[PATCH 00/38] Reorganize setup code and merge 32 and 64-bit setup_arch()

[PATCH 00/38] Reorganize setup code and merge 32 and 64-bit setup_arch()

[Benjamin Herrenschmidt]

This series reorganizes the setup code, moving probe_machine() later than
when it's currently done, and sync'ing 32-bit and 64-bit enough to merge
their implementations of setup_arch(). We get rid of ppc64 setup_system()
which is subsumed by the new setup_arch().

Note: The first 2 patches should go separately. The first one is a pre
requisite for the ppc32 move of probe_machine(). The second one a general
bug fix for recent kernels which I included here so my test kernels
could build.

I test built all ppc64 platforms and the STD_MMU ppc32 ones. I test
booted pseries and powermac in qemu.

[PATCH 01/38] dt: Add of_device_compatible_match()

[Benjamin Herrenschmidt]

This provides an equivalent of of_fdt_match() for non-flat trees.

This is more practical than matching an array of of_device_id structs
when converting a bunch of existing users of of_fdt_match().

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
---
 drivers/of/base.c  | 22 ++++++++++++++++++++++
 include/linux/of.h |  2 ++
 2 files changed, 24 insertions(+)

diff --git a/drivers/of/base.c b/drivers/of/base.c
index ebf84e3..429c594 100644
--- a/drivers/of/base.c
+++ b/drivers/of/base.c
@@ -493,6 +493,28 @@ int of_device_is_compatible(const struct device_node *device,
 }
 EXPORT_SYMBOL(of_device_is_compatible);
 
+/** Checks if the device is compatible with any of the entries in
+ *  a NULL terminated array of strings. Returns the best match
+ *  score or 0.
+ */
+int of_device_compatible_match(struct device_node *device,
+			       const char *const *compat)
+{
+	unsigned int tmp, score = 0;
+
+	if (!compat)
+		return 0;
+
+	while (*compat) {
+		tmp = of_device_is_compatible(device, *compat);
+		if (tmp && (score == 0 || (tmp < score)))
+			score = tmp;
+		compat++;
+	}
+
+	return score;
+}
+
 /**
  * of_machine_is_compatible - Test root of device tree for a given compatible value
  * @compat: compatible string to look for in root node's compatible property.
diff --git a/include/linux/of.h b/include/linux/of.h
index 74eb28c..33c184d 100644
--- a/include/linux/of.h
+++ b/include/linux/of.h
@@ -324,6 +324,8 @@ extern int of_property_read_string_helper(const struct device_node *np,
 					      const char **out_strs, size_t sz, int index);
 extern int of_device_is_compatible(const struct device_node *device,
 				   const char *);
+extern int of_device_compatible_match(struct device_node *device,
+				      const char *const *compat);
 extern bool of_device_is_available(const struct device_node *device);
 extern bool of_device_is_big_endian(const struct device_node *device);
 extern const void *of_get_property(const struct device_node *node,
-- 
2.7.4

[PATCH 02/38] drm: Fix broken use of _PAGE_NO_CACHE on powerpc

[Benjamin Herrenschmidt]

That constant no longer exist. Use the proper accessor instead

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
---
 drivers/gpu/drm/drm_memory.c  | 2 +-
 drivers/gpu/drm/drm_scatter.c | 2 +-
 drivers/gpu/drm/drm_vm.c      | 4 ++--
 3 files changed, 4 insertions(+), 4 deletions(-)

diff --git a/drivers/gpu/drm/drm_memory.c b/drivers/gpu/drm/drm_memory.c
index 87a8cb7..fc0ebd2 100644
--- a/drivers/gpu/drm/drm_memory.c
+++ b/drivers/gpu/drm/drm_memory.c
@@ -44,7 +44,7 @@
 # include <asm/agp.h>
 #else
 # ifdef __powerpc__
-#  define PAGE_AGP	__pgprot(_PAGE_KERNEL | _PAGE_NO_CACHE)
+#  define PAGE_AGP	pgprot_noncached_wc(PAGE_KERNEL)
 # else
 #  define PAGE_AGP	PAGE_KERNEL
 # endif
diff --git a/drivers/gpu/drm/drm_scatter.c b/drivers/gpu/drm/drm_scatter.c
index 4f0f3b3..bf70431 100644
--- a/drivers/gpu/drm/drm_scatter.c
+++ b/drivers/gpu/drm/drm_scatter.c
@@ -41,7 +41,7 @@
 static inline void *drm_vmalloc_dma(unsigned long size)
 {
 #if defined(__powerpc__) && defined(CONFIG_NOT_COHERENT_CACHE)
-	return __vmalloc(size, GFP_KERNEL, PAGE_KERNEL | _PAGE_NO_CACHE);
+	return __vmalloc(size, GFP_KERNEL, pgprot_noncached_wc(PAGE_KERNEL));
 #else
 	return vmalloc_32(size);
 #endif
diff --git a/drivers/gpu/drm/drm_vm.c b/drivers/gpu/drm/drm_vm.c
index ac9f4b3..7e9f642 100644
--- a/drivers/gpu/drm/drm_vm.c
+++ b/drivers/gpu/drm/drm_vm.c
@@ -80,7 +80,7 @@ static pgprot_t drm_dma_prot(uint32_t map_type, struct vm_area_struct *vma)
 	pgprot_t tmp = vm_get_page_prot(vma->vm_flags);
 
 #if defined(__powerpc__) && defined(CONFIG_NOT_COHERENT_CACHE)
-	tmp |= _PAGE_NO_CACHE;
+	tmp = pgprot_noncached_wc(tmp);
 #endif
 	return tmp;
 }
@@ -593,7 +593,7 @@ static int drm_mmap_locked(struct file *filp, struct vm_area_struct *vma)
 			 * pages and mappings in fault()
 			 */
 #if defined(__powerpc__)
-			pgprot_val(vma->vm_page_prot) |= _PAGE_NO_CACHE;
+			vma->vm_page_prot = pgprot_noncached(vma->vm_page_prot);
 #endif
 			vma->vm_ops = &drm_vm_ops;
 			break;
-- 
2.7.4

[PATCH 03/38] powerpc/prom_init: PTRRELOC is not needed

[Benjamin Herrenschmidt]

There's one line of code still using PTRRELOC in prom_init, it
shouldn't be necessary since that code is relocatable. Take it
out.

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
---
 arch/powerpc/kernel/prom_init.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/arch/powerpc/kernel/prom_init.c b/arch/powerpc/kernel/prom_init.c
index 6ee4b72..54e450c 100644
--- a/arch/powerpc/kernel/prom_init.c
+++ b/arch/powerpc/kernel/prom_init.c
@@ -1980,7 +1980,7 @@ static void __init prom_check_displays(void)
 				break;
 
 #ifdef CONFIG_LOGO_LINUX_CLUT224
-		clut = PTRRELOC(logo_linux_clut224.clut);
+		clut = logo_linux_clut224.clut;
 		for (i = 0; i < logo_linux_clut224.clutsize; i++, clut += 3)
 			if (prom_set_color(ih, i + 32, clut[0], clut[1],
 					   clut[2]) != 0)
-- 
2.7.4

[PATCH 04/38] powerpc: Make PTRRELOC() 32-bit only

[Benjamin Herrenschmidt]

PTRRELOC is only used by 32-bit code since on 32-bit systems, early_init
can be run with relocation off and running at an offset. Define it to
identity on 64-bit

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
---
 arch/powerpc/include/asm/setup.h | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/arch/powerpc/include/asm/setup.h b/arch/powerpc/include/asm/setup.h
index e9d384c..fa0687e 100644
--- a/arch/powerpc/include/asm/setup.h
+++ b/arch/powerpc/include/asm/setup.h
@@ -19,7 +19,11 @@ extern unsigned long reloc_offset(void);
 extern unsigned long add_reloc_offset(unsigned long);
 extern void reloc_got2(unsigned long);
 
+#ifdef CONFIG_PPC32
 #define PTRRELOC(x)	((typeof(x)) add_reloc_offset((unsigned long)(x)))
+#else
+#define PTRRELOC(x)	(x)
+#endif
 
 void check_for_initrd(void);
 void initmem_init(void);
-- 
2.7.4

[PATCH 05/38] powerpc: Factor do_feature_fixup calls

[Benjamin Herrenschmidt]

32 and 64-bit do a similar set of calls early on, we move it all to
a single common function to make the boot code more readable.

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
---
 arch/powerpc/include/asm/feature-fixups.h |  4 ++++
 arch/powerpc/include/asm/synch.h          |  1 -
 arch/powerpc/kernel/setup_32.c            | 17 ++---------------
 arch/powerpc/kernel/setup_64.c            | 13 +------------
 arch/powerpc/lib/feature-fixups.c         | 31 +++++++++++++++++++++++++++++--
 5 files changed, 36 insertions(+), 30 deletions(-)

diff --git a/arch/powerpc/include/asm/feature-fixups.h b/arch/powerpc/include/asm/feature-fixups.h
index 9a67a38..57fec8a 100644
--- a/arch/powerpc/include/asm/feature-fixups.h
+++ b/arch/powerpc/include/asm/feature-fixups.h
@@ -184,4 +184,8 @@ label##3:					       	\
 	FTR_ENTRY_OFFSET label##1b-label##3b;		\
 	.popsection;
 
+#ifndef __ASSEMBLY__
+void apply_feature_fixups(void);
+#endif
+
 #endif /* __ASM_POWERPC_FEATURE_FIXUPS_H */
diff --git a/arch/powerpc/include/asm/synch.h b/arch/powerpc/include/asm/synch.h
index c508686..78efe8d 100644
--- a/arch/powerpc/include/asm/synch.h
+++ b/arch/powerpc/include/asm/synch.h
@@ -13,7 +13,6 @@
 extern unsigned int __start___lwsync_fixup, __stop___lwsync_fixup;
 extern void do_lwsync_fixups(unsigned long value, void *fixup_start,
 			     void *fixup_end);
-extern void do_final_fixups(void);
 
 static inline void eieio(void)
 {
diff --git a/arch/powerpc/kernel/setup_32.c b/arch/powerpc/kernel/setup_32.c
index d544fa3..3d06fab 100644
--- a/arch/powerpc/kernel/setup_32.c
+++ b/arch/powerpc/kernel/setup_32.c
@@ -73,7 +73,6 @@ int ucache_bsize;
 notrace unsigned long __init early_init(unsigned long dt_ptr)
 {
 	unsigned long offset = reloc_offset();
-	struct cpu_spec *spec;
 
 	/* First zero the BSS -- use memset_io, some platforms don't have
 	 * caches on yet */
@@ -84,21 +83,9 @@ notrace unsigned long __init early_init(unsigned long dt_ptr)
 	 * Identify the CPU type and fix up code sections
 	 * that depend on which cpu we have.
 	 */
-	spec = identify_cpu(offset, mfspr(SPRN_PVR));
+	identify_cpu(offset, mfspr(SPRN_PVR));
 
-	do_feature_fixups(spec->cpu_features,
-			  PTRRELOC(&__start___ftr_fixup),
-			  PTRRELOC(&__stop___ftr_fixup));
-
-	do_feature_fixups(spec->mmu_features,
-			  PTRRELOC(&__start___mmu_ftr_fixup),
-			  PTRRELOC(&__stop___mmu_ftr_fixup));
-
-	do_lwsync_fixups(spec->cpu_features,
-			 PTRRELOC(&__start___lwsync_fixup),
-			 PTRRELOC(&__stop___lwsync_fixup));
-
-	do_final_fixups();
+	apply_feature_fixups();
 
 	return KERNELBASE + offset;
 }
diff --git a/arch/powerpc/kernel/setup_64.c b/arch/powerpc/kernel/setup_64.c
index 96d4a2b..996bed2 100644
--- a/arch/powerpc/kernel/setup_64.c
+++ b/arch/powerpc/kernel/setup_64.c
@@ -468,18 +468,7 @@ void __init setup_system(void)
 {
 	DBG(" -> setup_system()\n");
 
-	/* Apply the CPUs-specific and firmware specific fixups to kernel
-	 * text (nop out sections not relevant to this CPU or this firmware)
-	 */
-	do_feature_fixups(cur_cpu_spec->cpu_features,
-			  &__start___ftr_fixup, &__stop___ftr_fixup);
-	do_feature_fixups(cur_cpu_spec->mmu_features,
-			  &__start___mmu_ftr_fixup, &__stop___mmu_ftr_fixup);
-	do_feature_fixups(powerpc_firmware_features,
-			  &__start___fw_ftr_fixup, &__stop___fw_ftr_fixup);
-	do_lwsync_fixups(cur_cpu_spec->cpu_features,
-			 &__start___lwsync_fixup, &__stop___lwsync_fixup);
-	do_final_fixups();
+	apply_feature_fixups();
 
 	/*
 	 * Unflatten the device-tree passed by prom_init or kexec
diff --git a/arch/powerpc/lib/feature-fixups.c b/arch/powerpc/lib/feature-fixups.c
index 7ce3870..d0c72aa 100644
--- a/arch/powerpc/lib/feature-fixups.c
+++ b/arch/powerpc/lib/feature-fixups.c
@@ -20,7 +20,8 @@
 #include <asm/code-patching.h>
 #include <asm/page.h>
 #include <asm/sections.h>
-
+#include <asm/setup.h>
+#include <asm/firmware.h>
 
 struct fixup_entry {
 	unsigned long	mask;
@@ -130,7 +131,7 @@ void do_lwsync_fixups(unsigned long value, void *fixup_start, void *fixup_end)
 	}
 }
 
-void do_final_fixups(void)
+static void do_final_fixups(void)
 {
 #if defined(CONFIG_PPC64) && defined(CONFIG_RELOCATABLE)
 	int *src, *dest;
@@ -151,6 +152,32 @@ void do_final_fixups(void)
 #endif
 }
 
+void apply_feature_fixups(void)
+{
+	struct cpu_spec *spec = *PTRRELOC(&cur_cpu_spec);
+
+	/* Apply the CPUs-specific and firmware specific fixups to kernel
+	 * text (nop out sections not relevant to this CPU or this firmware)
+	 */
+	do_feature_fixups(spec->cpu_features,
+			  PTRRELOC(&__start___ftr_fixup),
+			  PTRRELOC(&__stop___ftr_fixup));
+
+	do_feature_fixups(spec->mmu_features,
+			  PTRRELOC(&__start___mmu_ftr_fixup),
+			  PTRRELOC(&__stop___mmu_ftr_fixup));
+
+	do_lwsync_fixups(spec->cpu_features,
+			 PTRRELOC(&__start___lwsync_fixup),
+			 PTRRELOC(&__stop___lwsync_fixup));
+
+#ifdef CONFIG_PPC64
+	do_feature_fixups(powerpc_firmware_features,
+			  &__start___fw_ftr_fixup, &__stop___fw_ftr_fixup);
+#endif
+	do_final_fixups();
+}
+
 #ifdef CONFIG_FTR_FIXUP_SELFTEST
 
 #define check(x)	\
-- 
2.7.4

[PATCH 06/38] powerpc: Move 64-bit feature fixup earlier

[Benjamin Herrenschmidt]

Make it part of early_setup() as we really want the feature fixups
to be applied before we turn on the MMU since they can have an impact
on the various assembly path related to MMU management and interrupts.

This makes 64-bit match what 32-bit does.

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
---
 arch/powerpc/kernel/setup_64.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/arch/powerpc/kernel/setup_64.c b/arch/powerpc/kernel/setup_64.c
index 996bed2..e79cf1b 100644
--- a/arch/powerpc/kernel/setup_64.c
+++ b/arch/powerpc/kernel/setup_64.c
@@ -286,6 +286,9 @@ void __init early_setup(unsigned long dt_ptr)
 	/* Initialize the hash table or TLB handling */
 	early_init_mmu();
 
+	/* Apply all the dynamic patching */
+	apply_feature_fixups();
+
 	/*
 	 * At this point, we can let interrupts switch to virtual mode
 	 * (the MMU has been setup), so adjust the MSR in the PACA to
@@ -468,8 +471,6 @@ void __init setup_system(void)
 {
 	DBG(" -> setup_system()\n");
 
-	apply_feature_fixups();
-
 	/*
 	 * Unflatten the device-tree passed by prom_init or kexec
 	 */
-- 
2.7.4

[PATCH 07/38] powerpc: Move 64-bit memory reserves to setup_arch()

[Benjamin Herrenschmidt]

There is really no need to do them that early, early_setup() runs
before MMU is on, we should do the strict minimum there to get the
MMU going.

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
---
 arch/powerpc/kernel/setup_64.c | 22 +++++++++++-----------
 1 file changed, 11 insertions(+), 11 deletions(-)

diff --git a/arch/powerpc/kernel/setup_64.c b/arch/powerpc/kernel/setup_64.c
index e79cf1b..bcc95ac 100644
--- a/arch/powerpc/kernel/setup_64.c
+++ b/arch/powerpc/kernel/setup_64.c
@@ -296,16 +296,6 @@ void __init early_setup(unsigned long dt_ptr)
 	 */
 	cpu_ready_for_interrupts();
 
-	/* Reserve large chunks of memory for use by CMA for KVM */
-	kvm_cma_reserve();
-
-	/*
-	 * Reserve any gigantic pages requested on the command line.
-	 * memblock needs to have been initialized by the time this is
-	 * called since this will reserve memory.
-	 */
-	reserve_hugetlb_gpages();
-
 	DBG(" <- early_setup()\n");
 
 #ifdef CONFIG_PPC_EARLY_DEBUG_BOOTX
@@ -688,6 +678,17 @@ void __init setup_arch(char **cmdline_p)
 	dcache_bsize = ppc64_caches.dline_size;
 	icache_bsize = ppc64_caches.iline_size;
 
+
+	/* Reserve large chunks of memory for use by CMA for KVM */
+	kvm_cma_reserve();
+
+	/*
+	 * Reserve any gigantic pages requested on the command line.
+	 * memblock needs to have been initialized by the time this is
+	 * called since this will reserve memory.
+	 */
+	reserve_hugetlb_gpages();
+
 	if (ppc_md.panic)
 		setup_panic();
 
@@ -712,7 +713,6 @@ void __init setup_arch(char **cmdline_p)
 #ifdef CONFIG_DUMMY_CONSOLE
 	conswitchp = &dummy_con;
 #endif
-
 	if (ppc_md.setup_arch)
 		ppc_md.setup_arch();
 
-- 
2.7.4

[PATCH 08/38] powerpc: Move epapr_paravirt_early_init() to early_init_devtree()

[Benjamin Herrenschmidt]

The function is called by both 32-bit and 64-bit early setup right
after early_init_devtree(). All it does is run yet another early
DT parser which is precisely what early_init_devtree() is about,
so move it in there.

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
---
 arch/powerpc/kernel/prom.c     | 2 ++
 arch/powerpc/kernel/setup_32.c | 3 ---
 arch/powerpc/kernel/setup_64.c | 3 ---
 3 files changed, 2 insertions(+), 6 deletions(-)

diff --git a/arch/powerpc/kernel/prom.c b/arch/powerpc/kernel/prom.c
index 946e34f..48434be 100644
--- a/arch/powerpc/kernel/prom.c
+++ b/arch/powerpc/kernel/prom.c
@@ -56,6 +56,7 @@
 #include <asm/opal.h>
 #include <asm/fadump.h>
 #include <asm/debug.h>
+#include <asm/epapr_hcalls.h>
 
 #include <mm/mmu_decl.h>
 
@@ -739,6 +740,7 @@ void __init early_init_devtree(void *params)
 	/* Scan and build the list of machine check recoverable ranges */
 	of_scan_flat_dt(early_init_dt_scan_recoverable_ranges, NULL);
 #endif
+	epapr_paravirt_early_init();
 
 	DBG(" <- early_init_devtree()\n");
 }
diff --git a/arch/powerpc/kernel/setup_32.c b/arch/powerpc/kernel/setup_32.c
index 3d06fab..34e61d68 100644
--- a/arch/powerpc/kernel/setup_32.c
+++ b/arch/powerpc/kernel/setup_32.c
@@ -37,7 +37,6 @@
 #include <asm/serial.h>
 #include <asm/udbg.h>
 #include <asm/mmu_context.h>
-#include <asm/epapr_hcalls.h>
 #include <asm/code-patching.h>
 
 #define DBG(fmt...)
@@ -110,8 +109,6 @@ notrace void __init machine_init(u64 dt_ptr)
 	/* Do some early initialization based on the flat device tree */
 	early_init_devtree(__va(dt_ptr));
 
-	epapr_paravirt_early_init();
-
 	early_init_mmu();
 
 	probe_machine();
diff --git a/arch/powerpc/kernel/setup_64.c b/arch/powerpc/kernel/setup_64.c
index bcc95ac..e1b2e6f 100644
--- a/arch/powerpc/kernel/setup_64.c
+++ b/arch/powerpc/kernel/setup_64.c
@@ -68,7 +68,6 @@
 #include <asm/code-patching.h>
 #include <asm/kvm_ppc.h>
 #include <asm/hugetlb.h>
-#include <asm/epapr_hcalls.h>
 #include <asm/livepatch.h>
 
 #ifdef DEBUG
@@ -270,8 +269,6 @@ void __init early_setup(unsigned long dt_ptr)
 	 */
 	early_init_devtree(__va(dt_ptr));
 
-	epapr_paravirt_early_init();
-
 	/* Now we know the logical id of our boot cpu, setup the paca. */
 	setup_paca(&paca[boot_cpuid]);
 	fixup_boot_paca();
-- 
2.7.4

[PATCH 09/38] powerpc: Update obsolete comments in setup_32.c about entry conditions

[Benjamin Herrenschmidt]

early_init() is called in-place before kernel relocation and using
whatever MMU setup exists at the point the kernel is entered.

machine_init() is called after relocation and after some initial
mapping of PAGE_OFFSET has been established (typically using BATs
on 6xx/7xx/7xxx processors or some form of bolted TLB on others).

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
---
 arch/powerpc/kernel/setup_32.c | 8 +++++---
 1 file changed, 5 insertions(+), 3 deletions(-)

diff --git a/arch/powerpc/kernel/setup_32.c b/arch/powerpc/kernel/setup_32.c
index 34e61d68..3f0aca2 100644
--- a/arch/powerpc/kernel/setup_32.c
+++ b/arch/powerpc/kernel/setup_32.c
@@ -61,9 +61,7 @@ int icache_bsize;
 int ucache_bsize;
 
 /*
- * We're called here very early in the boot.  We determine the machine
- * type and call the appropriate low-level setup functions.
- *  -- Cort <cort@fsmlabs.com>
+ * We're called here very early in the boot.
  *
  * Note that the kernel may be running at an address which is different
  * from the address that it was linked at, so we must use RELOC/PTRRELOC
@@ -91,6 +89,10 @@ notrace unsigned long __init early_init(unsigned long dt_ptr)
 
 
 /*
+ * This is run before start_kernel(), the kernel has been relocated
+ * and we are running with enough of the MMU enabled to have our
+ * proper kernel virtual addresses
+ *
  * Find out what kind of machine we're on and save any data we need
  * from the early boot process (devtree is copied on pmac by prom_init()).
  * This is called very early on the boot process, after a minimal
-- 
2.7.4

[PATCH 10/38] powerpc: Add comment explaining the purpose of setup_kdump_trampoline()

[Benjamin Herrenschmidt]

Anything in early_setup() needs to be justified to be there, in
this case, we need the trampolines before we can take exceptions
and thus before we turn on the MMU.

Also remove a pretty meaningless and misplaced debug message

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
---
 arch/powerpc/kernel/setup_64.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/arch/powerpc/kernel/setup_64.c b/arch/powerpc/kernel/setup_64.c
index e1b2e6f..a641753 100644
--- a/arch/powerpc/kernel/setup_64.c
+++ b/arch/powerpc/kernel/setup_64.c
@@ -276,10 +276,11 @@ void __init early_setup(unsigned long dt_ptr)
 	/* Probe the machine type */
 	probe_machine();
 
+	/* Setup the trampolines from the lowmem exception vectors
+	 * to the kdump kernel when not using a relocatable kernel.
+	 */
 	setup_kdump_trampoline();
 
-	DBG("Found, Initializing memory management...\n");
-
 	/* Initialize the hash table or TLB handling */
 	early_init_mmu();
 
-- 
2.7.4

[PATCH 11/38] powerpc/dart: Use a cachable DART

[Benjamin Herrenschmidt]

Instead of punching a hole in the linear mapping, just use normal
cachable memory, and apply the flush sequence documented in the
CPC625 (aka U3) user manual.

This allows us to remove quite a bit of code related to the early
allocation of the DART and the hole in the linear mapping. We can
also get rid of the copy of the DART for suspend/resume as the
original memory can just be saved/restored now, as long as we
properly sync the caches.

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
---
 arch/powerpc/include/asm/iommu.h        |   1 -
 arch/powerpc/mm/hash_utils_64.c         |  32 ------
 arch/powerpc/platforms/maple/setup.c    |   7 --
 arch/powerpc/platforms/powermac/setup.c |   8 --
 arch/powerpc/sysdev/dart_iommu.c        | 180 +++++++++++++++-----------------
 5 files changed, 84 insertions(+), 144 deletions(-)

diff --git a/arch/powerpc/include/asm/iommu.h b/arch/powerpc/include/asm/iommu.h
index 7b87bab..f49a72a 100644
--- a/arch/powerpc/include/asm/iommu.h
+++ b/arch/powerpc/include/asm/iommu.h
@@ -273,7 +273,6 @@ extern void iommu_init_early_pSeries(void);
 extern void iommu_init_early_dart(struct pci_controller_ops *controller_ops);
 extern void iommu_init_early_pasemi(void);
 
-extern void alloc_dart_table(void);
 #if defined(CONFIG_PPC64) && defined(CONFIG_PM)
 static inline void iommu_save(void)
 {
diff --git a/arch/powerpc/mm/hash_utils_64.c b/arch/powerpc/mm/hash_utils_64.c
index b2740c6..2bea864 100644
--- a/arch/powerpc/mm/hash_utils_64.c
+++ b/arch/powerpc/mm/hash_utils_64.c
@@ -87,10 +87,6 @@
  *
  */
 
-#ifdef CONFIG_U3_DART
-extern unsigned long dart_tablebase;
-#endif /* CONFIG_U3_DART */
-
 static unsigned long _SDR1;
 struct mmu_psize_def mmu_psize_defs[MMU_PAGE_COUNT];
 EXPORT_SYMBOL_GPL(mmu_psize_defs);
@@ -828,34 +824,6 @@ static void __init htab_initialize(void)
 		DBG("creating mapping for region: %lx..%lx (prot: %lx)\n",
 		    base, size, prot);
 
-#ifdef CONFIG_U3_DART
-		/* Do not map the DART space. Fortunately, it will be aligned
-		 * in such a way that it will not cross two memblock regions and
-		 * will fit within a single 16Mb page.
-		 * The DART space is assumed to be a full 16Mb region even if
-		 * we only use 2Mb of that space. We will use more of it later
-		 * for AGP GART. We have to use a full 16Mb large page.
-		 */
-		DBG("DART base: %lx\n", dart_tablebase);
-
-		if (dart_tablebase != 0 && dart_tablebase >= base
-		    && dart_tablebase < (base + size)) {
-			unsigned long dart_table_end = dart_tablebase + 16 * MB;
-			if (base != dart_tablebase)
-				BUG_ON(htab_bolt_mapping(base, dart_tablebase,
-							__pa(base), prot,
-							mmu_linear_psize,
-							mmu_kernel_ssize));
-			if ((base + size) > dart_table_end)
-				BUG_ON(htab_bolt_mapping(dart_tablebase+16*MB,
-							base + size,
-							__pa(dart_table_end),
-							 prot,
-							 mmu_linear_psize,
-							 mmu_kernel_ssize));
-			continue;
-		}
-#endif /* CONFIG_U3_DART */
 		BUG_ON(htab_bolt_mapping(base, base + size, __pa(base),
 				prot, mmu_linear_psize, mmu_kernel_ssize));
 	}
diff --git a/arch/powerpc/platforms/maple/setup.c b/arch/powerpc/platforms/maple/setup.c
index a837188..3cd625d 100644
--- a/arch/powerpc/platforms/maple/setup.c
+++ b/arch/powerpc/platforms/maple/setup.c
@@ -303,13 +303,6 @@ static int __init maple_probe(void)
 	if (!of_flat_dt_is_compatible(root, "Momentum,Maple") &&
 	    !of_flat_dt_is_compatible(root, "Momentum,Apache"))
 		return 0;
-	/*
-	 * On U3, the DART (iommu) must be allocated now since it
-	 * has an impact on htab_initialize (due to the large page it
-	 * occupies having to be broken up so the DART itself is not
-	 * part of the cacheable linar mapping
-	 */
-	alloc_dart_table();
 
 	hpte_init_native();
 	pm_power_off = maple_power_off;
diff --git a/arch/powerpc/platforms/powermac/setup.c b/arch/powerpc/platforms/powermac/setup.c
index 8dd78f4..19de197 100644
--- a/arch/powerpc/platforms/powermac/setup.c
+++ b/arch/powerpc/platforms/powermac/setup.c
@@ -605,14 +605,6 @@ static int __init pmac_probe(void)
 		return 0;
 
 #ifdef CONFIG_PPC64
-	/*
-	 * On U3, the DART (iommu) must be allocated now since it
-	 * has an impact on htab_initialize (due to the large page it
-	 * occupies having to be broken up so the DART itself is not
-	 * part of the cacheable linar mapping
-	 */
-	alloc_dart_table();
-
 	hpte_init_native();
 #endif
 
diff --git a/arch/powerpc/sysdev/dart_iommu.c b/arch/powerpc/sysdev/dart_iommu.c
index b734863..53f862e 100644
--- a/arch/powerpc/sysdev/dart_iommu.c
+++ b/arch/powerpc/sysdev/dart_iommu.c
@@ -48,16 +48,10 @@
 
 #include "dart.h"
 
-/* Physical base address and size of the DART table */
-unsigned long dart_tablebase; /* exported to htab_initialize */
+/* DART table address and size */
+static u32 *dart_tablebase;
 static unsigned long dart_tablesize;
 
-/* Virtual base address of the DART table */
-static u32 *dart_vbase;
-#ifdef CONFIG_PM
-static u32 *dart_copy;
-#endif
-
 /* Mapped base address for the dart */
 static unsigned int __iomem *dart;
 
@@ -151,6 +145,32 @@ wait_more:
 	spin_unlock_irqrestore(&invalidate_lock, flags);
 }
 
+static void dart_cache_sync(unsigned int *base, unsigned int count)
+{
+	/* We add 1 to the number of entries to flush, following a
+	 * comment in Darwin indicating that the memory controller
+	 * can prefetch unmapped memory under some circumstances
+	 */
+	unsigned long start = (unsigned long)base;
+	unsigned long end = start + (count + 1) * sizeof(unsigned int);
+	unsigned int tmp;
+
+	/* Perform a standard cache flush */
+	flush_inval_dcache_range(start, end);
+
+	/* Perform the sequence described in the CPC925 manual to
+	 * ensure all the data gets to a point the cache incoherent
+	 * DART hardware will see
+	 */
+	asm volatile(" sync;"
+		     " isync;"
+		     " dcbf 0,%1;"
+		     " sync;"
+		     " isync;"
+		     " lwz %0,0(%1);"
+		     " isync" : "=r" (tmp) : "r" (end) : "memory");
+}
+
 static void dart_flush(struct iommu_table *tbl)
 {
 	mb();
@@ -165,13 +185,13 @@ static int dart_build(struct iommu_table *tbl, long index,
 		       enum dma_data_direction direction,
 		       struct dma_attrs *attrs)
 {
-	unsigned int *dp;
+	unsigned int *dp, *orig_dp;
 	unsigned int rpn;
 	long l;
 
 	DBG("dart: build at: %lx, %lx, addr: %x\n", index, npages, uaddr);
 
-	dp = ((unsigned int*)tbl->it_base) + index;
+	orig_dp = dp = ((unsigned int*)tbl->it_base) + index;
 
 	/* On U3, all memory is contiguous, so we can move this
 	 * out of the loop.
@@ -184,11 +204,7 @@ static int dart_build(struct iommu_table *tbl, long index,
 
 		uaddr += DART_PAGE_SIZE;
 	}
-
-	/* make sure all updates have reached memory */
-	mb();
-	in_be32((unsigned __iomem *)dp);
-	mb();
+	dart_cache_sync(orig_dp, npages);
 
 	if (dart_is_u4) {
 		rpn = index;
@@ -203,7 +219,8 @@ static int dart_build(struct iommu_table *tbl, long index,
 
 static void dart_free(struct iommu_table *tbl, long index, long npages)
 {
-	unsigned int *dp;
+	unsigned int *dp, *orig_dp;
+	long orig_npages = npages;
 
 	/* We don't worry about flushing the TLB cache. The only drawback of
 	 * not doing it is that we won't catch buggy device drivers doing
@@ -212,34 +229,29 @@ static void dart_free(struct iommu_table *tbl, long index, long npages)
 
 	DBG("dart: free at: %lx, %lx\n", index, npages);
 
-	dp  = ((unsigned int *)tbl->it_base) + index;
+	orig_dp = dp  = ((unsigned int *)tbl->it_base) + index;
 
 	while (npages--)
 		*(dp++) = dart_emptyval;
-}
 
+	dart_cache_sync(orig_dp, orig_npages);
+}
 
-static int __init dart_init(struct device_node *dart_node)
+static void allocate_dart(void)
 {
-	unsigned int i;
-	unsigned long tmp, base, size;
-	struct resource r;
+	unsigned long tmp;
 
-	if (dart_tablebase == 0 || dart_tablesize == 0) {
-		printk(KERN_INFO "DART: table not allocated, using "
-		       "direct DMA\n");
-		return -ENODEV;
-	}
-
-	if (of_address_to_resource(dart_node, 0, &r))
-		panic("DART: can't get register base ! ");
+	/* 512 pages (2MB) is max DART tablesize. */
+	dart_tablesize = 1UL << 21;
 
-	/* Make sure nothing from the DART range remains in the CPU cache
-	 * from a previous mapping that existed before the kernel took
-	 * over
+	/* 16MB (1 << 24) alignment. We allocate a full 16Mb chuck since we
+	 * will blow up an entire large page anyway in the kernel mapping
 	 */
-	flush_dcache_phys_range(dart_tablebase,
-				dart_tablebase + dart_tablesize);
+	dart_tablebase = __va(memblock_alloc_base(1UL<<24,
+						  1UL<<24, 0x80000000L));
+
+	/* There is no point scanning the DART space for leaks*/
+	kmemleak_no_scan((void *)dart_tablebase);
 
 	/* Allocate a spare page to map all invalid DART pages. We need to do
 	 * that to work around what looks like a problem with the HT bridge
@@ -249,20 +261,50 @@ static int __init dart_init(struct device_node *dart_node)
 	dart_emptyval = DARTMAP_VALID | ((tmp >> DART_PAGE_SHIFT) &
 					 DARTMAP_RPNMASK);
 
+	printk(KERN_INFO "DART table allocated at: %p\n", dart_tablebase);
+}
+
+static int __init dart_init(struct device_node *dart_node)
+{
+	unsigned int i;
+	unsigned long base, size;
+	struct resource r;
+
+	/* IOMMU disabled by the user ? bail out */
+	if (iommu_is_off)
+		return 0;
+
+	/* Only use the DART if the machine has more than 1GB of RAM
+	 * or if requested with iommu=on on cmdline.
+	 *
+	 * 1GB of RAM is picked as limit because some default devices
+	 * (i.e. Airport Extreme) have 30 bit address range limits.
+	 */
+
+	if (!iommu_force_on && memblock_end_of_DRAM() <= 0x40000000ull)
+		return 0;
+
+	/* Get DART registers */
+	if (of_address_to_resource(dart_node, 0, &r))
+		panic("DART: can't get register base ! ");
+
 	/* Map in DART registers */
 	dart = ioremap(r.start, resource_size(&r));
 	if (dart == NULL)
 		panic("DART: Cannot map registers!");
 
-	/* Map in DART table */
-	dart_vbase = ioremap(__pa(dart_tablebase), dart_tablesize);
+	/* Allocate the DART and dummy page */
+	allocate_dart();
 
 	/* Fill initial table */
 	for (i = 0; i < dart_tablesize/4; i++)
-		dart_vbase[i] = dart_emptyval;
+		dart_tablebase[i] = dart_emptyval;
+
+	/* Push to memory */
+	dart_cache_sync(dart_tablebase, dart_tablesize / sizeof(u32));
 
 	/* Initialize DART with table base and enable it. */
-	base = dart_tablebase >> DART_PAGE_SHIFT;
+	base = ((unsigned long)dart_tablebase) >> DART_PAGE_SHIFT;
 	size = dart_tablesize >> DART_PAGE_SHIFT;
 	if (dart_is_u4) {
 		size &= DART_SIZE_U4_SIZE_MASK;
@@ -301,7 +343,7 @@ static void iommu_table_dart_setup(void)
 	iommu_table_dart.it_page_shift = IOMMU_PAGE_SHIFT_4K;
 
 	/* Initialize the common IOMMU code */
-	iommu_table_dart.it_base = (unsigned long)dart_vbase;
+	iommu_table_dart.it_base = (unsigned long)dart_tablebase;
 	iommu_table_dart.it_index = 0;
 	iommu_table_dart.it_blocksize = 1;
 	iommu_table_dart.it_ops = &iommu_dart_ops;
@@ -404,75 +446,21 @@ void __init iommu_init_early_dart(struct pci_controller_ops *controller_ops)
 }
 
 #ifdef CONFIG_PM
-static void iommu_dart_save(void)
-{
-	memcpy(dart_copy, dart_vbase, 2*1024*1024);
-}
-
 static void iommu_dart_restore(void)
 {
-	memcpy(dart_vbase, dart_copy, 2*1024*1024);
+	dart_cache_sync(dart_tablebase, dart_tablesize / sizeof(u32));
 	dart_tlb_invalidate_all();
 }
 
 static int __init iommu_init_late_dart(void)
 {
-	unsigned long tbasepfn;
-	struct page *p;
-
-	/* if no dart table exists then we won't need to save it
-	 * and the area has also not been reserved */
 	if (!dart_tablebase)
 		return 0;
 
-	tbasepfn = __pa(dart_tablebase) >> PAGE_SHIFT;
-	register_nosave_region_late(tbasepfn,
-				    tbasepfn + ((1<<24) >> PAGE_SHIFT));
-
-	/* For suspend we need to copy the dart contents because
-	 * it is not part of the regular mapping (see above) and
-	 * thus not saved automatically. The memory for this copy
-	 * must be allocated early because we need 2 MB. */
-	p = alloc_pages(GFP_KERNEL, 21 - PAGE_SHIFT);
-	BUG_ON(!p);
-	dart_copy = page_address(p);
-
-	ppc_md.iommu_save = iommu_dart_save;
 	ppc_md.iommu_restore = iommu_dart_restore;
 
 	return 0;
 }
 
 late_initcall(iommu_init_late_dart);
-#endif
-
-void __init alloc_dart_table(void)
-{
-	/* Only reserve DART space if machine has more than 1GB of RAM
-	 * or if requested with iommu=on on cmdline.
-	 *
-	 * 1GB of RAM is picked as limit because some default devices
-	 * (i.e. Airport Extreme) have 30 bit address range limits.
-	 */
-
-	if (iommu_is_off)
-		return;
-
-	if (!iommu_force_on && memblock_end_of_DRAM() <= 0x40000000ull)
-		return;
-
-	/* 512 pages (2MB) is max DART tablesize. */
-	dart_tablesize = 1UL << 21;
-	/* 16MB (1 << 24) alignment. We allocate a full 16Mb chuck since we
-	 * will blow up an entire large page anyway in the kernel mapping
-	 */
-	dart_tablebase = (unsigned long)
-		__va(memblock_alloc_base(1UL<<24, 1UL<<24, 0x80000000L));
-	/*
-	 * The DART space is later unmapped from the kernel linear mapping and
-	 * accessing dart_tablebase during kmemleak scanning will fault.
-	 */
-	kmemleak_no_scan((void *)dart_tablebase);
-
-	printk(KERN_INFO "DART table allocated at: %lx\n", dart_tablebase);
-}
+#endif /* CONFIG_PM */
-- 
2.7.4

[PATCH 12/38] powerpc: Move FW feature probing out of pseries probe()

[Benjamin Herrenschmidt]

We move the function itself to pseries/firmware.c and call it along
with almost all other flat device-tree parsers from early_init_devtree()

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
---
 arch/powerpc/include/asm/firmware.h       |  4 +++
 arch/powerpc/kernel/prom.c                |  6 +++++
 arch/powerpc/platforms/pseries/firmware.c | 44 +++++++++++++++++++++++++++++--
 arch/powerpc/platforms/pseries/pseries.h  |  5 ----
 arch/powerpc/platforms/pseries/setup.c    | 41 ----------------------------
 5 files changed, 52 insertions(+), 48 deletions(-)

diff --git a/arch/powerpc/include/asm/firmware.h b/arch/powerpc/include/asm/firmware.h
index b062924..ae0d457 100644
--- a/arch/powerpc/include/asm/firmware.h
+++ b/arch/powerpc/include/asm/firmware.h
@@ -126,6 +126,10 @@ extern int fwnmi_active;
 
 extern unsigned int __start___fw_ftr_fixup, __stop___fw_ftr_fixup;
 
+int pseries_probe_fw_features(unsigned long node,
+			      const char *uname, int depth,
+			      void *data);
+
 #endif /* __ASSEMBLY__ */
 #endif /* __KERNEL__ */
 #endif /* __ASM_POWERPC_FIRMWARE_H */
diff --git a/arch/powerpc/kernel/prom.c b/arch/powerpc/kernel/prom.c
index 48434be..d10e786 100644
--- a/arch/powerpc/kernel/prom.c
+++ b/arch/powerpc/kernel/prom.c
@@ -57,6 +57,7 @@
 #include <asm/fadump.h>
 #include <asm/debug.h>
 #include <asm/epapr_hcalls.h>
+#include <asm/firmware.h>
 
 #include <mm/mmu_decl.h>
 
@@ -740,8 +741,13 @@ void __init early_init_devtree(void *params)
 	/* Scan and build the list of machine check recoverable ranges */
 	of_scan_flat_dt(early_init_dt_scan_recoverable_ranges, NULL);
 #endif
+
 	epapr_paravirt_early_init();
 
+#ifdef CONFIG_PPC_PSERIES
+	/* Now try to figure out if we are running on LPAR */
+	of_scan_flat_dt(pseries_probe_fw_features, NULL);
+#endif
 	DBG(" <- early_init_devtree()\n");
 }
 
diff --git a/arch/powerpc/platforms/pseries/firmware.c b/arch/powerpc/platforms/pseries/firmware.c
index 8c80588..d9a7f62 100644
--- a/arch/powerpc/platforms/pseries/firmware.c
+++ b/arch/powerpc/platforms/pseries/firmware.c
@@ -69,7 +69,8 @@ hypertas_fw_features_table[] = {
  * device-tree/ibm,hypertas-functions.  Ultimately this functionality may
  * be moved into prom.c prom_init().
  */
-void __init fw_hypertas_feature_init(const char *hypertas, unsigned long len)
+static void __init fw_hypertas_feature_init(const char *hypertas,
+					    unsigned long len)
 {
 	const char *s;
 	int i;
@@ -113,7 +114,7 @@ vec5_fw_features_table[] = {
 	{FW_FEATURE_PRRN,		OV5_PRRN},
 };
 
-void __init fw_vec5_feature_init(const char *vec5, unsigned long len)
+static void __init fw_vec5_feature_init(const char *vec5, unsigned long len)
 {
 	unsigned int index, feat;
 	int i;
@@ -131,3 +132,42 @@ void __init fw_vec5_feature_init(const char *vec5, unsigned long len)
 
 	pr_debug(" <- fw_vec5_feature_init()\n");
 }
+
+/*
+ * Called very early, MMU is off, device-tree isn't unflattened
+ */
+
+int __init pseries_probe_fw_features(unsigned long node,
+				     const char *uname, int depth,
+				     void *data)
+{
+	const char *prop;
+	int len;
+	static int hypertas_found;
+	static int vec5_found;
+
+	if (depth != 1)
+		return 0;
+
+	if (!strcmp(uname, "rtas") || !strcmp(uname, "rtas@0")) {
+		prop = of_get_flat_dt_prop(node, "ibm,hypertas-functions",
+					   &len);
+		if (prop) {
+			powerpc_firmware_features |= FW_FEATURE_LPAR;
+			fw_hypertas_feature_init(prop, len);
+		}
+
+		hypertas_found = 1;
+	}
+
+	if (!strcmp(uname, "chosen")) {
+		prop = of_get_flat_dt_prop(node, "ibm,architecture-vec-5",
+					   &len);
+		if (prop)
+			fw_vec5_feature_init(prop, len);
+
+		vec5_found = 1;
+	}
+
+	return hypertas_found && vec5_found;
+}
diff --git a/arch/powerpc/platforms/pseries/pseries.h b/arch/powerpc/platforms/pseries/pseries.h
index 7aa83f0..58a89f9 100644
--- a/arch/powerpc/platforms/pseries/pseries.h
+++ b/arch/powerpc/platforms/pseries/pseries.h
@@ -20,11 +20,6 @@ extern void request_event_sources_irqs(struct device_node *np,
 
 #include <linux/of.h>
 
-extern void __init fw_hypertas_feature_init(const char *hypertas,
-					    unsigned long len);
-extern void __init fw_vec5_feature_init(const char *hypertas,
-					unsigned long len);
-
 struct pt_regs;
 
 extern int pSeries_system_reset_exception(struct pt_regs *regs);
diff --git a/arch/powerpc/platforms/pseries/setup.c b/arch/powerpc/platforms/pseries/setup.c
index 9883bc7..f1fe7aa 100644
--- a/arch/powerpc/platforms/pseries/setup.c
+++ b/arch/powerpc/platforms/pseries/setup.c
@@ -732,45 +732,6 @@ static void pseries_power_off(void)
 	for (;;);
 }
 
-/*
- * Called very early, MMU is off, device-tree isn't unflattened
- */
-
-static int __init pseries_probe_fw_features(unsigned long node,
-					    const char *uname, int depth,
-					    void *data)
-{
-	const char *prop;
-	int len;
-	static int hypertas_found;
-	static int vec5_found;
-
-	if (depth != 1)
-		return 0;
-
-	if (!strcmp(uname, "rtas") || !strcmp(uname, "rtas@0")) {
-		prop = of_get_flat_dt_prop(node, "ibm,hypertas-functions",
-					   &len);
-		if (prop) {
-			powerpc_firmware_features |= FW_FEATURE_LPAR;
-			fw_hypertas_feature_init(prop, len);
-		}
-
-		hypertas_found = 1;
-	}
-
-	if (!strcmp(uname, "chosen")) {
-		prop = of_get_flat_dt_prop(node, "ibm,architecture-vec-5",
-					   &len);
-		if (prop)
-			fw_vec5_feature_init(prop, len);
-
-		vec5_found = 1;
-	}
-
-	return hypertas_found && vec5_found;
-}
-
 static int __init pSeries_probe(void)
 {
 	unsigned long root = of_get_flat_dt_root();
@@ -790,8 +751,6 @@ static int __init pSeries_probe(void)
 
 	pr_debug("pSeries detected, looking for LPAR capability...\n");
 
-	/* Now try to figure out if we are running on LPAR */
-	of_scan_flat_dt(pseries_probe_fw_features, NULL);
 
 #ifdef __LITTLE_ENDIAN__
 	if (firmware_has_feature(FW_FEATURE_SET_MODE)) {
-- 
2.7.4

[PATCH 13/38] powerpc: Put exception configuration in a common place

[Benjamin Herrenschmidt]

The various calls to establish exception endianness and AIL are
now done from a single point using already established CPU and FW
feature bits to decide what to do.

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
---
 arch/powerpc/include/asm/opal.h        |  1 +
 arch/powerpc/kernel/setup_64.c         | 68 +++++++++++++++++++++++++++-------
 arch/powerpc/platforms/powernv/opal.c  | 13 +++----
 arch/powerpc/platforms/pseries/setup.c | 29 ---------------
 4 files changed, 60 insertions(+), 51 deletions(-)

diff --git a/arch/powerpc/include/asm/opal.h b/arch/powerpc/include/asm/opal.h
index 9d86c66..6135816 100644
--- a/arch/powerpc/include/asm/opal.h
+++ b/arch/powerpc/include/asm/opal.h
@@ -215,6 +215,7 @@ extern int early_init_dt_scan_opal(unsigned long node, const char *uname,
 				   int depth, void *data);
 extern int early_init_dt_scan_recoverable_ranges(unsigned long node,
 				 const char *uname, int depth, void *data);
+extern void opal_configure_cores(void);
 
 extern int opal_get_chars(uint32_t vtermno, char *buf, int count);
 extern int opal_put_chars(uint32_t vtermno, const char *buf, int total_len);
diff --git a/arch/powerpc/kernel/setup_64.c b/arch/powerpc/kernel/setup_64.c
index a641753..47a2706 100644
--- a/arch/powerpc/kernel/setup_64.c
+++ b/arch/powerpc/kernel/setup_64.c
@@ -69,6 +69,7 @@
 #include <asm/kvm_ppc.h>
 #include <asm/hugetlb.h>
 #include <asm/livepatch.h>
+#include <asm/opal.h>
 
 #ifdef DEBUG
 #define DBG(fmt...) udbg_printf(fmt)
@@ -205,21 +206,60 @@ static void fixup_boot_paca(void)
 	get_paca()->data_offset = 0;
 }
 
+static void configure_exceptions(void)
+{
+	/* Setup the trampolines from the lowmem exception vectors
+	 * to the kdump kernel when not using a relocatable kernel.
+	 */
+	setup_kdump_trampoline();
+
+	/* Under a PAPR hypervisor, we need hypercalls */
+	if (firmware_has_feature(FW_FEATURE_SET_MODE)) {
+		long rc;
+
+		/* Enable AIL */
+		rc = pSeries_enable_reloc_on_exc();
+		if (rc == H_P2) {
+			pr_info("Relocation on exceptions not supported\n");
+		} else if (rc != H_SUCCESS) {
+			pr_warn("Unable to enable relocation on exceptions: "
+				"%ld\n", rc);
+		}
+
+		/*
+		 * Tell the hypervisor that we want our exceptions to
+		 * be taken in little endian mode. If this fails we don't
+		 * want to use BUG() because it will trigger an exception.
+		 *
+		 * We don't call this for big endian as our calling convention
+		 * makes us always enter in BE, and the call may fail under
+		 * some circumstances with kdump.
+		 */
+#ifdef __LITTLE_ENDIAN__
+		rc = pseries_little_endian_exceptions();
+		if (rc) {
+			ppc_md.progress("H_SET_MODE LE exception fail", 0);
+			panic("Could not enable little endian exceptions");
+		}
+#endif
+	} else {
+		/* Set endian mode using OPAL */
+		if (firmware_has_feature(FW_FEATURE_OPAL))
+			opal_configure_cores();
+
+		/* Enable AIL if supported, and we are in hypervisor mode */
+		if (cpu_has_feature(CPU_FTR_HVMODE) &&
+		    cpu_has_feature(CPU_FTR_ARCH_207S)) {
+			unsigned long lpcr = mfspr(SPRN_LPCR);
+			mtspr(SPRN_LPCR, lpcr | LPCR_AIL_3);
+		}
+	}
+}
+
 static void cpu_ready_for_interrupts(void)
 {
 	/* Set IR and DR in PACA MSR */
 	get_paca()->kernel_msr = MSR_KERNEL;
-
-	/*
-	 * Enable AIL if supported, and we are in hypervisor mode. If we are
-	 * not in hypervisor mode, we enable relocation-on interrupts later
-	 * in pSeries_setup_arch() using the H_SET_MODE hcall.
-	 */
-	if (cpu_has_feature(CPU_FTR_HVMODE) &&
-	    cpu_has_feature(CPU_FTR_ARCH_207S)) {
-		unsigned long lpcr = mfspr(SPRN_LPCR);
-		mtspr(SPRN_LPCR, lpcr | LPCR_AIL_3);
-	}
 }
 
 /*
@@ -276,10 +316,10 @@ void __init early_setup(unsigned long dt_ptr)
 	/* Probe the machine type */
 	probe_machine();
 
-	/* Setup the trampolines from the lowmem exception vectors
-	 * to the kdump kernel when not using a relocatable kernel.
+	/* Configure exception handlers. This include setting up trampolines
+	 * if needed, setting exception endian mode, etc...
 	 */
-	setup_kdump_trampoline();
+	configure_exceptions();
 
 	/* Initialize the hash table or TLB handling */
 	early_init_mmu();
diff --git a/arch/powerpc/platforms/powernv/opal.c b/arch/powerpc/platforms/powernv/opal.c
index 0256d07..802f3b7 100644
--- a/arch/powerpc/platforms/powernv/opal.c
+++ b/arch/powerpc/platforms/powernv/opal.c
@@ -56,7 +56,7 @@ static DEFINE_SPINLOCK(opal_write_lock);
 static struct atomic_notifier_head opal_msg_notifier_head[OPAL_MSG_TYPE_MAX];
 static uint32_t opal_heartbeat;
 
-static void opal_reinit_cores(void)
+void opal_configure_cores(void)
 {
 	/* Do the actual re-init, This will clobber all FPRs, VRs, etc...
 	 *
@@ -69,6 +69,10 @@ static void opal_reinit_cores(void)
 #else
 	opal_reinit_cpus(OPAL_REINIT_CPUS_HILE_LE);
 #endif
+
+	/* Restore some bits */
+	if (cur_cpu_spec->cpu_restore)
+		cur_cpu_spec->cpu_restore();
 }
 
 int __init early_init_dt_scan_opal(unsigned long node,
@@ -105,13 +109,6 @@ int __init early_init_dt_scan_opal(unsigned long node,
 		panic("OPAL != V3 detected, no longer supported.\n");
 	}
 
-	/* Reinit all cores with the right endian */
-	opal_reinit_cores();
-
-	/* Restore some bits */
-	if (cur_cpu_spec->cpu_restore)
-		cur_cpu_spec->cpu_restore();
-
 	return 1;
 }
 
diff --git a/arch/powerpc/platforms/pseries/setup.c b/arch/powerpc/platforms/pseries/setup.c
index f1fe7aa..6dc91a4 100644
--- a/arch/powerpc/platforms/pseries/setup.c
+++ b/arch/powerpc/platforms/pseries/setup.c
@@ -537,18 +537,6 @@ static void __init pSeries_setup_arch(void)
 	}
 
 	ppc_md.pcibios_root_bridge_prepare = pseries_root_bridge_prepare;
-
-	if (firmware_has_feature(FW_FEATURE_SET_MODE)) {
-		long rc;
-
-		rc = pSeries_enable_reloc_on_exc();
-		if (rc == H_P2) {
-			pr_info("Relocation on exceptions not supported\n");
-		} else if (rc != H_SUCCESS) {
-			pr_warn("Unable to enable relocation on exceptions: "
-				"%ld\n", rc);
-		}
-	}
 }
 
 static int __init pSeries_init_panel(void)
@@ -751,23 +739,6 @@ static int __init pSeries_probe(void)
 
 	pr_debug("pSeries detected, looking for LPAR capability...\n");
 
-
-#ifdef __LITTLE_ENDIAN__
-	if (firmware_has_feature(FW_FEATURE_SET_MODE)) {
-		long rc;
-		/*
-		 * Tell the hypervisor that we want our exceptions to
-		 * be taken in little endian mode. If this fails we don't
-		 * want to use BUG() because it will trigger an exception.
-		 */
-		rc = pseries_little_endian_exceptions();
-		if (rc) {
-			ppc_md.progress("H_SET_MODE LE exception fail", 0);
-			panic("Could not enable little endian exceptions");
-		}
-	}
-#endif
-
 	if (firmware_has_feature(FW_FEATURE_LPAR))
 		hpte_init_lpar();
 	else
-- 
2.7.4

[PATCH 14/38] powerpc/pmac: Remove early allocation of the SMU command buffer

[Benjamin Herrenschmidt]

The SMU command buffer needs to be allocated below 2G using memblock.

In the past, this had to be done very early from the arch code as
memblock wasn't available past that point. That is no longer the
case though, smu_init() is called from setup_arch() when memblock
is still functional these days. So move the allocation to the
SMU driver itself.

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
---
 arch/powerpc/include/asm/smu.h          |  7 -------
 arch/powerpc/platforms/powermac/setup.c | 16 ----------------
 drivers/macintosh/smu.c                 |  9 ++++++++-
 3 files changed, 8 insertions(+), 24 deletions(-)

diff --git a/arch/powerpc/include/asm/smu.h b/arch/powerpc/include/asm/smu.h
index f280dd1..ddee470 100644
--- a/arch/powerpc/include/asm/smu.h
+++ b/arch/powerpc/include/asm/smu.h
@@ -471,13 +471,6 @@ extern int smu_get_rtc_time(struct rtc_time *time, int spinwait);
 extern int smu_set_rtc_time(struct rtc_time *time, int spinwait);
 
 /*
- * SMU command buffer absolute address, exported by pmac_setup,
- * this is allocated very early during boot.
- */
-extern unsigned long smu_cmdbuf_abs;
-
-
-/*
  * Kernel asynchronous i2c interface
  */
 
diff --git a/arch/powerpc/platforms/powermac/setup.c b/arch/powerpc/platforms/powermac/setup.c
index 19de197..43d02c2 100644
--- a/arch/powerpc/platforms/powermac/setup.c
+++ b/arch/powerpc/platforms/powermac/setup.c
@@ -52,7 +52,6 @@
 #include <linux/suspend.h>
 #include <linux/of_device.h>
 #include <linux/of_platform.h>
-#include <linux/memblock.h>
 
 #include <asm/reg.h>
 #include <asm/sections.h>
@@ -97,11 +96,6 @@ int sccdbg;
 sys_ctrler_t sys_ctrler = SYS_CTRLER_UNKNOWN;
 EXPORT_SYMBOL(sys_ctrler);
 
-#ifdef CONFIG_PMAC_SMU
-unsigned long smu_cmdbuf_abs;
-EXPORT_SYMBOL(smu_cmdbuf_abs);
-#endif
-
 static void pmac_show_cpuinfo(struct seq_file *m)
 {
 	struct device_node *np;
@@ -325,7 +319,6 @@ static void __init pmac_setup_arch(void)
     defined(CONFIG_PPC64)
 	pmac_nvram_init();
 #endif
-
 #ifdef CONFIG_PPC32
 #ifdef CONFIG_BLK_DEV_INITRD
 	if (initrd_start)
@@ -615,15 +608,6 @@ static int __init pmac_probe(void)
 	DMA_MODE_WRITE = 2;
 #endif /* CONFIG_PPC32 */
 
-#ifdef CONFIG_PMAC_SMU
-	/*
-	 * SMU based G5s need some memory below 2Gb, at least the current
-	 * driver needs that. We have to allocate it now. We allocate 4k
-	 * (1 small page) for now.
-	 */
-	smu_cmdbuf_abs = memblock_alloc_base(4096, 4096, 0x80000000UL);
-#endif /* CONFIG_PMAC_SMU */
-
 	pm_power_off = pmac_power_off;
 
 	return 1;
diff --git a/drivers/macintosh/smu.c b/drivers/macintosh/smu.c
index d531f80..d6f72c8 100644
--- a/drivers/macintosh/smu.c
+++ b/drivers/macintosh/smu.c
@@ -38,6 +38,7 @@
 #include <linux/of_irq.h>
 #include <linux/of_platform.h>
 #include <linux/slab.h>
+#include <linux/memblock.h>
 
 #include <asm/byteorder.h>
 #include <asm/io.h>
@@ -99,6 +100,7 @@ static DEFINE_MUTEX(smu_mutex);
 static struct smu_device	*smu;
 static DEFINE_MUTEX(smu_part_access);
 static int smu_irq_inited;
+static unsigned long smu_cmdbuf_abs;
 
 static void smu_i2c_retry(unsigned long data);
 
@@ -479,8 +481,13 @@ int __init smu_init (void)
 
 	printk(KERN_INFO "SMU: Driver %s %s\n", VERSION, AUTHOR);
 
+	/*
+	 * SMU based G5s need some memory below 2Gb. Thankfully this is
+	 * called at a time where memblock is still available.
+	 */
+	smu_cmdbuf_abs = memblock_alloc_base(4096, 4096, 0x80000000UL);
 	if (smu_cmdbuf_abs == 0) {
-		printk(KERN_ERR "SMU: Command buffer not allocated !\n");
+		printk(KERN_ERR "SMU: Command buffer allocation failed !\n");
 		ret = -EINVAL;
 		goto fail_np;
 	}
-- 
2.7.4

[PATCH 15/38] powerpc/64: Move MMU backend selection out of platform code

[Benjamin Herrenschmidt]

We move it into early_mmu_init() based on firmware features. For PS3,
we have to move the setting of these into early_init_devtree().

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
---
 arch/powerpc/include/asm/ps3.h          |  2 ++
 arch/powerpc/kernel/prom.c              |  7 +++++++
 arch/powerpc/mm/hash_utils_64.c         |  9 +++++++++
 arch/powerpc/mm/pgtable-radix.c         |  1 +
 arch/powerpc/platforms/cell/setup.c     |  1 -
 arch/powerpc/platforms/maple/setup.c    |  1 -
 arch/powerpc/platforms/pasemi/setup.c   |  2 --
 arch/powerpc/platforms/powermac/setup.c |  4 ----
 arch/powerpc/platforms/powernv/setup.c  |  5 -----
 arch/powerpc/platforms/ps3/setup.c      | 15 +++++++++------
 arch/powerpc/platforms/pseries/setup.c  |  7 -------
 11 files changed, 28 insertions(+), 26 deletions(-)

diff --git a/arch/powerpc/include/asm/ps3.h b/arch/powerpc/include/asm/ps3.h
index a1bc7e7..a19f831 100644
--- a/arch/powerpc/include/asm/ps3.h
+++ b/arch/powerpc/include/asm/ps3.h
@@ -526,4 +526,6 @@ void ps3_sync_irq(int node);
 u32 ps3_get_hw_thread_id(int cpu);
 u64 ps3_get_spe_id(void *arg);
 
+void ps3_early_mm_init(void);
+
 #endif
diff --git a/arch/powerpc/kernel/prom.c b/arch/powerpc/kernel/prom.c
index d10e786..5d59f11 100644
--- a/arch/powerpc/kernel/prom.c
+++ b/arch/powerpc/kernel/prom.c
@@ -748,6 +748,13 @@ void __init early_init_devtree(void *params)
 	/* Now try to figure out if we are running on LPAR */
 	of_scan_flat_dt(pseries_probe_fw_features, NULL);
 #endif
+
+#ifdef CONFIG_PPC_PS3
+	/* Identify PS3 firmware */
+	if (of_flat_dt_is_compatible(of_get_flat_dt_root(), "sony,ps3"))
+		powerpc_firmware_features |= FW_FEATURE_PS3_POSSIBLE;
+#endif
+
 	DBG(" <- early_init_devtree()\n");
 }
 
diff --git a/arch/powerpc/mm/hash_utils_64.c b/arch/powerpc/mm/hash_utils_64.c
index 2bea864..e4f796f 100644
--- a/arch/powerpc/mm/hash_utils_64.c
+++ b/arch/powerpc/mm/hash_utils_64.c
@@ -58,6 +58,7 @@
 #include <asm/firmware.h>
 #include <asm/tm.h>
 #include <asm/trace.h>
+#include <asm/ps3.h>
 
 #ifdef DEBUG
 #define DBG(fmt...) udbg_printf(fmt)
@@ -886,6 +887,14 @@ void __init hash__early_init_mmu(void)
 	vmemmap = (struct page *)H_VMEMMAP_BASE;
 	ioremap_bot = IOREMAP_BASE;
 
+	/* Select appropriate backend */
+	if (firmware_has_feature(FW_FEATURE_PS3_LV1))
+		ps3_early_mm_init();
+	else if (firmware_has_feature(FW_FEATURE_LPAR))
+		hpte_init_lpar();
+	else
+		hpte_init_native();
+
 	/* Initialize the MMU Hash table and create the linear mapping
 	 * of memory. Has to be done before SLB initialization as this is
 	 * currently where the page size encoding is obtained.
diff --git a/arch/powerpc/mm/pgtable-radix.c b/arch/powerpc/mm/pgtable-radix.c
index c939e6e..6624b7b 100644
--- a/arch/powerpc/mm/pgtable-radix.c
+++ b/arch/powerpc/mm/pgtable-radix.c
@@ -339,6 +339,7 @@ void __init radix__early_init_mmu(void)
 
 	radix_init_page_sizes();
 	if (!firmware_has_feature(FW_FEATURE_LPAR)) {
+		radix_init_native();
 		lpcr = mfspr(SPRN_LPCR);
 		mtspr(SPRN_LPCR, lpcr | LPCR_UPRT);
 		radix_init_partition_table();
diff --git a/arch/powerpc/platforms/cell/setup.c b/arch/powerpc/platforms/cell/setup.c
index 36cff28..e342f78 100644
--- a/arch/powerpc/platforms/cell/setup.c
+++ b/arch/powerpc/platforms/cell/setup.c
@@ -261,7 +261,6 @@ static int __init cell_probe(void)
 	    !of_flat_dt_is_compatible(root, "IBM,CPBW-1.0"))
 		return 0;
 
-	hpte_init_native();
 	pm_power_off = rtas_power_off;
 
 	return 1;
diff --git a/arch/powerpc/platforms/maple/setup.c b/arch/powerpc/platforms/maple/setup.c
index 3cd625d..a56828c 100644
--- a/arch/powerpc/platforms/maple/setup.c
+++ b/arch/powerpc/platforms/maple/setup.c
@@ -304,7 +304,6 @@ static int __init maple_probe(void)
 	    !of_flat_dt_is_compatible(root, "Momentum,Apache"))
 		return 0;
 
-	hpte_init_native();
 	pm_power_off = maple_power_off;
 
 	return 1;
diff --git a/arch/powerpc/platforms/pasemi/setup.c b/arch/powerpc/platforms/pasemi/setup.c
index d71b2c7..7605bc6 100644
--- a/arch/powerpc/platforms/pasemi/setup.c
+++ b/arch/powerpc/platforms/pasemi/setup.c
@@ -426,8 +426,6 @@ static int __init pas_probe(void)
 	    !of_flat_dt_is_compatible(root, "pasemi,pwrficient"))
 		return 0;
 
-	hpte_init_native();
-
 	alloc_iobmap_l2();
 
 	return 1;
diff --git a/arch/powerpc/platforms/powermac/setup.c b/arch/powerpc/platforms/powermac/setup.c
index 43d02c2..5801889 100644
--- a/arch/powerpc/platforms/powermac/setup.c
+++ b/arch/powerpc/platforms/powermac/setup.c
@@ -597,10 +597,6 @@ static int __init pmac_probe(void)
 	    !of_flat_dt_is_compatible(root, "MacRISC"))
 		return 0;
 
-#ifdef CONFIG_PPC64
-	hpte_init_native();
-#endif
-
 #ifdef CONFIG_PPC32
 	/* isa_io_base gets set in pmac_pci_init */
 	ISA_DMA_THRESHOLD = ~0L;
diff --git a/arch/powerpc/platforms/powernv/setup.c b/arch/powerpc/platforms/powernv/setup.c
index ee6430b..15e3b8b 100644
--- a/arch/powerpc/platforms/powernv/setup.c
+++ b/arch/powerpc/platforms/powernv/setup.c
@@ -273,11 +273,6 @@ static int __init pnv_probe(void)
 	if (!of_flat_dt_is_compatible(root, "ibm,powernv"))
 		return 0;
 
-	if (IS_ENABLED(CONFIG_PPC_RADIX_MMU) && radix_enabled())
-		radix_init_native();
-	else if (IS_ENABLED(CONFIG_PPC_STD_MMU_64))
-		hpte_init_native();
-
 	if (firmware_has_feature(FW_FEATURE_OPAL))
 		pnv_setup_machdep_opal();
 
diff --git a/arch/powerpc/platforms/ps3/setup.c b/arch/powerpc/platforms/ps3/setup.c
index 799c858..e5ee80d 100644
--- a/arch/powerpc/platforms/ps3/setup.c
+++ b/arch/powerpc/platforms/ps3/setup.c
@@ -226,9 +226,17 @@ static void __init ps3_progress(char *s, unsigned short hex)
 	printk("*** %04x : %s\n", hex, s ? s : "");
 }
 
-static int __init ps3_probe(void)
+void __init ps3_early_mm_init(void)
 {
 	unsigned long htab_size;
+
+	ps3_mm_init();
+	ps3_mm_vas_create(&htab_size);
+	ps3_hpte_init(htab_size);
+}
+
+static int __init ps3_probe(void)
+{
 	unsigned long dt_root;
 
 	DBG(" -> %s:%d\n", __func__, __LINE__);
@@ -237,12 +245,7 @@ static int __init ps3_probe(void)
 	if (!of_flat_dt_is_compatible(dt_root, "sony,ps3"))
 		return 0;
 
-	powerpc_firmware_features |= FW_FEATURE_PS3_POSSIBLE;
-
 	ps3_os_area_save_params();
-	ps3_mm_init();
-	ps3_mm_vas_create(&htab_size);
-	ps3_hpte_init(htab_size);
 	pm_power_off = ps3_power_off;
 
 	DBG(" <- %s:%d\n", __func__, __LINE__);
diff --git a/arch/powerpc/platforms/pseries/setup.c b/arch/powerpc/platforms/pseries/setup.c
index 6dc91a4..a8e33424 100644
--- a/arch/powerpc/platforms/pseries/setup.c
+++ b/arch/powerpc/platforms/pseries/setup.c
@@ -737,13 +737,6 @@ static int __init pSeries_probe(void)
 	    of_flat_dt_is_compatible(root, "IBM,CBEA"))
 		return 0;
 
-	pr_debug("pSeries detected, looking for LPAR capability...\n");
-
-	if (firmware_has_feature(FW_FEATURE_LPAR))
-		hpte_init_lpar();
-	else
-		hpte_init_native();
-
 	pm_power_off = pseries_power_off;
 
 	pr_debug("Machine is%s LPAR !\n",
-- 
2.7.4

[PATCH 16/38] powerpc/pasemi: Remove IOBMAP allocation from platform probe()

[Benjamin Herrenschmidt]

These days, memblocks is available later, so we can just allocate it
as part of iob_init.

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
---
 arch/powerpc/platforms/pasemi/iommu.c  | 15 +++++----------
 arch/powerpc/platforms/pasemi/pasemi.h |  1 -
 arch/powerpc/platforms/pasemi/setup.c  |  2 --
 3 files changed, 5 insertions(+), 13 deletions(-)

diff --git a/arch/powerpc/platforms/pasemi/iommu.c b/arch/powerpc/platforms/pasemi/iommu.c
index c929644..43dd3fb 100644
--- a/arch/powerpc/platforms/pasemi/iommu.c
+++ b/arch/powerpc/platforms/pasemi/iommu.c
@@ -202,6 +202,11 @@ int __init iob_init(struct device_node *dn)
 
 	pr_debug(" -> %s\n", __func__);
 
+	/* For 2G space, 8x64 pages (2^21 bytes) is max total l2 size */
+	iob_l2_base = (u32 *)__va(memblock_alloc_base(1UL<<21, 1UL<<21, 0x80000000));
+
+	printk(KERN_INFO "IOBMAP L2 allocated at: %p\n", iob_l2_base);
+
 	/* Allocate a spare page to map all invalid IOTLB pages. */
 	tmp = memblock_alloc(IOBMAP_PAGE_SIZE, IOBMAP_PAGE_SIZE);
 	if (!tmp)
@@ -260,13 +265,3 @@ void __init iommu_init_early_pasemi(void)
 	set_pci_dma_ops(&dma_iommu_ops);
 }
 
-void __init alloc_iobmap_l2(void)
-{
-#ifndef CONFIG_PPC_PASEMI_IOMMU
-	return;
-#endif
-	/* For 2G space, 8x64 pages (2^21 bytes) is max total l2 size */
-	iob_l2_base = (u32 *)__va(memblock_alloc_base(1UL<<21, 1UL<<21, 0x80000000));
-
-	printk(KERN_INFO "IOBMAP L2 allocated at: %p\n", iob_l2_base);
-}
diff --git a/arch/powerpc/platforms/pasemi/pasemi.h b/arch/powerpc/platforms/pasemi/pasemi.h
index 11f230a..74cbcb3 100644
--- a/arch/powerpc/platforms/pasemi/pasemi.h
+++ b/arch/powerpc/platforms/pasemi/pasemi.h
@@ -8,7 +8,6 @@ extern void pas_pci_dma_dev_setup(struct pci_dev *dev);
 
 extern void __iomem *pasemi_pci_getcfgaddr(struct pci_dev *dev, int offset);
 
-extern void __init alloc_iobmap_l2(void);
 extern void __init pasemi_map_registers(void);
 
 /* Power savings modes, implemented in asm */
diff --git a/arch/powerpc/platforms/pasemi/setup.c b/arch/powerpc/platforms/pasemi/setup.c
index 7605bc6..9a881be 100644
--- a/arch/powerpc/platforms/pasemi/setup.c
+++ b/arch/powerpc/platforms/pasemi/setup.c
@@ -426,8 +426,6 @@ static int __init pas_probe(void)
 	    !of_flat_dt_is_compatible(root, "pasemi,pwrficient"))
 		return 0;
 
-	alloc_iobmap_l2();
-
 	return 1;
 }
 
-- 
2.7.4

[PATCH 17/38] powerpc/mm/hash: Don't use machine_is() early during boot

[Benjamin Herrenschmidt]

Use the device-tree instead as we'll be moving probe_machine()
out of early_setup

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
---
 arch/powerpc/mm/hash_utils_64.c | 25 ++++++++++++++++---------
 1 file changed, 16 insertions(+), 9 deletions(-)

diff --git a/arch/powerpc/mm/hash_utils_64.c b/arch/powerpc/mm/hash_utils_64.c
index e4f796f..a70d0a3 100644
--- a/arch/powerpc/mm/hash_utils_64.c
+++ b/arch/powerpc/mm/hash_utils_64.c
@@ -34,6 +34,7 @@
 #include <linux/signal.h>
 #include <linux/memblock.h>
 #include <linux/context_tracking.h>
+#include <linux/libfdt.h>
 
 #include <asm/processor.h>
 #include <asm/pgtable.h>
@@ -731,7 +732,7 @@ static void __init htab_initialize(void)
 	unsigned long table;
 	unsigned long pteg_count;
 	unsigned long prot;
-	unsigned long base = 0, size = 0, limit;
+	unsigned long base = 0, size = 0;
 	struct memblock_region *reg;
 
 	DBG(" -> htab_initialize()\n");
@@ -757,7 +758,8 @@ static void __init htab_initialize(void)
 
 	htab_hash_mask = pteg_count - 1;
 
-	if (firmware_has_feature(FW_FEATURE_LPAR)) {
+	if (firmware_has_feature(FW_FEATURE_LPAR) ||
+	    firmware_has_feature(FW_FEATURE_PS3_LV1)) {
 		/* Using a hypervisor which owns the htab */
 		htab_address = NULL;
 		_SDR1 = 0; 
@@ -772,16 +774,21 @@ static void __init htab_initialize(void)
 			ppc_md.hpte_clear_all();
 #endif
 	} else {
-		/* Find storage for the HPT.  Must be contiguous in
-		 * the absolute address space. On cell we want it to be
-		 * in the first 2 Gig so we can use it for IOMMU hacks.
+		unsigned long limit = MEMBLOCK_ALLOC_ANYWHERE;
+
+#ifdef CONFIG_PPC_CELL
+		/* Cell may require the hash table down low when using the
+		 * Axon IOMMU in order to fit the dynamic region over it, see
+		 * comments in cell/iommu.c
 		 */
-		if (machine_is(cell))
+		if (fdt_subnode_offset(initial_boot_params, 0, "axon") > 0) {
 			limit = 0x80000000;
-		else
-			limit = MEMBLOCK_ALLOC_ANYWHERE;
+			pr_info("Hash table forced below 2G for Axon IOMMU\n");
+		}
+#endif /* CONFIG_PPC_CELL */
 
-		table = memblock_alloc_base(htab_size_bytes, htab_size_bytes, limit);
+		table = memblock_alloc_base(htab_size_bytes, htab_size_bytes,
+					    limit);
 
 		DBG("Hash table allocated at %lx, size: %lx\n", table,
 		    htab_size_bytes);
-- 
2.7.4

[PATCH 18/38] powerpc/rtas: Don't test for machine type in rtas_initialize()

[Benjamin Herrenschmidt]

The test is unnecessary, the FW_FEATURE_LPAR is sufficient as there
exist no other LPAR type that has RTAS.

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
---
 arch/powerpc/kernel/rtas.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/arch/powerpc/kernel/rtas.c b/arch/powerpc/kernel/rtas.c
index 28736ff..030a4d5 100644
--- a/arch/powerpc/kernel/rtas.c
+++ b/arch/powerpc/kernel/rtas.c
@@ -1174,7 +1174,7 @@ void __init rtas_initialize(void)
 	 * the stop-self token if any
 	 */
 #ifdef CONFIG_PPC64
-	if (machine_is(pseries) && firmware_has_feature(FW_FEATURE_LPAR)) {
+	if (firmware_has_feature(FW_FEATURE_LPAR)) {
 		rtas_region = min(ppc64_rma_size, RTAS_INSTANTIATE_MAX);
 		ibm_suspend_me_token = rtas_token("ibm,suspend-me");
 	}
-- 
2.7.4

[PATCH 19/38] powerpc: Don't test for machine type in smp_setup_cpu_maps()

[Benjamin Herrenschmidt]

The subsequent test for RTAS along with the LPAR test are sufficient

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
---
 arch/powerpc/kernel/setup-common.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/arch/powerpc/kernel/setup-common.c b/arch/powerpc/kernel/setup-common.c
index 8ca79b7..2a3564c 100644
--- a/arch/powerpc/kernel/setup-common.c
+++ b/arch/powerpc/kernel/setup-common.c
@@ -494,7 +494,7 @@ void __init smp_setup_cpu_maps(void)
 	 * On pSeries LPAR, we need to know how many cpus
 	 * could possibly be added to this partition.
 	 */
-	if (machine_is(pseries) && firmware_has_feature(FW_FEATURE_LPAR) &&
+	if (firmware_has_feature(FW_FEATURE_LPAR) &&
 	    (dn = of_find_node_by_path("/rtas"))) {
 		int num_addr_cell, num_size_cell, maxcpus;
 		const __be32 *ireg;
-- 
2.7.4

[PATCH 20/38] powerpc/mm/hash64: Don't test for machine type to detect HEA special case

[Benjamin Herrenschmidt]

Instead, check for FW_FEATURE_SPLPAR. This should be roughtly equivalent
as all pseries machiens that can have an HEA also support SPLPAR and
no other machine type does.

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
---
 arch/powerpc/mm/hash_utils_64.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/arch/powerpc/mm/hash_utils_64.c b/arch/powerpc/mm/hash_utils_64.c
index a70d0a3..164c2da 100644
--- a/arch/powerpc/mm/hash_utils_64.c
+++ b/arch/powerpc/mm/hash_utils_64.c
@@ -508,7 +508,8 @@ static bool might_have_hea(void)
 	 * we will never see an HEA ethernet device.
 	 */
 #ifdef CONFIG_IBMEBUS
-	return !cpu_has_feature(CPU_FTR_ARCH_207S);
+	return !cpu_has_feature(CPU_FTR_ARCH_207S) &&
+		!firmware_has_feature(FW_FEATURE_SPLPAR);
 #else
 	return false;
 #endif
@@ -574,7 +575,7 @@ found:
 			 * would stop us accessing the HEA ethernet. So if we
 			 * have the chance of ever seeing one, stay at 4k.
 			 */
-			if (!might_have_hea() || !machine_is(pseries))
+			if (!might_have_hea())
 				mmu_io_psize = MMU_PAGE_64K;
 		} else
 			mmu_ci_restrictions = 1;
-- 
2.7.4

[PATCH 21/38] powerpc/pmac: Remove spurrious machine type test

[Benjamin Herrenschmidt]

pmac_declare_of_platform_devices() is already a machine initcall, thus
it won't be called on a non-powermac machine. Testing for chrp there
is pointless.

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
---
 arch/powerpc/platforms/powermac/setup.c | 3 ---
 1 file changed, 3 deletions(-)

diff --git a/arch/powerpc/platforms/powermac/setup.c b/arch/powerpc/platforms/powermac/setup.c
index 5801889..1e42104 100644
--- a/arch/powerpc/platforms/powermac/setup.c
+++ b/arch/powerpc/platforms/powermac/setup.c
@@ -482,9 +482,6 @@ static int __init pmac_declare_of_platform_devices(void)
 {
 	struct device_node *np;
 
-	if (machine_is(chrp))
-		return -1;
-
 	np = of_find_node_by_name(NULL, "valkyrie");
 	if (np) {
 		of_platform_device_create(np, "valkyrie", NULL);
-- 
2.7.4

[PATCH 22/38] powerpc/mm: Move hash table ops to a separate structure

[Benjamin Herrenschmidt]

Moving probe_machine() to after mmu init will cause the ppc_md
fields relative to the hash table management to be overwritten.

Since we have essentially disconnected the machine type from
the hash backend ops, finish the job by moving them to a different
structure.

The only callback that didn't quite fix is update_partition_table
since this is not specific to hash, so I moved it to a standalone
variable for now. We can revisit later if needed.

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
---
 arch/powerpc/include/asm/book3s/64/mmu-hash.h | 39 ++++++++++++++++++++
 arch/powerpc/include/asm/book3s/64/mmu.h      |  2 ++
 arch/powerpc/include/asm/machdep.h            | 37 -------------------
 arch/powerpc/kernel/machine_kexec_64.c        |  4 +--
 arch/powerpc/kernel/misc_64.S                 |  2 +-
 arch/powerpc/kvm/book3s_64_mmu_host.c         | 18 +++++-----
 arch/powerpc/mm/hash64_4k.c                   | 18 +++++-----
 arch/powerpc/mm/hash64_64k.c                  | 39 +++++++++++---------
 arch/powerpc/mm/hash_native_64.c              | 18 +++++-----
 arch/powerpc/mm/hash_utils_64.c               | 51 +++++++++++++++------------
 arch/powerpc/mm/hugepage-hash64.c             | 17 ++++-----
 arch/powerpc/mm/hugetlbpage-hash64.c          |  4 +--
 arch/powerpc/mm/pgtable-book3s64.c            |  2 ++
 arch/powerpc/mm/pgtable-radix.c               |  7 ++--
 arch/powerpc/platforms/ps3/htab.c             | 12 +++----
 arch/powerpc/platforms/pseries/lpar.c         | 18 +++++-----
 16 files changed, 156 insertions(+), 132 deletions(-)

diff --git a/arch/powerpc/include/asm/book3s/64/mmu-hash.h b/arch/powerpc/include/asm/book3s/64/mmu-hash.h
index 290157e..c2baabe 100644
--- a/arch/powerpc/include/asm/book3s/64/mmu-hash.h
+++ b/arch/powerpc/include/asm/book3s/64/mmu-hash.h
@@ -123,6 +123,45 @@
 
 #ifndef __ASSEMBLY__
 
+struct mmu_hash_ops {
+	void            (*hpte_invalidate)(unsigned long slot,
+					   unsigned long vpn,
+					   int bpsize, int apsize,
+					   int ssize, int local);
+	long		(*hpte_updatepp)(unsigned long slot,
+					 unsigned long newpp,
+					 unsigned long vpn,
+					 int bpsize, int apsize,
+					 int ssize, unsigned long flags);
+	void            (*hpte_updateboltedpp)(unsigned long newpp,
+					       unsigned long ea,
+					       int psize, int ssize);
+	long		(*hpte_insert)(unsigned long hpte_group,
+				       unsigned long vpn,
+				       unsigned long prpn,
+				       unsigned long rflags,
+				       unsigned long vflags,
+				       int psize, int apsize,
+				       int ssize);
+	long		(*hpte_remove)(unsigned long hpte_group);
+	int             (*hpte_removebolted)(unsigned long ea,
+					     int psize, int ssize);
+	void		(*flush_hash_range)(unsigned long number, int local);
+	void		(*hugepage_invalidate)(unsigned long vsid,
+					       unsigned long addr,
+					       unsigned char *hpte_slot_array,
+					       int psize, int ssize, int local);
+	/*
+	 * Special for kexec.
+	 * To be called in real mode with interrupts disabled. No locks are
+	 * taken as such, concurrent access on pre POWER5 hardware could result
+	 * in a deadlock.
+	 * The linear mapping is destroyed as well.
+	 */
+	void		(*hpte_clear_all)(void);
+};
+extern struct mmu_hash_ops mmu_hash_ops;
+
 struct hash_pte {
 	__be64 v;
 	__be64 r;
diff --git a/arch/powerpc/include/asm/book3s/64/mmu.h b/arch/powerpc/include/asm/book3s/64/mmu.h
index 5854263..6a90efe 100644
--- a/arch/powerpc/include/asm/book3s/64/mmu.h
+++ b/arch/powerpc/include/asm/book3s/64/mmu.h
@@ -25,6 +25,8 @@ extern struct mmu_psize_def mmu_psize_defs[MMU_PAGE_COUNT];
 
 #define radix_enabled() mmu_has_feature(MMU_FTR_RADIX)
 
+extern int (*ppc64_update_partition_table)(u64);
+
 #endif /* __ASSEMBLY__ */
 
 /* 64-bit classic hash table MMU */
diff --git a/arch/powerpc/include/asm/machdep.h b/arch/powerpc/include/asm/machdep.h
index 6bdcd0d..62b1461 100644
--- a/arch/powerpc/include/asm/machdep.h
+++ b/arch/powerpc/include/asm/machdep.h
@@ -34,42 +34,6 @@ struct pci_host_bridge;
 struct machdep_calls {
 	char		*name;
 #ifdef CONFIG_PPC64
-	void            (*hpte_invalidate)(unsigned long slot,
-					   unsigned long vpn,
-					   int bpsize, int apsize,
-					   int ssize, int local);
-	long		(*hpte_updatepp)(unsigned long slot, 
-					 unsigned long newpp, 
-					 unsigned long vpn,
-					 int bpsize, int apsize,
-					 int ssize, unsigned long flags);
-	void            (*hpte_updateboltedpp)(unsigned long newpp, 
-					       unsigned long ea,
-					       int psize, int ssize);
-	long		(*hpte_insert)(unsigned long hpte_group,
-				       unsigned long vpn,
-				       unsigned long prpn,
-				       unsigned long rflags,
-				       unsigned long vflags,
-				       int psize, int apsize,
-				       int ssize);
-	long		(*hpte_remove)(unsigned long hpte_group);
-	int             (*hpte_removebolted)(unsigned long ea,
-					     int psize, int ssize);
-	void		(*flush_hash_range)(unsigned long number, int local);
-	void		(*hugepage_invalidate)(unsigned long vsid,
-					       unsigned long addr,
-					       unsigned char *hpte_slot_array,
-					       int psize, int ssize, int local);
-	/*
-	 * Special for kexec.
-	 * To be called in real mode with interrupts disabled. No locks are
-	 * taken as such, concurrent access on pre POWER5 hardware could result
-	 * in a deadlock.
-	 * The linear mapping is destroyed as well.
-	 */
-	void		(*hpte_clear_all)(void);
-
 	void __iomem *	(*ioremap)(phys_addr_t addr, unsigned long size,
 				   unsigned long flags, void *caller);
 	void		(*iounmap)(volatile void __iomem *token);
@@ -256,7 +220,6 @@ struct machdep_calls {
 #ifdef CONFIG_ARCH_RANDOM
 	int (*get_random_seed)(unsigned long *v);
 #endif
-	int (*update_partition_table)(u64);
 };
 
 extern void e500_idle(void);
diff --git a/arch/powerpc/kernel/machine_kexec_64.c b/arch/powerpc/kernel/machine_kexec_64.c
index b8c202d..021cc7c 100644
--- a/arch/powerpc/kernel/machine_kexec_64.c
+++ b/arch/powerpc/kernel/machine_kexec_64.c
@@ -54,7 +54,7 @@ int default_machine_kexec_prepare(struct kimage *image)
 	const unsigned long *basep;
 	const unsigned int *sizep;
 
-	if (!ppc_md.hpte_clear_all)
+	if (!mmu_hash_ops.hpte_clear_all)
 		return -ENOENT;
 
 	/*
@@ -379,7 +379,7 @@ void default_machine_kexec(struct kimage *image)
 	 */
 	kexec_sequence(&kexec_stack, image->start, image,
 			page_address(image->control_code_page),
-			ppc_md.hpte_clear_all);
+			mmu_hash_ops.hpte_clear_all);
 	/* NOTREACHED */
 }
 
diff --git a/arch/powerpc/kernel/misc_64.S b/arch/powerpc/kernel/misc_64.S
index f28754c..5974837 100644
--- a/arch/powerpc/kernel/misc_64.S
+++ b/arch/powerpc/kernel/misc_64.S
@@ -667,7 +667,7 @@ _GLOBAL(kexec_sequence)
 	mr	r12,r27
 #endif
 	mtctr	r12
-	bctrl				/* ppc_md.hpte_clear_all(void); */
+	bctrl				/* mmu_hash_ops.hpte_clear_all(void); */
 #endif /* !CONFIG_PPC_BOOK3E */
 
 /*
diff --git a/arch/powerpc/kvm/book3s_64_mmu_host.c b/arch/powerpc/kvm/book3s_64_mmu_host.c
index 114edac..8164792 100644
--- a/arch/powerpc/kvm/book3s_64_mmu_host.c
+++ b/arch/powerpc/kvm/book3s_64_mmu_host.c
@@ -34,9 +34,9 @@
 
 void kvmppc_mmu_invalidate_pte(struct kvm_vcpu *vcpu, struct hpte_cache *pte)
 {
-	ppc_md.hpte_invalidate(pte->slot, pte->host_vpn,
-			       pte->pagesize, pte->pagesize, MMU_SEGSIZE_256M,
-			       false);
+	mmu_hash_ops.hpte_invalidate(pte->slot, pte->host_vpn,
+				     pte->pagesize, pte->pagesize,
+				     MMU_SEGSIZE_256M, false);
 }
 
 /* We keep 512 gvsid->hvsid entries, mapping the guest ones to the array using
@@ -169,13 +169,13 @@ map_again:
 
 	/* In case we tried normal mapping already, let's nuke old entries */
 	if (attempt > 1)
-		if (ppc_md.hpte_remove(hpteg) < 0) {
+		if (mmu_hash_ops.hpte_remove(hpteg) < 0) {
 			r = -1;
 			goto out_unlock;
 		}
 
-	ret = ppc_md.hpte_insert(hpteg, vpn, hpaddr, rflags, vflags,
-				 hpsize, hpsize, MMU_SEGSIZE_256M);
+	ret = mmu_hash_ops.hpte_insert(hpteg, vpn, hpaddr, rflags, vflags,
+				       hpsize, hpsize, MMU_SEGSIZE_256M);
 
 	if (ret < 0) {
 		/* If we couldn't map a primary PTE, try a secondary */
@@ -187,8 +187,10 @@ map_again:
 		trace_kvm_book3s_64_mmu_map(rflags, hpteg,
 					    vpn, hpaddr, orig_pte);
 
-		/* The ppc_md code may give us a secondary entry even though we
-		   asked for a primary. Fix up. */
+		/* The mmu_hash_ops code may give us a secondary entry even
+		 * though we
+		 * asked for a primary. Fix up.
+		 */
 		if ((ret & _PTEIDX_SECONDARY) && !(vflags & HPTE_V_SECONDARY)) {
 			hash = ~hash;
 			hpteg = ((hash & htab_hash_mask) * HPTES_PER_GROUP);
diff --git a/arch/powerpc/mm/hash64_4k.c b/arch/powerpc/mm/hash64_4k.c
index 6333b27..42c702b 100644
--- a/arch/powerpc/mm/hash64_4k.c
+++ b/arch/powerpc/mm/hash64_4k.c
@@ -70,8 +70,8 @@ int __hash_page_4K(unsigned long ea, unsigned long access, unsigned long vsid,
 		slot = (hash & htab_hash_mask) * HPTES_PER_GROUP;
 		slot += (old_pte & H_PAGE_F_GIX) >> H_PAGE_F_GIX_SHIFT;
 
-		if (ppc_md.hpte_updatepp(slot, rflags, vpn, MMU_PAGE_4K,
-					 MMU_PAGE_4K, ssize, flags) == -1)
+		if (mmu_hash_ops.hpte_updatepp(slot, rflags, vpn, MMU_PAGE_4K,
+					       MMU_PAGE_4K, ssize, flags) == -1)
 			old_pte &= ~_PAGE_HPTEFLAGS;
 	}
 
@@ -84,21 +84,23 @@ repeat:
 		hpte_group = ((hash & htab_hash_mask) * HPTES_PER_GROUP) & ~0x7UL;
 
 		/* Insert into the hash table, primary slot */
-		slot = ppc_md.hpte_insert(hpte_group, vpn, pa, rflags, 0,
-				  MMU_PAGE_4K, MMU_PAGE_4K, ssize);
+		slot = mmu_hash_ops.hpte_insert(hpte_group, vpn, pa, rflags, 0,
+						MMU_PAGE_4K, MMU_PAGE_4K, ssize);
 		/*
 		 * Primary is full, try the secondary
 		 */
 		if (unlikely(slot == -1)) {
 			hpte_group = ((~hash & htab_hash_mask) * HPTES_PER_GROUP) & ~0x7UL;
-			slot = ppc_md.hpte_insert(hpte_group, vpn, pa,
-						  rflags, HPTE_V_SECONDARY,
-						  MMU_PAGE_4K, MMU_PAGE_4K, ssize);
+			slot = mmu_hash_ops.hpte_insert(hpte_group, vpn, pa,
+							rflags,
+							HPTE_V_SECONDARY,
+							MMU_PAGE_4K,
+							MMU_PAGE_4K, ssize);
 			if (slot == -1) {
 				if (mftb() & 0x1)
 					hpte_group = ((hash & htab_hash_mask) *
 						      HPTES_PER_GROUP) & ~0x7UL;
-				ppc_md.hpte_remove(hpte_group);
+				mmu_hash_ops.hpte_remove(hpte_group);
 				/*
 				 * FIXME!! Should be try the group from which we removed ?
 				 */
diff --git a/arch/powerpc/mm/hash64_64k.c b/arch/powerpc/mm/hash64_64k.c
index 16644e1..3bbbea0 100644
--- a/arch/powerpc/mm/hash64_64k.c
+++ b/arch/powerpc/mm/hash64_64k.c
@@ -133,9 +133,9 @@ int __hash_page_4K(unsigned long ea, unsigned long access, unsigned long vsid,
 		slot = (hash & htab_hash_mask) * HPTES_PER_GROUP;
 		slot += hidx & _PTEIDX_GROUP_IX;
 
-		ret = ppc_md.hpte_updatepp(slot, rflags, vpn,
-					   MMU_PAGE_4K, MMU_PAGE_4K,
-					   ssize, flags);
+		ret = mmu_hash_ops.hpte_updatepp(slot, rflags, vpn,
+						 MMU_PAGE_4K, MMU_PAGE_4K,
+						 ssize, flags);
 		/*
 		 *if we failed because typically the HPTE wasn't really here
 		 * we try an insertion.
@@ -166,21 +166,22 @@ repeat:
 	hpte_group = ((hash & htab_hash_mask) * HPTES_PER_GROUP) & ~0x7UL;
 
 	/* Insert into the hash table, primary slot */
-	slot = ppc_md.hpte_insert(hpte_group, vpn, pa, rflags, 0,
-				  MMU_PAGE_4K, MMU_PAGE_4K, ssize);
+	slot = mmu_hash_ops.hpte_insert(hpte_group, vpn, pa, rflags, 0,
+					MMU_PAGE_4K, MMU_PAGE_4K, ssize);
 	/*
 	 * Primary is full, try the secondary
 	 */
 	if (unlikely(slot == -1)) {
 		hpte_group = ((~hash & htab_hash_mask) * HPTES_PER_GROUP) & ~0x7UL;
-		slot = ppc_md.hpte_insert(hpte_group, vpn, pa,
-					  rflags, HPTE_V_SECONDARY,
-					  MMU_PAGE_4K, MMU_PAGE_4K, ssize);
+		slot = mmu_hash_ops.hpte_insert(hpte_group, vpn, pa,
+						rflags, HPTE_V_SECONDARY,
+						MMU_PAGE_4K, MMU_PAGE_4K,
+						ssize);
 		if (slot == -1) {
 			if (mftb() & 0x1)
 				hpte_group = ((hash & htab_hash_mask) *
 					      HPTES_PER_GROUP) & ~0x7UL;
-			ppc_md.hpte_remove(hpte_group);
+			mmu_hash_ops.hpte_remove(hpte_group);
 			/*
 			 * FIXME!! Should be try the group from which we removed ?
 			 */
@@ -272,8 +273,9 @@ int __hash_page_64K(unsigned long ea, unsigned long access,
 		slot = (hash & htab_hash_mask) * HPTES_PER_GROUP;
 		slot += (old_pte & H_PAGE_F_GIX) >> H_PAGE_F_GIX_SHIFT;
 
-		if (ppc_md.hpte_updatepp(slot, rflags, vpn, MMU_PAGE_64K,
-					 MMU_PAGE_64K, ssize, flags) == -1)
+		if (mmu_hash_ops.hpte_updatepp(slot, rflags, vpn, MMU_PAGE_64K,
+					       MMU_PAGE_64K, ssize,
+					       flags) == -1)
 			old_pte &= ~_PAGE_HPTEFLAGS;
 	}
 
@@ -286,21 +288,24 @@ repeat:
 		hpte_group = ((hash & htab_hash_mask) * HPTES_PER_GROUP) & ~0x7UL;
 
 		/* Insert into the hash table, primary slot */
-		slot = ppc_md.hpte_insert(hpte_group, vpn, pa, rflags, 0,
-				  MMU_PAGE_64K, MMU_PAGE_64K, ssize);
+		slot = mmu_hash_ops.hpte_insert(hpte_group, vpn, pa, rflags, 0,
+						MMU_PAGE_64K, MMU_PAGE_64K,
+						ssize);
 		/*
 		 * Primary is full, try the secondary
 		 */
 		if (unlikely(slot == -1)) {
 			hpte_group = ((~hash & htab_hash_mask) * HPTES_PER_GROUP) & ~0x7UL;
-			slot = ppc_md.hpte_insert(hpte_group, vpn, pa,
-						  rflags, HPTE_V_SECONDARY,
-						  MMU_PAGE_64K, MMU_PAGE_64K, ssize);
+			slot = mmu_hash_ops.hpte_insert(hpte_group, vpn, pa,
+							rflags,
+							HPTE_V_SECONDARY,
+							MMU_PAGE_64K,
+							MMU_PAGE_64K, ssize);
 			if (slot == -1) {
 				if (mftb() & 0x1)
 					hpte_group = ((hash & htab_hash_mask) *
 						      HPTES_PER_GROUP) & ~0x7UL;
-				ppc_md.hpte_remove(hpte_group);
+				mmu_hash_ops.hpte_remove(hpte_group);
 				/*
 				 * FIXME!! Should be try the group from which we removed ?
 				 */
diff --git a/arch/powerpc/mm/hash_native_64.c b/arch/powerpc/mm/hash_native_64.c
index 40e05e7..09ec60f 100644
--- a/arch/powerpc/mm/hash_native_64.c
+++ b/arch/powerpc/mm/hash_native_64.c
@@ -731,15 +731,15 @@ static int native_update_partition_table(u64 patb1)
 
 void __init hpte_init_native(void)
 {
-	ppc_md.hpte_invalidate	= native_hpte_invalidate;
-	ppc_md.hpte_updatepp	= native_hpte_updatepp;
-	ppc_md.hpte_updateboltedpp = native_hpte_updateboltedpp;
-	ppc_md.hpte_insert	= native_hpte_insert;
-	ppc_md.hpte_remove	= native_hpte_remove;
-	ppc_md.hpte_clear_all	= native_hpte_clear;
-	ppc_md.flush_hash_range = native_flush_hash_range;
-	ppc_md.hugepage_invalidate   = native_hugepage_invalidate;
+	mmu_hash_ops.hpte_invalidate	= native_hpte_invalidate;
+	mmu_hash_ops.hpte_updatepp	= native_hpte_updatepp;
+	mmu_hash_ops.hpte_updateboltedpp = native_hpte_updateboltedpp;
+	mmu_hash_ops.hpte_insert	= native_hpte_insert;
+	mmu_hash_ops.hpte_remove	= native_hpte_remove;
+	mmu_hash_ops.hpte_clear_all	= native_hpte_clear;
+	mmu_hash_ops.flush_hash_range = native_flush_hash_range;
+	mmu_hash_ops.hugepage_invalidate   = native_hugepage_invalidate;
 
 	if (cpu_has_feature(CPU_FTR_ARCH_300))
-		ppc_md.update_partition_table = native_update_partition_table;
+		ppc64_update_partition_table = native_update_partition_table;
 }
diff --git a/arch/powerpc/mm/hash_utils_64.c b/arch/powerpc/mm/hash_utils_64.c
index 164c2da..54b6eb9 100644
--- a/arch/powerpc/mm/hash_utils_64.c
+++ b/arch/powerpc/mm/hash_utils_64.c
@@ -118,6 +118,8 @@ static u8 *linear_map_hash_slots;
 static unsigned long linear_map_hash_count;
 static DEFINE_SPINLOCK(linear_map_hash_lock);
 #endif /* CONFIG_DEBUG_PAGEALLOC */
+struct mmu_hash_ops mmu_hash_ops;
+EXPORT_SYMBOL(mmu_hash_ops);
 
 /* There are definitions of page sizes arrays to be used when none
  * is provided by the firmware.
@@ -272,9 +274,10 @@ int htab_bolt_mapping(unsigned long vstart, unsigned long vend,
 		hash = hpt_hash(vpn, shift, ssize);
 		hpteg = ((hash & htab_hash_mask) * HPTES_PER_GROUP);
 
-		BUG_ON(!ppc_md.hpte_insert);
-		ret = ppc_md.hpte_insert(hpteg, vpn, paddr, tprot,
-					 HPTE_V_BOLTED, psize, psize, ssize);
+		BUG_ON(!mmu_hash_ops.hpte_insert);
+		ret = mmu_hash_ops.hpte_insert(hpteg, vpn, paddr, tprot,
+					       HPTE_V_BOLTED, psize, psize,
+					       ssize);
 
 		if (ret < 0)
 			break;
@@ -299,11 +302,11 @@ int htab_remove_mapping(unsigned long vstart, unsigned long vend,
 	shift = mmu_psize_defs[psize].shift;
 	step = 1 << shift;
 
-	if (!ppc_md.hpte_removebolted)
+	if (!mmu_hash_ops.hpte_removebolted)
 		return -ENODEV;
 
 	for (vaddr = vstart; vaddr < vend; vaddr += step) {
-		rc = ppc_md.hpte_removebolted(vaddr, psize, ssize);
+		rc = mmu_hash_ops.hpte_removebolted(vaddr, psize, ssize);
 		if (rc == -ENOENT) {
 			ret = -ENOENT;
 			continue;
@@ -771,8 +774,8 @@ static void __init htab_initialize(void)
 		 * Clear the htab if firmware assisted dump is active so
 		 * that we dont end up using old mappings.
 		 */
-		if (is_fadump_active() && ppc_md.hpte_clear_all)
-			ppc_md.hpte_clear_all();
+		if (is_fadump_active() && mmu_hash_ops.hpte_clear_all)
+			mmu_hash_ops.hpte_clear_all();
 #endif
 	} else {
 		unsigned long limit = MEMBLOCK_ALLOC_ANYWHERE;
@@ -1451,7 +1454,8 @@ void flush_hash_page(unsigned long vpn, real_pte_t pte, int psize, int ssize,
 		 * We use same base page size and actual psize, because we don't
 		 * use these functions for hugepage
 		 */
-		ppc_md.hpte_invalidate(slot, vpn, psize, psize, ssize, local);
+		mmu_hash_ops.hpte_invalidate(slot, vpn, psize, psize,
+					     ssize, local);
 	} pte_iterate_hashed_end();
 
 #ifdef CONFIG_PPC_TRANSACTIONAL_MEM
@@ -1492,9 +1496,9 @@ void flush_hash_hugepage(unsigned long vsid, unsigned long addr,
 	if (!hpte_slot_array)
 		return;
 
-	if (ppc_md.hugepage_invalidate) {
-		ppc_md.hugepage_invalidate(vsid, s_addr, hpte_slot_array,
-					   psize, ssize, local);
+	if (mmu_hash_ops.hugepage_invalidate) {
+		mmu_hash_ops.hugepage_invalidate(vsid, s_addr, hpte_slot_array,
+						 psize, ssize, local);
 		goto tm_abort;
 	}
 	/*
@@ -1521,8 +1525,8 @@ void flush_hash_hugepage(unsigned long vsid, unsigned long addr,
 
 		slot = (hash & htab_hash_mask) * HPTES_PER_GROUP;
 		slot += hidx & _PTEIDX_GROUP_IX;
-		ppc_md.hpte_invalidate(slot, vpn, psize,
-				       MMU_PAGE_16M, ssize, local);
+		mmu_hash_ops.hpte_invalidate(slot, vpn, psize,
+					     MMU_PAGE_16M, ssize, local);
 	}
 tm_abort:
 #ifdef CONFIG_PPC_TRANSACTIONAL_MEM
@@ -1546,8 +1550,8 @@ tm_abort:
 
 void flush_hash_range(unsigned long number, int local)
 {
-	if (ppc_md.flush_hash_range)
-		ppc_md.flush_hash_range(number, local);
+	if (mmu_hash_ops.flush_hash_range)
+		mmu_hash_ops.flush_hash_range(number, local);
 	else {
 		int i;
 		struct ppc64_tlb_batch *batch =
@@ -1592,22 +1596,22 @@ repeat:
 		       HPTES_PER_GROUP) & ~0x7UL;
 
 	/* Insert into the hash table, primary slot */
-	slot = ppc_md.hpte_insert(hpte_group, vpn, pa, rflags, vflags,
-				  psize, psize, ssize);
+	slot = mmu_hash_ops.hpte_insert(hpte_group, vpn, pa, rflags, vflags,
+					psize, psize, ssize);
 
 	/* Primary is full, try the secondary */
 	if (unlikely(slot == -1)) {
 		hpte_group = ((~hash & htab_hash_mask) *
 			      HPTES_PER_GROUP) & ~0x7UL;
-		slot = ppc_md.hpte_insert(hpte_group, vpn, pa, rflags,
-					  vflags | HPTE_V_SECONDARY,
-					  psize, psize, ssize);
+		slot = mmu_hash_ops.hpte_insert(hpte_group, vpn, pa, rflags,
+						vflags | HPTE_V_SECONDARY,
+						psize, psize, ssize);
 		if (slot == -1) {
 			if (mftb() & 0x1)
 				hpte_group = ((hash & htab_hash_mask) *
 					      HPTES_PER_GROUP)&~0x7UL;
 
-			ppc_md.hpte_remove(hpte_group);
+			mmu_hash_ops.hpte_remove(hpte_group);
 			goto repeat;
 		}
 	}
@@ -1657,8 +1661,9 @@ static void kernel_unmap_linear_page(unsigned long vaddr, unsigned long lmi)
 		hash = ~hash;
 	slot = (hash & htab_hash_mask) * HPTES_PER_GROUP;
 	slot += hidx & _PTEIDX_GROUP_IX;
-	ppc_md.hpte_invalidate(slot, vpn, mmu_linear_psize, mmu_linear_psize,
-			       mmu_kernel_ssize, 0);
+	mmu_hash_ops.hpte_invalidate(slot, vpn, mmu_linear_psize,
+				     mmu_linear_psize,
+				     mmu_kernel_ssize, 0);
 }
 
 void __kernel_map_pages(struct page *page, int numpages, int enable)
diff --git a/arch/powerpc/mm/hugepage-hash64.c b/arch/powerpc/mm/hugepage-hash64.c
index ba3fc22..f20d16f 100644
--- a/arch/powerpc/mm/hugepage-hash64.c
+++ b/arch/powerpc/mm/hugepage-hash64.c
@@ -103,8 +103,8 @@ int __hash_page_thp(unsigned long ea, unsigned long access, unsigned long vsid,
 		slot = (hash & htab_hash_mask) * HPTES_PER_GROUP;
 		slot += hidx & _PTEIDX_GROUP_IX;
 
-		ret = ppc_md.hpte_updatepp(slot, rflags, vpn,
-					   psize, lpsize, ssize, flags);
+		ret = mmu_hash_ops.hpte_updatepp(slot, rflags, vpn,
+						 psize, lpsize, ssize, flags);
 		/*
 		 * We failed to update, try to insert a new entry.
 		 */
@@ -131,23 +131,24 @@ repeat:
 		hpte_group = ((hash & htab_hash_mask) * HPTES_PER_GROUP) & ~0x7UL;
 
 		/* Insert into the hash table, primary slot */
-		slot = ppc_md.hpte_insert(hpte_group, vpn, pa, rflags, 0,
-					  psize, lpsize, ssize);
+		slot = mmu_hash_ops.hpte_insert(hpte_group, vpn, pa, rflags, 0,
+						psize, lpsize, ssize);
 		/*
 		 * Primary is full, try the secondary
 		 */
 		if (unlikely(slot == -1)) {
 			hpte_group = ((~hash & htab_hash_mask) *
 				      HPTES_PER_GROUP) & ~0x7UL;
-			slot = ppc_md.hpte_insert(hpte_group, vpn, pa,
-						  rflags, HPTE_V_SECONDARY,
-						  psize, lpsize, ssize);
+			slot = mmu_hash_ops.hpte_insert(hpte_group, vpn, pa,
+							rflags,
+							HPTE_V_SECONDARY,
+							psize, lpsize, ssize);
 			if (slot == -1) {
 				if (mftb() & 0x1)
 					hpte_group = ((hash & htab_hash_mask) *
 						      HPTES_PER_GROUP) & ~0x7UL;
 
-				ppc_md.hpte_remove(hpte_group);
+				mmu_hash_ops.hpte_remove(hpte_group);
 				goto repeat;
 			}
 		}
diff --git a/arch/powerpc/mm/hugetlbpage-hash64.c b/arch/powerpc/mm/hugetlbpage-hash64.c
index 3058560..d5026f3 100644
--- a/arch/powerpc/mm/hugetlbpage-hash64.c
+++ b/arch/powerpc/mm/hugetlbpage-hash64.c
@@ -79,8 +79,8 @@ int __hash_page_huge(unsigned long ea, unsigned long access, unsigned long vsid,
 		slot = (hash & htab_hash_mask) * HPTES_PER_GROUP;
 		slot += (old_pte & H_PAGE_F_GIX) >> H_PAGE_F_GIX_SHIFT;
 
-		if (ppc_md.hpte_updatepp(slot, rflags, vpn, mmu_psize,
-					 mmu_psize, ssize, flags) == -1)
+		if (mmu_hash_ops.hpte_updatepp(slot, rflags, vpn, mmu_psize,
+					       mmu_psize, ssize, flags) == -1)
 			old_pte &= ~_PAGE_HPTEFLAGS;
 	}
 
diff --git a/arch/powerpc/mm/pgtable-book3s64.c b/arch/powerpc/mm/pgtable-book3s64.c
index 6703187..85c6070 100644
--- a/arch/powerpc/mm/pgtable-book3s64.c
+++ b/arch/powerpc/mm/pgtable-book3s64.c
@@ -14,6 +14,8 @@
 #include "mmu_decl.h"
 #include <trace/events/thp.h>
 
+int (*ppc64_update_partition_table)(u64);
+
 #ifdef CONFIG_TRANSPARENT_HUGEPAGE
 /*
  * This is called when relaxing access to a hugepage. It's also called in the page
diff --git a/arch/powerpc/mm/pgtable-radix.c b/arch/powerpc/mm/pgtable-radix.c
index 6624b7b..3ababda 100644
--- a/arch/powerpc/mm/pgtable-radix.c
+++ b/arch/powerpc/mm/pgtable-radix.c
@@ -114,6 +114,9 @@ static void __init radix_init_pgtable(void)
 
 	/* We don't support slb for radix */
 	mmu_slb_size = 0;
+
+	BUG_ON(!ppc64_update_partition_table);
+
 	/*
 	 * Create the linear mapping, using standard page size for now
 	 */
@@ -169,7 +172,7 @@ redo:
 	 * of process table here. But our linear mapping also enable us to use
 	 * physical address here.
 	 */
-	ppc_md.update_partition_table(__pa(process_tb) | (PRTB_SIZE_SHIFT - 12) | PATB_GR);
+	ppc64_update_partition_table(__pa(process_tb) | (PRTB_SIZE_SHIFT - 12) | PATB_GR);
 	pr_info("Process table %p and radix root for kernel: %p\n", process_tb, init_mm.pgd);
 }
 
@@ -197,7 +200,7 @@ static void __init radix_init_partition_table(void)
 
 void __init radix_init_native(void)
 {
-	ppc_md.update_partition_table = native_update_partition_table;
+	ppc64_update_partition_table = native_update_partition_table;
 }
 
 static int __init get_idx_from_shift(unsigned int shift)
diff --git a/arch/powerpc/platforms/ps3/htab.c b/arch/powerpc/platforms/ps3/htab.c
index c9a3e67..cb3c503 100644
--- a/arch/powerpc/platforms/ps3/htab.c
+++ b/arch/powerpc/platforms/ps3/htab.c
@@ -195,12 +195,12 @@ static void ps3_hpte_clear(void)
 
 void __init ps3_hpte_init(unsigned long htab_size)
 {
-	ppc_md.hpte_invalidate = ps3_hpte_invalidate;
-	ppc_md.hpte_updatepp = ps3_hpte_updatepp;
-	ppc_md.hpte_updateboltedpp = ps3_hpte_updateboltedpp;
-	ppc_md.hpte_insert = ps3_hpte_insert;
-	ppc_md.hpte_remove = ps3_hpte_remove;
-	ppc_md.hpte_clear_all = ps3_hpte_clear;
+	mmu_hash_ops.hpte_invalidate = ps3_hpte_invalidate;
+	mmu_hash_ops.hpte_updatepp = ps3_hpte_updatepp;
+	mmu_hash_ops.hpte_updateboltedpp = ps3_hpte_updateboltedpp;
+	mmu_hash_ops.hpte_insert = ps3_hpte_insert;
+	mmu_hash_ops.hpte_remove = ps3_hpte_remove;
+	mmu_hash_ops.hpte_clear_all = ps3_hpte_clear;
 
 	ppc64_pft_size = __ilog2(htab_size);
 }
diff --git a/arch/powerpc/platforms/pseries/lpar.c b/arch/powerpc/platforms/pseries/lpar.c
index 7f6100d..f37bba9 100644
--- a/arch/powerpc/platforms/pseries/lpar.c
+++ b/arch/powerpc/platforms/pseries/lpar.c
@@ -606,15 +606,15 @@ __setup("bulk_remove=", disable_bulk_remove);
 
 void __init hpte_init_lpar(void)
 {
-	ppc_md.hpte_invalidate	= pSeries_lpar_hpte_invalidate;
-	ppc_md.hpte_updatepp	= pSeries_lpar_hpte_updatepp;
-	ppc_md.hpte_updateboltedpp = pSeries_lpar_hpte_updateboltedpp;
-	ppc_md.hpte_insert	= pSeries_lpar_hpte_insert;
-	ppc_md.hpte_remove	= pSeries_lpar_hpte_remove;
-	ppc_md.hpte_removebolted = pSeries_lpar_hpte_removebolted;
-	ppc_md.flush_hash_range	= pSeries_lpar_flush_hash_range;
-	ppc_md.hpte_clear_all   = pSeries_lpar_hptab_clear;
-	ppc_md.hugepage_invalidate = pSeries_lpar_hugepage_invalidate;
+	mmu_hash_ops.hpte_invalidate	 = pSeries_lpar_hpte_invalidate;
+	mmu_hash_ops.hpte_updatepp	 = pSeries_lpar_hpte_updatepp;
+	mmu_hash_ops.hpte_updateboltedpp = pSeries_lpar_hpte_updateboltedpp;
+	mmu_hash_ops.hpte_insert	 = pSeries_lpar_hpte_insert;
+	mmu_hash_ops.hpte_remove	 = pSeries_lpar_hpte_remove;
+	mmu_hash_ops.hpte_removebolted   = pSeries_lpar_hpte_removebolted;
+	mmu_hash_ops.flush_hash_range	 = pSeries_lpar_flush_hash_range;
+	mmu_hash_ops.hpte_clear_all      = pSeries_lpar_hptab_clear;
+	mmu_hash_ops.hugepage_invalidate = pSeries_lpar_hugepage_invalidate;
 }
 
 #ifdef CONFIG_PPC_SMLPAR
-- 
2.7.4

[PATCH 23/38] powerpc: Ensure that ppc_md is empty before probing for machine type

[Benjamin Herrenschmidt]

Anything in there will be overwritten, so it helps catching nasty
bugs if we check that it's indeed full of NULL's before we do so.

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
---
 arch/powerpc/kernel/setup-common.c | 11 +++++++++++
 1 file changed, 11 insertions(+)

diff --git a/arch/powerpc/kernel/setup-common.c b/arch/powerpc/kernel/setup-common.c
index 2a3564c..7e85fac 100644
--- a/arch/powerpc/kernel/setup-common.c
+++ b/arch/powerpc/kernel/setup-common.c
@@ -575,6 +575,7 @@ void probe_machine(void)
 {
 	extern struct machdep_calls __machine_desc_start;
 	extern struct machdep_calls __machine_desc_end;
+	unsigned int i;
 
 	/*
 	 * Iterate all ppc_md structures until we find the proper
@@ -582,6 +583,16 @@ void probe_machine(void)
 	 */
 	DBG("Probing machine type ...\n");
 
+	/* Check ppc_md is empty, if not we have a bug, ie, we setup an
+	 * entry before probe_machine() which will be overwritten
+	 */
+	for (i = 0; i < (sizeof(ppc_md) / sizeof(void *)); i++) {
+		if (((void **)&ppc_md)[i]) {
+			printk(KERN_ERR "Entry %d in ppc_md non empty before"
+			       " machine probe !\n", i);
+		}
+	}
+
 	for (machine_id = &__machine_desc_start;
 	     machine_id < &__machine_desc_end;
 	     machine_id++) {
-- 
2.7.4

[PATCH 24/38] powerpc: Move 64-bit probe_machine() to later in the boot process

[Benjamin Herrenschmidt]

We no long need the machine type that early, so we can move probe_machine()
to after the device-tree has been expanded. This will allow further
consolidation.

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
---
 arch/powerpc/kernel/setup_64.c          | 6 +++---
 arch/powerpc/platforms/cell/setup.c     | 6 ++----
 arch/powerpc/platforms/maple/setup.c    | 6 ++----
 arch/powerpc/platforms/pasemi/setup.c   | 6 ++----
 arch/powerpc/platforms/powermac/setup.c | 6 ++++++
 arch/powerpc/platforms/powernv/setup.c  | 4 +---
 arch/powerpc/platforms/ps3/setup.c      | 6 ++----
 arch/powerpc/platforms/pseries/setup.c  | 7 +++----
 8 files changed, 21 insertions(+), 26 deletions(-)

diff --git a/arch/powerpc/kernel/setup_64.c b/arch/powerpc/kernel/setup_64.c
index 47a2706..1f20f99 100644
--- a/arch/powerpc/kernel/setup_64.c
+++ b/arch/powerpc/kernel/setup_64.c
@@ -313,9 +313,6 @@ void __init early_setup(unsigned long dt_ptr)
 	setup_paca(&paca[boot_cpuid]);
 	fixup_boot_paca();
 
-	/* Probe the machine type */
-	probe_machine();
-
 	/* Configure exception handlers. This include setting up trampolines
 	 * if needed, setting exception endian mode, etc...
 	 */
@@ -522,6 +519,9 @@ void __init setup_system(void)
 	 */
 	check_for_initrd();
 
+	/* Probe the machine type */
+	probe_machine();
+
 	/*
 	 * Do some platform specific early initializations, that includes
 	 * setting up the hash table pointers. It also sets up some interrupt-mapping
diff --git a/arch/powerpc/platforms/cell/setup.c b/arch/powerpc/platforms/cell/setup.c
index e342f78..d3543e6 100644
--- a/arch/powerpc/platforms/cell/setup.c
+++ b/arch/powerpc/platforms/cell/setup.c
@@ -255,10 +255,8 @@ static void __init cell_setup_arch(void)
 
 static int __init cell_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	if (!of_flat_dt_is_compatible(root, "IBM,CBEA") &&
-	    !of_flat_dt_is_compatible(root, "IBM,CPBW-1.0"))
+	if (!of_machine_is_compatible("IBM,CBEA") &&
+	    !of_machine_is_compatible("IBM,CPBW-1.0"))
 		return 0;
 
 	pm_power_off = rtas_power_off;
diff --git a/arch/powerpc/platforms/maple/setup.c b/arch/powerpc/platforms/maple/setup.c
index a56828c..a1ecbc9 100644
--- a/arch/powerpc/platforms/maple/setup.c
+++ b/arch/powerpc/platforms/maple/setup.c
@@ -298,10 +298,8 @@ static void __init maple_progress(char *s, unsigned short hex)
  */
 static int __init maple_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	if (!of_flat_dt_is_compatible(root, "Momentum,Maple") &&
-	    !of_flat_dt_is_compatible(root, "Momentum,Apache"))
+	if (!of_machine_is_compatible("Momentum,Maple") &&
+	    !of_machine_is_compatible("Momentum,Apache"))
 		return 0;
 
 	pm_power_off = maple_power_off;
diff --git a/arch/powerpc/platforms/pasemi/setup.c b/arch/powerpc/platforms/pasemi/setup.c
index 9a881be..8f5e291 100644
--- a/arch/powerpc/platforms/pasemi/setup.c
+++ b/arch/powerpc/platforms/pasemi/setup.c
@@ -420,10 +420,8 @@ machine_device_initcall(pasemi, pasemi_publish_devices);
  */
 static int __init pas_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	if (!of_flat_dt_is_compatible(root, "PA6T-1682M") &&
-	    !of_flat_dt_is_compatible(root, "pasemi,pwrficient"))
+	if (!of_machine_is_compatible("PA6T-1682M") &&
+	    !of_machine_is_compatible("pasemi,pwrficient"))
 		return 0;
 
 	return 1;
diff --git a/arch/powerpc/platforms/powermac/setup.c b/arch/powerpc/platforms/powermac/setup.c
index 1e42104..128ce76 100644
--- a/arch/powerpc/platforms/powermac/setup.c
+++ b/arch/powerpc/platforms/powermac/setup.c
@@ -588,11 +588,17 @@ console_initcall(check_pmac_serial_console);
  */
 static int __init pmac_probe(void)
 {
+#ifdef CONFIG_PPC32
 	unsigned long root = of_get_flat_dt_root();
 
 	if (!of_flat_dt_is_compatible(root, "Power Macintosh") &&
 	    !of_flat_dt_is_compatible(root, "MacRISC"))
 		return 0;
+#else
+	if (!of_machine_is_compatible("Power Macintosh") &&
+	    !of_machine_is_compatible("MacRISC"))
+		return 0;
+#endif
 
 #ifdef CONFIG_PPC32
 	/* isa_io_base gets set in pmac_pci_init */
diff --git a/arch/powerpc/platforms/powernv/setup.c b/arch/powerpc/platforms/powernv/setup.c
index 15e3b8b..ed99177 100644
--- a/arch/powerpc/platforms/powernv/setup.c
+++ b/arch/powerpc/platforms/powernv/setup.c
@@ -268,9 +268,7 @@ static void __init pnv_setup_machdep_opal(void)
 
 static int __init pnv_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	if (!of_flat_dt_is_compatible(root, "ibm,powernv"))
+	if (!of_machine_is_compatible("ibm,powernv"))
 		return 0;
 
 	if (firmware_has_feature(FW_FEATURE_OPAL))
diff --git a/arch/powerpc/platforms/ps3/setup.c b/arch/powerpc/platforms/ps3/setup.c
index e5ee80d..af18deb 100644
--- a/arch/powerpc/platforms/ps3/setup.c
+++ b/arch/powerpc/platforms/ps3/setup.c
@@ -237,15 +237,13 @@ void __init ps3_early_mm_init(void)
 
 static int __init ps3_probe(void)
 {
-	unsigned long dt_root;
-
 	DBG(" -> %s:%d\n", __func__, __LINE__);
 
-	dt_root = of_get_flat_dt_root();
-	if (!of_flat_dt_is_compatible(dt_root, "sony,ps3"))
+	if (!of_machine_is_compatible("sony,ps3"))
 		return 0;
 
 	ps3_os_area_save_params();
+
 	pm_power_off = ps3_power_off;
 
 	DBG(" <- %s:%d\n", __func__, __LINE__);
diff --git a/arch/powerpc/platforms/pseries/setup.c b/arch/powerpc/platforms/pseries/setup.c
index a8e33424..eacfba7 100644
--- a/arch/powerpc/platforms/pseries/setup.c
+++ b/arch/powerpc/platforms/pseries/setup.c
@@ -722,8 +722,7 @@ static void pseries_power_off(void)
 
 static int __init pSeries_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-	const char *dtype = of_get_flat_dt_prop(root, "device_type", NULL);
+	const char *dtype = of_get_property(of_root, "device_type", NULL);
 
  	if (dtype == NULL)
  		return 0;
@@ -733,8 +732,8 @@ static int __init pSeries_probe(void)
 	/* Cell blades firmware claims to be chrp while it's not. Until this
 	 * is fixed, we need to avoid those here.
 	 */
-	if (of_flat_dt_is_compatible(root, "IBM,CPBW-1.0") ||
-	    of_flat_dt_is_compatible(root, "IBM,CBEA"))
+	if (of_machine_is_compatible("IBM,CPBW-1.0") ||
+	    of_machine_is_compatible("IBM,CBEA"))
 		return 0;
 
 	pm_power_off = pseries_power_off;
-- 
2.7.4

[PATCH 25/38] powerpc/cell: Don't use flat device-tree after boot

[Benjamin Herrenschmidt]

Some bit of SPU code was using the FDT rather than the expanded
device-tree. Fix it.

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
---
 arch/powerpc/platforms/cell/spu_manage.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/arch/powerpc/platforms/cell/spu_manage.c b/arch/powerpc/platforms/cell/spu_manage.c
index c3327f3..21b4bfb 100644
--- a/arch/powerpc/platforms/cell/spu_manage.c
+++ b/arch/powerpc/platforms/cell/spu_manage.c
@@ -535,8 +535,7 @@ static int __init init_affinity(void)
 	if (of_has_vicinity()) {
 		init_affinity_fw();
 	} else {
-		long root = of_get_flat_dt_root();
-		if (of_flat_dt_is_compatible(root, "IBM,CPBW-1.0"))
+		if (of_machine_is_compatible("IBM,CPBW-1.0"))
 			init_affinity_qs20_harcoded();
 		else
 			printk("No affinity configuration found\n");
-- 
2.7.4

[PATCH 26/38] powerpc/85xx/ge_imp3a: Don't use the flat device-tree after boot

[Benjamin Herrenschmidt]

ge_imp3a_pic_init() is called way beyond the unflattening of
the tree, it shouldn't be using of_flat_dt_*

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
---
 arch/powerpc/platforms/85xx/ge_imp3a.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/arch/powerpc/platforms/85xx/ge_imp3a.c b/arch/powerpc/platforms/85xx/ge_imp3a.c
index 11790e0..55eefef 100644
--- a/arch/powerpc/platforms/85xx/ge_imp3a.c
+++ b/arch/powerpc/platforms/85xx/ge_imp3a.c
@@ -47,9 +47,8 @@ void __init ge_imp3a_pic_init(void)
 	struct mpic *mpic;
 	struct device_node *np;
 	struct device_node *cascade_node = NULL;
-	unsigned long root = of_get_flat_dt_root();
 
-	if (of_flat_dt_is_compatible(root, "fsl,MPC8572DS-CAMP")) {
+	if (of_machine_is_compatible("fsl,MPC8572DS-CAMP")) {
 		mpic = mpic_alloc(NULL, 0,
 			MPIC_NO_RESET |
 			MPIC_BIG_ENDIAN |
-- 
2.7.4

[PATCH 27/38] powerpc/85xx/mpc85xx_ds: Don't use the flat device-tree after boot

[Benjamin Herrenschmidt]

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
---
 arch/powerpc/platforms/85xx/mpc85xx_ds.c | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

diff --git a/arch/powerpc/platforms/85xx/mpc85xx_ds.c b/arch/powerpc/platforms/85xx/mpc85xx_ds.c
index f858306..64a7e8c 100644
--- a/arch/powerpc/platforms/85xx/mpc85xx_ds.c
+++ b/arch/powerpc/platforms/85xx/mpc85xx_ds.c
@@ -66,9 +66,7 @@ void __init mpc85xx_ds_pic_init(void)
 	struct device_node *cascade_node = NULL;
 	int cascade_irq;
 #endif
-	unsigned long root = of_get_flat_dt_root();
-
-	if (of_flat_dt_is_compatible(root, "fsl,MPC8572DS-CAMP")) {
+	if (of_machine_is_compatible("fsl,MPC8572DS-CAMP")) {
 		mpic = mpic_alloc(NULL, 0,
 			MPIC_NO_RESET |
 			MPIC_BIG_ENDIAN |
-- 
2.7.4

[PATCH 28/38] powerpc/85xx/mpc85xx_rdb: Don't use the flat device-tree after boot

[Benjamin Herrenschmidt]

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
---
 arch/powerpc/platforms/85xx/mpc85xx_rdb.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/arch/powerpc/platforms/85xx/mpc85xx_rdb.c b/arch/powerpc/platforms/85xx/mpc85xx_rdb.c
index 3f4dad1..761e504 100644
--- a/arch/powerpc/platforms/85xx/mpc85xx_rdb.c
+++ b/arch/powerpc/platforms/85xx/mpc85xx_rdb.c
@@ -47,13 +47,12 @@
 void __init mpc85xx_rdb_pic_init(void)
 {
 	struct mpic *mpic;
-	unsigned long root = of_get_flat_dt_root();
 
 #ifdef CONFIG_QUICC_ENGINE
 	struct device_node *np;
 #endif
 
-	if (of_flat_dt_is_compatible(root, "fsl,MPC85XXRDB-CAMP")) {
+	if (of_machine_is_compatible("fsl,MPC85XXRDB-CAMP")) {
 		mpic = mpic_alloc(NULL, 0, MPIC_NO_RESET |
 			MPIC_BIG_ENDIAN |
 			MPIC_SINGLE_DEST_CPU,
-- 
2.7.4

[PATCH 29/38] powerpc: Move 32-bit probe() machine to later in the boot process

[Benjamin Herrenschmidt]

This converts all the 32-bit platforms to use the expanded device-tree
which is a pretty mechanical change. Unlike 64-bit, the 32-bit kernel
didn't rely on platform initializations to setup the MMU since it
sets it up entirely before probe_machine() so the move has comparatively
less consequences though it's a bigger patch.

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
---
 arch/powerpc/kernel/setup_32.c                    | 35 +++++++++++---------
 arch/powerpc/platforms/40x/ep405.c                |  4 +--
 arch/powerpc/platforms/40x/ppc40x_simple.c        |  2 +-
 arch/powerpc/platforms/40x/virtex.c               |  4 +--
 arch/powerpc/platforms/40x/walnut.c               |  4 +--
 arch/powerpc/platforms/44x/canyonlands.c          |  5 ++-
 arch/powerpc/platforms/44x/ebony.c                |  4 +--
 arch/powerpc/platforms/44x/iss4xx.c               |  4 +--
 arch/powerpc/platforms/44x/ppc44x_simple.c        |  3 +-
 arch/powerpc/platforms/44x/ppc476.c               |  6 ++--
 arch/powerpc/platforms/44x/sam440ep.c             |  4 +--
 arch/powerpc/platforms/44x/virtex.c               |  4 +--
 arch/powerpc/platforms/44x/warp.c                 |  4 +--
 arch/powerpc/platforms/512x/mpc5121_ads.c         |  4 +--
 arch/powerpc/platforms/512x/mpc512x_generic.c     |  2 +-
 arch/powerpc/platforms/512x/pdm360ng.c            |  4 +--
 arch/powerpc/platforms/52xx/efika.c               |  3 +-
 arch/powerpc/platforms/52xx/lite5200.c            |  2 +-
 arch/powerpc/platforms/52xx/media5200.c           |  2 +-
 arch/powerpc/platforms/52xx/mpc5200_simple.c      |  2 +-
 arch/powerpc/platforms/82xx/ep8248e.c             |  3 +-
 arch/powerpc/platforms/82xx/km82xx.c              |  3 +-
 arch/powerpc/platforms/82xx/mpc8272_ads.c         |  3 +-
 arch/powerpc/platforms/82xx/pq2fads.c             |  3 +-
 arch/powerpc/platforms/83xx/asp834x.c             |  3 +-
 arch/powerpc/platforms/83xx/km83xx.c              |  3 +-
 arch/powerpc/platforms/83xx/mpc830x_rdb.c         |  2 +-
 arch/powerpc/platforms/83xx/mpc831x_rdb.c         |  2 +-
 arch/powerpc/platforms/83xx/mpc832x_mds.c         |  4 +--
 arch/powerpc/platforms/83xx/mpc832x_rdb.c         |  4 +--
 arch/powerpc/platforms/83xx/mpc834x_itx.c         |  4 +--
 arch/powerpc/platforms/83xx/mpc834x_mds.c         |  4 +--
 arch/powerpc/platforms/83xx/mpc836x_mds.c         |  4 +--
 arch/powerpc/platforms/83xx/mpc836x_rdk.c         |  4 +--
 arch/powerpc/platforms/83xx/mpc837x_mds.c         |  4 +--
 arch/powerpc/platforms/83xx/mpc837x_rdb.c         |  2 +-
 arch/powerpc/platforms/83xx/sbc834x.c             |  4 +--
 arch/powerpc/platforms/85xx/bsc913x_qds.c         |  4 +--
 arch/powerpc/platforms/85xx/bsc913x_rdb.c         |  4 +--
 arch/powerpc/platforms/85xx/c293pcie.c            |  4 +--
 arch/powerpc/platforms/85xx/corenet_generic.c     |  5 ++-
 arch/powerpc/platforms/85xx/ge_imp3a.c            |  4 +--
 arch/powerpc/platforms/85xx/ksi8560.c             |  4 +--
 arch/powerpc/platforms/85xx/mpc8536_ds.c          |  4 +--
 arch/powerpc/platforms/85xx/mpc85xx_ads.c         |  4 +--
 arch/powerpc/platforms/85xx/mpc85xx_cds.c         |  4 +--
 arch/powerpc/platforms/85xx/mpc85xx_ds.c          | 12 ++-----
 arch/powerpc/platforms/85xx/mpc85xx_mds.c         | 12 ++-----
 arch/powerpc/platforms/85xx/mpc85xx_rdb.c         | 40 ++++++-----------------
 arch/powerpc/platforms/85xx/mvme2500.c            |  4 +--
 arch/powerpc/platforms/85xx/p1010rdb.c            |  6 ++--
 arch/powerpc/platforms/85xx/p1022_ds.c            |  4 +--
 arch/powerpc/platforms/85xx/p1022_rdk.c           |  4 +--
 arch/powerpc/platforms/85xx/p1023_rdb.c           |  4 +--
 arch/powerpc/platforms/85xx/ppa8548.c             |  4 +--
 arch/powerpc/platforms/85xx/qemu_e500.c           |  4 +--
 arch/powerpc/platforms/85xx/sbc8548.c             |  4 +--
 arch/powerpc/platforms/85xx/socrates.c            |  4 +--
 arch/powerpc/platforms/85xx/stx_gp3.c             |  4 +--
 arch/powerpc/platforms/85xx/tqm85xx.c             |  2 +-
 arch/powerpc/platforms/85xx/twr_p102x.c           |  4 +--
 arch/powerpc/platforms/85xx/xes_mpc85xx.c         | 12 ++-----
 arch/powerpc/platforms/86xx/gef_ppc9a.c           |  4 +--
 arch/powerpc/platforms/86xx/gef_sbc310.c          |  4 +--
 arch/powerpc/platforms/86xx/gef_sbc610.c          |  4 +--
 arch/powerpc/platforms/86xx/mpc8610_hpcd.c        |  4 +--
 arch/powerpc/platforms/86xx/mpc86xx_hpcn.c        |  6 ++--
 arch/powerpc/platforms/86xx/sbc8641d.c            |  4 +--
 arch/powerpc/platforms/8xx/adder875.c             |  3 +-
 arch/powerpc/platforms/8xx/ep88xc.c               |  3 +-
 arch/powerpc/platforms/8xx/mpc86xads_setup.c      |  3 +-
 arch/powerpc/platforms/8xx/mpc885ads_setup.c      |  3 +-
 arch/powerpc/platforms/8xx/tqm8xx_setup.c         |  4 +--
 arch/powerpc/platforms/amigaone/setup.c           |  4 +--
 arch/powerpc/platforms/embedded6xx/c2k.c          |  4 +--
 arch/powerpc/platforms/embedded6xx/gamecube.c     |  5 +--
 arch/powerpc/platforms/embedded6xx/holly.c        |  4 +--
 arch/powerpc/platforms/embedded6xx/linkstation.c  |  6 +---
 arch/powerpc/platforms/embedded6xx/mpc7448_hpc2.c |  4 +--
 arch/powerpc/platforms/embedded6xx/mvme5100.c     |  4 +--
 arch/powerpc/platforms/embedded6xx/storcenter.c   |  4 +--
 arch/powerpc/platforms/embedded6xx/wii.c          |  5 +--
 arch/powerpc/platforms/powermac/setup.c           |  8 -----
 83 files changed, 120 insertions(+), 289 deletions(-)

diff --git a/arch/powerpc/kernel/setup_32.c b/arch/powerpc/kernel/setup_32.c
index 3f0aca2..e7bb4e7 100644
--- a/arch/powerpc/kernel/setup_32.c
+++ b/arch/powerpc/kernel/setup_32.c
@@ -113,23 +113,7 @@ notrace void __init machine_init(u64 dt_ptr)
 
 	early_init_mmu();
 
-	probe_machine();
-
 	setup_kdump_trampoline();
-
-#ifdef CONFIG_6xx
-	if (cpu_has_feature(CPU_FTR_CAN_DOZE) ||
-	    cpu_has_feature(CPU_FTR_CAN_NAP))
-		ppc_md.power_save = ppc6xx_idle;
-#endif
-
-#ifdef CONFIG_E500
-	if (cpu_has_feature(CPU_FTR_CAN_DOZE) ||
-	    cpu_has_feature(CPU_FTR_CAN_NAP))
-		ppc_md.power_save = e500_idle;
-#endif
-	if (ppc_md.progress)
-		ppc_md.progress("id mach(): done", 0x200);
 }
 
 /* Checks "l2cr=xxxx" command-line option */
@@ -249,6 +233,21 @@ static void __init exc_lvl_early_init(void)
 #define exc_lvl_early_init()
 #endif
 
+static void setup_power_save(void)
+{
+#ifdef CONFIG_6xx
+	if (cpu_has_feature(CPU_FTR_CAN_DOZE) ||
+	    cpu_has_feature(CPU_FTR_CAN_NAP))
+		ppc_md.power_save = ppc6xx_idle;
+#endif
+
+#ifdef CONFIG_E500
+	if (cpu_has_feature(CPU_FTR_CAN_DOZE) ||
+	    cpu_has_feature(CPU_FTR_CAN_NAP))
+		ppc_md.power_save = e500_idle;
+#endif
+}
+
 /* Warning, IO base is not yet inited */
 void __init setup_arch(char **cmdline_p)
 {
@@ -260,6 +259,10 @@ void __init setup_arch(char **cmdline_p)
 	unflatten_device_tree();
 	check_for_initrd();
 
+	probe_machine();
+
+	setup_power_save();
+
 	if (ppc_md.init_early)
 		ppc_md.init_early();
 
diff --git a/arch/powerpc/platforms/40x/ep405.c b/arch/powerpc/platforms/40x/ep405.c
index ddc12a1..1c8aec6 100644
--- a/arch/powerpc/platforms/40x/ep405.c
+++ b/arch/powerpc/platforms/40x/ep405.c
@@ -105,9 +105,7 @@ static void __init ep405_setup_arch(void)
 
 static int __init ep405_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	if (!of_flat_dt_is_compatible(root, "ep405"))
+	if (!of_machine_is_compatible("ep405"))
 		return 0;
 
 	return 1;
diff --git a/arch/powerpc/platforms/40x/ppc40x_simple.c b/arch/powerpc/platforms/40x/ppc40x_simple.c
index b0c4637..2a05000 100644
--- a/arch/powerpc/platforms/40x/ppc40x_simple.c
+++ b/arch/powerpc/platforms/40x/ppc40x_simple.c
@@ -63,7 +63,7 @@ static const char * const board[] __initconst = {
 
 static int __init ppc40x_probe(void)
 {
-	if (of_flat_dt_match(of_get_flat_dt_root(), board)) {
+	if (of_device_compatible_match(of_root, board)) {
 		pci_set_flags(PCI_REASSIGN_ALL_RSRC);
 		return 1;
 	}
diff --git a/arch/powerpc/platforms/40x/virtex.c b/arch/powerpc/platforms/40x/virtex.c
index 9aa7ae2..91a08ea 100644
--- a/arch/powerpc/platforms/40x/virtex.c
+++ b/arch/powerpc/platforms/40x/virtex.c
@@ -37,9 +37,7 @@ machine_device_initcall(virtex, virtex_device_probe);
 
 static int __init virtex_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	if (!of_flat_dt_is_compatible(root, "xlnx,virtex"))
+	if (!of_machine_is_compatible("xlnx,virtex"))
 		return 0;
 
 	return 1;
diff --git a/arch/powerpc/platforms/40x/walnut.c b/arch/powerpc/platforms/40x/walnut.c
index f7ac2d0..e579781 100644
--- a/arch/powerpc/platforms/40x/walnut.c
+++ b/arch/powerpc/platforms/40x/walnut.c
@@ -46,9 +46,7 @@ machine_device_initcall(walnut, walnut_device_probe);
 
 static int __init walnut_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	if (!of_flat_dt_is_compatible(root, "ibm,walnut"))
+	if (!of_machine_is_compatible("ibm,walnut"))
 		return 0;
 
 	pci_set_flags(PCI_REASSIGN_ALL_RSRC);
diff --git a/arch/powerpc/platforms/44x/canyonlands.c b/arch/powerpc/platforms/44x/canyonlands.c
index 22ca543..157f4ce 100644
--- a/arch/powerpc/platforms/44x/canyonlands.c
+++ b/arch/powerpc/platforms/44x/canyonlands.c
@@ -53,11 +53,10 @@ machine_device_initcall(canyonlands, ppc460ex_device_probe);
 
 static int __init ppc460ex_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-	if (of_flat_dt_is_compatible(root, "amcc,canyonlands")) {
+	if (of_machine_is_compatible("amcc,canyonlands")) {
 		pci_set_flags(PCI_REASSIGN_ALL_RSRC);
 		return 1;
-		}
+	}
 	return 0;
 }
 
diff --git a/arch/powerpc/platforms/44x/ebony.c b/arch/powerpc/platforms/44x/ebony.c
index ae89322..1070225 100644
--- a/arch/powerpc/platforms/44x/ebony.c
+++ b/arch/powerpc/platforms/44x/ebony.c
@@ -49,9 +49,7 @@ machine_device_initcall(ebony, ebony_device_probe);
  */
 static int __init ebony_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	if (!of_flat_dt_is_compatible(root, "ibm,ebony"))
+	if (!of_machine_is_compatible("ibm,ebony"))
 		return 0;
 
 	pci_set_flags(PCI_REASSIGN_ALL_RSRC);
diff --git a/arch/powerpc/platforms/44x/iss4xx.c b/arch/powerpc/platforms/44x/iss4xx.c
index c7c6758..5f296dd 100644
--- a/arch/powerpc/platforms/44x/iss4xx.c
+++ b/arch/powerpc/platforms/44x/iss4xx.c
@@ -149,9 +149,7 @@ static void __init iss4xx_setup_arch(void)
  */
 static int __init iss4xx_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	if (!of_flat_dt_is_compatible(root, "ibm,iss-4xx"))
+	if (!of_machine_is_compatible("ibm,iss-4xx"))
 		return 0;
 
 	return 1;
diff --git a/arch/powerpc/platforms/44x/ppc44x_simple.c b/arch/powerpc/platforms/44x/ppc44x_simple.c
index 573c3d2..8d6e4da 100644
--- a/arch/powerpc/platforms/44x/ppc44x_simple.c
+++ b/arch/powerpc/platforms/44x/ppc44x_simple.c
@@ -67,11 +67,10 @@ static char *board[] __initdata = {
 
 static int __init ppc44x_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
 	int i = 0;
 
 	for (i = 0; i < ARRAY_SIZE(board); i++) {
-		if (of_flat_dt_is_compatible(root, board[i])) {
+		if (of_machine_is_compatible(board[i])) {
 			pci_set_flags(PCI_REASSIGN_ALL_RSRC);
 			return 1;
 		}
diff --git a/arch/powerpc/platforms/44x/ppc476.c b/arch/powerpc/platforms/44x/ppc476.c
index c11ce65..be4bf64 100644
--- a/arch/powerpc/platforms/44x/ppc476.c
+++ b/arch/powerpc/platforms/44x/ppc476.c
@@ -275,12 +275,10 @@ static void ppc47x_pci_irq_fixup(struct pci_dev *dev)
  */
 static int __init ppc47x_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	if (of_flat_dt_is_compatible(root, "ibm,akebono"))
+	if (of_machine_is_compatible("ibm,akebono"))
 		return 1;
 
-	if (of_flat_dt_is_compatible(root, "ibm,currituck")) {
+	if (of_machine_is_compatible("ibm,currituck")) {
 		ppc_md.pci_irq_fixup = ppc47x_pci_irq_fixup;
 		return 1;
 	}
diff --git a/arch/powerpc/platforms/44x/sam440ep.c b/arch/powerpc/platforms/44x/sam440ep.c
index 3ee4a03..688ffea 100644
--- a/arch/powerpc/platforms/44x/sam440ep.c
+++ b/arch/powerpc/platforms/44x/sam440ep.c
@@ -46,9 +46,7 @@ machine_device_initcall(sam440ep, sam440ep_device_probe);
 
 static int __init sam440ep_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	if (!of_flat_dt_is_compatible(root, "acube,sam440ep"))
+	if (!of_machine_is_compatible("acube,sam440ep"))
 		return 0;
 
 	pci_set_flags(PCI_REASSIGN_ALL_RSRC);
diff --git a/arch/powerpc/platforms/44x/virtex.c b/arch/powerpc/platforms/44x/virtex.c
index ad272c1..a7e0802 100644
--- a/arch/powerpc/platforms/44x/virtex.c
+++ b/arch/powerpc/platforms/44x/virtex.c
@@ -43,9 +43,7 @@ machine_device_initcall(virtex, virtex_device_probe);
 
 static int __init virtex_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	if (!of_flat_dt_is_compatible(root, "xlnx,virtex440"))
+	if (!of_machine_is_compatible("xlnx,virtex440"))
 		return 0;
 
 	return 1;
diff --git a/arch/powerpc/platforms/44x/warp.c b/arch/powerpc/platforms/44x/warp.c
index 501333c..6cf9bea 100644
--- a/arch/powerpc/platforms/44x/warp.c
+++ b/arch/powerpc/platforms/44x/warp.c
@@ -44,9 +44,7 @@ machine_device_initcall(warp, warp_device_probe);
 
 static int __init warp_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	if (!of_flat_dt_is_compatible(root, "pika,warp"))
+	if (!of_machine_compatible("pika,warp"))
 		return 0;
 
 	/* For __dma_alloc_coherent */
diff --git a/arch/powerpc/platforms/512x/mpc5121_ads.c b/arch/powerpc/platforms/512x/mpc5121_ads.c
index 3e90ece..4e03f04 100644
--- a/arch/powerpc/platforms/512x/mpc5121_ads.c
+++ b/arch/powerpc/platforms/512x/mpc5121_ads.c
@@ -57,9 +57,7 @@ static void __init mpc5121_ads_init_IRQ(void)
  */
 static int __init mpc5121_ads_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	return of_flat_dt_is_compatible(root, "fsl,mpc5121ads");
+	return of_machine_is_compatible("fsl,mpc5121ads");
 }
 
 define_machine(mpc5121_ads) {
diff --git a/arch/powerpc/platforms/512x/mpc512x_generic.c b/arch/powerpc/platforms/512x/mpc512x_generic.c
index ce71408..87eba17 100644
--- a/arch/powerpc/platforms/512x/mpc512x_generic.c
+++ b/arch/powerpc/platforms/512x/mpc512x_generic.c
@@ -38,7 +38,7 @@ static const char * const board[] __initconst = {
  */
 static int __init mpc512x_generic_probe(void)
 {
-	return of_flat_dt_match(of_get_flat_dt_root(), board);
+	return of_device_compatible_match(of_root, board);
 }
 
 define_machine(mpc512x_generic) {
diff --git a/arch/powerpc/platforms/512x/pdm360ng.c b/arch/powerpc/platforms/512x/pdm360ng.c
index 116f2325..f9cad19 100644
--- a/arch/powerpc/platforms/512x/pdm360ng.c
+++ b/arch/powerpc/platforms/512x/pdm360ng.c
@@ -113,9 +113,7 @@ void __init pdm360ng_init(void)
 
 static int __init pdm360ng_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	return of_flat_dt_is_compatible(root, "ifm,pdm360ng");
+	return of_machine_is_compatible("ifm,pdm360ng");
 }
 
 define_machine(pdm360ng) {
diff --git a/arch/powerpc/platforms/52xx/efika.c b/arch/powerpc/platforms/52xx/efika.c
index 6af651e..39b4982 100644
--- a/arch/powerpc/platforms/52xx/efika.c
+++ b/arch/powerpc/platforms/52xx/efika.c
@@ -200,8 +200,7 @@ static void __init efika_setup_arch(void)
 
 static int __init efika_probe(void)
 {
-	const char *model = of_get_flat_dt_prop(of_get_flat_dt_root(),
-						"model", NULL);
+	const char *model = of_get_property(of_root, "model", NULL);
 
 	if (model == NULL)
 		return 0;
diff --git a/arch/powerpc/platforms/52xx/lite5200.c b/arch/powerpc/platforms/52xx/lite5200.c
index 7492de3..c94c385 100644
--- a/arch/powerpc/platforms/52xx/lite5200.c
+++ b/arch/powerpc/platforms/52xx/lite5200.c
@@ -183,7 +183,7 @@ static const char * const board[] __initconst = {
  */
 static int __init lite5200_probe(void)
 {
-	return of_flat_dt_match(of_get_flat_dt_root(), board);
+	return of_device_compatible_match(of_root, board);
 }
 
 define_machine(lite5200) {
diff --git a/arch/powerpc/platforms/52xx/media5200.c b/arch/powerpc/platforms/52xx/media5200.c
index 8fb9548..a322704 100644
--- a/arch/powerpc/platforms/52xx/media5200.c
+++ b/arch/powerpc/platforms/52xx/media5200.c
@@ -242,7 +242,7 @@ static const char * const board[] __initconst = {
  */
 static int __init media5200_probe(void)
 {
-	return of_flat_dt_match(of_get_flat_dt_root(), board);
+	return of_device_compatible_match(of_root, board);
 }
 
 define_machine(media5200_platform) {
diff --git a/arch/powerpc/platforms/52xx/mpc5200_simple.c b/arch/powerpc/platforms/52xx/mpc5200_simple.c
index 792a301..a80c627 100644
--- a/arch/powerpc/platforms/52xx/mpc5200_simple.c
+++ b/arch/powerpc/platforms/52xx/mpc5200_simple.c
@@ -70,7 +70,7 @@ static const char *board[] __initdata = {
  */
 static int __init mpc5200_simple_probe(void)
 {
-	return of_flat_dt_match(of_get_flat_dt_root(), board);
+	return of_device_compatible_match(of_root, board);
 }
 
 define_machine(mpc5200_simple_platform) {
diff --git a/arch/powerpc/platforms/82xx/ep8248e.c b/arch/powerpc/platforms/82xx/ep8248e.c
index 6781bda..cdab847 100644
--- a/arch/powerpc/platforms/82xx/ep8248e.c
+++ b/arch/powerpc/platforms/82xx/ep8248e.c
@@ -309,8 +309,7 @@ machine_device_initcall(ep8248e, declare_of_platform_devices);
  */
 static int __init ep8248e_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-	return of_flat_dt_is_compatible(root, "fsl,ep8248e");
+	return of_machine_is_compatible("fsl,ep8248e");
 }
 
 define_machine(ep8248e)
diff --git a/arch/powerpc/platforms/82xx/km82xx.c b/arch/powerpc/platforms/82xx/km82xx.c
index 387b446..28860e4 100644
--- a/arch/powerpc/platforms/82xx/km82xx.c
+++ b/arch/powerpc/platforms/82xx/km82xx.c
@@ -198,8 +198,7 @@ machine_device_initcall(km82xx, declare_of_platform_devices);
  */
 static int __init km82xx_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-	return of_flat_dt_is_compatible(root, "keymile,km82xx");
+	return of_machine_is_compatible("keymile,km82xx");
 }
 
 define_machine(km82xx)
diff --git a/arch/powerpc/platforms/82xx/mpc8272_ads.c b/arch/powerpc/platforms/82xx/mpc8272_ads.c
index d24deac..d23c10a 100644
--- a/arch/powerpc/platforms/82xx/mpc8272_ads.c
+++ b/arch/powerpc/platforms/82xx/mpc8272_ads.c
@@ -201,8 +201,7 @@ machine_device_initcall(mpc8272_ads, declare_of_platform_devices);
  */
 static int __init mpc8272_ads_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-	return of_flat_dt_is_compatible(root, "fsl,mpc8272ads");
+	return of_machine_is_compatible("fsl,mpc8272ads");
 }
 
 define_machine(mpc8272_ads)
diff --git a/arch/powerpc/platforms/82xx/pq2fads.c b/arch/powerpc/platforms/82xx/pq2fads.c
index 3a5164a..6c654dc 100644
--- a/arch/powerpc/platforms/82xx/pq2fads.c
+++ b/arch/powerpc/platforms/82xx/pq2fads.c
@@ -164,8 +164,7 @@ static void __init pq2fads_setup_arch(void)
  */
 static int __init pq2fads_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-	return of_flat_dt_is_compatible(root, "fsl,pq2fads");
+	return of_machine_is_compatible("fsl,pq2fads");
 }
 
 static const struct of_device_id of_bus_ids[] __initconst = {
diff --git a/arch/powerpc/platforms/83xx/asp834x.c b/arch/powerpc/platforms/83xx/asp834x.c
index 464ea8e..17e5433 100644
--- a/arch/powerpc/platforms/83xx/asp834x.c
+++ b/arch/powerpc/platforms/83xx/asp834x.c
@@ -43,8 +43,7 @@ machine_device_initcall(asp834x, mpc83xx_declare_of_platform_devices);
  */
 static int __init asp834x_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-	return of_flat_dt_is_compatible(root, "analogue-and-micro,asp8347e");
+	return of_machine_is_compatible("analogue-and-micro,asp8347e");
 }
 
 define_machine(asp834x) {
diff --git a/arch/powerpc/platforms/83xx/km83xx.c b/arch/powerpc/platforms/83xx/km83xx.c
index 4bc6bbb..e7fbd63 100644
--- a/arch/powerpc/platforms/83xx/km83xx.c
+++ b/arch/powerpc/platforms/83xx/km83xx.c
@@ -171,11 +171,10 @@ static char *board[] __initdata = {
  */
 static int __init mpc83xx_km_probe(void)
 {
-	unsigned long node = of_get_flat_dt_root();
 	int i = 0;
 
 	while (board[i]) {
-		if (of_flat_dt_is_compatible(node, board[i]))
+		if (of_machine_is_compatible(board[i]))
 			break;
 		i++;
 	}
diff --git a/arch/powerpc/platforms/83xx/mpc830x_rdb.c b/arch/powerpc/platforms/83xx/mpc830x_rdb.c
index 4f2d9fe..040d5d0 100644
--- a/arch/powerpc/platforms/83xx/mpc830x_rdb.c
+++ b/arch/powerpc/platforms/83xx/mpc830x_rdb.c
@@ -46,7 +46,7 @@ static const char *board[] __initdata = {
  */
 static int __init mpc830x_rdb_probe(void)
 {
-	return of_flat_dt_match(of_get_flat_dt_root(), board);
+	return of_device_compatible_match(of_root, board);
 }
 
 machine_device_initcall(mpc830x_rdb, mpc83xx_declare_of_platform_devices);
diff --git a/arch/powerpc/platforms/83xx/mpc831x_rdb.c b/arch/powerpc/platforms/83xx/mpc831x_rdb.c
index fa25977..40e0d83 100644
--- a/arch/powerpc/platforms/83xx/mpc831x_rdb.c
+++ b/arch/powerpc/platforms/83xx/mpc831x_rdb.c
@@ -46,7 +46,7 @@ static const char *board[] __initdata = {
  */
 static int __init mpc831x_rdb_probe(void)
 {
-	return of_flat_dt_match(of_get_flat_dt_root(), board);
+	return of_device_compatible_match(of_root, board);
 }
 
 machine_device_initcall(mpc831x_rdb, mpc83xx_declare_of_platform_devices);
diff --git a/arch/powerpc/platforms/83xx/mpc832x_mds.c b/arch/powerpc/platforms/83xx/mpc832x_mds.c
index a973b2a..375a58d 100644
--- a/arch/powerpc/platforms/83xx/mpc832x_mds.c
+++ b/arch/powerpc/platforms/83xx/mpc832x_mds.c
@@ -102,9 +102,7 @@ machine_device_initcall(mpc832x_mds, mpc83xx_declare_of_platform_devices);
  */
 static int __init mpc832x_sys_probe(void)
 {
-        unsigned long root = of_get_flat_dt_root();
-
-        return of_flat_dt_is_compatible(root, "MPC832xMDS");
+        return of_machine_is_compatible("MPC832xMDS");
 }
 
 define_machine(mpc832x_mds) {
diff --git a/arch/powerpc/platforms/83xx/mpc832x_rdb.c b/arch/powerpc/platforms/83xx/mpc832x_rdb.c
index ea2b87d..2ef03e7 100644
--- a/arch/powerpc/platforms/83xx/mpc832x_rdb.c
+++ b/arch/powerpc/platforms/83xx/mpc832x_rdb.c
@@ -220,9 +220,7 @@ machine_device_initcall(mpc832x_rdb, mpc83xx_declare_of_platform_devices);
  */
 static int __init mpc832x_rdb_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	return of_flat_dt_is_compatible(root, "MPC832xRDB");
+	return of_machine_is_compatible("MPC832xRDB");
 }
 
 define_machine(mpc832x_rdb) {
diff --git a/arch/powerpc/platforms/83xx/mpc834x_itx.c b/arch/powerpc/platforms/83xx/mpc834x_itx.c
index 80aea8c..88657a4 100644
--- a/arch/powerpc/platforms/83xx/mpc834x_itx.c
+++ b/arch/powerpc/platforms/83xx/mpc834x_itx.c
@@ -70,9 +70,7 @@ static void __init mpc834x_itx_setup_arch(void)
  */
 static int __init mpc834x_itx_probe(void)
 {
-        unsigned long root = of_get_flat_dt_root();
-
-        return of_flat_dt_is_compatible(root, "MPC834xMITX");
+        return of_machine_is_compatible("MPC834xMITX");
 }
 
 define_machine(mpc834x_itx) {
diff --git a/arch/powerpc/platforms/83xx/mpc834x_mds.c b/arch/powerpc/platforms/83xx/mpc834x_mds.c
index 553e793..eeaee61 100644
--- a/arch/powerpc/platforms/83xx/mpc834x_mds.c
+++ b/arch/powerpc/platforms/83xx/mpc834x_mds.c
@@ -91,9 +91,7 @@ machine_device_initcall(mpc834x_mds, mpc83xx_declare_of_platform_devices);
  */
 static int __init mpc834x_mds_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	return of_flat_dt_is_compatible(root, "MPC834xMDS");
+	return of_machine_is_compatible("MPC834xMDS");
 }
 
 define_machine(mpc834x_mds) {
diff --git a/arch/powerpc/platforms/83xx/mpc836x_mds.c b/arch/powerpc/platforms/83xx/mpc836x_mds.c
index dd70b85..ada6e5e 100644
--- a/arch/powerpc/platforms/83xx/mpc836x_mds.c
+++ b/arch/powerpc/platforms/83xx/mpc836x_mds.c
@@ -209,9 +209,7 @@ machine_arch_initcall(mpc836x_mds, mpc836x_usb_cfg);
  */
 static int __init mpc836x_mds_probe(void)
 {
-        unsigned long root = of_get_flat_dt_root();
-
-        return of_flat_dt_is_compatible(root, "MPC836xMDS");
+        return of_machine_is_compatible("MPC836xMDS");
 }
 
 define_machine(mpc836x_mds) {
diff --git a/arch/powerpc/platforms/83xx/mpc836x_rdk.c b/arch/powerpc/platforms/83xx/mpc836x_rdk.c
index 4cd7153..cf67ac9 100644
--- a/arch/powerpc/platforms/83xx/mpc836x_rdk.c
+++ b/arch/powerpc/platforms/83xx/mpc836x_rdk.c
@@ -42,9 +42,7 @@ static void __init mpc836x_rdk_setup_arch(void)
  */
 static int __init mpc836x_rdk_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	return of_flat_dt_is_compatible(root, "fsl,mpc8360rdk");
+	return of_machine_is_compatible("fsl,mpc8360rdk");
 }
 
 define_machine(mpc836x_rdk) {
diff --git a/arch/powerpc/platforms/83xx/mpc837x_mds.c b/arch/powerpc/platforms/83xx/mpc837x_mds.c
index e53a60b..199f6ee 100644
--- a/arch/powerpc/platforms/83xx/mpc837x_mds.c
+++ b/arch/powerpc/platforms/83xx/mpc837x_mds.c
@@ -93,9 +93,7 @@ machine_device_initcall(mpc837x_mds, mpc83xx_declare_of_platform_devices);
  */
 static int __init mpc837x_mds_probe(void)
 {
-        unsigned long root = of_get_flat_dt_root();
-
-        return of_flat_dt_is_compatible(root, "fsl,mpc837xmds");
+        return of_machine_is_compatible("fsl,mpc837xmds");
 }
 
 define_machine(mpc837x_mds) {
diff --git a/arch/powerpc/platforms/83xx/mpc837x_rdb.c b/arch/powerpc/platforms/83xx/mpc837x_rdb.c
index 9813c81..667731d 100644
--- a/arch/powerpc/platforms/83xx/mpc837x_rdb.c
+++ b/arch/powerpc/platforms/83xx/mpc837x_rdb.c
@@ -73,7 +73,7 @@ static const char * const board[] __initconst = {
  */
 static int __init mpc837x_rdb_probe(void)
 {
-	return of_flat_dt_match(of_get_flat_dt_root(), board);
+	return of_device_compatible_match(of_root, board);
 }
 
 define_machine(mpc837x_rdb) {
diff --git a/arch/powerpc/platforms/83xx/sbc834x.c b/arch/powerpc/platforms/83xx/sbc834x.c
index 26cb3e9..b867e88 100644
--- a/arch/powerpc/platforms/83xx/sbc834x.c
+++ b/arch/powerpc/platforms/83xx/sbc834x.c
@@ -60,9 +60,7 @@ machine_device_initcall(sbc834x, mpc83xx_declare_of_platform_devices);
  */
 static int __init sbc834x_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	return of_flat_dt_is_compatible(root, "SBC834xE");
+	return of_machine_is_compatible("SBC834xE");
 }
 
 define_machine(sbc834x) {
diff --git a/arch/powerpc/platforms/85xx/bsc913x_qds.c b/arch/powerpc/platforms/85xx/bsc913x_qds.c
index dcfafd6..07dd6ae 100644
--- a/arch/powerpc/platforms/85xx/bsc913x_qds.c
+++ b/arch/powerpc/platforms/85xx/bsc913x_qds.c
@@ -60,9 +60,7 @@ machine_arch_initcall(bsc9132_qds, mpc85xx_common_publish_devices);
 
 static int __init bsc9132_qds_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	return of_flat_dt_is_compatible(root, "fsl,bsc9132qds");
+	return of_machine_is_compatible("fsl,bsc9132qds");
 }
 
 define_machine(bsc9132_qds) {
diff --git a/arch/powerpc/platforms/85xx/bsc913x_rdb.c b/arch/powerpc/platforms/85xx/bsc913x_rdb.c
index 9d57bed..e48f671 100644
--- a/arch/powerpc/platforms/85xx/bsc913x_rdb.c
+++ b/arch/powerpc/platforms/85xx/bsc913x_rdb.c
@@ -50,9 +50,7 @@ machine_device_initcall(bsc9131_rdb, mpc85xx_common_publish_devices);
 
 static int __init bsc9131_rdb_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	return of_flat_dt_is_compatible(root, "fsl,bsc9131rdb");
+	return of_machine_is_compatible("fsl,bsc9131rdb");
 }
 
 define_machine(bsc9131_rdb) {
diff --git a/arch/powerpc/platforms/85xx/c293pcie.c b/arch/powerpc/platforms/85xx/c293pcie.c
index 61bc851..3b9e3f0 100644
--- a/arch/powerpc/platforms/85xx/c293pcie.c
+++ b/arch/powerpc/platforms/85xx/c293pcie.c
@@ -54,9 +54,7 @@ machine_arch_initcall(c293_pcie, mpc85xx_common_publish_devices);
  */
 static int __init c293_pcie_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	if (of_flat_dt_is_compatible(root, "fsl,C293PCIE"))
+	if (of_machine_is_compatible("fsl,C293PCIE"))
 		return 1;
 	return 0;
 }
diff --git a/arch/powerpc/platforms/85xx/corenet_generic.c b/arch/powerpc/platforms/85xx/corenet_generic.c
index a2b0bc8..3a6a84f 100644
--- a/arch/powerpc/platforms/85xx/corenet_generic.c
+++ b/arch/powerpc/platforms/85xx/corenet_generic.c
@@ -170,20 +170,19 @@ static const char * const boards[] __initconst = {
  */
 static int __init corenet_generic_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
 	char hv_compat[24];
 	int i;
 #ifdef CONFIG_SMP
 	extern struct smp_ops_t smp_85xx_ops;
 #endif
 
-	if (of_flat_dt_match(root, boards))
+	if (of_device_compatible_match(of_root, boards))
 		return 1;
 
 	/* Check if we're running under the Freescale hypervisor */
 	for (i = 0; boards[i]; i++) {
 		snprintf(hv_compat, sizeof(hv_compat), "%s-hv", boards[i]);
-		if (of_flat_dt_is_compatible(root, hv_compat)) {
+		if (of_machine_is_compatible(hv_compat)) {
 			ppc_md.init_IRQ = ehv_pic_init;
 
 			ppc_md.get_irq = ehv_pic_get_irq;
diff --git a/arch/powerpc/platforms/85xx/ge_imp3a.c b/arch/powerpc/platforms/85xx/ge_imp3a.c
index 55eefef..14af36a 100644
--- a/arch/powerpc/platforms/85xx/ge_imp3a.c
+++ b/arch/powerpc/platforms/85xx/ge_imp3a.c
@@ -197,9 +197,7 @@ static void ge_imp3a_show_cpuinfo(struct seq_file *m)
  */
 static int __init ge_imp3a_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	return of_flat_dt_is_compatible(root, "ge,IMP3A");
+	return of_machine_is_compatible("ge,IMP3A");
 }
 
 machine_arch_initcall(ge_imp3a, mpc85xx_common_publish_devices);
diff --git a/arch/powerpc/platforms/85xx/ksi8560.c b/arch/powerpc/platforms/85xx/ksi8560.c
index 3dc1bda..e841675 100644
--- a/arch/powerpc/platforms/85xx/ksi8560.c
+++ b/arch/powerpc/platforms/85xx/ksi8560.c
@@ -176,9 +176,7 @@ machine_device_initcall(ksi8560, mpc85xx_common_publish_devices);
  */
 static int __init ksi8560_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	return of_flat_dt_is_compatible(root, "emerson,KSI8560");
+	return of_machine_is_compatible("emerson,KSI8560");
 }
 
 define_machine(ksi8560) {
diff --git a/arch/powerpc/platforms/85xx/mpc8536_ds.c b/arch/powerpc/platforms/85xx/mpc8536_ds.c
index a378ba3..6ba687f 100644
--- a/arch/powerpc/platforms/85xx/mpc8536_ds.c
+++ b/arch/powerpc/platforms/85xx/mpc8536_ds.c
@@ -64,9 +64,7 @@ machine_arch_initcall(mpc8536_ds, swiotlb_setup_bus_notifier);
  */
 static int __init mpc8536_ds_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	return of_flat_dt_is_compatible(root, "fsl,mpc8536ds");
+	return of_machine_is_compatible("fsl,mpc8536ds");
 }
 
 define_machine(mpc8536_ds) {
diff --git a/arch/powerpc/platforms/85xx/mpc85xx_ads.c b/arch/powerpc/platforms/85xx/mpc85xx_ads.c
index de72a5f..0fba649 100644
--- a/arch/powerpc/platforms/85xx/mpc85xx_ads.c
+++ b/arch/powerpc/platforms/85xx/mpc85xx_ads.c
@@ -160,9 +160,7 @@ machine_arch_initcall(mpc85xx_ads, mpc85xx_common_publish_devices);
  */
 static int __init mpc85xx_ads_probe(void)
 {
-        unsigned long root = of_get_flat_dt_root();
-
-        return of_flat_dt_is_compatible(root, "MPC85xxADS");
+        return of_machine_is_compatible("MPC85xxADS");
 }
 
 define_machine(mpc85xx_ads) {
diff --git a/arch/powerpc/platforms/85xx/mpc85xx_cds.c b/arch/powerpc/platforms/85xx/mpc85xx_cds.c
index d7e87ff..827155a 100644
--- a/arch/powerpc/platforms/85xx/mpc85xx_cds.c
+++ b/arch/powerpc/platforms/85xx/mpc85xx_cds.c
@@ -367,9 +367,7 @@ static void mpc85xx_cds_show_cpuinfo(struct seq_file *m)
  */
 static int __init mpc85xx_cds_probe(void)
 {
-        unsigned long root = of_get_flat_dt_root();
-
-        return of_flat_dt_is_compatible(root, "MPC85xxCDS");
+        return of_machine_is_compatible("MPC85xxCDS");
 }
 
 machine_arch_initcall(mpc85xx_cds, mpc85xx_common_publish_devices);
diff --git a/arch/powerpc/platforms/85xx/mpc85xx_ds.c b/arch/powerpc/platforms/85xx/mpc85xx_ds.c
index 64a7e8c..6bc07d8 100644
--- a/arch/powerpc/platforms/85xx/mpc85xx_ds.c
+++ b/arch/powerpc/platforms/85xx/mpc85xx_ds.c
@@ -167,9 +167,7 @@ static void __init mpc85xx_ds_setup_arch(void)
  */
 static int __init mpc8544_ds_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	return !!of_flat_dt_is_compatible(root, "MPC8544DS");
+	return !!of_machine_is_compatible("MPC8544DS");
 }
 
 machine_arch_initcall(mpc8544_ds, mpc85xx_common_publish_devices);
@@ -185,9 +183,7 @@ machine_arch_initcall(p2020_ds, swiotlb_setup_bus_notifier);
  */
 static int __init mpc8572_ds_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	return !!of_flat_dt_is_compatible(root, "fsl,MPC8572DS");
+	return !!of_machine_is_compatible("fsl,MPC8572DS");
 }
 
 /*
@@ -195,9 +191,7 @@ static int __init mpc8572_ds_probe(void)
  */
 static int __init p2020_ds_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	return !!of_flat_dt_is_compatible(root, "fsl,P2020DS");
+	return !!of_machine_is_compatible("fsl,P2020DS");
 }
 
 define_machine(mpc8544_ds) {
diff --git a/arch/powerpc/platforms/85xx/mpc85xx_mds.c b/arch/powerpc/platforms/85xx/mpc85xx_mds.c
index f61cbe2..c26db01 100644
--- a/arch/powerpc/platforms/85xx/mpc85xx_mds.c
+++ b/arch/powerpc/platforms/85xx/mpc85xx_mds.c
@@ -376,9 +376,7 @@ static void __init mpc85xx_mds_pic_init(void)
 
 static int __init mpc85xx_mds_probe(void)
 {
-        unsigned long root = of_get_flat_dt_root();
-
-        return of_flat_dt_is_compatible(root, "MPC85xxMDS");
+        return of_machine_is_compatible("MPC85xxMDS");
 }
 
 define_machine(mpc8568_mds) {
@@ -398,9 +396,7 @@ define_machine(mpc8568_mds) {
 
 static int __init mpc8569_mds_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	return of_flat_dt_is_compatible(root, "fsl,MPC8569EMDS");
+	return of_machine_is_compatible("fsl,MPC8569EMDS");
 }
 
 define_machine(mpc8569_mds) {
@@ -420,9 +416,7 @@ define_machine(mpc8569_mds) {
 
 static int __init p1021_mds_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	return of_flat_dt_is_compatible(root, "fsl,P1021MDS");
+	return of_machine_is_compatible("fsl,P1021MDS");
 
 }
 
diff --git a/arch/powerpc/platforms/85xx/mpc85xx_rdb.c b/arch/powerpc/platforms/85xx/mpc85xx_rdb.c
index 761e504..c1499cb 100644
--- a/arch/powerpc/platforms/85xx/mpc85xx_rdb.c
+++ b/arch/powerpc/platforms/85xx/mpc85xx_rdb.c
@@ -147,80 +147,60 @@ machine_arch_initcall(p1024_rdb, mpc85xx_common_publish_devices);
  */
 static int __init p2020_rdb_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	if (of_flat_dt_is_compatible(root, "fsl,P2020RDB"))
+	if (of_machine_is_compatible("fsl,P2020RDB"))
 		return 1;
 	return 0;
 }
 
 static int __init p1020_rdb_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	if (of_flat_dt_is_compatible(root, "fsl,P1020RDB"))
+	if (of_machine_is_compatible("fsl,P1020RDB"))
 		return 1;
 	return 0;
 }
 
 static int __init p1020_rdb_pc_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	return of_flat_dt_is_compatible(root, "fsl,P1020RDB-PC");
+	return of_machine_is_compatible("fsl,P1020RDB-PC");
 }
 
 static int __init p1020_rdb_pd_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	return of_flat_dt_is_compatible(root, "fsl,P1020RDB-PD");
+	return of_machine_is_compatible("fsl,P1020RDB-PD");
 }
 
 static int __init p1021_rdb_pc_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	if (of_flat_dt_is_compatible(root, "fsl,P1021RDB-PC"))
+	if (of_machine_is_compatible("fsl,P1021RDB-PC"))
 		return 1;
 	return 0;
 }
 
 static int __init p2020_rdb_pc_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	if (of_flat_dt_is_compatible(root, "fsl,P2020RDB-PC"))
+	if (of_machine_is_compatible("fsl,P2020RDB-PC"))
 		return 1;
 	return 0;
 }
 
 static int __init p1025_rdb_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	return of_flat_dt_is_compatible(root, "fsl,P1025RDB");
+	return of_machine_is_compatible("fsl,P1025RDB");
 }
 
 static int __init p1020_mbg_pc_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	return of_flat_dt_is_compatible(root, "fsl,P1020MBG-PC");
+	return of_machine_is_compatible("fsl,P1020MBG-PC");
 }
 
 static int __init p1020_utm_pc_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	return of_flat_dt_is_compatible(root, "fsl,P1020UTM-PC");
+	return of_machine_is_compatible("fsl,P1020UTM-PC");
 }
 
 static int __init p1024_rdb_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	return of_flat_dt_is_compatible(root, "fsl,P1024RDB");
+	return of_machine_is_compatible("fsl,P1024RDB");
 }
 
 define_machine(p2020_rdb) {
diff --git a/arch/powerpc/platforms/85xx/mvme2500.c b/arch/powerpc/platforms/85xx/mvme2500.c
index 1233050..acc3d0d 100644
--- a/arch/powerpc/platforms/85xx/mvme2500.c
+++ b/arch/powerpc/platforms/85xx/mvme2500.c
@@ -53,9 +53,7 @@ machine_arch_initcall(mvme2500, mpc85xx_common_publish_devices);
  */
 static int __init mvme2500_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	return of_flat_dt_is_compatible(root, "artesyn,MVME2500");
+	return of_machine_is_compatible("artesyn,MVME2500");
 }
 
 define_machine(mvme2500) {
diff --git a/arch/powerpc/platforms/85xx/p1010rdb.c b/arch/powerpc/platforms/85xx/p1010rdb.c
index ad1a3d4..661d7b5 100644
--- a/arch/powerpc/platforms/85xx/p1010rdb.c
+++ b/arch/powerpc/platforms/85xx/p1010rdb.c
@@ -62,11 +62,9 @@ machine_arch_initcall(p1010_rdb, swiotlb_setup_bus_notifier);
  */
 static int __init p1010_rdb_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	if (of_flat_dt_is_compatible(root, "fsl,P1010RDB"))
+	if (of_machine_is_compatible("fsl,P1010RDB"))
 		return 1;
-	if (of_flat_dt_is_compatible(root, "fsl,P1010RDB-PB"))
+	if (of_machine_is_compatible("fsl,P1010RDB-PB"))
 		return 1;
 	return 0;
 }
diff --git a/arch/powerpc/platforms/85xx/p1022_ds.c b/arch/powerpc/platforms/85xx/p1022_ds.c
index 371df82..63568d6 100644
--- a/arch/powerpc/platforms/85xx/p1022_ds.c
+++ b/arch/powerpc/platforms/85xx/p1022_ds.c
@@ -555,9 +555,7 @@ machine_arch_initcall(p1022_ds, swiotlb_setup_bus_notifier);
  */
 static int __init p1022_ds_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	return of_flat_dt_is_compatible(root, "fsl,p1022ds");
+	return of_machine_is_compatible("fsl,p1022ds");
 }
 
 define_machine(p1022_ds) {
diff --git a/arch/powerpc/platforms/85xx/p1022_rdk.c b/arch/powerpc/platforms/85xx/p1022_rdk.c
index 5087bec..2f29436 100644
--- a/arch/powerpc/platforms/85xx/p1022_rdk.c
+++ b/arch/powerpc/platforms/85xx/p1022_rdk.c
@@ -135,9 +135,7 @@ machine_arch_initcall(p1022_rdk, swiotlb_setup_bus_notifier);
  */
 static int __init p1022_rdk_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	return of_flat_dt_is_compatible(root, "fsl,p1022rdk");
+	return of_machine_is_compatible("fsl,p1022rdk");
 }
 
 define_machine(p1022_rdk) {
diff --git a/arch/powerpc/platforms/85xx/p1023_rdb.c b/arch/powerpc/platforms/85xx/p1023_rdb.c
index d5b7509..40d8de5 100644
--- a/arch/powerpc/platforms/85xx/p1023_rdb.c
+++ b/arch/powerpc/platforms/85xx/p1023_rdb.c
@@ -100,9 +100,7 @@ static void __init mpc85xx_rdb_pic_init(void)
 
 static int __init p1023_rdb_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	return of_flat_dt_is_compatible(root, "fsl,P1023RDB");
+	return of_machine_is_compatible("fsl,P1023RDB");
 
 }
 
diff --git a/arch/powerpc/platforms/85xx/ppa8548.c b/arch/powerpc/platforms/85xx/ppa8548.c
index 12019f1..2410167 100644
--- a/arch/powerpc/platforms/85xx/ppa8548.c
+++ b/arch/powerpc/platforms/85xx/ppa8548.c
@@ -81,9 +81,7 @@ machine_device_initcall(ppa8548, declare_of_platform_devices);
  */
 static int __init ppa8548_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	return of_flat_dt_is_compatible(root, "ppa8548");
+	return of_machine_is_compatible("ppa8548");
 }
 
 define_machine(ppa8548) {
diff --git a/arch/powerpc/platforms/85xx/qemu_e500.c b/arch/powerpc/platforms/85xx/qemu_e500.c
index 8ad2fe6..50d7458 100644
--- a/arch/powerpc/platforms/85xx/qemu_e500.c
+++ b/arch/powerpc/platforms/85xx/qemu_e500.c
@@ -62,9 +62,7 @@ static void __init qemu_e500_setup_arch(void)
  */
 static int __init qemu_e500_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	return !!of_flat_dt_is_compatible(root, "fsl,qemu-e500");
+	return !!of_machine_is_compatible("fsl,qemu-e500");
 }
 
 machine_arch_initcall(qemu_e500, mpc85xx_common_publish_devices);
diff --git a/arch/powerpc/platforms/85xx/sbc8548.c b/arch/powerpc/platforms/85xx/sbc8548.c
index b072146..c967180 100644
--- a/arch/powerpc/platforms/85xx/sbc8548.c
+++ b/arch/powerpc/platforms/85xx/sbc8548.c
@@ -120,9 +120,7 @@ machine_arch_initcall(sbc8548, mpc85xx_common_publish_devices);
  */
 static int __init sbc8548_probe(void)
 {
-        unsigned long root = of_get_flat_dt_root();
-
-        return of_flat_dt_is_compatible(root, "SBC8548");
+        return of_machine_is_compatible("SBC8548");
 }
 
 define_machine(sbc8548) {
diff --git a/arch/powerpc/platforms/85xx/socrates.c b/arch/powerpc/platforms/85xx/socrates.c
index ae368e0..cd255ac 100644
--- a/arch/powerpc/platforms/85xx/socrates.c
+++ b/arch/powerpc/platforms/85xx/socrates.c
@@ -79,9 +79,7 @@ machine_arch_initcall(socrates, mpc85xx_common_publish_devices);
  */
 static int __init socrates_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	if (of_flat_dt_is_compatible(root, "abb,socrates"))
+	if (of_machine_is_compatible("abb,socrates"))
 		return 1;
 
 	return 0;
diff --git a/arch/powerpc/platforms/85xx/stx_gp3.c b/arch/powerpc/platforms/85xx/stx_gp3.c
index 6f4939b..91b824c 100644
--- a/arch/powerpc/platforms/85xx/stx_gp3.c
+++ b/arch/powerpc/platforms/85xx/stx_gp3.c
@@ -93,9 +93,7 @@ machine_arch_initcall(stx_gp3, mpc85xx_common_publish_devices);
  */
 static int __init stx_gp3_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	return of_flat_dt_is_compatible(root, "stx,gp3-8560");
+	return of_machine_is_compatible("stx,gp3-8560");
 }
 
 define_machine(stx_gp3) {
diff --git a/arch/powerpc/platforms/85xx/tqm85xx.c b/arch/powerpc/platforms/85xx/tqm85xx.c
index ec0b727..b7c5445 100644
--- a/arch/powerpc/platforms/85xx/tqm85xx.c
+++ b/arch/powerpc/platforms/85xx/tqm85xx.c
@@ -122,7 +122,7 @@ static const char * const board[] __initconst = {
  */
 static int __init tqm85xx_probe(void)
 {
-	return of_flat_dt_match(of_get_flat_dt_root(), board);
+	return of_device_compatible_match(of_root, board);
 }
 
 define_machine(tqm85xx) {
diff --git a/arch/powerpc/platforms/85xx/twr_p102x.c b/arch/powerpc/platforms/85xx/twr_p102x.c
index 71bc255..6407d51 100644
--- a/arch/powerpc/platforms/85xx/twr_p102x.c
+++ b/arch/powerpc/platforms/85xx/twr_p102x.c
@@ -127,10 +127,8 @@ static void __init twr_p1025_setup_arch(void)
 machine_arch_initcall(twr_p1025, mpc85xx_common_publish_devices);
 
 static int __init twr_p1025_probe(void)
-{
-	unsigned long root = of_get_flat_dt_root();
 
-	return of_flat_dt_is_compatible(root, "fsl,TWR-P1025");
+	return of_machine_is_compatible("fsl,TWR-P1025");
 }
 
 define_machine(twr_p1025) {
diff --git a/arch/powerpc/platforms/85xx/xes_mpc85xx.c b/arch/powerpc/platforms/85xx/xes_mpc85xx.c
index 1a9c108..cf0c70f 100644
--- a/arch/powerpc/platforms/85xx/xes_mpc85xx.c
+++ b/arch/powerpc/platforms/85xx/xes_mpc85xx.c
@@ -144,23 +144,17 @@ machine_arch_initcall(xes_mpc8540, mpc85xx_common_publish_devices);
  */
 static int __init xes_mpc8572_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	return of_flat_dt_is_compatible(root, "xes,MPC8572");
+	return of_machine_is_compatible("xes,MPC8572");
 }
 
 static int __init xes_mpc8548_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	return of_flat_dt_is_compatible(root, "xes,MPC8548");
+	return of_machine_is_compatible("xes,MPC8548");
 }
 
 static int __init xes_mpc8540_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	return of_flat_dt_is_compatible(root, "xes,MPC8540");
+	return of_machine_is_compatible("xes,MPC8540");
 }
 
 define_machine(xes_mpc8572) {
diff --git a/arch/powerpc/platforms/86xx/gef_ppc9a.c b/arch/powerpc/platforms/86xx/gef_ppc9a.c
index 8e63b75..ef684af 100644
--- a/arch/powerpc/platforms/86xx/gef_ppc9a.c
+++ b/arch/powerpc/platforms/86xx/gef_ppc9a.c
@@ -189,9 +189,7 @@ DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_NEC, PCI_DEVICE_ID_NEC_USB,
  */
 static int __init gef_ppc9a_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	if (of_flat_dt_is_compatible(root, "gef,ppc9a"))
+	if (of_machine_is_compatible("gef,ppc9a"))
 		return 1;
 
 	return 0;
diff --git a/arch/powerpc/platforms/86xx/gef_sbc310.c b/arch/powerpc/platforms/86xx/gef_sbc310.c
index 0e0be94..67dd0c2 100644
--- a/arch/powerpc/platforms/86xx/gef_sbc310.c
+++ b/arch/powerpc/platforms/86xx/gef_sbc310.c
@@ -176,9 +176,7 @@ DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_NEC, PCI_DEVICE_ID_NEC_USB,
  */
 static int __init gef_sbc310_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	if (of_flat_dt_is_compatible(root, "gef,sbc310"))
+	if (of_machine_is_compatible("gef,sbc310"))
 		return 1;
 
 	return 0;
diff --git a/arch/powerpc/platforms/86xx/gef_sbc610.c b/arch/powerpc/platforms/86xx/gef_sbc610.c
index e8292b4..8050269 100644
--- a/arch/powerpc/platforms/86xx/gef_sbc610.c
+++ b/arch/powerpc/platforms/86xx/gef_sbc610.c
@@ -166,9 +166,7 @@ DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_NEC, PCI_DEVICE_ID_NEC_USB,
  */
 static int __init gef_sbc610_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	if (of_flat_dt_is_compatible(root, "gef,sbc610"))
+	if (of_machine_is_compatible("gef,sbc610"))
 		return 1;
 
 	return 0;
diff --git a/arch/powerpc/platforms/86xx/mpc8610_hpcd.c b/arch/powerpc/platforms/86xx/mpc8610_hpcd.c
index 957473e..fef0582 100644
--- a/arch/powerpc/platforms/86xx/mpc8610_hpcd.c
+++ b/arch/powerpc/platforms/86xx/mpc8610_hpcd.c
@@ -319,9 +319,7 @@ static void __init mpc86xx_hpcd_setup_arch(void)
  */
 static int __init mpc86xx_hpcd_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	if (of_flat_dt_is_compatible(root, "fsl,MPC8610HPCD"))
+	if (of_machine_is_compatible("fsl,MPC8610HPCD"))
 		return 1;	/* Looks good */
 
 	return 0;
diff --git a/arch/powerpc/platforms/86xx/mpc86xx_hpcn.c b/arch/powerpc/platforms/86xx/mpc86xx_hpcn.c
index e508481..5ae42a0 100644
--- a/arch/powerpc/platforms/86xx/mpc86xx_hpcn.c
+++ b/arch/powerpc/platforms/86xx/mpc86xx_hpcn.c
@@ -96,13 +96,11 @@ mpc86xx_hpcn_show_cpuinfo(struct seq_file *m)
  */
 static int __init mpc86xx_hpcn_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	if (of_flat_dt_is_compatible(root, "fsl,mpc8641hpcn"))
+	if (of_machine_is_compatible("fsl,mpc8641hpcn"))
 		return 1;	/* Looks good */
 
 	/* Be nice and don't give silent boot death.  Delete this in 2.6.27 */
-	if (of_flat_dt_is_compatible(root, "mpc86xx")) {
+	if (of_machine_is_compatible("mpc86xx")) {
 		pr_warning("WARNING: your dts/dtb is old. You must update before the next kernel release\n");
 		return 1;
 	}
diff --git a/arch/powerpc/platforms/86xx/sbc8641d.c b/arch/powerpc/platforms/86xx/sbc8641d.c
index 2a9cf27..52af573 100644
--- a/arch/powerpc/platforms/86xx/sbc8641d.c
+++ b/arch/powerpc/platforms/86xx/sbc8641d.c
@@ -67,9 +67,7 @@ sbc8641_show_cpuinfo(struct seq_file *m)
  */
 static int __init sbc8641_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	if (of_flat_dt_is_compatible(root, "wind,sbc8641"))
+	if (of_machine_is_compatible("wind,sbc8641"))
 		return 1;	/* Looks good */
 
 	return 0;
diff --git a/arch/powerpc/platforms/8xx/adder875.c b/arch/powerpc/platforms/8xx/adder875.c
index 61cae4c..333dece 100644
--- a/arch/powerpc/platforms/8xx/adder875.c
+++ b/arch/powerpc/platforms/8xx/adder875.c
@@ -88,8 +88,7 @@ static void __init adder875_setup(void)
 
 static int __init adder875_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-	return of_flat_dt_is_compatible(root, "analogue-and-micro,adder875");
+	return of_machine_is_compatible("analogue-and-micro,adder875");
 }
 
 static const struct of_device_id of_bus_ids[] __initconst = {
diff --git a/arch/powerpc/platforms/8xx/ep88xc.c b/arch/powerpc/platforms/8xx/ep88xc.c
index 2bedeb7..cd0d90f 100644
--- a/arch/powerpc/platforms/8xx/ep88xc.c
+++ b/arch/powerpc/platforms/8xx/ep88xc.c
@@ -143,8 +143,7 @@ static void __init ep88xc_setup_arch(void)
 
 static int __init ep88xc_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-	return of_flat_dt_is_compatible(root, "fsl,ep88xc");
+	return of_machine_is_compatible("fsl,ep88xc");
 }
 
 static const struct of_device_id of_bus_ids[] __initconst = {
diff --git a/arch/powerpc/platforms/8xx/mpc86xads_setup.c b/arch/powerpc/platforms/8xx/mpc86xads_setup.c
index 78180c5..8d02f5f 100644
--- a/arch/powerpc/platforms/8xx/mpc86xads_setup.c
+++ b/arch/powerpc/platforms/8xx/mpc86xads_setup.c
@@ -118,8 +118,7 @@ static void __init mpc86xads_setup_arch(void)
 
 static int __init mpc86xads_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-	return of_flat_dt_is_compatible(root, "fsl,mpc866ads");
+	return of_machine_is_compatible("fsl,mpc866ads");
 }
 
 static const struct of_device_id of_bus_ids[] __initconst = {
diff --git a/arch/powerpc/platforms/8xx/mpc885ads_setup.c b/arch/powerpc/platforms/8xx/mpc885ads_setup.c
index 4d62bf9..e821a42 100644
--- a/arch/powerpc/platforms/8xx/mpc885ads_setup.c
+++ b/arch/powerpc/platforms/8xx/mpc885ads_setup.c
@@ -193,8 +193,7 @@ static void __init mpc885ads_setup_arch(void)
 
 static int __init mpc885ads_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-	return of_flat_dt_is_compatible(root, "fsl,mpc885ads");
+	return of_machine_is_compatible("fsl,mpc885ads");
 }
 
 static const struct of_device_id of_bus_ids[] __initconst = {
diff --git a/arch/powerpc/platforms/8xx/tqm8xx_setup.c b/arch/powerpc/platforms/8xx/tqm8xx_setup.c
index bee47a2..4cea8b1 100644
--- a/arch/powerpc/platforms/8xx/tqm8xx_setup.c
+++ b/arch/powerpc/platforms/8xx/tqm8xx_setup.c
@@ -119,9 +119,7 @@ static void __init tqm8xx_setup_arch(void)
 
 static int __init tqm8xx_probe(void)
 {
-	unsigned long node = of_get_flat_dt_root();
-
-	return of_flat_dt_is_compatible(node, "tqc,tqm8xx");
+	return of_machine_is_compatible("tqc,tqm8xx");
 }
 
 static const struct of_device_id of_bus_ids[] __initconst = {
diff --git a/arch/powerpc/platforms/amigaone/setup.c b/arch/powerpc/platforms/amigaone/setup.c
index 2fe1204..a030d3b 100644
--- a/arch/powerpc/platforms/amigaone/setup.c
+++ b/arch/powerpc/platforms/amigaone/setup.c
@@ -143,9 +143,7 @@ void amigaone_restart(char *cmd)
 
 static int __init amigaone_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	if (of_flat_dt_is_compatible(root, "eyetech,amigaone")) {
+	if (of_machine_is_compatible("eyetech,amigaone")) {
 		/*
 		 * Coherent memory access cause complete system lockup! Thus
 		 * disable this CPU feature, even if the CPU needs it.
diff --git a/arch/powerpc/platforms/embedded6xx/c2k.c b/arch/powerpc/platforms/embedded6xx/c2k.c
index ebd3963..45197b5 100644
--- a/arch/powerpc/platforms/embedded6xx/c2k.c
+++ b/arch/powerpc/platforms/embedded6xx/c2k.c
@@ -123,9 +123,7 @@ void c2k_show_cpuinfo(struct seq_file *m)
  */
 static int __init c2k_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	if (!of_flat_dt_is_compatible(root, "GEFanuc,C2K"))
+	if (!of_machine_is_compatible("GEFanuc,C2K"))
 		return 0;
 
 	printk(KERN_INFO "Detected a GEFanuc C2K board\n");
diff --git a/arch/powerpc/platforms/embedded6xx/gamecube.c b/arch/powerpc/platforms/embedded6xx/gamecube.c
index fe0ed6e..defcef8 100644
--- a/arch/powerpc/platforms/embedded6xx/gamecube.c
+++ b/arch/powerpc/platforms/embedded6xx/gamecube.c
@@ -61,10 +61,7 @@ static void __init gamecube_init_early(void)
 
 static int __init gamecube_probe(void)
 {
-	unsigned long dt_root;
-
-	dt_root = of_get_flat_dt_root();
-	if (!of_flat_dt_is_compatible(dt_root, "nintendo,gamecube"))
+	if (!of_machine_is_compatible("nintendo,gamecube"))
 		return 0;
 
 	pm_power_off = gamecube_power_off;
diff --git a/arch/powerpc/platforms/embedded6xx/holly.c b/arch/powerpc/platforms/embedded6xx/holly.c
index 8c305c7..a7a0ef9 100644
--- a/arch/powerpc/platforms/embedded6xx/holly.c
+++ b/arch/powerpc/platforms/embedded6xx/holly.c
@@ -250,9 +250,7 @@ void holly_halt(void)
  */
 static int __init holly_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	if (!of_flat_dt_is_compatible(root, "ibm,holly"))
+	if (!of_machine_is_compatible("ibm,holly"))
 		return 0;
 	return 1;
 }
diff --git a/arch/powerpc/platforms/embedded6xx/linkstation.c b/arch/powerpc/platforms/embedded6xx/linkstation.c
index 540eeb5..1032655 100644
--- a/arch/powerpc/platforms/embedded6xx/linkstation.c
+++ b/arch/powerpc/platforms/embedded6xx/linkstation.c
@@ -141,11 +141,7 @@ static void linkstation_show_cpuinfo(struct seq_file *m)
 
 static int __init linkstation_probe(void)
 {
-	unsigned long root;
-
-	root = of_get_flat_dt_root();
-
-	if (!of_flat_dt_is_compatible(root, "linkstation"))
+	if (!of_machine_is_compatible("linkstation"))
 		return 0;
 
 	pm_power_off = linkstation_power_off;
diff --git a/arch/powerpc/platforms/embedded6xx/mpc7448_hpc2.c b/arch/powerpc/platforms/embedded6xx/mpc7448_hpc2.c
index df4ad95..eb5bfcf 100644
--- a/arch/powerpc/platforms/embedded6xx/mpc7448_hpc2.c
+++ b/arch/powerpc/platforms/embedded6xx/mpc7448_hpc2.c
@@ -161,9 +161,7 @@ void mpc7448_hpc2_restart(char *cmd)
  */
 static int __init mpc7448_hpc2_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	if (!of_flat_dt_is_compatible(root, "mpc74xx"))
+	if (!of_machine_is_compatible("mpc74xx"))
 		return 0;
 	return 1;
 }
diff --git a/arch/powerpc/platforms/embedded6xx/mvme5100.c b/arch/powerpc/platforms/embedded6xx/mvme5100.c
index 8f65aa3..b5ae04e 100644
--- a/arch/powerpc/platforms/embedded6xx/mvme5100.c
+++ b/arch/powerpc/platforms/embedded6xx/mvme5100.c
@@ -194,9 +194,7 @@ static void mvme5100_restart(char *cmd)
  */
 static int __init mvme5100_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	return of_flat_dt_is_compatible(root, "MVME5100");
+	return of_machine_is_compatible("MVME5100");
 }
 
 static int __init probe_of_platform_devices(void)
diff --git a/arch/powerpc/platforms/embedded6xx/storcenter.c b/arch/powerpc/platforms/embedded6xx/storcenter.c
index d572833..20a2f91 100644
--- a/arch/powerpc/platforms/embedded6xx/storcenter.c
+++ b/arch/powerpc/platforms/embedded6xx/storcenter.c
@@ -109,9 +109,7 @@ static void storcenter_restart(char *cmd)
 
 static int __init storcenter_probe(void)
 {
-	unsigned long root = of_get_flat_dt_root();
-
-	return of_flat_dt_is_compatible(root, "iomega,storcenter");
+	return of_machine_is_compatible("iomega,storcenter");
 }
 
 define_machine(storcenter){
diff --git a/arch/powerpc/platforms/embedded6xx/wii.c b/arch/powerpc/platforms/embedded6xx/wii.c
index 352592d..cc65d68 100644
--- a/arch/powerpc/platforms/embedded6xx/wii.c
+++ b/arch/powerpc/platforms/embedded6xx/wii.c
@@ -205,10 +205,7 @@ static void __init wii_pic_probe(void)
 
 static int __init wii_probe(void)
 {
-	unsigned long dt_root;
-
-	dt_root = of_get_flat_dt_root();
-	if (!of_flat_dt_is_compatible(dt_root, "nintendo,wii"))
+	if (!of_machine_is_compatible("nintendo,wii"))
 		return 0;
 
 	pm_power_off = wii_power_off;
diff --git a/arch/powerpc/platforms/powermac/setup.c b/arch/powerpc/platforms/powermac/setup.c
index 128ce76..509b18f 100644
--- a/arch/powerpc/platforms/powermac/setup.c
+++ b/arch/powerpc/platforms/powermac/setup.c
@@ -588,17 +588,9 @@ console_initcall(check_pmac_serial_console);
  */
 static int __init pmac_probe(void)
 {
-#ifdef CONFIG_PPC32
-	unsigned long root = of_get_flat_dt_root();
-
-	if (!of_flat_dt_is_compatible(root, "Power Macintosh") &&
-	    !of_flat_dt_is_compatible(root, "MacRISC"))
-		return 0;
-#else
 	if (!of_machine_is_compatible("Power Macintosh") &&
 	    !of_machine_is_compatible("MacRISC"))
 		return 0;
-#endif
 
 #ifdef CONFIG_PPC32
 	/* isa_io_base gets set in pmac_pci_init */
-- 
2.7.4

[PATCH 30/38] powerpc: Get rid of ppc_md.init_early()

[Benjamin Herrenschmidt]

It is now called right after platform probe, so the probe function
can just do the job.

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
---
 arch/powerpc/include/asm/machdep.h            |  1 -
 arch/powerpc/kernel/setup_32.c                |  3 ---
 arch/powerpc/kernel/setup_64.c                |  8 --------
 arch/powerpc/platforms/512x/mpc5121_ads.c     |  8 ++++++--
 arch/powerpc/platforms/512x/mpc512x_generic.c |  8 ++++++--
 arch/powerpc/platforms/512x/pdm360ng.c        |  8 ++++++--
 arch/powerpc/platforms/chrp/setup.c           |  5 +++--
 arch/powerpc/platforms/embedded6xx/c2k.c      |  4 +++-
 arch/powerpc/platforms/embedded6xx/gamecube.c |  8 ++------
 arch/powerpc/platforms/embedded6xx/wii.c      |  8 ++------
 arch/powerpc/platforms/maple/setup.c          | 15 ++-------------
 arch/powerpc/platforms/pasemi/setup.c         |  8 ++------
 arch/powerpc/platforms/powermac/setup.c       |  5 +++--
 arch/powerpc/platforms/powernv/setup.c        |  5 +++--
 arch/powerpc/platforms/pseries/setup.c        |  9 +++++----
 15 files changed, 43 insertions(+), 60 deletions(-)

diff --git a/arch/powerpc/include/asm/machdep.h b/arch/powerpc/include/asm/machdep.h
index 62b1461..29bc9a3 100644
--- a/arch/powerpc/include/asm/machdep.h
+++ b/arch/powerpc/include/asm/machdep.h
@@ -53,7 +53,6 @@ struct machdep_calls {
 
 	int		(*probe)(void);
 	void		(*setup_arch)(void); /* Optional, may be NULL */
-	void		(*init_early)(void);
 	/* Optional, may be NULL. */
 	void		(*show_cpuinfo)(struct seq_file *m);
 	void		(*show_percpuinfo)(struct seq_file *m, int i);
diff --git a/arch/powerpc/kernel/setup_32.c b/arch/powerpc/kernel/setup_32.c
index e7bb4e7..22347e87 100644
--- a/arch/powerpc/kernel/setup_32.c
+++ b/arch/powerpc/kernel/setup_32.c
@@ -263,9 +263,6 @@ void __init setup_arch(char **cmdline_p)
 
 	setup_power_save();
 
-	if (ppc_md.init_early)
-		ppc_md.init_early();
-
 	find_legacy_serial_ports();
 
 	smp_setup_cpu_maps();
diff --git a/arch/powerpc/kernel/setup_64.c b/arch/powerpc/kernel/setup_64.c
index 1f20f99..afb22a6 100644
--- a/arch/powerpc/kernel/setup_64.c
+++ b/arch/powerpc/kernel/setup_64.c
@@ -522,14 +522,6 @@ void __init setup_system(void)
 	/* Probe the machine type */
 	probe_machine();
 
-	/*
-	 * Do some platform specific early initializations, that includes
-	 * setting up the hash table pointers. It also sets up some interrupt-mapping
-	 * related options that will be used by finish_device_tree()
-	 */
-	if (ppc_md.init_early)
-		ppc_md.init_early();
-
  	/*
 	 * We can discover serial ports now since the above did setup the
 	 * hash table management for us, thus ioremap works. We do that early
diff --git a/arch/powerpc/platforms/512x/mpc5121_ads.c b/arch/powerpc/platforms/512x/mpc5121_ads.c
index 4e03f04..f65d503 100644
--- a/arch/powerpc/platforms/512x/mpc5121_ads.c
+++ b/arch/powerpc/platforms/512x/mpc5121_ads.c
@@ -57,7 +57,12 @@ static void __init mpc5121_ads_init_IRQ(void)
  */
 static int __init mpc5121_ads_probe(void)
 {
-	return of_machine_is_compatible("fsl,mpc5121ads");
+	if (!of_machine_is_compatible("fsl,mpc5121ads"))
+		return 0;
+
+	mpc512x_init_early();
+
+	return 1;
 }
 
 define_machine(mpc5121_ads) {
@@ -65,7 +70,6 @@ define_machine(mpc5121_ads) {
 	.probe			= mpc5121_ads_probe,
 	.setup_arch		= mpc5121_ads_setup_arch,
 	.init			= mpc512x_init,
-	.init_early		= mpc512x_init_early,
 	.init_IRQ		= mpc5121_ads_init_IRQ,
 	.get_irq		= ipic_get_irq,
 	.calibrate_decr		= generic_calibrate_decr,
diff --git a/arch/powerpc/platforms/512x/mpc512x_generic.c b/arch/powerpc/platforms/512x/mpc512x_generic.c
index 87eba17..bf884d3 100644
--- a/arch/powerpc/platforms/512x/mpc512x_generic.c
+++ b/arch/powerpc/platforms/512x/mpc512x_generic.c
@@ -38,14 +38,18 @@ static const char * const board[] __initconst = {
  */
 static int __init mpc512x_generic_probe(void)
 {
-	return of_device_compatible_match(of_root, board);
+	if (!of_device_compatible_match(of_root, board))
+		return 0;
+
+	mpc512x_init_early();
+
+	return 1;
 }
 
 define_machine(mpc512x_generic) {
 	.name			= "MPC512x generic",
 	.probe			= mpc512x_generic_probe,
 	.init			= mpc512x_init,
-	.init_early		= mpc512x_init_early,
 	.setup_arch		= mpc512x_setup_arch,
 	.init_IRQ		= mpc512x_init_IRQ,
 	.get_irq		= ipic_get_irq,
diff --git a/arch/powerpc/platforms/512x/pdm360ng.c b/arch/powerpc/platforms/512x/pdm360ng.c
index f9cad19..dc81f05 100644
--- a/arch/powerpc/platforms/512x/pdm360ng.c
+++ b/arch/powerpc/platforms/512x/pdm360ng.c
@@ -113,7 +113,12 @@ void __init pdm360ng_init(void)
 
 static int __init pdm360ng_probe(void)
 {
-	return of_machine_is_compatible("ifm,pdm360ng");
+	if (!of_machine_is_compatible("ifm,pdm360ng"))
+		return 0;
+
+	mpc512x_init_early();
+
+	return 1;
 }
 
 define_machine(pdm360ng) {
@@ -121,7 +126,6 @@ define_machine(pdm360ng) {
 	.probe			= pdm360ng_probe,
 	.setup_arch		= mpc512x_setup_arch,
 	.init			= pdm360ng_init,
-	.init_early		= mpc512x_init_early,
 	.init_IRQ		= mpc512x_init_IRQ,
 	.get_irq		= ipic_get_irq,
 	.calibrate_decr		= generic_calibrate_decr,
diff --git a/arch/powerpc/platforms/chrp/setup.c b/arch/powerpc/platforms/chrp/setup.c
index 987d1b8..4974fc3 100644
--- a/arch/powerpc/platforms/chrp/setup.c
+++ b/arch/powerpc/platforms/chrp/setup.c
@@ -253,7 +253,7 @@ static void briq_restart(char *cmd)
  * But unfortunately, the firmware does not connect /chosen/{stdin,stdout}
  * the the built-in serial node. Instead, a /failsafe node is created.
  */
-static __init void chrp_init_early(void)
+static __init void chrp_init(void)
 {
 	struct device_node *node;
 	const char *property;
@@ -587,6 +587,8 @@ static int __init chrp_probe(void)
 
 	pm_power_off = rtas_power_off;
 
+	chrp_init();
+
 	return 1;
 }
 
@@ -595,7 +597,6 @@ define_machine(chrp) {
 	.probe			= chrp_probe,
 	.setup_arch		= chrp_setup_arch,
 	.init			= chrp_init2,
-	.init_early		= chrp_init_early,
 	.show_cpuinfo		= chrp_show_cpuinfo,
 	.init_IRQ		= chrp_init_IRQ,
 	.restart		= rtas_restart,
diff --git a/arch/powerpc/platforms/embedded6xx/c2k.c b/arch/powerpc/platforms/embedded6xx/c2k.c
index 45197b5..10a6445 100644
--- a/arch/powerpc/platforms/embedded6xx/c2k.c
+++ b/arch/powerpc/platforms/embedded6xx/c2k.c
@@ -130,6 +130,9 @@ static int __init c2k_probe(void)
 
 	_set_L2CR(0);
 	_set_L2CR(L2CR_L2E | L2CR_L2PE | L2CR_L2I);
+
+	mv64x60_init_early();
+
 	return 1;
 }
 
@@ -137,7 +140,6 @@ define_machine(c2k) {
 	.name			= "C2K",
 	.probe			= c2k_probe,
 	.setup_arch		= c2k_setup_arch,
-	.init_early		= mv64x60_init_early,
 	.show_cpuinfo		= c2k_show_cpuinfo,
 	.init_IRQ		= mv64x60_init_irq,
 	.get_irq		= mv64x60_get_irq,
diff --git a/arch/powerpc/platforms/embedded6xx/gamecube.c b/arch/powerpc/platforms/embedded6xx/gamecube.c
index defcef8..18a2ba3 100644
--- a/arch/powerpc/platforms/embedded6xx/gamecube.c
+++ b/arch/powerpc/platforms/embedded6xx/gamecube.c
@@ -54,11 +54,6 @@ static void gamecube_halt(void)
 	gamecube_restart(NULL);
 }
 
-static void __init gamecube_init_early(void)
-{
-	ug_udbg_init();
-}
-
 static int __init gamecube_probe(void)
 {
 	if (!of_machine_is_compatible("nintendo,gamecube"))
@@ -66,6 +61,8 @@ static int __init gamecube_probe(void)
 
 	pm_power_off = gamecube_power_off;
 
+	ug_udbg_init();
+
 	return 1;
 }
 
@@ -77,7 +74,6 @@ static void gamecube_shutdown(void)
 define_machine(gamecube) {
 	.name			= "gamecube",
 	.probe			= gamecube_probe,
-	.init_early		= gamecube_init_early,
 	.restart		= gamecube_restart,
 	.halt			= gamecube_halt,
 	.init_IRQ		= flipper_pic_probe,
diff --git a/arch/powerpc/platforms/embedded6xx/wii.c b/arch/powerpc/platforms/embedded6xx/wii.c
index cc65d68..9676b09 100644
--- a/arch/powerpc/platforms/embedded6xx/wii.c
+++ b/arch/powerpc/platforms/embedded6xx/wii.c
@@ -192,11 +192,6 @@ static void wii_halt(void)
 	wii_spin();
 }
 
-static void __init wii_init_early(void)
-{
-	ug_udbg_init();
-}
-
 static void __init wii_pic_probe(void)
 {
 	flipper_pic_probe();
@@ -210,6 +205,8 @@ static int __init wii_probe(void)
 
 	pm_power_off = wii_power_off;
 
+	ug_udbg_init();
+
 	return 1;
 }
 
@@ -222,7 +219,6 @@ static void wii_shutdown(void)
 define_machine(wii) {
 	.name			= "wii",
 	.probe			= wii_probe,
-	.init_early		= wii_init_early,
 	.setup_arch		= wii_setup_arch,
 	.restart		= wii_restart,
 	.halt			= wii_halt,
diff --git a/arch/powerpc/platforms/maple/setup.c b/arch/powerpc/platforms/maple/setup.c
index a1ecbc9..98ae9f5 100644
--- a/arch/powerpc/platforms/maple/setup.c
+++ b/arch/powerpc/platforms/maple/setup.c
@@ -196,18 +196,6 @@ void __init maple_setup_arch(void)
 	mmio_nvram_init();
 }
 
-/* 
- * Early initialization.
- */
-static void __init maple_init_early(void)
-{
-	DBG(" -> maple_init_early\n");
-
-	iommu_init_early_dart(&maple_pci_controller_ops);
-
-	DBG(" <- maple_init_early\n");
-}
-
 /*
  * This is almost identical to pSeries and CHRP. We need to make that
  * code generic at one point, with appropriate bits in the device-tree to
@@ -304,6 +292,8 @@ static int __init maple_probe(void)
 
 	pm_power_off = maple_power_off;
 
+	iommu_init_early_dart(&maple_pci_controller_ops);
+
 	return 1;
 }
 
@@ -311,7 +301,6 @@ define_machine(maple) {
 	.name			= "Maple",
 	.probe			= maple_probe,
 	.setup_arch		= maple_setup_arch,
-	.init_early		= maple_init_early,
 	.init_IRQ		= maple_init_IRQ,
 	.pci_irq_fixup		= maple_pci_irq_fixup,
 	.pci_get_legacy_ide_irq	= maple_pci_get_legacy_ide_irq,
diff --git a/arch/powerpc/platforms/pasemi/setup.c b/arch/powerpc/platforms/pasemi/setup.c
index 8f5e291..e7ad8a4 100644
--- a/arch/powerpc/platforms/pasemi/setup.c
+++ b/arch/powerpc/platforms/pasemi/setup.c
@@ -339,11 +339,6 @@ out:
 	return !!(srr1 & 0x2);
 }
 
-static void __init pas_init_early(void)
-{
-	iommu_init_early_pasemi();
-}
-
 #ifdef CONFIG_PCMCIA
 static int pcmcia_notify(struct notifier_block *nb, unsigned long action,
 			 void *data)
@@ -424,6 +419,8 @@ static int __init pas_probe(void)
 	    !of_machine_is_compatible("pasemi,pwrficient"))
 		return 0;
 
+	iommu_init_early_pasemi();
+
 	return 1;
 }
 
@@ -431,7 +428,6 @@ define_machine(pasemi) {
 	.name			= "PA Semi PWRficient",
 	.probe			= pas_probe,
 	.setup_arch		= pas_setup_arch,
-	.init_early		= pas_init_early,
 	.init_IRQ		= pas_init_IRQ,
 	.get_irq		= mpic_get_irq,
 	.restart		= pas_restart,
diff --git a/arch/powerpc/platforms/powermac/setup.c b/arch/powerpc/platforms/powermac/setup.c
index 509b18f..3830582 100644
--- a/arch/powerpc/platforms/powermac/setup.c
+++ b/arch/powerpc/platforms/powermac/setup.c
@@ -450,7 +450,7 @@ pmac_halt(void)
 /* 
  * Early initialization.
  */
-static void __init pmac_init_early(void)
+static void __init pmac_init(void)
 {
 	/* Enable early btext debug if requested */
 	if (strstr(boot_command_line, "btextdbg")) {
@@ -601,6 +601,8 @@ static int __init pmac_probe(void)
 
 	pm_power_off = pmac_power_off;
 
+	pmac_init();
+
 	return 1;
 }
 
@@ -608,7 +610,6 @@ define_machine(powermac) {
 	.name			= "PowerMac",
 	.probe			= pmac_probe,
 	.setup_arch		= pmac_setup_arch,
-	.init_early		= pmac_init_early,
 	.show_cpuinfo		= pmac_show_cpuinfo,
 	.init_IRQ		= pmac_pic_init,
 	.get_irq		= NULL,	/* changed later */
diff --git a/arch/powerpc/platforms/powernv/setup.c b/arch/powerpc/platforms/powernv/setup.c
index ed99177..e8f7782 100644
--- a/arch/powerpc/platforms/powernv/setup.c
+++ b/arch/powerpc/platforms/powernv/setup.c
@@ -58,7 +58,7 @@ static void __init pnv_setup_arch(void)
 	/* XXX PMCS */
 }
 
-static void __init pnv_init_early(void)
+static void __init pnv_init(void)
 {
 	/*
 	 * Initialize the LPC bus now so that legacy serial
@@ -276,6 +276,8 @@ static int __init pnv_probe(void)
 
 	pr_debug("PowerNV detected !\n");
 
+	pnv_init();
+
 	return 1;
 }
 
@@ -301,7 +303,6 @@ static unsigned long pnv_get_proc_freq(unsigned int cpu)
 define_machine(powernv) {
 	.name			= "PowerNV",
 	.probe			= pnv_probe,
-	.init_early		= pnv_init_early,
 	.setup_arch		= pnv_setup_arch,
 	.init_IRQ		= pnv_init_IRQ,
 	.show_cpuinfo		= pnv_show_cpuinfo,
diff --git a/arch/powerpc/platforms/pseries/setup.c b/arch/powerpc/platforms/pseries/setup.c
index eacfba7..657e85b 100644
--- a/arch/powerpc/platforms/pseries/setup.c
+++ b/arch/powerpc/platforms/pseries/setup.c
@@ -670,9 +670,9 @@ static void pSeries_cmo_feature_init(void)
 /*
  * Early initialization.  Relocation is on but do not reference unbolted pages
  */
-static void __init pSeries_init_early(void)
+static void __init pSeries_init(void)
 {
-	pr_debug(" -> pSeries_init_early()\n");
+	pr_debug(" -> pSeries_init()\n");
 
 #ifdef CONFIG_HVC_CONSOLE
 	if (firmware_has_feature(FW_FEATURE_LPAR))
@@ -689,7 +689,7 @@ static void __init pSeries_init_early(void)
 	pSeries_cmo_feature_init();
 	iommu_init_early_pSeries();
 
-	pr_debug(" <- pSeries_init_early()\n");
+	pr_debug(" <- pSeries_init()\n");
 }
 
 /**
@@ -741,6 +741,8 @@ static int __init pSeries_probe(void)
 	pr_debug("Machine is%s LPAR !\n",
 	         (powerpc_firmware_features & FW_FEATURE_LPAR) ? "" : " not");
 
+	pSeries_init();
+
 	return 1;
 }
 
@@ -759,7 +761,6 @@ define_machine(pseries) {
 	.name			= "pSeries",
 	.probe			= pSeries_probe,
 	.setup_arch		= pSeries_setup_arch,
-	.init_early		= pSeries_init_early,
 	.show_cpuinfo		= pSeries_show_cpuinfo,
 	.log_error		= pSeries_log_error,
 	.pcibios_fixup		= pSeries_final_fixup,
-- 
2.7.4

[PATCH 31/38] powerpc/64: Move the boot time info banner to a separate function

[Benjamin Herrenschmidt]

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
---
 arch/powerpc/kernel/setup_64.c | 66 ++++++++++++++++++++++--------------------
 1 file changed, 35 insertions(+), 31 deletions(-)

diff --git a/arch/powerpc/kernel/setup_64.c b/arch/powerpc/kernel/setup_64.c
index afb22a6..8c8f440 100644
--- a/arch/powerpc/kernel/setup_64.c
+++ b/arch/powerpc/kernel/setup_64.c
@@ -487,6 +487,37 @@ static void __init initialize_cache_info(void)
 	DBG(" <- initialize_cache_info()\n");
 }
 
+static __init void print_system_info(void)
+{
+	pr_info("-----------------------------------------------------\n");
+	pr_info("ppc64_pft_size    = 0x%llx\n", ppc64_pft_size);
+	pr_info("phys_mem_size     = 0x%llx\n", memblock_phys_mem_size());
+
+	if (ppc64_caches.dline_size != 0x80)
+		pr_info("dcache_line_size  = 0x%x\n", ppc64_caches.dline_size);
+	if (ppc64_caches.iline_size != 0x80)
+		pr_info("icache_line_size  = 0x%x\n", ppc64_caches.iline_size);
+
+	pr_info("cpu_features      = 0x%016lx\n", cur_cpu_spec->cpu_features);
+	pr_info("  possible        = 0x%016lx\n", CPU_FTRS_POSSIBLE);
+	pr_info("  always          = 0x%016lx\n", CPU_FTRS_ALWAYS);
+	pr_info("cpu_user_features = 0x%08x 0x%08x\n", cur_cpu_spec->cpu_user_features,
+		cur_cpu_spec->cpu_user_features2);
+	pr_info("mmu_features      = 0x%08x\n", cur_cpu_spec->mmu_features);
+	pr_info("firmware_features = 0x%016lx\n", powerpc_firmware_features);
+
+#ifdef CONFIG_PPC_STD_MMU_64
+	if (htab_address)
+		pr_info("htab_address      = 0x%p\n", htab_address);
+
+	pr_info("htab_hash_mask    = 0x%lx\n", htab_hash_mask);
+#endif
+
+	if (PHYSICAL_START > 0)
+		pr_info("physical_start    = 0x%llx\n",
+		       (unsigned long long)PHYSICAL_START);
+	pr_info("-----------------------------------------------------\n");
+}
 
 /*
  * Do some initial setup of the system.  The parameters are those which 
@@ -554,37 +585,10 @@ void __init setup_system(void)
 	smp_release_cpus();
 #endif
 
-	pr_info("Starting Linux %s %s\n", init_utsname()->machine,
-		 init_utsname()->version);
-
-	pr_info("-----------------------------------------------------\n");
-	pr_info("ppc64_pft_size    = 0x%llx\n", ppc64_pft_size);
-	pr_info("phys_mem_size     = 0x%llx\n", memblock_phys_mem_size());
-
-	if (ppc64_caches.dline_size != 0x80)
-		pr_info("dcache_line_size  = 0x%x\n", ppc64_caches.dline_size);
-	if (ppc64_caches.iline_size != 0x80)
-		pr_info("icache_line_size  = 0x%x\n", ppc64_caches.iline_size);
-
-	pr_info("cpu_features      = 0x%016lx\n", cur_cpu_spec->cpu_features);
-	pr_info("  possible        = 0x%016lx\n", CPU_FTRS_POSSIBLE);
-	pr_info("  always          = 0x%016lx\n", CPU_FTRS_ALWAYS);
-	pr_info("cpu_user_features = 0x%08x 0x%08x\n", cur_cpu_spec->cpu_user_features,
-		cur_cpu_spec->cpu_user_features2);
-	pr_info("mmu_features      = 0x%08x\n", cur_cpu_spec->mmu_features);
-	pr_info("firmware_features = 0x%016lx\n", powerpc_firmware_features);
-
-#ifdef CONFIG_PPC_STD_MMU_64
-	if (htab_address)
-		pr_info("htab_address      = 0x%p\n", htab_address);
-
-	pr_info("htab_hash_mask    = 0x%lx\n", htab_hash_mask);
-#endif
-
-	if (PHYSICAL_START > 0)
-		pr_info("physical_start    = 0x%llx\n",
-		       (unsigned long long)PHYSICAL_START);
-	pr_info("-----------------------------------------------------\n");
+	/* Print various info about the machine that gave been gathered
+	 * so far
+	 */
+	print_system_info();
 
 	DBG(" <- setup_system()\n");
 }
-- 
2.7.4

[PATCH 32/38] powerpc/64: Move setting of {i, d}cache_bsize to initialize_cache_info()

[Benjamin Herrenschmidt]

Also remove the completely osbolete comment. We *do* look in the
device-tree.

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
---
 arch/powerpc/kernel/setup_64.c | 13 ++++---------
 1 file changed, 4 insertions(+), 9 deletions(-)

diff --git a/arch/powerpc/kernel/setup_64.c b/arch/powerpc/kernel/setup_64.c
index 8c8f440..29bbe8c 100644
--- a/arch/powerpc/kernel/setup_64.c
+++ b/arch/powerpc/kernel/setup_64.c
@@ -484,6 +484,10 @@ static void __init initialize_cache_info(void)
 		}
 	}
 
+	/* For use by binfmt_elf */
+	dcache_bsize = ppc64_caches.dline_size;
+	icache_bsize = ppc64_caches.iline_size;
+
 	DBG(" <- initialize_cache_info()\n");
 }
 
@@ -704,15 +708,6 @@ void __init setup_arch(char **cmdline_p)
 {
 	*cmdline_p = boot_command_line;
 
-	/*
-	 * Set cache line size based on type of cpu as a default.
-	 * Systems with OF can look in the properties on the cpu node(s)
-	 * for a possibly more accurate value.
-	 */
-	dcache_bsize = ppc64_caches.dline_size;
-	icache_bsize = ppc64_caches.iline_size;
-
-
 	/* Reserve large chunks of memory for use by CMA for KVM */
 	kvm_cma_reserve();
 
-- 
2.7.4

[PATCH 33/38] powerpc/64: Move the content of setup_system() to setup_arch()

[Benjamin Herrenschmidt]

And kill setup_system()

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
---
 arch/powerpc/kernel/head_64.S  |   5 +-
 arch/powerpc/kernel/setup_64.c | 137 +++++++++++++++++++----------------------
 2 files changed, 64 insertions(+), 78 deletions(-)

diff --git a/arch/powerpc/kernel/head_64.S b/arch/powerpc/kernel/head_64.S
index 2d14774..3109d0c 100644
--- a/arch/powerpc/kernel/head_64.S
+++ b/arch/powerpc/kernel/head_64.S
@@ -941,7 +941,7 @@ start_here_multiplatform:
 	mtspr	SPRN_SRR1,r4
 	RFI
 	b	.	/* prevent speculative execution */
-	
+
 	/* This is where all platforms converge execution */
 
 start_here_common:
@@ -951,9 +951,6 @@ start_here_common:
 	/* Load the TOC (virtual address) */
 	ld	r2,PACATOC(r13)
 
-	/* Do more system initializations in virtual mode */
-	bl	setup_system
-
 	/* Mark interrupts soft and hard disabled (they might be enabled
 	 * in the PACA when doing hotplug)
 	 */
diff --git a/arch/powerpc/kernel/setup_64.c b/arch/powerpc/kernel/setup_64.c
index 29bbe8c..108035b 100644
--- a/arch/powerpc/kernel/setup_64.c
+++ b/arch/powerpc/kernel/setup_64.c
@@ -523,80 +523,6 @@ static __init void print_system_info(void)
 	pr_info("-----------------------------------------------------\n");
 }
 
-/*
- * Do some initial setup of the system.  The parameters are those which 
- * were passed in from the bootloader.
- */
-void __init setup_system(void)
-{
-	DBG(" -> setup_system()\n");
-
-	/*
-	 * Unflatten the device-tree passed by prom_init or kexec
-	 */
-	unflatten_device_tree();
-
-	/*
-	 * Fill the ppc64_caches & systemcfg structures with informations
- 	 * retrieved from the device-tree.
-	 */
-	initialize_cache_info();
-
-#ifdef CONFIG_PPC_RTAS
-	/*
-	 * Initialize RTAS if available
-	 */
-	rtas_initialize();
-#endif /* CONFIG_PPC_RTAS */
-
-	/*
-	 * Check if we have an initrd provided via the device-tree
-	 */
-	check_for_initrd();
-
-	/* Probe the machine type */
-	probe_machine();
-
- 	/*
-	 * We can discover serial ports now since the above did setup the
-	 * hash table management for us, thus ioremap works. We do that early
-	 * so that further code can be debugged
-	 */
-	find_legacy_serial_ports();
-
-	/*
-	 * Register early console
-	 */
-	register_early_udbg_console();
-
-	/*
-	 * Initialize xmon
-	 */
-	xmon_setup();
-
-	smp_setup_cpu_maps();
-	check_smt_enabled();
-	setup_tlb_core_data();
-
-	/*
-	 * Freescale Book3e parts spin in a loop provided by firmware,
-	 * so smp_release_cpus() does nothing for them
-	 */
-#if defined(CONFIG_SMP)
-	/* Release secondary cpus out of their spinloops at 0x60 now that
-	 * we can map physical -> logical CPU ids
-	 */
-	smp_release_cpus();
-#endif
-
-	/* Print various info about the machine that gave been gathered
-	 * so far
-	 */
-	print_system_info();
-
-	DBG(" <- setup_system()\n");
-}
-
 /* This returns the limit below which memory accesses to the linear
  * mapping are guarnateed not to cause a TLB or SLB miss. This is
  * used to allocate interrupt or emergency stacks for which our
@@ -708,6 +634,69 @@ void __init setup_arch(char **cmdline_p)
 {
 	*cmdline_p = boot_command_line;
 
+	/*
+	 * Unflatten the device-tree passed by prom_init or kexec
+	 */
+	unflatten_device_tree();
+
+	/*
+	 * Fill the ppc64_caches & systemcfg structures with informations
+ 	 * retrieved from the device-tree.
+	 */
+	initialize_cache_info();
+
+#ifdef CONFIG_PPC_RTAS
+	/*
+	 * Initialize RTAS if available
+	 */
+	rtas_initialize();
+#endif /* CONFIG_PPC_RTAS */
+
+	/*
+	 * Check if we have an initrd provided via the device-tree
+	 */
+	check_for_initrd();
+
+	/* Probe the machine type */
+	probe_machine();
+
+ 	/*
+	 * We can discover serial ports now since the above did setup the
+	 * hash table management for us, thus ioremap works. We do that early
+	 * so that further code can be debugged
+	 */
+	find_legacy_serial_ports();
+
+	/*
+	 * Register early console
+	 */
+	register_early_udbg_console();
+
+	/*
+	 * Initialize xmon
+	 */
+	xmon_setup();
+
+	smp_setup_cpu_maps();
+	check_smt_enabled();
+	setup_tlb_core_data();
+
+	/*
+	 * Freescale Book3e parts spin in a loop provided by firmware,
+	 * so smp_release_cpus() does nothing for them
+	 */
+#if defined(CONFIG_SMP)
+	/* Release secondary cpus out of their spinloops at 0x60 now that
+	 * we can map physical -> logical CPU ids
+	 */
+	smp_release_cpus();
+#endif
+
+	/* Print various info about the machine that gave been gathered
+	 * so far
+	 */
+	print_system_info();
+
 	/* Reserve large chunks of memory for use by CMA for KVM */
 	kvm_cma_reserve();
 
-- 
2.7.4

[PATCH 34/38] powerpc/32: Move cache info inits to a separate function

[Benjamin Herrenschmidt]

Matches 64-bit. Also move the call to the same spot as ppc64

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
---
 arch/powerpc/kernel/setup_32.c | 27 ++++++++++++++++-----------
 1 file changed, 16 insertions(+), 11 deletions(-)

diff --git a/arch/powerpc/kernel/setup_32.c b/arch/powerpc/kernel/setup_32.c
index 22347e87..5457911 100644
--- a/arch/powerpc/kernel/setup_32.c
+++ b/arch/powerpc/kernel/setup_32.c
@@ -248,6 +248,21 @@ static void setup_power_save(void)
 #endif
 }
 
+static __init void initialize_cache_info(void)
+{
+	/*
+	 * Set cache line size based on type of cpu as a default.
+	 * Systems with OF can look in the properties on the cpu node(s)
+	 * for a possibly more accurate value.
+	 */
+	dcache_bsize = cur_cpu_spec->dcache_bsize;
+	icache_bsize = cur_cpu_spec->icache_bsize;
+	ucache_bsize = 0;
+	if (cpu_has_feature(CPU_FTR_UNIFIED_ID_CACHE))
+		ucache_bsize = icache_bsize = dcache_bsize;
+}
+
+
 /* Warning, IO base is not yet inited */
 void __init setup_arch(char **cmdline_p)
 {
@@ -257,6 +272,7 @@ void __init setup_arch(char **cmdline_p)
 	loops_per_jiffy = 500000000 / HZ;
 
 	unflatten_device_tree();
+	initialize_cache_info();
 	check_for_initrd();
 
 	probe_machine();
@@ -272,17 +288,6 @@ void __init setup_arch(char **cmdline_p)
 
 	xmon_setup();
 
-	/*
-	 * Set cache line size based on type of cpu as a default.
-	 * Systems with OF can look in the properties on the cpu node(s)
-	 * for a possibly more accurate value.
-	 */
-	dcache_bsize = cur_cpu_spec->dcache_bsize;
-	icache_bsize = cur_cpu_spec->icache_bsize;
-	ucache_bsize = 0;
-	if (cpu_has_feature(CPU_FTR_UNIFIED_ID_CACHE))
-		ucache_bsize = icache_bsize = dcache_bsize;
-
 	if (ppc_md.panic)
 		setup_panic();
 
-- 
2.7.4

[PATCH 35/38] powerpc: Re-order the call to smp_setup_cpu_maps()

[Benjamin Herrenschmidt]

It makes more sense to do it before intializing xmon() as xmon might
use the info in there. We do want to register the console early
though in case we want some functioning printk's in the cpu map setup.

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
---
 arch/powerpc/kernel/setup_32.c | 4 ++--
 arch/powerpc/kernel/setup_64.c | 3 ++-
 2 files changed, 4 insertions(+), 3 deletions(-)

diff --git a/arch/powerpc/kernel/setup_32.c b/arch/powerpc/kernel/setup_32.c
index 5457911..58674b6 100644
--- a/arch/powerpc/kernel/setup_32.c
+++ b/arch/powerpc/kernel/setup_32.c
@@ -281,11 +281,11 @@ void __init setup_arch(char **cmdline_p)
 
 	find_legacy_serial_ports();
 
-	smp_setup_cpu_maps();
-
 	/* Register early console */
 	register_early_udbg_console();
 
+	smp_setup_cpu_maps();
+
 	xmon_setup();
 
 	if (ppc_md.panic)
diff --git a/arch/powerpc/kernel/setup_64.c b/arch/powerpc/kernel/setup_64.c
index 108035b..df09d33 100644
--- a/arch/powerpc/kernel/setup_64.c
+++ b/arch/powerpc/kernel/setup_64.c
@@ -672,12 +672,13 @@ void __init setup_arch(char **cmdline_p)
 	 */
 	register_early_udbg_console();
 
+	smp_setup_cpu_maps();
+
 	/*
 	 * Initialize xmon
 	 */
 	xmon_setup();
 
-	smp_setup_cpu_maps();
 	check_smt_enabled();
 	setup_tlb_core_data();
 
-- 
2.7.4

[PATCH 36/38] powerpc: Re-order setup_panic()

[Benjamin Herrenschmidt]

Do it right after probe_machine() since it's about testing ppc_md,
and put the test in the common code.

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
---
 arch/powerpc/kernel/setup-common.c | 2 ++
 arch/powerpc/kernel/setup_32.c     | 5 ++---
 arch/powerpc/kernel/setup_64.c     | 5 ++---
 3 files changed, 6 insertions(+), 6 deletions(-)

diff --git a/arch/powerpc/kernel/setup-common.c b/arch/powerpc/kernel/setup-common.c
index 7e85fac..d0256a3 100644
--- a/arch/powerpc/kernel/setup-common.c
+++ b/arch/powerpc/kernel/setup-common.c
@@ -687,6 +687,8 @@ static struct notifier_block ppc_panic_block = {
 
 void __init setup_panic(void)
 {
+	if (!ppc_md.panic)
+		return;
 	atomic_notifier_chain_register(&panic_notifier_list, &ppc_panic_block);
 }
 
diff --git a/arch/powerpc/kernel/setup_32.c b/arch/powerpc/kernel/setup_32.c
index 58674b6..6247a3a 100644
--- a/arch/powerpc/kernel/setup_32.c
+++ b/arch/powerpc/kernel/setup_32.c
@@ -277,6 +277,8 @@ void __init setup_arch(char **cmdline_p)
 
 	probe_machine();
 
+	setup_panic();
+
 	setup_power_save();
 
 	find_legacy_serial_ports();
@@ -288,9 +290,6 @@ void __init setup_arch(char **cmdline_p)
 
 	xmon_setup();
 
-	if (ppc_md.panic)
-		setup_panic();
-
 	init_mm.start_code = (unsigned long)_stext;
 	init_mm.end_code = (unsigned long) _etext;
 	init_mm.end_data = (unsigned long) _edata;
diff --git a/arch/powerpc/kernel/setup_64.c b/arch/powerpc/kernel/setup_64.c
index df09d33..b41756e 100644
--- a/arch/powerpc/kernel/setup_64.c
+++ b/arch/powerpc/kernel/setup_64.c
@@ -660,6 +660,8 @@ void __init setup_arch(char **cmdline_p)
 	/* Probe the machine type */
 	probe_machine();
 
+	setup_panic();
+
  	/*
 	 * We can discover serial ports now since the above did setup the
 	 * hash table management for us, thus ioremap works. We do that early
@@ -708,9 +710,6 @@ void __init setup_arch(char **cmdline_p)
 	 */
 	reserve_hugetlb_gpages();
 
-	if (ppc_md.panic)
-		setup_panic();
-
 	klp_init_thread_info(&init_thread_info);
 
 	init_mm.start_code = (unsigned long)_stext;
-- 
2.7.4

[PATCH 37/38] powerpc/64: Make a few boot functions __init

[Benjamin Herrenschmidt]

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
---
 arch/powerpc/kernel/setup_64.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/arch/powerpc/kernel/setup_64.c b/arch/powerpc/kernel/setup_64.c
index b41756e..9ca2df9 100644
--- a/arch/powerpc/kernel/setup_64.c
+++ b/arch/powerpc/kernel/setup_64.c
@@ -198,7 +198,7 @@ early_param("smt-enabled", early_smt_enabled);
 #endif /* CONFIG_SMP */
 
 /** Fix up paca fields required for the boot cpu */
-static void fixup_boot_paca(void)
+static void __init fixup_boot_paca(void)
 {
 	/* The boot cpu is started */
 	get_paca()->cpu_start = 1;
@@ -206,7 +206,7 @@ static void fixup_boot_paca(void)
 	get_paca()->data_offset = 0;
 }
 
-static void configure_exceptions(void)
+static void __init configure_exceptions(void)
 {
 	/* Setup the trampolines from the lowmem exception vectors
 	 * to the kdump kernel when not using a relocatable kernel.
@@ -528,7 +528,7 @@ static __init void print_system_info(void)
  * used to allocate interrupt or emergency stacks for which our
  * exception entry path doesn't deal with being interrupted.
  */
-static u64 safe_stack_limit(void)
+static __init u64 safe_stack_limit(void)
 {
 #ifdef CONFIG_PPC_BOOK3E
 	/* Freescale BookE bolts the entire linear mapping */
-- 
2.7.4

[PATCH 38/38] powerpc: Merge 32-bit and 64-bit setup_arch()

[Benjamin Herrenschmidt]

There is little enough differences now. Use a few __weak stubs
for functions that may not be around in order to avoid too many
ifdefs.

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
---
 arch/powerpc/include/asm/setup.h   |   4 +
 arch/powerpc/kernel/setup-common.c | 202 +++++++++++++++++++++++++++++++++++++
 arch/powerpc/kernel/setup_32.c     |  65 +-----------
 arch/powerpc/kernel/setup_64.c     | 178 ++------------------------------
 4 files changed, 218 insertions(+), 231 deletions(-)

diff --git a/arch/powerpc/include/asm/setup.h b/arch/powerpc/include/asm/setup.h
index fa0687e..a98f497 100644
--- a/arch/powerpc/include/asm/setup.h
+++ b/arch/powerpc/include/asm/setup.h
@@ -19,6 +19,10 @@ extern unsigned long reloc_offset(void);
 extern unsigned long add_reloc_offset(unsigned long);
 extern void reloc_got2(unsigned long);
 
+extern void initialize_cache_info(void);
+extern void irqstack_early_init(void);
+extern void emergency_stack_init(void);
+
 #ifdef CONFIG_PPC32
 #define PTRRELOC(x)	((typeof(x)) add_reloc_offset((unsigned long)(x)))
 #else
diff --git a/arch/powerpc/kernel/setup-common.c b/arch/powerpc/kernel/setup-common.c
index d0256a3..12e3f5e 100644
--- a/arch/powerpc/kernel/setup-common.c
+++ b/arch/powerpc/kernel/setup-common.c
@@ -35,6 +35,7 @@
 #include <linux/percpu.h>
 #include <linux/memblock.h>
 #include <linux/of_platform.h>
+#include <linux/hugetlb.h>
 #include <asm/io.h>
 #include <asm/paca.h>
 #include <asm/prom.h>
@@ -61,6 +62,11 @@
 #include <asm/cputhreads.h>
 #include <mm/mmu_decl.h>
 #include <asm/fadump.h>
+#include <asm/udbg.h>
+#include <asm/kvm_ppc.h>
+#include <asm/hugetlb.h>
+#include <asm/livepatch.h>
+#include <asm/mmu_context.h>
 
 #ifdef DEBUG
 #include <asm/udbg.h>
@@ -757,3 +763,199 @@ void arch_setup_pdev_archdata(struct platform_device *pdev)
 	pdev->dev.dma_mask = &pdev->archdata.dma_mask;
  	set_dma_ops(&pdev->dev, &dma_direct_ops);
 }
+
+static __init void print_system_info(void)
+{
+	pr_info("-----------------------------------------------------\n");
+#ifdef CONFIG_PPC_STD_MMU_64
+	pr_info("ppc64_pft_size    = 0x%llx\n", ppc64_pft_size);
+#endif
+#ifdef CONFIG_PPC_STD_MMU_32
+	pr_info("Hash_size         = 0x%lx\n", Hash_size);
+#endif
+	pr_info("phys_mem_size     = 0x%llx\n",
+		(unsigned long long)memblock_phys_mem_size());
+
+	pr_info("dcache_bsize      = 0x%x\n", dcache_bsize);
+	pr_info("icache_bsize      = 0x%x\n", icache_bsize);
+	if (ucache_bsize != 0)
+		pr_info("ucache_bsize      = 0x%x\n", ucache_bsize);
+
+	pr_info("cpu_features      = 0x%016lx\n", cur_cpu_spec->cpu_features);
+	pr_info("  possible        = 0x%016lx\n",
+		(unsigned long)CPU_FTRS_POSSIBLE);
+	pr_info("  always          = 0x%016lx\n",
+		(unsigned long)CPU_FTRS_ALWAYS);
+	pr_info("cpu_user_features = 0x%08x 0x%08x\n",
+		cur_cpu_spec->cpu_user_features,
+		cur_cpu_spec->cpu_user_features2);
+	pr_info("mmu_features      = 0x%08x\n", cur_cpu_spec->mmu_features);
+#ifdef CONFIG_PPC64
+	pr_info("firmware_features = 0x%016lx\n", powerpc_firmware_features);
+#endif
+
+#ifdef CONFIG_PPC_STD_MMU_64
+	if (htab_address)
+		pr_info("htab_address      = 0x%p\n", htab_address);
+	if (htab_hash_mask)
+		pr_info("htab_hash_mask    = 0x%lx\n", htab_hash_mask);
+#endif
+#ifdef CONFIG_PPC_STD_MMU_32
+	if (Hash)
+		pr_info("Hash              = 0x%p\n", Hash);
+	if (Hash_mask)
+		pr_info("Hash_mask         = 0x%lx\n", Hash_mask);
+#endif
+
+	if (PHYSICAL_START > 0)
+		pr_info("physical_start    = 0x%llx\n",
+		       (unsigned long long)PHYSICAL_START);
+	pr_info("-----------------------------------------------------\n");
+}
+
+/*
+ * Here are a bunch of weak stubs for initialization functions
+ * that may or may not be present in the current configuration
+ */
+void __init __weak setup_power_save(void)
+{
+}
+
+void __init __weak check_smt_enabled(void)
+{
+}
+
+void __init __weak setup_tlb_core_data(void)
+{
+}
+
+void __init __weak exc_lvl_early_init(void)
+{
+}
+
+void __init __weak emergency_stack_init(void)
+{
+}
+
+void __init __weak rtas_initialize(void)
+{
+}
+
+void __weak smp_release_cpus(void)
+{
+}
+
+/*
+ * Called into from start_kernel this initializes memblock, which is used
+ * to manage page allocation until mem_init is called.
+ */
+void __init setup_arch(char **cmdline_p)
+{
+	*cmdline_p = boot_command_line;
+
+	/* Set a half-reasonable default so udelay does something sensible */
+	loops_per_jiffy = 500000000 / HZ;
+
+	/* Unflatten the device-tree passed by prom_init or kexec */
+	unflatten_device_tree();
+
+	/* Initialize cache line/block info from device-tree (on ppc64) or
+	 * just cputable (on ppc32).
+	 */
+	initialize_cache_info();
+
+	/* Initialize RTAS if available */
+	rtas_initialize();
+
+	/* Check if we have an initrd provided via the device-tree */
+	check_for_initrd();
+
+	/* Probe the machine type, establish ppc_md */
+	probe_machine();
+
+	/* Setup panic notifier if requested by the platform */
+	setup_panic();
+
+	/*
+	 * Configure ppc_md.power_save (ppc32 only, 64-bit machines do
+	 * it from their respective probe() function
+	 */
+	setup_power_save();
+
+	/* Discover standard serial ports */
+	find_legacy_serial_ports();
+
+	/* Register early console with the printk subsystem */
+	register_early_udbg_console();
+
+	/* Setup the various CPU maps based on the device-tree */
+	smp_setup_cpu_maps();
+
+	/* Initialize xmon */
+	xmon_setup();
+
+	/* Check the SMT related command line arguments (ppc64) */
+	check_smt_enabled();
+
+	/* On BookE, setup per-core TLB data structures */
+	setup_tlb_core_data();
+
+	/* Release secondary cpus out of their spinloops at 0x60 now that
+	 * we can map physical -> logical CPU ids
+	 *
+	 * Freescale Book3e parts spin in a loop provided by firmware,
+	 * so smp_release_cpus() does nothing for them
+	 */
+	smp_release_cpus();
+
+	/* Print various info about the machine that gave been gathered
+	 * so far
+	 */
+	print_system_info();
+
+	/* Reserve large chunks of memory for use by CMA for KVM */
+	kvm_cma_reserve();
+
+	/*
+	 * Reserve any gigantic pages requested on the command line.
+	 * memblock needs to have been initialized by the time this is
+	 * called since this will reserve memory.
+	 */
+	reserve_hugetlb_gpages();
+
+	klp_init_thread_info(&init_thread_info);
+
+	init_mm.start_code = (unsigned long)_stext;
+	init_mm.end_code = (unsigned long) _etext;
+	init_mm.end_data = (unsigned long) _edata;
+	init_mm.brk = klimit;
+#ifdef CONFIG_PPC_64K_PAGES
+	init_mm.context.pte_frag = NULL;
+#endif
+#ifdef CONFIG_SPAPR_TCE_IOMMU
+	mm_iommu_init(&init_mm.context);
+#endif
+	irqstack_early_init();
+	exc_lvl_early_init();
+	emergency_stack_init();
+
+	initmem_init();
+
+#ifdef CONFIG_DUMMY_CONSOLE
+	conswitchp = &dummy_con;
+#endif
+	if (ppc_md.setup_arch)
+		ppc_md.setup_arch();
+
+	paging_init();
+
+	/* Initialize the MMU context management stuff */
+	mmu_context_init();
+
+#ifdef CONFIG_PPC64
+	/* Interrupt code needs to be 64K-aligned */
+	if ((unsigned long)_stext & 0xffff)
+		panic("Kernelbase not 64K-aligned (0x%lx)!\n",
+		      (unsigned long)_stext);
+#endif
+}
diff --git a/arch/powerpc/kernel/setup_32.c b/arch/powerpc/kernel/setup_32.c
index 6247a3a..00f5775 100644
--- a/arch/powerpc/kernel/setup_32.c
+++ b/arch/powerpc/kernel/setup_32.c
@@ -36,7 +36,6 @@
 #include <asm/time.h>
 #include <asm/serial.h>
 #include <asm/udbg.h>
-#include <asm/mmu_context.h>
 #include <asm/code-patching.h>
 
 #define DBG(fmt...)
@@ -191,7 +190,7 @@ int __init ppc_init(void)
 
 arch_initcall(ppc_init);
 
-static void __init irqstack_early_init(void)
+void __init irqstack_early_init(void)
 {
 	unsigned int i;
 
@@ -206,7 +205,7 @@ static void __init irqstack_early_init(void)
 }
 
 #if defined(CONFIG_BOOKE) || defined(CONFIG_40x)
-static void __init exc_lvl_early_init(void)
+void __init exc_lvl_early_init(void)
 {
 	unsigned int i, hw_cpu;
 
@@ -229,11 +228,9 @@ static void __init exc_lvl_early_init(void)
 #endif
 	}
 }
-#else
-#define exc_lvl_early_init()
 #endif
 
-static void setup_power_save(void)
+void __init setup_power_save(void)
 {
 #ifdef CONFIG_6xx
 	if (cpu_has_feature(CPU_FTR_CAN_DOZE) ||
@@ -248,7 +245,7 @@ static void setup_power_save(void)
 #endif
 }
 
-static __init void initialize_cache_info(void)
+__init void initialize_cache_info(void)
 {
 	/*
 	 * Set cache line size based on type of cpu as a default.
@@ -261,57 +258,3 @@ static __init void initialize_cache_info(void)
 	if (cpu_has_feature(CPU_FTR_UNIFIED_ID_CACHE))
 		ucache_bsize = icache_bsize = dcache_bsize;
 }
-
-
-/* Warning, IO base is not yet inited */
-void __init setup_arch(char **cmdline_p)
-{
-	*cmdline_p = boot_command_line;
-
-	/* so udelay does something sensible, assume <= 1000 bogomips */
-	loops_per_jiffy = 500000000 / HZ;
-
-	unflatten_device_tree();
-	initialize_cache_info();
-	check_for_initrd();
-
-	probe_machine();
-
-	setup_panic();
-
-	setup_power_save();
-
-	find_legacy_serial_ports();
-
-	/* Register early console */
-	register_early_udbg_console();
-
-	smp_setup_cpu_maps();
-
-	xmon_setup();
-
-	init_mm.start_code = (unsigned long)_stext;
-	init_mm.end_code = (unsigned long) _etext;
-	init_mm.end_data = (unsigned long) _edata;
-	init_mm.brk = klimit;
-
-	exc_lvl_early_init();
-
-	irqstack_early_init();
-
-	initmem_init();
-	if ( ppc_md.progress ) ppc_md.progress("setup_arch: initmem", 0x3eab);
-
-#ifdef CONFIG_DUMMY_CONSOLE
-	conswitchp = &dummy_con;
-#endif
-
-	if (ppc_md.setup_arch)
-		ppc_md.setup_arch();
-	if ( ppc_md.progress ) ppc_md.progress("arch: exit", 0x3eab);
-
-	paging_init();
-
-	/* Initialize the MMU context management stuff */
-	mmu_context_init();
-}
diff --git a/arch/powerpc/kernel/setup_64.c b/arch/powerpc/kernel/setup_64.c
index 9ca2df9..132f32e 100644
--- a/arch/powerpc/kernel/setup_64.c
+++ b/arch/powerpc/kernel/setup_64.c
@@ -35,7 +35,6 @@
 #include <linux/pci.h>
 #include <linux/lockdep.h>
 #include <linux/memblock.h>
-#include <linux/hugetlb.h>
 #include <linux/memory.h>
 #include <linux/nmi.h>
 
@@ -64,12 +63,10 @@
 #include <asm/xmon.h>
 #include <asm/udbg.h>
 #include <asm/kexec.h>
-#include <asm/mmu_context.h>
 #include <asm/code-patching.h>
-#include <asm/kvm_ppc.h>
-#include <asm/hugetlb.h>
-#include <asm/livepatch.h>
 #include <asm/opal.h>
+#include <asm/cputhreads.h>
+#include <asm/livepatch.h>
 
 #ifdef DEBUG
 #define DBG(fmt...) udbg_printf(fmt)
@@ -100,7 +97,7 @@ int icache_bsize;
 int ucache_bsize;
 
 #if defined(CONFIG_PPC_BOOK3E) && defined(CONFIG_SMP)
-static void setup_tlb_core_data(void)
+void __init setup_tlb_core_data(void)
 {
 	int cpu;
 
@@ -133,10 +130,6 @@ static void setup_tlb_core_data(void)
 		}
 	}
 }
-#else
-static void setup_tlb_core_data(void)
-{
-}
 #endif
 
 #ifdef CONFIG_SMP
@@ -144,7 +137,7 @@ static void setup_tlb_core_data(void)
 static char *smt_enabled_cmdline;
 
 /* Look for ibm,smt-enabled OF option */
-static void check_smt_enabled(void)
+void __init check_smt_enabled(void)
 {
 	struct device_node *dn;
 	const char *smt_option;
@@ -193,8 +186,6 @@ static int __init early_smt_enabled(char *p)
 }
 early_param("smt-enabled", early_smt_enabled);
 
-#else
-#define check_smt_enabled()
 #endif /* CONFIG_SMP */
 
 /** Fix up paca fields required for the boot cpu */
@@ -419,7 +410,7 @@ void smp_release_cpus(void)
  * cache informations about the CPU that will be used by cache flush
  * routines and/or provided to userland
  */
-static void __init initialize_cache_info(void)
+void __init initialize_cache_info(void)
 {
 	struct device_node *np;
 	unsigned long num_cpus = 0;
@@ -491,38 +482,6 @@ static void __init initialize_cache_info(void)
 	DBG(" <- initialize_cache_info()\n");
 }
 
-static __init void print_system_info(void)
-{
-	pr_info("-----------------------------------------------------\n");
-	pr_info("ppc64_pft_size    = 0x%llx\n", ppc64_pft_size);
-	pr_info("phys_mem_size     = 0x%llx\n", memblock_phys_mem_size());
-
-	if (ppc64_caches.dline_size != 0x80)
-		pr_info("dcache_line_size  = 0x%x\n", ppc64_caches.dline_size);
-	if (ppc64_caches.iline_size != 0x80)
-		pr_info("icache_line_size  = 0x%x\n", ppc64_caches.iline_size);
-
-	pr_info("cpu_features      = 0x%016lx\n", cur_cpu_spec->cpu_features);
-	pr_info("  possible        = 0x%016lx\n", CPU_FTRS_POSSIBLE);
-	pr_info("  always          = 0x%016lx\n", CPU_FTRS_ALWAYS);
-	pr_info("cpu_user_features = 0x%08x 0x%08x\n", cur_cpu_spec->cpu_user_features,
-		cur_cpu_spec->cpu_user_features2);
-	pr_info("mmu_features      = 0x%08x\n", cur_cpu_spec->mmu_features);
-	pr_info("firmware_features = 0x%016lx\n", powerpc_firmware_features);
-
-#ifdef CONFIG_PPC_STD_MMU_64
-	if (htab_address)
-		pr_info("htab_address      = 0x%p\n", htab_address);
-
-	pr_info("htab_hash_mask    = 0x%lx\n", htab_hash_mask);
-#endif
-
-	if (PHYSICAL_START > 0)
-		pr_info("physical_start    = 0x%llx\n",
-		       (unsigned long long)PHYSICAL_START);
-	pr_info("-----------------------------------------------------\n");
-}
-
 /* This returns the limit below which memory accesses to the linear
  * mapping are guarnateed not to cause a TLB or SLB miss. This is
  * used to allocate interrupt or emergency stacks for which our
@@ -544,7 +503,7 @@ static __init u64 safe_stack_limit(void)
 #endif
 }
 
-static void __init irqstack_early_init(void)
+void __init irqstack_early_init(void)
 {
 	u64 limit = safe_stack_limit();
 	unsigned int i;
@@ -564,7 +523,7 @@ static void __init irqstack_early_init(void)
 }
 
 #ifdef CONFIG_PPC_BOOK3E
-static void __init exc_lvl_early_init(void)
+void __init exc_lvl_early_init(void)
 {
 	unsigned int i;
 	unsigned long sp;
@@ -586,8 +545,6 @@ static void __init exc_lvl_early_init(void)
 	if (cpu_has_feature(CPU_FTR_DEBUG_LVL_EXC))
 		patch_exception(0x040, exc_debug_debug_book3e);
 }
-#else
-#define exc_lvl_early_init()
 #endif
 
 /*
@@ -595,7 +552,7 @@ static void __init exc_lvl_early_init(void)
  * early in SMP boots before relocation is enabled. Exclusive emergency
  * stack for machine checks.
  */
-static void __init emergency_stack_init(void)
+void __init emergency_stack_init(void)
 {
 	u64 limit;
 	unsigned int i;
@@ -626,125 +583,6 @@ static void __init emergency_stack_init(void)
 	}
 }
 
-/*
- * Called into from start_kernel this initializes memblock, which is used
- * to manage page allocation until mem_init is called.
- */
-void __init setup_arch(char **cmdline_p)
-{
-	*cmdline_p = boot_command_line;
-
-	/*
-	 * Unflatten the device-tree passed by prom_init or kexec
-	 */
-	unflatten_device_tree();
-
-	/*
-	 * Fill the ppc64_caches & systemcfg structures with informations
- 	 * retrieved from the device-tree.
-	 */
-	initialize_cache_info();
-
-#ifdef CONFIG_PPC_RTAS
-	/*
-	 * Initialize RTAS if available
-	 */
-	rtas_initialize();
-#endif /* CONFIG_PPC_RTAS */
-
-	/*
-	 * Check if we have an initrd provided via the device-tree
-	 */
-	check_for_initrd();
-
-	/* Probe the machine type */
-	probe_machine();
-
-	setup_panic();
-
- 	/*
-	 * We can discover serial ports now since the above did setup the
-	 * hash table management for us, thus ioremap works. We do that early
-	 * so that further code can be debugged
-	 */
-	find_legacy_serial_ports();
-
-	/*
-	 * Register early console
-	 */
-	register_early_udbg_console();
-
-	smp_setup_cpu_maps();
-
-	/*
-	 * Initialize xmon
-	 */
-	xmon_setup();
-
-	check_smt_enabled();
-	setup_tlb_core_data();
-
-	/*
-	 * Freescale Book3e parts spin in a loop provided by firmware,
-	 * so smp_release_cpus() does nothing for them
-	 */
-#if defined(CONFIG_SMP)
-	/* Release secondary cpus out of their spinloops at 0x60 now that
-	 * we can map physical -> logical CPU ids
-	 */
-	smp_release_cpus();
-#endif
-
-	/* Print various info about the machine that gave been gathered
-	 * so far
-	 */
-	print_system_info();
-
-	/* Reserve large chunks of memory for use by CMA for KVM */
-	kvm_cma_reserve();
-
-	/*
-	 * Reserve any gigantic pages requested on the command line.
-	 * memblock needs to have been initialized by the time this is
-	 * called since this will reserve memory.
-	 */
-	reserve_hugetlb_gpages();
-
-	klp_init_thread_info(&init_thread_info);
-
-	init_mm.start_code = (unsigned long)_stext;
-	init_mm.end_code = (unsigned long) _etext;
-	init_mm.end_data = (unsigned long) _edata;
-	init_mm.brk = klimit;
-#ifdef CONFIG_PPC_64K_PAGES
-	init_mm.context.pte_frag = NULL;
-#endif
-#ifdef CONFIG_SPAPR_TCE_IOMMU
-	mm_iommu_init(&init_mm.context);
-#endif
-	irqstack_early_init();
-	exc_lvl_early_init();
-	emergency_stack_init();
-
-	initmem_init();
-
-#ifdef CONFIG_DUMMY_CONSOLE
-	conswitchp = &dummy_con;
-#endif
-	if (ppc_md.setup_arch)
-		ppc_md.setup_arch();
-
-	paging_init();
-
-	/* Initialize the MMU context management stuff */
-	mmu_context_init();
-
-	/* Interrupt code needs to be 64K-aligned */
-	if ((unsigned long)_stext & 0xffff)
-		panic("Kernelbase not 64K-aligned (0x%lx)!\n",
-		      (unsigned long)_stext);
-}
-
 #ifdef CONFIG_SMP
 #define PCPU_DYN_SIZE		()
 
-- 
2.7.4

Re: [PATCH 29/38] powerpc: Move 32-bit probe() machine to later in the boot process

[Gerhard Pircher]

Am 2016-06-27 um 13:29 schrieb Benjamin Herrenschmidt:
> This converts all the 32-bit platforms to use the expanded device-tree
> which is a pretty mechanical change. Unlike 64-bit, the 32-bit kernel
> didn't rely on platform initializations to setup the MMU since it
> sets it up entirely before probe_machine() so the move has comparatively
> less consequences though it's a bigger patch.
> 
> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

[...]

> diff --git a/arch/powerpc/platforms/amigaone/setup.c b/arch/powerpc/platforms/amigaone/setup.c
> index 2fe1204..a030d3b 100644
> --- a/arch/powerpc/platforms/amigaone/setup.c
> +++ b/arch/powerpc/platforms/amigaone/setup.c
> @@ -143,9 +143,7 @@ void amigaone_restart(char *cmd)
>  
>  static int __init amigaone_probe(void)
>  {
> -	unsigned long root = of_get_flat_dt_root();
> -
> -	if (of_flat_dt_is_compatible(root, "eyetech,amigaone")) {
> +	if (of_machine_is_compatible("eyetech,amigaone")) {
>  		/*
>  		 * Coherent memory access cause complete system lockup! Thus
>  		 * disable this CPU feature, even if the CPU needs it.
This patch series reminds me of a long-standing issue with the
AmigaOne platform code, which would have to clear the coherence
(M) flag for every memory mapping to avoid lockups (especially
on G4 CPUs) - as you can read in the comment below your patch
above.

Now the follow-up code in amigaone/setup.c to clear the
corresponding CPU feature flag is essentially dead code (it used
to work with a second call of do_feature_fixups() before the
handling of the CPU_FTR_NEED_COHERENT flag was changed/inverted
somewhere around v2.6.30). Also it doesn't seem to fix all
deadlocks, as e.g. KVM virtual machines can still inject memory
mappings with the M flag set AFAICT.

Now I wonder, if there is a reasonably clean way to clear the M
flag in all hardware page table and BAT entries without defacing
generic kernel code too much. Any ideas?

BR,
Gerhard

Re: [PATCH 29/38] powerpc: Move 32-bit probe() machine to later in the boot process

[Benjamin Herrenschmidt]

On Mon, 2016-06-27 at 22:42 +0200, Gerhard Pircher wrote:
> This patch series reminds me of a long-standing issue with the
> AmigaOne platform code, which would have to clear the coherence
> (M) flag for every memory mapping to avoid lockups (especially
> on G4 CPUs) - as you can read in the comment below your patch
> above.
> 
> Now the follow-up code in amigaone/setup.c to clear the
> corresponding CPU feature flag is essentially dead code (it used
> to work with a second call of do_feature_fixups() before the
> handling of the CPU_FTR_NEED_COHERENT flag was changed/inverted
> somewhere around v2.6.30). Also it doesn't seem to fix all
> deadlocks, as e.g. KVM virtual machines can still inject memory
> mappings with the M flag set AFAICT.
> 
> Now I wonder, if there is a reasonably clean way to clear the M
> flag in all hardware page table and BAT entries without defacing
> generic kernel code too much. Any ideas?

Other than a compile time option ? Do we know where the lockups come
from btw ? A problem with the northbridge ?

Cheers,
Ben.

Re: [PATCH 06/38] powerpc: Move 64-bit feature fixup earlier

[Aneesh Kumar K.V]

Benjamin Herrenschmidt <benh@kernel.crashing.org> writes:

> Make it part of early_setup() as we really want the feature fixups
> to be applied before we turn on the MMU since they can have an impact
> on the various assembly path related to MMU management and interrupts.
>
> This makes 64-bit match what 32-bit does.
>
> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
> ---
>  arch/powerpc/kernel/setup_64.c | 5 +++--
>  1 file changed, 3 insertions(+), 2 deletions(-)
>
> diff --git a/arch/powerpc/kernel/setup_64.c b/arch/powerpc/kernel/setup_64.c
> index 996bed2..e79cf1b 100644
> --- a/arch/powerpc/kernel/setup_64.c
> +++ b/arch/powerpc/kernel/setup_64.c
> @@ -286,6 +286,9 @@ void __init early_setup(unsigned long dt_ptr)
>  	/* Initialize the hash table or TLB handling */
>  	early_init_mmu();
>  
> +	/* Apply all the dynamic patching */
> +	apply_feature_fixups();
> +
>  	/*
>  	 * At this point, we can let interrupts switch to virtual mode
>  	 * (the MMU has been setup), so adjust the MSR in the PACA to
> @@ -468,8 +471,6 @@ void __init setup_system(void)
>  {
>  	DBG(" -> setup_system()\n");
>  
> -	apply_feature_fixups();
> -
>  	/*
>  	 * Unflatten the device-tree passed by prom_init or kexec
>  	 */

Why not move it immediately after we finish all the device tress scan.

ie,
	early_init_devtree(__va(dt_ptr));

	epapr_paravirt_early_init();

  +      apply_feature_fixups();
	/* Now we know the logical id of our boot cpu, setup the paca. */
	setup_paca(&paca[boot_cpuid]);
	fixup_boot_paca();

Also with a comment explaining why we can't move it further up ?


-aneesh

Re: [PATCH 06/38] powerpc: Move 64-bit feature fixup earlier

[Benjamin Herrenschmidt]

On Tue, 2016-06-28 at 16:35 +0530, Aneesh Kumar K.V wrote:
> Why not move it immediately after we finish all the device tress
> scan.

Do you need it in early_init_mmu ? The important thing is to have it
done before we *turn on* the MMU but yeah, there's no big deal
moving it even further up I think..

> ie,
>         early_init_devtree(__va(dt_ptr));
> 
>         epapr_paravirt_early_init();
> 
>   +      apply_feature_fixups();
>         /* Now we know the logical id of our boot cpu, setup the
> paca. */
>         setup_paca(&paca[boot_cpuid]);
>         fixup_boot_paca();
> 
> Also with a comment explaining why we can't move it further up ?

Yup, adding some comments about the why things are done in that order
would be good, I'll improve that

Cheers,
Ben.

Re: [PATCH 29/38] powerpc: Move 32-bit probe() machine to later in the boot process

[Gerhard Pircher]

On Mon, 2016-06-27 at 23:40 +0200, Benjamin Herrenschmidt wrote:
> On Mon, 2016-06-27 at 22:42 +0200, Gerhard Pircher wrote:
> > This patch series reminds me of a long-standing issue with the
> > AmigaOne platform code, which would have to clear the coherence
> > (M) flag for every memory mapping to avoid lockups (especially
> > on G4 CPUs) - as you can read in the comment below your patch
> > above.
> > 
> > Now the follow-up code in amigaone/setup.c to clear the
> > corresponding CPU feature flag is essentially dead code (it used
> > to work with a second call of do_feature_fixups() before the
> > handling of the CPU_FTR_NEED_COHERENT flag was changed/inverted
> > somewhere around v2.6.30). Also it doesn't seem to fix all
> > deadlocks, as e.g. KVM virtual machines can still inject memory
> > mappings with the M flag set AFAICT.
> > 
> > Now I wonder, if there is a reasonably clean way to clear the M
> > flag in all hardware page table and BAT entries without defacing
> > generic kernel code too much. Any ideas?
> 
> Other than a compile time option ?
Well, I guess a runtime option would always be preferable, but I don't
see how this could be implemented without moving probe_machine() in
front of identify_cpu() (which doesn't look easy either). On the other
side the AmigaOne also requires CONFIG_NOT_COHERENT_CACHE set and this
one is still a compile time option, too.
The question is, if a compile time option that simply clears
CPU_FTR_NEED_COHERENT after identify_cpu() would be acceptable. And
then I still wonder why KVM needs its own similar fix to work on the
AmigaOne. I specifically had to remove/clear the PTE_M bit in
arch/powerpc/kvm/book3s_32_mmu_host.c for this:

206         pteg0 = ((eaddr & 0x0fffffff) >> 22) | (vsid << 7) | PTE_V |
207                 (primary ? 0 : PTE_SEC);
208         pteg1 = hpaddr | PTE_M | PTE_R | PTE_C;
                             ^^^^^

> Do we know where the lockups come from btw ? A problem with the
> northbridge ?
Yes, it's a problem with the northbridge...as usual. :-(

> Cheers,
> Ben.

Thanks!

BR,
Gerhard

Re: [PATCH 06/38] powerpc: Move 64-bit feature fixup earlier

[Aneesh Kumar K.V]

Benjamin Herrenschmidt <benh@kernel.crashing.org> writes:

> On Tue, 2016-06-28 at 16:35 +0530, Aneesh Kumar K.V wrote:
>> Why not move it immediately after we finish all the device tress
>> scan.
>
> Do you need it in early_init_mmu ? The important thing is to have it
> done before we *turn on* the MMU but yeah, there's no big deal
> moving it even further up I think..

If we can do feature fixup early. Then this patch series become simpler.

https://lkml.kernel.org/r/1465887288-12952-1-git-send-email-aneesh.kumar@li=
nux.vnet.ibm.com

ie, we can now start using cpu_has_feature/mmu_has_feature early instead
of using __cpu/mmu_has_feature.

>
>> ie,
>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0early_init_devtree(__va(=
dt_ptr));
>>=20
>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0epapr_paravirt_early_ini=
t();
>>=20
>> =C2=A0 +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 apply_feature_fixups();
>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0/* Now we know the logic=
al id of our boot cpu, setup the
>> paca. */
>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0setup_paca(&paca[boot_cp=
uid]);
>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0fixup_boot_paca();
>>=20
>> Also with a comment explaining why we can't move it further up ?
>
> Yup, adding some comments about the why things are done in that order
> would be good, I'll improve that
>
> Cheers,
> Ben.

-aneesh

Re: [PATCH 06/38] powerpc: Move 64-bit feature fixup earlier

[Benjamin Herrenschmidt]

On Tue, 2016-06-28 at 17:19 +0530, Aneesh Kumar K.V wrote:
> If we can do feature fixup early. Then this patch series become
> simpler.
> 
> https://lkml.kernel.org/r/1465887288-12952-1-git-send-email-aneesh.ku
> mar@linux.vnet.ibm.com
> 
> ie, we can now start using cpu_has_feature/mmu_has_feature early
> instead of using __cpu/mmu_has_feature.

Ok, I'll look into it tomorrow.

Cheers,
Ben.

Re: [PATCH 29/38] powerpc: Move 32-bit probe() machine to later in the boot process

[Benjamin Herrenschmidt]

On Tue, 2016-06-28 at 13:42 +0200, Gerhard Pircher wrote:
> The question is, if a compile time option that simply clear ?s
> CPU_FTR_NEED_COHERENT after identify_cpu() would be acceptable. And
> then I still wonder why KVM needs its own similar fix to work on the
> AmigaOne. I specifically had to remove/clear the PTE_M bit
> inarch/powerpc/kvm/book3s_32_mmu_host.c for th

No we could just add something to early_init_devtree that does clear
that bit if it sees the relevant piece of broken HW, it's not the AmigaOne per-se, it's the northbridge right ?

We could also make non-coherent cache a runtime option if we really
wanted, it's just more churn ...

> 206         pteg0 = ((eaddr & 0x0fffffff) >> 22) | (vsid << 7) |
> PTE_V |
> 207                 (primary ? 0 : PTE_SEC);
> 208         pteg1 = hpaddr | PTE_M | PTE_R | PTE_C;
>                              ^^^^^
> 
> > Do we know where the lockups come from btw ? A problem with the
> > northbridge ?
> Yes, it's a problem with the northbridge...as usual. :-(

Marvell ?

Cheers,
Ben.

Re: [PATCH 29/38] powerpc: Move 32-bit probe() machine to later in the boot process

[Gerhard Pircher]

On Tue, 2016-06-28 at 14:01 +0200, Benjamin Herrenschmidt wrote:
> On Tue, 2016-06-28 at 13:42 +0200, Gerhard Pircher wrote:
> > The question is, if a compile time option that simply clear ?s
> > CPU_FTR_NEED_COHERENT after identify_cpu() would be acceptable. And
> > then I still wonder why KVM needs its own similar fix to work on the
> > AmigaOne. I specifically had to remove/clear the PTE_M bit
> > inarch/powerpc/kvm/book3s_32_mmu_host.c for th
> 
> No we could just add something to early_init_devtree that does clear
> that bit if it sees the relevant piece of broken HW, it's not the AmigaOne per-se, it's the northbridge right ?
Does that work for CPU_FTR_NEED_COHERENT, given that the feature fixups
are done right after identify_cpu() and instructions within BEGIN_FTR_SECTION
and END_FTR_SECTION* are patched out (if I understand this mechanism
correctly) !?
But yes, the specific northbridge is the problem here.

> We could also make non-coherent cache a runtime option if we really
> wanted, it's just more churn ...
Yeah, I remember a discussion or two about this topic and I always wondered
how driver code would differentiate then between non-coherent and coherent
DMA, given that they sometimes require different handling of the DMA addresses
(e.g. if I look at /sound/core/pcm_native.c or the DRM code).

> > > Do we know where the lockups come from btw ? A problem with the
> > > northbridge ?
> > Yes, it's a problem with the northbridge...as usual. :-(
> 
> Marvell ?
No, the one with the dreaded "A" name: MAI Logic's "Articia S" :-)

BR,
Gerhard

Re: [PATCH 29/38] powerpc: Move 32-bit probe() machine to later in the boot process

[Benjamin Herrenschmidt]

On Tue, 2016-06-28 at 15:25 +0200, Gerhard Pircher wrote:
> > No we could just add something to early_init_devtree that does clear
> > that bit if it sees the relevant piece of broken HW, it's not the AmigaOne per-se, it's the northbridge right ?
> Does that work for CPU_FTR_NEED_COHERENT, given that the feature fixups
> are done right after identify_cpu() and instructions within BEGIN_FTR_SECTION
> and END_FTR_SECTION* are patched out (if I understand this mechanism
> correctly) !?

Right that's a problem. We do the fixup before early_init_devtree() on
ppc32 (we don't on ppc64) ... ugh. I'll have to think about it. We can
try to make them match by moving the fixup but that means making
sure that the transition to the "initial" MMU mapping use for
machine_init() can be done without the fixups being applied.

I can have a quick sweep of the init code later today or tomorrow to
see how bad that would be.

Otherwise, compile option...

> But yes, the specific northbridge is the problem here.
> 
> > We could also make non-coherent cache a runtime option if we really
> > wanted, it's just more churn ...
> Yeah, I remember a discussion or two about this topic and I always wondered
> how driver code would differentiate then between non-coherent and coherent
> DMA, given that they sometimes require different handling of the DMA addresses
> (e.g. if I look at /sound/core/pcm_native.c or the DRM code).
> 
> > > > Do we know where the lockups come from btw ? A problem with the
> > > > northbridge ?
> > > Yes, it's a problem with the northbridge...as usual. :-(
> > 
> > Marvell ?
> No, the one with the dreaded "A" name: MAI Logic's "Articia S" :-)

Heh ok.

Cheers,
Ben.

[PATCH v2 13/38] powerpc: Put exception configuration in a common place

[Benjamin Herrenschmidt]

The various calls to establish exception endianness and AIL are
now done from a single point using already established CPU and FW
feature bits to decide what to do.

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
---
 
v2: Add/fix prototypes of exported function, remove "static"

 arch/powerpc/include/asm/hvcall.h      |  8 ++--
 arch/powerpc/include/asm/opal.h        |  1 +
 arch/powerpc/kernel/setup_64.c         | 68 +++++++++++++++++++++++++++-------
 arch/powerpc/platforms/powernv/opal.c  | 13 +++----
 arch/powerpc/platforms/pseries/setup.c | 31 +---------------
 5 files changed, 66 insertions(+), 55 deletions(-)

diff --git a/arch/powerpc/include/asm/hvcall.h b/arch/powerpc/include/asm/hvcall.h
index 0bc9c28..b88efbb 100644
--- a/arch/powerpc/include/asm/hvcall.h
+++ b/arch/powerpc/include/asm/hvcall.h
@@ -434,13 +434,15 @@ static inline unsigned long cmo_get_page_size(void)
 
 extern long pSeries_enable_reloc_on_exc(void);
 extern long pSeries_disable_reloc_on_exc(void);
-
 extern long pseries_big_endian_exceptions(void);
+extern long pseries_little_endian_exceptions(void);
 
 #else
 
-#define pSeries_enable_reloc_on_exc()  do {} while (0)
-#define pSeries_disable_reloc_on_exc() do {} while (0)
+#define pSeries_enable_reloc_on_exc()		(0)
+#define pSeries_disable_reloc_on_exc()		(0)
+#define pseries_big_endian_exceptions()		(0)
+#define pseries_little_endian_exceptions()	(0)
 
 #endif /* CONFIG_PPC_PSERIES */
 
diff --git a/arch/powerpc/include/asm/opal.h b/arch/powerpc/include/asm/opal.h
index 9d86c66..6135816 100644
--- a/arch/powerpc/include/asm/opal.h
+++ b/arch/powerpc/include/asm/opal.h
@@ -215,6 +215,7 @@ extern int early_init_dt_scan_opal(unsigned long node, const char *uname,
 				   int depth, void *data);
 extern int early_init_dt_scan_recoverable_ranges(unsigned long node,
 				 const char *uname, int depth, void *data);
+extern void opal_configure_cores(void);
 
 extern int opal_get_chars(uint32_t vtermno, char *buf, int count);
 extern int opal_put_chars(uint32_t vtermno, const char *buf, int total_len);
diff --git a/arch/powerpc/kernel/setup_64.c b/arch/powerpc/kernel/setup_64.c
index a641753..47a2706 100644
--- a/arch/powerpc/kernel/setup_64.c
+++ b/arch/powerpc/kernel/setup_64.c
@@ -69,6 +69,7 @@
 #include <asm/kvm_ppc.h>
 #include <asm/hugetlb.h>
 #include <asm/livepatch.h>
+#include <asm/opal.h>
 
 #ifdef DEBUG
 #define DBG(fmt...) udbg_printf(fmt)
@@ -205,21 +206,60 @@ static void fixup_boot_paca(void)
 	get_paca()->data_offset = 0;
 }
 
+static void configure_exceptions(void)
+{
+	/* Setup the trampolines from the lowmem exception vectors
+	 * to the kdump kernel when not using a relocatable kernel.
+	 */
+	setup_kdump_trampoline();
+
+	/* Under a PAPR hypervisor, we need hypercalls */
+	if (firmware_has_feature(FW_FEATURE_SET_MODE)) {
+		long rc;
+
+		/* Enable AIL */
+		rc = pSeries_enable_reloc_on_exc();
+		if (rc == H_P2) {
+			pr_info("Relocation on exceptions not supported\n");
+		} else if (rc != H_SUCCESS) {
+			pr_warn("Unable to enable relocation on exceptions: "
+				"%ld\n", rc);
+		}
+
+		/*
+		 * Tell the hypervisor that we want our exceptions to
+		 * be taken in little endian mode. If this fails we don't
+		 * want to use BUG() because it will trigger an exception.
+		 *
+		 * We don't call this for big endian as our calling convention
+		 * makes us always enter in BE, and the call may fail under
+		 * some circumstances with kdump.
+		 */
+#ifdef __LITTLE_ENDIAN__
+		rc = pseries_little_endian_exceptions();
+		if (rc) {
+			ppc_md.progress("H_SET_MODE LE exception fail", 0);
+			panic("Could not enable little endian exceptions");
+		}
+#endif
+	} else {
+		/* Set endian mode using OPAL */
+		if (firmware_has_feature(FW_FEATURE_OPAL))
+			opal_configure_cores();
+
+		/* Enable AIL if supported, and we are in hypervisor mode */
+		if (cpu_has_feature(CPU_FTR_HVMODE) &&
+		    cpu_has_feature(CPU_FTR_ARCH_207S)) {
+			unsigned long lpcr = mfspr(SPRN_LPCR);
+			mtspr(SPRN_LPCR, lpcr | LPCR_AIL_3);
+		}
+	}
+}
+
 static void cpu_ready_for_interrupts(void)
 {
 	/* Set IR and DR in PACA MSR */
 	get_paca()->kernel_msr = MSR_KERNEL;
-
-	/*
-	 * Enable AIL if supported, and we are in hypervisor mode. If we are
-	 * not in hypervisor mode, we enable relocation-on interrupts later
-	 * in pSeries_setup_arch() using the H_SET_MODE hcall.
-	 */
-	if (cpu_has_feature(CPU_FTR_HVMODE) &&
-	    cpu_has_feature(CPU_FTR_ARCH_207S)) {
-		unsigned long lpcr = mfspr(SPRN_LPCR);
-		mtspr(SPRN_LPCR, lpcr | LPCR_AIL_3);
-	}
 }
 
 /*
@@ -276,10 +316,10 @@ void __init early_setup(unsigned long dt_ptr)
 	/* Probe the machine type */
 	probe_machine();
 
-	/* Setup the trampolines from the lowmem exception vectors
-	 * to the kdump kernel when not using a relocatable kernel.
+	/* Configure exception handlers. This include setting up trampolines
+	 * if needed, setting exception endian mode, etc...
 	 */
-	setup_kdump_trampoline();
+	configure_exceptions();
 
 	/* Initialize the hash table or TLB handling */
 	early_init_mmu();
diff --git a/arch/powerpc/platforms/powernv/opal.c b/arch/powerpc/platforms/powernv/opal.c
index 0256d07..802f3b7 100644
--- a/arch/powerpc/platforms/powernv/opal.c
+++ b/arch/powerpc/platforms/powernv/opal.c
@@ -56,7 +56,7 @@ static DEFINE_SPINLOCK(opal_write_lock);
 static struct atomic_notifier_head opal_msg_notifier_head[OPAL_MSG_TYPE_MAX];
 static uint32_t opal_heartbeat;
 
-static void opal_reinit_cores(void)
+void opal_configure_cores(void)
 {
 	/* Do the actual re-init, This will clobber all FPRs, VRs, etc...
 	 *
@@ -69,6 +69,10 @@ static void opal_reinit_cores(void)
 #else
 	opal_reinit_cpus(OPAL_REINIT_CPUS_HILE_LE);
 #endif
+
+	/* Restore some bits */
+	if (cur_cpu_spec->cpu_restore)
+		cur_cpu_spec->cpu_restore();
 }
 
 int __init early_init_dt_scan_opal(unsigned long node,
@@ -105,13 +109,6 @@ int __init early_init_dt_scan_opal(unsigned long node,
 		panic("OPAL != V3 detected, no longer supported.\n");
 	}
 
-	/* Reinit all cores with the right endian */
-	opal_reinit_cores();
-
-	/* Restore some bits */
-	if (cur_cpu_spec->cpu_restore)
-		cur_cpu_spec->cpu_restore();
-
 	return 1;
 }
 
diff --git a/arch/powerpc/platforms/pseries/setup.c b/arch/powerpc/platforms/pseries/setup.c
index f1fe7aa..ffffd77 100644
--- a/arch/powerpc/platforms/pseries/setup.c
+++ b/arch/powerpc/platforms/pseries/setup.c
@@ -458,7 +458,7 @@ long pseries_big_endian_exceptions(void)
 	}
 }
 
-static long pseries_little_endian_exceptions(void)
+long pseries_little_endian_exceptions(void)
 {
 	long rc;
 
@@ -537,18 +537,6 @@ static void __init pSeries_setup_arch(void)
 	}
 
 	ppc_md.pcibios_root_bridge_prepare = pseries_root_bridge_prepare;
-
-	if (firmware_has_feature(FW_FEATURE_SET_MODE)) {
-		long rc;
-
-		rc = pSeries_enable_reloc_on_exc();
-		if (rc == H_P2) {
-			pr_info("Relocation on exceptions not supported\n");
-		} else if (rc != H_SUCCESS) {
-			pr_warn("Unable to enable relocation on exceptions: "
-				"%ld\n", rc);
-		}
-	}
 }
 
 static int __init pSeries_init_panel(void)
@@ -751,23 +739,6 @@ static int __init pSeries_probe(void)
 
 	pr_debug("pSeries detected, looking for LPAR capability...\n");
 
-
-#ifdef __LITTLE_ENDIAN__
-	if (firmware_has_feature(FW_FEATURE_SET_MODE)) {
-		long rc;
-		/*
-		 * Tell the hypervisor that we want our exceptions to
-		 * be taken in little endian mode. If this fails we don't
-		 * want to use BUG() because it will trigger an exception.
-		 */
-		rc = pseries_little_endian_exceptions();
-		if (rc) {
-			ppc_md.progress("H_SET_MODE LE exception fail", 0);
-			panic("Could not enable little endian exceptions");
-		}
-	}
-#endif
-
 	if (firmware_has_feature(FW_FEATURE_LPAR))
 		hpte_init_lpar();
 	else

Re: [PATCH 13/38] powerpc: Put exception configuration in a common place

[Anton Blanchard]

Hi Ben,

> The various calls to establish exception endianness and AIL are
> now done from a single point using already established CPU and FW
> feature bits to decide what to do.
> 
> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

...

+static void configure_exceptions(void)
+{
+	/* Setup the trampolines from the lowmem exception vectors
+	 * to the kdump kernel when not using a relocatable kernel.
+	 */
+	setup_kdump_trampoline();
+
+	/* Under a PAPR hypervisor, we need hypercalls */
+	if (firmware_has_feature(FW_FEATURE_SET_MODE)) {
+		long rc;
+
+		/* Enable AIL */
+		rc = pSeries_enable_reloc_on_exc();
+		if (rc == H_P2) {
+			pr_info("Relocation on exceptions not supported\n");
+		} else if (rc != H_SUCCESS) {
+			pr_warn("Unable to enable relocation on exceptions: "
+				"%ld\n", rc);
+		}
+
+		/*
+		 * Tell the hypervisor that we want our exceptions to
+		 * be taken in little endian mode. If this fails we don't
+		 * want to use BUG() because it will trigger an exception.
+		 *
+		 * We don't call this for big endian as our calling convention
+		 * makes us always enter in BE, and the call may fail under
+		 * some circumstances with kdump.
+		 */
+#ifdef __LITTLE_ENDIAN__
+		rc = pseries_little_endian_exceptions();
+		if (rc) {
+			ppc_md.progress("H_SET_MODE LE exception fail", 0);
+			panic("Could not enable little endian exceptions");
+		}
+#endif
+	} else {
+		/* Set endian mode using OPAL */
+		if (firmware_has_feature(FW_FEATURE_OPAL))
+			opal_configure_cores();
+
+		/* Enable AIL if supported, and we are in hypervisor mode */
+		if (cpu_has_feature(CPU_FTR_HVMODE) &&
+		    cpu_has_feature(CPU_FTR_ARCH_207S)) {
+			unsigned long lpcr = mfspr(SPRN_LPCR);
+			mtspr(SPRN_LPCR, lpcr | LPCR_AIL_3);
+		}
+	}
+}

It looks like we only set LPCR_AIL_3 on the boot CPU after this change,
is that expected? Before that we did it in cpu_ready_for_interrupts()
which is called for the primary and all secondary CPUs.

Anton

Re: [PATCH 13/38] powerpc: Put exception configuration in a common place

[Benjamin Herrenschmidt]

On Fri, 2016-11-11 at 21:06 +1100, Anton Blanchard wrote:
> It looks like we only set LPCR_AIL_3 on the boot CPU after this change,
> is that expected? Before that we did it in cpu_ready_for_interrupts()
> which is called for the primary and all secondary CPUs.

That would be a mistake... I suppose I didn't realize we had to call
these things for every CPU.

I'll try to cook up a fix next week.

Cheers,
Ben.