GET_GLOBAL_FIELD(b, CR, DCDEE)
+#define GET_CLIENTPD(b) GET_GLOBAL_FIELD(b, CR, CLIENTPD)
+#define GET_STALLD(b) GET_GLOBAL_FIELD(b, CR, STALLD)
+#define GET_TLBLKCRWE(b) GET_GLOBAL_FIELD(b, CR, TLBLKCRWE)
+#define GET_CR_TLBIALLCFG(b) GET_GLOBAL_FIELD(b, CR, CR_TLBIALLCFG)
+#define GET_TLBIVMIDCFG(b) GET_GLOBAL_FIELD(b, CR, TLBIVMIDCFG)
+#define GET_CR_HUME(b) GET_GLOBAL_FIELD(b, CR, CR_HUME)
+
+
+/* ESR */
+#define GET_CFG(b) GET_GLOBAL_FIELD(b, ESR, CFG)
+#define GET_BYPASS(b) GET_GLOBAL_FIELD(b, ESR, BYPASS)
+#define GET_ESR_MULTI(b) GET_GLOBAL_FIELD(b, ESR, ESR_MULTI)
+
+
+/* ESYNR0 */
+#define GET_ESYNR0_AMID(b) GET_GLOBAL_FIELD(b, ESYNR0, ESYNR0_AMID)
+#define GET_ESYNR0_APID(b) GET_GLOBAL_FIELD(b, ESYNR0, ESYNR0_APID)
+#define GET_ESYNR0_ABID(b) GET_GLOBAL_FIELD(b, ESYNR0, ESYNR0_ABID)
+#define GET_ESYNR0_AVMID(b) GET_GLOBAL_FIELD(b, ESYNR0, ESYNR0_AVMID)
+#define GET_ESYNR0_ATID(b) GET_GLOBAL_FIELD(b, ESYNR0, ESYNR0_ATID)
+
+
+/* ESYNR1 */
+#define GET_ESYNR1_AMEMTYPE(b) GET_GLOBAL_FIELD(b, ESYNR1, ESYNR1_AMEMTYPE)
+#define GET_ESYNR1_ASHARED(b) GET_GLOBAL_FIELD(b, ESYNR1, ESYNR1_ASHARED)
+#define GET_ESYNR1_AINNERSHARED(b) \
+ GET_GLOBAL_FIELD(b, ESYNR1, ESYNR1_AINNERSHARED)
+#define GET_ESYNR1_APRIV(b) GET_GLOBAL_FIELD(b, ESYNR1, ESYNR1_APRIV)
+#define GET_ESYNR1_APROTNS(b) GET_GLOBAL_FIELD(b, ESYNR1, ESYNR1_APROTNS)
+#define GET_ESYNR1_AINST(b) GET_GLOBAL_FIELD(b, ESYNR1, ESYNR1_AINST)
+#define GET_ESYNR1_AWRITE(b) GET_GLOBAL_FIELD(b, ESYNR1, ESYNR1_AWRITE)
+#define GET_ESYNR1_ABURST(b) GET_GLOBAL_FIELD(b, ESYNR1, ESYNR1_ABURST)
+#define GET_ESYNR1_ALEN(b) GET_GLOBAL_FIELD(b, ESYNR1, ESYNR1_ALEN)
+#define GET_ESYNR1_ASIZE(b) GET_GLOBAL_FIELD(b, ESYNR1, ESYNR1_ASIZE)
+#define GET_ESYNR1_ALOCK(b) GET_GLOBAL_FIELD(b, ESYNR1, ESYNR1_ALOCK)
+#define GET_ESYNR1_AOOO(b) GET_GLOBAL_FIELD(b, ESYNR1, ESYNR1_AOOO)
+#define GET_ESYNR1_AFULL(b) GET_GLOBAL_FIELD(b, ESYNR1, ESYNR1_AFULL)
+#define GET_ESYNR1_AC(b) GET_GLOBAL_FIELD(b, ESYNR1, ESYNR1_AC)
+#define GET_ESYNR1_DCD(b) GET_GLOBAL_FIELD(b, ESYNR1, ESYNR1_DCD)
+
+
+/* IDR */
+#define GET_NM2VCBMT(b) GET_GLOBAL_FIELD(b, IDR, NM2VCBMT)
+#define GET_HTW(b) GET_GLOBAL_FIELD(b, IDR, HTW)
+#define GET_HUM(b) GET_GLOBAL_FIELD(b, IDR, HUM)
+#define GET_TLBSIZE(b) GET_GLOBAL_FIELD(b, IDR, TLBSIZE)
+#define GET_NCB(b) GET_GLOBAL_FIELD(b, IDR, NCB)
+#define GET_NIRPT(b) GET_GLOBAL_FIELD(b, IDR, NIRPT)
+
+
+/* REV */
+#define GET_MAJOR(b) GET_GLOBAL_FIELD(b, REV, MAJOR)
+#define GET_MINOR(b) GET_GLOBAL_FIELD(b, REV, MINOR)
+
+
+/* TESTBUSCR */
+#define GET_TBE(b) GET_GLOBAL_FIELD(b, TESTBUSCR, TBE)
+#define GET_SPDMBE(b) GET_GLOBAL_FIELD(b, TESTBUSCR, SPDMBE)
+#define GET_WGSEL(b) GET_GLOBAL_FIELD(b, TESTBUSCR, WGSEL)
+#define GET_TBLSEL(b) GET_GLOBAL_FIELD(b, TESTBUSCR, TBLSEL)
+#define GET_TBHSEL(b) GET_GLOBAL_FIELD(b, TESTBUSCR, TBHSEL)
+#define GET_SPDM0SEL(b) GET_GLOBAL_FIELD(b, TESTBUSCR, SPDM0SEL)
+#define GET_SPDM1SEL(b) GET_GLOBAL_FIELD(b, TESTBUSCR, SPDM1SEL)
+#define GET_SPDM2SEL(b) GET_GLOBAL_FIELD(b, TESTBUSCR, SPDM2SEL)
+#define GET_SPDM3SEL(b) GET_GLOBAL_FIELD(b, TESTBUSCR, SPDM3SEL)
+
+
+/* TLBIVMID */
+#define GET_TLBIVMID_VMID(b) GET_GLOBAL_FIELD(b, TLBIVMID, TLBIVMID_VMID)
+
+
+/* TLBTR0 */
+#define GET_PR(b) GET_GLOBAL_FIELD(b, TLBTR0, PR)
+#define GET_PW(b) GET_GLOBAL_FIELD(b, TLBTR0, PW)
+#define GET_UR(b) GET_GLOBAL_FIELD(b, TLBTR0, UR)
+#define GET_UW(b) GET_GLOBAL_FIELD(b, TLBTR0, UW)
+#define GET_XN(b) GET_GLOBAL_FIELD(b, TLBTR0, XN)
+#define GET_NSDESC(b) GET_GLOBAL_FIELD(b, TLBTR0, NSDESC)
+#define GET_ISH(b) GET_GLOBAL_FIELD(b, TLBTR0, ISH)
+#define GET_SH(b) GET_GLOBAL_FIELD(b, TLBTR0, SH)
+#define GET_MT(b) GET_GLOBAL_FIELD(b, TLBTR0, MT)
+#define GET_DPSIZR(b) GET_GLOBAL_FIELD(b, TLBTR0, DPSIZR)
+#define GET_DPSIZC(b) GET_GLOBAL_FIELD(b, TLBTR0, DPSIZC)
+
+
+/* TLBTR1 */
+#define GET_TLBTR1_VMID(b) GET_GLOBAL_FIELD(b, TLBTR1, TLBTR1_VMID)
+#define GET_TLBTR1_PA(b) GET_GLOBAL_FIELD(b, TLBTR1, TLBTR1_PA)
+
+
+/* TLBTR2 */
+#define GET_TLBTR2_ASID(b) GET_GLOBAL_FIELD(b, TLBTR2, TLBTR2_ASID)
+#define GET_TLBTR2_V(b) GET_GLOBAL_FIELD(b, TLBTR2, TLBTR2_V)
+#define GET_TLBTR2_NSTID(b) GET_GLOBAL_FIELD(b, TLBTR2, TLBTR2_NSTID)
+#define GET_TLBTR2_NV(b) GET_GLOBAL_FIELD(b, TLBTR2, TLBTR2_NV)
+#define GET_TLBTR2_VA(b) GET_GLOBAL_FIELD(b, TLBTR2, TLBTR2_VA)
+
+
+/* Context Register setters / getters */
+/* Context Register setters */
+/* ACTLR */
+#define SET_CFERE(b, c, v) SET_CONTEXT_FIELD(b, c, ACTLR, CFERE, v)
+#define SET_CFEIE(b, c, v) SET_CONTEXT_FIELD(b, c, ACTLR, CFEIE, v)
+#define SET_PTSHCFG(b, c, v) SET_CONTEXT_FIELD(b, c, ACTLR, PTSHCFG, v)
+#define SET_RCOSH(b, c, v) SET_CONTEXT_FIELD(b, c, ACTLR, RCOSH, v)
+#define SET_RCISH(b, c, v) SET_CONTEXT_FIELD(b, c, ACTLR, RCISH, v)
+#define SET_RCNSH(b, c, v) SET_CONTEXT_FIELD(b, c, ACTLR, RCNSH, v)
+#define SET_PRIVCFG(b, c, v) SET_CONTEXT_FIELD(b, c, ACTLR, PRIVCFG, v)
+#define SET_DNA(b, c, v) SET_CONTEXT_FIELD(b, c, ACTLR, DNA, v)
+#define SET_DNLV2PA(b, c, v) SET_CONTEXT_FIELD(b, c, ACTLR, DNLV2PA, v)
+#define SET_TLBMCFG(b, c, v) SET_CONTEXT_FIELD(b, c, ACTLR, TLBMCFG, v)
+#define SET_CFCFG(b, c, v) SET_CONTEXT_FIELD(b, c, ACTLR, CFCFG, v)
+#define SET_TIPCF(b, c, v) SET_CONTEXT_FIELD(b, c, ACTLR, TIPCF, v)
+#define SET_V2PCFG(b, c, v) SET_CONTEXT_FIELD(b, c, ACTLR, V2PCFG, v)
+#define SET_HUME(b, c, v) SET_CONTEXT_FIELD(b, c, ACTLR, HUME, v)
+#define SET_PTMTCFG(b, c, v) SET_CONTEXT_FIELD(b, c, ACTLR, PTMTCFG, v)
+#define SET_PTMEMTYPE(b, c, v) SET_CONTEXT_FIELD(b, c, ACTLR, PTMEMTYPE, v)
+
+
+/* BFBCR */
+#define SET_BFBDFE(b, c, v) SET_CONTEXT_FIELD(b, c, BFBCR, BFBDFE, v)
+#define SET_BFBSFE(b, c, v) SET_CONTEXT_FIELD(b, c, BFBCR, BFBSFE, v)
+#define SET_SFVS(b, c, v) SET_CONTEXT_FIELD(b, c, BFBCR, SFVS, v)
+#define SET_FLVIC(b, c, v) SET_CONTEXT_FIELD(b, c, BFBCR, FLVIC, v)
+#define SET_SLVIC(b, c, v) SET_CONTEXT_FIELD(b, c, BFBCR, SLVIC, v)
+
+
+/* CONTEXTIDR */
+#define SET_CONTEXTIDR_ASID(b, c, v) \
+ SET_CONTEXT_FIELD(b, c, CONTEXTIDR, CONTEXTIDR_ASID, v)
+#define SET_CONTEXTIDR_PROCID(b, c, v) \
+ SET_CONTEXT_FIELD(b, c, CONTEXTIDR, PROCID, v)
+
+
+/* FSR */
+#define SET_TF(b, c, v) SET_CONTEXT_FIELD(b, c, FSR, TF, v)
+#define SET_AFF(b, c, v) SET_CONTEXT_FIELD(b, c, FSR, AFF, v)
+#define SET_APF(b, c, v) SET_CONTEXT_FIELD(b, c, FSR, APF, v)
+#define SET_TLBMF(b, c, v) SET_CONTEXT_FIELD(b, c, FSR, TLBMF, v)
+#define SET_HTWDEEF(b, c, v) SET_CONTEXT_FIELD(b, c, FSR, HTWDEEF, v)
+#define SET_HTWSEEF(b, c, v) SET_CONTEXT_FIELD(b, c, FSR, HTWSEEF, v)
+#define SET_MHF(b, c, v) SET_CONTEXT_FIELD(b, c, FSR, MHF, v)
+#define SET_SL(b, c, v) SET_CONTEXT_FIELD(b, c, FSR, SL, v)
+#define SET_SS(b, c, v) SET_CONTEXT_FIELD(b, c, FSR, SS, v)
+#define SET_MULTI(b, c, v) SET_CONTEXT_FIELD(b, c, FSR, MULTI, v)
+
+
+/* FSYNR0 */
+#define SET_AMID(b, c, v) SET_CONTEXT_FIELD(b, c, FSYNR0, AMID, v)
+#define SET_APID(b, c, v) SET_CONTEXT_FIELD(b, c, FSYNR0, APID, v)
+#define SET_ABID(b, c, v) SET_CONTEXT_FIELD(b, c, FSYNR0, ABID, v)
+#define SET_ATID(b, c, v) SET_CONTEXT_FIELD(b, c, FSYNR0, ATID, v)
+
+
+/* FSYNR1 */
+#define SET_AMEMTYPE(b, c, v) SET_CONTEXT_FIELD(b, c, FSYNR1, AMEMTYPE, v)
+#define SET_ASHARED(b, c, v) SET_CONTEXT_FIELD(b, c, FSYNR1, ASHARED, v)
+#define SET_AINNERSHARED(b, c, v) \
+ SET_CONTEXT_FIELD(b, c, FSYNR1, AINNERSHARED, v)
+#define SET_APRIV(b, c, v) SET_CONTEXT_FIELD(b, c, FSYNR1, APRIV, v)
+#define SET_APROTNS(b, c, v) SET_CONTEXT_FIELD(b, c, FSYNR1, APROTNS, v)
+#define SET_AINST(b, c, v) SET_CONTEXT_FIELD(b, c, FSYNR1, AINST, v)
+#define SET_AWRITE(b, c, v) SET_CONTEXT_FIELD(b, c, FSYNR1, AWRITE, v)
+#define SET_ABURST(b, c, v) SET_CONTEXT_FIELD(b, c, FSYNR1, ABURST, v)
+#define SET_ALEN(b, c, v) SET_CONTEXT_FIELD(b, c, FSYNR1, ALEN, v)
+#define SET_FSYNR1_ASIZE(b, c, v) \
+ SET_CONTEXT_FIELD(b, c, FSYNR1, FSYNR1_ASIZE, v)
+#define SET_ALOCK(b, c, v) SET_CONTEXT_FIELD(b, c, FSYNR1, ALOCK, v)
+#define SET_AFULL(b, c, v) SET_CONTEXT_FIELD(b, c, FSYNR1, AFULL, v)
+
+
+/* NMRR */
+#define SET_ICPC0(b, c, v) SET_CONTEXT_FIELD(b, c, NMRR, ICPC0, v)
+#define SET_ICPC1(b, c, v) SET_CONTEXT_FIELD(b, c, NMRR, ICPC1, v)
+#define SET_ICPC2(b, c, v) SET_CONTEXT_FIELD(b, c, NMRR, ICPC2, v)
+#define SET_ICPC3(b, c, v) SET_CONTEXT_FIELD(b, c, NMRR, ICPC3, v)
+#define SET_ICPC4(b, c, v) SET_CONTEXT_FIELD(b, c, NMRR, ICPC4, v)
+#define SET_ICPC5(b, c, v) SET_CONTEXT_FIELD(b, c, NMRR, ICPC5, v)
+#define SET_ICPC6(b, c, v) SET_CONTEXT_FIELD(b, c, NMRR, ICPC6, v)
+#define SET_ICPC7(b, c, v) SET_CONTEXT_FIELD(b, c, NMRR, ICPC7, v)
+#define SET_OCPC0(b, c, v) SET_CONTEXT_FIELD(b, c, NMRR, OCPC0, v)
+#define SET_OCPC1(b, c, v) SET_CONTEXT_FIELD(b, c, NMRR, OCPC1, v)
+#define SET_OCPC2(b, c, v) SET_CONTEXT_FIELD(b, c, NMRR, OCPC2, v)
+#define SET_OCPC3(b, c, v) SET_CONTEXT_FIELD(b, c, NMRR, OCPC3, v)
+#define SET_OCPC4(b, c, v) SET_CONTEXT_FIELD(b, c, NMRR, OCPC4, v)
+#define SET_OCPC5(b, c, v) SET_CONTEXT_FIELD(b, c, NMRR, OCPC5, v)
+#define SET_OCPC6(b, c, v) SET_CONTEXT_FIELD(b, c, NMRR, OCPC6, v)
+#define SET_OCPC7(b, c, v) SET_CONTEXT_FIELD(b, c, NMRR, OCPC7, v)
+
+
+/* PAR */
+#define SET_FAULT(b, c, v) SET_CONTEXT_FIELD(b, c, PAR, FAULT, v)
+
+#define SET_FAULT_TF(b, c, v) SET_CONTEXT_FIELD(b, c, PAR, FAULT_TF, v)
+#define SET_FAULT_AFF(b, c, v) SET_CONTEXT_FIELD(b, c, PAR, FAULT_AFF, v)
+#define SET_FAULT_APF(b, c, v) SET_CONTEXT_FIELD(b, c, PAR, FAULT_APF, v)
+#define SET_FAULT_TLBMF(b, c, v) SET_CONTEXT_FIELD(b, c, PAR, FAULT_TLBMF, v)
+#define SET_FAULT_HTWDEEF(b, c, v) \
+ SET_CONTEXT_FIELD(b, c, PAR, FAULT_HTWDEEF, v)
+#define SET_FAULT_HTWSEEF(b, c, v) \
+ SET_CONTEXT_FIELD(b, c, PAR, FAULT_HTWSEEF, v)
+#define SET_FAULT_MHF(b, c, v) SET_CONTEXT_FIELD(b, c, PAR, FAULT_MHF, v)
+#define SET_FAULT_SL(b, c, v) SET_CONTEXT_FIELD(b, c, PAR, FAULT_SL, v)
+#define SET_FAULT_SS(b, c, v) SET_CONTEXT_FIELD(b, c, PAR, FAULT_SS, v)
+
+#define SET_NOFAULT_SS(b, c, v) SET_CONTEXT_FIELD(b, c, PAR, NOFAULT_SS, v)
+#define SET_NOFAULT_MT(b, c, v) SET_CONTEXT_FIELD(b, c, PAR, NOFAULT_MT, v)
+#define SET_NOFAULT_SH(b, c, v) SET_CONTEXT_FIELD(b, c, PAR, NOFAULT_SH, v)
+#define SET_NOFAULT_NS(b, c, v) SET_CONTEXT_FIELD(b, c, PAR, NOFAULT_NS, v)
+#define SET_NOFAULT_NOS(b, c, v) SET_CONTEXT_FIELD(b, c, PAR, NOFAULT_NOS, v)
+#define SET_NPFAULT_PA(b, c, v) SET_CONTEXT_FIELD(b, c, PAR, NPFAULT_PA, v)
+
+
+/* PRRR */
+#define SET_MTC0(b, c, v) SET_CONTEXT_FIELD(b, c, PRRR, MTC0, v)
+#define SET_MTC1(b, c, v) SET_CONTEXT_FIELD(b, c, PRRR, MTC1, v)
+#define SET_MTC2(b, c, v) SET_CONTEXT_FIELD(b, c, PRRR, MTC2, v)
+#define SET_MTC3(b, c, v) SET_CONTEXT_FIELD(b, c, PRRR, MTC3, v)
+#define SET_MTC4(b, c, v) SET_CONTEXT_FIELD(b, c, PRRR, MTC4, v)
+#define SET_MTC5(b, c, v) SET_CONTEXT_FIELD(b, c, PRRR, MTC5, v)
+#define SET_MTC6(b, c, v) SET_CONTEXT_FIELD(b, c, PRRR, MTC6, v)
+#define SET_MTC7(b, c, v) SET_CONTEXT_FIELD(b, c, PRRR, MTC7, v)
+#define SET_SHDSH0(b, c, v) SET_CONTEXT_FIELD(b, c, PRRR, SHDSH0, v)
+#define SET_SHDSH1(b, c, v) SET_CONTEXT_FIELD(b, c, PRRR, SHDSH1, v)
+#define SET_SHNMSH0(b, c, v) SET_CONTEXT_FIELD(b, c, PRRR, SHNMSH0, v)
+#define SET_SHNMSH1(b, c, v) SET_CONTEXT_FIELD(b, c, PRRR, SHNMSH1, v)
+#define SET_NOS0(b, c, v) SET_CONTEXT_FIELD(b, c, PRRR, NOS0, v)
+#define SET_NOS1(b, c, v) SET_CONTEXT_FIELD(b, c, PRRR, NOS1, v)
+#define SET_NOS2(b, c, v) SET_CONTEXT_FIELD(b, c, PRRR, NOS2, v)
+#define SET_NOS3(b, c, v) SET_CONTEXT_FIELD(b, c, PRRR, NOS3, v)
+#define SET_NOS4(b, c, v) SET_CONTEXT_FIELD(b, c, PRRR, NOS4, v)
+#define SET_NOS5(b, c, v) SET_CONTEXT_FIELD(b, c, PRRR, NOS5, v)
+#define SET_NOS6(b, c, v) SET_CONTEXT_FIELD(b, c, PRRR, NOS6, v)
+#define SET_NOS7(b, c, v) SET_CONTEXT_FIELD(b, c, PRRR, NOS7, v)
+
+
+/* RESUME */
+#define SET_TNR(b, c, v) SET_CONTEXT_FIELD(b, c, RESUME, TNR, v)
+
+
+/* SCTLR */
+#define SET_M(b, c, v) SET_CONTEXT_FIELD(b, c, SCTLR, M, v)
+#define SET_TRE(b, c, v) SET_CONTEXT_FIELD(b, c, SCTLR, TRE, v)
+#define SET_AFE(b, c, v) SET_CONTEXT_FIELD(b, c, SCTLR, AFE, v)
+#define SET_HAF(b, c, v) SET_CONTEXT_FIELD(b, c, SCTLR, HAF, v)
+#define SET_BE(b, c, v) SET_CONTEXT_FIELD(b, c, SCTLR, BE, v)
+#define SET_AFFD(b, c, v) SET_CONTEXT_FIELD(b, c, SCTLR, AFFD, v)
+
+
+/* TLBLKCR */
+#define SET_LKE(b, c, v) SET_CONTEXT_FIELD(b, c, TLBLKCR, LKE, v)
+#define SET_TLBLKCR_TLBIALLCFG(b, c, v) \
+ SET_CONTEXT_FIELD(b, c, TLBLKCR, TLBLCKR_TLBIALLCFG, v)
+#define SET_TLBIASIDCFG(b, c, v) \
+ SET_CONTEXT_FIELD(b, c, TLBLKCR, TLBIASIDCFG, v)
+#define SET_TLBIVAACFG(b, c, v) SET_CONTEXT_FIELD(b, c, TLBLKCR, TLBIVAACFG, v)
+#define SET_FLOOR(b, c, v) SET_CONTEXT_FIELD(b, c, TLBLKCR, FLOOR, v)
+#define SET_VICTIM(b, c, v) SET_CONTEXT_FIELD(b, c, TLBLKCR, VICTIM, v)
+
+
+/* TTBCR */
+#define SET_N(b, c, v) SET_CONTEXT_FIELD(b, c, TTBCR, N, v)
+#define SET_PD0(b, c, v) SET_CONTEXT_FIELD(b, c, TTBCR, PD0, v)
+#define SET_PD1(b, c, v) SET_CONTEXT_FIELD(b, c, TTBCR, PD1, v)
+
+
+/* TTBR0 */
+#define SET_TTBR0_IRGNH(b, c, v) SET_CONTEXT_FIELD(b, c, TTBR0, TTBR0_IRGNH, v)
+#define SET_TTBR0_SH(b, c, v) SET_CONTEXT_FIELD(b, c, TTBR0, TTBR0_SH, v)
+#define SET_TTBR0_ORGN(b, c, v) SET_CONTEXT_FIELD(b, c, TTBR0, TTBR0_ORGN, v)
+#define SET_TTBR0_NOS(b, c, v) SET_CONTEXT_FIELD(b, c, TTBR0, TTBR0_NOS, v)
+#define SET_TTBR0_IRGNL(b, c, v) SET_CONTEXT_FIELD(b, c, TTBR0, TTBR0_IRGNL, v)
+#define SET_TTBR0_PA(b, c, v) SET_CONTEXT_FIELD(b, c, TTBR0, TTBR0_PA, v)
+
+
+/* TTBR1 */
+#define SET_TTBR1_IRGNH(b, c, v) SET_CONTEXT_FIELD(b, c, TTBR1, TTBR1_IRGNH, v)
+#define SET_TTBR1_SH(b, c, v) SET_CONTEXT_FIELD(b, c, TTBR1, TTBR1_SH, v)
+#define SET_TTBR1_ORGN(b, c, v) SET_CONTEXT_FIELD(b, c, TTBR1, TTBR1_ORGN, v)
+#define SET_TTBR1_NOS(b, c, v) SET_CONTEXT_FIELD(b, c, TTBR1, TTBR1_NOS, v)
+#define SET_TTBR1_IRGNL(b, c, v) SET_CONTEXT_FIELD(b, c, TTBR1, TTBR1_IRGNL, v)
+#define SET_TTBR1_PA(b, c, v) SET_CONTEXT_FIELD(b, c, TTBR1, TTBR1_PA, v)
+
+
+/* V2PSR */
+#define SET_HIT(b, c, v) SET_CONTEXT_FIELD(b, c, V2PSR, HIT, v)
+#define SET_INDEX(b, c, v) SET_CONTEXT_FIELD(b, c, V2PSR, INDEX, v)
+
+
+/* V2Pxx UW UR PW PR */
+#define SET_V2PUW_INDEX(b, c, v) SET_CONTEXT_FIELD(b, c, V2PUW, V2Pxx_INDEX, v)
+#define SET_V2PUW_VA(b, c, v) SET_CONTEXT_FIELD(b, c, V2PUW, V2Pxx_VA, v)
+
+#define SET_V2PUR_INDEX(b, c, v) SET_CONTEXT_FIELD(b, c, V2PUR, V2Pxx_INDEX, v)
+#define SET_V2PUR_VA(b, c, v) SET_CONTEXT_FIELD(b, c, V2PUR, V2Pxx_VA, v)
+
+#define SET_V2PPW_INDEX(b, c, v) SET_CONTEXT_FIELD(b, c, V2PPW, V2Pxx_INDEX, v)
+#define SET_V2PPW_VA(b, c, v) SET_CONTEXT_FIELD(b, c, V2PPW, V2Pxx_VA, v)
+
+#define SET_V2PPR_INDEX(b, c, v) SET_CONTEXT_FIELD(b, c, V2PPR, V2Pxx_INDEX, v)
+#define SET_V2PPR_VA(b, c, v) SET_CONTEXT_FIELD(b, c, V2PPR, V2Pxx_VA, v)
+
+
+/* Context Register getters */
+/* ACTLR */
+#define GET_CFERE(b, c) GET_CONTEXT_FIELD(b, c, ACTLR, CFERE)
+#define GET_CFEIE(b, c) GET_CONTEXT_FIELD(b, c, ACTLR, CFEIE)
+#define GET_PTSHCFG(b, c) GET_CONTEXT_FIELD(b, c, ACTLR, PTSHCFG)
+#define GET_RCOSH(b, c) GET_CONTEXT_FIELD(b, c, ACTLR, RCOSH)
+#define GET_RCISH(b, c) GET_CONTEXT_FIELD(b, c, ACTLR, RCISH)
+#define GET_RCNSH(b, c) GET_CONTEXT_FIELD(b, c, ACTLR, RCNSH)
+#define GET_PRIVCFG(b, c) GET_CONTEXT_FIELD(b, c, ACTLR, PRIVCFG)
+#define GET_DNA(b, c) GET_CONTEXT_FIELD(b, c, ACTLR, DNA)
+#define GET_DNLV2PA(b, c) GET_CONTEXT_FIELD(b, c, ACTLR, DNLV2PA)
+#define GET_TLBMCFG(b, c) GET_CONTEXT_FIELD(b, c, ACTLR, TLBMCFG)
+#define GET_CFCFG(b, c) GET_CONTEXT_FIELD(b, c, ACTLR, CFCFG)
+#define GET_TIPCF(b, c) GET_CONTEXT_FIELD(b, c, ACTLR, TIPCF)
+#define GET_V2PCFG(b, c) GET_CONTEXT_FIELD(b, c, ACTLR, V2PCFG)
+#define GET_HUME(b, c) GET_CONTEXT_FIELD(b, c, ACTLR, HUME)
+#define GET_PTMTCFG(b, c) GET_CONTEXT_FIELD(b, c, ACTLR, PTMTCFG)
+#define GET_PTMEMTYPE(b, c) GET_CONTEXT_FIELD(b, c, ACTLR, PTMEMTYPE)
+
+/* BFBCR */
+#define GET_BFBDFE(b, c) GET_CONTEXT_FIELD(b, c, BFBCR, BFBDFE)
+#define GET_BFBSFE(b, c) GET_CONTEXT_FIELD(b, c, BFBCR, BFBSFE)
+#define GET_SFVS(b, c) GET_CONTEXT_FIELD(b, c, BFBCR, SFVS)
+#define GET_FLVIC(b, c) GET_CONTEXT_FIELD(b, c, BFBCR, FLVIC)
+#define GET_SLVIC(b, c) GET_CONTEXT_FIELD(b, c, BFBCR, SLVIC)
+
+
+/* CONTEXTIDR */
+#define GET_CONTEXTIDR_ASID(b, c) \
+ GET_CONTEXT_FIELD(b, c, CONTEXTIDR, CONTEXTIDR_ASID)
+#define GET_CONTEXTIDR_PROCID(b, c) GET_CONTEXT_FIELD(b, c, CONTEXTIDR, PROCID)
+
+
+/* FSR */
+#define GET_TF(b, c) GET_CONTEXT_FIELD(b, c, FSR, TF)
+#define GET_AFF(b, c) GET_CONTEXT_FIELD(b, c, FSR, AFF)
+#define GET_APF(b, c) GET_CONTEXT_FIELD(b, c, FSR, APF)
+#define GET_TLBMF(b, c) GET_CONTEXT_FIELD(b, c, FSR, TLBMF)
+#define GET_HTWDEEF(b, c) GET_CONTEXT_FIELD(b, c, FSR, HTWDEEF)
+#define GET_HTWSEEF(b, c) GET_CONTEXT_FIELD(b, c, FSR, HTWSEEF)
+#define GET_MHF(b, c) GET_CONTEXT_FIELD(b, c, FSR, MHF)
+#define GET_SL(b, c) GET_CONTEXT_FIELD(b, c, FSR, SL)
+#define GET_SS(b, c) GET_CONTEXT_FIELD(b, c, FSR, SS)
+#define GET_MULTI(b, c) GET_CONTEXT_FIELD(b, c, FSR, MULTI)
+
+
+/* FSYNR0 */
+#define GET_AMID(b, c) GET_CONTEXT_FIELD(b, c, FSYNR0, AMID)
+#define GET_APID(b, c) GET_CONTEXT_FIELD(b, c, FSYNR0, APID)
+#define GET_ABID(b, c) GET_CONTEXT_FIELD(b, c, FSYNR0, ABID)
+#define GET_ATID(b, c) GET_CONTEXT_FIELD(b, c, FSYNR0, ATID)
+
+
+/* FSYNR1 */
+#define GET_AMEMTYPE(b, c) GET_CONTEXT_FIELD(b, c, FSYNR1, AMEMTYPE)
+#define GET_ASHARED(b, c) GET_CONTEXT_FIELD(b, c, FSYNR1, ASHARED)
+#define GET_AINNERSHARED(b, c) GET_CONTEXT_FIELD(b, c, FSYNR1, AINNERSHARED)
+#define GET_APRIV(b, c) GET_CONTEXT_FIELD(b, c, FSYNR1, APRIV)
+#define GET_APROTNS(b, c) GET_CONTEXT_FIELD(b, c, FSYNR1, APROTNS)
+#define GET_AINST(b, c) GET_CONTEXT_FIELD(b, c, FSYNR1, AINST)
+#define GET_AWRITE(b, c) GET_CONTEXT_FIELD(b, c, FSYNR1, AWRITE)
+#define GET_ABURST(b, c) GET_CONTEXT_FIELD(b, c, FSYNR1, ABURST)
+#define GET_ALEN(b, c) GET_CONTEXT_FIELD(b, c, FSYNR1, ALEN)
+#define GET_FSYNR1_ASIZE(b, c) GET_CONTEXT_FIELD(b, c, FSYNR1, FSYNR1_ASIZE)
+#define GET_ALOCK(b, c) GET_CONTEXT_FIELD(b, c, FSYNR1, ALOCK)
+#define GET_AFULL(b, c) GET_CONTEXT_FIELD(b, c, FSYNR1, AFULL)
+
+
+/* NMRR */
+#define GET_ICPC0(b, c) GET_CONTEXT_FIELD(b, c, NMRR, ICPC0)
+#define GET_ICPC1(b, c) GET_CONTEXT_FIELD(b, c, NMRR, ICPC1)
+#define GET_ICPC2(b, c) GET_CONTEXT_FIELD(b, c, NMRR, ICPC2)
+#define GET_ICPC3(b, c) GET_CONTEXT_FIELD(b, c, NMRR, ICPC3)
+#define GET_ICPC4(b, c) GET_CONTEXT_FIELD(b, c, NMRR, ICPC4)
+#define GET_ICPC5(b, c) GET_CONTEXT_FIELD(b, c, NMRR, ICPC5)
+#define GET_ICPC6(b, c) GET_CONTEXT_FIELD(b, c, NMRR, ICPC6)
+#define GET_ICPC7(b, c) GET_CONTEXT_FIELD(b, c, NMRR, ICPC7)
+#define GET_OCPC0(b, c) GET_CONTEXT_FIELD(b, c, NMRR, OCPC0)
+#define GET_OCPC1(b, c) GET_CONTEXT_FIELD(b, c, NMRR, OCPC1)
+#define GET_OCPC2(b, c) GET_CONTEXT_FIELD(b, c, NMRR, OCPC2)
+#define GET_OCPC3(b, c) GET_CONTEXT_FIELD(b, c, NMRR, OCPC3)
+#define GET_OCPC4(b, c) GET_CONTEXT_FIELD(b, c, NMRR, OCPC4)
+#define GET_OCPC5(b, c) GET_CONTEXT_FIELD(b, c, NMRR, OCPC5)
+#define GET_OCPC6(b, c) GET_CONTEXT_FIELD(b, c, NMRR, OCPC6)
+#define GET_OCPC7(b, c) GET_CONTEXT_FIELD(b, c, NMRR, OCPC7)
+
+
+/* PAR */
+#define GET_FAULT(b, c) GET_CONTEXT_FIELD(b, c, PAR, FAULT)
+
+#define GET_FAULT_TF(b, c) GET_CONTEXT_FIELD(b, c, PAR, FAULT_TF)
+#define GET_FAULT_AFF(b, c) GET_CONTEXT_FIELD(b, c, PAR, FAULT_AFF)
+#define GET_FAULT_APF(b, c) GET_CONTEXT_FIELD(b, c, PAR, FAULT_APF)
+#define GET_FAULT_TLBMF(b, c) GET_CONTEXT_FIELD(b, c, PAR, FAULT_TLBMF)
+#define GET_FAULT_HTWDEEF(b, c) GET_CONTEXT_FIELD(b, c, PAR, FAULT_HTWDEEF)
+#define GET_FAULT_HTWSEEF(b, c) GET_CONTEXT_FIELD(b, c, PAR, FAULT_HTWSEEF)
+#define GET_FAULT_MHF(b, c) GET_CONTEXT_FIELD(b, c, PAR, FAULT_MHF)
+#define GET_FAULT_SL(b, c) GET_CONTEXT_FIELD(b, c, PAR, FAULT_SL)
+#define GET_FAULT_SS(b, c) GET_CONTEXT_FIELD(b, c, PAR, FAULT_SS)
+
+#define GET_NOFAULT_SS(b, c) GET_CONTEXT_FIELD(b, c, PAR, PAR_NOFAULT_SS)
+#define GET_NOFAULT_MT(b, c) GET_CONTEXT_FIELD(b, c, PAR, PAR_NOFAULT_MT)
+#define GET_NOFAULT_SH(b, c) GET_CONTEXT_FIELD(b, c, PAR, PAR_NOFAULT_SH)
+#define GET_NOFAULT_NS(b, c) GET_CONTEXT_FIELD(b, c, PAR, PAR_NOFAULT_NS)
+#define GET_NOFAULT_NOS(b, c) GET_CONTEXT_FIELD(b, c, PAR, PAR_NOFAULT_NOS)
+#define GET_NPFAULT_PA(b, c) GET_CONTEXT_FIELD(b, c, PAR, PAR_NPFAULT_PA)
+
+
+/* PRRR */
+#define GET_MTC0(b, c) GET_CONTEXT_FIELD(b, c, PRRR, MTC0)
+#define GET_MTC1(b, c) GET_CONTEXT_FIELD(b, c, PRRR, MTC1)
+#define GET_MTC2(b, c) GET_CONTEXT_FIELD(b, c, PRRR, MTC2)
+#define GET_MTC3(b, c) GET_CONTEXT_FIELD(b, c, PRRR, MTC3)
+#define GET_MTC4(b, c) GET_CONTEXT_FIELD(b, c, PRRR, MTC4)
+#define GET_MTC5(b, c) GET_CONTEXT_FIELD(b, c, PRRR, MTC5)
+#define GET_MTC6(b, c) GET_CONTEXT_FIELD(b, c, PRRR, MTC6)
+#define GET_MTC7(b, c) GET_CONTEXT_FIELD(b, c, PRRR, MTC7)
+#define GET_SHDSH0(b, c) GET_CONTEXT_FIELD(b, c, PRRR, SHDSH0)
+#define GET_SHDSH1(b, c) GET_CONTEXT_FIELD(b, c, PRRR, SHDSH1)
+#define GET_SHNMSH0(b, c) GET_CONTEXT_FIELD(b, c, PRRR, SHNMSH0)
+#define GET_SHNMSH1(b, c) GET_CONTEXT_FIELD(b, c, PRRR, SHNMSH1)
+#define GET_NOS0(b, c) GET_CONTEXT_FIELD(b, c, PRRR, NOS0)
+#define GET_NOS1(b, c) GET_CONTEXT_FIELD(b, c, PRRR, NOS1)
+#define GET_NOS2(b, c) GET_CONTEXT_FIELD(b, c, PRRR, NOS2)
+#define GET_NOS3(b, c) GET_CONTEXT_FIELD(b, c, PRRR, NOS3)
+#define GET_NOS4(b, c) GET_CONTEXT_FIELD(b, c, PRRR, NOS4)
+#define GET_NOS5(b, c) GET_CONTEXT_FIELD(b, c, PRRR, NOS5)
+#define GET_NOS6(b, c) GET_CONTEXT_FIELD(b, c, PRRR, NOS6)
+#define GET_NOS7(b, c) GET_CONTEXT_FIELD(b, c, PRRR, NOS7)
+
+
+/* RESUME */
+#define GET_TNR(b, c) GET_CONTEXT_FIELD(b, c, RESUME, TNR)
+
+
+/* SCTLR */
+#define GET_M(b, c) GET_CONTEXT_FIELD(b, c, SCTLR, M)
+#define GET_TRE(b, c) GET_CONTEXT_FIELD(b, c, SCTLR, TRE)
+#define GET_AFE(b, c) GET_CONTEXT_FIELD(b, c, SCTLR, AFE)
+#define GET_HAF(b, c) GET_CONTEXT_FIELD(b, c, SCTLR, HAF)
+#define GET_BE(b, c) GET_CONTEXT_FIELD(b, c, SCTLR, BE)
+#define GET_AFFD(b, c) GET_CONTEXT_FIELD(b, c, SCTLR, AFFD)
+
+
+/* TLBLKCR */
+#define GET_LKE(b, c) GET_CONTEXT_FIELD(b, c, TLBLKCR, LKE)
+#define GET_TLBLCKR_TLBIALLCFG(b, c) \
+ GET_CONTEXT_FIELD(b, c, TLBLKCR, TLBLCKR_TLBIALLCFG)
+#define GET_TLBIASIDCFG(b, c) GET_CONTEXT_FIELD(b, c, TLBLKCR, TLBIASIDCFG)
+#define GET_TLBIVAACFG(b, c) GET_CONTEXT_FIELD(b, c, TLBLKCR, TLBIVAACFG)
+#define GET_FLOOR(b, c) GET_CONTEXT_FIELD(b, c, TLBLKCR, FLOOR)
+#define GET_VICTIM(b, c) GET_CONTEXT_FIELD(b, c, TLBLKCR, VICTIM)
+
+
+/* TTBCR */
+#define GET_N(b, c) GET_CONTEXT_FIELD(b, c, TTBCR, N)
+#define GET_PD0(b, c) GET_CONTEXT_FIELD(b, c, TTBCR, PD0)
+#define GET_PD1(b, c) GET_CONTEXT_FIELD(b, c, TTBCR, PD1)
+
+
+/* TTBR0 */
+#define GET_TTBR0_IRGNH(b, c) GET_CONTEXT_FIELD(b, c, TTBR0, TTBR0_IRGNH)
+#define GET_TTBR0_SH(b, c) GET_CONTEXT_FIELD(b, c, TTBR0, TTBR0_SH)
+#define GET_TTBR0_ORGN(b, c) GET_CONTEXT_FIELD(b, c, TTBR0, TTBR0_ORGN)
+#define GET_TTBR0_NOS(b, c) GET_CONTEXT_FIELD(b, c, TTBR0, TTBR0_NOS)
+#define GET_TTBR0_IRGNL(b, c) GET_CONTEXT_FIELD(b, c, TTBR0, TTBR0_IRGNL)
+#define GET_TTBR0_PA(b, c) GET_CONTEXT_FIELD(b, c, TTBR0, TTBR0_PA)
+
+
+/* TTBR1 */
+#define GET_TTBR1_IRGNH(b, c) GET_CONTEXT_FIELD(b, c, TTBR1, TTBR1_IRGNH)
+#define GET_TTBR1_SH(b, c) GET_CONTEXT_FIELD(b, c, TTBR1, TTBR1_SH)
+#define GET_TTBR1_ORGN(b, c) GET_CONTEXT_FIELD(b, c, TTBR1, TTBR1_ORGN)
+#define GET_TTBR1_NOS(b, c) GET_CONTEXT_FIELD(b, c, TTBR1, TTBR1_NOS)
+#define GET_TTBR1_IRGNL(b, c) GET_CONTEXT_FIELD(b, c, TTBR1, TTBR1_IRGNL)
+#define GET_TTBR1_PA(b, c) GET_CONTEXT_FIELD(b, c, TTBR1, TTBR1_PA)
+
+
+/* V2PSR */
+#define GET_HIT(b, c) GET_CONTEXT_FIELD(b, c, V2PSR, HIT)
+#define GET_INDEX(b, c) GET_CONTEXT_FIELD(b, c, V2PSR, INDEX)
+
+
+/* V2Pxx UW UR PW PR */
+#define GET_V2PUW_INDEX(b, c) GET_CONTEXT_FIELD(b, c, V2PUW, V2Pxx_INDEX)
+#define GET_V2PUW_VA(b, c) GET_CONTEXT_FIELD(b, c, V2PUW, V2Pxx_VA)
+
+#define GET_V2PUR_INDEX(b, c) GET_CONTEXT_FIELD(b, c, V2PUR, V2Pxx_INDEX)
+#define GET_V2PUR_VA(b, c) GET_CONTEXT_FIELD(b, c, V2PUR, V2Pxx_VA)
+
+#define GET_V2PPW_INDEX(b, c) GET_CONTEXT_FIELD(b, c, V2PPW, V2Pxx_INDEX)
+#define GET_V2PPW_VA(b, c) GET_CONTEXT_FIELD(b, c, V2PPW, V2Pxx_VA)
+
+#define GET_V2PPR_INDEX(b, c) GET_CONTEXT_FIELD(b, c, V2PPR, V2Pxx_INDEX)
+#define GET_V2PPR_VA(b, c) GET_CONTEXT_FIELD(b, c, V2PPR, V2Pxx_VA)
+
+
+/* Global Registers */
+#define M2VCBR_N (0xFF000)
+#define CBACR_N (0xFF800)
+#define TLBRSW (0xFFE00)
+#define TLBTR0 (0xFFE80)
+#define TLBTR1 (0xFFE84)
+#define TLBTR2 (0xFFE88)
+#define TESTBUSCR (0xFFE8C)
+#define GLOBAL_TLBIALL (0xFFF00)
+#define TLBIVMID (0xFFF04)
+#define CR (0xFFF80)
+#define EAR (0xFFF84)
+#define ESR (0xFFF88)
+#define ESRRESTORE (0xFFF8C)
+#define ESYNR0 (0xFFF90)
+#define ESYNR1 (0xFFF94)
+#define REV (0xFFFF4)
+#define IDR (0xFFFF8)
+#define RPU_ACR (0xFFFFC)
+
+
+/* Context Bank Registers */
+#define SCTLR (0x000)
+#define ACTLR (0x004)
+#define CONTEXTIDR (0x008)
+#define TTBR0 (0x010)
+#define TTBR1 (0x014)
+#define TTBCR (0x018)
+#define PAR (0x01C)
+#define FSR (0x020)
+#define FSRRESTORE (0x024)
+#define FAR (0x028)
+#define FSYNR0 (0x02C)
+#define FSYNR1 (0x030)
+#define PRRR (0x034)
+#define NMRR (0x038)
+#define TLBLCKR (0x03C)
+#define V2PSR (0x040)
+#define TLBFLPTER (0x044)
+#define TLBSLPTER (0x048)
+#define BFBCR (0x04C)
+#define CTX_TLBIALL (0x800)
+#define TLBIASID (0x804)
+#define TLBIVA (0x808)
+#define TLBIVAA (0x80C)
+#define V2PPR (0x810)
+#define V2PPW (0x814)
+#define V2PUR (0x818)
+#define V2PUW (0x81C)
+#define RESUME (0x820)
+
+
+/* Global Register Fields */
+/* CBACRn */
+#define RWVMID (RWVMID_MASK << RWVMID_SHIFT)
+#define RWE (RWE_MASK << RWE_SHIFT)
+#define RWGE (RWGE_MASK << RWGE_SHIFT)
+#define CBVMID (CBVMID_MASK << CBVMID_SHIFT)
+#define IRPTNDX (IRPTNDX_MASK << IRPTNDX_SHIFT)
+
+
+/* CR */
+#define RPUE (RPUE_MASK << RPUE_SHIFT)
+#define RPUERE (RPUERE_MASK << RPUERE_SHIFT)
+#define RPUEIE (RPUEIE_MASK << RPUEIE_SHIFT)
+#define DCDEE (DCDEE_MASK << DCDEE_SHIFT)
+#define CLIENTPD (CLIENTPD_MASK << CLIENTPD_SHIFT)
+#define STALLD (STALLD_MASK << STALLD_SHIFT)
+#define TLBLKCRWE (TLBLKCRWE_MASK << TLBLKCRWE_SHIFT)
+#define CR_TLBIALLCFG (CR_TLBIALLCFG_MASK << CR_TLBIALLCFG_SHIFT)
+#define TLBIVMIDCFG (TLBIVMIDCFG_MASK << TLBIVMIDCFG_SHIFT)
+#define CR_HUME (CR_HUME_MASK << CR_HUME_SHIFT)
+
+
+/* ESR */
+#define CFG (CFG_MASK << CFG_SHIFT)
+#define BYPASS (BYPASS_MASK << BYPASS_SHIFT)
+#define ESR_MULTI (ESR_MULTI_MASK << ESR_MULTI_SHIFT)
+
+
+/* ESYNR0 */
+#define ESYNR0_AMID (ESYNR0_AMID_MASK << ESYNR0_AMID_SHIFT)
+#define ESYNR0_APID (ESYNR0_APID_MASK << ESYNR0_APID_SHIFT)
+#define ESYNR0_ABID (ESYNR0_ABID_MASK << ESYNR0_ABID_SHIFT)
+#define ESYNR0_AVMID (ESYNR0_AVMID_MASK << ESYNR0_AVMID_SHIFT)
+#define ESYNR0_ATID (ESYNR0_ATID_MASK << ESYNR0_ATID_SHIFT)
+
+
+/* ESYNR1 */
+#define ESYNR1_AMEMTYPE (ESYNR1_AMEMTYPE_MASK << ESYNR1_AMEMTYPE_SHIFT)
+#define ESYNR1_ASHARED (ESYNR1_ASHARED_MASK << ESYNR1_ASHARED_SHIFT)
+#define ESYNR1_AINNERSHARED (ESYNR1_AINNERSHARED_MASK<< \
+ ESYNR1_AINNERSHARED_SHIFT)
+#define ESYNR1_APRIV (ESYNR1_APRIV_MASK << ESYNR1_APRIV_SHIFT)
+#define ESYNR1_APROTNS (ESYNR1_APROTNS_MASK << ESYNR1_APROTNS_SHIFT)
+#define ESYNR1_AINST (ESYNR1_AINST_MASK << ESYNR1_AINST_SHIFT)
+#define ESYNR1_AWRITE (ESYNR1_AWRITE_MASK << ESYNR1_AWRITE_SHIFT)
+#define ESYNR1_ABURST (ESYNR1_ABURST_MASK << ESYNR1_ABURST_SHIFT)
+#define ESYNR1_ALEN (ESYNR1_ALEN_MASK << ESYNR1_ALEN_SHIFT)
+#define ESYNR1_ASIZE (ESYNR1_ASIZE_MASK << ESYNR1_ASIZE_SHIFT)
+#define ESYNR1_ALOCK (ESYNR1_ALOCK_MASK << ESYNR1_ALOCK_SHIFT)
+#define ESYNR1_AOOO (ESYNR1_AOOO_MASK << ESYNR1_AOOO_SHIFT)
+#define ESYNR1_AFULL (ESYNR1_AFULL_MASK << ESYNR1_AFULL_SHIFT)
+#define ESYNR1_AC (ESYNR1_AC_MASK << ESYNR1_AC_SHIFT)
+#define ESYNR1_DCD (ESYNR1_DCD_MASK << ESYNR1_DCD_SHIFT)
+
+
+/* IDR */
+#define NM2VCBMT (NM2VCBMT_MASK << NM2VCBMT_SHIFT)
+#define HTW (HTW_MASK << HTW_SHIFT)
+#define HUM (HUM_MASK << HUM_SHIFT)
+#define TLBSIZE (TLBSIZE_MASK << TLBSIZE_SHIFT)
+#define NCB (NCB_MASK << NCB_SHIFT)
+#define NIRPT (NIRPT_MASK << NIRPT_SHIFT)
+
+
+/* M2VCBRn */
+#define VMID (VMID_MASK << VMID_SHIFT)
+#define CBNDX (CBNDX_MASK << CBNDX_SHIFT)
+#define BYPASSD (BYPASSD_MASK << BYPASSD_SHIFT)
+#define BPRCOSH (BPRCOSH_MASK << BPRCOSH_SHIFT)
+#define BPRCISH (BPRCISH_MASK << BPRCISH_SHIFT)
+#define BPRCNSH (BPRCNSH_MASK << BPRCNSH_SHIFT)
+#define BPSHCFG (BPSHCFG_MASK << BPSHCFG_SHIFT)
+#define NSCFG (NSCFG_MASK << NSCFG_SHIFT)
+#define BPMTCFG (BPMTCFG_MASK << BPMTCFG_SHIFT)
+#define BPMEMTYPE (BPMEMTYPE_MASK << BPMEMTYPE_SHIFT)
+
+
+/* REV */
+#define IDR_MINOR (MINOR_MASK << MINOR_SHIFT)
+#define IDR_MAJOR (MAJOR_MASK << MAJOR_SHIFT)
+
+
+/* TESTBUSCR */
+#define TBE (TBE_MASK << TBE_SHIFT)
+#define SPDMBE (SPDMBE_MASK << SPDMBE_SHIFT)
+#define WGSEL (WGSEL_MASK << WGSEL_SHIFT)
+#define TBLSEL (TBLSEL_MASK << TBLSEL_SHIFT)
+#define TBHSEL (TBHSEL_MASK << TBHSEL_SHIFT)
+#define SPDM0SEL (SPDM0SEL_MASK << SPDM0SEL_SHIFT)
+#define SPDM1SEL (SPDM1SEL_MASK << SPDM1SEL_SHIFT)
+#define SPDM2SEL (SPDM2SEL_MASK << SPDM2SEL_SHIFT)
+#define SPDM3SEL (SPDM3SEL_MASK << SPDM3SEL_SHIFT)
+
+
+/* TLBIVMID */
+#define TLBIVMID_VMID (TLBIVMID_VMID_MASK << TLBIVMID_VMID_SHIFT)
+
+
+/* TLBRSW */
+#define TLBRSW_INDEX (TLBRSW_INDEX_MASK << TLBRSW_INDEX_SHIFT)
+#define TLBBFBS (TLBBFBS_MASK << TLBBFBS_SHIFT)
+
+
+/* TLBTR0 */
+#define PR (PR_MASK << PR_SHIFT)
+#define PW (PW_MASK << PW_SHIFT)
+#define UR (UR_MASK << UR_SHIFT)
+#define UW (UW_MASK << UW_SHIFT)
+#define XN (XN_MASK << XN_SHIFT)
+#define NSDESC (NSDESC_MASK << NSDESC_SHIFT)
+#define ISH (ISH_MASK << ISH_SHIFT)
+#define SH (SH_MASK << SH_SHIFT)
+#define MT (MT_MASK << MT_SHIFT)
+#define DPSIZR (DPSIZR_MASK << DPSIZR_SHIFT)
+#define DPSIZC (DPSIZC_MASK << DPSIZC_SHIFT)
+
+
+/* TLBTR1 */
+#define TLBTR1_VMID (TLBTR1_VMID_MASK << TLBTR1_VMID_SHIFT)
+#define TLBTR1_PA (TLBTR1_PA_MASK << TLBTR1_PA_SHIFT)
+
+
+/* TLBTR2 */
+#define TLBTR2_ASID (TLBTR2_ASID_MASK << TLBTR2_ASID_SHIFT)
+#define TLBTR2_V (TLBTR2_V_MASK << TLBTR2_V_SHIFT)
+#define TLBTR2_NSTID (TLBTR2_NSTID_MASK << TLBTR2_NSTID_SHIFT)
+#define TLBTR2_NV (TLBTR2_NV_MASK << TLBTR2_NV_SHIFT)
+#define TLBTR2_VA (TLBTR2_VA_MASK << TLBTR2_VA_SHIFT)
+
+
+/* Context Register Fields */
+/* ACTLR */
+#define CFERE (CFERE_MASK << CFERE_SHIFT)
+#define CFEIE (CFEIE_MASK << CFEIE_SHIFT)
+#define PTSHCFG (PTSHCFG_MASK << PTSHCFG_SHIFT)
+#define RCOSH (RCOSH_MASK << RCOSH_SHIFT)
+#define RCISH (RCISH_MASK << RCISH_SHIFT)
+#define RCNSH (RCNSH_MASK << RCNSH_SHIFT)
+#define PRIVCFG (PRIVCFG_MASK << PRIVCFG_SHIFT)
+#define DNA (DNA_MASK << DNA_SHIFT)
+#define DNLV2PA (DNLV2PA_MASK << DNLV2PA_SHIFT)
+#define TLBMCFG (TLBMCFG_MASK << TLBMCFG_SHIFT)
+#define CFCFG (CFCFG_MASK << CFCFG_SHIFT)
+#define TIPCF (TIPCF_MASK << TIPCF_SHIFT)
+#define V2PCFG (V2PCFG_MASK << V2PCFG_SHIFT)
+#define HUME (HUME_MASK << HUME_SHIFT)
+#define PTMTCFG (PTMTCFG_MASK << PTMTCFG_SHIFT)
+#define PTMEMTYPE (PTMEMTYPE_MASK << PTMEMTYPE_SHIFT)
+
+
+/* BFBCR */
+#define BFBDFE (BFBDFE_MASK << BFBDFE_SHIFT)
+#define BFBSFE (BFBSFE_MASK << BFBSFE_SHIFT)
+#define SFVS (SFVS_MASK << SFVS_SHIFT)
+#define FLVIC (FLVIC_MASK << FLVIC_SHIFT)
+#define SLVIC (SLVIC_MASK << SLVIC_SHIFT)
+
+
+/* CONTEXTIDR */
+#define CONTEXTIDR_ASID (CONTEXTIDR_ASID_MASK << CONTEXTIDR_ASID_SHIFT)
+#define PROCID (PROCID_MASK << PROCID_SHIFT)
+
+
+/* FSR */
+#define TF (TF_MASK << TF_SHIFT)
+#define AFF (AFF_MASK << AFF_SHIFT)
+#define APF (APF_MASK << APF_SHIFT)
+#define TLBMF (TLBMF_MASK << TLBMF_SHIFT)
+#define HTWDEEF (HTWDEEF_MASK << HTWDEEF_SHIFT)
+#define HTWSEEF (HTWSEEF_MASK << HTWSEEF_SHIFT)
+#define MHF (MHF_MASK << MHF_SHIFT)
+#define SL (SL_MASK << SL_SHIFT)
+#define SS (SS_MASK << SS_SHIFT)
+#define MULTI (MULTI_MASK << MULTI_SHIFT)
+
+
+/* FSYNR0 */
+#define AMID (AMID_MASK << AMID_SHIFT)
+#define APID (APID_MASK << APID_SHIFT)
+#define ABID (ABID_MASK << ABID_SHIFT)
+#define ATID (ATID_MASK << ATID_SHIFT)
+
+
+/* FSYNR1 */
+#define AMEMTYPE (AMEMTYPE_MASK << AMEMTYPE_SHIFT)
+#define ASHARED (ASHARED_MASK << ASHARED_SHIFT)
+#define AINNERSHARED (AINNERSHARED_MASK << AINNERSHARED_SHIFT)
+#define APRIV (APRIV_MASK << APRIV_SHIFT)
+#define APROTNS (APROTNS_MASK << APROTNS_SHIFT)
+#define AINST (AINST_MASK << AINST_SHIFT)
+#define AWRITE (AWRITE_MASK << AWRITE_SHIFT)
+#define ABURST (ABURST_MASK << ABURST_SHIFT)
+#define ALEN (ALEN_MASK << ALEN_SHIFT)
+#define FSYNR1_ASIZE (FSYNR1_ASIZE_MASK << FSYNR1_ASIZE_SHIFT)
+#define ALOCK (ALOCK_MASK << ALOCK_SHIFT)
+#define AFULL (AFULL_MASK << AFULL_SHIFT)
+
+
+/* NMRR */
+#define ICPC0 (ICPC0_MASK << ICPC0_SHIFT)
+#define ICPC1 (ICPC1_MASK << ICPC1_SHIFT)
+#define ICPC2 (ICPC2_MASK << ICPC2_SHIFT)
+#define ICPC3 (ICPC3_MASK << ICPC3_SHIFT)
+#define ICPC4 (ICPC4_MASK << ICPC4_SHIFT)
+#define ICPC5 (ICPC5_MASK << ICPC5_SHIFT)
+#define ICPC6 (ICPC6_MASK << ICPC6_SHIFT)
+#define ICPC7 (ICPC7_MASK << ICPC7_SHIFT)
+#define OCPC0 (OCPC0_MASK << OCPC0_SHIFT)
+#define OCPC1 (OCPC1_MASK << OCPC1_SHIFT)
+#define OCPC2 (OCPC2_MASK << OCPC2_SHIFT)
+#define OCPC3 (OCPC3_MASK << OCPC3_SHIFT)
+#define OCPC4 (OCPC4_MASK << OCPC4_SHIFT)
+#define OCPC5 (OCPC5_MASK << OCPC5_SHIFT)
+#define OCPC6 (OCPC6_MASK << OCPC6_SHIFT)
+#define OCPC7 (OCPC7_MASK << OCPC7_SHIFT)
+
+
+/* PAR */
+#define FAULT (FAULT_MASK << FAULT_SHIFT)
+/* If a fault is present, these are the
+same as the fault fields in the FAR */
+#define FAULT_TF (FAULT_TF_MASK << FAULT_TF_SHIFT)
+#define FAULT_AFF (FAULT_AFF_MASK << FAULT_AFF_SHIFT)
+#define FAULT_APF (FAULT_APF_MASK << FAULT_APF_SHIFT)
+#define FAULT_TLBMF (FAULT_TLBMF_MASK << FAULT_TLBMF_SHIFT)
+#define FAULT_HTWDEEF (FAULT_HTWDEEF_MASK << FAULT_HTWDEEF_SHIFT)
+#define FAULT_HTWSEEF (FAULT_HTWSEEF_MASK << FAULT_HTWSEEF_SHIFT)
+#define FAULT_MHF (FAULT_MHF_MASK << FAULT_MHF_SHIFT)
+#define FAULT_SL (FAULT_SL_MASK << FAULT_SL_SHIFT)
+#define FAULT_SS (FAULT_SS_MASK << FAULT_SS_SHIFT)
+
+/* If NO fault is present, the following fields are in effect */
+/* (FAULT remains as before) */
+#define PAR_NOFAULT_SS (PAR_NOFAULT_SS_MASK << PAR_NOFAULT_SS_SHIFT)
+#define PAR_NOFAULT_MT (PAR_NOFAULT_MT_MASK << PAR_NOFAULT_MT_SHIFT)
+#define PAR_NOFAULT_SH (PAR_NOFAULT_SH_MASK << PAR_NOFAULT_SH_SHIFT)
+#define PAR_NOFAULT_NS (PAR_NOFAULT_NS_MASK << PAR_NOFAULT_NS_SHIFT)
+#define PAR_NOFAULT_NOS (PAR_NOFAULT_NOS_MASK << PAR_NOFAULT_NOS_SHIFT)
+#define PAR_NPFAULT_PA (PAR_NPFAULT_PA_MASK << PAR_NPFAULT_PA_SHIFT)
+
+
+/* PRRR */
+#define MTC0 (MTC0_MASK << MTC0_SHIFT)
+#define MTC1 (MTC1_MASK << MTC1_SHIFT)
+#define MTC2 (MTC2_MASK << MTC2_SHIFT)
+#define MTC3 (MTC3_MASK << MTC3_SHIFT)
+#define MTC4 (MTC4_MASK << MTC4_SHIFT)
+#define MTC5 (MTC5_MASK << MTC5_SHIFT)
+#define MTC6 (MTC6_MASK << MTC6_SHIFT)
+#define MTC7 (MTC7_MASK << MTC7_SHIFT)
+#define SHDSH0 (SHDSH0_MASK << SHDSH0_SHIFT)
+#define SHDSH1 (SHDSH1_MASK << SHDSH1_SHIFT)
+#define SHNMSH0 (SHNMSH0_MASK << SHNMSH0_SHIFT)
+#define SHNMSH1 (SHNMSH1_MASK << SHNMSH1_SHIFT)
+#define NOS0 (NOS0_MASK << NOS0_SHIFT)
+#define NOS1 (NOS1_MASK << NOS1_SHIFT)
+#define NOS2 (NOS2_MASK << NOS2_SHIFT)
+#define NOS3 (NOS3_MASK << NOS3_SHIFT)
+#define NOS4 (NOS4_MASK << NOS4_SHIFT)
+#define NOS5 (NOS5_MASK << NOS5_SHIFT)
+#define NOS6 (NOS6_MASK << NOS6_SHIFT)
+#define NOS7 (NOS7_MASK << NOS7_SHIFT)
+
+
+/* RESUME */
+#define TNR (TNR_MASK << TNR_SHIFT)
+
+
+/* SCTLR */
+#define M (M_MASK << M_SHIFT)
+#define TRE (TRE_MASK << TRE_SHIFT)
+#define AFE (AFE_MASK << AFE_SHIFT)
+#define HAF (HAF_MASK << HAF_SHIFT)
+#define BE (BE_MASK << BE_SHIFT)
+#define AFFD (AFFD_MASK << AFFD_SHIFT)
+
+
+/* TLBIASID */
+#define TLBIASID_ASID (TLBIASID_ASID_MASK << TLBIASID_ASID_SHIFT)
+
+
+/* TLBIVA */
+#define TLBIVA_ASID (TLBIVA_ASID_MASK << TLBIVA_ASID_SHIFT)
+#define TLBIVA_VA (TLBIVA_VA_MASK << TLBIVA_VA_SHIFT)
+
+
+/* TLBIVAA */
+#define TLBIVAA_VA (TLBIVAA_VA_MASK << TLBIVAA_VA_SHIFT)
+
+
+/* TLBLCKR */
+#define LKE (LKE_MASK << LKE_SHIFT)
+#define TLBLCKR_TLBIALLCFG (TLBLCKR_TLBIALLCFG_MASK<<TLBLCKR_TLBIALLCFG_SHIFT)
+#define TLBIASIDCFG (TLBIASIDCFG_MASK << TLBIASIDCFG_SHIFT)
+#define TLBIVAACFG (TLBIVAACFG_MASK << TLBIVAACFG_SHIFT)
+#define FLOOR (FLOOR_MASK << FLOOR_SHIFT)
+#define VICTIM (VICTIM_MASK << VICTIM_SHIFT)
+
+
+/* TTBCR */
+#define N (N_MASK << N_SHIFT)
+#define PD0 (PD0_MASK << PD0_SHIFT)
+#define PD1 (PD1_MASK << PD1_SHIFT)
+
+
+/* TTBR0 */
+#define TTBR0_IRGNH (TTBR0_IRGNH_MASK << TTBR0_IRGNH_SHIFT)
+#define TTBR0_SH (TTBR0_SH_MASK << TTBR0_SH_SHIFT)
+#define TTBR0_ORGN (TTBR0_ORGN_MASK << TTBR0_ORGN_SHIFT)
+#define TTBR0_NOS (TTBR0_NOS_MASK << TTBR0_NOS_SHIFT)
+#define TTBR0_IRGNL (TTBR0_IRGNL_MASK << TTBR0_IRGNL_SHIFT)
+#define TTBR0_PA (TTBR0_PA_MASK << TTBR0_PA_SHIFT)
+
+
+/* TTBR1 */
+#define TTBR1_IRGNH (TTBR1_IRGNH_MASK << TTBR1_IRGNH_SHIFT)
+#define TTBR1_SH (TTBR1_SH_MASK << TTBR1_SH_SHIFT)
+#define TTBR1_ORGN (TTBR1_ORGN_MASK << TTBR1_ORGN_SHIFT)
+#define TTBR1_NOS (TTBR1_NOS_MASK << TTBR1_NOS_SHIFT)
+#define TTBR1_IRGNL (TTBR1_IRGNL_MASK << TTBR1_IRGNL_SHIFT)
+#define TTBR1_PA (TTBR1_PA_MASK << TTBR1_PA_SHIFT)
+
+
+/* V2PSR */
+#define HIT (HIT_MASK << HIT_SHIFT)
+#define INDEX (INDEX_MASK << INDEX_SHIFT)
+
+
+/* V2Pxx */
+#define V2Pxx_INDEX (V2Pxx_INDEX_MASK << V2Pxx_INDEX_SHIFT)
+#define V2Pxx_VA (V2Pxx_VA_MASK << V2Pxx_VA_SHIFT)
+
+
+/* Global Register Masks */
+/* CBACRn */
+#define RWVMID_MASK 0x1F
+#define RWE_MASK 0x01
+#define RWGE_MASK 0x01
+#define CBVMID_MASK 0x1F
+#define IRPTNDX_MASK 0xFF
+
+
+/* CR */
+#define RPUE_MASK 0x01
+#define RPUERE_MASK 0x01
+#define RPUEIE_MASK 0x01
+#define DCDEE_MASK 0x01
+#define CLIENTPD_MASK 0x01
+#define STALLD_MASK 0x01
+#define TLBLKCRWE_MASK 0x01
+#define CR_TLBIALLCFG_MASK 0x01
+#define TLBIVMIDCFG_MASK 0x01
+#define CR_HUME_MASK 0x01
+
+
+/* ESR */
+#define CFG_MASK 0x01
+#define BYPASS_MASK 0x01
+#define ESR_MULTI_MASK 0x01
+
+
+/* ESYNR0 */
+#define ESYNR0_AMID_MASK 0xFF
+#define ESYNR0_APID_MASK 0x1F
+#define ESYNR0_ABID_MASK 0x07
+#define ESYNR0_AVMID_MASK 0x1F
+#define ESYNR0_ATID_MASK 0xFF
+
+
+/* ESYNR1 */
+#define ESYNR1_AMEMTYPE_MASK 0x07
+#define ESYNR1_ASHARED_MASK 0x01
+#define ESYNR1_AINNERSHARED_MASK 0x01
+#define ESYNR1_APRIV_MASK 0x01
+#define ESYNR1_APROTNS_MASK 0x01
+#define ESYNR1_AINST_MASK 0x01
+#define ESYNR1_AWRITE_MASK 0x01
+#define ESYNR1_ABURST_MASK 0x01
+#define ESYNR1_ALEN_MASK 0x0F
+#define ESYNR1_ASIZE_MASK 0x01
+#define ESYNR1_ALOCK_MASK 0x03
+#define ESYNR1_AOOO_MASK 0x01
+#define ESYNR1_AFULL_MASK 0x01
+#define ESYNR1_AC_MASK 0x01
+#define ESYNR1_DCD_MASK 0x01
+
+
+/* IDR */
+#define NM2VCBMT_MASK 0x1FF
+#define HTW_MASK 0x01
+#define HUM_MASK 0x01
+#define TLBSIZE_MASK 0x0F
+#define NCB_MASK 0xFF
+#define NIRPT_MASK 0xFF
+
+
+/* M2VCBRn */
+#define VMID_MASK 0x1F
+#define CBNDX_MASK 0xFF
+#define BYPASSD_MASK 0x01
+#define BPRCOSH_MASK 0x01
+#define BPRCISH_MASK 0x01
+#define BPRCNSH_MASK 0x01
+#define BPSHCFG_MASK 0x03
+#define NSCFG_MASK 0x03
+#define BPMTCFG_MASK 0x01
+#define BPMEMTYPE_MASK 0x07
+
+
+/* REV */
+#define MINOR_MASK 0x0F
+#define MAJOR_MASK 0x0F
+
+
+/* TESTBUSCR */
+#define TBE_MASK 0x01
+#define SPDMBE_MASK 0x01
+#define WGSEL_MASK 0x03
+#define TBLSEL_MASK 0x03
+#define TBHSEL_MASK 0x03
+#define SPDM0SEL_MASK 0x0F
+#define SPDM1SEL_MASK 0x0F
+#define SPDM2SEL_MASK 0x0F
+#define SPDM3SEL_MASK 0x0F
+
+
+/* TLBIMID */
+#define TLBIVMID_VMID_MASK 0x1F
+
+
+/* TLBRSW */
+#define TLBRSW_INDEX_MASK 0xFF
+#define TLBBFBS_MASK 0x03
+
+
+/* TLBTR0 */
+#define PR_MASK 0x01
+#define PW_MASK 0x01
+#define UR_MASK 0x01
+#define UW_MASK 0x01
+#define XN_MASK 0x01
+#define NSDESC_MASK 0x01
+#define ISH_MASK 0x01
+#define SH_MASK 0x01
+#define MT_MASK 0x07
+#define DPSIZR_MASK 0x07
+#define DPSIZC_MASK 0x07
+
+
+/* TLBTR1 */
+#define TLBTR1_VMID_MASK 0x1F
+#define TLBTR1_PA_MASK 0x000FFFFF
+
+
+/* TLBTR2 */
+#define TLBTR2_ASID_MASK 0xFF
+#define TLBTR2_V_MASK 0x01
+#define TLBTR2_NSTID_MASK 0x01
+#define TLBTR2_NV_MASK 0x01
+#define TLBTR2_VA_MASK 0x000FFFFF
+
+
+/* Global Register Shifts */
+/* CBACRn */
+#define RWVMID_SHIFT 0
+#define RWE_SHIFT 8
+#define RWGE_SHIFT 9
+#define CBVMID_SHIFT 16
+#define IRPTNDX_SHIFT 24
+
+
+/* CR */
+#define RPUE_SHIFT 0
+#define RPUERE_SHIFT 1
+#define RPUEIE_SHIFT 2
+#define DCDEE_SHIFT 3
+#define CLIENTPD_SHIFT 4
+#define STALLD_SHIFT 5
+#define TLBLKCRWE_SHIFT 6
+#define CR_TLBIALLCFG_SHIFT 7
+#define TLBIVMIDCFG_SHIFT 8
+#define CR_HUME_SHIFT 9
+
+
+/* ESR */
+#define CFG_SHIFT 0
+#define BYPASS_SHIFT 1
+#define ESR_MULTI_SHIFT 31
+
+
+/* ESYNR0 */
+#define ESYNR0_AMID_SHIFT 0
+#define ESYNR0_APID_SHIFT 8
+#define ESYNR0_ABID_SHIFT 13
+#define ESYNR0_AVMID_SHIFT 16
+#define ESYNR0_ATID_SHIFT 24
+
+
+/* ESYNR1 */
+#define ESYNR1_AMEMTYPE_SHIFT 0
+#define ESYNR1_ASHARED_SHIFT 3
+#define ESYNR1_AINNERSHARED_SHIFT 4
+#define ESYNR1_APRIV_SHIFT 5
+#define ESYNR1_APROTNS_SHIFT 6
+#define ESYNR1_AINST_SHIFT 7
+#define ESYNR1_AWRITE_SHIFT 8
+#define ESYNR1_ABURST_SHIFT 10
+#define ESYNR1_ALEN_SHIFT 12
+#define ESYNR1_ASIZE_SHIFT 16
+#define ESYNR1_ALOCK_SHIFT 20
+#define ESYNR1_AOOO_SHIFT 22
+#define ESYNR1_AFULL_SHIFT 24
+#define ESYNR1_AC_SHIFT 30
+#define ESYNR1_DCD_SHIFT 31
+
+
+/* IDR */
+#define NM2VCBMT_SHIFT 0
+#define HTW_SHIFT 9
+#define HUM_SHIFT 10
+#define TLBSIZE_SHIFT 12
+#define NCB_SHIFT 16
+#define NIRPT_SHIFT 24
+
+
+/* M2VCBRn */
+#define VMID_SHIFT 0
+#define CBNDX_SHIFT 8
+#define BYPASSD_SHIFT 16
+#define BPRCOSH_SHIFT 17
+#define BPRCISH_SHIFT 18
+#define BPRCNSH_SHIFT 19
+#define BPSHCFG_SHIFT 20
+#define NSCFG_SHIFT 22
+#define BPMTCFG_SHIFT 24
+#define BPMEMTYPE_SHIFT 25
+
+
+/* REV */
+#define MINOR_SHIFT 0
+#define MAJOR_SHIFT 4
+
+
+/* TESTBUSCR */
+#define TBE_SHIFT 0
+#define SPDMBE_SHIFT 1
+#define WGSEL_SHIFT 8
+#define TBLSEL_SHIFT 12
+#define TBHSEL_SHIFT 14
+#define SPDM0SEL_SHIFT 16
+#define SPDM1SEL_SHIFT 20
+#define SPDM2SEL_SHIFT 24
+#define SPDM3SEL_SHIFT 28
+
+
+/* TLBIMID */
+#define TLBIVMID_VMID_SHIFT 0
+
+
+/* TLBRSW */
+#define TLBRSW_INDEX_SHIFT 0
+#define TLBBFBS_SHIFT 8
+
+
+/* TLBTR0 */
+#define PR_SHIFT 0
+#define PW_SHIFT 1
+#define UR_SHIFT 2
+#define UW_SHIFT 3
+#define XN_SHIFT 4
+#define NSDESC_SHIFT 6
+#define ISH_SHIFT 7
+#define SH_SHIFT 8
+#define MT_SHIFT 9
+#define DPSIZR_SHIFT 16
+#define DPSIZC_SHIFT 20
+
+
+/* TLBTR1 */
+#define TLBTR1_VMID_SHIFT 0
+#define TLBTR1_PA_SHIFT 12
+
+
+/* TLBTR2 */
+#define TLBTR2_ASID_SHIFT 0
+#define TLBTR2_V_SHIFT 8
+#define TLBTR2_NSTID_SHIFT 9
+#define TLBTR2_NV_SHIFT 10
+#define TLBTR2_VA_SHIFT 12
+
+
+/* Context Register Masks */
+/* ACTLR */
+#define CFERE_MASK 0x01
+#define CFEIE_MASK 0x01
+#define PTSHCFG_MASK 0x03
+#define RCOSH_MASK 0x01
+#define RCISH_MASK 0x01
+#define RCNSH_MASK 0x01
+#define PRIVCFG_MASK 0x03
+#define DNA_MASK 0x01
+#define DNLV2PA_MASK 0x01
+#define TLBMCFG_MASK 0x03
+#define CFCFG_MASK 0x01
+#define TIPCF_MASK 0x01
+#define V2PCFG_MASK 0x03
+#define HUME_MASK 0x01
+#define PTMTCFG_MASK 0x01
+#define PTMEMTYPE_MASK 0x07
+
+
+/* BFBCR */
+#define BFBDFE_MASK 0x01
+#define BFBSFE_MASK 0x01
+#define SFVS_MASK 0x01
+#define FLVIC_MASK 0x0F
+#define SLVIC_MASK 0x0F
+
+
+/* CONTEXTIDR */
+#define CONTEXTIDR_ASID_MASK 0xFF
+#define PROCID_MASK 0x00FFFFFF
+
+
+/* FSR */
+#define TF_MASK 0x01
+#define AFF_MASK 0x01
+#define APF_MASK 0x01
+#define TLBMF_MASK 0x01
+#define HTWDEEF_MASK 0x01
+#define HTWSEEF_MASK 0x01
+#define MHF_MASK 0x01
+#define SL_MASK 0x01
+#define SS_MASK 0x01
+#define MULTI_MASK 0x01
+
+
+/* FSYNR0 */
+#define AMID_MASK 0xFF
+#define APID_MASK 0x1F
+#define ABID_MASK 0x07
+#define ATID_MASK 0xFF
+
+
+/* FSYNR1 */
+#define AMEMTYPE_MASK 0x07
+#define ASHARED_MASK 0x01
+#define AINNERSHARED_MASK 0x01
+#define APRIV_MASK 0x01
+#define APROTNS_MASK 0x01
+#define AINST_MASK 0x01
+#define AWRITE_MASK 0x01
+#define ABURST_MASK 0x01
+#define ALEN_MASK 0x0F
+#define FSYNR1_ASIZE_MASK 0x07
+#define ALOCK_MASK 0x03
+#define AFULL_MASK 0x01
+
+
+/* NMRR */
+#define ICPC0_MASK 0x03
+#define ICPC1_MASK 0x03
+#define ICPC2_MASK 0x03
+#define ICPC3_MASK 0x03
+#define ICPC4_MASK 0x03
+#define ICPC5_MASK 0x03
+#define ICPC6_MASK 0x03
+#define ICPC7_MASK 0x03
+#define OCPC0_MASK 0x03
+#define OCPC1_MASK 0x03
+#define OCPC2_MASK 0x03
+#define OCPC3_MASK 0x03
+#define OCPC4_MASK 0x03
+#define OCPC5_MASK 0x03
+#define OCPC6_MASK 0x03
+#define OCPC7_MASK 0x03
+
+
+/* PAR */
+#define FAULT_MASK 0x01
+/* If a fault is present, these are the
+same as the fault fields in the FAR */
+#define FAULT_TF_MASK 0x01
+#define FAULT_AFF_MASK 0x01
+#define FAULT_APF_MASK 0x01
+#define FAULT_TLBMF_MASK 0x01
+#define FAULT_HTWDEEF_MASK 0x01
+#define FAULT_HTWSEEF_MASK 0x01
+#define FAULT_MHF_MASK 0x01
+#define FAULT_SL_MASK 0x01
+#define FAULT_SS_MASK 0x01
+
+/* If NO fault is present, the following
+ fields are in effect */
+/* (FAULT remains as before) */
+#define PAR_NOFAULT_SS_MASK 0x01
+#define PAR_NOFAULT_MT_MASK 0x07
+#define PAR_NOFAULT_SH_MASK 0x01
+#define PAR_NOFAULT_NS_MASK 0x01
+#define PAR_NOFAULT_NOS_MASK 0x01
+#define PAR_NPFAULT_PA_MASK 0x000FFFFF
+
+
+/* PRRR */
+#define MTC0_MASK 0x03
+#define MTC1_MASK 0x03
+#define MTC2_MASK 0x03
+#define MTC3_MASK 0x03
+#define MTC4_MASK 0x03
+#define MTC5_MASK 0x03
+#define MTC6_MASK 0x03
+#define MTC7_MASK 0x03
+#define SHDSH0_MASK 0x01
+#define SHDSH1_MASK 0x01
+#define SHNMSH0_MASK 0x01
+#define SHNMSH1_MASK 0x01
+#define NOS0_MASK 0x01
+#define NOS1_MASK 0x01
+#define NOS2_MASK 0x01
+#define NOS3_MASK 0x01
+#define NOS4_MASK 0x01
+#define NOS5_MASK 0x01
+#define NOS6_MASK 0x01
+#define NOS7_MASK 0x01
+
+
+/* RESUME */
+#define TNR_MASK 0x01
+
+
+/* SCTLR */
+#define M_MASK 0x01
+#define TRE_MASK 0x01
+#define AFE_MASK 0x01
+#define HAF_MASK 0x01
+#define BE_MASK 0x01
+#define AFFD_MASK 0x01
+
+
+/* TLBIASID */
+#define TLBIASID_ASID_MASK 0xFF
+
+
+/* TLBIVA */
+#define TLBIVA_ASID_MASK 0xFF
+#define TLBIVA_VA_MASK 0x000FFFFF
+
+
+/* TLBIVAA */
+#define TLBIVAA_VA_MASK 0x000FFFFF
+
+
+/* TLBLCKR */
+#define LKE_MASK 0x01
+#define TLBLCKR_TLBIALLCFG_MASK 0x01
+#define TLBIASIDCFG_MASK 0x01
+#define TLBIVAACFG_MASK 0x01
+#define FLOOR_MASK 0xFF
+#define VICTIM_MASK 0xFF
+
+
+/* TTBCR */
+#define N_MASK 0x07
+#define PD0_MASK 0x01
+#define PD1_MASK 0x01
+
+
+/* TTBR0 */
+#define TTBR0_IRGNH_MASK 0x01
+#define TTBR0_SH_MASK 0x01
+#define TTBR0_ORGN_MASK 0x03
+#define TTBR0_NOS_MASK 0x01
+#define TTBR0_IRGNL_MASK 0x01
+#define TTBR0_PA_MASK 0x0003FFFF
+
+
+/* TTBR1 */
+#define TTBR1_IRGNH_MASK 0x01
+#define TTBR1_SH_MASK 0x01
+#define TTBR1_ORGN_MASK 0x03
+#define TTBR1_NOS_MASK 0x01
+#define TTBR1_IRGNL_MASK 0x01
+#define TTBR1_PA_MASK 0x0003FFFF
+
+
+/* V2PSR */
+#define HIT_MASK 0x01
+#define INDEX_MASK 0xFF
+
+
+/* V2Pxx */
+#define V2Pxx_INDEX_MASK 0xFF
+#define V2Pxx_VA_MASK 0x000FFFFF
+
+
+/* Context Register Shifts */
+/* ACTLR */
+#define CFERE_SHIFT 0
+#define CFEIE_SHIFT 1
+#define PTSHCFG_SHIFT 2
+#define RCOSH_SHIFT 4
+#define RCISH_SHIFT 5
+#define RCNSH_SHIFT 6
+#define PRIVCFG_SHIFT 8
+#define DNA_SHIFT 10
+#define DNLV2PA_SHIFT 11
+#define TLBMCFG_SHIFT 12
+#define CFCFG_SHIFT 14
+#define TIPCF_SHIFT 15
+#define V2PCFG_SHIFT 16
+#define HUME_SHIFT 18
+#define PTMTCFG_SHIFT 20
+#define PTMEMTYPE_SHIFT 21
+
+
+/* BFBCR */
+#define BFBDFE_SHIFT 0
+#define BFBSFE_SHIFT 1
+#define SFVS_SHIFT 2
+#define FLVIC_SHIFT 4
+#define SLVIC_SHIFT 8
+
+
+/* CONTEXTIDR */
+#define CONTEXTIDR_ASID_SHIFT 0
+#define PROCID_SHIFT 8
+
+
+/* FSR */
+#define TF_SHIFT 1
+#define AFF_SHIFT 2
+#define APF_SHIFT 3
+#define TLBMF_SHIFT 4
+#define HTWDEEF_SHIFT 5
+#define HTWSEEF_SHIFT 6
+#define MHF_SHIFT 7
+#define SL_SHIFT 16
+#define SS_SHIFT 30
+#define MULTI_SHIFT 31
+
+
+/* FSYNR0 */
+#define AMID_SHIFT 0
+#define APID_SHIFT 8
+#define ABID_SHIFT 13
+#define ATID_SHIFT 24
+
+
+/* FSYNR1 */
+#define AMEMTYPE_SHIFT 0
+#define ASHARED_SHIFT 3
+#define AINNERSHARED_SHIFT 4
+#define APRIV_SHIFT 5
+#define APROTNS_SHIFT 6
+#define AINST_SHIFT 7
+#define AWRITE_SHIFT 8
+#define ABURST_SHIFT 10
+#define ALEN_SHIFT 12
+#define FSYNR1_ASIZE_SHIFT 16
+#define ALOCK_SHIFT 20
+#define AFULL_SHIFT 24
+
+
+/* NMRR */
+#define ICPC0_SHIFT 0
+#define ICPC1_SHIFT 2
+#define ICPC2_SHIFT 4
+#define ICPC3_SHIFT 6
+#define ICPC4_SHIFT 8
+#define ICPC5_SHIFT 10
+#define ICPC6_SHIFT 12
+#define ICPC7_SHIFT 14
+#define OCPC0_SHIFT 16
+#define OCPC1_SHIFT 18
+#define OCPC2_SHIFT 20
+#define OCPC3_SHIFT 22
+#define OCPC4_SHIFT 24
+#define OCPC5_SHIFT 26
+#define OCPC6_SHIFT 28
+#define OCPC7_SHIFT 30
+
+
+/* PAR */
+#define FAULT_SHIFT 0
+/* If a fault is present, these are the
+same as the fault fields in the FAR */
+#define FAULT_TF_SHIFT 1
+#define FAULT_AFF_SHIFT 2
+#define FAULT_APF_SHIFT 3
+#define FAULT_TLBMF_SHIFT 4
+#define FAULT_HTWDEEF_SHIFT 5
+#define FAULT_HTWSEEF_SHIFT 6
+#define FAULT_MHF_SHIFT 7
+#define FAULT_SL_SHIFT 16
+#define FAULT_SS_SHIFT 30
+
+/* If NO fault is present, the following
+ fields are in effect */
+/* (FAULT remains as before) */
+#define PAR_NOFAULT_SS_SHIFT 1
+#define PAR_NOFAULT_MT_SHIFT 4
+#define PAR_NOFAULT_SH_SHIFT 7
+#define PAR_NOFAULT_NS_SHIFT 9
+#define PAR_NOFAULT_NOS_SHIFT 10
+#define PAR_NPFAULT_PA_SHIFT 12
+
+
+/* PRRR */
+#define MTC0_SHIFT 0
+#define MTC1_SHIFT 2
+#define MTC2_SHIFT 4
+#define MTC3_SHIFT 6
+#define MTC4_SHIFT 8
+#define MTC5_SHIFT 10
+#define MTC6_SHIFT 12
+#define MTC7_SHIFT 14
+#define SHDSH0_SHIFT 16
+#define SHDSH1_SHIFT 17
+#define SHNMSH0_SHIFT 18
+#define SHNMSH1_SHIFT 19
+#define NOS0_SHIFT 24
+#define NOS1_SHIFT 25
+#define NOS2_SHIFT 26
+#define NOS3_SHIFT 27
+#define NOS4_SHIFT 28
+#define NOS5_SHIFT 29
+#define NOS6_SHIFT 30
+#define NOS7_SHIFT 31
+
+
+/* RESUME */
+#define TNR_SHIFT 0
+
+
+/* SCTLR */
+#define M_SHIFT 0
+#define TRE_SHIFT 1
+#define AFE_SHIFT 2
+#define HAF_SHIFT 3
+#define BE_SHIFT 4
+#define AFFD_SHIFT 5
+
+
+/* TLBIASID */
+#define TLBIASID_ASID_SHIFT 0
+
+
+/* TLBIVA */
+#define TLBIVA_ASID_SHIFT 0
+#define TLBIVA_VA_SHIFT 12
+
+
+/* TLBIVAA */
+#define TLBIVAA_VA_SHIFT 12
+
+
+/* TLBLCKR */
+#define LKE_SHIFT 0
+#define TLBLCKR_TLBIALLCFG_SHIFT 1
+#define TLBIASIDCFG_SHIFT 2
+#define TLBIVAACFG_SHIFT 3
+#define FLOOR_SHIFT 8
+#define VICTIM_SHIFT 8
+
+
+/* TTBCR */
+#define N_SHIFT 3
+#define PD0_SHIFT 4
+#define PD1_SHIFT 5
+
+
+/* TTBR0 */
+#define TTBR0_IRGNH_SHIFT 0
+#define TTBR0_SH_SHIFT 1
+#define TTBR0_ORGN_SHIFT 3
+#define TTBR0_NOS_SHIFT 5
+#define TTBR0_IRGNL_SHIFT 6
+#define TTBR0_PA_SHIFT 14
+
+
+/* TTBR1 */
+#define TTBR1_IRGNH_SHIFT 0
+#define TTBR1_SH_SHIFT 1
+#define TTBR1_ORGN_SHIFT 3
+#define TTBR1_NOS_SHIFT 5
+#define TTBR1_IRGNL_SHIFT 6
+#define TTBR1_PA_SHIFT 14
+
+
+/* V2PSR */
+#define HIT_SHIFT 0
+#define INDEX_SHIFT 8
+
+
+/* V2Pxx */
+#define V2Pxx_INDEX_SHIFT 0
+#define V2Pxx_VA_SHIFT 12
+
+#endif
diff --git a/arch/arm/mach-msm/iommu.c b/arch/arm/mach-msm/iommu.c
new file mode 100644
index 0000000..f71747d
--- /dev/null
+++ b/arch/arm/mach-msm/iommu.c
@@ -0,0 +1,597 @@
+/* Copyright (c) 2010, Code Aurora Forum. All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 and
+ * only version 2 as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
+ * 02110-1301, USA.
+ */
+
+#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/errno.h>
+#include <linux/io.h>
+#include <linux/interrupt.h>
+#include <linux/list.h>
+#include <linux/spinlock.h>
+#include <linux/slab.h>
+#include <linux/iommu.h>
+
+#include <asm/cacheflush.h>
+#include <asm/sizes.h>
+
+#include <mach/iommu_hw-8xxx.h>
+#include <mach/iommu.h>
+
+DEFINE_SPINLOCK(msm_iommu_lock);
+
+struct msm_priv {
+ unsigned long *pgtable;
+ struct list_head list_attached;
+};
+
+static void __flush_iotlb(struct iommu_domain *domain)
+{
+ struct msm_priv *priv = domain->priv;
+ struct msm_iommu_drvdata *iommu_drvdata;
+ struct msm_iommu_ctx_drvdata *ctx_drvdata;
+
+#ifndef CONFIG_IOMMU_PGTABLES_L2
+ unsigned long *fl_table = priv->pgtable;
+ int i;
+
+ dmac_flush_range(fl_table, fl_table + SZ_16K);
+
+ for (i = 0; i < NUM_FL_PTE; i++)
+ if ((fl_table[i] & 0x03) == FL_TYPE_TABLE) {
+ void *sl_table = __va(fl_table[i] & FL_BASE_MASK);
+ dmac_flush_range(sl_table, sl_table + SZ_4K);
+ }
+#endif
+
+ list_for_each_entry(ctx_drvdata, &priv->list_attached, attached_elm) {
+ if (!ctx_drvdata->pdev || !ctx_drvdata->pdev->dev.parent)
+ BUG();
+
+ iommu_drvdata = dev_get_drvdata(ctx_drvdata->pdev->dev.parent);
+ SET_CTX_TLBIALL(iommu_drvdata->base, ctx_drvdata->num, 0);
+ }
+}
+
+static void __reset_context(void __iomem *base, int ctx)
+{
+ SET_BPRCOSH(base, ctx, 0);
+ SET_BPRCISH(base, ctx, 0);
+ SET_BPRCNSH(base, ctx, 0);
+ SET_BPSHCFG(base, ctx, 0);
+ SET_BPMTCFG(base, ctx, 0);
+ SET_ACTLR(base, ctx, 0);
+ SET_SCTLR(base, ctx, 0);
+ SET_FSRRESTORE(base, ctx, 0);
+ SET_TTBR0(base, ctx, 0);
+ SET_TTBR1(base, ctx, 0);
+ SET_TTBCR(base, ctx, 0);
+ SET_BFBCR(base, ctx, 0);
+ SET_PAR(base, ctx, 0);
+ SET_FAR(base, ctx, 0);
+ SET_CTX_TLBIALL(base, ctx, 0);
+ SET_TLBFLPTER(base, ctx, 0);
+ SET_TLBSLPTER(base, ctx, 0);
+ SET_TLBLKCR(base, ctx, 0);
+ SET_PRRR(base, ctx, 0);
+ SET_NMRR(base, ctx, 0);
+ SET_CONTEXTIDR(base, ctx, 0);
+}
+
+static void __program_context(void __iomem *base, int ctx, phys_addr_t pgtable)
+{
+ __reset_context(base, ctx);
+
+ /* Set up HTW mode */
+ /* TLB miss configuration: perform HTW on miss */
+ SET_TLBMCFG(base, ctx, 0x3);
+
+ /* V2P configuration: HTW for access */
+ SET_V2PCFG(base, ctx, 0x3);
+
+ SET_TTBCR(base, ctx, 0);
+ SET_TTBR0_PA(base, ctx, (pgtable >> 14));
+
+ /* Invalidate the TLB for this context */
+ SET_CTX_TLBIALL(base, ctx, 0);
+
+ /* Set interrupt number to "secure" interrupt */
+ SET_IRPTNDX(base, ctx, 0);
+
+ /* Enable context fault interrupt */
+ SET_CFEIE(base, ctx, 1);
+
+ /* Stall access on a context fault and let the handler deal with it */
+ SET_CFCFG(base, ctx, 1);
+
+ /* Redirect all cacheable requests to L2 slave port. */
+ SET_RCISH(base, ctx, 1);
+ SET_RCOSH(base, ctx, 1);
+ SET_RCNSH(base, ctx, 1);
+
+ /* Turn on TEX Remap */
+ SET_TRE(base, ctx, 1);
+
+ /* Do not configure PRRR / NMRR on the IOMMU for now. We will assume
+ * TEX class 0 for everything until attributes are properly worked out
+ */
+ SET_PRRR(base, ctx, 0);
+ SET_NMRR(base, ctx, 0);
+
+ /* Turn on BFB prefetch */
+ SET_BFBDFE(base, ctx, 1);
+
+#ifdef CONFIG_IOMMU_PGTABLES_L2
+ /* Configure page tables as inner-cacheable and shareable to reduce
+ * the TLB miss penalty.
+ */
+ SET_TTBR0_SH(base, ctx, 1);
+ SET_TTBR1_SH(base, ctx, 1);
+
+ SET_TTBR0_NOS(base, ctx, 1);
+ SET_TTBR1_NOS(base, ctx, 1);
+
+ SET_TTBR0_IRGNH(base, ctx, 0); /* WB, WA */
+ SET_TTBR0_IRGNL(base, ctx, 1);
+
+ SET_TTBR1_IRGNH(base, ctx, 0); /* WB, WA */
+ SET_TTBR1_IRGNL(base, ctx, 1);
+
+ SET_TTBR0_ORGN(base, ctx, 1); /* WB, WA */
+ SET_TTBR1_ORGN(base, ctx, 1); /* WB, WA */
+#endif
+
+ /* Enable the MMU */
+ SET_M(base, ctx, 1);
+}
+
+static int msm_iommu_domain_init(struct iommu_domain *domain)
+{
+ struct msm_priv *priv = kzalloc(sizeof(*priv), GFP_KERNEL);
+
+ if (!priv)
+ goto fail_nomem;
+
+ INIT_LIST_HEAD(&priv->list_attached);
+ priv->pgtable = (unsigned long *)__get_free_pages(GFP_KERNEL,
+ get_order(SZ_16K));
+
+ if (!priv->pgtable)
+ goto fail_nomem;
+
+ memset(priv->pgtable, 0, SZ_16K);
+ domain->priv = priv;
+ return 0;
+
+fail_nomem:
+ kfree(priv);
+ return -ENOMEM;
+}
+
+static void msm_iommu_domain_destroy(struct iommu_domain *domain)
+{
+ struct msm_priv *priv;
+ unsigned long flags;
+ unsigned long *fl_table;
+ int i;
+
+ spin_lock_irqsave(&msm_iommu_lock, flags);
+ priv = domain->priv;
+ domain->priv = NULL;
+
+ if (priv) {
+ fl_table = priv->pgtable;
+
+ for (i = 0; i < NUM_FL_PTE; i++)
+ if ((fl_table[i] & 0x03) == FL_TYPE_TABLE)
+ free_page((unsigned long) __va(((fl_table[i]) &
+ FL_BASE_MASK)));
+
+ free_pages((unsigned long)priv->pgtable, get_order(SZ_16K));
+ priv->pgtable = NULL;
+ }
+
+ kfree(priv);
+ spin_unlock_irqrestore(&msm_iommu_lock, flags);
+}
+
+static int msm_iommu_attach_dev(struct iommu_domain *domain, struct device *dev)
+{
+ struct msm_priv *priv;
+ struct msm_iommu_ctx_dev *ctx_dev;
+ struct msm_iommu_drvdata *iommu_drvdata;
+ struct msm_iommu_ctx_drvdata *ctx_drvdata;
+ struct msm_iommu_ctx_drvdata *tmp_drvdata;
+ int ret = 0;
+ unsigned long flags;
+
+ spin_lock_irqsave(&msm_iommu_lock, flags);
+
+ priv = domain->priv;
+
+ if (!priv || !dev) {
+ ret = -EINVAL;
+ goto fail;
+ }
+
+ iommu_drvdata = dev_get_drvdata(dev->parent);
+ ctx_drvdata = dev_get_drvdata(dev);
+ ctx_dev = dev->platform_data;
+
+ if (!iommu_drvdata || !ctx_drvdata || !ctx_dev) {
+ ret = -EINVAL;
+ goto fail;
+ }
+
+ list_for_each_entry(tmp_drvdata, &priv->list_attached, attached_elm)
+ if (tmp_drvdata == ctx_drvdata) {
+ ret = -EBUSY;
+ goto fail;
+ }
+
+ __program_context(iommu_drvdata->base, ctx_dev->num,
+ __pa(priv->pgtable));
+
+ list_add(&(ctx_drvdata->attached_elm), &priv->list_attached);
+ __flush_iotlb(domain);
+
+fail:
+ spin_unlock_irqrestore(&msm_iommu_lock, flags);
+ return ret;
+}
+
+static void msm_iommu_detach_dev(struct iommu_domain *domain,
+ struct device *dev)
+{
+ struct msm_priv *priv;
+ struct msm_iommu_ctx_dev *ctx_dev;
+ struct msm_iommu_drvdata *iommu_drvdata;
+ struct msm_iommu_ctx_drvdata *ctx_drvdata;
+ unsigned long flags;
+
+ spin_lock_irqsave(&msm_iommu_lock, flags);
+ priv = domain->priv;
+
+ if (!priv || !dev)
+ goto fail;
+
+ iommu_drvdata = dev_get_drvdata(dev->parent);
+ ctx_drvdata = dev_get_drvdata(dev);
+ ctx_dev = dev->platform_data;
+
+ if (!iommu_drvdata || !ctx_drvdata || !ctx_dev)
+ goto fail;
+
+ __flush_iotlb(domain);
+ __reset_context(iommu_drvdata->base, ctx_dev->num);
+ list_del_init(&ctx_drvdata->attached_elm);
+
+fail:
+ spin_unlock_irqrestore(&msm_iommu_lock, flags);
+}
+
+static int msm_iommu_map(struct iommu_domain *domain, unsigned long va,
+ phys_addr_t pa, int order, int prot)
+{
+ struct msm_priv *priv;
+ unsigned long flags;
+ unsigned long *fl_table;
+ unsigned long *fl_pte;
+ unsigned long fl_offset;
+ unsigned long *sl_table;
+ unsigned long *sl_pte;
+ unsigned long sl_offset;
+ size_t len = 0x1000UL << order;
+ int ret = 0;
+
+ spin_lock_irqsave(&msm_iommu_lock, flags);
+ priv = domain->priv;
+
+ if (!priv) {
+ ret = -EINVAL;
+ goto fail;
+ }
+
+ fl_table = priv->pgtable;
+
+ if (len != SZ_16M && len != SZ_1M &&
+ len != SZ_64K && len != SZ_4K) {
+ pr_debug("Bad size: %d\n", len);
+ ret = -EINVAL;
+ goto fail;
+ }
+
+ if (!fl_table) {
+ pr_debug("Null page table\n");
+ ret = -EINVAL;
+ goto fail;
+ }
+
+ fl_offset = FL_OFFSET(va); /* Upper 12 bits */
+ fl_pte = fl_table + fl_offset; /* int pointers, 4 bytes */
+
+ if (len == SZ_16M) {
+ int i = 0;
+ for (i = 0; i < 16; i++)
+ *(fl_pte+i) = (pa & 0xFF000000) | FL_SUPERSECTION |
+ FL_AP_READ | FL_AP_WRITE | FL_TYPE_SECT |
+ FL_SHARED;
+ }
+
+ if (len == SZ_1M)
+ *fl_pte = (pa & 0xFFF00000) | FL_AP_READ | FL_AP_WRITE |
+ FL_TYPE_SECT | FL_SHARED;
+
+ /* Need a 2nd level table */
+ if ((len == SZ_4K || len == SZ_64K) && (*fl_pte) == 0) {
+ unsigned long *sl;
+ sl = (unsigned long *) __get_free_pages(GFP_KERNEL,
+ get_order(SZ_4K));
+
+ if (!sl) {
+ pr_debug("Could not allocate second level table\n");
+ ret = -ENOMEM;
+ goto fail;
+ }
+
+ memset(sl, 0, SZ_4K);
+ *fl_pte = ((((int)__pa(sl)) & FL_BASE_MASK) | FL_TYPE_TABLE);
+ }
+
+ sl_table = (unsigned long *) __va(((*fl_pte) & FL_BASE_MASK));
+ sl_offset = SL_OFFSET(va);
+ sl_pte = sl_table + sl_offset;
+
+
+ if (len == SZ_4K)
+ *sl_pte = (pa & SL_BASE_MASK_SMALL) | SL_AP0 | SL_AP1 |
+ SL_SHARED | SL_TYPE_SMALL;
+
+ if (len == SZ_64K) {
+ int i;
+
+ for (i = 0; i < 16; i++)
+ *(sl_pte+i) = (pa & SL_BASE_MASK_LARGE) | SL_AP0 |
+ SL_AP1 | SL_SHARED | SL_TYPE_LARGE;
+ }
+
+ __flush_iotlb(domain);
+fail:
+ spin_unlock_irqrestore(&msm_iommu_lock, flags);
+ return ret;
+}
+
+static int msm_iommu_unmap(struct iommu_domain *domain, unsigned long va,
+ int order)
+{
+ struct msm_priv *priv;
+ unsigned long flags;
+ unsigned long *fl_table;
+ unsigned long *fl_pte;
+ unsigned long fl_offset;
+ unsigned long *sl_table;
+ unsigned long *sl_pte;
+ unsigned long sl_offset;
+ size_t len = 0x1000UL << order;
+ int i, ret = 0;
+
+ spin_lock_irqsave(&msm_iommu_lock, flags);
+
+ priv = domain->priv;
+
+ if (!priv) {
+ ret = -ENODEV;
+ goto fail;
+ }
+
+ fl_table = priv->pgtable;
+
+ if (len != SZ_16M && len != SZ_1M &&
+ len != SZ_64K && len != SZ_4K) {
+ pr_debug("Bad length: %d\n", len);
+ ret = -EINVAL;
+ goto fail;
+ }
+
+ if (!fl_table) {
+ pr_debug("Null page table\n");
+ ret = -EINVAL;
+ goto fail;
+ }
+
+ fl_offset = FL_OFFSET(va); /* Upper 12 bits */
+ fl_pte = fl_table + fl_offset; /* int pointers, 4 bytes */
+
+ if (*fl_pte == 0) {
+ pr_debug("First level PTE is 0\n");
+ ret = -ENODEV;
+ goto fail;
+ }
+
+ /* Unmap supersection */
+ if (len == SZ_16M)
+ for (i = 0; i < 16; i++)
+ *(fl_pte+i) = 0;
+
+ if (len == SZ_1M)
+ *fl_pte = 0;
+
+ sl_table = (unsigned long *) __va(((*fl_pte) & FL_BASE_MASK));
+ sl_offset = SL_OFFSET(va);
+ sl_pte = sl_table + sl_offset;
+
+ if (len == SZ_64K) {
+ for (i = 0; i < 16; i++)
+ *(sl_pte+i) = 0;
+ }
+
+ if (len == SZ_4K)
+ *sl_pte = 0;
+
+ if (len == SZ_4K || len == SZ_64K) {
+ int used = 0;
+
+ for (i = 0; i < NUM_SL_PTE; i++)
+ if (sl_table[i])
+ used = 1;
+ if (!used) {
+ free_page((unsigned long)sl_table);
+ *fl_pte = 0;
+ }
+ }
+
+ __flush_iotlb(domain);
+fail:
+ spin_unlock_irqrestore(&msm_iommu_lock, flags);
+ return ret;
+}
+
+static phys_addr_t msm_iommu_iova_to_phys(struct iommu_domain *domain,
+ unsigned long va)
+{
+ struct msm_priv *priv;
+ struct msm_iommu_drvdata *iommu_drvdata;
+ struct msm_iommu_ctx_drvdata *ctx_drvdata;
+ unsigned int par;
+ unsigned long flags;
+ void __iomem *base;
+ phys_addr_t ret = 0;
+ int ctx;
+
+ spin_lock_irqsave(&msm_iommu_lock, flags);
+
+ priv = domain->priv;
+ if (list_empty(&priv->list_attached))
+ goto fail;
+
+ ctx_drvdata = list_entry(priv->list_attached.next,
+ struct msm_iommu_ctx_drvdata, attached_elm);
+ iommu_drvdata = dev_get_drvdata(ctx_drvdata->pdev->dev.parent);
+
+ base = iommu_drvdata->base;
+ ctx = ctx_drvdata->num;
+
+ /* Invalidate context TLB */
+ SET_CTX_TLBIALL(base, ctx, 0);
+ SET_V2PPR_VA(base, ctx, va >> V2Pxx_VA_SHIFT);
+
+ if (GET_FAULT(base, ctx))
+ goto fail;
+
+ par = GET_PAR(base, ctx);
+
+ /* We are dealing with a supersection */
+ if (GET_NOFAULT_SS(base, ctx))
+ ret = (par & 0xFF000000) | (va & 0x00FFFFFF);
+ else /* Upper 20 bits from PAR, lower 12 from VA */
+ ret = (par & 0xFFFFF000) | (va & 0x00000FFF);
+
+fail:
+ spin_unlock_irqrestore(&msm_iommu_lock, flags);
+ return ret;
+}
+
+static int msm_iommu_domain_has_cap(struct iommu_domain *domain,
+ unsigned long cap)
+{
+ return 0;
+}
+
+static void print_ctx_regs(void __iomem *base, int ctx)
+{
+ unsigned int fsr = GET_FSR(base, ctx);
+ pr_err("FAR = %08x PAR = %08x\n",
+ GET_FAR(base, ctx), GET_PAR(base, ctx));
+ pr_err("FSR = %08x [%s%s%s%s%s%s%s%s%s%s]\n", fsr,
+ (fsr & 0x02) ? "TF " : "",
+ (fsr & 0x04) ? "AFF " : "",
+ (fsr & 0x08) ? "APF " : "",
+ (fsr & 0x10) ? "TLBMF " : "",
+ (fsr & 0x20) ? "HTWDEEF " : "",
+ (fsr & 0x40) ? "HTWSEEF " : "",
+ (fsr & 0x80) ? "MHF " : "",
+ (fsr & 0x10000) ? "SL " : "",
+ (fsr & 0x40000000) ? "SS " : "",
+ (fsr & 0x80000000) ? "MULTI " : "");
+
+ pr_err("FSYNR0 = %08x FSYNR1 = %08x\n",
+ GET_FSYNR0(base, ctx), GET_FSYNR1(base, ctx));
+ pr_err("TTBR0 = %08x TTBR1 = %08x\n",
+ GET_TTBR0(base, ctx), GET_TTBR1(base, ctx));
+ pr_err("SCTLR = %08x ACTLR = %08x\n",
+ GET_SCTLR(base, ctx), GET_ACTLR(base, ctx));
+ pr_err("PRRR = %08x NMRR = %08x\n",
+ GET_PRRR(base, ctx), GET_NMRR(base, ctx));
+}
+
+irqreturn_t msm_iommu_fault_handler(int irq, void *dev_id)
+{
+ struct msm_iommu_drvdata *drvdata = dev_id;
+ void __iomem *base;
+ unsigned int fsr = 0;
+ int ncb = 0, i = 0;
+
+ spin_lock(&msm_iommu_lock);
+
+ if (!drvdata) {
+ pr_err("Invalid device ID in context interrupt handler\n");
+ goto fail;
+ }
+
+ base = drvdata->base;
+
+ pr_err("===== WOAH! =====\n");
+ pr_err("Unexpected IOMMU page fault!\n");
+ pr_err("base = %08x\n", (unsigned int) base);
+
+ ncb = GET_NCB(base)+1;
+ for (i = 0; i < ncb; i++) {
+ fsr = GET_FSR(base, i);
+ if (fsr) {
+ pr_err("Fault occurred in context %d.\n", i);
+ pr_err("Interesting registers:\n");
+ print_ctx_regs(base, i);
+ SET_FSR(base, i, 0x4000000F);
+ }
+ }
+fail:
+ spin_unlock(&msm_iommu_lock);
+ return 0;
+}
+
+static struct iommu_ops msm_iommu_ops = {
+ .domain_init = msm_iommu_domain_init,
+ .domain_destroy = msm_iommu_domain_destroy,
+ .attach_dev = msm_iommu_attach_dev,
+ .detach_dev = msm_iommu_detach_dev,
+ .map = msm_iommu_map,
+ .unmap = msm_iommu_unmap,
+ .iova_to_phys = msm_iommu_iova_to_phys,
+ .domain_has_cap = msm_iommu_domain_has_cap
+};
+
+static int msm_iommu_init(void)
+{
+ register_iommu(&msm_iommu_ops);
+ return 0;
+}
+
+subsys_initcall(msm_iommu_init);
+
+MODULE_LICENSE("GPL v2");
+MODULE_AUTHOR("Stepan Moskovchenko <[email protected]>");
--
1.7.2.1
Sent by an employee of the Qualcomm Innovation Center, Inc.
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum.
From: Abhijeet Dharmapurikar <[email protected]>
Private Peripheral interrupts could be edge triggered or level triggered
depending on the platform. Initialize handlers for these in board file.
Signed-off-by: Abhijeet Dharmapurikar <[email protected]>
---
arch/arm/mach-msm/board-msm8x60.c | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/arch/arm/mach-msm/board-msm8x60.c b/arch/arm/mach-msm/board-msm8x60.c
index c6bf8e3..e2e5a5c 100644
--- a/arch/arm/mach-msm/board-msm8x60.c
+++ b/arch/arm/mach-msm/board-msm8x60.c
@@ -46,7 +46,7 @@ static void __init msm8x60_init_irq(void)
{
unsigned int i;
- gic_dist_init(0, MSM_QGIC_DIST_BASE, 1);
+ gic_dist_init(0, MSM_QGIC_DIST_BASE, GIC_PPI_START);
gic_cpu_base_addr = (void *)MSM_QGIC_CPU_BASE;
gic_cpu_init(0, MSM_QGIC_CPU_BASE);
--
1.7.2.1
Sent by an employee of the Qualcomm Innovation Center, Inc.
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum.
Signed-off-by: Daniel Walker <[email protected]>
Signed-off-by: Jeff Ohlstein <[email protected]>
---
arch/arm/mach-msm/include/mach/dma.h | 10 +++++++++-
1 files changed, 9 insertions(+), 1 deletions(-)
diff --git a/arch/arm/mach-msm/include/mach/dma.h b/arch/arm/mach-msm/include/mach/dma.h
index 00f9bbf..05583f5 100644
--- a/arch/arm/mach-msm/include/mach/dma.h
+++ b/arch/arm/mach-msm/include/mach/dma.h
@@ -32,10 +32,18 @@ struct msm_dmov_cmd {
void *data;
};
+#ifndef CONFIG_ARCH_MSM8X60
void msm_dmov_enqueue_cmd(unsigned id, struct msm_dmov_cmd *cmd);
void msm_dmov_stop_cmd(unsigned id, struct msm_dmov_cmd *cmd, int graceful);
int msm_dmov_exec_cmd(unsigned id, unsigned int cmdptr);
-
+#else
+static inline
+void msm_dmov_enqueue_cmd(unsigned id, struct msm_dmov_cmd *cmd) { }
+static inline
+void msm_dmov_stop_cmd(unsigned id, struct msm_dmov_cmd *cmd, int graceful) { }
+static inline
+int msm_dmov_exec_cmd(unsigned id, unsigned int cmdptr) { return -EIO; }
+#endif
#define DMOV_SD0(off, ch) (MSM_DMOV_BASE + 0x0000 + (off) + ((ch) << 2))
--
1.7.2.1
Sent by an employee of the Qualcomm Innovation Center, Inc.
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum.
From: Daniel Walker <[email protected]>
We don't actually do anything for hotplug yet. This just lets
everything compile, even when hotplug is enabled.
Signed-off-by: Daniel Walker <[email protected]>
Signed-off-by: Jeff Ohlstein <[email protected]>
---
arch/arm/mach-msm/Makefile | 1 +
arch/arm/mach-msm/hotplug.c | 22 ++++++++++++++++++++++
2 files changed, 23 insertions(+), 0 deletions(-)
create mode 100644 arch/arm/mach-msm/hotplug.c
diff --git a/arch/arm/mach-msm/Makefile b/arch/arm/mach-msm/Makefile
index ff19e01..8a01f8a 100644
--- a/arch/arm/mach-msm/Makefile
+++ b/arch/arm/mach-msm/Makefile
@@ -21,4 +21,5 @@ obj-y += gpiomux.o
obj-$(CONFIG_ARCH_MSM7X30) += gpiomux-7x30.o gpiomux-7x00.o
obj-$(CONFIG_ARCH_QSD8X50) += gpiomux-8x50.o gpiomux-7x00.o
obj-$(CONFIG_ARCH_MSM8X60) += gpiomux-8x60.o
+obj-$(CONFIG_HOTPLUG_CPU) += hotplug.o
obj-$(CONFIG_MACH_MSM8X60_RUMI3) += board-msm8x60.o
diff --git a/arch/arm/mach-msm/hotplug.c b/arch/arm/mach-msm/hotplug.c
new file mode 100644
index 0000000..ba0803a
--- /dev/null
+++ b/arch/arm/mach-msm/hotplug.c
@@ -0,0 +1,22 @@
+/*
+ * Do nothing for now.
+ */
+
+int platform_cpu_kill(unsigned int cpu)
+{
+ return 0;
+}
+
+void platform_cpu_die(unsigned int cpu)
+{
+}
+
+int platform_cpu_disable(unsigned int cpu)
+{
+ return 0;
+}
+
+void local_timer_stop(void)
+{
+}
+
--
1.7.2.1
Sent by an employee of the Qualcomm Innovation Center, Inc.
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum.
From: Daniel Walker <[email protected]>
Some MSM targets don't select the debug UART in this way. For those we
need to disable this selection mechanism.
Signed-off-by: Daniel Walker <[email protected]>
Signed-off-by: Jeff Ohlstein <[email protected]>
---
arch/arm/mach-msm/Kconfig | 8 +++++++-
arch/arm/mach-msm/include/mach/debug-macro.S | 2 +-
2 files changed, 8 insertions(+), 2 deletions(-)
diff --git a/arch/arm/mach-msm/Kconfig b/arch/arm/mach-msm/Kconfig
index 60ad211..27fc183 100644
--- a/arch/arm/mach-msm/Kconfig
+++ b/arch/arm/mach-msm/Kconfig
@@ -11,6 +11,7 @@ config ARCH_MSM7X00A
select MSM_SMD_PKG3
select CPU_V6
select MSM_PROC_COMM
+ select HAS_MSM_DEBUG_UART_PHYS
config ARCH_MSM7X30
bool "MSM7x30"
@@ -20,6 +21,7 @@ config ARCH_MSM7X30
select CPU_V7
select MSM_REMOTE_SPINLOCK_DEKKERS
select MSM_PROC_COMM
+ select HAS_MSM_DEBUG_UART_PHYS
config ARCH_QSD8X50
bool "QSD8X50"
@@ -29,6 +31,7 @@ config ARCH_QSD8X50
select CPU_V7
select MSM_REMOTE_SPINLOCK_LDREX
select MSM_PROC_COMM
+ select HAS_MSM_DEBUG_UART_PHYS
endchoice
@@ -41,7 +44,8 @@ config ARCH_MSM_SCORPION
bool
config ARCH_MSM_SCORPIONMP
bool
-
+config HAS_MSM_DEBUG_UART_PHYS
+ bool
config MSM_VIC
bool
@@ -95,6 +99,7 @@ config MSM_DEBUG_UART
default 2 if MSM_DEBUG_UART2
default 3 if MSM_DEBUG_UART3
+if HAS_MSM_DEBUG_UART_PHYS
choice
prompt "Debug UART"
@@ -112,6 +117,7 @@ choice
config MSM_DEBUG_UART3
bool "UART3"
endchoice
+endif
config MSM_SMD_PKG3
bool
diff --git a/arch/arm/mach-msm/include/mach/debug-macro.S b/arch/arm/mach-msm/include/mach/debug-macro.S
index 528750f..238c4f1 100644
--- a/arch/arm/mach-msm/include/mach/debug-macro.S
+++ b/arch/arm/mach-msm/include/mach/debug-macro.S
@@ -19,7 +19,7 @@
#include <mach/hardware.h>
#include <mach/msm_iomap.h>
-#ifdef CONFIG_MSM_DEBUG_UART
+#ifdef CONFIG_HAS_MSM_DEBUG_UART_PHYS
.macro addruart, rx, tmp
@ see if the MMU is enabled and select appropriate base address
mrc p15, 0, \rx, c1, c0
--
1.7.2.1
Sent by an employee of the Qualcomm Innovation Center, Inc.
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum.
From: Steve Muckle <[email protected]>
The existing MSM irq entry macro is specific to a VIC
implementation. Renaming this makes room for irq support based on
other interrupt controllers.
Signed-off-by: Steve Muckle <[email protected]>
---
arch/arm/mach-msm/include/mach/entry-macro-qgic.S | 96 +++++++++++++++++++++
arch/arm/mach-msm/include/mach/entry-macro-vic.S | 37 ++++++++
arch/arm/mach-msm/include/mach/entry-macro.S | 43 +++------
arch/arm/mach-msm/include/mach/smp.h | 39 +++++++++
4 files changed, 186 insertions(+), 29 deletions(-)
create mode 100644 arch/arm/mach-msm/include/mach/entry-macro-qgic.S
create mode 100644 arch/arm/mach-msm/include/mach/entry-macro-vic.S
create mode 100644 arch/arm/mach-msm/include/mach/smp.h
diff --git a/arch/arm/mach-msm/include/mach/entry-macro-qgic.S b/arch/arm/mach-msm/include/mach/entry-macro-qgic.S
new file mode 100644
index 0000000..092e48e
--- /dev/null
+++ b/arch/arm/mach-msm/include/mach/entry-macro-qgic.S
@@ -0,0 +1,96 @@
+/*
+ * Low-level IRQ helper macros
+ *
+ * Copyright (c) 2010, Code Aurora Forum. All rights reserved.
+ *
+ * This file is licensed under the terms of the GNU General Public
+ * License version 2. This program is licensed "as is" without any
+ * warranty of any kind, whether express or implied.
+ */
+
+#include <mach/hardware.h>
+#include <asm/hardware/gic.h>
+
+ .macro disable_fiq
+ .endm
+
+ .macro get_irqnr_preamble, base, tmp
+ ldr \base, =gic_cpu_base_addr
+ ldr \base, [\base]
+ .endm
+
+ .macro arch_ret_to_user, tmp1, tmp2
+ .endm
+
+ /*
+ * The interrupt numbering scheme is defined in the
+ * interrupt controller spec. To wit:
+ *
+ * Migrated the code from ARM MP port to be more consistant
+ * with interrupt processing , the following still holds true
+ * however, all interrupts are treated the same regardless of
+ * if they are local IPI or PPI
+ *
+ * Interrupts 0-15 are IPI
+ * 16-31 are PPI
+ * (16-18 are the timers)
+ * 32-1020 are global
+ * 1021-1022 are reserved
+ * 1023 is "spurious" (no interrupt)
+ *
+ * A simple read from the controller will tell us the number of the
+ * highest priority enabled interrupt. We then just need to check
+ * whether it is in the valid range for an IRQ (0-1020 inclusive).
+ *
+ * Base ARM code assumes that the local (private) peripheral interrupts
+ * are not valid, we treat them differently, in that the privates are
+ * handled like normal shared interrupts with the exception that only
+ * one processor can register the interrupt and the handler must be
+ * the same for all processors.
+ */
+
+ .macro get_irqnr_and_base, irqnr, irqstat, base, tmp
+
+ ldr \irqstat, [\base, #GIC_CPU_INTACK] /* bits 12-10 =srcCPU,
+ 9-0 =int # */
+
+ bic \irqnr, \irqstat, #0x1c00 @mask src
+#ifdef CONFIG_REQUEST_IPI
+ cmp \irqnr, #0
+#else
+ cmp \irqnr, #15
+#endif
+ ldr \tmp, =1021
+ cmpcc \irqnr, \irqnr
+ cmpne \irqnr, \tmp
+ cmpcs \irqnr, \irqnr
+
+ .endm
+
+ /* We assume that irqstat (the raw value of the IRQ acknowledge
+ * register) is preserved from the macro above.
+ * If there is an IPI, we immediately signal end of interrupt on the
+ * controller, since this requires the original irqstat value which
+ * we won't easily be able to recreate later.
+ */
+ .macro test_for_ipi, irqnr, irqstat, base, tmp
+#ifndef CONFIG_REQUEST_IPI
+ bic \irqnr, \irqstat, #0x1c00
+ cmp \irqnr, #16
+ strcc \irqstat, [\base, #GIC_CPU_EOI]
+ cmpcs \irqnr, \irqnr
+#endif
+ .endm
+
+ /* As above, this assumes that irqstat and base are preserved.. */
+
+ .macro test_for_ltirq, irqnr, irqstat, base, tmp
+#ifndef CONFIG_REQUEST_IPI
+ bic \irqnr, \irqstat, #0x1c00
+ mov \tmp, #0
+ cmp \irqnr, #16
+ moveq \tmp, #1
+ streq \irqstat, [\base, #GIC_CPU_EOI]
+ cmp \tmp, #0
+#endif
+ .endm
diff --git a/arch/arm/mach-msm/include/mach/entry-macro-vic.S b/arch/arm/mach-msm/include/mach/entry-macro-vic.S
new file mode 100644
index 0000000..70563ed
--- /dev/null
+++ b/arch/arm/mach-msm/include/mach/entry-macro-vic.S
@@ -0,0 +1,37 @@
+/*
+ * Copyright (C) 2007 Google, Inc.
+ * Author: Brian Swetland <[email protected]>
+ *
+ * This software is licensed under the terms of the GNU General Public
+ * License version 2, as published by the Free Software Foundation, and
+ * may be copied, distributed, and modified under those terms.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ */
+
+#include <mach/msm_iomap.h>
+
+ .macro disable_fiq
+ .endm
+
+ .macro get_irqnr_preamble, base, tmp
+ @ enable imprecise aborts
+ cpsie a
+ mov \base, #MSM_VIC_BASE
+ .endm
+
+ .macro arch_ret_to_user, tmp1, tmp2
+ .endm
+
+ .macro get_irqnr_and_base, irqnr, irqstat, base, tmp
+ @ 0xD0 has irq# or old irq# if the irq has been handled
+ @ 0xD4 has irq# or -1 if none pending *but* if you just
+ @ read 0xD4 you never get the first irq for some reason
+ ldr \irqnr, [\base, #0xD0]
+ ldr \irqnr, [\base, #0xD4]
+ cmp \irqnr, #0xffffffff
+ .endm
diff --git a/arch/arm/mach-msm/include/mach/entry-macro.S b/arch/arm/mach-msm/include/mach/entry-macro.S
index d225948..b16f082 100644
--- a/arch/arm/mach-msm/include/mach/entry-macro.S
+++ b/arch/arm/mach-msm/include/mach/entry-macro.S
@@ -1,38 +1,23 @@
-/* arch/arm/mach-msm7200/include/mach/entry-macro.S
+/* Copyright (c) 2010, Code Aurora Forum. All rights reserved.
*
- * Copyright (C) 2007 Google, Inc.
- * Author: Brian Swetland <[email protected]>
- *
- * This software is licensed under the terms of the GNU General Public
- * License version 2, as published by the Free Software Foundation, and
- * may be copied, distributed, and modified under those terms.
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 and
+ * only version 2 as published by the Free Software Foundation.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
+ * 02110-1301, USA.
+ *
*/
-#include <mach/msm_iomap.h>
-
- .macro disable_fiq
- .endm
-
- .macro get_irqnr_preamble, base, tmp
- @ enable imprecise aborts
- cpsie a
- mov \base, #MSM_VIC_BASE
- .endm
-
- .macro arch_ret_to_user, tmp1, tmp2
- .endm
-
- .macro get_irqnr_and_base, irqnr, irqstat, base, tmp
- @ 0xD0 has irq# or old irq# if the irq has been handled
- @ 0xD4 has irq# or -1 if none pending *but* if you just
- @ read 0xD4 you never get the first irq for some reason
- ldr \irqnr, [\base, #0xD0]
- ldr \irqnr, [\base, #0xD4]
- cmp \irqnr, #0xffffffff
- .endm
+#if defined(CONFIG_ARM_GIC)
+#include <mach/entry-macro-qgic.S>
+#else
+#include <mach/entry-macro-vic.S>
+#endif
diff --git a/arch/arm/mach-msm/include/mach/smp.h b/arch/arm/mach-msm/include/mach/smp.h
new file mode 100644
index 0000000..3ff7bf5
--- /dev/null
+++ b/arch/arm/mach-msm/include/mach/smp.h
@@ -0,0 +1,39 @@
+/* Copyright (c) 2010, Code Aurora Forum. All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions are met:
+ * * Redistributions of source code must retain the above copyright
+ * notice, this list of conditions and the following disclaimer.
+ * * Redistributions in binary form must reproduce the above copyright
+ * notice, this list of conditions and the following disclaimer in the
+ * documentation and/or other materials provided with the distribution.
+ * * Neither the name of Code Aurora nor
+ * the names of its contributors may be used to endorse or promote
+ * products derived from this software without specific prior written
+ * permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NON-INFRINGEMENT ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR
+ * CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
+ * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
+ * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS;
+ * OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,
+ * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR
+ * OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF
+ * ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ *
+ */
+
+#ifndef __ASM_ARCH_MSM_SMP_H
+#define __ASM_ARCH_MSM_SMP_H
+
+#include <asm/hardware/gic.h>
+
+static inline void smp_cross_call(const struct cpumask *mask)
+{
+ gic_raise_softirq(mask, 1);
+}
+
+#endif
--
1.7.2.1
Sent by an employee of the Qualcomm Innovation Center, Inc.
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum.
From: Steve Muckle <[email protected]>
On RUMI platform STIs are not enabled by default, contrary to the
GIC spec. The bits for STIs in the enable/enable clear registers
are also RW instead of RO. STIs need to be enabled at initialization
time.
Signed-off-by: Steve Muckle <[email protected]>
---
arch/arm/mach-msm/board-msm8x60.c | 21 +++++++++++++++++++++
1 files changed, 21 insertions(+), 0 deletions(-)
diff --git a/arch/arm/mach-msm/board-msm8x60.c b/arch/arm/mach-msm/board-msm8x60.c
index 3ab4bd9..c6bf8e3 100644
--- a/arch/arm/mach-msm/board-msm8x60.c
+++ b/arch/arm/mach-msm/board-msm8x60.c
@@ -18,6 +18,8 @@
#include <linux/kernel.h>
#include <linux/platform_device.h>
+#include <linux/irq.h>
+#include <linux/io.h>
#include <asm/mach-types.h>
#include <asm/mach/arch.h>
@@ -42,9 +44,28 @@ static void __init msm8x60_map_io(void)
static void __init msm8x60_init_irq(void)
{
+ unsigned int i;
+
gic_dist_init(0, MSM_QGIC_DIST_BASE, 1);
gic_cpu_base_addr = (void *)MSM_QGIC_CPU_BASE;
gic_cpu_init(0, MSM_QGIC_CPU_BASE);
+
+ /* Edge trigger PPIs except AVS_SVICINT and AVS_SVICINTSWDONE */
+ writel(0xFFFFD7FF, MSM_QGIC_DIST_BASE + GIC_DIST_CONFIG + 4);
+
+ /* RUMI does not adhere to GIC spec by enabling STIs by default.
+ * Enable/clear is supposed to be RO for STIs, but is RW on RUMI.
+ */
+ writel(0x0000FFFF, MSM_QGIC_DIST_BASE + GIC_DIST_ENABLE_SET);
+
+ /* FIXME: Not installing AVS_SVICINT and AVS_SVICINTSWDONE yet
+ * as they are configured as level, which does not play nice with
+ * handle_percpu_irq.
+ */
+ for (i = GIC_PPI_START; i < GIC_SPI_START; i++) {
+ if (i != AVS_SVICINT && i != AVS_SVICINTSWDONE)
+ set_irq_handler(i, handle_percpu_irq);
+ }
}
static void __init msm8x60_init(void)
--
1.7.2.1
Sent by an employee of the Qualcomm Innovation Center, Inc.
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum.
From: Steve Muckle <[email protected]>
Board configuration for MSM8X60 emulation on RUMI3.
Signed-off-by: Steve Muckle <[email protected]>
---
arch/arm/mach-msm/Kconfig | 7 ++++
arch/arm/mach-msm/Makefile | 2 +-
arch/arm/mach-msm/board-msm8x60.c | 59 +++++++++++++++++++++++++++++++++++++
3 files changed, 67 insertions(+), 1 deletions(-)
create mode 100644 arch/arm/mach-msm/board-msm8x60.c
diff --git a/arch/arm/mach-msm/Kconfig b/arch/arm/mach-msm/Kconfig
index 3b9bece..60ad211 100644
--- a/arch/arm/mach-msm/Kconfig
+++ b/arch/arm/mach-msm/Kconfig
@@ -80,6 +80,13 @@ config MACH_QSD8X50A_ST1_5
help
Support for the Qualcomm ST1.5.
+config MACH_MSM8X60_RUMI3
+ depends on ARCH_MSM8X60
+ default n
+ bool "MSM8x60 RUMI3"
+ help
+ Support for the Qualcomm MSM8x60 RUMI3 emulator.
+
endmenu
config MSM_DEBUG_UART
diff --git a/arch/arm/mach-msm/Makefile b/arch/arm/mach-msm/Makefile
index 78fb536..64a95ae 100644
--- a/arch/arm/mach-msm/Makefile
+++ b/arch/arm/mach-msm/Makefile
@@ -16,8 +16,8 @@ obj-$(CONFIG_MACH_TROUT) += board-trout.o board-trout-gpio.o board-trout-mmc.o d
obj-$(CONFIG_MACH_HALIBUT) += board-halibut.o devices-msm7x00.o
obj-$(CONFIG_ARCH_MSM7X30) += board-msm7x30.o devices-msm7x30.o
obj-$(CONFIG_ARCH_QSD8X50) += board-qsd8x50.o devices-qsd8x50.o
-
obj-y += gpiomux.o
obj-$(CONFIG_ARCH_MSM7X30) += gpiomux-7x30.o gpiomux-7x00.o
obj-$(CONFIG_ARCH_QSD8X50) += gpiomux-8x50.o gpiomux-7x00.o
obj-$(CONFIG_ARCH_MSM8X60) += gpiomux-8x60.o
+obj-$(CONFIG_MACH_MSM8X60_RUMI3) += board-msm8x60.o
diff --git a/arch/arm/mach-msm/board-msm8x60.c b/arch/arm/mach-msm/board-msm8x60.c
new file mode 100644
index 0000000..3ab4bd9
--- /dev/null
+++ b/arch/arm/mach-msm/board-msm8x60.c
@@ -0,0 +1,59 @@
+/* Copyright (c) 2010, Code Aurora Forum. All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 and
+ * only version 2 as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
+ * 02110-1301, USA.
+ *
+ */
+
+#include <linux/kernel.h>
+#include <linux/platform_device.h>
+
+#include <asm/mach-types.h>
+#include <asm/mach/arch.h>
+#include <asm/hardware/gic.h>
+
+#include <mach/board.h>
+#include <mach/msm_iomap.h>
+#include <mach/io.h>
+
+extern struct sys_timer msm_timer;
+void __iomem *gic_cpu_base_addr;
+
+unsigned long clk_get_max_axi_khz(void)
+{
+ return 0;
+}
+
+static void __init msm8x60_map_io(void)
+{
+ msm_map_msm8x60_io();
+}
+
+static void __init msm8x60_init_irq(void)
+{
+ gic_dist_init(0, MSM_QGIC_DIST_BASE, 1);
+ gic_cpu_base_addr = (void *)MSM_QGIC_CPU_BASE;
+ gic_cpu_init(0, MSM_QGIC_CPU_BASE);
+}
+
+static void __init msm8x60_init(void)
+{
+}
+
+MACHINE_START(MSM8X60_RUMI3, "QCT MSM8X60 RUMI3")
+ .map_io = msm8x60_map_io,
+ .init_irq = msm8x60_init_irq,
+ .init_machine = msm8x60_init,
+ .timer = &msm_timer,
+MACHINE_END
--
1.7.2.1
Sent by an employee of the Qualcomm Innovation Center, Inc.
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum.
From: Abhijeet Dharmapurikar <[email protected]>
Define the interrupt map in irq-8x60.h
Signed-off-by: Abhijeet Dharmapurikar <[email protected]>
---
arch/arm/mach-msm/include/mach/irqs-8x60.h | 225 ++++++++++++++++++++++++++++
1 files changed, 225 insertions(+), 0 deletions(-)
diff --git a/arch/arm/mach-msm/include/mach/irqs-8x60.h b/arch/arm/mach-msm/include/mach/irqs-8x60.h
index bef47c9..ab4ad27 100644
--- a/arch/arm/mach-msm/include/mach/irqs-8x60.h
+++ b/arch/arm/mach-msm/include/mach/irqs-8x60.h
@@ -35,6 +35,231 @@
16-31: PPI (private peripheral interrupts)
32+: SPI (shared peripheral interrupts) */
+#define GIC_PPI_START 16
+#define GIC_SPI_START 32
+
+#define INT_DEBUG_TIMER_EXP (GIC_PPI_START + 0)
+#define INT_GP_TIMER_EXP (GIC_PPI_START + 1)
+#define INT_GP_TIMER2_EXP (GIC_PPI_START + 2)
+#define WDT0_ACCSCSSNBARK_INT (GIC_PPI_START + 3)
+#define WDT1_ACCSCSSNBARK_INT (GIC_PPI_START + 4)
+#define AVS_SVICINT (GIC_PPI_START + 5)
+#define AVS_SVICINTSWDONE (GIC_PPI_START + 6)
+#define CPU_DBGCPUXCOMMRXFULL (GIC_PPI_START + 7)
+#define CPU_DBGCPUXCOMMTXEMPTY (GIC_PPI_START + 8)
+#define CPU_SICCPUXPERFMONIRPTREQ (GIC_PPI_START + 9)
+#define SC_AVSCPUXDOWN (GIC_PPI_START + 10)
+#define SC_AVSCPUXUP (GIC_PPI_START + 11)
+#define SC_SICCPUXACGIRPTREQ (GIC_PPI_START + 12)
+/* PPI 13 to 15 are unused */
+
+
+#define SC_SICMPUIRPTREQ (GIC_SPI_START + 0)
+#define SC_SICL2IRPTREQ (GIC_SPI_START + 1)
+#define SC_SICL2ACGIRPTREQ (GIC_SPI_START + 2)
+#define NC (GIC_SPI_START + 3)
+#define TLMM_SCSS_DIR_CONN_IRQ_0 (GIC_SPI_START + 4)
+#define TLMM_SCSS_DIR_CONN_IRQ_1 (GIC_SPI_START + 5)
+#define TLMM_SCSS_DIR_CONN_IRQ_2 (GIC_SPI_START + 6)
+#define TLMM_SCSS_DIR_CONN_IRQ_3 (GIC_SPI_START + 7)
+#define TLMM_SCSS_DIR_CONN_IRQ_4 (GIC_SPI_START + 8)
+#define TLMM_SCSS_DIR_CONN_IRQ_5 (GIC_SPI_START + 9)
+#define TLMM_SCSS_DIR_CONN_IRQ_6 (GIC_SPI_START + 10)
+#define TLMM_SCSS_DIR_CONN_IRQ_7 (GIC_SPI_START + 11)
+#define TLMM_SCSS_DIR_CONN_IRQ_8 (GIC_SPI_START + 12)
+#define TLMM_SCSS_DIR_CONN_IRQ_9 (GIC_SPI_START + 13)
+#define PM8058_SEC_IRQ_N (GIC_SPI_START + 14)
+#define PM8901_SEC_IRQ_N (GIC_SPI_START + 15)
+#define TLMM_SCSS_SUMMARY_IRQ (GIC_SPI_START + 16)
+#define SPDM_RT_1_IRQ (GIC_SPI_START + 17)
+#define SPDM_DIAG_IRQ (GIC_SPI_START + 18)
+#define RPM_SCSS_CPU0_GP_HIGH_IRQ (GIC_SPI_START + 19)
+#define RPM_SCSS_CPU0_GP_MEDIUM_IRQ (GIC_SPI_START + 20)
+#define RPM_SCSS_CPU0_GP_LOW_IRQ (GIC_SPI_START + 21)
+#define RPM_SCSS_CPU0_WAKE_UP_IRQ (GIC_SPI_START + 22)
+#define RPM_SCSS_CPU1_GP_HIGH_IRQ (GIC_SPI_START + 23)
+#define RPM_SCSS_CPU1_GP_MEDIUM_IRQ (GIC_SPI_START + 24)
+#define RPM_SCSS_CPU1_GP_LOW_IRQ (GIC_SPI_START + 25)
+#define RPM_SCSS_CPU1_WAKE_UP_IRQ (GIC_SPI_START + 26)
+#define SSBI2_2_SC_CPU0_SECURE_INT (GIC_SPI_START + 27)
+#define SSBI2_2_SC_CPU0_NON_SECURE_INT (GIC_SPI_START + 28)
+#define SSBI2_1_SC_CPU0_SECURE_INT (GIC_SPI_START + 29)
+#define SSBI2_1_SC_CPU0_NON_SECURE_INT (GIC_SPI_START + 30)
+#define MSMC_SC_SEC_CE_IRQ (GIC_SPI_START + 31)
+#define MSMC_SC_PRI_CE_IRQ (GIC_SPI_START + 32)
+#define MARM_FIQ (GIC_SPI_START + 33)
+#define MARM_IRQ (GIC_SPI_START + 34)
+#define MARM_L2CC_IRQ (GIC_SPI_START + 35)
+#define MARM_WDOG_EXPIRED (GIC_SPI_START + 36)
+#define MARM_SCSS_GP_IRQ_0 (GIC_SPI_START + 37)
+#define MARM_SCSS_GP_IRQ_1 (GIC_SPI_START + 38)
+#define MARM_SCSS_GP_IRQ_2 (GIC_SPI_START + 39)
+#define MARM_SCSS_GP_IRQ_3 (GIC_SPI_START + 40)
+#define MARM_SCSS_GP_IRQ_4 (GIC_SPI_START + 41)
+#define MARM_SCSS_GP_IRQ_5 (GIC_SPI_START + 42)
+#define MARM_SCSS_GP_IRQ_6 (GIC_SPI_START + 43)
+#define MARM_SCSS_GP_IRQ_7 (GIC_SPI_START + 44)
+#define MARM_SCSS_GP_IRQ_8 (GIC_SPI_START + 45)
+#define MARM_SCSS_GP_IRQ_9 (GIC_SPI_START + 46)
+#define VPE_IRQ (GIC_SPI_START + 47)
+#define VFE_IRQ (GIC_SPI_START + 48)
+#define VCODEC_IRQ (GIC_SPI_START + 49)
+#define TV_ENC_IRQ (GIC_SPI_START + 50)
+#define SMMU_VPE_CB_SC_SECURE_IRQ (GIC_SPI_START + 51)
+#define SMMU_VPE_CB_SC_NON_SECURE_IRQ (GIC_SPI_START + 52)
+#define SMMU_VFE_CB_SC_SECURE_IRQ (GIC_SPI_START + 53)
+#define SMMU_VFE_CB_SC_NON_SECURE_IRQ (GIC_SPI_START + 54)
+#define SMMU_VCODEC_B_CB_SC_SECURE_IRQ (GIC_SPI_START + 55)
+#define SMMU_VCODEC_B_CB_SC_NON_SECURE_IRQ (GIC_SPI_START + 56)
+#define SMMU_VCODEC_A_CB_SC_SECURE_IRQ (GIC_SPI_START + 57)
+#define SMMU_VCODEC_A_CB_SC_NON_SECURE_IRQ (GIC_SPI_START + 58)
+#define SMMU_ROT_CB_SC_SECURE_IRQ (GIC_SPI_START + 59)
+#define SMMU_ROT_CB_SC_NON_SECURE_IRQ (GIC_SPI_START + 60)
+#define SMMU_MDP1_CB_SC_SECURE_IRQ (GIC_SPI_START + 61)
+#define SMMU_MDP1_CB_SC_NON_SECURE_IRQ (GIC_SPI_START + 62)
+#define SMMU_MDP0_CB_SC_SECURE_IRQ (GIC_SPI_START + 63)
+#define SMMU_MDP0_CB_SC_NON_SECURE_IRQ (GIC_SPI_START + 64)
+#define SMMU_JPEGD_CB_SC_SECURE_IRQ (GIC_SPI_START + 65)
+#define SMMU_JPEGD_CB_SC_NON_SECURE_IRQ (GIC_SPI_START + 66)
+#define SMMU_IJPEG_CB_SC_SECURE_IRQ (GIC_SPI_START + 67)
+#define SMMU_IJPEG_CB_SC_NON_SECURE_IRQ (GIC_SPI_START + 68)
+#define SMMU_GFX3D_CB_SC_SECURE_IRQ (GIC_SPI_START + 69)
+#define SMMU_GFX3D_CB_SC_NON_SECURE_IRQ (GIC_SPI_START + 70)
+#define SMMU_GFX2D0_CB_SC_SECURE_IRQ (GIC_SPI_START + 71)
+#define SMMU_GFX2D0_CB_SC_NON_SECURE_IRQ (GIC_SPI_START + 72)
+#define ROT_IRQ (GIC_SPI_START + 73)
+#define MMSS_FABRIC_IRQ (GIC_SPI_START + 74)
+#define MDP_IRQ (GIC_SPI_START + 75)
+#define JPEGD_IRQ (GIC_SPI_START + 76)
+#define JPEG_IRQ (GIC_SPI_START + 77)
+#define MMSS_IMEM_IRQ (GIC_SPI_START + 78)
+#define HDMI_IRQ (GIC_SPI_START + 79)
+#define GFX3D_IRQ (GIC_SPI_START + 80)
+#define GFX2D0_IRQ (GIC_SPI_START + 81)
+#define DSI_IRQ (GIC_SPI_START + 82)
+#define CSI_1_IRQ (GIC_SPI_START + 83)
+#define CSI_0_IRQ (GIC_SPI_START + 84)
+#define LPASS_SCSS_AUDIO_IF_OUT0_IRQ (GIC_SPI_START + 85)
+#define LPASS_SCSS_MIDI_IRQ (GIC_SPI_START + 86)
+#define LPASS_Q6SS_WDOG_EXPIRED (GIC_SPI_START + 87)
+#define LPASS_SCSS_GP_LOW_IRQ (GIC_SPI_START + 88)
+#define LPASS_SCSS_GP_MEDIUM_IRQ (GIC_SPI_START + 89)
+#define LPASS_SCSS_GP_HIGH_IRQ (GIC_SPI_START + 90)
+#define TOP_IMEM_IRQ (GIC_SPI_START + 91)
+#define FABRIC_SYS_IRQ (GIC_SPI_START + 92)
+#define FABRIC_APPS_IRQ (GIC_SPI_START + 93)
+#define USB1_HS_BAM_IRQ (GIC_SPI_START + 94)
+#define SDC4_BAM_IRQ (GIC_SPI_START + 95)
+#define SDC3_BAM_IRQ (GIC_SPI_START + 96)
+#define SDC2_BAM_IRQ (GIC_SPI_START + 97)
+#define SDC1_BAM_IRQ (GIC_SPI_START + 98)
+#define FABRIC_SPS_IRQ (GIC_SPI_START + 99)
+#define USB1_HS_IRQ (GIC_SPI_START + 100)
+#define SDC4_IRQ_0 (GIC_SPI_START + 101)
+#define SDC3_IRQ_0 (GIC_SPI_START + 102)
+#define SDC2_IRQ_0 (GIC_SPI_START + 103)
+#define SDC1_IRQ_0 (GIC_SPI_START + 104)
+#define SPS_BAM_DMA_IRQ (GIC_SPI_START + 105)
+#define SPS_SEC_VIOL_IRQ (GIC_SPI_START + 106)
+#define SPS_MTI_0 (GIC_SPI_START + 107)
+#define SPS_MTI_1 (GIC_SPI_START + 108)
+#define SPS_MTI_2 (GIC_SPI_START + 109)
+#define SPS_MTI_3 (GIC_SPI_START + 110)
+#define SPS_MTI_4 (GIC_SPI_START + 111)
+#define SPS_MTI_5 (GIC_SPI_START + 112)
+#define SPS_MTI_6 (GIC_SPI_START + 113)
+#define SPS_MTI_7 (GIC_SPI_START + 114)
+#define SPS_MTI_8 (GIC_SPI_START + 115)
+#define SPS_MTI_9 (GIC_SPI_START + 116)
+#define SPS_MTI_10 (GIC_SPI_START + 117)
+#define SPS_MTI_11 (GIC_SPI_START + 118)
+#define SPS_MTI_12 (GIC_SPI_START + 119)
+#define SPS_MTI_13 (GIC_SPI_START + 120)
+#define SPS_MTI_14 (GIC_SPI_START + 121)
+#define SPS_MTI_15 (GIC_SPI_START + 122)
+#define SPS_MTI_16 (GIC_SPI_START + 123)
+#define SPS_MTI_17 (GIC_SPI_START + 124)
+#define SPS_MTI_18 (GIC_SPI_START + 125)
+#define SPS_MTI_19 (GIC_SPI_START + 126)
+#define SPS_MTI_20 (GIC_SPI_START + 127)
+#define SPS_MTI_21 (GIC_SPI_START + 128)
+#define SPS_MTI_22 (GIC_SPI_START + 129)
+#define SPS_MTI_23 (GIC_SPI_START + 130)
+#define SPS_MTI_24 (GIC_SPI_START + 131)
+#define SPS_MTI_25 (GIC_SPI_START + 132)
+#define SPS_MTI_26 (GIC_SPI_START + 133)
+#define SPS_MTI_27 (GIC_SPI_START + 134)
+#define SPS_MTI_28 (GIC_SPI_START + 135)
+#define SPS_MTI_29 (GIC_SPI_START + 136)
+#define SPS_MTI_30 (GIC_SPI_START + 137)
+#define SPS_MTI_31 (GIC_SPI_START + 138)
+#define UXMC_EBI2_WR_ER_DONE_IRQ (GIC_SPI_START + 139)
+#define UXMC_EBI2_OP_DONE_IRQ (GIC_SPI_START + 140)
+#define USB2_IRQ (GIC_SPI_START + 141)
+#define USB1_IRQ (GIC_SPI_START + 142)
+#define TSSC_SSBI_IRQ (GIC_SPI_START + 143)
+#define TSSC_SAMPLE_IRQ (GIC_SPI_START + 144)
+#define TSSC_PENUP_IRQ (GIC_SPI_START + 145)
+#define INT_UART1DM_IRQ (GIC_SPI_START + 146)
+#define GSBI1_QUP_IRQ (GIC_SPI_START + 147)
+#define INT_UART2DM_IRQ (GIC_SPI_START + 148)
+#define GSBI2_QUP_IRQ (GIC_SPI_START + 149)
+#define INT_UART3DM_IRQ (GIC_SPI_START + 150)
+#define GSBI3_QUP_IRQ (GIC_SPI_START + 151)
+#define INT_UART4DM_IRQ (GIC_SPI_START + 152)
+#define GSBI4_QUP_IRQ (GIC_SPI_START + 153)
+#define INT_UART5DM_IRQ (GIC_SPI_START + 154)
+#define GSBI5_QUP_IRQ (GIC_SPI_START + 155)
+#define INT_UART6DM_IRQ (GIC_SPI_START + 156)
+#define GSBI6_QUP_IRQ (GIC_SPI_START + 157)
+#define INT_UART7DM_IRQ (GIC_SPI_START + 158)
+#define GSBI7_QUP_IRQ (GIC_SPI_START + 159)
+#define INT_UART8DM_IRQ (GIC_SPI_START + 160)
+#define GSBI8_QUP_IRQ (GIC_SPI_START + 161)
+#define TSIF_TSPP_IRQ (GIC_SPI_START + 162)
+#define TSIF_BAM_IRQ (GIC_SPI_START + 163)
+#define TSIF2_IRQ (GIC_SPI_START + 164)
+#define TSIF1_IRQ (GIC_SPI_START + 165)
+#define INT_ADM1_MASTER (GIC_SPI_START + 166)
+#define INT_ADM1_AARM (GIC_SPI_START + 167)
+#define INT_ADM1_SD2 (GIC_SPI_START + 168)
+#define INT_ADM1_SD3 (GIC_SPI_START + 169)
+#define INT_ADM0_MASTER (GIC_SPI_START + 170)
+#define INT_ADM0_AARM (GIC_SPI_START + 171)
+#define INT_ADM0_SD2 (GIC_SPI_START + 172)
+#define INT_ADM0_SD3 (GIC_SPI_START + 173)
+#define CC_SCSS_WDT1CPU1BITEEXPIRED (GIC_SPI_START + 174)
+#define CC_SCSS_WDT1CPU0BITEEXPIRED (GIC_SPI_START + 175)
+#define CC_SCSS_WDT0CPU1BITEEXPIRED (GIC_SPI_START + 176)
+#define CC_SCSS_WDT0CPU0BITEEXPIRED (GIC_SPI_START + 177)
+#define TSENS_UPPER_LOWER_INT (GIC_SPI_START + 178)
+#define SSBI2_2_SC_CPU1_SECURE_INT (GIC_SPI_START + 179)
+#define SSBI2_2_SC_CPU1_NON_SECURE_INT (GIC_SPI_START + 180)
+#define SSBI2_1_SC_CPU1_SECURE_INT (GIC_SPI_START + 181)
+#define SSBI2_1_SC_CPU1_NON_SECURE_INT (GIC_SPI_START + 182)
+#define XPU_SUMMARY_IRQ (GIC_SPI_START + 183)
+#define BUS_EXCEPTION_SUMMARY_IRQ (GIC_SPI_START + 184)
+#define HSDDRX_SMICH0_IRQ (GIC_SPI_START + 185)
+#define HSDDRX_EBI1_IRQ (GIC_SPI_START + 186)
+#define SDC5_BAM_IRQ (GIC_SPI_START + 187)
+#define SDC5_IRQ_0 (GIC_SPI_START + 188)
+#define INT_UART9DM_IRQ (GIC_SPI_START + 189)
+#define GSBI9_QUP_IRQ (GIC_SPI_START + 190)
+#define INT_UART10DM_IRQ (GIC_SPI_START + 191)
+#define GSBI10_QUP_IRQ (GIC_SPI_START + 192)
+#define INT_UART11DM_IRQ (GIC_SPI_START + 193)
+#define GSBI11_QUP_IRQ (GIC_SPI_START + 194)
+#define INT_UART12DM_IRQ (GIC_SPI_START + 195)
+#define GSBI12_QUP_IRQ (GIC_SPI_START + 196)
+/*SPI 197 to 216 arent used in 8x60*/
+#define SMPSS_SPARE_1 (GIC_SPI_START + 217)
+#define SMPSS_SPARE_2 (GIC_SPI_START + 218)
+#define SMPSS_SPARE_3 (GIC_SPI_START + 219)
+#define SMPSS_SPARE_4 (GIC_SPI_START + 220)
+#define SMPSS_SPARE_5 (GIC_SPI_START + 221)
+#define SMPSS_SPARE_6 (GIC_SPI_START + 222)
+#define SMPSS_SPARE_7 (GIC_SPI_START + 223)
+
#define NR_GPIO_IRQS 173
#define NR_MSM_IRQS 256
#define NR_BOARD_IRQS 0
--
1.7.2.1
Sent by an employee of the Qualcomm Innovation Center, Inc.
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum.
Signed-off-by: Jeff Ohlstein <[email protected]>
---
arch/arm/mach-msm/include/mach/msm_iomap-8x60.h | 10 +-
arch/arm/mach-msm/timer.c | 226 +++++++++++++++++------
2 files changed, 177 insertions(+), 59 deletions(-)
diff --git a/arch/arm/mach-msm/include/mach/msm_iomap-8x60.h b/arch/arm/mach-msm/include/mach/msm_iomap-8x60.h
index ee58da5..1f15bbb 100644
--- a/arch/arm/mach-msm/include/mach/msm_iomap-8x60.h
+++ b/arch/arm/mach-msm/include/mach/msm_iomap-8x60.h
@@ -35,7 +35,6 @@
*
*/
-
#define MSM_QGIC_DIST_BASE IOMEM(0xF0000000)
#define MSM_QGIC_DIST_PHYS 0x02080000
#define MSM_QGIC_DIST_SIZE SZ_4K
@@ -56,7 +55,14 @@
#define MSM_TLMM_PHYS 0x00800000
#define MSM_TLMM_SIZE SZ_16K
-#define MSM_SHARED_RAM_BASE IOMEM(0xF0100000)
+#define MSM_TMR_BASE IOMEM(0xF0100000)
+#define MSM_TMR_PHYS 0x02000000
+#define MSM_TMR_SIZE (SZ_1M)
+
+#define MSM_GPT_BASE (MSM_TMR_BASE + 0x4)
+#define MSM_DGT_BASE (MSM_TMR_BASE + 0x24)
+
+#define MSM_SHARED_RAM_BASE IOMEM(0xF0200000)
#define MSM_SHARED_RAM_SIZE SZ_1M
diff --git a/arch/arm/mach-msm/timer.c b/arch/arm/mach-msm/timer.c
index dec5ca6..e76d869 100644
--- a/arch/arm/mach-msm/timer.c
+++ b/arch/arm/mach-msm/timer.c
@@ -1,6 +1,6 @@
-/* linux/arch/arm/mach-msm/timer.c
- *
+/*
* Copyright (C) 2007 Google, Inc.
+ * Copyright (c) 2009-2010, Code Aurora Forum. All rights reserved.
*
* This software is licensed under the terms of the GNU General Public
* License version 2, as published by the Free Software Foundation, and
@@ -21,6 +21,7 @@
#include <linux/clockchips.h>
#include <linux/delay.h>
#include <linux/io.h>
+#include <linux/percpu.h>
#include <asm/mach/time.h>
#include <mach/msm_iomap.h>
@@ -28,7 +29,20 @@
#ifndef MSM_DGT_BASE
#define MSM_DGT_BASE (MSM_GPT_BASE + 0x10)
#endif
+
+#ifdef CONFIG_MSM7X00A_USE_GP_TIMER
+ #define DG_TIMER_RATING 100
+ #define MSM_GLOBAL_TIMER MSM_CLOCK_GPT
+#else
+ #define DG_TIMER_RATING 300
+ #define MSM_GLOBAL_TIMER MSM_CLOCK_DGT
+#endif
+
+#if defined(CONFIG_ARCH_MSM_ARM11)
#define MSM_DGT_SHIFT (5)
+#else
+#define MSM_DGT_SHIFT (0)
+#endif
#define TIMER_MATCH_VAL 0x0000
#define TIMER_COUNT_VAL 0x0004
@@ -40,8 +54,32 @@
#define CSR_PROTECTION 0x0020
#define CSR_PROTECTION_EN 1
+#define LOCAL_TIMER 0
+#define GLOBAL_TIMER 1
+
+#ifdef CONFIG_ARCH_MSM8X60
+#define MSM_TMR_BASE_CPU0 0x40000
+#else
+#define MSM_TMR_BASE_CPU0 0
+#endif
+
+#define NR_TIMERS ARRAY_SIZE(msm_clocks)
+
#define GPT_HZ 32768
-#define DGT_HZ 19200000 /* 19.2 MHz or 600 KHz after shift */
+
+#if defined(CONFIG_ARCH_QSD8X50) || defined(CONFIG_ARCH_MSM8X60)
+#define DGT_HZ 4800000 /* Uses TCXO/4 (19.2 MHz / 4) */
+#else
+#define DGT_HZ 19200000 /* Uses TCXO (19.2 MHz) */
+#endif
+
+static irqreturn_t msm_timer_interrupt(int irq, void *dev_id);
+static cycle_t msm_gpt_read(struct clocksource *cs);
+static cycle_t msm_dgt_read(struct clocksource *cs);
+static void msm_timer_set_mode(enum clock_event_mode mode,
+ struct clock_event_device *evt);
+static int msm_timer_set_next_event(unsigned long cycles,
+ struct clock_event_device *evt);
struct msm_clock {
struct clock_event_device clockevent;
@@ -52,60 +90,10 @@ struct msm_clock {
uint32_t shift;
};
-static irqreturn_t msm_timer_interrupt(int irq, void *dev_id)
-{
- struct clock_event_device *evt = dev_id;
- evt->event_handler(evt);
- return IRQ_HANDLED;
-}
-
-static cycle_t msm_gpt_read(struct clocksource *cs)
-{
- return readl(MSM_GPT_BASE + TIMER_COUNT_VAL);
-}
-
-static cycle_t msm_dgt_read(struct clocksource *cs)
-{
- return readl(MSM_DGT_BASE + TIMER_COUNT_VAL) >> MSM_DGT_SHIFT;
-}
-
-static int msm_timer_set_next_event(unsigned long cycles,
- struct clock_event_device *evt)
-{
- struct msm_clock *clock = container_of(evt, struct msm_clock, clockevent);
- uint32_t now = readl(clock->regbase + TIMER_COUNT_VAL);
- uint32_t alarm = now + (cycles << clock->shift);
- int late;
-
- writel(alarm, clock->regbase + TIMER_MATCH_VAL);
- now = readl(clock->regbase + TIMER_COUNT_VAL);
- late = now - alarm;
- if (late >= (-2 << clock->shift) && late < DGT_HZ*5) {
- printk(KERN_NOTICE "msm_timer_set_next_event(%lu) clock %s, "
- "alarm already expired, now %x, alarm %x, late %d\n",
- cycles, clock->clockevent.name, now, alarm, late);
- return -ETIME;
- }
- return 0;
-}
-
-static void msm_timer_set_mode(enum clock_event_mode mode,
- struct clock_event_device *evt)
-{
- struct msm_clock *clock = container_of(evt, struct msm_clock, clockevent);
- switch (mode) {
- case CLOCK_EVT_MODE_RESUME:
- case CLOCK_EVT_MODE_PERIODIC:
- break;
- case CLOCK_EVT_MODE_ONESHOT:
- writel(TIMER_ENABLE_EN, clock->regbase + TIMER_ENABLE);
- break;
- case CLOCK_EVT_MODE_UNUSED:
- case CLOCK_EVT_MODE_SHUTDOWN:
- writel(0, clock->regbase + TIMER_ENABLE);
- break;
- }
-}
+enum {
+ MSM_CLOCK_GPT,
+ MSM_CLOCK_DGT,
+};
static struct msm_clock msm_clocks[] = {
{
@@ -165,6 +153,89 @@ static struct msm_clock msm_clocks[] = {
}
};
+static struct clock_event_device *local_clock_event;
+
+static irqreturn_t msm_timer_interrupt(int irq, void *dev_id)
+{
+ struct clock_event_device *evt = dev_id;
+ if (smp_processor_id() != 0)
+ evt = local_clock_event;
+ if (evt->event_handler == NULL)
+ return IRQ_HANDLED;
+ evt->event_handler(evt);
+ return IRQ_HANDLED;
+}
+
+static cycle_t msm_gpt_read(struct clocksource *cs)
+{
+ struct msm_clock *clock =
+ container_of(cs, struct msm_clock, clocksource);
+ return readl(clock->regbase + TIMER_COUNT_VAL + MSM_TMR_BASE_CPU0);
+}
+
+static cycle_t msm_dgt_read(struct clocksource *cs)
+{
+ struct msm_clock *clock =
+ container_of(cs, struct msm_clock, clocksource);
+ return readl(clock->regbase + TIMER_COUNT_VAL + MSM_TMR_BASE_CPU0)
+ >> MSM_DGT_SHIFT;
+}
+
+static int msm_timer_set_next_event(unsigned long cycles,
+ struct clock_event_device *evt)
+{
+ struct msm_clock *clock;
+ uint32_t now;
+ uint32_t alarm;
+ int late;
+
+#ifdef CONFIG_SMP
+ clock = &msm_clocks[MSM_GLOBAL_TIMER];
+#else
+ clock = container_of(evt, struct msm_clock, clockevent);
+#endif
+ now = readl(clock->regbase + TIMER_COUNT_VAL);
+ alarm = now + (cycles << clock->shift);
+ writel(alarm, clock->regbase + TIMER_MATCH_VAL);
+ late = now - alarm;
+ if (late >= (-2 << clock->shift) && late < DGT_HZ*5) {
+ printk(KERN_NOTICE "msm_timer_set_next_event(%lu) clock %s, "
+ "alarm already expired, now %x, alarm %x, late %d\n",
+ cycles, clock->clockevent.name, now, alarm, late);
+ return -ETIME;
+ }
+ return 0;
+}
+
+static void msm_timer_set_mode(enum clock_event_mode mode,
+ struct clock_event_device *evt)
+{
+ struct msm_clock *clock;
+ unsigned long irq_flags;
+
+#ifdef CONFIG_SMP
+ clock = &msm_clocks[MSM_GLOBAL_TIMER];
+#else
+ clock = container_of(evt, struct msm_clock, clockevent);
+#endif
+ local_irq_save(irq_flags);
+
+ switch (mode) {
+ case CLOCK_EVT_MODE_RESUME:
+ case CLOCK_EVT_MODE_PERIODIC:
+ break;
+ case CLOCK_EVT_MODE_ONESHOT:
+ writel(TIMER_ENABLE_EN, clock->regbase + TIMER_ENABLE);
+ break;
+ case CLOCK_EVT_MODE_UNUSED:
+ case CLOCK_EVT_MODE_SHUTDOWN:
+ writel(0, clock->regbase + TIMER_ENABLE);
+ break;
+ }
+
+ local_irq_restore(irq_flags);
+}
+
static void __init msm_timer_init(void)
{
int i;
@@ -201,6 +272,47 @@ static void __init msm_timer_init(void)
}
}
+#ifdef CONFIG_SMP
+void local_timer_setup(struct clock_event_device *evt)
+{
+ unsigned long flags;
+ struct msm_clock *clock = &msm_clocks[MSM_GLOBAL_TIMER];
+
+ if (!local_clock_event) {
+ writel(0, clock->regbase + TIMER_ENABLE);
+ writel(1, clock->regbase + TIMER_CLEAR);
+ writel(0, clock->regbase + TIMER_COUNT_VAL);
+ writel(~0, clock->regbase + TIMER_MATCH_VAL);
+ }
+ evt->irq = clock->irq.irq;
+ evt->name = "local_timer";
+ evt->features = CLOCK_EVT_FEAT_ONESHOT;
+ evt->rating = clock->clockevent.rating;
+ evt->set_mode = msm_timer_set_mode;
+ evt->set_next_event = msm_timer_set_next_event;
+ evt->shift = clock->clockevent.shift;
+ evt->mult = div_sc(clock->freq, NSEC_PER_SEC, evt->shift);
+ evt->max_delta_ns =
+ clockevent_delta2ns(0xf0000000 >> clock->shift, evt);
+ evt->min_delta_ns = clockevent_delta2ns(4, evt);
+ evt->cpumask = cpumask_of(smp_processor_id());
+
+ local_clock_event = evt;
+
+ local_irq_save(flags);
+ get_irq_chip(clock->irq.irq)->unmask(clock->irq.irq);
+ local_irq_restore(flags);
+
+ clockevents_register_device(evt);
+}
+
+int local_timer_ack(void)
+{
+ return 1;
+}
+
+#endif
+
struct sys_timer msm_timer = {
.init = msm_timer_init
};
--
1.7.2.1
Sent by an employee of the Qualcomm Innovation Center, Inc.
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum.
From: Steve Muckle <[email protected]>
ScorpionMP is a configuration of Scorpion processors. It supports local
timers.
Signed-off-by: Steve Muckle <[email protected]>
Signed-off-by: Jeff Ohlstein <[email protected]>
---
arch/arm/Kconfig | 3 ++-
arch/arm/mach-msm/Kconfig | 2 ++
2 files changed, 4 insertions(+), 1 deletions(-)
diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig
index 8036e46..d4d9f55 100644
--- a/arch/arm/Kconfig
+++ b/arch/arm/Kconfig
@@ -1173,7 +1173,8 @@ config HOTPLUG_CPU
config LOCAL_TIMERS
bool "Use local timer interrupts"
depends on SMP && (REALVIEW_EB_ARM11MP || MACH_REALVIEW_PB11MP || \
- REALVIEW_EB_A9MP || MACH_REALVIEW_PBX || ARCH_OMAP4 || ARCH_U8500)
+ REALVIEW_EB_A9MP || MACH_REALVIEW_PBX || ARCH_OMAP4 || ARCH_U8500 || \
+ ARCH_MSM_SCORPIONMP)
default y
select HAVE_ARM_TWD if (ARCH_REALVIEW || ARCH_OMAP4 || ARCH_U8500)
help
diff --git a/arch/arm/mach-msm/Kconfig b/arch/arm/mach-msm/Kconfig
index 08ed76c..3b9bece 100644
--- a/arch/arm/mach-msm/Kconfig
+++ b/arch/arm/mach-msm/Kconfig
@@ -39,6 +39,8 @@ config ARCH_MSM_ARM11
bool
config ARCH_MSM_SCORPION
bool
+config ARCH_MSM_SCORPIONMP
+ bool
config MSM_VIC
bool
--
1.7.2.1
Sent by an employee of the Qualcomm Innovation Center, Inc.
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum.
From: Steve Muckle <[email protected]>
MSM8X60 has different IO mappings than previous MSMs.
Signed-off-by: Steve Muckle <[email protected]>
---
arch/arm/mach-msm/include/mach/io.h | 1 +
arch/arm/mach-msm/include/mach/msm_iomap-8x60.h | 21 +++++++++++++++++++++
arch/arm/mach-msm/io.c | 15 +++++++++++++++
3 files changed, 37 insertions(+), 0 deletions(-)
diff --git a/arch/arm/mach-msm/include/mach/io.h b/arch/arm/mach-msm/include/mach/io.h
index c35b29f..7386e73 100644
--- a/arch/arm/mach-msm/include/mach/io.h
+++ b/arch/arm/mach-msm/include/mach/io.h
@@ -28,6 +28,7 @@ void __iomem *__msm_ioremap(unsigned long phys_addr, size_t size, unsigned int m
void msm_map_qsd8x50_io(void);
void msm_map_msm7x30_io(void);
+void msm_map_msm8x60_io(void);
extern unsigned int msm_shared_ram_phys;
diff --git a/arch/arm/mach-msm/include/mach/msm_iomap-8x60.h b/arch/arm/mach-msm/include/mach/msm_iomap-8x60.h
index 17209f7..ee58da5 100644
--- a/arch/arm/mach-msm/include/mach/msm_iomap-8x60.h
+++ b/arch/arm/mach-msm/include/mach/msm_iomap-8x60.h
@@ -35,8 +35,29 @@
*
*/
+
+#define MSM_QGIC_DIST_BASE IOMEM(0xF0000000)
+#define MSM_QGIC_DIST_PHYS 0x02080000
+#define MSM_QGIC_DIST_SIZE SZ_4K
+
+#define MSM_QGIC_CPU_BASE IOMEM(0xF0001000)
+#define MSM_QGIC_CPU_PHYS 0x02081000
+#define MSM_QGIC_CPU_SIZE SZ_4K
+
+#define MSM_ACC_BASE IOMEM(0xF0002000)
+#define MSM_ACC_PHYS 0x02001000
+#define MSM_ACC_SIZE SZ_4K
+
+#define MSM_GCC_BASE IOMEM(0xF0003000)
+#define MSM_GCC_PHYS 0x02082000
+#define MSM_GCC_SIZE SZ_4K
+
#define MSM_TLMM_BASE IOMEM(0xF0004000)
#define MSM_TLMM_PHYS 0x00800000
#define MSM_TLMM_SIZE SZ_16K
+#define MSM_SHARED_RAM_BASE IOMEM(0xF0100000)
+#define MSM_SHARED_RAM_SIZE SZ_1M
+
+
#endif
diff --git a/arch/arm/mach-msm/io.c b/arch/arm/mach-msm/io.c
index 1c05060..d36b610 100644
--- a/arch/arm/mach-msm/io.c
+++ b/arch/arm/mach-msm/io.c
@@ -100,6 +100,21 @@ void __init msm_map_qsd8x50_io(void)
}
#endif /* CONFIG_ARCH_QSD8X50 */
+#ifdef CONFIG_ARCH_MSM8X60
+static struct map_desc msm8x60_io_desc[] __initdata = {
+ MSM_DEVICE(QGIC_DIST),
+ MSM_DEVICE(QGIC_CPU),
+ MSM_DEVICE(TMR),
+ MSM_DEVICE(ACC),
+ MSM_DEVICE(GCC),
+};
+
+void __init msm_map_msm8x60_io(void)
+{
+ iotable_init(msm8x60_io_desc, ARRAY_SIZE(msm8x60_io_desc));
+}
+#endif /* CONFIG_ARCH_MSM8X60 */
+
#ifdef CONFIG_ARCH_MSM7X30
static struct map_desc msm7x30_io_desc[] __initdata = {
MSM_DEVICE(VIC),
--
1.7.2.1
Sent by an employee of the Qualcomm Innovation Center, Inc.
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum.
From: Daniel Walker <[email protected]>
ScorpionMP does not have the SMP/nAMP and TLB ops broadcasting bits in
ACTLR.
Signed-off-by: Daniel Walker <[email protected]>
Signed-off-by: Jeff Ohlstein <[email protected]>
---
arch/arm/mm/proc-v7.S | 24 ++++++++++++++++++++++++
1 files changed, 24 insertions(+), 0 deletions(-)
diff --git a/arch/arm/mm/proc-v7.S b/arch/arm/mm/proc-v7.S
index 7aaf88a..98fd7e5 100644
--- a/arch/arm/mm/proc-v7.S
+++ b/arch/arm/mm/proc-v7.S
@@ -196,6 +196,7 @@ __v7_setup:
orreq r0, r0, #(1 << 6) | (1 << 0) @ Enable SMP/nAMP mode and
mcreq p15, 0, r0, c1, c0, 1 @ TLB ops broadcasting
#endif
+__v7_msm_setup:
adr r12, __v7_setup_stack @ the local stack
stmia r12, {r0-r5, r7, r9, r11, lr}
bl v7_flush_dcache_all
@@ -326,6 +327,29 @@ cpu_elf_name:
.section ".proc.info.init", #alloc, #execinstr
+ .type __v7_msm_proc_info, #object
+__v7_msm_proc_info:
+ .long 0x510f0000 @ Required ID value
+ .long 0xff0f0000 @ Mask for ID
+ .long PMD_TYPE_SECT | \
+ PMD_SECT_AP_WRITE | \
+ PMD_SECT_AP_READ | \
+ PMD_FLAGS
+ .long PMD_TYPE_SECT | \
+ PMD_SECT_XN | \
+ PMD_SECT_AP_WRITE | \
+ PMD_SECT_AP_READ
+ b __v7_msm_setup
+ .long cpu_arch_name
+ .long cpu_elf_name
+ .long HWCAP_SWP|HWCAP_HALF|HWCAP_THUMB|HWCAP_FAST_MULT|HWCAP_EDSP
+ .long cpu_v7_name
+ .long v7_processor_functions
+ .long v7wbi_tlb_fns
+ .long v6_user_fns
+ .long v7_cache_fns
+ .size __v7_msm_proc_info, . - __v7_msm_proc_info
+
/*
* Match any ARMv7 processor core.
*/
--
1.7.2.1
Sent by an employee of the Qualcomm Innovation Center, Inc.
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum.
From: Steve Muckle <[email protected]>
IRQ assignments are different for MSM8X60 than other existing MSMs.
Signed-off-by: Steve Muckle <[email protected]>
---
arch/arm/mach-msm/include/mach/irqs-8x60.h | 42 ++++++++++++++++++++++++++++
arch/arm/mach-msm/include/mach/irqs.h | 2 +
2 files changed, 44 insertions(+), 0 deletions(-)
create mode 100644 arch/arm/mach-msm/include/mach/irqs-8x60.h
diff --git a/arch/arm/mach-msm/include/mach/irqs-8x60.h b/arch/arm/mach-msm/include/mach/irqs-8x60.h
new file mode 100644
index 0000000..bef47c9
--- /dev/null
+++ b/arch/arm/mach-msm/include/mach/irqs-8x60.h
@@ -0,0 +1,42 @@
+/* Copyright (c) 2010 Code Aurora Forum. All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions are met:
+ * * Redistributions of source code must retain the above copyright
+ * notice, this list of conditions and the following disclaimer.
+ * * Redistributions in binary form must reproduce the above copyright
+ * notice, this list of conditions and the following disclaimer in the
+ * documentation and/or other materials provided with the distribution.
+ * * Neither the name of Code Aurora nor
+ * the names of its contributors may be used to endorse or promote
+ * products derived from this software without specific prior written
+ * permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NON-INFRINGEMENT ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR
+ * CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
+ * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
+ * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS;
+ * OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,
+ * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR
+ * OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF
+ * ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ *
+ */
+
+#ifndef __ASM_ARCH_MSM_IRQS_8X60_H
+#define __ASM_ARCH_MSM_IRQS_8X60_H
+
+/* MSM ACPU Interrupt Numbers */
+
+/* 0-15: STI/SGI (software triggered/generated interrupts)
+ 16-31: PPI (private peripheral interrupts)
+ 32+: SPI (shared peripheral interrupts) */
+
+#define NR_GPIO_IRQS 173
+#define NR_MSM_IRQS 256
+#define NR_BOARD_IRQS 0
+
+#endif
diff --git a/arch/arm/mach-msm/include/mach/irqs.h b/arch/arm/mach-msm/include/mach/irqs.h
index 164d355..8679a45 100644
--- a/arch/arm/mach-msm/include/mach/irqs.h
+++ b/arch/arm/mach-msm/include/mach/irqs.h
@@ -24,6 +24,8 @@
#elif defined(CONFIG_ARCH_QSD8X50)
#include "irqs-8x50.h"
#include "sirc.h"
+#elif defined(CONFIG_ARCH_MSM8X60)
+#include "irqs-8x60.h"
#elif defined(CONFIG_ARCH_MSM_ARM11)
#include "irqs-7x00.h"
#else
--
1.7.2.1
Sent by an employee of the Qualcomm Innovation Center, Inc.
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum.
From: Abhijeet Dharmapurikar <[email protected]>
Unless gic_ack_irq is called from __do_IRQ, interrupt should not
be disabled in the ack function.
Signed-off-by: Abhijeet Dharmapurikar <[email protected]>
Signed-off-by: Jeff Ohlstein <[email protected]>
---
arch/arm/common/gic.c | 14 --------------
1 files changed, 0 insertions(+), 14 deletions(-)
diff --git a/arch/arm/common/gic.c b/arch/arm/common/gic.c
index 337741f..7d8ca5a 100644
--- a/arch/arm/common/gic.c
+++ b/arch/arm/common/gic.c
@@ -67,25 +67,11 @@ static inline unsigned int gic_irq(unsigned int irq)
/*
* Routines to acknowledge, disable and enable interrupts
- *
- * Linux assumes that when we're done with an interrupt we need to
- * unmask it, in the same way we need to unmask an interrupt when
- * we first enable it.
- *
- * The GIC has a separate notion of "end of interrupt" to re-enable
- * an interrupt after handling, in order to support hardware
- * prioritisation.
- *
- * We can make the GIC behave in the way that Linux expects by making
- * our "acknowledge" routine disable the interrupt, then mark it as
- * complete.
*/
static void gic_ack_irq(unsigned int irq)
{
- u32 mask = 1 << (irq % 32);
spin_lock(&irq_controller_lock);
- writel(mask, gic_dist_base(irq) + GIC_DIST_ENABLE_CLEAR + (gic_irq(irq) / 32) * 4);
writel(gic_irq(irq), gic_cpu_base(irq) + GIC_CPU_EOI);
spin_unlock(&irq_controller_lock);
}
--
1.7.2.1
Sent by an employee of the Qualcomm Innovation Center, Inc.
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum.
On Wed, 2010-08-25 at 05:57 +0100, Jeff Ohlstein wrote:
> From: Daniel Walker <[email protected]>
>
> ScorpionMP does not have the SMP/nAMP and TLB ops broadcasting bits in
> ACTLR.
>
> Signed-off-by: Daniel Walker <[email protected]>
> Signed-off-by: Jeff Ohlstein <[email protected]>
> ---
> arch/arm/mm/proc-v7.S | 24 ++++++++++++++++++++++++
> 1 files changed, 24 insertions(+), 0 deletions(-)
>
> diff --git a/arch/arm/mm/proc-v7.S b/arch/arm/mm/proc-v7.S
> index 7aaf88a..98fd7e5 100644
> --- a/arch/arm/mm/proc-v7.S
> +++ b/arch/arm/mm/proc-v7.S
> @@ -196,6 +196,7 @@ __v7_setup:
> orreq r0, r0, #(1 << 6) | (1 << 0) @ Enable SMP/nAMP mode and
> mcreq p15, 0, r0, c1, c0, 1 @ TLB ops broadcasting
> #endif
> +__v7_msm_setup:
> adr r12, __v7_setup_stack @ the local stack
> stmia r12, {r0-r5, r7, r9, r11, lr}
> bl v7_flush_dcache_all
It may be better to move the generic __v7_setup here and have a specific
entry for ARM Ltd cores (mask 0xff0f0000 and value 0x410f0000).
In general, I would have liked a different way (still ID-based) of
handling such issues without duplicating the whole structure. But I see
this becoming the standard as proc-v6.S follows it as well.
--
Catalin
On Fri, 2010-08-27 at 14:54 +0100, Catalin Marinas wrote:
> On Wed, 2010-08-25 at 05:57 +0100, Jeff Ohlstein wrote:
> > From: Daniel Walker <[email protected]>
> >
> > ScorpionMP does not have the SMP/nAMP and TLB ops broadcasting bits in
> > ACTLR.
> >
> > Signed-off-by: Daniel Walker <[email protected]>
> > Signed-off-by: Jeff Ohlstein <[email protected]>
> > ---
> > arch/arm/mm/proc-v7.S | 24 ++++++++++++++++++++++++
> > 1 files changed, 24 insertions(+), 0 deletions(-)
> >
> > diff --git a/arch/arm/mm/proc-v7.S b/arch/arm/mm/proc-v7.S
> > index 7aaf88a..98fd7e5 100644
> > --- a/arch/arm/mm/proc-v7.S
> > +++ b/arch/arm/mm/proc-v7.S
> > @@ -196,6 +196,7 @@ __v7_setup:
> > orreq r0, r0, #(1 << 6) | (1 << 0) @ Enable SMP/nAMP mode and
> > mcreq p15, 0, r0, c1, c0, 1 @ TLB ops broadcasting
> > #endif
> > +__v7_msm_setup:
> > adr r12, __v7_setup_stack @ the local stack
> > stmia r12, {r0-r5, r7, r9, r11, lr}
> > bl v7_flush_dcache_all
>
> It may be better to move the generic __v7_setup here and have a specific
> entry for ARM Ltd cores (mask 0xff0f0000 and value 0x410f0000).
I'm not sure I follow you.. Are you saying there are other cpu's that
don't need the SMP stuff at the top of __v7_setup ?
> In general, I would have liked a different way (still ID-based) of
> handling such issues without duplicating the whole structure. But I see
> this becoming the standard as proc-v6.S follows it as well.
This whole thing was something Russell specifically asked for (and
practically wrote for me).
Daniel
--
Sent by an consultant of the Qualcomm Innovation Center, Inc.
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora
Forum.
On Fri, 2010-08-27 at 16:29 +0100, Daniel Walker wrote:
> On Fri, 2010-08-27 at 14:54 +0100, Catalin Marinas wrote:
> > On Wed, 2010-08-25 at 05:57 +0100, Jeff Ohlstein wrote:
> > > From: Daniel Walker <[email protected]>
> > >
> > > ScorpionMP does not have the SMP/nAMP and TLB ops broadcasting bits in
> > > ACTLR.
> > >
> > > Signed-off-by: Daniel Walker <[email protected]>
> > > Signed-off-by: Jeff Ohlstein <[email protected]>
> > > ---
> > > arch/arm/mm/proc-v7.S | 24 ++++++++++++++++++++++++
> > > 1 files changed, 24 insertions(+), 0 deletions(-)
> > >
> > > diff --git a/arch/arm/mm/proc-v7.S b/arch/arm/mm/proc-v7.S
> > > index 7aaf88a..98fd7e5 100644
> > > --- a/arch/arm/mm/proc-v7.S
> > > +++ b/arch/arm/mm/proc-v7.S
> > > @@ -196,6 +196,7 @@ __v7_setup:
> > > orreq r0, r0, #(1 << 6) | (1 << 0) @ Enable SMP/nAMP mode and
> > > mcreq p15, 0, r0, c1, c0, 1 @ TLB ops broadcasting
> > > #endif
> > > +__v7_msm_setup:
> > > adr r12, __v7_setup_stack @ the local stack
> > > stmia r12, {r0-r5, r7, r9, r11, lr}
> > > bl v7_flush_dcache_all
> >
> > It may be better to move the generic __v7_setup here and have a specific
> > entry for ARM Ltd cores (mask 0xff0f0000 and value 0x410f0000).
>
> I'm not sure I follow you.. Are you saying there are other cpu's that
> don't need the SMP stuff at the top of __v7_setup ?
The SMP/nAMP bit is something specific to the ARM11MPCore and Cortex-A9
processors produced by ARM. If you core doesn't need this, it should use
a default __v7_setup without any additional bits being set.
I'm not aware of other MP ARM processors produced by other companies
that would require exactly the same bit being set.
> > In general, I would have liked a different way (still ID-based) of
> > handling such issues without duplicating the whole structure. But I see
> > this becoming the standard as proc-v6.S follows it as well.
>
> This whole thing was something Russell specifically asked for (and
> practically wrote for me).
Yes, I know and I'm not saying that you should change it in this patch.
It's just that I would have liked to only duplicate the setup function
rather than the whole procinfo structure.
--
Catalin
On Fri, 2010-08-27 at 17:04 +0100, Catalin Marinas wrote:
> On Fri, 2010-08-27 at 16:29 +0100, Daniel Walker wrote:
> > On Fri, 2010-08-27 at 14:54 +0100, Catalin Marinas wrote:
> > > On Wed, 2010-08-25 at 05:57 +0100, Jeff Ohlstein wrote:
> > > > From: Daniel Walker <[email protected]>
> > > >
> > > > ScorpionMP does not have the SMP/nAMP and TLB ops broadcasting bits in
> > > > ACTLR.
> > > >
> > > > Signed-off-by: Daniel Walker <[email protected]>
> > > > Signed-off-by: Jeff Ohlstein <[email protected]>
> > > > ---
> > > > arch/arm/mm/proc-v7.S | 24 ++++++++++++++++++++++++
> > > > 1 files changed, 24 insertions(+), 0 deletions(-)
> > > >
> > > > diff --git a/arch/arm/mm/proc-v7.S b/arch/arm/mm/proc-v7.S
> > > > index 7aaf88a..98fd7e5 100644
> > > > --- a/arch/arm/mm/proc-v7.S
> > > > +++ b/arch/arm/mm/proc-v7.S
> > > > @@ -196,6 +196,7 @@ __v7_setup:
> > > > orreq r0, r0, #(1 << 6) | (1 << 0) @ Enable SMP/nAMP mode and
> > > > mcreq p15, 0, r0, c1, c0, 1 @ TLB ops broadcasting
> > > > #endif
> > > > +__v7_msm_setup:
> > > > adr r12, __v7_setup_stack @ the local stack
> > > > stmia r12, {r0-r5, r7, r9, r11, lr}
> > > > bl v7_flush_dcache_all
> > >
> > > It may be better to move the generic __v7_setup here and have a specific
> > > entry for ARM Ltd cores (mask 0xff0f0000 and value 0x410f0000).
> >
> > I'm not sure I follow you.. Are you saying there are other cpu's that
> > don't need the SMP stuff at the top of __v7_setup ?
>
> The SMP/nAMP bit is something specific to the ARM11MPCore and Cortex-A9
> processors produced by ARM. If you core doesn't need this, it should use
> a default __v7_setup without any additional bits being set.
>
> I'm not aware of other MP ARM processors produced by other companies
> that would require exactly the same bit being set.
So your saying it makes more sense to change the msm entry into the
default entry, and make the current default into the
ARM11MPCore/Cortex-A9 entry?
There's 4 or 5 other cpu's that have SMP but none have had to jump over
those bits AFAIK ..
Daniel
--
Sent by an consultant of the Qualcomm Innovation Center, Inc.
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora
Forum.
On Fri, 2010-08-27 at 17:33 +0100, Daniel Walker wrote:
> On Fri, 2010-08-27 at 17:04 +0100, Catalin Marinas wrote:
> > On Fri, 2010-08-27 at 16:29 +0100, Daniel Walker wrote:
> > > On Fri, 2010-08-27 at 14:54 +0100, Catalin Marinas wrote:
> > > > On Wed, 2010-08-25 at 05:57 +0100, Jeff Ohlstein wrote:
> > > > > From: Daniel Walker <[email protected]>
> > > > >
> > > > > ScorpionMP does not have the SMP/nAMP and TLB ops broadcasting bits in
> > > > > ACTLR.
> > > > >
> > > > > Signed-off-by: Daniel Walker <[email protected]>
> > > > > Signed-off-by: Jeff Ohlstein <[email protected]>
> > > > > ---
> > > > > arch/arm/mm/proc-v7.S | 24 ++++++++++++++++++++++++
> > > > > 1 files changed, 24 insertions(+), 0 deletions(-)
> > > > >
> > > > > diff --git a/arch/arm/mm/proc-v7.S b/arch/arm/mm/proc-v7.S
> > > > > index 7aaf88a..98fd7e5 100644
> > > > > --- a/arch/arm/mm/proc-v7.S
> > > > > +++ b/arch/arm/mm/proc-v7.S
> > > > > @@ -196,6 +196,7 @@ __v7_setup:
> > > > > orreq r0, r0, #(1 << 6) | (1 << 0) @ Enable SMP/nAMP mode and
> > > > > mcreq p15, 0, r0, c1, c0, 1 @ TLB ops broadcasting
> > > > > #endif
> > > > > +__v7_msm_setup:
> > > > > adr r12, __v7_setup_stack @ the local stack
> > > > > stmia r12, {r0-r5, r7, r9, r11, lr}
> > > > > bl v7_flush_dcache_all
> > > >
> > > > It may be better to move the generic __v7_setup here and have a specific
> > > > entry for ARM Ltd cores (mask 0xff0f0000 and value 0x410f0000).
> > >
> > > I'm not sure I follow you.. Are you saying there are other cpu's that
> > > don't need the SMP stuff at the top of __v7_setup ?
> >
> > The SMP/nAMP bit is something specific to the ARM11MPCore and Cortex-A9
> > processors produced by ARM. If you core doesn't need this, it should use
> > a default __v7_setup without any additional bits being set.
> >
> > I'm not aware of other MP ARM processors produced by other companies
> > that would require exactly the same bit being set.
>
> So your saying it makes more sense to change the msm entry into the
> default entry, and make the current default into the
> ARM11MPCore/Cortex-A9 entry?
So my opinion is to not add any specific msm code but make the
__v7_setup skip the ACTLR bit setting. Then add an entry for Cortex-A9
to set those bits.
> There's 4 or 5 other cpu's that have SMP but none have had to jump over
> those bits AFAIK ..
Maybe they don't do anything with those bits and writing to them is
harmless. They'll complain if their code stops working but it is usually
unlikely that people implementing their own v7MP would use the same bits
in ACTLR as Cortex-A9.
I'm not sure whether someone licensing and manufacturing a Cortex-A9
would use the same manufacturer id (TI folks would know better). If not
we could check against 0x000fc090.
--
Catalin
On Fri, 2010-08-27 at 17:49 +0100, Catalin Marinas wrote:
> > So your saying it makes more sense to change the msm entry into the
> > default entry, and make the current default into the
> > ARM11MPCore/Cortex-A9 entry?
>
> So my opinion is to not add any specific msm code but make the
> __v7_setup skip the ACTLR bit setting. Then add an entry for Cortex-A9
> to set those bits.
>
how about this? Naming is of course flexible ..
ScorpionMP does not have the SMP/nAMP and TLB ops broadcasting bits in
ACTLR. These bits are only used on ARM11MPCore and Cortex-A9 from arm.
This patch just makes it so no other arm core ends up getting these
bits set.
Signed-off-by: Daniel Walker <[email protected]>
---
arch/arm/mm/proc-v7.S | 26 +++++++++++++++++++++++++-
1 files changed, 25 insertions(+), 1 deletions(-)
diff --git a/arch/arm/mm/proc-v7.S b/arch/arm/mm/proc-v7.S
index 7aaf88a..e1b5492 100644
--- a/arch/arm/mm/proc-v7.S
+++ b/arch/arm/mm/proc-v7.S
@@ -189,13 +189,14 @@ cpu_v7_name:
* It is assumed that:
* - cache type register is implemented
*/
-__v7_setup:
+__v7_armmp_setup:
#ifdef CONFIG_SMP
mrc p15, 0, r0, c1, c0, 1
tst r0, #(1 << 6) @ SMP/nAMP mode enabled?
orreq r0, r0, #(1 << 6) | (1 << 0) @ Enable SMP/nAMP mode and
mcreq p15, 0, r0, c1, c0, 1 @ TLB ops broadcasting
#endif
+__v7_setup:
adr r12, __v7_setup_stack @ the local stack
stmia r12, {r0-r5, r7, r9, r11, lr}
bl v7_flush_dcache_all
@@ -326,6 +327,29 @@ cpu_elf_name:
.section ".proc.info.init", #alloc, #execinstr
+ .type __v7_armmp_proc_info, #object
+__v7_armmp_proc_info:
+ .long 0x410f0000 @ Required ID value
+ .long 0xff0f0000 @ Mask for ID
+ .long PMD_TYPE_SECT | \
+ PMD_SECT_AP_WRITE | \
+ PMD_SECT_AP_READ | \
+ PMD_FLAGS
+ .long PMD_TYPE_SECT | \
+ PMD_SECT_XN | \
+ PMD_SECT_AP_WRITE | \
+ PMD_SECT_AP_READ
+ b __v7_armmp_setup
+ .long cpu_arch_name
+ .long cpu_elf_name
+ .long HWCAP_SWP|HWCAP_HALF|HWCAP_THUMB|HWCAP_FAST_MULT|HWCAP_EDSP
+ .long cpu_v7_name
+ .long v7_processor_functions
+ .long v7wbi_tlb_fns
+ .long v6_user_fns
+ .long v7_cache_fns
+ .size __v7_armmp_proc_info, . - __v7_armmp_proc_info
+
/*
* Match any ARMv7 processor core.
*/
--
1.7.0.4
--
Sent by an consultant of the Qualcomm Innovation Center, Inc.
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora
Forum.
On Fri, 2010-08-27 at 20:53 +0100, Daniel Walker wrote:
> On Fri, 2010-08-27 at 17:49 +0100, Catalin Marinas wrote:
>
> > > So your saying it makes more sense to change the msm entry into the
> > > default entry, and make the current default into the
> > > ARM11MPCore/Cortex-A9 entry?
> >
> > So my opinion is to not add any specific msm code but make the
> > __v7_setup skip the ACTLR bit setting. Then add an entry for Cortex-A9
> > to set those bits.
> >
>
> how about this? Naming is of course flexible ..
I cc'ed Santosh as well. I'm not sure whether TI are using a Cortex-A9
but with different manufacturer field (I suspect it's still 0x41).
> ScorpionMP does not have the SMP/nAMP and TLB ops broadcasting bits in
> ACTLR. These bits are only used on ARM11MPCore and Cortex-A9 from arm.
> This patch just makes it so no other arm core ends up getting these
> bits set.
>
> Signed-off-by: Daniel Walker <[email protected]>
>
> ---
> arch/arm/mm/proc-v7.S | 26 +++++++++++++++++++++++++-
> 1 files changed, 25 insertions(+), 1 deletions(-)
>
> diff --git a/arch/arm/mm/proc-v7.S b/arch/arm/mm/proc-v7.S
> index 7aaf88a..e1b5492 100644
> --- a/arch/arm/mm/proc-v7.S
> +++ b/arch/arm/mm/proc-v7.S
> @@ -189,13 +189,14 @@ cpu_v7_name:
> * It is assumed that:
> * - cache type register is implemented
> */
> -__v7_setup:
> +__v7_armmp_setup:
Maybe __v7_ca9mp_setup. Future MP processors from ARM may not need this
bit set.
> +__v7_armmp_proc_info:
> + .long 0x410f0000 @ Required ID value
> + .long 0xff0f0000 @ Mask for ID
We could restrict it to:
.long 0x410fc090
.long 0xff0ffff0
Otherwise the patch looks fine. Thanks.
--
Catalin
On Tue, 2010-08-31 at 13:18 +0100, Catalin Marinas wrote:
> On Fri, 2010-08-27 at 20:53 +0100, Daniel Walker wrote:
> > On Fri, 2010-08-27 at 17:49 +0100, Catalin Marinas wrote:
> >
> > > > So your saying it makes more sense to change the msm entry into the
> > > > default entry, and make the current default into the
> > > > ARM11MPCore/Cortex-A9 entry?
> > >
> > > So my opinion is to not add any specific msm code but make the
> > > __v7_setup skip the ACTLR bit setting. Then add an entry for Cortex-A9
> > > to set those bits.
> > >
> >
> > how about this? Naming is of course flexible ..
>
> I cc'ed Santosh as well. I'm not sure whether TI are using a Cortex-A9
> but with different manufacturer field (I suspect it's still 0x41).
>
> > ScorpionMP does not have the SMP/nAMP and TLB ops broadcasting bits in
> > ACTLR. These bits are only used on ARM11MPCore and Cortex-A9 from arm.
> > This patch just makes it so no other arm core ends up getting these
> > bits set.
> >
> > Signed-off-by: Daniel Walker <[email protected]>
> >
> > ---
> > arch/arm/mm/proc-v7.S | 26 +++++++++++++++++++++++++-
> > 1 files changed, 25 insertions(+), 1 deletions(-)
> >
> > diff --git a/arch/arm/mm/proc-v7.S b/arch/arm/mm/proc-v7.S
> > index 7aaf88a..e1b5492 100644
> > --- a/arch/arm/mm/proc-v7.S
> > +++ b/arch/arm/mm/proc-v7.S
> > @@ -189,13 +189,14 @@ cpu_v7_name:
> > * It is assumed that:
> > * - cache type register is implemented
> > */
> > -__v7_setup:
> > +__v7_armmp_setup:
>
> Maybe __v7_ca9mp_setup. Future MP processors from ARM may not need this
> bit set.
Ok ..
> > +__v7_armmp_proc_info:
> > + .long 0x410f0000 @ Required ID value
> > + .long 0xff0f0000 @ Mask for ID
>
> We could restrict it to:
>
> .long 0x410fc090
> .long 0xff0ffff0
>
> Otherwise the patch looks fine. Thanks.
Ok, I'll add this.
Daniel
--
Sent by a consultant of the Qualcomm Innovation Center, Inc.
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum.