Hi Everyone,
This patchset is intended to implement sparsemem on RISC-V.
The first few patches are introducing a common helper used by the
sparesmem implementation in other architectures and the final
patch is the actual RISC-V implementation.
This is the first small step in supporting P2P on RISC-V.
Thanks,
Logan
--
Logan Gunthorpe (5):
mm/sparse: add common helper to mark all memblocks present
ARM: mm: make use of new memblocks_present() helper
arm64: mm: make use of new memblocks_present() helper
sh: mm: make use of new memblocks_present() helper
RISC-V: Implement sparsemem
arch/arm/mm/init.c | 17 +----------------
arch/arm64/mm/init.c | 20 +-------------------
arch/riscv/Kconfig | 23 +++++++++++++++++++++++
arch/riscv/include/asm/pgtable.h | 24 ++++++++++++++++++++----
arch/riscv/include/asm/sparsemem.h | 11 +++++++++++
arch/riscv/kernel/setup.c | 4 +++-
arch/riscv/mm/init.c | 8 ++++++++
arch/sh/mm/init.c | 7 +------
include/linux/mmzone.h | 6 ++++++
mm/sparse.c | 15 +++++++++++++++
10 files changed, 89 insertions(+), 46 deletions(-)
create mode 100644 arch/riscv/include/asm/sparsemem.h
--
2.19.0
This patch implements sparsemem support for risc-v which helps pave the
way for memory hotplug and eventually P2P support.
We introduce Kconfig options for virtual and physical address bits which
are used to calculate the size of the vmemmap and set the
MAX_PHYSMEM_BITS.
The vmemmap is located directly before the VMALLOC region and sized
such that we can allocate enough pages to populate all the virtual
address space in the system (similar to the way it's done in arm64).
During initialization, call memblocks_present() and sparse_init(),
and provide a stub for vmemmap_populate() (all of which is similar to
arm64).
Signed-off-by: Logan Gunthorpe <[email protected]>
Cc: Palmer Dabbelt <[email protected]>
Cc: Albert Ou <[email protected]>
Cc: Andrew Waterman <[email protected]>
Cc: Olof Johansson <[email protected]>
Cc: Michael Clark <[email protected]>
Cc: Rob Herring <[email protected]>
Cc: Zong Li <[email protected]>
---
arch/riscv/Kconfig | 23 +++++++++++++++++++++++
arch/riscv/include/asm/pgtable.h | 24 ++++++++++++++++++++----
arch/riscv/include/asm/sparsemem.h | 11 +++++++++++
arch/riscv/kernel/setup.c | 4 +++-
arch/riscv/mm/init.c | 8 ++++++++
5 files changed, 65 insertions(+), 5 deletions(-)
create mode 100644 arch/riscv/include/asm/sparsemem.h
diff --git a/arch/riscv/Kconfig b/arch/riscv/Kconfig
index a344980287a5..a1b5d758a542 100644
--- a/arch/riscv/Kconfig
+++ b/arch/riscv/Kconfig
@@ -52,12 +52,32 @@ config ZONE_DMA32
bool
default y if 64BIT
+config VA_BITS
+ int
+ default 32 if 32BIT
+ default 39 if 64BIT
+
+config PA_BITS
+ int
+ default 34 if 32BIT
+ default 56 if 64BIT
+
config PAGE_OFFSET
hex
default 0xC0000000 if 32BIT && MAXPHYSMEM_2GB
default 0xffffffff80000000 if 64BIT && MAXPHYSMEM_2GB
default 0xffffffe000000000 if 64BIT && MAXPHYSMEM_128GB
+config ARCH_FLATMEM_ENABLE
+ def_bool y
+
+config ARCH_SPARSEMEM_ENABLE
+ def_bool y
+ select SPARSEMEM_VMEMMAP_ENABLE
+
+config ARCH_SELECT_MEMORY_MODEL
+ def_bool ARCH_SPARSEMEM_ENABLE
+
config STACKTRACE_SUPPORT
def_bool y
@@ -92,6 +112,9 @@ config PGTABLE_LEVELS
config HAVE_KPROBES
def_bool n
+config HAVE_ARCH_PFN_VALID
+ def_bool y
+
menu "Platform type"
choice
diff --git a/arch/riscv/include/asm/pgtable.h b/arch/riscv/include/asm/pgtable.h
index 16301966d65b..20c49cded686 100644
--- a/arch/riscv/include/asm/pgtable.h
+++ b/arch/riscv/include/asm/pgtable.h
@@ -89,6 +89,26 @@ extern pgd_t swapper_pg_dir[];
#define __S110 PAGE_SHARED_EXEC
#define __S111 PAGE_SHARED_EXEC
+#define VMALLOC_SIZE (KERN_VIRT_SIZE >> 1)
+#define VMALLOC_END (PAGE_OFFSET - 1)
+#define VMALLOC_START (PAGE_OFFSET - VMALLOC_SIZE)
+
+/*
+ * Log2 of the upper bound of the size of a struct page. Used for sizing
+ * the vmemmap region only, does not affect actual memory footprint.
+ * We don't use sizeof(struct page) directly since taking its size here
+ * requires its definition to be available at this point in the inclusion
+ * chain, and it may not be a power of 2 in the first place.
+ */
+#define STRUCT_PAGE_MAX_SHIFT 6
+
+#define VMEMMAP_SIZE (UL(1) << (CONFIG_VA_BITS - PAGE_SHIFT - 1 + \
+ STRUCT_PAGE_MAX_SHIFT))
+#define VMEMMAP_END (VMALLOC_START - 1)
+#define VMEMMAP_START (VMALLOC_START - VMEMMAP_SIZE)
+
+#define vmemmap ((struct page *)VMEMMAP_START)
+
/*
* ZERO_PAGE is a global shared page that is always zero,
* used for zero-mapped memory areas, etc.
@@ -411,10 +431,6 @@ static inline void pgtable_cache_init(void)
/* No page table caches to initialize */
}
-#define VMALLOC_SIZE (KERN_VIRT_SIZE >> 1)
-#define VMALLOC_END (PAGE_OFFSET - 1)
-#define VMALLOC_START (PAGE_OFFSET - VMALLOC_SIZE)
-
/*
* Task size is 0x40000000000 for RV64 or 0xb800000 for RV32.
* Note that PGDIR_SIZE must evenly divide TASK_SIZE.
diff --git a/arch/riscv/include/asm/sparsemem.h b/arch/riscv/include/asm/sparsemem.h
new file mode 100644
index 000000000000..4563e806c788
--- /dev/null
+++ b/arch/riscv/include/asm/sparsemem.h
@@ -0,0 +1,11 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+
+#ifndef __ASM_SPARSEMEM_H
+#define __ASM_SPARSEMEM_H
+
+#ifdef CONFIG_SPARSEMEM
+#define MAX_PHYSMEM_BITS CONFIG_PA_BITS
+#define SECTION_SIZE_BITS 30
+#endif
+
+#endif
diff --git a/arch/riscv/kernel/setup.c b/arch/riscv/kernel/setup.c
index aee603123030..89fa781a9bf8 100644
--- a/arch/riscv/kernel/setup.c
+++ b/arch/riscv/kernel/setup.c
@@ -205,6 +205,9 @@ static void __init setup_bootmem(void)
PFN_PHYS(end_pfn - start_pfn),
&memblock.memory, 0);
}
+
+ memblocks_present();
+ sparse_init();
}
void __init setup_arch(char **cmdline_p)
@@ -239,4 +242,3 @@ void __init setup_arch(char **cmdline_p)
riscv_fill_hwcap();
}
-
diff --git a/arch/riscv/mm/init.c b/arch/riscv/mm/init.c
index 58a522f9bcc3..5d529878667c 100644
--- a/arch/riscv/mm/init.c
+++ b/arch/riscv/mm/init.c
@@ -70,3 +70,11 @@ void free_initrd_mem(unsigned long start, unsigned long end)
{
}
#endif /* CONFIG_BLK_DEV_INITRD */
+
+#ifdef CONFIG_SPARSEMEM
+int __meminit vmemmap_populate(unsigned long start, unsigned long end, int node,
+ struct vmem_altmap *altmap)
+{
+ return vmemmap_populate_basepages(start, end, node);
+}
+#endif
--
2.19.0
Cleanup the arm_memory_present() function seeing it's very
similar to other arches.
The new memblocks_present() helper checks for node ids which the
arm version did not. However, this is equivalent seeing
HAVE_MEMBLOCK_NODE_MAP should be false in this arch and therefore
memblock_get_region_node() should return 0.
Signed-off-by: Logan Gunthorpe <[email protected]>
Cc: Russell King <[email protected]>
Cc: Kees Cook <[email protected]>
Cc: Philip Derrin <[email protected]>
Cc: "Steven Rostedt (VMware)" <[email protected]>
Cc: Nicolas Pitre <[email protected]>
---
arch/arm/mm/init.c | 17 +----------------
1 file changed, 1 insertion(+), 16 deletions(-)
diff --git a/arch/arm/mm/init.c b/arch/arm/mm/init.c
index 0cc8e04295a4..e2710dd7446f 100644
--- a/arch/arm/mm/init.c
+++ b/arch/arm/mm/init.c
@@ -201,21 +201,6 @@ int pfn_valid(unsigned long pfn)
EXPORT_SYMBOL(pfn_valid);
#endif
-#ifndef CONFIG_SPARSEMEM
-static void __init arm_memory_present(void)
-{
-}
-#else
-static void __init arm_memory_present(void)
-{
- struct memblock_region *reg;
-
- for_each_memblock(memory, reg)
- memory_present(0, memblock_region_memory_base_pfn(reg),
- memblock_region_memory_end_pfn(reg));
-}
-#endif
-
static bool arm_memblock_steal_permitted = true;
phys_addr_t __init arm_memblock_steal(phys_addr_t size, phys_addr_t align)
@@ -317,7 +302,7 @@ void __init bootmem_init(void)
* Sparsemem tries to allocate bootmem in memory_present(),
* so must be done after the fixed reservations
*/
- arm_memory_present();
+ memblocks_present();
/*
* sparse_init() needs the bootmem allocator up and running.
--
2.19.0
Presently the arches arm64, arm, sh have a function which loops through
each memblock and calls memory present. riscv will require a similar
function.
Introduce a common memblocks_present() function that can be used by
all the arches. Subsequent patches will cleanup the arches that
make use of this.
Signed-off-by: Logan Gunthorpe <[email protected]>
Cc: Andrew Morton <[email protected]>
Cc: Michal Hocko <[email protected]>
Cc: Vlastimil Babka <[email protected]>
Cc: Pavel Tatashin <[email protected]>
Cc: Oscar Salvador <[email protected]>
---
include/linux/mmzone.h | 6 ++++++
mm/sparse.c | 15 +++++++++++++++
2 files changed, 21 insertions(+)
diff --git a/include/linux/mmzone.h b/include/linux/mmzone.h
index 1e22d96734e0..a10fc3c18b07 100644
--- a/include/linux/mmzone.h
+++ b/include/linux/mmzone.h
@@ -794,6 +794,12 @@ void memory_present(int nid, unsigned long start, unsigned long end);
static inline void memory_present(int nid, unsigned long start, unsigned long end) {}
#endif
+#if defined(CONFIG_SPARSEMEM) && defined(CONFIG_HAVE_MEMBLOCK)
+void memblocks_present(void);
+#else
+static inline void memblocks_present(void) {}
+#endif
+
#ifdef CONFIG_HAVE_MEMORYLESS_NODES
int local_memory_node(int node_id);
#else
diff --git a/mm/sparse.c b/mm/sparse.c
index 10b07eea9a6e..109159574208 100644
--- a/mm/sparse.c
+++ b/mm/sparse.c
@@ -5,6 +5,7 @@
#include <linux/mm.h>
#include <linux/slab.h>
#include <linux/mmzone.h>
+#include <linux/memblock.h>
#include <linux/bootmem.h>
#include <linux/compiler.h>
#include <linux/highmem.h>
@@ -238,6 +239,20 @@ void __init memory_present(int nid, unsigned long start, unsigned long end)
}
}
+#ifdef CONFIG_HAVE_MEMBLOCK
+void __init memblocks_present(void)
+{
+ struct memblock_region *reg;
+
+ for_each_memblock(memory, reg) {
+ int nid = memblock_get_region_node(reg);
+
+ memory_present(nid, memblock_region_memory_base_pfn(reg),
+ memblock_region_memory_end_pfn(reg));
+ }
+}
+#endif
+
/*
* Subtle, we encode the real pfn into the mem_map such that
* the identity pfn - section_mem_map will return the actual
--
2.19.0
Cleanup the open coded for_each_memblock() loop that is equivalent
to the new memblocks_present() helper.
Signed-off-by: Logan Gunthorpe <[email protected]>
Cc: Yoshinori Sato <[email protected]>
Cc: Rich Felker <[email protected]>
Cc: Dan Williams <[email protected]>
Cc: Rob Herring <[email protected]>
---
arch/sh/mm/init.c | 7 +------
1 file changed, 1 insertion(+), 6 deletions(-)
diff --git a/arch/sh/mm/init.c b/arch/sh/mm/init.c
index 7713c084d040..f601f96408ee 100644
--- a/arch/sh/mm/init.c
+++ b/arch/sh/mm/init.c
@@ -235,12 +235,7 @@ static void __init do_init_bootmem(void)
plat_mem_setup();
- for_each_memblock(memory, reg) {
- int nid = memblock_get_region_node(reg);
-
- memory_present(nid, memblock_region_memory_base_pfn(reg),
- memblock_region_memory_end_pfn(reg));
- }
+ memblocks_present();
sparse_init();
}
--
2.19.0
Cleanup the arm64_memory_present() function seeing it's very
similar to other arches.
memblocks_present() is a direct replacement of arm64_memory_present()
Signed-off-by: Logan Gunthorpe <[email protected]>
---
arch/arm64/mm/init.c | 20 +-------------------
1 file changed, 1 insertion(+), 19 deletions(-)
diff --git a/arch/arm64/mm/init.c b/arch/arm64/mm/init.c
index 787e27964ab9..63fa9653f281 100644
--- a/arch/arm64/mm/init.c
+++ b/arch/arm64/mm/init.c
@@ -296,24 +296,6 @@ int pfn_valid(unsigned long pfn)
EXPORT_SYMBOL(pfn_valid);
#endif
-#ifndef CONFIG_SPARSEMEM
-static void __init arm64_memory_present(void)
-{
-}
-#else
-static void __init arm64_memory_present(void)
-{
- struct memblock_region *reg;
-
- for_each_memblock(memory, reg) {
- int nid = memblock_get_region_node(reg);
-
- memory_present(nid, memblock_region_memory_base_pfn(reg),
- memblock_region_memory_end_pfn(reg));
- }
-}
-#endif
-
static phys_addr_t memory_limit = PHYS_ADDR_MAX;
/*
@@ -506,7 +488,7 @@ void __init bootmem_init(void)
* Sparsemem tries to allocate bootmem in memory_present(), so must be
* done after the fixed reservations.
*/
- arm64_memory_present();
+ memblocks_present();
sparse_init();
zone_sizes_init(min, max);
--
2.19.0
On Fri, Oct 05, 2018 at 10:16:40AM -0600, Logan Gunthorpe wrote:
> Cleanup the arm64_memory_present() function seeing it's very
> similar to other arches.
>
> memblocks_present() is a direct replacement of arm64_memory_present()
>
> Signed-off-by: Logan Gunthorpe <[email protected]>
> ---
> arch/arm64/mm/init.c | 20 +-------------------
> 1 file changed, 1 insertion(+), 19 deletions(-)
Acked-by: Catalin Marinas <[email protected]>
On Fri, 05 Oct 2018 09:16:42 PDT (-0700), [email protected] wrote:
> This patch implements sparsemem support for risc-v which helps pave the
> way for memory hotplug and eventually P2P support.
>
> We introduce Kconfig options for virtual and physical address bits which
> are used to calculate the size of the vmemmap and set the
> MAX_PHYSMEM_BITS.
>
> The vmemmap is located directly before the VMALLOC region and sized
> such that we can allocate enough pages to populate all the virtual
> address space in the system (similar to the way it's done in arm64).
>
> During initialization, call memblocks_present() and sparse_init(),
> and provide a stub for vmemmap_populate() (all of which is similar to
> arm64).
>
> Signed-off-by: Logan Gunthorpe <[email protected]>
> Cc: Palmer Dabbelt <[email protected]>
> Cc: Albert Ou <[email protected]>
> Cc: Andrew Waterman <[email protected]>
> Cc: Olof Johansson <[email protected]>
> Cc: Michael Clark <[email protected]>
> Cc: Rob Herring <[email protected]>
> Cc: Zong Li <[email protected]>
> ---
> arch/riscv/Kconfig | 23 +++++++++++++++++++++++
> arch/riscv/include/asm/pgtable.h | 24 ++++++++++++++++++++----
> arch/riscv/include/asm/sparsemem.h | 11 +++++++++++
> arch/riscv/kernel/setup.c | 4 +++-
> arch/riscv/mm/init.c | 8 ++++++++
> 5 files changed, 65 insertions(+), 5 deletions(-)
> create mode 100644 arch/riscv/include/asm/sparsemem.h
I don't really know anything about this, but you're welcome to add a
Reviewed-by: Palmer Dabbelt <[email protected]>
if you think it'll help. I'm assuming you're targeting a different tree for
the patch set, in which case it's probably best to keep this together with the
rest of it.
Thanks for porting your stuff to RISC-V!
> diff --git a/arch/riscv/Kconfig b/arch/riscv/Kconfig
> index a344980287a5..a1b5d758a542 100644
> --- a/arch/riscv/Kconfig
> +++ b/arch/riscv/Kconfig
> @@ -52,12 +52,32 @@ config ZONE_DMA32
> bool
> default y if 64BIT
>
> +config VA_BITS
> + int
> + default 32 if 32BIT
> + default 39 if 64BIT
> +
> +config PA_BITS
> + int
> + default 34 if 32BIT
> + default 56 if 64BIT
> +
> config PAGE_OFFSET
> hex
> default 0xC0000000 if 32BIT && MAXPHYSMEM_2GB
> default 0xffffffff80000000 if 64BIT && MAXPHYSMEM_2GB
> default 0xffffffe000000000 if 64BIT && MAXPHYSMEM_128GB
>
> +config ARCH_FLATMEM_ENABLE
> + def_bool y
> +
> +config ARCH_SPARSEMEM_ENABLE
> + def_bool y
> + select SPARSEMEM_VMEMMAP_ENABLE
> +
> +config ARCH_SELECT_MEMORY_MODEL
> + def_bool ARCH_SPARSEMEM_ENABLE
> +
> config STACKTRACE_SUPPORT
> def_bool y
>
> @@ -92,6 +112,9 @@ config PGTABLE_LEVELS
> config HAVE_KPROBES
> def_bool n
>
> +config HAVE_ARCH_PFN_VALID
> + def_bool y
> +
> menu "Platform type"
>
> choice
> diff --git a/arch/riscv/include/asm/pgtable.h b/arch/riscv/include/asm/pgtable.h
> index 16301966d65b..20c49cded686 100644
> --- a/arch/riscv/include/asm/pgtable.h
> +++ b/arch/riscv/include/asm/pgtable.h
> @@ -89,6 +89,26 @@ extern pgd_t swapper_pg_dir[];
> #define __S110 PAGE_SHARED_EXEC
> #define __S111 PAGE_SHARED_EXEC
>
> +#define VMALLOC_SIZE (KERN_VIRT_SIZE >> 1)
> +#define VMALLOC_END (PAGE_OFFSET - 1)
> +#define VMALLOC_START (PAGE_OFFSET - VMALLOC_SIZE)
> +
> +/*
> + * Log2 of the upper bound of the size of a struct page. Used for sizing
> + * the vmemmap region only, does not affect actual memory footprint.
> + * We don't use sizeof(struct page) directly since taking its size here
> + * requires its definition to be available at this point in the inclusion
> + * chain, and it may not be a power of 2 in the first place.
> + */
> +#define STRUCT_PAGE_MAX_SHIFT 6
> +
> +#define VMEMMAP_SIZE (UL(1) << (CONFIG_VA_BITS - PAGE_SHIFT - 1 + \
> + STRUCT_PAGE_MAX_SHIFT))
> +#define VMEMMAP_END (VMALLOC_START - 1)
> +#define VMEMMAP_START (VMALLOC_START - VMEMMAP_SIZE)
> +
> +#define vmemmap ((struct page *)VMEMMAP_START)
> +
> /*
> * ZERO_PAGE is a global shared page that is always zero,
> * used for zero-mapped memory areas, etc.
> @@ -411,10 +431,6 @@ static inline void pgtable_cache_init(void)
> /* No page table caches to initialize */
> }
>
> -#define VMALLOC_SIZE (KERN_VIRT_SIZE >> 1)
> -#define VMALLOC_END (PAGE_OFFSET - 1)
> -#define VMALLOC_START (PAGE_OFFSET - VMALLOC_SIZE)
> -
> /*
> * Task size is 0x40000000000 for RV64 or 0xb800000 for RV32.
> * Note that PGDIR_SIZE must evenly divide TASK_SIZE.
> diff --git a/arch/riscv/include/asm/sparsemem.h b/arch/riscv/include/asm/sparsemem.h
> new file mode 100644
> index 000000000000..4563e806c788
> --- /dev/null
> +++ b/arch/riscv/include/asm/sparsemem.h
> @@ -0,0 +1,11 @@
> +/* SPDX-License-Identifier: GPL-2.0 */
> +
> +#ifndef __ASM_SPARSEMEM_H
> +#define __ASM_SPARSEMEM_H
> +
> +#ifdef CONFIG_SPARSEMEM
> +#define MAX_PHYSMEM_BITS CONFIG_PA_BITS
> +#define SECTION_SIZE_BITS 30
> +#endif
> +
> +#endif
> diff --git a/arch/riscv/kernel/setup.c b/arch/riscv/kernel/setup.c
> index aee603123030..89fa781a9bf8 100644
> --- a/arch/riscv/kernel/setup.c
> +++ b/arch/riscv/kernel/setup.c
> @@ -205,6 +205,9 @@ static void __init setup_bootmem(void)
> PFN_PHYS(end_pfn - start_pfn),
> &memblock.memory, 0);
> }
> +
> + memblocks_present();
> + sparse_init();
> }
>
> void __init setup_arch(char **cmdline_p)
> @@ -239,4 +242,3 @@ void __init setup_arch(char **cmdline_p)
>
> riscv_fill_hwcap();
> }
> -
> diff --git a/arch/riscv/mm/init.c b/arch/riscv/mm/init.c
> index 58a522f9bcc3..5d529878667c 100644
> --- a/arch/riscv/mm/init.c
> +++ b/arch/riscv/mm/init.c
> @@ -70,3 +70,11 @@ void free_initrd_mem(unsigned long start, unsigned long end)
> {
> }
> #endif /* CONFIG_BLK_DEV_INITRD */
> +
> +#ifdef CONFIG_SPARSEMEM
> +int __meminit vmemmap_populate(unsigned long start, unsigned long end, int node,
> + struct vmem_altmap *altmap)
> +{
> + return vmemmap_populate_basepages(start, end, node);
> +}
> +#endif
Palmer
> I don't really know anything about this, but you're welcome to add a
>
> Reviewed-by: Palmer Dabbelt <[email protected]>
Thanks. I think it would be good to get someone who's familiar with linux/mm to take a look.
> if you think it'll help. I'm assuming you're targeting a different tree for
> the patch set, in which case it's probably best to keep this together with the
> rest of it.
No I think this series should be pulled by the RISC-V maintainer. The other patches in this series just refactor some code and need to be ACK'ed by their ARCH developers but I suspect the series should be pulled into RISC-V. That said since it does touch other arch should it be pulled by mm?
BTW note that RISC-V SPARSEMEM support is pretty useful for all manner of things and not just the p2pdma discussed in the cover.
> Thanks for porting your stuff to RISC-V!
You bet ;-)
Stephen
> + for_each_memblock(memory, reg) {
> + int nid = memblock_get_region_node(reg);
> +
> + memory_present(nid, memblock_region_memory_base_pfn(reg),
> + memblock_region_memory_end_pfn(reg));
Any reason you have a local variable for the node id while you happily
get away without one for the others? Why not simply:
for_each_memblock(memory, reg) {
memory_present(memblock_get_region_node(reg)
memblock_region_memory_base_pfn(reg),
memblock_region_memory_end_pfn(reg));
}
> +/*
> + * Log2 of the upper bound of the size of a struct page. Used for sizing
> + * the vmemmap region only, does not affect actual memory footprint.
> + * We don't use sizeof(struct page) directly since taking its size here
> + * requires its definition to be available at this point in the inclusion
> + * chain, and it may not be a power of 2 in the first place.
> + */
> +#define STRUCT_PAGE_MAX_SHIFT 6
I know this is copied from arm64, but wouldn't this be a good time
to move this next to the struct page defintion?
Also this:
arch/arm64/mm/init.c: BUILD_BUG_ON(sizeof(struct page) > (1 << STRUCT_PAGE_MAX_SHIFT));
should move to comment code (or would have to be duplicated for riscv)
> +#define VMEMMAP_SIZE (UL(1) << (CONFIG_VA_BITS - PAGE_SHIFT - 1 + \
> + STRUCT_PAGE_MAX_SHIFT))
Might be more readable with a another define, and without abuse of the
horrible UL macro:
#define VMEMMAP_SHIFT \
(CONFIG_VA_BITS - PAGE_SHIFT - 1 + STRUCT_PAGE_MAX_SHIFT)
#define VMEMMAP_SIZE (1UL << VMEMMAP_SHIFT)
> +#define VMEMMAP_END (VMALLOC_START - 1)
> +#define VMEMMAP_START (VMALLOC_START - VMEMMAP_SIZE)
> +
> +#define vmemmap ((struct page *)VMEMMAP_START)
This could also use some comments..
> @@ -0,0 +1,11 @@
> +/* SPDX-License-Identifier: GPL-2.0 */
> +
> +#ifndef __ASM_SPARSEMEM_H
> +#define __ASM_SPARSEMEM_H
> +
> +#ifdef CONFIG_SPARSEMEM
> +#define MAX_PHYSMEM_BITS CONFIG_PA_BITS
> +#define SECTION_SIZE_BITS 30
> +#endif
> +
> +#endif
For potentially wide-spanning ifdefs like inclusion headers it always
is nice to have a comment with the symbol on the endif line.
On 2018-10-11 7:37 a.m., Christoph Hellwig wrote:
>> +/*
>> + * Log2 of the upper bound of the size of a struct page. Used for sizing
>> + * the vmemmap region only, does not affect actual memory footprint.
>> + * We don't use sizeof(struct page) directly since taking its size here
>> + * requires its definition to be available at this point in the inclusion
>> + * chain, and it may not be a power of 2 in the first place.
>> + */
>> +#define STRUCT_PAGE_MAX_SHIFT 6
>
> I know this is copied from arm64, but wouldn't this be a good time
> to move this next to the struct page defintion?
>
> Also this:
>
> arch/arm64/mm/init.c: BUILD_BUG_ON(sizeof(struct page) > (1 << STRUCT_PAGE_MAX_SHIFT));
>
> should move to comment code (or would have to be duplicated for riscv)
Makes sense. Where is a good place for the BUILD_BUG_ON in common code?
I've queued up changes for your other feedback.
Thanks,
Logan
On 2018-10-11 10:24 a.m., Logan Gunthorpe wrote:
>
>
> On 2018-10-11 7:37 a.m., Christoph Hellwig wrote:
>>> +/*
>>> + * Log2 of the upper bound of the size of a struct page. Used for sizing
>>> + * the vmemmap region only, does not affect actual memory footprint.
>>> + * We don't use sizeof(struct page) directly since taking its size here
>>> + * requires its definition to be available at this point in the inclusion
>>> + * chain, and it may not be a power of 2 in the first place.
>>> + */
>>> +#define STRUCT_PAGE_MAX_SHIFT 6
>>
>> I know this is copied from arm64, but wouldn't this be a good time
>> to move this next to the struct page defintion?
Ok, I spoke too soon...
Having this define next to the struct page definition works great for
riscv. However, making that happen in arm64 seems to be a nightmare. The
include chain in arm64 is tangled up so much that including mm_types
where this is needed seems to be extremely difficult.
Unless you have any ideas, this might not be possible.
Logan
On 2018-10-11 10:24 a.m., Logan Gunthorpe wrote:
> On 2018-10-11 7:37 a.m., Christoph Hellwig wrote:
>>> +/*
>>> + * Log2 of the upper bound of the size of a struct page. Used for sizing
>>> + * the vmemmap region only, does not affect actual memory footprint.
>>> + * We don't use sizeof(struct page) directly since taking its size here
>>> + * requires its definition to be available at this point in the inclusion
>>> + * chain, and it may not be a power of 2 in the first place.
>>> + */
>>> +#define STRUCT_PAGE_MAX_SHIFT 6
>>
>> I know this is copied from arm64, but wouldn't this be a good time
>> to move this next to the struct page defintion?
>>
>> Also this:
>>
>> arch/arm64/mm/init.c: BUILD_BUG_ON(sizeof(struct page) > (1 << STRUCT_PAGE_MAX_SHIFT));
>>
>> should move to comment code (or would have to be duplicated for riscv)
>
> Makes sense. Where is a good place for the BUILD_BUG_ON in common code?
Never mind. Seems like it's pretty trivial to do this:
#define STRUCT_PAGE_MAX_SHIFT \
ilog2(roundup_pow_of_two(sizeof(struct page)))
So the BUILD_BUG_ON becomes unnecessary.
The comment saying it can't be done is really misleading as it wasn't
actually difficult.
Logan
On 2018-10-11 12:45 p.m., Logan Gunthorpe wrote:
> Ok, I spoke too soon...
>
> Having this define next to the struct page definition works great for
> riscv. However, making that happen in arm64 seems to be a nightmare. The
> include chain in arm64 is tangled up so much that including mm_types
> where this is needed seems to be extremely difficult.
Sorry for all the unnecessary churn but I've figured it out. Just had to
realize we only need mm_types.h to be included where
STRUCT_PAGE_MAX_SHIFT is finally expanded. Thus we only need it in one
more spot (fixmap.h). See below.
Thanks,
Logan
--
diff --git a/arch/arm64/include/asm/memory.h
b/arch/arm64/include/asm/memory.h
index b96442960aea..f0a5c9531e8b 100644
--- a/arch/arm64/include/asm/memory.h
+++ b/arch/arm64/include/asm/memory.h
@@ -34,15 +34,6 @@
*/
#define PCI_IO_SIZE SZ_16M
-/*
- * Log2 of the upper bound of the size of a struct page. Used for sizing
- * the vmemmap region only, does not affect actual memory footprint.
- * We don't use sizeof(struct page) directly since taking its size here
- * requires its definition to be available at this point in the inclusion
- * chain, and it may not be a power of 2 in the first place.
- */
-#define STRUCT_PAGE_MAX_SHIFT 6
-
/*
* VMEMMAP_SIZE - allows the whole linear region to be covered by
* a struct page array
diff --git a/include/asm-generic/fixmap.h b/include/asm-generic/fixmap.h
index 827e4d3bbc7a..8cc7b09c1bc7 100644
--- a/include/asm-generic/fixmap.h
+++ b/include/asm-generic/fixmap.h
@@ -16,6 +16,7 @@
#define __ASM_GENERIC_FIXMAP_H
#include <linux/bug.h>
+#include <linux/mm_types.h>
#define __fix_to_virt(x) (FIXADDR_TOP - ((x) << PAGE_SHIFT))
#define __virt_to_fix(x) ((FIXADDR_TOP - ((x)&PAGE_MASK)) >>
PAGE_SHIFT)
diff --git a/include/linux/mm_types.h b/include/linux/mm_types.h
index 5ed8f6292a53..d1c3cde8c201 100644
--- a/include/linux/mm_types.h
+++ b/include/linux/mm_types.h
@@ -206,6 +206,11 @@ struct page {
#endif
} _struct_page_alignment;
+/*
+ * Used for sizing the vmemmap region on some architectures.
+ */
+#define STRUCT_PAGE_MAX_SHIFT ilog2(roundup_pow_of_two(sizeof(struct
page)))
+
#define PAGE_FRAG_CACHE_MAX_SIZE __ALIGN_MASK(32768, ~PAGE_MASK)
#define PAGE_FRAG_CACHE_MAX_ORDER get_order(PAGE_FRAG_CACHE_MAX_SIZE)
On Thu, 11 Oct 2018 05:18:20 PDT (-0700), [email protected] wrote:
> Palmer
>
>> I don't really know anything about this, but you're welcome to add a
>>
>> Reviewed-by: Palmer Dabbelt <[email protected]>
>
> Thanks. I think it would be good to get someone who's familiar with linux/mm to take a look.
>
>> if you think it'll help. I'm assuming you're targeting a different tree for
>> the patch set, in which case it's probably best to keep this together with the
>> rest of it.
>
> No I think this series should be pulled by the RISC-V maintainer. The other patches in this series just refactor some code and need to be ACK'ed by their ARCH developers but I suspect the series should be pulled into RISC-V. That said since it does touch other arch should it be pulled by mm?
>
> BTW note that RISC-V SPARSEMEM support is pretty useful for all manner of things and not just the p2pdma discussed in the cover.
Ah, OK -- I thought this was adding the support everywhere. Do you mind
re-sending the patches with the various acks/reviews and I'll put in on
for-next?
>
>> Thanks for porting your stuff to RISC-V!
>
> You bet ;-)