mm: Enable CONFIG_NODES_SPAN_OTHER_NODES by default for NUMA
Subsystems (10)
v3 cover mm: Enable CONFIG_NODES_SPAN_OTHER_NODES by default for NUMA
In NUMA layout which nodes have memory ranges that span across other nodes, the mm driver can detect the memory node id incorrectly.
For example, with layout below
Node 0 address: 0000 xxxx 0000 xxxx
Node 1 address: xxxx 1111 xxxx 1111
Note:
- Memory from low to high
- 0/1: Node id
- x: Invalid memory of a node
When mm probes the memory map, without CONFIG_NODES_SPAN_OTHER_NODES config, mm only checks the memory validity but not the node id. Because of that, Node 1 also detects the memory from node 0 as below when it scans from the start address to the end address of node 1.
Node 0 address: 0000 xxxx xxxx xxxx
Node 1 address: xxxx 1111 1111 1111
This layout could occur on any architecture. Most of them enables this config by default with CONFIG_NUMA. This patch, by default, enables CONFIG_NODES_SPAN_OTHER_NODES or uses early_pfn_in_nid() for NUMA.
v3:
* Revise the patch description
V2:
* Revise the patch description
Hoan Tran (5):
mm: Enable CONFIG_NODES_SPAN_OTHER_NODES by default for NUMA
powerpc: Kconfig: Remove CONFIG_NODES_SPAN_OTHER_NODES
x86: Kconfig: Remove CONFIG_NODES_SPAN_OTHER_NODES
sparc: Kconfig: Remove CONFIG_NODES_SPAN_OTHER_NODES
s390: Kconfig: Remove CONFIG_NODES_SPAN_OTHER_NODES
arch/powerpc/Kconfig | 9 --------- arch/s390/Kconfig | 8 -------- arch/sparc/Kconfig | 9 --------- arch/x86/Kconfig | 9 --------- mm/page_alloc.c | 2 +- 5 files changed, 1 insertion(+), 36 deletions(-)
Signature · 1 line
1.8.3.1
PATCH v3 1/5 mm: Enable CONFIG_NODES_SPAN_OTHER_NODES by default for NUMA
In NUMA layout which nodes have memory ranges that span across other nodes, the mm driver can detect the memory node id incorrectly.
For example, with layout below
Node 0 address: 0000 xxxx 0000 xxxx
Node 1 address: xxxx 1111 xxxx 1111
Note:
- Memory from low to high
- 0/1: Node id
- x: Invalid memory of a node
When mm probes the memory map, without CONFIG_NODES_SPAN_OTHER_NODES config, mm only checks the memory validity but not the node id. Because of that, Node 1 also detects the memory from node 0 as below when it scans from the start address to the end address of node 1.
Node 0 address: 0000 xxxx xxxx xxxx
Node 1 address: xxxx 1111 1111 1111
This layout could occur on any architecture. Most of them enables this config by default with CONFIG_NUMA. This patch, by default, enables CONFIG_NODES_SPAN_OTHER_NODES or uses early_pfn_in_nid() for NUMA.
--- mm/page_alloc.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
mm/page_alloc.c +1 -1
@@ -1467,7 +1467,7 @@ int __meminit early_pfn_to_nid(unsigned long pfn) }#endif -#ifdef CONFIG_NODES_SPAN_OTHER_NODES +#ifdef CONFIG_NUMA /* Only safe to use early in boot when initialisation is single-threaded */ static inline bool __meminit early_pfn_in_nid (unsigned long pfn,int node) {
Signature · 1 line
1.8.3.1
PATCH v3 2/5 powerpc: Kconfig: Remove CONFIG_NODES_SPAN_OTHER_NODES
In reply to Hoan TranRemove CONFIG_NODES_SPAN_OTHER_NODES as it's enabled by default with NUMA.
--- arch/powerpc/Kconfig | 9 --------- 1 file changed, 9 deletions(-)
arch/powerpc/Kconfig +0 -9
@@ -686,15 +686,6 @@ config ARCH_MEMORY_PROBE def_bool y depends on MEMORY_HOTPLUG -# Some NUMA nodes have memory ranges that span -# other nodes. Even though a pfn is valid and -# between a node's start and end pfns, it may not -# reside on that node. See memmap_init_zone() -# for details. -config NODES_SPAN_OTHER_NODES - def_bool y - depends on NEED_MULTIPLE_NODES - config STDBINUTILS bool "Using standard binutils settings" depends on 44x
Signature · 1 line
1.8.3.1
PATCH v3 3/5 x86: Kconfig: Remove CONFIG_NODES_SPAN_OTHER_NODES
In reply to Hoan TranRemove CONFIG_NODES_SPAN_OTHER_NODES as it's enabled by default. It is now enabled for x86(32 bit) configurations and do not depend on X64_64_ACPI_NUMA config. Because of that, on NUMA enabled system, early_pfn_in_nid() function is called by memmap_init_zone() during boot-time. It doesn't affect the performance at run-time.
--- arch/x86/Kconfig | 9 --------- 1 file changed, 9 deletions(-)
arch/x86/Kconfig +0 -9
@@ -1581,15 +1581,6 @@ config X86_64_ACPI_NUMA ---help--- Enable ACPI SRAT based node topology detection. -# Some NUMA nodes have memory ranges that span -# other nodes. Even though a pfn is valid and -# between a node's start and end pfns, it may not -# reside on that node. See memmap_init_zone() -# for details. -config NODES_SPAN_OTHER_NODES - def_bool y - depends on X86_64_ACPI_NUMA - config NUMA_EMU bool "NUMA emulation" depends on NUMA
Signature · 1 line
1.8.3.1
PATCH v3 4/5 sparc: Kconfig: Remove CONFIG_NODES_SPAN_OTHER_NODES
In reply to Hoan TranRemove CONFIG_NODES_SPAN_OTHER_NODES as it's enabled by default with NUMA.
--- arch/sparc/Kconfig | 9 --------- 1 file changed, 9 deletions(-)
arch/sparc/Kconfig +0 -9
@@ -292,15 +292,6 @@ config NODES_SHIFT Specify the maximum number of NUMA Nodes available on the target system. Increases memory reserved to accommodate various tables. -# Some NUMA nodes have memory ranges that span -# other nodes. Even though a pfn is valid and -# between a node's start and end pfns, it may not -# reside on that node. See memmap_init_zone() -# for details. -config NODES_SPAN_OTHER_NODES - def_bool y - depends on NEED_MULTIPLE_NODES - config ARCH_SPARSEMEM_ENABLE def_bool y if SPARC64 select SPARSEMEM_VMEMMAP_ENABLE
Signature · 1 line
1.8.3.1
PATCH v3 5/5 s390: Kconfig: Remove CONFIG_NODES_SPAN_OTHER_NODES
In reply to Hoan TranRemove CONFIG_NODES_SPAN_OTHER_NODES as it's enabled by default with NUMA.
--- arch/s390/Kconfig | 8 -------- 1 file changed, 8 deletions(-)
arch/s390/Kconfig +0 -8
@@ -449,14 +449,6 @@ config NR_CPUS config HOTPLUG_CPU def_bool y -# Some NUMA nodes have memory ranges that span -# other nodes. Even though a pfn is valid and -# between a node's start and end pfns, it may not -# reside on that node. See memmap_init_zone() -# for details. <- They meant memory holes! -config NODES_SPAN_OTHER_NODES - def_bool NUMA - config NUMA bool "NUMA support" depends on SCHED_TOPOLOGY
Signature · 1 line
1.8.3.1
Quoted 6 lines · Hoan Tran
In NUMA layout which nodes have memory ranges that span across other nodes, the mm driver can detect the memory node id incorrectly.
For example, with layout below
Node 0 address: 0000 xxxx 0000 xxxx
Node 1 address: xxxx 1111 xxxx 1111
Sorry, I read this example several times, but still don't get what it means. Can it be given with real hex number address as an exmaple? I mean just using the memory layout you have seen from some systems. The change looks interesting though.
Quoted 42 lines
Note:
- Memory from low to high
- 0/1: Node id
- x: Invalid memory of a nodeWhen mm probes the memory map, without
CONFIG_NODES_SPAN_OTHER_NODESconfig, mm only checks the memory validity but not the node id. Because of that, Node 1 also detects the memory from node 0 as below when it scans from the start address to the end address of node 1.Node 0 address: 0000 xxxx xxxx xxxx
Node 1 address: xxxx 1111 1111 1111This layout could occur on any architecture. Most of them enables this config by default with
CONFIG_NUMA. This patch, by default, enablesCONFIG_NODES_SPAN_OTHER_NODESor usesearly_pfn_in_nid()for NUMA.v3:
* Revise the patch descriptionV2:
* Revise the patch descriptionHoan Tran (5):
mm: EnableCONFIG_NODES_SPAN_OTHER_NODESby default for NUMA
powerpc:Kconfig: RemoveCONFIG_NODES_SPAN_OTHER_NODES
x86:Kconfig: RemoveCONFIG_NODES_SPAN_OTHER_NODES
sparc:Kconfig: RemoveCONFIG_NODES_SPAN_OTHER_NODES
s390:Kconfig: RemoveCONFIG_NODES_SPAN_OTHER_NODESarch/powerpc/Kconfig | 9 --------- arch/s390/Kconfig | 8 -------- arch/sparc/Kconfig | 9 --------- arch/x86/Kconfig | 9 --------- mm/page_alloc.c | 2 +- 5 files changed, 1 insertion(+), 36 deletions(-)Signature · 1 line
1.8.3.1
Quoted 4 lines · Baoquan He
Quoted 3 lines · Hoan Tran
In NUMA layout which nodes have memory ranges that span across other nodes, the mm driver can detect the memory node id incorrectly.
Quoted 8 lines
Quoted 3 lines
For example, with layout below
Node 0 address: 0000 xxxx 0000 xxxx
Node 1 address: xxxx 1111 xxxx 1111Sorry, I read this example several times, but still don't get what it means. Can it be given with real hex number address as an exmaple? I mean just using the memory layout you have seen from some systems. The change looks interesting though.
Does this make it more clear?
physical address range and its node associaion
[0 0 0 0 1 1 1 1 0 0 0 0 1 1 1 1 0 0 0 0 1 1 1 1]
Signature · 2 lines
Michal Hocko SUSE Labs
Quoted 9 lines · Michal Hocko
Quoted 8 lines · Baoquan He
Quoted 6 lines · Hoan Tran
In NUMA layout which nodes have memory ranges that span across other nodes, the mm driver can detect the memory node id incorrectly.
For example, with layout below
Node 0 address: 0000 xxxx 0000 xxxx
Node 1 address: xxxx 1111 xxxx 1111
Quoted 8 lines
Quoted 4 lines
Sorry, I read this example several times, but still don't get what it means. Can it be given with real hex number address as an exmaple? I mean just using the memory layout you have seen from some systems. The change looks interesting though.
Does this make it more clear?
physical address range and its node associaion
[0 0 0 0 1 1 1 1 0 0 0 0 1 1 1 1 0 0 0 0 1 1 1 1]
I later read it again, have got what Hoan is trying to say, thanks.
I think the change in this patchset makes sense, still have some concern though, let me add comment in other thread.
Quoted 17 lines · Hoan Tran
In NUMA layout which nodes have memory ranges that span across other nodes, the mm driver can detect the memory node id incorrectly.
For example, with layout below
Node 0 address: 0000 xxxx 0000 xxxx
Node 1 address: xxxx 1111 xxxx 1111Note:
- Memory from low to high
- 0/1: Node id
- x: Invalid memory of a nodeWhen mm probes the memory map, without
CONFIG_NODES_SPAN_OTHER_NODESconfig, mm only checks the memory validity but not the node id. Because of that, Node 1 also detects the memory from node 0 as below when it scans from the start address to the end address of node 1.
Quoted 6 lines
Node 0 address: 0000 xxxx xxxx xxxx
Node 1 address: xxxx 1111 1111 1111This layout could occur on any architecture. Most of them enables this config by default with
CONFIG_NUMA. This patch, by default, enablesCONFIG_NODES_SPAN_OTHER_NODESor usesearly_pfn_in_nid()for NUMA.
I am not opposed to this at all. It reduces the config space and that is a good thing on its own. The history has shown that meory layout might be really wild wrt NUMA. The config is only used for early_pfn_in_nid which is clearly an overkill.
Your description doesn't really explain why this is safe though. The history of this config is somehow messy, though. Mike has tried to remove it a94b3ab7eab4 ("[PATCH] mm: remove arch independent NODES_SPAN_OTHER_NODES") just to be reintroduced by 7516795739bd ("[PATCH] Reintroduce NODES_SPAN_OTHER_NODES for powerpc") without any reasoning what so ever. This doesn't make it really easy see whether reasons for reintroduction are still there. Maybe there are some subtle dependencies. I do not see any TBH but that might be burried deep in an arch specific code.
Quoted 23 lines
v3:
* Revise the patch descriptionV2:
* Revise the patch descriptionHoan Tran (5):
mm: EnableCONFIG_NODES_SPAN_OTHER_NODESby default for NUMA
powerpc:Kconfig: RemoveCONFIG_NODES_SPAN_OTHER_NODES
x86:Kconfig: RemoveCONFIG_NODES_SPAN_OTHER_NODES
sparc:Kconfig: RemoveCONFIG_NODES_SPAN_OTHER_NODES
s390:Kconfig: RemoveCONFIG_NODES_SPAN_OTHER_NODESarch/powerpc/Kconfig | 9 --------- arch/s390/Kconfig | 8 -------- arch/sparc/Kconfig | 9 --------- arch/x86/Kconfig | 9 --------- mm/page_alloc.c | 2 +- 5 files changed, 1 insertion(+), 36 deletions(-)Signature · 1 line
1.8.3.1
Signature · 2 lines
Michal Hocko SUSE Labs
Quoted 30 lines · Michal Hocko
Quoted 17 lines · Hoan Tran
In NUMA layout which nodes have memory ranges that span across other nodes, the mm driver can detect the memory node id incorrectly.
For example, with layout below
Node 0 address: 0000 xxxx 0000 xxxx
Node 1 address: xxxx 1111 xxxx 1111Note:
- Memory from low to high
- 0/1: Node id
- x: Invalid memory of a nodeWhen mm probes the memory map, without
CONFIG_NODES_SPAN_OTHER_NODESconfig, mm only checks the memory validity but not the node id. Because of that, Node 1 also detects the memory from node 0 as below when it scans from the start address to the end address of node 1.Quoted 6 lines
Node 0 address: 0000 xxxx xxxx xxxx
Node 1 address: xxxx 1111 1111 1111This layout could occur on any architecture. Most of them enables this config by default with
CONFIG_NUMA. This patch, by default, enablesCONFIG_NODES_SPAN_OTHER_NODESor usesearly_pfn_in_nid()for NUMA.I am not opposed to this at all. It reduces the config space and that is a good thing on its own. The history has shown that meory layout might be really wild wrt NUMA. The config is only used for
early_pfn_in_nidwhich is clearly an overkill.
Quoted 9 lines
Your description doesn't really explain why this is safe though. The history of this config is somehow messy, though. Mike has tried to remove it
a94b3ab7eab4("[PATCH] mm: remove arch independentNODES_SPAN_OTHER_NODES") just to be reintroduced by7516795739bd("[PATCH] ReintroduceNODES_SPAN_OTHER_NODESfor powerpc") without any reasoning what so ever. This doesn't make it really easy see whether reasons for reintroduction are still there. Maybe there are some subtle dependencies. I do not see any TBH but that might be burried deep in an arch specific code.
Yeah, since early_pfnnid_cache was added, we do not need worry about the performance. But when I read the mem init code on x86 again, I do see there are codes to handle the node overlapping, e.g in numa_cleanup_meminfo(), when store node id into memblock. But the thing is if we have encountered the node overlapping, we just return ahead of time, leave something uninitialized. I am wondering if the system with node overlapping can still run heathily.
Quoted 29 lines
Quoted 24 lines
v3:
* Revise the patch descriptionV2:
* Revise the patch descriptionHoan Tran (5):
mm: EnableCONFIG_NODES_SPAN_OTHER_NODESby default for NUMA
powerpc:Kconfig: RemoveCONFIG_NODES_SPAN_OTHER_NODES
x86:Kconfig: RemoveCONFIG_NODES_SPAN_OTHER_NODES
sparc:Kconfig: RemoveCONFIG_NODES_SPAN_OTHER_NODES
s390:Kconfig: RemoveCONFIG_NODES_SPAN_OTHER_NODESarch/powerpc/Kconfig | 9 --------- arch/s390/Kconfig | 8 -------- arch/sparc/Kconfig | 9 --------- arch/x86/Kconfig | 9 --------- mm/page_alloc.c | 2 +- 5 files changed, 1 insertion(+), 36 deletions(-)Signature · 1 line
1.8.3.1Signature · 2 lines
Michal Hocko SUSE Labs
Quoted 41 lines · Baoquan He
Quoted 30 lines · Michal Hocko
Quoted 17 lines · Hoan Tran
In NUMA layout which nodes have memory ranges that span across other nodes, the mm driver can detect the memory node id incorrectly.
For example, with layout below
Node 0 address: 0000 xxxx 0000 xxxx
Node 1 address: xxxx 1111 xxxx 1111Note:
- Memory from low to high
- 0/1: Node id
- x: Invalid memory of a nodeWhen mm probes the memory map, without
CONFIG_NODES_SPAN_OTHER_NODESconfig, mm only checks the memory validity but not the node id. Because of that, Node 1 also detects the memory from node 0 as below when it scans from the start address to the end address of node 1.Quoted 6 lines
Node 0 address: 0000 xxxx xxxx xxxx
Node 1 address: xxxx 1111 1111 1111This layout could occur on any architecture. Most of them enables this config by default with
CONFIG_NUMA. This patch, by default, enablesCONFIG_NODES_SPAN_OTHER_NODESor usesearly_pfn_in_nid()for NUMA.I am not opposed to this at all. It reduces the config space and that is a good thing on its own. The history has shown that meory layout might be really wild wrt NUMA. The config is only used for
early_pfn_in_nidwhich is clearly an overkill.Quoted 9 lines
Your description doesn't really explain why this is safe though. The history of this config is somehow messy, though. Mike has tried to remove it
a94b3ab7eab4("[PATCH] mm: remove arch independentNODES_SPAN_OTHER_NODES") just to be reintroduced by7516795739bd("[PATCH] ReintroduceNODES_SPAN_OTHER_NODESfor powerpc") without any reasoning what so ever. This doesn't make it really easy see whether reasons for reintroduction are still there. Maybe there are some subtle dependencies. I do not see any TBH but that might be burried deep in an arch specific code.
Quoted 7 lines
Yeah, since
early_pfnnid_cachewas added, we do not need worry about the performance. But when I read the mem init code on x86 again, I do see there are codes to handle the node overlapping, e.g innuma_cleanup_meminfo(), when store node id into memblock. But the thing is if we have encountered the node overlapping, we just return ahead of time, leave something uninitialized. I am wondering if the system with node overlapping can still run heathily.
Ok, I didn't read code carefully. That is handling case where memblock with different node id overlap, it needs return. In the example Hoan gave, it has no problem, system can run well. Please ignore above comment.
Quoted 30 lines · Michal Hocko
Quoted 17 lines · Hoan Tran
In NUMA layout which nodes have memory ranges that span across other nodes, the mm driver can detect the memory node id incorrectly.
For example, with layout below
Node 0 address: 0000 xxxx 0000 xxxx
Node 1 address: xxxx 1111 xxxx 1111Note:
- Memory from low to high
- 0/1: Node id
- x: Invalid memory of a nodeWhen mm probes the memory map, without
CONFIG_NODES_SPAN_OTHER_NODESconfig, mm only checks the memory validity but not the node id. Because of that, Node 1 also detects the memory from node 0 as below when it scans from the start address to the end address of node 1.Quoted 6 lines
Node 0 address: 0000 xxxx xxxx xxxx
Node 1 address: xxxx 1111 1111 1111This layout could occur on any architecture. Most of them enables this config by default with
CONFIG_NUMA. This patch, by default, enablesCONFIG_NODES_SPAN_OTHER_NODESor usesearly_pfn_in_nid()for NUMA.I am not opposed to this at all. It reduces the config space and that is a good thing on its own. The history has shown that meory layout might be really wild wrt NUMA. The config is only used for
early_pfn_in_nidwhich is clearly an overkill.
Quoted 9 lines
Your description doesn't really explain why this is safe though. The history of this config is somehow messy, though. Mike has tried to remove it
a94b3ab7eab4("[PATCH] mm: remove arch independentNODES_SPAN_OTHER_NODES") just to be reintroduced by7516795739bd("[PATCH] ReintroduceNODES_SPAN_OTHER_NODESfor powerpc") without any reasoning what so ever. This doesn't make it really easy see whether reasons for reintroduction are still there. Maybe there are some subtle dependencies. I do not see any TBH but that might be burried deep in an arch specific code.
Well, back then early_pfn_in_nid() was arch-dependant, today everyone except ia64 rely on HAVE_MEMBLOCK_NODE_MAP. So, if the memblock node map is correct, that using CONFIG_NUMA instead of CONFIG_NODES_SPAN_OTHER_NODES would only mean that early_pfn_in_nid() will cost several cycles more on architectures that didn't select CONFIG_NODES_SPAN_OTHER_NODES (i.e. arm64 and sh).
Agian, ia64 is an exception here.
Quoted 27 lines
Quoted 23 lines
v3:
* Revise the patch descriptionV2:
* Revise the patch descriptionHoan Tran (5):
mm: EnableCONFIG_NODES_SPAN_OTHER_NODESby default for NUMA
powerpc:Kconfig: RemoveCONFIG_NODES_SPAN_OTHER_NODES
x86:Kconfig: RemoveCONFIG_NODES_SPAN_OTHER_NODES
sparc:Kconfig: RemoveCONFIG_NODES_SPAN_OTHER_NODES
s390:Kconfig: RemoveCONFIG_NODES_SPAN_OTHER_NODESarch/powerpc/Kconfig | 9 --------- arch/s390/Kconfig | 8 -------- arch/sparc/Kconfig | 9 --------- arch/x86/Kconfig | 9 --------- mm/page_alloc.c | 2 +- 5 files changed, 1 insertion(+), 36 deletions(-)Signature · 1 line
1.8.3.1Signature · 2 lines
Michal Hocko SUSE Labs
Signature · 2 lines
Sincerely yours, Mike.
Quoted 41 lines · Mike Rapoport
Quoted 30 lines · Michal Hocko
Quoted 17 lines · Hoan Tran
In NUMA layout which nodes have memory ranges that span across other nodes, the mm driver can detect the memory node id incorrectly.
For example, with layout below
Node 0 address: 0000 xxxx 0000 xxxx
Node 1 address: xxxx 1111 xxxx 1111Note:
- Memory from low to high
- 0/1: Node id
- x: Invalid memory of a nodeWhen mm probes the memory map, without
CONFIG_NODES_SPAN_OTHER_NODESconfig, mm only checks the memory validity but not the node id. Because of that, Node 1 also detects the memory from node 0 as below when it scans from the start address to the end address of node 1.Quoted 6 lines
Node 0 address: 0000 xxxx xxxx xxxx
Node 1 address: xxxx 1111 1111 1111This layout could occur on any architecture. Most of them enables this config by default with
CONFIG_NUMA. This patch, by default, enablesCONFIG_NODES_SPAN_OTHER_NODESor usesearly_pfn_in_nid()for NUMA.I am not opposed to this at all. It reduces the config space and that is a good thing on its own. The history has shown that meory layout might be really wild wrt NUMA. The config is only used for
early_pfn_in_nidwhich is clearly an overkill.Quoted 9 lines
Your description doesn't really explain why this is safe though. The history of this config is somehow messy, though. Mike has tried to remove it
a94b3ab7eab4("[PATCH] mm: remove arch independentNODES_SPAN_OTHER_NODES") just to be reintroduced by7516795739bd("[PATCH] ReintroduceNODES_SPAN_OTHER_NODESfor powerpc") without any reasoning what so ever. This doesn't make it really easy see whether reasons for reintroduction are still there. Maybe there are some subtle dependencies. I do not see any TBH but that might be burried deep in an arch specific code.
Quoted 2 lines
Well, back then
early_pfn_in_nid()was arch-dependant, today everyone except ia64 rely onHAVE_MEMBLOCK_NODE_MAP.
What would it take to make ia64 use HAVE_MEMBLOCK_NODE_MAP? I would really love to see that thing go away. It is causing problems when people try to use memblock api.
Quoted 5 lines
So, if the memblock node map
is correct, that usingCONFIG_NUMAinstead ofCONFIG_NODES_SPAN_OTHER_NODESwould only mean thatearly_pfn_in_nid()will cost several cycles more on architectures that didn't selectCONFIG_NODES_SPAN_OTHER_NODES(i.e. arm64 and sh).
Do we have any idea on how much of an overhead that is? Because this is per each pfn so it can accumulate a lot!
Quoted 1 line
Agian, ia64 is an exception here.
Thanks for the clarification!
Signature · 2 lines
Michal Hocko SUSE Labs
Quoted 42 lines · Michal Hocko
Quoted 41 lines · Mike Rapoport
Quoted 39 lines · Michal Hocko
Quoted 17 lines · Hoan Tran
In NUMA layout which nodes have memory ranges that span across other nodes, the mm driver can detect the memory node id incorrectly.
For example, with layout below
Node 0 address: 0000 xxxx 0000 xxxx
Node 1 address: xxxx 1111 xxxx 1111Note:
- Memory from low to high
- 0/1: Node id
- x: Invalid memory of a nodeWhen mm probes the memory map, without
CONFIG_NODES_SPAN_OTHER_NODESconfig, mm only checks the memory validity but not the node id. Because of that, Node 1 also detects the memory from node 0 as below when it scans from the start address to the end address of node 1.Quoted 6 lines
Node 0 address: 0000 xxxx xxxx xxxx
Node 1 address: xxxx 1111 1111 1111This layout could occur on any architecture. Most of them enables this config by default with
CONFIG_NUMA. This patch, by default, enablesCONFIG_NODES_SPAN_OTHER_NODESor usesearly_pfn_in_nid()for NUMA.I am not opposed to this at all. It reduces the config space and that is a good thing on its own. The history has shown that meory layout might be really wild wrt NUMA. The config is only used for
early_pfn_in_nidwhich is clearly an overkill.Your description doesn't really explain why this is safe though. The history of this config is somehow messy, though. Mike has tried to remove it
a94b3ab7eab4("[PATCH] mm: remove arch independentNODES_SPAN_OTHER_NODES") just to be reintroduced by7516795739bd("[PATCH] ReintroduceNODES_SPAN_OTHER_NODESfor powerpc") without any reasoning what so ever. This doesn't make it really easy see whether reasons for reintroduction are still there. Maybe there are some subtle dependencies. I do not see any TBH but that might be burried deep in an arch specific code.
Quoted 6 lines
Quoted 2 lines
Well, back then
early_pfn_in_nid()was arch-dependant, today everyone except ia64 rely onHAVE_MEMBLOCK_NODE_MAP.What would it take to make ia64 use
HAVE_MEMBLOCK_NODE_MAP? I would really love to see that thing go away. It is causing problems when people try to use memblock api.
Sorry, my bad, ia64 does not have NODES_SPAN_OTHER_NODES, but it does have HAVE_MEMBLOCK_NODE_MAP.
I remember I've tried killing HAVE_MEMBLOCK_NODE_MAP, but I've run into some problems and then I've got distracted. I too would like to have HAVE_MEMBLOCK_NODE_MAP go away, maybe I'll take another look at it.
Quoted 8 lines
Quoted 5 lines
So, if the memblock node map
is correct, that usingCONFIG_NUMAinstead ofCONFIG_NODES_SPAN_OTHER_NODESwould only mean thatearly_pfn_in_nid()will cost several cycles more on architectures that didn't selectCONFIG_NODES_SPAN_OTHER_NODES(i.e. arm64 and sh).Do we have any idea on how much of an overhead that is? Because this is per each pfn so it can accumulate a lot!
It's O(log(N)) where N is the amount of the memory banks (ie. memblock.memory.cnt)
Quoted 6 lines
Quoted 1 line
Agian, ia64 is an exception here.
Thanks for the clarification!
Signature · 2 lines
Michal Hocko SUSE Labs
Signature · 2 lines
Sincerely yours, Mike.
Quoted 63 lines · Mike Rapoport
Quoted 42 lines · Michal Hocko
Quoted 41 lines · Mike Rapoport
Quoted 39 lines · Michal Hocko
Quoted 17 lines · Hoan Tran
In NUMA layout which nodes have memory ranges that span across other nodes, the mm driver can detect the memory node id incorrectly.
For example, with layout below
Node 0 address: 0000 xxxx 0000 xxxx
Node 1 address: xxxx 1111 xxxx 1111Note:
- Memory from low to high
- 0/1: Node id
- x: Invalid memory of a nodeWhen mm probes the memory map, without
CONFIG_NODES_SPAN_OTHER_NODESconfig, mm only checks the memory validity but not the node id. Because of that, Node 1 also detects the memory from node 0 as below when it scans from the start address to the end address of node 1.Quoted 6 lines
Node 0 address: 0000 xxxx xxxx xxxx
Node 1 address: xxxx 1111 1111 1111This layout could occur on any architecture. Most of them enables this config by default with
CONFIG_NUMA. This patch, by default, enablesCONFIG_NODES_SPAN_OTHER_NODESor usesearly_pfn_in_nid()for NUMA.I am not opposed to this at all. It reduces the config space and that is a good thing on its own. The history has shown that meory layout might be really wild wrt NUMA. The config is only used for
early_pfn_in_nidwhich is clearly an overkill.Your description doesn't really explain why this is safe though. The history of this config is somehow messy, though. Mike has tried to remove it
a94b3ab7eab4("[PATCH] mm: remove arch independentNODES_SPAN_OTHER_NODES") just to be reintroduced by7516795739bd("[PATCH] ReintroduceNODES_SPAN_OTHER_NODESfor powerpc") without any reasoning what so ever. This doesn't make it really easy see whether reasons for reintroduction are still there. Maybe there are some subtle dependencies. I do not see any TBH but that might be burried deep in an arch specific code.Quoted 6 lines
Quoted 2 lines
Well, back then
early_pfn_in_nid()was arch-dependant, today everyone except ia64 rely onHAVE_MEMBLOCK_NODE_MAP.What would it take to make ia64 use
HAVE_MEMBLOCK_NODE_MAP? I would really love to see that thing go away. It is causing problems when people try to use memblock api.Sorry, my bad, ia64 does not have
NODES_SPAN_OTHER_NODES, but it does haveHAVE_MEMBLOCK_NODE_MAP.I remember I've tried killing
HAVE_MEMBLOCK_NODE_MAP, but I've run into some problems and then I've got distracted. I too would like to haveHAVE_MEMBLOCK_NODE_MAPgo away, maybe I'll take another look at it.Quoted 6 lines
Quoted 5 lines
So, if the memblock node map
is correct, that usingCONFIG_NUMAinstead ofCONFIG_NODES_SPAN_OTHER_NODESwould only mean thatearly_pfn_in_nid()will cost several cycles more on architectures that didn't selectCONFIG_NODES_SPAN_OTHER_NODES(i.e. arm64 and sh).
Quoted 4 lines
Quoted 2 lines
Do we have any idea on how much of an overhead that is? Because this is per each pfn so it can accumulate a lot!
It's O(log(N)) where N is the amount of the memory banks (ie. memblock.memory.cnt)
This is for the Node id searching. But early_pfn_in_nid() is calling for each pfn, this is the big one, I think. Otherwise, it may be optimized as no-op.
Quoted 13 lines
Quoted 7 lines
Quoted 2 lines
Agian, ia64 is an exception here.
Thanks for the clarification!
Signature · 2 lines
Michal Hocko SUSE LabsSignature · 2 lines
Sincerely yours, Mike.
RFC PATCH 1/1 mm: remove CONFIG_HAVE_MEMBLOCK_NODE_MAP (was: Re: [PATCH v3 0/5] mm: Enable CONFIG_NODES_SPAN_OTHER_NODES by default for NUMA)
In reply to Michal HockoQuoted 4 lines · Michal Hocko
What would it take to make ia64 use
HAVE_MEMBLOCK_NODE_MAP? I would really love to see that thing go away. It is causing problems when people try to use memblock api.
Well, it's a small patch in the end :)
Currently all NUMA architectures currently enable CONFIG_HAVE_MEMBLOCK_NODE_MAP and use free_area_init_nodes() to initialize nodes and zones structures.
On the other hand, the systems that don't have CONFIG_HAVE_MEMBLOCK_NODE_MAP use free_area_init() or free_area_init_node() for this purpose.
With this assumptions, it's possible to make selection of the functions that calculate spanned and absent pages at runtime.
This patch builds for arm and x86-64 and boots on qemu-system for both.
From f907df987db4d6735c4940b30cfb4764fc0007d4 Mon Sep 17 00:00:00 2001
From: Mike Rapoport <[email protected]>
Date: Wed, 1 Apr 2020 00:27:17 +0300
Subject: [PATCH RFC] mm: remove CONFIG_HAVE_MEMBLOCK_NODE_MAP option
The CONFIG_HAVE_MEMBLOCK_NODE_MAP is used to differentiate initialization of nodes and zones structures between the systems that have region to node mapping in memblock and those that don't.
Currently all the NUMA architectures enable this option and for the non-NUMA systems we can presume that all the memory belongs to node 0 and therefore the compile time configuration option is not required.
Still, free_area_init_node() must have a backward compatible version because its semantics with and without CONFIG_HAVE_MEMBLOCK_NODE_MAP is different. Once all the architectures will be converted from free_area_init() and free_area_init_node() to free_area_init_nodes(), the entire compatibility layer can be dropped.
--- include/linux/memblock.h | 15 ------ include/linux/mm.h | 16 +----- include/linux/mmzone.h | 13 ----- mm/memblock.c | 9 +--- mm/memory_hotplug.c | 4 -- mm/page_alloc.c | 103 ++++++++++++++++++++++----------------- 6 files changed, 61 insertions(+), 99 deletions(-)
include/linux/memblock.h +0 -15
@@ -50,9 +50,7 @@ struct memblock_region {phys_addr_t base;phys_addr_t size;enum memblock_flags flags; -#ifdef CONFIG_HAVE_MEMBLOCK_NODE_MAP int nid; -#endif };/**
@@ -215,7 +213,6 @@ static inline bool memblock_is_nomap(struct memblock_region *m)return m->flags &MEMBLOCK_NOMAP ; } -#ifdef CONFIG_HAVE_MEMBLOCK_NODE_MAP int memblock_search_pfn_nid (unsigned long pfn,unsigned long *start_pfn,unsigned long *end_pfn); void __next_mem_pfn_range(int *idx, int nid, unsigned long *out_start_pfn,
@@ -234,7 +231,6 @@ void __next_mem_pfn_range(int *idx, int nid, unsigned long *out_start_pfn,#define for_each_mem_pfn_range (i, nid, p_start, p_end, p_nid) \ for (i = -1, __next_mem_pfn_range(&i, nid, p_start, p_end, p_nid); \ i >= 0; __next_mem_pfn_range(&i, nid, p_start, p_end, p_nid)) -#endif /* CONFIG_HAVE_MEMBLOCK_NODE_MAP */ #ifdef CONFIG_DEFERRED_STRUCT_PAGE_INIT void __next_mem_pfn_range_in_zone(u64 *idx, struct zone *zone,
@@ -310,7 +306,6 @@ void __next_mem_pfn_range_in_zone(u64 *idx, struct zone *zone,for_each_mem_range_rev (i, &memblock.memory, &memblock.reserved, \ nid, flags, p_start, p_end, p_nid) -#ifdef CONFIG_HAVE_MEMBLOCK_NODE_MAP int memblock_set_node (phys_addr_t base,phys_addr_t size,struct memblock_type *type,int nid);
@@ -323,16 +318,6 @@ static inline int memblock_get_region_node(const struct memblock_region *r) {return r->nid; } -#else -static inline void memblock_set_region_node (struct memblock_region *r,int nid) -{ -} - -static inline int memblock_get_region_node (const struct memblock_region *r) -{ -return 0 ; -} -#endif /* CONFIG_HAVE_MEMBLOCK_NODE_MAP */ /* Flags for memblock allocation APIs */ #define MEMBLOCK_ALLOC_ANYWHERE (~(phys_addr_t)0)
include/linux/mm.h +1 -15
@@ -2125,9 +2125,8 @@ static inline unsigned long get_num_physpages(void)return phys_pages; } -#ifdef CONFIG_HAVE_MEMBLOCK_NODE_MAP /* -* With CONFIG_HAVE_MEMBLOCK_NODE_MAP set, an architecture may initialise its +* Using memblock node mappings, an architecture may initialise its * zones, allocate the backing mem_map and account for memory holes in a more * architecture independent manner. This is a substitute for creating the * zone_sizes[] and zholes_size[] arrays and passing them to
@@ -2148,9 +2147,6 @@ static inline unsigned long get_num_physpages(void)* registered physical page range. Similarly * sparse_memory_present_with_active_regions() calls memory_present() for * each range when SPARSEMEM is enabled. -* -* See mm/page_alloc.c for more information on each function exposed by -* CONFIG_HAVE_MEMBLOCK_NODE_MAP. */ extern void free_area_init_nodes (unsigned long *max_zone_pfn);unsigned long node_map_pfn_alignment (void );
@@ -2165,22 +2161,12 @@ extern void free_bootmem_with_active_regions(int nid, unsigned long max_low_pfn);extern void sparse_memory_present_with_active_regions (int nid); -#endif /* CONFIG_HAVE_MEMBLOCK_NODE_MAP */ -#if !defined(CONFIG_HAVE_MEMBLOCK_NODE_MAP ) && \ - !defined(CONFIG_HAVE_ARCH_EARLY_PFN_TO_NID ) -static inline int __early_pfn_to_nid (unsigned long pfn, -struct mminit_pfnnid_cache *state) -{ -return 0 ; -} -#else /* please see mm/page_alloc.c */ extern int __meminit early_pfn_to_nid (unsigned long pfn);/* there is a per-arch backend function. */ extern int __meminit __early_pfn_to_nid (unsigned long pfn,struct mminit_pfnnid_cache *state); -#endif extern void set_dma_reserve (unsigned long new_dma_reserve); extern void memmap_init_zone(unsigned long, int, unsigned long, unsigned long,
include/linux/mmzone.h +0 -13
@@ -917,11 +917,7 @@ extern int movable_zone;#ifdef CONFIG_HIGHMEM static inline int zone_movable_is_highmem (void ) { -#ifdef CONFIG_HAVE_MEMBLOCK_NODE_MAP return movable_zone =ZONE_HIGHMEM ; -#else -return (ZONE_MOVABLE -1 ) =ZONE_HIGHMEM ; -#endif }#endif
@@ -1121,15 +1117,6 @@ static inline struct zoneref *first_zones_zonelist(struct zonelist *zonelist,#include <asm/sparsemem.h> #endif -#if !defined(CONFIG_HAVE_ARCH_EARLY_PFN_TO_NID ) && \ - !defined(CONFIG_HAVE_MEMBLOCK_NODE_MAP ) -static inline unsigned long early_pfn_to_nid (unsigned long pfn) -{ -BUILD_BUG_ON (IS_ENABLED (CONFIG_NUMA )); -return 0 ; -} -#endif -#ifdef CONFIG_FLATMEM #define pfn_to_nid (pfn) (0)#endif
mm/memblock.c +1 -8
@@ -620,9 +620,7 @@ static int __init_memblock memblock_add_range(struct memblock_type *type,* area, insert that portion. */ if (rbase > base) { -#ifdef CONFIG_HAVE_MEMBLOCK_NODE_MAP WARN_ON (nid !=memblock_get_region_node (rgn)); -#endif WARN_ON (flags != rgn->flags); nr_new++;if (insert)
@@ -1197,7 +1195,6 @@ void __init_memblock __next_mem_range_rev(u64 *idx, int nid, *idx =ULLONG_MAX ; } -#ifdef CONFIG_HAVE_MEMBLOCK_NODE_MAP /* * Common iterator interface used to define for_each_mem_pfn_range(). */
@@ -1258,7 +1255,7 @@ int __init_memblock memblock_set_node(phys_addr_t base, phys_addr_t size,memblock_merge_regions (type);return 0 ; } -#endif /* CONFIG_HAVE_MEMBLOCK_NODE_MAP */ +#ifdef CONFIG_DEFERRED_STRUCT_PAGE_INIT /** * __next_mem_pfn_range_in_zone - iterator for for_each_*_range_in_zone()
@@ -1797,7 +1794,6 @@ bool __init_memblock memblock_is_map_memory(phys_addr_t addr)return !memblock_is_nomap (&memblock.memory.regions[i]); } -#ifdef CONFIG_HAVE_MEMBLOCK_NODE_MAP int __init_memblock memblock_search_pfn_nid(unsigned long pfn,unsigned long *start_pfn,unsigned long *end_pfn) {
@@ -1812,7 +1808,6 @@ int __init_memblock memblock_search_pfn_nid(unsigned long pfn,return type->regions[mid].nid; } -#endif /** * memblock_is_region_memory - check if a region is a subset of memory
@@ -1903,11 +1898,9 @@ static void __init_memblock memblock_dump(struct memblock_type *type) size = rgn->size; end = base + size -1 ; flags = rgn->flags; -#ifdef CONFIG_HAVE_MEMBLOCK_NODE_MAP if (memblock_get_region_node (rgn) !=MAX_NUMNODES )snprintf (nid_buf,sizeof (nid_buf)," on node %d" ,memblock_get_region_node (rgn)); -#endif pr_info (" %s[%#x]\t[%pa-%pa], %pa bytes%s flags: %#x\n" , type->name, idx, &base, &end, &size, nid_buf, flags); }
mm/memory_hotplug.c +0 -4
@@ -1366,11 +1366,7 @@ check_pages_isolated_cb(unsigned long start_pfn, unsigned long nr_pages,static int __init cmdline_parse_movable_node (char *p) { -#ifdef CONFIG_HAVE_MEMBLOCK_NODE_MAP movable_node_enabled = true; -#else -pr_warn ("movable_node parameter depends on CONFIG_HAVE_MEMBLOCK_NODE_MAP to work properly\n" ); -#endif return 0 ; }early_param ("movable_node" , cmdline_parse_movable_node);
mm/page_alloc.c +59 -44
220 lines not on this page · raw message
Signature · 1 line
2.25.1
Re: [PATCH RFC] mm: remove CONFIG_HAVE_MEMBLOCK_NODE_MAP (was: Re: [PATCH v3 0/5] mm: Enable CONFIG_NODES_SPAN_OTHER_NODES by default for NUMA)
Quoted 11 lines · Mike Rapoport
Quoted 4 lines · Michal Hocko
What would it take to make ia64 use
HAVE_MEMBLOCK_NODE_MAP? I would really love to see that thing go away. It is causing problems when people try to use memblock api.Well, it's a small patch in the end :)
Currently all NUMA architectures currently enable
CONFIG_HAVE_MEMBLOCK_NODE_MAPand usefree_area_init_nodes()to initialize nodes and zones structures.
I did some investigation, there are nine ARCHes having NUMA config. And among them, alpha doesn't have HAVE_MEMBLOCK_NODE_MAP support. While the interesting thing is there are two ARCHes which have HAVE_MEMBLOCK_NODE_MAP config, but don't have NUMA config adding, they are microblaze and riscv. Obviously it was not carefully considered to add HAVE_MEMBLOCK_NODE_MAP config into riscv and microblaze.
arch/alpha/Kconfig:config NUMA
arch/arm64/Kconfig:config NUMA
arch/ia64/Kconfig:config NUMA
arch/mips/Kconfig:config NUMA
arch/powerpc/Kconfig:config NUMA
arch/s390/Kconfig:config NUMA
arch/sh/mm/Kconfig:config NUMA
arch/sparc/Kconfig:config NUMA
arch/x86/Kconfig:config NUMA
From above information, we can remove HAVE_MEMBLOCK_NODE_MAP, and replace it with CONFIG_NUMA. That sounds more sensible to store nid into memblock when NUMA support is enabled.
include/linux/memblock.h struct memblock_region { · line 55
Quoted 11 lines
include/linux/memblock.h +0
-2@@ -50,9 +50,7 @@ struct memblock_region {phys_addr_t base;phys_addr_t size;enum memblock_flags flags; -#ifdef CONFIG_HAVE_MEMBLOCK_NODE_MAP int nid; -#endif
I didn't look into other change very carefully, but feel enabling memblock node map for all ARCHes looks a little radical. After all, many ARCHes even don't have NUMA support.
Quoted 445 lines
include/linux/memblock.h +0
-13};/** @@ -215,7 +213,6 @@ static inline bool memblock_is_nomap(struct memblock_region *m)return m->flags &MEMBLOCK_NOMAP ; } -#ifdef CONFIG_HAVE_MEMBLOCK_NODE_MAP int memblock_search_pfn_nid (unsigned long pfn,unsigned long *start_pfn,unsigned long *end_pfn); void __next_mem_pfn_range(int *idx, int nid, unsigned long *out_start_pfn,@@ -234,7 +231,6 @@ void __next_mem_pfn_range(int *idx, int nid, unsigned long *out_start_pfn,#define for_each_mem_pfn_range (i, nid, p_start, p_end, p_nid) \ for (i = -1, __next_mem_pfn_range(&i, nid, p_start, p_end, p_nid); \ i >= 0; __next_mem_pfn_range(&i, nid, p_start, p_end, p_nid)) -#endif /* CONFIG_HAVE_MEMBLOCK_NODE_MAP */ #ifdef CONFIG_DEFERRED_STRUCT_PAGE_INIT void __next_mem_pfn_range_in_zone(u64 *idx, struct zone *zone,@@ -310,7 +306,6 @@ void __next_mem_pfn_range_in_zone(u64 *idx, struct zone *zone,for_each_mem_range_rev (i, &memblock.memory, &memblock.reserved, \ nid, flags, p_start, p_end, p_nid) -#ifdef CONFIG_HAVE_MEMBLOCK_NODE_MAP int memblock_set_node (phys_addr_t base,phys_addr_t size,struct memblock_type *type,int nid);@@ -323,16 +318,6 @@ static inline int memblock_get_region_node(const struct memblock_region *r) {return r->nid; } -#else -static inline void memblock_set_region_node (struct memblock_region *r,int nid) -{ -} - -static inline int memblock_get_region_node (const struct memblock_region *r) -{ -return 0 ; -} -#endif /* CONFIG_HAVE_MEMBLOCK_NODE_MAP */ /* Flags for memblock allocation APIs */ #define MEMBLOCK_ALLOC_ANYWHERE (~(phys_addr_t)0)include/linux/mm.h +1
-15@@ -2125,9 +2125,8 @@ static inline unsigned long get_num_physpages(void)return phys_pages; } -#ifdef CONFIG_HAVE_MEMBLOCK_NODE_MAP /* -* With CONFIG_HAVE_MEMBLOCK_NODE_MAP set, an architecture may initialise its +* Using memblock node mappings, an architecture may initialise its * zones, allocate the backing mem_map and account for memory holes in a more * architecture independent manner. This is a substitute for creating the * zone_sizes[] and zholes_size[] arrays and passing them to @@ -2148,9 +2147,6 @@ static inline unsigned long get_num_physpages(void)* registered physical page range. Similarly * sparse_memory_present_with_active_regions() calls memory_present() for * each range when SPARSEMEM is enabled. -* -* See mm/page_alloc.c for more information on each function exposed by -* CONFIG_HAVE_MEMBLOCK_NODE_MAP. */ extern void free_area_init_nodes (unsigned long *max_zone_pfn);unsigned long node_map_pfn_alignment (void );@@ -2165,22 +2161,12 @@ extern void free_bootmem_with_active_regions(int nid, unsigned long max_low_pfn);extern void sparse_memory_present_with_active_regions (int nid); -#endif /* CONFIG_HAVE_MEMBLOCK_NODE_MAP */ -#if !defined(CONFIG_HAVE_MEMBLOCK_NODE_MAP ) && \ - !defined(CONFIG_HAVE_ARCH_EARLY_PFN_TO_NID ) -static inline int __early_pfn_to_nid (unsigned long pfn, -struct mminit_pfnnid_cache *state) -{ -return 0 ; -} -#else /* please see mm/page_alloc.c */ extern int __meminit early_pfn_to_nid (unsigned long pfn);/* there is a per-arch backend function. */ extern int __meminit __early_pfn_to_nid (unsigned long pfn,struct mminit_pfnnid_cache *state); -#endif extern void set_dma_reserve (unsigned long new_dma_reserve); extern void memmap_init_zone(unsigned long, int, unsigned long, unsigned long,include/linux/mmzone.h +0
-13@@ -917,11 +917,7 @@ extern int movable_zone;#ifdef CONFIG_HIGHMEM static inline int zone_movable_is_highmem (void ) { -#ifdef CONFIG_HAVE_MEMBLOCK_NODE_MAP return movable_zone =ZONE_HIGHMEM ; -#else -return (ZONE_MOVABLE -1 ) =ZONE_HIGHMEM ; -#endif }#endif @@ -1121,15 +1117,6 @@ static inline struct zoneref *first_zones_zonelist(struct zonelist *zonelist,#include <asm/sparsemem.h> #endif -#if !defined(CONFIG_HAVE_ARCH_EARLY_PFN_TO_NID ) && \ - !defined(CONFIG_HAVE_MEMBLOCK_NODE_MAP ) -static inline unsigned long early_pfn_to_nid (unsigned long pfn) -{ -BUILD_BUG_ON (IS_ENABLED (CONFIG_NUMA )); -return 0 ; -} -#endif -#ifdef CONFIG_FLATMEM #define pfn_to_nid (pfn) (0)#endif mm/memblock.c +1
-8@@ -620,9 +620,7 @@ static int __init_memblock memblock_add_range(struct memblock_type *type,* area, insert that portion. */ if (rbase > base) { -#ifdef CONFIG_HAVE_MEMBLOCK_NODE_MAP WARN_ON (nid !=memblock_get_region_node (rgn)); -#endif WARN_ON (flags != rgn->flags); nr_new++;if (insert)@@ -1197,7 +1195,6 @@ void __init_memblock __next_mem_range_rev(u64 *idx, int nid, *idx =ULLONG_MAX ; } -#ifdef CONFIG_HAVE_MEMBLOCK_NODE_MAP /* * Common iterator interface used to define for_each_mem_pfn_range(). */ @@ -1258,7 +1255,7 @@ int __init_memblock memblock_set_node(phys_addr_t base, phys_addr_t size,memblock_merge_regions (type);return 0 ; } -#endif /* CONFIG_HAVE_MEMBLOCK_NODE_MAP */ +#ifdef CONFIG_DEFERRED_STRUCT_PAGE_INIT /** * __next_mem_pfn_range_in_zone - iterator for for_each_*_range_in_zone() @@ -1797,7 +1794,6 @@ bool __init_memblock memblock_is_map_memory(phys_addr_t addr)return !memblock_is_nomap (&memblock.memory.regions[i]); } -#ifdef CONFIG_HAVE_MEMBLOCK_NODE_MAP int __init_memblock memblock_search_pfn_nid(unsigned long pfn,unsigned long *start_pfn,unsigned long *end_pfn) {@@ -1812,7 +1808,6 @@ int __init_memblock memblock_search_pfn_nid(unsigned long pfn,return type->regions[mid].nid; } -#endif /** * memblock_is_region_memory - check if a region is a subset of memory @@ -1903,11 +1898,9 @@ static void __init_memblock memblock_dump(struct memblock_type *type) size = rgn->size; end = base + size -1 ; flags = rgn->flags; -#ifdef CONFIG_HAVE_MEMBLOCK_NODE_MAP if (memblock_get_region_node (rgn) !=MAX_NUMNODES )snprintf (nid_buf,sizeof (nid_buf)," on node %d" ,memblock_get_region_node (rgn)); -#endif pr_info (" %s[%#x]\t[%pa-%pa], %pa bytes%s flags: %#x\n" , type->name, idx, &base, &end, &size, nid_buf, flags); }mm/memory_hotplug.c +0
-4@@ -1366,11 +1366,7 @@ check_pages_isolated_cb(unsigned long start_pfn, unsigned long nr_pages,static int __init cmdline_parse_movable_node (char *p) { -#ifdef CONFIG_HAVE_MEMBLOCK_NODE_MAP movable_node_enabled = true; -#else -pr_warn ("movable_node parameter depends on CONFIG_HAVE_MEMBLOCK_NODE_MAP to work properly\n" ); -#endif return 0 ; }early_param ("movable_node" , cmdline_parse_movable_node);mm/page_alloc.c +59
-44@@ -335,7 +335,6 @@ static unsigned long nr_kernel_pages __initdata;static unsigned long nr_all_pages__initdata ;static unsigned long dma_reserve__initdata ; -#ifdef CONFIG_HAVE_MEMBLOCK_NODE_MAP static unsigned long arch_zone_lowest_possible_pfn[MAX_NR_ZONES ]__initdata ;static unsigned long arch_zone_highest_possible_pfn[MAX_NR_ZONES ]__initdata ;static unsigned long required_kernelcore__initdata ;@@ -348,7 +347,6 @@ static bool mirrored_kernelcore __meminitdata;/* movable_zone is the "real" zone pages in ZONE_MOVABLE are taken from */ int movable_zone;EXPORT_SYMBOL (movable_zone); -#endif /* CONFIG_HAVE_MEMBLOCK_NODE_MAP */ #if MAX_NUMNODES >1 unsigned int nr_node_ids__read_mostly =MAX_NUMNODES ;@@ -1447,9 +1445,6 @@ void __free_pages_core(struct page *page, unsigned int order)__free_pages (page, order); } -#if defined(CONFIG_HAVE_ARCH_EARLY_PFN_TO_NID ) || \ - defined(CONFIG_HAVE_MEMBLOCK_NODE_MAP ) -static struct mminit_pfnnid_cache early_pfnnid_cache__meminitdata ;int __meminit early_pfn_to_nid (unsigned long pfn)@@ -1465,7 +1460,6 @@ int __meminit early_pfn_to_nid(unsigned long pfn)return nid; } -#endif #ifdef CONFIG_NODES_SPAN_OTHER_NODES /* Only safe to use early in boot when initialisation is single-threaded */ @@ -5828,7 +5822,6 @@ void __ref build_all_zonelists(pg_data_t *pgdat)static bool __meminit overlap_memmap_init (unsigned long zone,unsigned long *pfn) { -#ifdef CONFIG_HAVE_MEMBLOCK_NODE_MAP static struct memblock_region *r;if (mirrored_kernelcore && zone =ZONE_MOVABLE ) {@@ -5844,7 +5837,7 @@ overlap_memmap_init(unsigned long zone, unsigned long *pfn)return true; } } -#endif +return false; }@@ -6227,7 +6220,6 @@ void __meminit init_currently_empty_zone(struct zone *zone, zone->initialized =1 ; } -#ifdef CONFIG_HAVE_MEMBLOCK_NODE_MAP #ifndef CONFIG_HAVE_ARCH_EARLY_PFN_TO_NID /* @@ -6503,8 +6495,7 @@ static unsigned long __init zone_absent_pages_in_node(int nid,return nr_absent; } -#else /* CONFIG_HAVE_MEMBLOCK_NODE_MAP */ -static inline unsigned long __init zone_spanned_pages_in_node(int nid, +static inline unsigned long __init compat_zone_spanned_pages_in_node(int nid, unsigned long zone_type, unsigned long node_start_pfn, unsigned long node_end_pfn,@@ -6523,7 +6514,7 @@ static inline unsigned long __init zone_spanned_pages_in_node(int nid,return zones_size[zone_type]; } -static inline unsigned long __init zone_absent_pages_in_node(int nid, +static inline unsigned long __init compat_zone_absent_pages_in_node(int nid, unsigned long zone_type, unsigned long node_start_pfn, unsigned long node_end_pfn,@@ -6535,13 +6526,12 @@ static inline unsigned long __init zone_absent_pages_in_node(int nid,return zholes_size[zone_type]; } -#endif /* CONFIG_HAVE_MEMBLOCK_NODE_MAP */ -static void __init calculate_node_totalpages (struct pglist_data *pgdat,unsigned long node_start_pfn,unsigned long node_end_pfn,unsigned long *zones_size, -unsigned long *zholes_size) +unsigned long *zholes_size, +bool compat) {unsigned long realtotalpages =0 , totalpages =0 ;enum zone_type i;@@ -6549,17 +6539,38 @@ static void __init calculate_node_totalpages(struct pglist_data *pgdat,for (i =0 ; i <MAX_NR_ZONES ; i++) {struct zone *zone = pgdat->node_zones + i;unsigned long zone_start_pfn, zone_end_pfn; +unsigned long spanned, absent;unsigned long size, real_size; - size =zone_spanned_pages_in_node (pgdat->node_id, i, - node_start_pfn, - node_end_pfn, - &zone_start_pfn, - &zone_end_pfn, - zones_size); - real_size = size -zone_absent_pages_in_node (pgdat->node_id, i, - node_start_pfn, node_end_pfn, - zholes_size); +if (compat) { + spanned =compat_zone_spanned_pages_in_node ( + pgdat->node_id, i, + node_start_pfn, + node_end_pfn, + &zone_start_pfn, + &zone_end_pfn, + zones_size); + absent =compat_zone_absent_pages_in_node ( + pgdat->node_id, i, + node_start_pfn, + node_end_pfn, + zholes_size); + }else { + spanned =zone_spanned_pages_in_node (pgdat->node_id, i, + node_start_pfn, + node_end_pfn, + &zone_start_pfn, + &zone_end_pfn, + zones_size); + absent =zone_absent_pages_in_node (pgdat->node_id, i, + node_start_pfn, + node_end_pfn, + zholes_size); + } + + size = spanned; + real_size = size - absent; +if (size) zone->zone_start_pfn = zone_start_pfn;else @@ -6859,10 +6870,8 @@ static void __ref alloc_node_mem_map(struct pglist_data *pgdat)*/ if (pgdat =NODE_DATA (0 )) { mem_map =NODE_DATA (0 )->node_mem_map; -#if defined(CONFIG_HAVE_MEMBLOCK_NODE_MAP ) || defined(CONFIG_FLATMEM )if (page_to_pfn (mem_map) != pgdat->node_start_pfn) mem_map -= offset; -#endif /* CONFIG_HAVE_MEMBLOCK_NODE_MAP */ }#endif }@@ -6879,9 +6888,10 @@ static inline void pgdat_set_deferred_range(pg_data_t *pgdat)static inline void pgdat_set_deferred_range (pg_data_t *pgdat) {}#endif -void __init free_area_init_node (int nid,unsigned long *zones_size, -unsigned long node_start_pfn, -unsigned long *zholes_size) +void __init __free_area_init_node (int nid,unsigned long *zones_size, +unsigned long node_start_pfn, +unsigned long *zholes_size, +bool compat) {pg_data_t *pgdat =NODE_DATA (nid);unsigned long start_pfn =0 ;@@ -6893,16 +6903,17 @@ void __init free_area_init_node(int nid, unsigned long *zones_size, pgdat->node_id = nid; pgdat->node_start_pfn = node_start_pfn; pgdat->per_cpu_nodestats =NULL ; -#ifdef CONFIG_HAVE_MEMBLOCK_NODE_MAP -get_pfn_range_for_nid (nid, &start_pfn, &end_pfn); -pr_info ("Initmem setup node %d [mem %#018Lx-%#018Lx]\n" , nid, - (u64 )start_pfn <<PAGE_SHIFT , - end_pfn ? ((u64 )end_pfn <<PAGE_SHIFT ) -1 :0 ); -#else - start_pfn = node_start_pfn; -#endif +if (!compat) { +get_pfn_range_for_nid (nid, &start_pfn, &end_pfn); +pr_info ("Initmem setup node %d [mem %#018Lx-%#018Lx]\n" , nid, + (u64 )start_pfn <<PAGE_SHIFT , + end_pfn ? ((u64 )end_pfn <<PAGE_SHIFT ) -1 :0 ); + }else { + start_pfn = node_start_pfn; + } +calculate_node_totalpages (pgdat, start_pfn, end_pfn, - zones_size, zholes_size); + zones_size, zholes_size, compat);alloc_node_mem_map (pgdat);pgdat_set_deferred_range (pgdat);@@ -6910,6 +6921,14 @@ void __init free_area_init_node(int nid, unsigned long *zones_size,free_area_init_core (pgdat); } +void __init free_area_init_node (int nid,unsigned long *zones_size, +unsigned long node_start_pfn, +unsigned long *zholes_size) +{ +__free_area_init_node (nid, zones_size, node_start_pfn, zholes_size, + true); +} +#if !defined(CONFIG_FLAT_NODE_MEM_MAP )/* * Initialize all valid struct pages in the range [spfn, epfn) and mark them @@ -6993,8 +7012,6 @@ static inline void __init init_unavailable_mem(void) }#endif /* !CONFIG_FLAT_NODE_MEM_MAP */ -#ifdef CONFIG_HAVE_MEMBLOCK_NODE_MAP -#if MAX_NUMNODES >1 /* * Figure out the number of possible node ids. @@ -7423,8 +7440,8 @@ void __init free_area_init_nodes(unsigned long *max_zone_pfn)init_unavailable_mem ();for_each_online_node (nid) {pg_data_t *pgdat =NODE_DATA (nid); -free_area_init_node (nid,NULL , -find_min_pfn_for_node (nid),NULL ); +__free_area_init_node (nid,NULL , +find_min_pfn_for_node (nid),NULL , false);/* Any memory on that node */ if (pgdat->node_present_pages)@@ -7489,8 +7506,6 @@ static int __init cmdline_parse_movablecore(char *p)early_param ("kernelcore" , cmdline_parse_kernelcore);early_param ("movablecore" , cmdline_parse_movablecore); -#endif /* CONFIG_HAVE_MEMBLOCK_NODE_MAP */ -void adjust_managed_page_count (struct page *page,long count) {atomic_long_add (count, &page_zone (page)->managed_pages);Signature · 1 line
2.25.1
Re: [PATCH RFC] mm: remove CONFIG_HAVE_MEMBLOCK_NODE_MAP (was: Re: [PATCH v3 0/5] mm: Enable CONFIG_NODES_SPAN_OTHER_NODES by default for NUMA)
Hi,
Quoted 7 lines · Baoquan He
Quoted 6 lines · Mike Rapoport
Quoted 4 lines · Michal Hocko
What would it take to make ia64 use
HAVE_MEMBLOCK_NODE_MAP? I would really love to see that thing go away. It is causing problems when people try to use memblock api.
Quoted 8 lines
Quoted 5 lines
Well, it's a small patch in the end :)
Currently all NUMA architectures currently enable
CONFIG_HAVE_MEMBLOCK_NODE_MAPand usefree_area_init_nodes()to initialize nodes and zones structures.I did some investigation, there are nine ARCHes having NUMA config. And among them, alpha doesn't have
HAVE_MEMBLOCK_NODE_MAPsupport.
alpha has NUMA marked as BROKEN for a long time now. I doubt anybody is going to fix that.
Quoted 4 lines
While the interesting thing is there are two ARCHes which have
HAVE_MEMBLOCK_NODE_MAPconfig, but don't have NUMA config adding, they are microblaze and riscv. Obviously it was not carefully considered to addHAVE_MEMBLOCK_NODE_MAPconfig into riscv and microblaze.
Well, microblaze probably copied it and forgot to remove and riscv might get actual NUMA at some point.
Quoted 10 lines
arch/alpha/Kconfig:config NUMA
arch/arm64/Kconfig:config NUMA
arch/ia64/Kconfig:config NUMA
arch/mips/Kconfig:config NUMA
arch/powerpc/Kconfig:config NUMA
arch/s390/Kconfig:config NUMA
arch/sh/mm/Kconfig:config NUMA
arch/sparc/Kconfig:config NUMA
arch/x86/Kconfig:config NUMA
Quoted 3 lines
From above information, we can remove
HAVE_MEMBLOCK_NODE_MAP, and replace it withCONFIG_NUMA. That sounds more sensible to store nid into memblock when NUMA support is enabled.
Replacing CONFIG_HAVE_MEMBLOCK_NODE_MAP with CONFIG_NUMA will work, but this will not help cleaning up the whole node/zone initialization mess and we'll be stuck with two implementations.
The overhead of enabling HAVE_MEMBLOCK_NODE_MAP is only for init time as most architectures will anyway discard the entire memblock, so having it in a UMA arch won't be a problem. The only exception is arm that uses memblock for pfn_valid(), here we may also think about a solution to compensate the addition of nid to the memblock structures.
We also may want to make all the movable_node and movable_core functionality dependent on CONFIG_NUMA.
Quoted 462 lines
include/linux/memblock.h struct memblock_region { · line 55
Quoted 11 lines
include/linux/memblock.h +0
-2@@ -50,9 +50,7 @@ struct memblock_region {phys_addr_t base;phys_addr_t size;enum memblock_flags flags; -#ifdef CONFIG_HAVE_MEMBLOCK_NODE_MAP int nid; -#endif I didn't look into other change very carefully, but feel enabling memblock node map for all ARCHes looks a little radical. After all, many ARCHes even don't have NUMA support.
Quoted 445 lines
include/linux/memblock.h +0
-13};/** @@ -215,7 +213,6 @@ static inline bool memblock_is_nomap(struct memblock_region *m)return m->flags &MEMBLOCK_NOMAP ; } -#ifdef CONFIG_HAVE_MEMBLOCK_NODE_MAP int memblock_search_pfn_nid (unsigned long pfn,unsigned long *start_pfn,unsigned long *end_pfn); void __next_mem_pfn_range(int *idx, int nid, unsigned long *out_start_pfn,@@ -234,7 +231,6 @@ void __next_mem_pfn_range(int *idx, int nid, unsigned long *out_start_pfn,#define for_each_mem_pfn_range (i, nid, p_start, p_end, p_nid) \ for (i = -1, __next_mem_pfn_range(&i, nid, p_start, p_end, p_nid); \ i >= 0; __next_mem_pfn_range(&i, nid, p_start, p_end, p_nid)) -#endif /* CONFIG_HAVE_MEMBLOCK_NODE_MAP */ #ifdef CONFIG_DEFERRED_STRUCT_PAGE_INIT void __next_mem_pfn_range_in_zone(u64 *idx, struct zone *zone,@@ -310,7 +306,6 @@ void __next_mem_pfn_range_in_zone(u64 *idx, struct zone *zone,for_each_mem_range_rev (i, &memblock.memory, &memblock.reserved, \ nid, flags, p_start, p_end, p_nid) -#ifdef CONFIG_HAVE_MEMBLOCK_NODE_MAP int memblock_set_node (phys_addr_t base,phys_addr_t size,struct memblock_type *type,int nid);@@ -323,16 +318,6 @@ static inline int memblock_get_region_node(const struct memblock_region *r) {return r->nid; } -#else -static inline void memblock_set_region_node (struct memblock_region *r,int nid) -{ -} - -static inline int memblock_get_region_node (const struct memblock_region *r) -{ -return 0 ; -} -#endif /* CONFIG_HAVE_MEMBLOCK_NODE_MAP */ /* Flags for memblock allocation APIs */ #define MEMBLOCK_ALLOC_ANYWHERE (~(phys_addr_t)0)include/linux/mm.h +1
-15@@ -2125,9 +2125,8 @@ static inline unsigned long get_num_physpages(void)return phys_pages; } -#ifdef CONFIG_HAVE_MEMBLOCK_NODE_MAP /* -* With CONFIG_HAVE_MEMBLOCK_NODE_MAP set, an architecture may initialise its +* Using memblock node mappings, an architecture may initialise its * zones, allocate the backing mem_map and account for memory holes in a more * architecture independent manner. This is a substitute for creating the * zone_sizes[] and zholes_size[] arrays and passing them to @@ -2148,9 +2147,6 @@ static inline unsigned long get_num_physpages(void)* registered physical page range. Similarly * sparse_memory_present_with_active_regions() calls memory_present() for * each range when SPARSEMEM is enabled. -* -* See mm/page_alloc.c for more information on each function exposed by -* CONFIG_HAVE_MEMBLOCK_NODE_MAP. */ extern void free_area_init_nodes (unsigned long *max_zone_pfn);unsigned long node_map_pfn_alignment (void );@@ -2165,22 +2161,12 @@ extern void free_bootmem_with_active_regions(int nid, unsigned long max_low_pfn);extern void sparse_memory_present_with_active_regions (int nid); -#endif /* CONFIG_HAVE_MEMBLOCK_NODE_MAP */ -#if !defined(CONFIG_HAVE_MEMBLOCK_NODE_MAP ) && \ - !defined(CONFIG_HAVE_ARCH_EARLY_PFN_TO_NID ) -static inline int __early_pfn_to_nid (unsigned long pfn, -struct mminit_pfnnid_cache *state) -{ -return 0 ; -} -#else /* please see mm/page_alloc.c */ extern int __meminit early_pfn_to_nid (unsigned long pfn);/* there is a per-arch backend function. */ extern int __meminit __early_pfn_to_nid (unsigned long pfn,struct mminit_pfnnid_cache *state); -#endif extern void set_dma_reserve (unsigned long new_dma_reserve); extern void memmap_init_zone(unsigned long, int, unsigned long, unsigned long,include/linux/mmzone.h +0
-13@@ -917,11 +917,7 @@ extern int movable_zone;#ifdef CONFIG_HIGHMEM static inline int zone_movable_is_highmem (void ) { -#ifdef CONFIG_HAVE_MEMBLOCK_NODE_MAP return movable_zone =ZONE_HIGHMEM ; -#else -return (ZONE_MOVABLE -1 ) =ZONE_HIGHMEM ; -#endif }#endif @@ -1121,15 +1117,6 @@ static inline struct zoneref *first_zones_zonelist(struct zonelist *zonelist,#include <asm/sparsemem.h> #endif -#if !defined(CONFIG_HAVE_ARCH_EARLY_PFN_TO_NID ) && \ - !defined(CONFIG_HAVE_MEMBLOCK_NODE_MAP ) -static inline unsigned long early_pfn_to_nid (unsigned long pfn) -{ -BUILD_BUG_ON (IS_ENABLED (CONFIG_NUMA )); -return 0 ; -} -#endif -#ifdef CONFIG_FLATMEM #define pfn_to_nid (pfn) (0)#endif mm/memblock.c +1
-8@@ -620,9 +620,7 @@ static int __init_memblock memblock_add_range(struct memblock_type *type,* area, insert that portion. */ if (rbase > base) { -#ifdef CONFIG_HAVE_MEMBLOCK_NODE_MAP WARN_ON (nid !=memblock_get_region_node (rgn)); -#endif WARN_ON (flags != rgn->flags); nr_new++;if (insert)@@ -1197,7 +1195,6 @@ void __init_memblock __next_mem_range_rev(u64 *idx, int nid, *idx =ULLONG_MAX ; } -#ifdef CONFIG_HAVE_MEMBLOCK_NODE_MAP /* * Common iterator interface used to define for_each_mem_pfn_range(). */ @@ -1258,7 +1255,7 @@ int __init_memblock memblock_set_node(phys_addr_t base, phys_addr_t size,memblock_merge_regions (type);return 0 ; } -#endif /* CONFIG_HAVE_MEMBLOCK_NODE_MAP */ +#ifdef CONFIG_DEFERRED_STRUCT_PAGE_INIT /** * __next_mem_pfn_range_in_zone - iterator for for_each_*_range_in_zone() @@ -1797,7 +1794,6 @@ bool __init_memblock memblock_is_map_memory(phys_addr_t addr)return !memblock_is_nomap (&memblock.memory.regions[i]); } -#ifdef CONFIG_HAVE_MEMBLOCK_NODE_MAP int __init_memblock memblock_search_pfn_nid(unsigned long pfn,unsigned long *start_pfn,unsigned long *end_pfn) {@@ -1812,7 +1808,6 @@ int __init_memblock memblock_search_pfn_nid(unsigned long pfn,return type->regions[mid].nid; } -#endif /** * memblock_is_region_memory - check if a region is a subset of memory @@ -1903,11 +1898,9 @@ static void __init_memblock memblock_dump(struct memblock_type *type) size = rgn->size; end = base + size -1 ; flags = rgn->flags; -#ifdef CONFIG_HAVE_MEMBLOCK_NODE_MAP if (memblock_get_region_node (rgn) !=MAX_NUMNODES )snprintf (nid_buf,sizeof (nid_buf)," on node %d" ,memblock_get_region_node (rgn)); -#endif pr_info (" %s[%#x]\t[%pa-%pa], %pa bytes%s flags: %#x\n" , type->name, idx, &base, &end, &size, nid_buf, flags); }mm/memory_hotplug.c +0
-4@@ -1366,11 +1366,7 @@ check_pages_isolated_cb(unsigned long start_pfn, unsigned long nr_pages,static int __init cmdline_parse_movable_node (char *p) { -#ifdef CONFIG_HAVE_MEMBLOCK_NODE_MAP movable_node_enabled = true; -#else -pr_warn ("movable_node parameter depends on CONFIG_HAVE_MEMBLOCK_NODE_MAP to work properly\n" ); -#endif return 0 ; }early_param ("movable_node" , cmdline_parse_movable_node);mm/page_alloc.c +59
-44@@ -335,7 +335,6 @@ static unsigned long nr_kernel_pages __initdata;static unsigned long nr_all_pages__initdata ;static unsigned long dma_reserve__initdata ; -#ifdef CONFIG_HAVE_MEMBLOCK_NODE_MAP static unsigned long arch_zone_lowest_possible_pfn[MAX_NR_ZONES ]__initdata ;static unsigned long arch_zone_highest_possible_pfn[MAX_NR_ZONES ]__initdata ;static unsigned long required_kernelcore__initdata ;@@ -348,7 +347,6 @@ static bool mirrored_kernelcore __meminitdata;/* movable_zone is the "real" zone pages in ZONE_MOVABLE are taken from */ int movable_zone;EXPORT_SYMBOL (movable_zone); -#endif /* CONFIG_HAVE_MEMBLOCK_NODE_MAP */ #if MAX_NUMNODES >1 unsigned int nr_node_ids__read_mostly =MAX_NUMNODES ;@@ -1447,9 +1445,6 @@ void __free_pages_core(struct page *page, unsigned int order)__free_pages (page, order); } -#if defined(CONFIG_HAVE_ARCH_EARLY_PFN_TO_NID ) || \ - defined(CONFIG_HAVE_MEMBLOCK_NODE_MAP ) -static struct mminit_pfnnid_cache early_pfnnid_cache__meminitdata ;int __meminit early_pfn_to_nid (unsigned long pfn)@@ -1465,7 +1460,6 @@ int __meminit early_pfn_to_nid(unsigned long pfn)return nid; } -#endif #ifdef CONFIG_NODES_SPAN_OTHER_NODES /* Only safe to use early in boot when initialisation is single-threaded */ @@ -5828,7 +5822,6 @@ void __ref build_all_zonelists(pg_data_t *pgdat)static bool __meminit overlap_memmap_init (unsigned long zone,unsigned long *pfn) { -#ifdef CONFIG_HAVE_MEMBLOCK_NODE_MAP static struct memblock_region *r;if (mirrored_kernelcore && zone =ZONE_MOVABLE ) {@@ -5844,7 +5837,7 @@ overlap_memmap_init(unsigned long zone, unsigned long *pfn)return true; } } -#endif +return false; }@@ -6227,7 +6220,6 @@ void __meminit init_currently_empty_zone(struct zone *zone, zone->initialized =1 ; } -#ifdef CONFIG_HAVE_MEMBLOCK_NODE_MAP #ifndef CONFIG_HAVE_ARCH_EARLY_PFN_TO_NID /* @@ -6503,8 +6495,7 @@ static unsigned long __init zone_absent_pages_in_node(int nid,return nr_absent; } -#else /* CONFIG_HAVE_MEMBLOCK_NODE_MAP */ -static inline unsigned long __init zone_spanned_pages_in_node(int nid, +static inline unsigned long __init compat_zone_spanned_pages_in_node(int nid, unsigned long zone_type, unsigned long node_start_pfn, unsigned long node_end_pfn,@@ -6523,7 +6514,7 @@ static inline unsigned long __init zone_spanned_pages_in_node(int nid,return zones_size[zone_type]; } -static inline unsigned long __init zone_absent_pages_in_node(int nid, +static inline unsigned long __init compat_zone_absent_pages_in_node(int nid, unsigned long zone_type, unsigned long node_start_pfn, unsigned long node_end_pfn,@@ -6535,13 +6526,12 @@ static inline unsigned long __init zone_absent_pages_in_node(int nid,return zholes_size[zone_type]; } -#endif /* CONFIG_HAVE_MEMBLOCK_NODE_MAP */ -static void __init calculate_node_totalpages (struct pglist_data *pgdat,unsigned long node_start_pfn,unsigned long node_end_pfn,unsigned long *zones_size, -unsigned long *zholes_size) +unsigned long *zholes_size, +bool compat) {unsigned long realtotalpages =0 , totalpages =0 ;enum zone_type i;@@ -6549,17 +6539,38 @@ static void __init calculate_node_totalpages(struct pglist_data *pgdat,for (i =0 ; i <MAX_NR_ZONES ; i++) {struct zone *zone = pgdat->node_zones + i;unsigned long zone_start_pfn, zone_end_pfn; +unsigned long spanned, absent;unsigned long size, real_size; - size =zone_spanned_pages_in_node (pgdat->node_id, i, - node_start_pfn, - node_end_pfn, - &zone_start_pfn, - &zone_end_pfn, - zones_size); - real_size = size -zone_absent_pages_in_node (pgdat->node_id, i, - node_start_pfn, node_end_pfn, - zholes_size); +if (compat) { + spanned =compat_zone_spanned_pages_in_node ( + pgdat->node_id, i, + node_start_pfn, + node_end_pfn, + &zone_start_pfn, + &zone_end_pfn, + zones_size); + absent =compat_zone_absent_pages_in_node ( + pgdat->node_id, i, + node_start_pfn, + node_end_pfn, + zholes_size); + }else { + spanned =zone_spanned_pages_in_node (pgdat->node_id, i, + node_start_pfn, + node_end_pfn, + &zone_start_pfn, + &zone_end_pfn, + zones_size); + absent =zone_absent_pages_in_node (pgdat->node_id, i, + node_start_pfn, + node_end_pfn, + zholes_size); + } + + size = spanned; + real_size = size - absent; +if (size) zone->zone_start_pfn = zone_start_pfn;else @@ -6859,10 +6870,8 @@ static void __ref alloc_node_mem_map(struct pglist_data *pgdat)*/ if (pgdat =NODE_DATA (0 )) { mem_map =NODE_DATA (0 )->node_mem_map; -#if defined(CONFIG_HAVE_MEMBLOCK_NODE_MAP ) || defined(CONFIG_FLATMEM )if (page_to_pfn (mem_map) != pgdat->node_start_pfn) mem_map -= offset; -#endif /* CONFIG_HAVE_MEMBLOCK_NODE_MAP */ }#endif }@@ -6879,9 +6888,10 @@ static inline void pgdat_set_deferred_range(pg_data_t *pgdat)static inline void pgdat_set_deferred_range (pg_data_t *pgdat) {}#endif -void __init free_area_init_node (int nid,unsigned long *zones_size, -unsigned long node_start_pfn, -unsigned long *zholes_size) +void __init __free_area_init_node (int nid,unsigned long *zones_size, +unsigned long node_start_pfn, +unsigned long *zholes_size, +bool compat) {pg_data_t *pgdat =NODE_DATA (nid);unsigned long start_pfn =0 ;@@ -6893,16 +6903,17 @@ void __init free_area_init_node(int nid, unsigned long *zones_size, pgdat->node_id = nid; pgdat->node_start_pfn = node_start_pfn; pgdat->per_cpu_nodestats =NULL ; -#ifdef CONFIG_HAVE_MEMBLOCK_NODE_MAP -get_pfn_range_for_nid (nid, &start_pfn, &end_pfn); -pr_info ("Initmem setup node %d [mem %#018Lx-%#018Lx]\n" , nid, - (u64 )start_pfn <<PAGE_SHIFT , - end_pfn ? ((u64 )end_pfn <<PAGE_SHIFT ) -1 :0 ); -#else - start_pfn = node_start_pfn; -#endif +if (!compat) { +get_pfn_range_for_nid (nid, &start_pfn, &end_pfn); +pr_info ("Initmem setup node %d [mem %#018Lx-%#018Lx]\n" , nid, + (u64 )start_pfn <<PAGE_SHIFT , + end_pfn ? ((u64 )end_pfn <<PAGE_SHIFT ) -1 :0 ); + }else { + start_pfn = node_start_pfn; + } +calculate_node_totalpages (pgdat, start_pfn, end_pfn, - zones_size, zholes_size); + zones_size, zholes_size, compat);alloc_node_mem_map (pgdat);pgdat_set_deferred_range (pgdat);@@ -6910,6 +6921,14 @@ void __init free_area_init_node(int nid, unsigned long *zones_size,free_area_init_core (pgdat); } +void __init free_area_init_node (int nid,unsigned long *zones_size, +unsigned long node_start_pfn, +unsigned long *zholes_size) +{ +__free_area_init_node (nid, zones_size, node_start_pfn, zholes_size, + true); +} +#if !defined(CONFIG_FLAT_NODE_MEM_MAP )/* * Initialize all valid struct pages in the range [spfn, epfn) and mark them @@ -6993,8 +7012,6 @@ static inline void __init init_unavailable_mem(void) }#endif /* !CONFIG_FLAT_NODE_MEM_MAP */ -#ifdef CONFIG_HAVE_MEMBLOCK_NODE_MAP -#if MAX_NUMNODES >1 /* * Figure out the number of possible node ids. @@ -7423,8 +7440,8 @@ void __init free_area_init_nodes(unsigned long *max_zone_pfn)init_unavailable_mem ();for_each_online_node (nid) {pg_data_t *pgdat =NODE_DATA (nid); -free_area_init_node (nid,NULL , -find_min_pfn_for_node (nid),NULL ); +__free_area_init_node (nid,NULL , +find_min_pfn_for_node (nid),NULL , false);/* Any memory on that node */ if (pgdat->node_present_pages)@@ -7489,8 +7506,6 @@ static int __init cmdline_parse_movablecore(char *p)early_param ("kernelcore" , cmdline_parse_kernelcore);early_param ("movablecore" , cmdline_parse_movablecore); -#endif /* CONFIG_HAVE_MEMBLOCK_NODE_MAP */ -void adjust_managed_page_count (struct page *page,long count) {atomic_long_add (count, &page_zone (page)->managed_pages);Signature · 1 line
2.25.1
Signature · 2 lines
Sincerely yours, Mike.
Re: [PATCH RFC] mm: remove CONFIG_HAVE_MEMBLOCK_NODE_MAP (was: Re: [PATCH v3 0/5] mm: Enable CONFIG_NODES_SPAN_OTHER_NODES by default for NUMA)
Quoted 3 lines · Mike Rapoport
Hi,
On Wed, Apr 01, 2020 at 01:42:27PM +0800, Baoquan He wrote:
[...]
Quoted 7 lines
Quoted 3 lines
From above information, we can remove
HAVE_MEMBLOCK_NODE_MAP, and replace it withCONFIG_NUMA. That sounds more sensible to store nid into memblock when NUMA support is enabled.Replacing
CONFIG_HAVE_MEMBLOCK_NODE_MAPwithCONFIG_NUMAwill work, but this will not help cleaning up the whole node/zone initialization mess and we'll be stuck with two implementations.
Yeah, this is far from optimal.
Quoted 5 lines
The overhead of enabling
HAVE_MEMBLOCK_NODE_MAPis only for init time as most architectures will anyway discard the entire memblock, so having it in a UMA arch won't be a problem. The only exception is arm that uses memblock forpfn_valid(), here we may also think about a solution to compensate the addition of nid to the memblock structures.
Well, we can make memblock_region->nid defined only for CONFIG_NUMA. memblock_get_region_node would then unconditionally return 0 on UMA. Essentially the same way we do NUMA for other MM code. I only see few direct usage of region->nid.
Signature · 2 lines
Michal Hocko SUSE Labs
Re: [PATCH RFC] mm: remove CONFIG_HAVE_MEMBLOCK_NODE_MAP (was: Re: [PATCH v3 0/5] mm: Enable CONFIG_NODES_SPAN_OTHER_NODES by default for NUMA)
Quoted 21 lines · Michal Hocko
Quoted 3 lines · Mike Rapoport
Hi,
On Wed, Apr 01, 2020 at 01:42:27PM +0800, Baoquan He wrote:
[...]
Quoted 7 lines
Quoted 3 lines
From above information, we can remove
HAVE_MEMBLOCK_NODE_MAP, and replace it withCONFIG_NUMA. That sounds more sensible to store nid into memblock when NUMA support is enabled.Replacing
CONFIG_HAVE_MEMBLOCK_NODE_MAPwithCONFIG_NUMAwill work, but this will not help cleaning up the whole node/zone initialization mess and we'll be stuck with two implementations.Yeah, this is far from optimal.
Quoted 5 lines
The overhead of enabling
HAVE_MEMBLOCK_NODE_MAPis only for init time as most architectures will anyway discard the entire memblock, so having it in a UMA arch won't be a problem. The only exception is arm that uses memblock forpfn_valid(), here we may also think about a solution to compensate the addition of nid to the memblock structures.
Quoted 4 lines
Well, we can make
memblock_region->niddefined only forCONFIG_NUMA.memblock_get_region_nodewould then unconditionally return 0 on UMA. Essentially the same way we do NUMA for other MM code. I only see few direct usage ofregion->nid.
Checked code again, seems HAVE_MEMBLOCK_NODE_MAP is selected directly in all ARCHes which support it. Means HAVE_MEMBLOCK_NODE_MAP is enabled by default on those ARCHes, and has no dependency on CONFIG_NUMA at all. E.g on x86, it just calls free_area_init_nodes() in generic code path, while free_area_init_nodes() is defined in CONFIG_HAVE_MEMBLOCK_NODE_MAP ifdeffery scope. So I tend to agree with Mike to remove HAVE_MEMBLOCK_NODE_MAP firstly on all ARCHes. We can check if it's worth only defining memblock_region->nid for CONFIG_NUMA case after HAVE_MEMBLOCK_NODE_MAP is removed.
config X86
def_bool y ... select HAVE_MEMBLOCK_NODE_MAP ...
Re: [PATCH RFC] mm: remove CONFIG_HAVE_MEMBLOCK_NODE_MAP (was: Re: [PATCH v3 0/5] mm: Enable CONFIG_NODES_SPAN_OTHER_NODES by default for NUMA)
Quoted 27 lines · Baoquan He
Quoted 21 lines · Michal Hocko
Quoted 3 lines · Mike Rapoport
Hi,
On Wed, Apr 01, 2020 at 01:42:27PM +0800, Baoquan He wrote:
[...]
Quoted 7 lines
Quoted 3 lines
From above information, we can remove
HAVE_MEMBLOCK_NODE_MAP, and replace it withCONFIG_NUMA. That sounds more sensible to store nid into memblock when NUMA support is enabled.Replacing
CONFIG_HAVE_MEMBLOCK_NODE_MAPwithCONFIG_NUMAwill work, but this will not help cleaning up the whole node/zone initialization mess and we'll be stuck with two implementations.Yeah, this is far from optimal.
Quoted 5 lines
The overhead of enabling
HAVE_MEMBLOCK_NODE_MAPis only for init time as most architectures will anyway discard the entire memblock, so having it in a UMA arch won't be a problem. The only exception is arm that uses memblock forpfn_valid(), here we may also think about a solution to compensate the addition of nid to the memblock structures.Quoted 4 lines
Well, we can make
memblock_region->niddefined only forCONFIG_NUMA.memblock_get_region_nodewould then unconditionally return 0 on UMA. Essentially the same way we do NUMA for other MM code. I only see few direct usage ofregion->nid.
Quoted 9 lines
Checked code again, seems
HAVE_MEMBLOCK_NODE_MAPis selected directly in all ARCHes which support it. MeansHAVE_MEMBLOCK_NODE_MAPis enabled by default on those ARCHes, and has no dependency onCONFIG_NUMAat all. E.g on x86, it just callsfree_area_init_nodes()in generic code path, whilefree_area_init_nodes()is defined inCONFIG_HAVE_MEMBLOCK_NODE_MAPifdeffery scope. So I tend to agree with Mike to removeHAVE_MEMBLOCK_NODE_MAPfirstly on all ARCHes. We can check if it's worth only definingmemblock_region->nidforCONFIG_NUMAcase afterHAVE_MEMBLOCK_NODE_MAPis removed.
This can surely go in separate patches. What I meant to say is the region->nid is by definition 0 on !CONFIG_NUMA.
Signature · 2 lines
Michal Hocko SUSE Labs
Re: [PATCH RFC] mm: remove CONFIG_HAVE_MEMBLOCK_NODE_MAP (was: Re: [PATCH v3 0/5] mm: Enable CONFIG_NODES_SPAN_OTHER_NODES by default for NUMA)
Quoted 38 lines · Michal Hocko
Quoted 27 lines · Baoquan He
Quoted 21 lines · Michal Hocko
Quoted 3 lines · Mike Rapoport
Hi,
On Wed, Apr 01, 2020 at 01:42:27PM +0800, Baoquan He wrote:
[...]
Quoted 7 lines
Quoted 3 lines
From above information, we can remove
HAVE_MEMBLOCK_NODE_MAP, and replace it withCONFIG_NUMA. That sounds more sensible to store nid into memblock when NUMA support is enabled.Replacing
CONFIG_HAVE_MEMBLOCK_NODE_MAPwithCONFIG_NUMAwill work, but this will not help cleaning up the whole node/zone initialization mess and we'll be stuck with two implementations.Yeah, this is far from optimal.
Quoted 5 lines
The overhead of enabling
HAVE_MEMBLOCK_NODE_MAPis only for init time as most architectures will anyway discard the entire memblock, so having it in a UMA arch won't be a problem. The only exception is arm that uses memblock forpfn_valid(), here we may also think about a solution to compensate the addition of nid to the memblock structures.Quoted 4 lines
Well, we can make
memblock_region->niddefined only forCONFIG_NUMA.memblock_get_region_nodewould then unconditionally return 0 on UMA. Essentially the same way we do NUMA for other MM code. I only see few direct usage ofregion->nid.Quoted 9 lines
Checked code again, seems
HAVE_MEMBLOCK_NODE_MAPis selected directly in all ARCHes which support it. MeansHAVE_MEMBLOCK_NODE_MAPis enabled by default on those ARCHes, and has no dependency onCONFIG_NUMAat all. E.g on x86, it just callsfree_area_init_nodes()in generic code path, whilefree_area_init_nodes()is defined inCONFIG_HAVE_MEMBLOCK_NODE_MAPifdeffery scope. So I tend to agree with Mike to removeHAVE_MEMBLOCK_NODE_MAPfirstly on all ARCHes. We can check if it's worth only definingmemblock_region->nidforCONFIG_NUMAcase afterHAVE_MEMBLOCK_NODE_MAPis removed.
Quoted 2 lines
This can surely go in separate patches. What I meant to say is the
region->nidis by definition 0 on !CONFIG_NUMA.
I see, thanks.
Quoted 30 lines · Michal Hocko
Quoted 17 lines · Hoan Tran
In NUMA layout which nodes have memory ranges that span across other nodes, the mm driver can detect the memory node id incorrectly.
For example, with layout below
Node 0 address: 0000 xxxx 0000 xxxx
Node 1 address: xxxx 1111 xxxx 1111Note:
- Memory from low to high
- 0/1: Node id
- x: Invalid memory of a nodeWhen mm probes the memory map, without
CONFIG_NODES_SPAN_OTHER_NODESconfig, mm only checks the memory validity but not the node id. Because of that, Node 1 also detects the memory from node 0 as below when it scans from the start address to the end address of node 1.Quoted 6 lines
Node 0 address: 0000 xxxx xxxx xxxx
Node 1 address: xxxx 1111 1111 1111This layout could occur on any architecture. Most of them enables this config by default with
CONFIG_NUMA. This patch, by default, enablesCONFIG_NODES_SPAN_OTHER_NODESor usesearly_pfn_in_nid()for NUMA.I am not opposed to this at all. It reduces the config space and that is a good thing on its own. The history has shown that meory layout might be really wild wrt NUMA. The config is only used for
early_pfn_in_nidwhich is clearly an overkill.
Quoted 9 lines
Your description doesn't really explain why this is safe though. The history of this config is somehow messy, though. Mike has tried to remove it
a94b3ab7eab4("[PATCH] mm: remove arch independentNODES_SPAN_OTHER_NODES") just to be reintroduced by7516795739bd("[PATCH] ReintroduceNODES_SPAN_OTHER_NODESfor powerpc") without any reasoning what so ever. This doesn't make it really easy see whether reasons for reintroduction are still there. Maybe there are some subtle dependencies. I do not see any TBH but that might be burried deep in an arch specific code.
Since on all ARCHes NODES_SPAN_OTHER_NODES has dependency on NUMA, replacing it with CONFIG_NUMA seems no risk. Just for those ARCHes which don't have CONFIG_NODES_SPAN_OTHER_NODES before, it involves a tiny performance degradation. Besides, s390 has removed support of NODES_SPAN_OTHER_NODES already.
commit 701dc81e7412daaf3c5bf4bc55d35c8b1525112a
Author: Heiko Carstens <[email protected]> Date: Wed Feb 19 13:29:15 2020 +0100
s390/mm: remove fake numa support
Quoted 29 lines
Quoted 24 lines
v3:
* Revise the patch descriptionV2:
* Revise the patch descriptionHoan Tran (5):
mm: EnableCONFIG_NODES_SPAN_OTHER_NODESby default for NUMA
powerpc:Kconfig: RemoveCONFIG_NODES_SPAN_OTHER_NODES
x86:Kconfig: RemoveCONFIG_NODES_SPAN_OTHER_NODES
sparc:Kconfig: RemoveCONFIG_NODES_SPAN_OTHER_NODES
s390:Kconfig: RemoveCONFIG_NODES_SPAN_OTHER_NODESarch/powerpc/Kconfig | 9 --------- arch/s390/Kconfig | 8 -------- arch/sparc/Kconfig | 9 --------- arch/x86/Kconfig | 9 --------- mm/page_alloc.c | 2 +- 5 files changed, 1 insertion(+), 36 deletions(-)Signature · 1 line
1.8.3.1Signature · 2 lines
Michal Hocko SUSE Labs
Quoted 30 lines · Michal Hocko
Quoted 17 lines · Hoan Tran
In NUMA layout which nodes have memory ranges that span across other nodes, the mm driver can detect the memory node id incorrectly.
For example, with layout below
Node 0 address: 0000 xxxx 0000 xxxx
Node 1 address: xxxx 1111 xxxx 1111Note:
- Memory from low to high
- 0/1: Node id
- x: Invalid memory of a nodeWhen mm probes the memory map, without
CONFIG_NODES_SPAN_OTHER_NODESconfig, mm only checks the memory validity but not the node id. Because of that, Node 1 also detects the memory from node 0 as below when it scans from the start address to the end address of node 1.Quoted 6 lines
Node 0 address: 0000 xxxx xxxx xxxx
Node 1 address: xxxx 1111 1111 1111This layout could occur on any architecture. Most of them enables this config by default with
CONFIG_NUMA. This patch, by default, enablesCONFIG_NODES_SPAN_OTHER_NODESor usesearly_pfn_in_nid()for NUMA.I am not opposed to this at all. It reduces the config space and that is a good thing on its own. The history has shown that meory layout might be really wild wrt NUMA. The config is only used for
early_pfn_in_nidwhich is clearly an overkill.
Quoted 9 lines
Your description doesn't really explain why this is safe though. The history of this config is somehow messy, though. Mike has tried to remove it
a94b3ab7eab4("[PATCH] mm: remove arch independentNODES_SPAN_OTHER_NODES") just to be reintroduced by7516795739bd("[PATCH] ReintroduceNODES_SPAN_OTHER_NODESfor powerpc") without any reasoning what so ever. This doesn't make it really easy see whether reasons for reintroduction are still there. Maybe there are some subtle dependencies. I do not see any TBH but that might be burried deep in an arch specific code.
I've looked at this a bit more and it seems that the check for early_pfn_in_nid() in memmap_init_zone() can be simply removed.
The commits you've mentioned were way before the addition of HAVE_MEMBLOCK_NODE_MAP and the whole infrastructure that calculates zone sizes and boundaries based on the memblock node map.
So, the memmap_init_zone() is called when zone boundaries are already within a node.
I don't have access to machines with memory layout that required this check at the first place, so if anybody who does could test the change below on such machine it would be great.
mm/page_alloc.c +0 -4
@@ -5908,10 +5908,6 @@ void __meminit memmap_init_zone(unsigned long size, int nid, unsigned long zone, pfn =next_pfn (pfn);continue ; } -if (!early_pfn_in_nid (pfn, nid)) { - pfn++; -continue ; - }if (overlap_memmap_init (zone, &pfn))continue ;if (defer_init (nid, pfn, end_pfn))
Quoted 27 lines
Quoted 23 lines
v3:
* Revise the patch descriptionV2:
* Revise the patch descriptionHoan Tran (5):
mm: EnableCONFIG_NODES_SPAN_OTHER_NODESby default for NUMA
powerpc:Kconfig: RemoveCONFIG_NODES_SPAN_OTHER_NODES
x86:Kconfig: RemoveCONFIG_NODES_SPAN_OTHER_NODES
sparc:Kconfig: RemoveCONFIG_NODES_SPAN_OTHER_NODES
s390:Kconfig: RemoveCONFIG_NODES_SPAN_OTHER_NODESarch/powerpc/Kconfig | 9 --------- arch/s390/Kconfig | 8 -------- arch/sparc/Kconfig | 9 --------- arch/x86/Kconfig | 9 --------- mm/page_alloc.c | 2 +- 5 files changed, 1 insertion(+), 36 deletions(-)Signature · 1 line
1.8.3.1Signature · 2 lines
Michal Hocko SUSE Labs
Signature · 2 lines
Sincerely yours, Mike.
Quoted 41 lines · Mike Rapoport
Quoted 30 lines · Michal Hocko
Quoted 17 lines · Hoan Tran
In NUMA layout which nodes have memory ranges that span across other nodes, the mm driver can detect the memory node id incorrectly.
For example, with layout below
Node 0 address: 0000 xxxx 0000 xxxx
Node 1 address: xxxx 1111 xxxx 1111Note:
- Memory from low to high
- 0/1: Node id
- x: Invalid memory of a nodeWhen mm probes the memory map, without
CONFIG_NODES_SPAN_OTHER_NODESconfig, mm only checks the memory validity but not the node id. Because of that, Node 1 also detects the memory from node 0 as below when it scans from the start address to the end address of node 1.Quoted 6 lines
Node 0 address: 0000 xxxx xxxx xxxx
Node 1 address: xxxx 1111 1111 1111This layout could occur on any architecture. Most of them enables this config by default with
CONFIG_NUMA. This patch, by default, enablesCONFIG_NODES_SPAN_OTHER_NODESor usesearly_pfn_in_nid()for NUMA.I am not opposed to this at all. It reduces the config space and that is a good thing on its own. The history has shown that meory layout might be really wild wrt NUMA. The config is only used for
early_pfn_in_nidwhich is clearly an overkill.Quoted 9 lines
Your description doesn't really explain why this is safe though. The history of this config is somehow messy, though. Mike has tried to remove it
a94b3ab7eab4("[PATCH] mm: remove arch independentNODES_SPAN_OTHER_NODES") just to be reintroduced by7516795739bd("[PATCH] ReintroduceNODES_SPAN_OTHER_NODESfor powerpc") without any reasoning what so ever. This doesn't make it really easy see whether reasons for reintroduction are still there. Maybe there are some subtle dependencies. I do not see any TBH but that might be burried deep in an arch specific code.
Quoted 8 lines
I've looked at this a bit more and it seems that the check for
early_pfn_in_nid()inmemmap_init_zone()can be simply removed.The commits you've mentioned were way before the addition of
HAVE_MEMBLOCK_NODE_MAPand the whole infrastructure that calculates zone sizes and boundaries based on the memblock node map.
So, thememmap_init_zone()is called when zone boundaries are already within a node.
But zones from different nodes might overlap in the pfn range. And this check is there to skip over those overlapping areas. The only way to skip over this check I can see is to do a different pfn walk and go through memblock ranges which are guaranteed to belong to a single node.
Signature · 2 lines
Michal Hocko SUSE Labs
Quoted 45 lines · Michal Hocko
Quoted 44 lines · Mike Rapoport
Quoted 30 lines · Michal Hocko
Quoted 17 lines · Hoan Tran
In NUMA layout which nodes have memory ranges that span across other nodes, the mm driver can detect the memory node id incorrectly.
For example, with layout below
Node 0 address: 0000 xxxx 0000 xxxx
Node 1 address: xxxx 1111 xxxx 1111Note:
- Memory from low to high
- 0/1: Node id
- x: Invalid memory of a nodeWhen mm probes the memory map, without
CONFIG_NODES_SPAN_OTHER_NODESconfig, mm only checks the memory validity but not the node id. Because of that, Node 1 also detects the memory from node 0 as below when it scans from the start address to the end address of node 1.Quoted 6 lines
Node 0 address: 0000 xxxx xxxx xxxx
Node 1 address: xxxx 1111 1111 1111This layout could occur on any architecture. Most of them enables this config by default with
CONFIG_NUMA. This patch, by default, enablesCONFIG_NODES_SPAN_OTHER_NODESor usesearly_pfn_in_nid()for NUMA.I am not opposed to this at all. It reduces the config space and that is a good thing on its own. The history has shown that meory layout might be really wild wrt NUMA. The config is only used for
early_pfn_in_nidwhich is clearly an overkill.Quoted 9 lines
Your description doesn't really explain why this is safe though. The history of this config is somehow messy, though. Mike has tried to remove it
a94b3ab7eab4("[PATCH] mm: remove arch independentNODES_SPAN_OTHER_NODES") just to be reintroduced by7516795739bd("[PATCH] ReintroduceNODES_SPAN_OTHER_NODESfor powerpc") without any reasoning what so ever. This doesn't make it really easy see whether reasons for reintroduction are still there. Maybe there are some subtle dependencies. I do not see any TBH but that might be burried deep in an arch specific code.I've looked at this a bit more and it seems that the check for
early_pfn_in_nid()inmemmap_init_zone()can be simply removed.
Quoted 8 lines
Quoted 5 lines
The commits you've mentioned were way before the addition of
HAVE_MEMBLOCK_NODE_MAPand the whole infrastructure that calculates zone sizes and boundaries based on the memblock node map.
So, thememmap_init_zone()is called when zone boundaries are already within a node.But zones from different nodes might overlap in the pfn range. And this check is there to skip over those overlapping areas.
Maybe I mis-read the code, but I don't see how this could happen. In the HAVE_MEMBLOCK_NODE_MAP=y case, free_area_init_node() calls calculate_node_totalpages() that ensures that node->node_zones are entirely within the node because this is checked in zone_spanned_pages_in_node(). So, for zones from different nodes to overlap in the pfn range the nodes themself should overlap. Is this even possible?
Quoted 6 lines
The only way to skip over this check I can see is to do a different pfn walk and go through memblock ranges which are guaranteed to belong to a single node.
Signature · 2 lines
Michal Hocko SUSE Labs
Signature · 2 lines
Sincerely yours, Mike.
Quoted 52 lines · Mike Rapoport
Quoted 51 lines · Michal Hocko
Quoted 49 lines · Mike Rapoport
Quoted 30 lines · Michal Hocko
Quoted 17 lines · Hoan Tran
In NUMA layout which nodes have memory ranges that span across other nodes, the mm driver can detect the memory node id incorrectly.
For example, with layout below
Node 0 address: 0000 xxxx 0000 xxxx
Node 1 address: xxxx 1111 xxxx 1111Note:
- Memory from low to high
- 0/1: Node id
- x: Invalid memory of a nodeWhen mm probes the memory map, without
CONFIG_NODES_SPAN_OTHER_NODESconfig, mm only checks the memory validity but not the node id. Because of that, Node 1 also detects the memory from node 0 as below when it scans from the start address to the end address of node 1.Quoted 6 lines
Node 0 address: 0000 xxxx xxxx xxxx
Node 1 address: xxxx 1111 1111 1111This layout could occur on any architecture. Most of them enables this config by default with
CONFIG_NUMA. This patch, by default, enablesCONFIG_NODES_SPAN_OTHER_NODESor usesearly_pfn_in_nid()for NUMA.I am not opposed to this at all. It reduces the config space and that is a good thing on its own. The history has shown that meory layout might be really wild wrt NUMA. The config is only used for
early_pfn_in_nidwhich is clearly an overkill.Quoted 9 lines
Your description doesn't really explain why this is safe though. The history of this config is somehow messy, though. Mike has tried to remove it
a94b3ab7eab4("[PATCH] mm: remove arch independentNODES_SPAN_OTHER_NODES") just to be reintroduced by7516795739bd("[PATCH] ReintroduceNODES_SPAN_OTHER_NODESfor powerpc") without any reasoning what so ever. This doesn't make it really easy see whether reasons for reintroduction are still there. Maybe there are some subtle dependencies. I do not see any TBH but that might be burried deep in an arch specific code.I've looked at this a bit more and it seems that the check for
early_pfn_in_nid()inmemmap_init_zone()can be simply removed.The commits you've mentioned were way before the addition of
HAVE_MEMBLOCK_NODE_MAPand the whole infrastructure that calculates zone sizes and boundaries based on the memblock node map.
So, thememmap_init_zone()is called when zone boundaries are already within a node.
Quoted 7 lines
Quoted 2 lines
But zones from different nodes might overlap in the pfn range. And this check is there to skip over those overlapping areas.
Maybe I mis-read the code, but I don't see how this could happen. In the HAVE_MEMBLOCK_NODE_MAP=y case,
free_area_init_node()callscalculate_node_totalpages()that ensures thatnode->node_zonesare entirely within the node because this is checked inzone_spanned_pages_in_node().
zone_spanned_pages_in_node does chech the zone boundaries are within the node boundaries. But that doesn't really tell anything about other potential zones interleaving with the physical memory range. zone->spanned_pages simply gives the physical range for the zone including holes. Interleaving nodes are essentially a hole (__absent_pages_in_range is going to skip those).
That means that when free_area_init_core simply goes over the whole physical zone range including holes and that is why we need to check both for physical and logical holes (aka other nodes).
The life would be so much easier if the whole thing would simply iterate over memblocks...
Signature · 2 lines
Michal Hocko SUSE Labs
Hi Michal,
Quoted 13 lines · Michal Hocko
Quoted 4 lines · Mike Rapoport
Maybe I mis-read the code, but I don't see how this could happen. In the HAVE_MEMBLOCK_NODE_MAP=y case,
free_area_init_node()callscalculate_node_totalpages()that ensures thatnode->node_zonesare entirely within the node because this is checked inzone_spanned_pages_in_node().
zone_spanned_pages_in_nodedoes chech the zone boundaries are within the node boundaries. But that doesn't really tell anything about other potential zones interleaving with the physical memory range.zone->spanned_pagessimply gives the physical range for the zone including holes. Interleaving nodes are essentially a hole (__absent_pages_in_rangeis going to skip those).
Quoted 6 lines
That means that when
free_area_init_coresimply goes over the whole physical zone range including holes and that is why we need to check both for physical and logical holes (aka other nodes).The life would be so much easier if the whole thing would simply iterate over memblocks...
The memblock iterating sounds a great idea. I tried with putting the memblock iterating in the upper layer, memmap_init(), which is used for boot mem only anyway. Do you think it's doable and OK? It yes, I can work out a formal patch to make this simpler as you said. The draft code is as below. Like this it uses the existing code and involves little change.
mm/page_alloc.c +10 -10
@@ -6007,14 +6007,6 @@ void __meminit memmap_init_zone(unsigned long size, int nid, unsigned long zone,* function. They do not exist on hotplugged memory. */ if (context =MEMMAP_EARLY ) { -if (!early_pfn_valid (pfn)) { - pfn =next_pfn (pfn); -continue ; - } -if (!early_pfn_in_nid (pfn, nid)) { - pfn++; -continue ; - }if (overlap_memmap_init (zone, &pfn))continue ;if (defer_init (nid, pfn, end_pfn))
@@ -6130,9 +6122,17 @@ static void __meminit zone_init_free_lists(struct zone *zone) }void __meminit __weak memmap_init (unsigned long size,int nid, -unsigned long zone,unsigned long start_pfn) +unsigned long zone,unsigned long range_start_pfn) { -memmap_init_zone (size, nid, zone, start_pfn,MEMMAP_EARLY ,NULL ); +unsigned long start_pfn, end_pfn; +unsigned long range_end_pfn = range_start_pfn + size; +int i; +for_each_mem_pfn_range (i, nid, &start_pfn, &end_pfn,NULL ) { + start_pfn =clamp (start_pfn, range_start_pfn, range_end_pfn); + end_pfn =clamp (end_pfn, range_start_pfn, range_end_pfn); +if (end_pfn > start_pfn) +memmap_init_zone (size, nid, zone, start_pfn,MEMMAP_EARLY ,NULL ); + } }static int zone_batchsize (struct zone *zone)
Quoted 20 lines · Baoquan He
Hi Michal,
Quoted 17 lines · Michal Hocko
Quoted 4 lines · Mike Rapoport
Maybe I mis-read the code, but I don't see how this could happen. In the HAVE_MEMBLOCK_NODE_MAP=y case,
free_area_init_node()callscalculate_node_totalpages()that ensures thatnode->node_zonesare entirely within the node because this is checked inzone_spanned_pages_in_node().
zone_spanned_pages_in_nodedoes chech the zone boundaries are within the node boundaries. But that doesn't really tell anything about other potential zones interleaving with the physical memory range.zone->spanned_pagessimply gives the physical range for the zone including holes. Interleaving nodes are essentially a hole (__absent_pages_in_rangeis going to skip those).That means that when
free_area_init_coresimply goes over the whole physical zone range including holes and that is why we need to check both for physical and logical holes (aka other nodes).
Quoted 8 lines
Quoted 2 lines
The life would be so much easier if the whole thing would simply iterate over memblocks...
The memblock iterating sounds a great idea. I tried with putting the memblock iterating in the upper layer,
memmap_init(), which is used for boot mem only anyway. Do you think it's doable and OK? It yes, I can work out a formal patch to make this simpler as you said. The draft code is as below. Like this it uses the existing code and involves little change.
Doing this would be a step in the right direction! I haven't checked the code very closely though. The below sounds way too simple to be truth I am afraid. First for_each_mem_pfn_range is available only for CONFIG_HAVE_MEMBLOCK_NODE_MAP (which is one of the reasons why I keep saying that I really hate that being conditional). Also I haven't really checked the deferred initialization path - I have a very vague recollection that it has been converted to the memblock api but I have happilly dropped all that memory.
Quoted 39 lines
mm/page_alloc.c +10
-10@@ -6007,14 +6007,6 @@ void __meminit memmap_init_zone(unsigned long size, int nid, unsigned long zone,* function. They do not exist on hotplugged memory. */ if (context =MEMMAP_EARLY ) { -if (!early_pfn_valid (pfn)) { - pfn =next_pfn (pfn); -continue ; - } -if (!early_pfn_in_nid (pfn, nid)) { - pfn++; -continue ; - }if (overlap_memmap_init (zone, &pfn))continue ;if (defer_init (nid, pfn, end_pfn))@@ -6130,9 +6122,17 @@ static void __meminit zone_init_free_lists(struct zone *zone) }void __meminit __weak memmap_init (unsigned long size,int nid, -unsigned long zone,unsigned long start_pfn) +unsigned long zone,unsigned long range_start_pfn) { -memmap_init_zone (size, nid, zone, start_pfn,MEMMAP_EARLY ,NULL ); +unsigned long start_pfn, end_pfn; +unsigned long range_end_pfn = range_start_pfn + size; +int i; +for_each_mem_pfn_range (i, nid, &start_pfn, &end_pfn,NULL ) { + start_pfn =clamp (start_pfn, range_start_pfn, range_end_pfn); + end_pfn =clamp (end_pfn, range_start_pfn, range_end_pfn); +if (end_pfn > start_pfn) +memmap_init_zone (size, nid, zone, start_pfn,MEMMAP_EARLY ,NULL ); + } }static int zone_batchsize (struct zone *zone)
Signature · 2 lines
Michal Hocko SUSE Labs
Quoted 30 lines · Michal Hocko
Quoted 20 lines · Baoquan He
Hi Michal,
Quoted 17 lines · Michal Hocko
Quoted 4 lines · Mike Rapoport
Maybe I mis-read the code, but I don't see how this could happen. In the HAVE_MEMBLOCK_NODE_MAP=y case,
free_area_init_node()callscalculate_node_totalpages()that ensures thatnode->node_zonesare entirely within the node because this is checked inzone_spanned_pages_in_node().
zone_spanned_pages_in_nodedoes chech the zone boundaries are within the node boundaries. But that doesn't really tell anything about other potential zones interleaving with the physical memory range.zone->spanned_pagessimply gives the physical range for the zone including holes. Interleaving nodes are essentially a hole (__absent_pages_in_rangeis going to skip those).That means that when
free_area_init_coresimply goes over the whole physical zone range including holes and that is why we need to check both for physical and logical holes (aka other nodes).Quoted 8 lines
Quoted 2 lines
The life would be so much easier if the whole thing would simply iterate over memblocks...
The memblock iterating sounds a great idea. I tried with putting the memblock iterating in the upper layer,
memmap_init(), which is used for boot mem only anyway. Do you think it's doable and OK? It yes, I can work out a formal patch to make this simpler as you said. The draft code is as below. Like this it uses the existing code and involves little change.
Quoted 8 lines
Doing this would be a step in the right direction! I haven't checked the code very closely though. The below sounds way too simple to be truth I am afraid. First
for_each_mem_pfn_rangeis available only forCONFIG_HAVE_MEMBLOCK_NODE_MAP(which is one of the reasons why I keep saying that I really hate that being conditional). Also I haven't really checked the deferred initialization path - I have a very vague recollection that it has been converted to the memblock api but I have happilly dropped all that memory.
Thanks for your quick response and pointing out the rest suspect aspects, I will investigate what you mentioned, see if they impact.
Quoted 45 lines
Quoted 40 lines
mm/page_alloc.c +10
-10@@ -6007,14 +6007,6 @@ void __meminit memmap_init_zone(unsigned long size, int nid, unsigned long zone,* function. They do not exist on hotplugged memory. */ if (context =MEMMAP_EARLY ) { -if (!early_pfn_valid (pfn)) { - pfn =next_pfn (pfn); -continue ; - } -if (!early_pfn_in_nid (pfn, nid)) { - pfn++; -continue ; - }if (overlap_memmap_init (zone, &pfn))continue ;if (defer_init (nid, pfn, end_pfn))@@ -6130,9 +6122,17 @@ static void __meminit zone_init_free_lists(struct zone *zone) }void __meminit __weak memmap_init (unsigned long size,int nid, -unsigned long zone,unsigned long start_pfn) +unsigned long zone,unsigned long range_start_pfn) { -memmap_init_zone (size, nid, zone, start_pfn,MEMMAP_EARLY ,NULL ); +unsigned long start_pfn, end_pfn; +unsigned long range_end_pfn = range_start_pfn + size; +int i; +for_each_mem_pfn_range (i, nid, &start_pfn, &end_pfn,NULL ) { + start_pfn =clamp (start_pfn, range_start_pfn, range_end_pfn); + end_pfn =clamp (end_pfn, range_start_pfn, range_end_pfn); +if (end_pfn > start_pfn) +memmap_init_zone (size, nid, zone, start_pfn,MEMMAP_EARLY ,NULL ); + } }static int zone_batchsize (struct zone *zone)Signature · 2 lines
Michal Hocko SUSE Labs
Hi All,
Quoted 40 lines · Baoquan He
Quoted 30 lines · Michal Hocko
Quoted 20 lines · Baoquan He
Hi Michal,
Quoted 17 lines · Michal Hocko
Quoted 4 lines · Mike Rapoport
Maybe I mis-read the code, but I don't see how this could happen. In the HAVE_MEMBLOCK_NODE_MAP=y case,
free_area_init_node()callscalculate_node_totalpages()that ensures thatnode->node_zonesare entirely within the node because this is checked inzone_spanned_pages_in_node().
zone_spanned_pages_in_nodedoes chech the zone boundaries are within the node boundaries. But that doesn't really tell anything about other potential zones interleaving with the physical memory range.zone->spanned_pagessimply gives the physical range for the zone including holes. Interleaving nodes are essentially a hole (__absent_pages_in_rangeis going to skip those).That means that when
free_area_init_coresimply goes over the whole physical zone range including holes and that is why we need to check both for physical and logical holes (aka other nodes).Quoted 8 lines
Quoted 2 lines
The life would be so much easier if the whole thing would simply iterate over memblocks...
The memblock iterating sounds a great idea. I tried with putting the memblock iterating in the upper layer,
memmap_init(), which is used for boot mem only anyway. Do you think it's doable and OK? It yes, I can work out a formal patch to make this simpler as you said. The draft code is as below. Like this it uses the existing code and involves little change.Quoted 8 lines
Doing this would be a step in the right direction! I haven't checked the code very closely though. The below sounds way too simple to be truth I am afraid. First
for_each_mem_pfn_rangeis available only forCONFIG_HAVE_MEMBLOCK_NODE_MAP(which is one of the reasons why I keep saying that I really hate that being conditional). Also I haven't really checked the deferred initialization path - I have a very vague recollection that it has been converted to the memblock api but I have happilly dropped all that memory.
Quoted 2 lines
Thanks for your quick response and pointing out the rest suspect aspects, I will investigate what you mentioned, see if they impact.
I would like to check if we still move on with my patch to remove CONFIG_NODES_SPAN_OTHER_NODES and have another patch on top it?
Thanks
Hoan
Quoted 47 lines
Quoted 46 lines
Quoted 41 lines
mm/page_alloc.c +10
-10@@ -6007,14 +6007,6 @@ void __meminit memmap_init_zone(unsigned long size, int nid, unsigned long zone,* function. They do not exist on hotplugged memory. */ if (context =MEMMAP_EARLY ) { -if (!early_pfn_valid (pfn)) { - pfn =next_pfn (pfn); -continue ; - } -if (!early_pfn_in_nid (pfn, nid)) { - pfn++; -continue ; - }if (overlap_memmap_init (zone, &pfn))continue ;if (defer_init (nid, pfn, end_pfn))@@ -6130,9 +6122,17 @@ static void __meminit zone_init_free_lists(struct zone *zone) }void __meminit __weak memmap_init (unsigned long size,int nid, -unsigned long zone,unsigned long start_pfn) +unsigned long zone,unsigned long range_start_pfn) { -memmap_init_zone (size, nid, zone, start_pfn,MEMMAP_EARLY ,NULL ); +unsigned long start_pfn, end_pfn; +unsigned long range_end_pfn = range_start_pfn + size; +int i; +for_each_mem_pfn_range (i, nid, &start_pfn, &end_pfn,NULL ) { + start_pfn =clamp (start_pfn, range_start_pfn, range_end_pfn); + end_pfn =clamp (end_pfn, range_start_pfn, range_end_pfn); +if (end_pfn > start_pfn) +memmap_init_zone (size, nid, zone, start_pfn,MEMMAP_EARLY ,NULL ); + } }static int zone_batchsize (struct zone *zone)Signature · 2 lines
Michal Hocko SUSE Labs
Quoted 43 lines · Hoan Tran
Hi All,
Quoted 40 lines · Baoquan He
Quoted 38 lines · Michal Hocko
Quoted 20 lines · Baoquan He
Hi Michal,
Quoted 17 lines · Michal Hocko
Quoted 4 lines · Mike Rapoport
Maybe I mis-read the code, but I don't see how this could happen. In the HAVE_MEMBLOCK_NODE_MAP=y case,
free_area_init_node()callscalculate_node_totalpages()that ensures thatnode->node_zonesare entirely within the node because this is checked inzone_spanned_pages_in_node().
zone_spanned_pages_in_nodedoes chech the zone boundaries are within the node boundaries. But that doesn't really tell anything about other potential zones interleaving with the physical memory range.zone->spanned_pagessimply gives the physical range for the zone including holes. Interleaving nodes are essentially a hole (__absent_pages_in_rangeis going to skip those).That means that when
free_area_init_coresimply goes over the whole physical zone range including holes and that is why we need to check both for physical and logical holes (aka other nodes).Quoted 8 lines
Quoted 2 lines
The life would be so much easier if the whole thing would simply iterate over memblocks...
The memblock iterating sounds a great idea. I tried with putting the memblock iterating in the upper layer,
memmap_init(), which is used for boot mem only anyway. Do you think it's doable and OK? It yes, I can work out a formal patch to make this simpler as you said. The draft code is as below. Like this it uses the existing code and involves little change.Doing this would be a step in the right direction! I haven't checked the code very closely though. The below sounds way too simple to be truth I am afraid. First
for_each_mem_pfn_rangeis available only forCONFIG_HAVE_MEMBLOCK_NODE_MAP(which is one of the reasons why I keep saying that I really hate that being conditional). Also I haven't really checked the deferred initialization path - I have a very vague recollection that it has been converted to the memblock api but I have happilly dropped all that memory.
Quoted 5 lines
Quoted 2 lines
Thanks for your quick response and pointing out the rest suspect aspects, I will investigate what you mentioned, see if they impact.
I would like to check if we still move on with my patch to remove
CONFIG_NODES_SPAN_OTHER_NODESand have another patch on top it?
I think we would like to replace CONFIG_NODES_SPAN_OTHER_NODES with CONFIG_NUMA, and just let UMA return 0 as node id, as Michal replied in another mail. Anyway, your patch 2~5 are still needed to sit on top of the change of this new plan.
Hi,
Quoted 47 lines · Baoquan He
Quoted 46 lines · Hoan Tran
Hi All,
Quoted 42 lines · Baoquan He
Quoted 30 lines · Michal Hocko
Quoted 20 lines · Baoquan He
Hi Michal,
Quoted 17 lines · Michal Hocko
Quoted 4 lines · Mike Rapoport
Maybe I mis-read the code, but I don't see how this could happen. In the HAVE_MEMBLOCK_NODE_MAP=y case,
free_area_init_node()callscalculate_node_totalpages()that ensures thatnode->node_zonesare entirely within the node because this is checked inzone_spanned_pages_in_node().
zone_spanned_pages_in_nodedoes chech the zone boundaries are within the node boundaries. But that doesn't really tell anything about other potential zones interleaving with the physical memory range.zone->spanned_pagessimply gives the physical range for the zone including holes. Interleaving nodes are essentially a hole (__absent_pages_in_rangeis going to skip those).That means that when
free_area_init_coresimply goes over the whole physical zone range including holes and that is why we need to check both for physical and logical holes (aka other nodes).Quoted 8 lines
Quoted 2 lines
The life would be so much easier if the whole thing would simply iterate over memblocks...
The memblock iterating sounds a great idea. I tried with putting the memblock iterating in the upper layer,
memmap_init(), which is used for boot mem only anyway. Do you think it's doable and OK? It yes, I can work out a formal patch to make this simpler as you said. The draft code is as below. Like this it uses the existing code and involves little change.Quoted 8 lines
Doing this would be a step in the right direction! I haven't checked the code very closely though. The below sounds way too simple to be truth I am afraid. First
for_each_mem_pfn_rangeis available only forCONFIG_HAVE_MEMBLOCK_NODE_MAP(which is one of the reasons why I keep saying that I really hate that being conditional). Also I haven't really checked the deferred initialization path - I have a very vague recollection that it has been converted to the memblock api but I have happilly dropped all that memory.Thanks for your quick response and pointing out the rest suspect aspects, I will investigate what you mentioned, see if they impact.
Quoted 7 lines
Quoted 2 lines
I would like to check if we still move on with my patch to remove
CONFIG_NODES_SPAN_OTHER_NODESand have another patch on top it?I think we would like to replace
CONFIG_NODES_SPAN_OTHER_NODESwithCONFIG_NUMA, and just let UMA return 0 as node id, as Michal replied in another mail. Anyway, your patch 2~5 are still needed to sit on top of the change of this new plan.
Got it. Thanks for quick response.
Regards
Hoan
Quoted 1 line
Quoted 30 lines · Michal Hocko
Quoted 20 lines · Baoquan He
Hi Michal,
Quoted 17 lines · Michal Hocko
Quoted 4 lines · Mike Rapoport
Maybe I mis-read the code, but I don't see how this could happen. In the HAVE_MEMBLOCK_NODE_MAP=y case,
free_area_init_node()callscalculate_node_totalpages()that ensures thatnode->node_zonesare entirely within the node because this is checked inzone_spanned_pages_in_node().
zone_spanned_pages_in_nodedoes chech the zone boundaries are within the node boundaries. But that doesn't really tell anything about other potential zones interleaving with the physical memory range.zone->spanned_pagessimply gives the physical range for the zone including holes. Interleaving nodes are essentially a hole (__absent_pages_in_rangeis going to skip those).That means that when
free_area_init_coresimply goes over the whole physical zone range including holes and that is why we need to check both for physical and logical holes (aka other nodes).Quoted 8 lines
Quoted 2 lines
The life would be so much easier if the whole thing would simply iterate over memblocks...
The memblock iterating sounds a great idea. I tried with putting the memblock iterating in the upper layer,
memmap_init(), which is used for boot mem only anyway. Do you think it's doable and OK? It yes, I can work out a formal patch to make this simpler as you said. The draft code is as below. Like this it uses the existing code and involves little change.
Quoted 8 lines
Doing this would be a step in the right direction! I haven't checked the code very closely though. The below sounds way too simple to be truth I am afraid. First
for_each_mem_pfn_rangeis available only forCONFIG_HAVE_MEMBLOCK_NODE_MAP(which is one of the reasons why I keep saying that I really hate that being conditional). Also I haven't really checked the deferred initialization path - I have a very vague recollection that it has been converted to the memblock api but I have happilly dropped all that memory.
The Baoquan's patch almost did it, at least for simple case of qemu with 2 nodes. It's only missing the adjustment to the size passed to memmap_init_zone() as it may change because of clamping.
I've drafted something that removes HAVE_MEMBLOCK_NODE_MAP and added this patch there [1]. For several memory configurations I could emulate with qemu it worked.
I'm going to wait a bit to see of kbuild is happy and then I'll send the patches.
Baoquan, I took liberty to add your SoB, hope you don't mind.
[1] https://git.kernel.org/pub/scm/linux/kernel/git/rppt/linux.git/log/?h=memblock/all-have-node-map
Quoted 43 lines
Quoted 39 lines
mm/page_alloc.c +10
-10@@ -6007,14 +6007,6 @@ void __meminit memmap_init_zone(unsigned long size, int nid, unsigned long zone,* function. They do not exist on hotplugged memory. */ if (context =MEMMAP_EARLY ) { -if (!early_pfn_valid (pfn)) { - pfn =next_pfn (pfn); -continue ; - } -if (!early_pfn_in_nid (pfn, nid)) { - pfn++; -continue ; - }if (overlap_memmap_init (zone, &pfn))continue ;if (defer_init (nid, pfn, end_pfn))@@ -6130,9 +6122,17 @@ static void __meminit zone_init_free_lists(struct zone *zone) }void __meminit __weak memmap_init (unsigned long size,int nid, -unsigned long zone,unsigned long start_pfn) +unsigned long zone,unsigned long range_start_pfn) { -memmap_init_zone (size, nid, zone, start_pfn,MEMMAP_EARLY ,NULL ); +unsigned long start_pfn, end_pfn; +unsigned long range_end_pfn = range_start_pfn + size; +int i; +for_each_mem_pfn_range (i, nid, &start_pfn, &end_pfn,NULL ) { + start_pfn =clamp (start_pfn, range_start_pfn, range_end_pfn); + end_pfn =clamp (end_pfn, range_start_pfn, range_end_pfn); +if (end_pfn > start_pfn) +memmap_init_zone (size, nid, zone, start_pfn,MEMMAP_EARLY ,NULL ); + } }static int zone_batchsize (struct zone *zone)Signature · 2 lines
Michal Hocko SUSE Labs
Signature · 2 lines
Sincerely yours, Mike.
Quoted 40 lines · Mike Rapoport
Quoted 30 lines · Michal Hocko
Quoted 20 lines · Baoquan He
Hi Michal,
Quoted 17 lines · Michal Hocko
Quoted 4 lines · Mike Rapoport
Maybe I mis-read the code, but I don't see how this could happen. In the HAVE_MEMBLOCK_NODE_MAP=y case,
free_area_init_node()callscalculate_node_totalpages()that ensures thatnode->node_zonesare entirely within the node because this is checked inzone_spanned_pages_in_node().
zone_spanned_pages_in_nodedoes chech the zone boundaries are within the node boundaries. But that doesn't really tell anything about other potential zones interleaving with the physical memory range.zone->spanned_pagessimply gives the physical range for the zone including holes. Interleaving nodes are essentially a hole (__absent_pages_in_rangeis going to skip those).That means that when
free_area_init_coresimply goes over the whole physical zone range including holes and that is why we need to check both for physical and logical holes (aka other nodes).Quoted 8 lines
Quoted 2 lines
The life would be so much easier if the whole thing would simply iterate over memblocks...
The memblock iterating sounds a great idea. I tried with putting the memblock iterating in the upper layer,
memmap_init(), which is used for boot mem only anyway. Do you think it's doable and OK? It yes, I can work out a formal patch to make this simpler as you said. The draft code is as below. Like this it uses the existing code and involves little change.Quoted 8 lines
Doing this would be a step in the right direction! I haven't checked the code very closely though. The below sounds way too simple to be truth I am afraid. First
for_each_mem_pfn_rangeis available only forCONFIG_HAVE_MEMBLOCK_NODE_MAP(which is one of the reasons why I keep saying that I really hate that being conditional). Also I haven't really checked the deferred initialization path - I have a very vague recollection that it has been converted to the memblock api but I have happilly dropped all that memory.
Quoted 3 lines
The Baoquan's patch almost did it, at least for simple case of qemu with 2 nodes. It's only missing the adjustment to the size passed to
memmap_init_zone()as it may change because of clamping.
Right, the size need be adjusted after start and end clamping.
Quoted 10 lines
I've drafted something that removes
HAVE_MEMBLOCK_NODE_MAPand added this patch there [1]. For several memory configurations I could emulate with qemu it worked.
I'm going to wait a bit to see of kbuild is happy and then I'll send the patches.Baoquan, I took liberty to add your SoB, hope you don't mind.
[1] https://git.kernel.org/pub/scm/linux/kernel/git/rppt/linux.git/log/?h=memblock/all-have-node-map
Of course not. Thanks for doing this, and look forward to seeing your formal patchset posting when it's ready.
Quoted 50 lines
Quoted 44 lines
Quoted 40 lines
mm/page_alloc.c +10
-10@@ -6007,14 +6007,6 @@ void __meminit memmap_init_zone(unsigned long size, int nid, unsigned long zone,* function. They do not exist on hotplugged memory. */ if (context =MEMMAP_EARLY ) { -if (!early_pfn_valid (pfn)) { - pfn =next_pfn (pfn); -continue ; - } -if (!early_pfn_in_nid (pfn, nid)) { - pfn++; -continue ; - }if (overlap_memmap_init (zone, &pfn))continue ;if (defer_init (nid, pfn, end_pfn))@@ -6130,9 +6122,17 @@ static void __meminit zone_init_free_lists(struct zone *zone) }void __meminit __weak memmap_init (unsigned long size,int nid, -unsigned long zone,unsigned long start_pfn) +unsigned long zone,unsigned long range_start_pfn) { -memmap_init_zone (size, nid, zone, start_pfn,MEMMAP_EARLY ,NULL ); +unsigned long start_pfn, end_pfn; +unsigned long range_end_pfn = range_start_pfn + size; +int i; +for_each_mem_pfn_range (i, nid, &start_pfn, &end_pfn,NULL ) { + start_pfn =clamp (start_pfn, range_start_pfn, range_end_pfn); + end_pfn =clamp (end_pfn, range_start_pfn, range_end_pfn); +if (end_pfn > start_pfn) +memmap_init_zone (size, nid, zone, start_pfn,MEMMAP_EARLY ,NULL ); + } }static int zone_batchsize (struct zone *zone)Signature · 2 lines
Michal Hocko SUSE LabsSignature · 2 lines
Sincerely yours, Mike.