License cleanup: add SPDX GPL-2.0 license identifier to files with no license
Many source files in the tree are missing licensing information, which
makes it harder for compliance tools to determine the correct license.
By default all files without license information are under the default
license of the kernel, which is GPL version 2.
Update the files which contain no license information with the 'GPL-2.0'
SPDX license identifier. The SPDX identifier is a legally binding
shorthand, which can be used instead of the full boiler plate text.
This patch is based on work done by Thomas Gleixner and Kate Stewart and
Philippe Ombredanne.
How this work was done:
Patches were generated and checked against linux-4.14-rc6 for a subset of
the use cases:
- file had no licensing information it it.
- file was a */uapi/* one with no licensing information in it,
- file was a */uapi/* one with existing licensing information,
Further patches will be generated in subsequent months to fix up cases
where non-standard license headers were used, and references to license
had to be inferred by heuristics based on keywords.
The analysis to determine which SPDX License Identifier to be applied to
a file was done in a spreadsheet of side by side results from of the
output of two independent scanners (ScanCode & Windriver) producing SPDX
tag:value files created by Philippe Ombredanne. Philippe prepared the
base worksheet, and did an initial spot review of a few 1000 files.
The 4.13 kernel was the starting point of the analysis with 60,537 files
assessed. Kate Stewart did a file by file comparison of the scanner
results in the spreadsheet to determine which SPDX license identifier(s)
to be applied to the file. She confirmed any determination that was not
immediately clear with lawyers working with the Linux Foundation.
Criteria used to select files for SPDX license identifier tagging was:
- Files considered eligible had to be source code files.
- Make and config files were included as candidates if they contained >5
lines of source
- File already had some variant of a license header in it (even if <5
lines).
All documentation files were explicitly excluded.
The following heuristics were used to determine which SPDX license
identifiers to apply.
- when both scanners couldn't find any license traces, file was
considered to have no license information in it, and the top level
COPYING file license applied.
For non */uapi/* files that summary was:
SPDX license identifier # files
---------------------------------------------------|-------
GPL-2.0 11139
and resulted in the first patch in this series.
If that file was a */uapi/* path one, it was "GPL-2.0 WITH
Linux-syscall-note" otherwise it was "GPL-2.0". Results of that was:
SPDX license identifier # files
---------------------------------------------------|-------
GPL-2.0 WITH Linux-syscall-note 930
and resulted in the second patch in this series.
- if a file had some form of licensing information in it, and was one
of the */uapi/* ones, it was denoted with the Linux-syscall-note if
any GPL family license was found in the file or had no licensing in
it (per prior point). Results summary:
SPDX license identifier # files
---------------------------------------------------|------
GPL-2.0 WITH Linux-syscall-note 270
GPL-2.0+ WITH Linux-syscall-note 169
((GPL-2.0 WITH Linux-syscall-note) OR BSD-2-Clause) 21
((GPL-2.0 WITH Linux-syscall-note) OR BSD-3-Clause) 17
LGPL-2.1+ WITH Linux-syscall-note 15
GPL-1.0+ WITH Linux-syscall-note 14
((GPL-2.0+ WITH Linux-syscall-note) OR BSD-3-Clause) 5
LGPL-2.0+ WITH Linux-syscall-note 4
LGPL-2.1 WITH Linux-syscall-note 3
((GPL-2.0 WITH Linux-syscall-note) OR MIT) 3
((GPL-2.0 WITH Linux-syscall-note) AND MIT) 1
and that resulted in the third patch in this series.
- when the two scanners agreed on the detected license(s), that became
the concluded license(s).
- when there was disagreement between the two scanners (one detected a
license but the other didn't, or they both detected different
licenses) a manual inspection of the file occurred.
- In most cases a manual inspection of the information in the file
resulted in a clear resolution of the license that should apply (and
which scanner probably needed to revisit its heuristics).
- When it was not immediately clear, the license identifier was
confirmed with lawyers working with the Linux Foundation.
- If there was any question as to the appropriate license identifier,
the file was flagged for further research and to be revisited later
in time.
In total, over 70 hours of logged manual review was done on the
spreadsheet to determine the SPDX license identifiers to apply to the
source files by Kate, Philippe, Thomas and, in some cases, confirmation
by lawyers working with the Linux Foundation.
Kate also obtained a third independent scan of the 4.13 code base from
FOSSology, and compared selected files where the other two scanners
disagreed against that SPDX file, to see if there was new insights. The
Windriver scanner is based on an older version of FOSSology in part, so
they are related.
Thomas did random spot checks in about 500 files from the spreadsheets
for the uapi headers and agreed with SPDX license identifier in the
files he inspected. For the non-uapi files Thomas did random spot checks
in about 15000 files.
In initial set of patches against 4.14-rc6, 3 files were found to have
copy/paste license identifier errors, and have been fixed to reflect the
correct identifier.
Additionally Philippe spent 10 hours this week doing a detailed manual
inspection and review of the 12,461 patched files from the initial patch
version early this week with:
- a full scancode scan run, collecting the matched texts, detected
license ids and scores
- reviewing anything where there was a license detected (about 500+
files) to ensure that the applied SPDX license was correct
- reviewing anything where there was no detection but the patch license
was not GPL-2.0 WITH Linux-syscall-note to ensure that the applied
SPDX license was correct
This produced a worksheet with 20 files needing minor correction. This
worksheet was then exported into 3 different .csv files for the
different types of files to be modified.
These .csv files were then reviewed by Greg. Thomas wrote a script to
parse the csv files and add the proper SPDX tag to the file, in the
format that the file expected. This script was further refined by Greg
based on the output to detect more types of files automatically and to
distinguish between header and source .c files (which need different
comment types.) Finally Greg ran the script using the .csv files to
generate the patches.
Reviewed-by: Kate Stewart <kstewart@linuxfoundation.org>
Reviewed-by: Philippe Ombredanne <pombredanne@nexb.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2017-11-01 08:07:57 -06:00
|
|
|
// SPDX-License-Identifier: GPL-2.0
|
[PATCH] sparsemem memory model
Sparsemem abstracts the use of discontiguous mem_maps[]. This kind of
mem_map[] is needed by discontiguous memory machines (like in the old
CONFIG_DISCONTIGMEM case) as well as memory hotplug systems. Sparsemem
replaces DISCONTIGMEM when enabled, and it is hoped that it can eventually
become a complete replacement.
A significant advantage over DISCONTIGMEM is that it's completely separated
from CONFIG_NUMA. When producing this patch, it became apparent in that NUMA
and DISCONTIG are often confused.
Another advantage is that sparse doesn't require each NUMA node's ranges to be
contiguous. It can handle overlapping ranges between nodes with no problems,
where DISCONTIGMEM currently throws away that memory.
Sparsemem uses an array to provide different pfn_to_page() translations for
each SECTION_SIZE area of physical memory. This is what allows the mem_map[]
to be chopped up.
In order to do quick pfn_to_page() operations, the section number of the page
is encoded in page->flags. Part of the sparsemem infrastructure enables
sharing of these bits more dynamically (at compile-time) between the
page_zone() and sparsemem operations. However, on 32-bit architectures, the
number of bits is quite limited, and may require growing the size of the
page->flags type in certain conditions. Several things might force this to
occur: a decrease in the SECTION_SIZE (if you want to hotplug smaller areas of
memory), an increase in the physical address space, or an increase in the
number of used page->flags.
One thing to note is that, once sparsemem is present, the NUMA node
information no longer needs to be stored in the page->flags. It might provide
speed increases on certain platforms and will be stored there if there is
room. But, if out of room, an alternate (theoretically slower) mechanism is
used.
This patch introduces CONFIG_FLATMEM. It is used in almost all cases where
there used to be an #ifndef DISCONTIG, because SPARSEMEM and DISCONTIGMEM
often have to compile out the same areas of code.
Signed-off-by: Andy Whitcroft <apw@shadowen.org>
Signed-off-by: Dave Hansen <haveblue@us.ibm.com>
Signed-off-by: Martin Bligh <mbligh@aracnet.com>
Signed-off-by: Adrian Bunk <bunk@stusta.de>
Signed-off-by: Yasunori Goto <y-goto@jp.fujitsu.com>
Signed-off-by: Bob Picco <bob.picco@hp.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-06-23 01:07:54 -06:00
|
|
|
/*
|
|
|
|
* sparse memory mappings.
|
|
|
|
*/
|
|
|
|
#include <linux/mm.h>
|
include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h
percpu.h is included by sched.h and module.h and thus ends up being
included when building most .c files. percpu.h includes slab.h which
in turn includes gfp.h making everything defined by the two files
universally available and complicating inclusion dependencies.
percpu.h -> slab.h dependency is about to be removed. Prepare for
this change by updating users of gfp and slab facilities include those
headers directly instead of assuming availability. As this conversion
needs to touch large number of source files, the following script is
used as the basis of conversion.
http://userweb.kernel.org/~tj/misc/slabh-sweep.py
The script does the followings.
* Scan files for gfp and slab usages and update includes such that
only the necessary includes are there. ie. if only gfp is used,
gfp.h, if slab is used, slab.h.
* When the script inserts a new include, it looks at the include
blocks and try to put the new include such that its order conforms
to its surrounding. It's put in the include block which contains
core kernel includes, in the same order that the rest are ordered -
alphabetical, Christmas tree, rev-Xmas-tree or at the end if there
doesn't seem to be any matching order.
* If the script can't find a place to put a new include (mostly
because the file doesn't have fitting include block), it prints out
an error message indicating which .h file needs to be added to the
file.
The conversion was done in the following steps.
1. The initial automatic conversion of all .c files updated slightly
over 4000 files, deleting around 700 includes and adding ~480 gfp.h
and ~3000 slab.h inclusions. The script emitted errors for ~400
files.
2. Each error was manually checked. Some didn't need the inclusion,
some needed manual addition while adding it to implementation .h or
embedding .c file was more appropriate for others. This step added
inclusions to around 150 files.
3. The script was run again and the output was compared to the edits
from #2 to make sure no file was left behind.
4. Several build tests were done and a couple of problems were fixed.
e.g. lib/decompress_*.c used malloc/free() wrappers around slab
APIs requiring slab.h to be added manually.
5. The script was run on all .h files but without automatically
editing them as sprinkling gfp.h and slab.h inclusions around .h
files could easily lead to inclusion dependency hell. Most gfp.h
inclusion directives were ignored as stuff from gfp.h was usually
wildly available and often used in preprocessor macros. Each
slab.h inclusion directive was examined and added manually as
necessary.
6. percpu.h was updated not to include slab.h.
7. Build test were done on the following configurations and failures
were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my
distributed build env didn't work with gcov compiles) and a few
more options had to be turned off depending on archs to make things
build (like ipr on powerpc/64 which failed due to missing writeq).
* x86 and x86_64 UP and SMP allmodconfig and a custom test config.
* powerpc and powerpc64 SMP allmodconfig
* sparc and sparc64 SMP allmodconfig
* ia64 SMP allmodconfig
* s390 SMP allmodconfig
* alpha SMP allmodconfig
* um on x86_64 SMP allmodconfig
8. percpu.h modifications were reverted so that it could be applied as
a separate patch and serve as bisection point.
Given the fact that I had only a couple of failures from tests on step
6, I'm fairly confident about the coverage of this conversion patch.
If there is a breakage, it's likely to be something in one of the arch
headers which should be easily discoverable easily on most builds of
the specific arch.
Signed-off-by: Tejun Heo <tj@kernel.org>
Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
2010-03-24 02:04:11 -06:00
|
|
|
#include <linux/slab.h>
|
[PATCH] sparsemem memory model
Sparsemem abstracts the use of discontiguous mem_maps[]. This kind of
mem_map[] is needed by discontiguous memory machines (like in the old
CONFIG_DISCONTIGMEM case) as well as memory hotplug systems. Sparsemem
replaces DISCONTIGMEM when enabled, and it is hoped that it can eventually
become a complete replacement.
A significant advantage over DISCONTIGMEM is that it's completely separated
from CONFIG_NUMA. When producing this patch, it became apparent in that NUMA
and DISCONTIG are often confused.
Another advantage is that sparse doesn't require each NUMA node's ranges to be
contiguous. It can handle overlapping ranges between nodes with no problems,
where DISCONTIGMEM currently throws away that memory.
Sparsemem uses an array to provide different pfn_to_page() translations for
each SECTION_SIZE area of physical memory. This is what allows the mem_map[]
to be chopped up.
In order to do quick pfn_to_page() operations, the section number of the page
is encoded in page->flags. Part of the sparsemem infrastructure enables
sharing of these bits more dynamically (at compile-time) between the
page_zone() and sparsemem operations. However, on 32-bit architectures, the
number of bits is quite limited, and may require growing the size of the
page->flags type in certain conditions. Several things might force this to
occur: a decrease in the SECTION_SIZE (if you want to hotplug smaller areas of
memory), an increase in the physical address space, or an increase in the
number of used page->flags.
One thing to note is that, once sparsemem is present, the NUMA node
information no longer needs to be stored in the page->flags. It might provide
speed increases on certain platforms and will be stored there if there is
room. But, if out of room, an alternate (theoretically slower) mechanism is
used.
This patch introduces CONFIG_FLATMEM. It is used in almost all cases where
there used to be an #ifndef DISCONTIG, because SPARSEMEM and DISCONTIGMEM
often have to compile out the same areas of code.
Signed-off-by: Andy Whitcroft <apw@shadowen.org>
Signed-off-by: Dave Hansen <haveblue@us.ibm.com>
Signed-off-by: Martin Bligh <mbligh@aracnet.com>
Signed-off-by: Adrian Bunk <bunk@stusta.de>
Signed-off-by: Yasunori Goto <y-goto@jp.fujitsu.com>
Signed-off-by: Bob Picco <bob.picco@hp.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-06-23 01:07:54 -06:00
|
|
|
#include <linux/mmzone.h>
|
|
|
|
#include <linux/bootmem.h>
|
2014-04-07 16:37:26 -06:00
|
|
|
#include <linux/compiler.h>
|
2005-10-29 19:16:55 -06:00
|
|
|
#include <linux/highmem.h>
|
2011-10-16 00:01:52 -06:00
|
|
|
#include <linux/export.h>
|
2005-09-03 16:54:29 -06:00
|
|
|
#include <linux/spinlock.h>
|
2005-10-29 19:16:55 -06:00
|
|
|
#include <linux/vmalloc.h>
|
2014-04-07 16:37:26 -06:00
|
|
|
|
2008-04-28 03:13:34 -06:00
|
|
|
#include "internal.h"
|
[PATCH] sparsemem memory model
Sparsemem abstracts the use of discontiguous mem_maps[]. This kind of
mem_map[] is needed by discontiguous memory machines (like in the old
CONFIG_DISCONTIGMEM case) as well as memory hotplug systems. Sparsemem
replaces DISCONTIGMEM when enabled, and it is hoped that it can eventually
become a complete replacement.
A significant advantage over DISCONTIGMEM is that it's completely separated
from CONFIG_NUMA. When producing this patch, it became apparent in that NUMA
and DISCONTIG are often confused.
Another advantage is that sparse doesn't require each NUMA node's ranges to be
contiguous. It can handle overlapping ranges between nodes with no problems,
where DISCONTIGMEM currently throws away that memory.
Sparsemem uses an array to provide different pfn_to_page() translations for
each SECTION_SIZE area of physical memory. This is what allows the mem_map[]
to be chopped up.
In order to do quick pfn_to_page() operations, the section number of the page
is encoded in page->flags. Part of the sparsemem infrastructure enables
sharing of these bits more dynamically (at compile-time) between the
page_zone() and sparsemem operations. However, on 32-bit architectures, the
number of bits is quite limited, and may require growing the size of the
page->flags type in certain conditions. Several things might force this to
occur: a decrease in the SECTION_SIZE (if you want to hotplug smaller areas of
memory), an increase in the physical address space, or an increase in the
number of used page->flags.
One thing to note is that, once sparsemem is present, the NUMA node
information no longer needs to be stored in the page->flags. It might provide
speed increases on certain platforms and will be stored there if there is
room. But, if out of room, an alternate (theoretically slower) mechanism is
used.
This patch introduces CONFIG_FLATMEM. It is used in almost all cases where
there used to be an #ifndef DISCONTIG, because SPARSEMEM and DISCONTIGMEM
often have to compile out the same areas of code.
Signed-off-by: Andy Whitcroft <apw@shadowen.org>
Signed-off-by: Dave Hansen <haveblue@us.ibm.com>
Signed-off-by: Martin Bligh <mbligh@aracnet.com>
Signed-off-by: Adrian Bunk <bunk@stusta.de>
Signed-off-by: Yasunori Goto <y-goto@jp.fujitsu.com>
Signed-off-by: Bob Picco <bob.picco@hp.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-06-23 01:07:54 -06:00
|
|
|
#include <asm/dma.h>
|
2007-10-16 02:24:13 -06:00
|
|
|
#include <asm/pgalloc.h>
|
|
|
|
#include <asm/pgtable.h>
|
[PATCH] sparsemem memory model
Sparsemem abstracts the use of discontiguous mem_maps[]. This kind of
mem_map[] is needed by discontiguous memory machines (like in the old
CONFIG_DISCONTIGMEM case) as well as memory hotplug systems. Sparsemem
replaces DISCONTIGMEM when enabled, and it is hoped that it can eventually
become a complete replacement.
A significant advantage over DISCONTIGMEM is that it's completely separated
from CONFIG_NUMA. When producing this patch, it became apparent in that NUMA
and DISCONTIG are often confused.
Another advantage is that sparse doesn't require each NUMA node's ranges to be
contiguous. It can handle overlapping ranges between nodes with no problems,
where DISCONTIGMEM currently throws away that memory.
Sparsemem uses an array to provide different pfn_to_page() translations for
each SECTION_SIZE area of physical memory. This is what allows the mem_map[]
to be chopped up.
In order to do quick pfn_to_page() operations, the section number of the page
is encoded in page->flags. Part of the sparsemem infrastructure enables
sharing of these bits more dynamically (at compile-time) between the
page_zone() and sparsemem operations. However, on 32-bit architectures, the
number of bits is quite limited, and may require growing the size of the
page->flags type in certain conditions. Several things might force this to
occur: a decrease in the SECTION_SIZE (if you want to hotplug smaller areas of
memory), an increase in the physical address space, or an increase in the
number of used page->flags.
One thing to note is that, once sparsemem is present, the NUMA node
information no longer needs to be stored in the page->flags. It might provide
speed increases on certain platforms and will be stored there if there is
room. But, if out of room, an alternate (theoretically slower) mechanism is
used.
This patch introduces CONFIG_FLATMEM. It is used in almost all cases where
there used to be an #ifndef DISCONTIG, because SPARSEMEM and DISCONTIGMEM
often have to compile out the same areas of code.
Signed-off-by: Andy Whitcroft <apw@shadowen.org>
Signed-off-by: Dave Hansen <haveblue@us.ibm.com>
Signed-off-by: Martin Bligh <mbligh@aracnet.com>
Signed-off-by: Adrian Bunk <bunk@stusta.de>
Signed-off-by: Yasunori Goto <y-goto@jp.fujitsu.com>
Signed-off-by: Bob Picco <bob.picco@hp.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-06-23 01:07:54 -06:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Permanent SPARSEMEM data:
|
|
|
|
*
|
|
|
|
* 1) mem_section - memory sections, mem_map's for valid memory
|
|
|
|
*/
|
2005-09-03 16:54:28 -06:00
|
|
|
#ifdef CONFIG_SPARSEMEM_EXTREME
|
2017-09-29 08:08:16 -06:00
|
|
|
struct mem_section **mem_section;
|
2005-09-03 16:54:28 -06:00
|
|
|
#else
|
|
|
|
struct mem_section mem_section[NR_SECTION_ROOTS][SECTIONS_PER_ROOT]
|
2006-01-08 02:01:27 -07:00
|
|
|
____cacheline_internodealigned_in_smp;
|
2005-09-03 16:54:28 -06:00
|
|
|
#endif
|
|
|
|
EXPORT_SYMBOL(mem_section);
|
|
|
|
|
2006-12-06 21:31:45 -07:00
|
|
|
#ifdef NODE_NOT_IN_PAGE_FLAGS
|
|
|
|
/*
|
|
|
|
* If we did not store the node number in the page then we have to
|
|
|
|
* do a lookup in the section_to_node_table in order to find which
|
|
|
|
* node the page belongs to.
|
|
|
|
*/
|
|
|
|
#if MAX_NUMNODES <= 256
|
|
|
|
static u8 section_to_node_table[NR_MEM_SECTIONS] __cacheline_aligned;
|
|
|
|
#else
|
|
|
|
static u16 section_to_node_table[NR_MEM_SECTIONS] __cacheline_aligned;
|
|
|
|
#endif
|
|
|
|
|
2011-07-25 18:11:51 -06:00
|
|
|
int page_to_nid(const struct page *page)
|
2006-12-06 21:31:45 -07:00
|
|
|
{
|
|
|
|
return section_to_node_table[page_to_section(page)];
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL(page_to_nid);
|
2007-08-22 15:01:03 -06:00
|
|
|
|
|
|
|
static void set_section_nid(unsigned long section_nr, int nid)
|
|
|
|
{
|
|
|
|
section_to_node_table[section_nr] = nid;
|
|
|
|
}
|
|
|
|
#else /* !NODE_NOT_IN_PAGE_FLAGS */
|
|
|
|
static inline void set_section_nid(unsigned long section_nr, int nid)
|
|
|
|
{
|
|
|
|
}
|
2006-12-06 21:31:45 -07:00
|
|
|
#endif
|
|
|
|
|
2005-09-03 16:54:28 -06:00
|
|
|
#ifdef CONFIG_SPARSEMEM_EXTREME
|
2016-08-02 15:03:33 -06:00
|
|
|
static noinline struct mem_section __ref *sparse_index_alloc(int nid)
|
2005-09-03 16:54:29 -06:00
|
|
|
{
|
|
|
|
struct mem_section *section = NULL;
|
|
|
|
unsigned long array_size = SECTIONS_PER_ROOT *
|
|
|
|
sizeof(struct mem_section);
|
|
|
|
|
2017-09-06 17:20:41 -06:00
|
|
|
if (slab_is_available())
|
|
|
|
section = kzalloc_node(array_size, GFP_KERNEL, nid);
|
|
|
|
else
|
2014-01-21 16:50:34 -07:00
|
|
|
section = memblock_virt_alloc_node(array_size, nid);
|
2005-09-03 16:54:29 -06:00
|
|
|
|
|
|
|
return section;
|
2005-09-03 16:54:28 -06:00
|
|
|
}
|
2005-09-03 16:54:26 -06:00
|
|
|
|
2007-05-08 01:23:07 -06:00
|
|
|
static int __meminit sparse_index_init(unsigned long section_nr, int nid)
|
2005-09-03 16:54:26 -06:00
|
|
|
{
|
2005-09-03 16:54:29 -06:00
|
|
|
unsigned long root = SECTION_NR_TO_ROOT(section_nr);
|
|
|
|
struct mem_section *section;
|
2005-09-03 16:54:26 -06:00
|
|
|
|
|
|
|
if (mem_section[root])
|
2005-09-03 16:54:29 -06:00
|
|
|
return -EEXIST;
|
2005-09-03 16:54:28 -06:00
|
|
|
|
2005-09-03 16:54:29 -06:00
|
|
|
section = sparse_index_alloc(nid);
|
2007-12-17 17:19:58 -07:00
|
|
|
if (!section)
|
|
|
|
return -ENOMEM;
|
2005-09-03 16:54:29 -06:00
|
|
|
|
|
|
|
mem_section[root] = section;
|
2012-07-31 17:46:06 -06:00
|
|
|
|
2013-05-17 08:10:38 -06:00
|
|
|
return 0;
|
2005-09-03 16:54:29 -06:00
|
|
|
}
|
|
|
|
#else /* !SPARSEMEM_EXTREME */
|
|
|
|
static inline int sparse_index_init(unsigned long section_nr, int nid)
|
|
|
|
{
|
|
|
|
return 0;
|
2005-09-03 16:54:26 -06:00
|
|
|
}
|
2005-09-03 16:54:29 -06:00
|
|
|
#endif
|
|
|
|
|
2016-07-28 16:48:35 -06:00
|
|
|
#ifdef CONFIG_SPARSEMEM_EXTREME
|
2005-10-29 19:16:51 -06:00
|
|
|
int __section_nr(struct mem_section* ms)
|
|
|
|
{
|
|
|
|
unsigned long root_nr;
|
2017-09-29 08:08:16 -06:00
|
|
|
struct mem_section *root = NULL;
|
2005-10-29 19:16:51 -06:00
|
|
|
|
2006-05-20 16:00:05 -06:00
|
|
|
for (root_nr = 0; root_nr < NR_SECTION_ROOTS; root_nr++) {
|
|
|
|
root = __nr_to_section(root_nr * SECTIONS_PER_ROOT);
|
2005-10-29 19:16:51 -06:00
|
|
|
if (!root)
|
|
|
|
continue;
|
|
|
|
|
|
|
|
if ((ms >= root) && (ms < (root + SECTIONS_PER_ROOT)))
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
2017-09-29 08:08:16 -06:00
|
|
|
VM_BUG_ON(!root);
|
2012-07-31 17:46:04 -06:00
|
|
|
|
2005-10-29 19:16:51 -06:00
|
|
|
return (root_nr * SECTIONS_PER_ROOT) + (ms - root);
|
|
|
|
}
|
2016-07-28 16:48:35 -06:00
|
|
|
#else
|
|
|
|
int __section_nr(struct mem_section* ms)
|
|
|
|
{
|
|
|
|
return (int)(ms - mem_section[0]);
|
|
|
|
}
|
|
|
|
#endif
|
2005-10-29 19:16:51 -06:00
|
|
|
|
2006-06-23 03:03:41 -06:00
|
|
|
/*
|
|
|
|
* During early boot, before section_mem_map is used for an actual
|
|
|
|
* mem_map, we use section_mem_map to store the section's NUMA
|
|
|
|
* node. This keeps us from having to use another data structure. The
|
|
|
|
* node information is cleared just before we store the real mem_map.
|
|
|
|
*/
|
|
|
|
static inline unsigned long sparse_encode_early_nid(int nid)
|
|
|
|
{
|
|
|
|
return (nid << SECTION_NID_SHIFT);
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline int sparse_early_nid(struct mem_section *section)
|
|
|
|
{
|
|
|
|
return (section->section_mem_map >> SECTION_NID_SHIFT);
|
|
|
|
}
|
|
|
|
|
2008-07-23 22:26:52 -06:00
|
|
|
/* Validate the physical addressing limitations of the model */
|
|
|
|
void __meminit mminit_validate_memmodel_limits(unsigned long *start_pfn,
|
|
|
|
unsigned long *end_pfn)
|
[PATCH] sparsemem memory model
Sparsemem abstracts the use of discontiguous mem_maps[]. This kind of
mem_map[] is needed by discontiguous memory machines (like in the old
CONFIG_DISCONTIGMEM case) as well as memory hotplug systems. Sparsemem
replaces DISCONTIGMEM when enabled, and it is hoped that it can eventually
become a complete replacement.
A significant advantage over DISCONTIGMEM is that it's completely separated
from CONFIG_NUMA. When producing this patch, it became apparent in that NUMA
and DISCONTIG are often confused.
Another advantage is that sparse doesn't require each NUMA node's ranges to be
contiguous. It can handle overlapping ranges between nodes with no problems,
where DISCONTIGMEM currently throws away that memory.
Sparsemem uses an array to provide different pfn_to_page() translations for
each SECTION_SIZE area of physical memory. This is what allows the mem_map[]
to be chopped up.
In order to do quick pfn_to_page() operations, the section number of the page
is encoded in page->flags. Part of the sparsemem infrastructure enables
sharing of these bits more dynamically (at compile-time) between the
page_zone() and sparsemem operations. However, on 32-bit architectures, the
number of bits is quite limited, and may require growing the size of the
page->flags type in certain conditions. Several things might force this to
occur: a decrease in the SECTION_SIZE (if you want to hotplug smaller areas of
memory), an increase in the physical address space, or an increase in the
number of used page->flags.
One thing to note is that, once sparsemem is present, the NUMA node
information no longer needs to be stored in the page->flags. It might provide
speed increases on certain platforms and will be stored there if there is
room. But, if out of room, an alternate (theoretically slower) mechanism is
used.
This patch introduces CONFIG_FLATMEM. It is used in almost all cases where
there used to be an #ifndef DISCONTIG, because SPARSEMEM and DISCONTIGMEM
often have to compile out the same areas of code.
Signed-off-by: Andy Whitcroft <apw@shadowen.org>
Signed-off-by: Dave Hansen <haveblue@us.ibm.com>
Signed-off-by: Martin Bligh <mbligh@aracnet.com>
Signed-off-by: Adrian Bunk <bunk@stusta.de>
Signed-off-by: Yasunori Goto <y-goto@jp.fujitsu.com>
Signed-off-by: Bob Picco <bob.picco@hp.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-06-23 01:07:54 -06:00
|
|
|
{
|
2008-07-23 22:26:52 -06:00
|
|
|
unsigned long max_sparsemem_pfn = 1UL << (MAX_PHYSMEM_BITS-PAGE_SHIFT);
|
[PATCH] sparsemem memory model
Sparsemem abstracts the use of discontiguous mem_maps[]. This kind of
mem_map[] is needed by discontiguous memory machines (like in the old
CONFIG_DISCONTIGMEM case) as well as memory hotplug systems. Sparsemem
replaces DISCONTIGMEM when enabled, and it is hoped that it can eventually
become a complete replacement.
A significant advantage over DISCONTIGMEM is that it's completely separated
from CONFIG_NUMA. When producing this patch, it became apparent in that NUMA
and DISCONTIG are often confused.
Another advantage is that sparse doesn't require each NUMA node's ranges to be
contiguous. It can handle overlapping ranges between nodes with no problems,
where DISCONTIGMEM currently throws away that memory.
Sparsemem uses an array to provide different pfn_to_page() translations for
each SECTION_SIZE area of physical memory. This is what allows the mem_map[]
to be chopped up.
In order to do quick pfn_to_page() operations, the section number of the page
is encoded in page->flags. Part of the sparsemem infrastructure enables
sharing of these bits more dynamically (at compile-time) between the
page_zone() and sparsemem operations. However, on 32-bit architectures, the
number of bits is quite limited, and may require growing the size of the
page->flags type in certain conditions. Several things might force this to
occur: a decrease in the SECTION_SIZE (if you want to hotplug smaller areas of
memory), an increase in the physical address space, or an increase in the
number of used page->flags.
One thing to note is that, once sparsemem is present, the NUMA node
information no longer needs to be stored in the page->flags. It might provide
speed increases on certain platforms and will be stored there if there is
room. But, if out of room, an alternate (theoretically slower) mechanism is
used.
This patch introduces CONFIG_FLATMEM. It is used in almost all cases where
there used to be an #ifndef DISCONTIG, because SPARSEMEM and DISCONTIGMEM
often have to compile out the same areas of code.
Signed-off-by: Andy Whitcroft <apw@shadowen.org>
Signed-off-by: Dave Hansen <haveblue@us.ibm.com>
Signed-off-by: Martin Bligh <mbligh@aracnet.com>
Signed-off-by: Adrian Bunk <bunk@stusta.de>
Signed-off-by: Yasunori Goto <y-goto@jp.fujitsu.com>
Signed-off-by: Bob Picco <bob.picco@hp.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-06-23 01:07:54 -06:00
|
|
|
|
2008-04-15 17:40:00 -06:00
|
|
|
/*
|
|
|
|
* Sanity checks - do not allow an architecture to pass
|
|
|
|
* in larger pfns than the maximum scope of sparsemem:
|
|
|
|
*/
|
2008-07-23 22:26:52 -06:00
|
|
|
if (*start_pfn > max_sparsemem_pfn) {
|
|
|
|
mminit_dprintk(MMINIT_WARNING, "pfnvalidation",
|
|
|
|
"Start of range %lu -> %lu exceeds SPARSEMEM max %lu\n",
|
|
|
|
*start_pfn, *end_pfn, max_sparsemem_pfn);
|
|
|
|
WARN_ON_ONCE(1);
|
|
|
|
*start_pfn = max_sparsemem_pfn;
|
|
|
|
*end_pfn = max_sparsemem_pfn;
|
2009-03-31 16:19:25 -06:00
|
|
|
} else if (*end_pfn > max_sparsemem_pfn) {
|
2008-07-23 22:26:52 -06:00
|
|
|
mminit_dprintk(MMINIT_WARNING, "pfnvalidation",
|
|
|
|
"End of range %lu -> %lu exceeds SPARSEMEM max %lu\n",
|
|
|
|
*start_pfn, *end_pfn, max_sparsemem_pfn);
|
|
|
|
WARN_ON_ONCE(1);
|
|
|
|
*end_pfn = max_sparsemem_pfn;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
mm, sparsemem: break out of loops early
There are a number of times that we loop over NR_MEM_SECTIONS, looking
for section_present() on each section. But, when we have very large
physical address spaces (large MAX_PHYSMEM_BITS), NR_MEM_SECTIONS
becomes very large, making the loops quite long.
With MAX_PHYSMEM_BITS=46 and a section size of 128MB, the current loops
are 512k iterations, which we barely notice on modern hardware. But,
raising MAX_PHYSMEM_BITS higher (like we will see on systems that
support 5-level paging) makes this 64x longer and we start to notice,
especially on slower systems like simulators. A 10-second delay for
512k iterations is annoying. But, a 640- second delay is crippling.
This does not help if we have extremely sparse physical address spaces,
but those are quite rare. We expect that most of the "slow" systems
where this matters will also be quite small and non-sparse.
To fix this, we track the highest section we've ever encountered. This
lets us know when we will *never* see another section_present(), and
lets us break out of the loops earlier.
Doing the whole for_each_present_section_nr() macro is probably
overkill, but it will ensure that any future loop iterations that we
grow are more likely to be correct.
Kirrill said "It shaved almost 40 seconds from boot time in qemu with
5-level paging enabled for me".
Link: http://lkml.kernel.org/r/20170504174434.C45A4735@viggo.jf.intel.com
Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com>
Tested-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2017-07-06 16:36:44 -06:00
|
|
|
/*
|
|
|
|
* There are a number of times that we loop over NR_MEM_SECTIONS,
|
|
|
|
* looking for section_present() on each. But, when we have very
|
|
|
|
* large physical address spaces, NR_MEM_SECTIONS can also be
|
|
|
|
* very large which makes the loops quite long.
|
|
|
|
*
|
|
|
|
* Keeping track of this gives us an easy way to break out of
|
|
|
|
* those loops early.
|
|
|
|
*/
|
|
|
|
int __highest_present_section_nr;
|
|
|
|
static void section_mark_present(struct mem_section *ms)
|
|
|
|
{
|
|
|
|
int section_nr = __section_nr(ms);
|
|
|
|
|
|
|
|
if (section_nr > __highest_present_section_nr)
|
|
|
|
__highest_present_section_nr = section_nr;
|
|
|
|
|
|
|
|
ms->section_mem_map |= SECTION_MARKED_PRESENT;
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline int next_present_section_nr(int section_nr)
|
|
|
|
{
|
|
|
|
do {
|
|
|
|
section_nr++;
|
|
|
|
if (present_section_nr(section_nr))
|
|
|
|
return section_nr;
|
2018-06-07 18:06:39 -06:00
|
|
|
} while ((section_nr <= __highest_present_section_nr));
|
mm, sparsemem: break out of loops early
There are a number of times that we loop over NR_MEM_SECTIONS, looking
for section_present() on each section. But, when we have very large
physical address spaces (large MAX_PHYSMEM_BITS), NR_MEM_SECTIONS
becomes very large, making the loops quite long.
With MAX_PHYSMEM_BITS=46 and a section size of 128MB, the current loops
are 512k iterations, which we barely notice on modern hardware. But,
raising MAX_PHYSMEM_BITS higher (like we will see on systems that
support 5-level paging) makes this 64x longer and we start to notice,
especially on slower systems like simulators. A 10-second delay for
512k iterations is annoying. But, a 640- second delay is crippling.
This does not help if we have extremely sparse physical address spaces,
but those are quite rare. We expect that most of the "slow" systems
where this matters will also be quite small and non-sparse.
To fix this, we track the highest section we've ever encountered. This
lets us know when we will *never* see another section_present(), and
lets us break out of the loops earlier.
Doing the whole for_each_present_section_nr() macro is probably
overkill, but it will ensure that any future loop iterations that we
grow are more likely to be correct.
Kirrill said "It shaved almost 40 seconds from boot time in qemu with
5-level paging enabled for me".
Link: http://lkml.kernel.org/r/20170504174434.C45A4735@viggo.jf.intel.com
Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com>
Tested-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2017-07-06 16:36:44 -06:00
|
|
|
|
|
|
|
return -1;
|
|
|
|
}
|
|
|
|
#define for_each_present_section_nr(start, section_nr) \
|
|
|
|
for (section_nr = next_present_section_nr(start-1); \
|
|
|
|
((section_nr >= 0) && \
|
|
|
|
(section_nr <= __highest_present_section_nr)); \
|
|
|
|
section_nr = next_present_section_nr(section_nr))
|
|
|
|
|
2008-07-23 22:26:52 -06:00
|
|
|
/* Record a memory area against a node. */
|
|
|
|
void __init memory_present(int nid, unsigned long start, unsigned long end)
|
|
|
|
{
|
|
|
|
unsigned long pfn;
|
2008-04-15 17:40:00 -06:00
|
|
|
|
2017-11-07 01:33:37 -07:00
|
|
|
#ifdef CONFIG_SPARSEMEM_EXTREME
|
|
|
|
if (unlikely(!mem_section)) {
|
|
|
|
unsigned long size, align;
|
|
|
|
|
2018-01-04 17:18:06 -07:00
|
|
|
size = sizeof(struct mem_section*) * NR_SECTION_ROOTS;
|
2017-11-07 01:33:37 -07:00
|
|
|
align = 1 << (INTERNODE_CACHE_SHIFT);
|
|
|
|
mem_section = memblock_virt_alloc(size, align);
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
|
[PATCH] sparsemem memory model
Sparsemem abstracts the use of discontiguous mem_maps[]. This kind of
mem_map[] is needed by discontiguous memory machines (like in the old
CONFIG_DISCONTIGMEM case) as well as memory hotplug systems. Sparsemem
replaces DISCONTIGMEM when enabled, and it is hoped that it can eventually
become a complete replacement.
A significant advantage over DISCONTIGMEM is that it's completely separated
from CONFIG_NUMA. When producing this patch, it became apparent in that NUMA
and DISCONTIG are often confused.
Another advantage is that sparse doesn't require each NUMA node's ranges to be
contiguous. It can handle overlapping ranges between nodes with no problems,
where DISCONTIGMEM currently throws away that memory.
Sparsemem uses an array to provide different pfn_to_page() translations for
each SECTION_SIZE area of physical memory. This is what allows the mem_map[]
to be chopped up.
In order to do quick pfn_to_page() operations, the section number of the page
is encoded in page->flags. Part of the sparsemem infrastructure enables
sharing of these bits more dynamically (at compile-time) between the
page_zone() and sparsemem operations. However, on 32-bit architectures, the
number of bits is quite limited, and may require growing the size of the
page->flags type in certain conditions. Several things might force this to
occur: a decrease in the SECTION_SIZE (if you want to hotplug smaller areas of
memory), an increase in the physical address space, or an increase in the
number of used page->flags.
One thing to note is that, once sparsemem is present, the NUMA node
information no longer needs to be stored in the page->flags. It might provide
speed increases on certain platforms and will be stored there if there is
room. But, if out of room, an alternate (theoretically slower) mechanism is
used.
This patch introduces CONFIG_FLATMEM. It is used in almost all cases where
there used to be an #ifndef DISCONTIG, because SPARSEMEM and DISCONTIGMEM
often have to compile out the same areas of code.
Signed-off-by: Andy Whitcroft <apw@shadowen.org>
Signed-off-by: Dave Hansen <haveblue@us.ibm.com>
Signed-off-by: Martin Bligh <mbligh@aracnet.com>
Signed-off-by: Adrian Bunk <bunk@stusta.de>
Signed-off-by: Yasunori Goto <y-goto@jp.fujitsu.com>
Signed-off-by: Bob Picco <bob.picco@hp.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-06-23 01:07:54 -06:00
|
|
|
start &= PAGE_SECTION_MASK;
|
2008-07-23 22:26:52 -06:00
|
|
|
mminit_validate_memmodel_limits(&start, &end);
|
[PATCH] sparsemem memory model
Sparsemem abstracts the use of discontiguous mem_maps[]. This kind of
mem_map[] is needed by discontiguous memory machines (like in the old
CONFIG_DISCONTIGMEM case) as well as memory hotplug systems. Sparsemem
replaces DISCONTIGMEM when enabled, and it is hoped that it can eventually
become a complete replacement.
A significant advantage over DISCONTIGMEM is that it's completely separated
from CONFIG_NUMA. When producing this patch, it became apparent in that NUMA
and DISCONTIG are often confused.
Another advantage is that sparse doesn't require each NUMA node's ranges to be
contiguous. It can handle overlapping ranges between nodes with no problems,
where DISCONTIGMEM currently throws away that memory.
Sparsemem uses an array to provide different pfn_to_page() translations for
each SECTION_SIZE area of physical memory. This is what allows the mem_map[]
to be chopped up.
In order to do quick pfn_to_page() operations, the section number of the page
is encoded in page->flags. Part of the sparsemem infrastructure enables
sharing of these bits more dynamically (at compile-time) between the
page_zone() and sparsemem operations. However, on 32-bit architectures, the
number of bits is quite limited, and may require growing the size of the
page->flags type in certain conditions. Several things might force this to
occur: a decrease in the SECTION_SIZE (if you want to hotplug smaller areas of
memory), an increase in the physical address space, or an increase in the
number of used page->flags.
One thing to note is that, once sparsemem is present, the NUMA node
information no longer needs to be stored in the page->flags. It might provide
speed increases on certain platforms and will be stored there if there is
room. But, if out of room, an alternate (theoretically slower) mechanism is
used.
This patch introduces CONFIG_FLATMEM. It is used in almost all cases where
there used to be an #ifndef DISCONTIG, because SPARSEMEM and DISCONTIGMEM
often have to compile out the same areas of code.
Signed-off-by: Andy Whitcroft <apw@shadowen.org>
Signed-off-by: Dave Hansen <haveblue@us.ibm.com>
Signed-off-by: Martin Bligh <mbligh@aracnet.com>
Signed-off-by: Adrian Bunk <bunk@stusta.de>
Signed-off-by: Yasunori Goto <y-goto@jp.fujitsu.com>
Signed-off-by: Bob Picco <bob.picco@hp.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-06-23 01:07:54 -06:00
|
|
|
for (pfn = start; pfn < end; pfn += PAGES_PER_SECTION) {
|
|
|
|
unsigned long section = pfn_to_section_nr(pfn);
|
2005-09-03 16:54:26 -06:00
|
|
|
struct mem_section *ms;
|
|
|
|
|
|
|
|
sparse_index_init(section, nid);
|
2007-08-22 15:01:03 -06:00
|
|
|
set_section_nid(section, nid);
|
2005-09-03 16:54:26 -06:00
|
|
|
|
|
|
|
ms = __nr_to_section(section);
|
mm, sparsemem: break out of loops early
There are a number of times that we loop over NR_MEM_SECTIONS, looking
for section_present() on each section. But, when we have very large
physical address spaces (large MAX_PHYSMEM_BITS), NR_MEM_SECTIONS
becomes very large, making the loops quite long.
With MAX_PHYSMEM_BITS=46 and a section size of 128MB, the current loops
are 512k iterations, which we barely notice on modern hardware. But,
raising MAX_PHYSMEM_BITS higher (like we will see on systems that
support 5-level paging) makes this 64x longer and we start to notice,
especially on slower systems like simulators. A 10-second delay for
512k iterations is annoying. But, a 640- second delay is crippling.
This does not help if we have extremely sparse physical address spaces,
but those are quite rare. We expect that most of the "slow" systems
where this matters will also be quite small and non-sparse.
To fix this, we track the highest section we've ever encountered. This
lets us know when we will *never* see another section_present(), and
lets us break out of the loops earlier.
Doing the whole for_each_present_section_nr() macro is probably
overkill, but it will ensure that any future loop iterations that we
grow are more likely to be correct.
Kirrill said "It shaved almost 40 seconds from boot time in qemu with
5-level paging enabled for me".
Link: http://lkml.kernel.org/r/20170504174434.C45A4735@viggo.jf.intel.com
Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com>
Tested-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2017-07-06 16:36:44 -06:00
|
|
|
if (!ms->section_mem_map) {
|
2017-07-06 16:37:56 -06:00
|
|
|
ms->section_mem_map = sparse_encode_early_nid(nid) |
|
|
|
|
SECTION_IS_ONLINE;
|
mm, sparsemem: break out of loops early
There are a number of times that we loop over NR_MEM_SECTIONS, looking
for section_present() on each section. But, when we have very large
physical address spaces (large MAX_PHYSMEM_BITS), NR_MEM_SECTIONS
becomes very large, making the loops quite long.
With MAX_PHYSMEM_BITS=46 and a section size of 128MB, the current loops
are 512k iterations, which we barely notice on modern hardware. But,
raising MAX_PHYSMEM_BITS higher (like we will see on systems that
support 5-level paging) makes this 64x longer and we start to notice,
especially on slower systems like simulators. A 10-second delay for
512k iterations is annoying. But, a 640- second delay is crippling.
This does not help if we have extremely sparse physical address spaces,
but those are quite rare. We expect that most of the "slow" systems
where this matters will also be quite small and non-sparse.
To fix this, we track the highest section we've ever encountered. This
lets us know when we will *never* see another section_present(), and
lets us break out of the loops earlier.
Doing the whole for_each_present_section_nr() macro is probably
overkill, but it will ensure that any future loop iterations that we
grow are more likely to be correct.
Kirrill said "It shaved almost 40 seconds from boot time in qemu with
5-level paging enabled for me".
Link: http://lkml.kernel.org/r/20170504174434.C45A4735@viggo.jf.intel.com
Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com>
Tested-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2017-07-06 16:36:44 -06:00
|
|
|
section_mark_present(ms);
|
|
|
|
}
|
[PATCH] sparsemem memory model
Sparsemem abstracts the use of discontiguous mem_maps[]. This kind of
mem_map[] is needed by discontiguous memory machines (like in the old
CONFIG_DISCONTIGMEM case) as well as memory hotplug systems. Sparsemem
replaces DISCONTIGMEM when enabled, and it is hoped that it can eventually
become a complete replacement.
A significant advantage over DISCONTIGMEM is that it's completely separated
from CONFIG_NUMA. When producing this patch, it became apparent in that NUMA
and DISCONTIG are often confused.
Another advantage is that sparse doesn't require each NUMA node's ranges to be
contiguous. It can handle overlapping ranges between nodes with no problems,
where DISCONTIGMEM currently throws away that memory.
Sparsemem uses an array to provide different pfn_to_page() translations for
each SECTION_SIZE area of physical memory. This is what allows the mem_map[]
to be chopped up.
In order to do quick pfn_to_page() operations, the section number of the page
is encoded in page->flags. Part of the sparsemem infrastructure enables
sharing of these bits more dynamically (at compile-time) between the
page_zone() and sparsemem operations. However, on 32-bit architectures, the
number of bits is quite limited, and may require growing the size of the
page->flags type in certain conditions. Several things might force this to
occur: a decrease in the SECTION_SIZE (if you want to hotplug smaller areas of
memory), an increase in the physical address space, or an increase in the
number of used page->flags.
One thing to note is that, once sparsemem is present, the NUMA node
information no longer needs to be stored in the page->flags. It might provide
speed increases on certain platforms and will be stored there if there is
room. But, if out of room, an alternate (theoretically slower) mechanism is
used.
This patch introduces CONFIG_FLATMEM. It is used in almost all cases where
there used to be an #ifndef DISCONTIG, because SPARSEMEM and DISCONTIGMEM
often have to compile out the same areas of code.
Signed-off-by: Andy Whitcroft <apw@shadowen.org>
Signed-off-by: Dave Hansen <haveblue@us.ibm.com>
Signed-off-by: Martin Bligh <mbligh@aracnet.com>
Signed-off-by: Adrian Bunk <bunk@stusta.de>
Signed-off-by: Yasunori Goto <y-goto@jp.fujitsu.com>
Signed-off-by: Bob Picco <bob.picco@hp.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-06-23 01:07:54 -06:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2005-06-23 01:08:00 -06:00
|
|
|
/*
|
|
|
|
* Subtle, we encode the real pfn into the mem_map such that
|
|
|
|
* the identity pfn - section_mem_map will return the actual
|
|
|
|
* physical page frame number.
|
|
|
|
*/
|
|
|
|
static unsigned long sparse_encode_mem_map(struct page *mem_map, unsigned long pnum)
|
|
|
|
{
|
2018-01-31 17:20:26 -07:00
|
|
|
unsigned long coded_mem_map =
|
|
|
|
(unsigned long)(mem_map - (section_nr_to_pfn(pnum)));
|
|
|
|
BUILD_BUG_ON(SECTION_MAP_LAST_BIT > (1UL<<PFN_SECTION_SHIFT));
|
|
|
|
BUG_ON(coded_mem_map & ~SECTION_MAP_MASK);
|
|
|
|
return coded_mem_map;
|
2005-06-23 01:08:00 -06:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
2008-04-28 03:12:01 -06:00
|
|
|
* Decode mem_map from the coded memmap
|
2005-06-23 01:08:00 -06:00
|
|
|
*/
|
|
|
|
struct page *sparse_decode_mem_map(unsigned long coded_mem_map, unsigned long pnum)
|
|
|
|
{
|
2008-04-28 03:12:01 -06:00
|
|
|
/* mask off the extra low bits of information */
|
|
|
|
coded_mem_map &= SECTION_MAP_MASK;
|
2005-06-23 01:08:00 -06:00
|
|
|
return ((struct page *)coded_mem_map) + section_nr_to_pfn(pnum);
|
|
|
|
}
|
|
|
|
|
2007-05-08 01:23:07 -06:00
|
|
|
static int __meminit sparse_init_one_section(struct mem_section *ms,
|
2007-10-16 02:25:56 -06:00
|
|
|
unsigned long pnum, struct page *mem_map,
|
|
|
|
unsigned long *pageblock_bitmap)
|
2005-06-23 01:08:00 -06:00
|
|
|
{
|
2007-10-16 02:24:11 -06:00
|
|
|
if (!present_section(ms))
|
2005-06-23 01:08:00 -06:00
|
|
|
return -EINVAL;
|
|
|
|
|
2006-06-23 03:03:41 -06:00
|
|
|
ms->section_mem_map &= ~SECTION_MAP_MASK;
|
2007-10-16 02:24:11 -06:00
|
|
|
ms->section_mem_map |= sparse_encode_mem_map(mem_map, pnum) |
|
|
|
|
SECTION_HAS_MEM_MAP;
|
2007-10-16 02:25:56 -06:00
|
|
|
ms->pageblock_flags = pageblock_bitmap;
|
2005-06-23 01:08:00 -06:00
|
|
|
|
|
|
|
return 1;
|
|
|
|
}
|
|
|
|
|
memory hotplug: register section/node id to free
This patch set is to free pages which is allocated by bootmem for
memory-hotremove. Some structures of memory management are allocated by
bootmem. ex) memmap, etc.
To remove memory physically, some of them must be freed according to
circumstance. This patch set makes basis to free those pages, and free
memmaps.
Basic my idea is using remain members of struct page to remember information
of users of bootmem (section number or node id). When the section is
removing, kernel can confirm it. By this information, some issues can be
solved.
1) When the memmap of removing section is allocated on other
section by bootmem, it should/can be free.
2) When the memmap of removing section is allocated on the
same section, it shouldn't be freed. Because the section has to be
logical memory offlined already and all pages must be isolated against
page allocater. If it is freed, page allocator may use it which will
be removed physically soon.
3) When removing section has other section's memmap,
kernel will be able to show easily which section should be removed
before it for user. (Not implemented yet)
4) When the above case 2), the page isolation will be able to check and skip
memmap's page when logical memory offline (offline_pages()).
Current page isolation code fails in this case because this page is
just reserved page and it can't distinguish this pages can be
removed or not. But, it will be able to do by this patch.
(Not implemented yet.)
5) The node information like pgdat has similar issues. But, this
will be able to be solved too by this.
(Not implemented yet, but, remembering node id in the pages.)
Fortunately, current bootmem allocator just keeps PageReserved flags,
and doesn't use any other members of page struct. The users of
bootmem doesn't use them too.
This patch:
This is to register information which is node or section's id. Kernel can
distinguish which node/section uses the pages allcated by bootmem. This is
basis for hot-remove sections or nodes.
Signed-off-by: Yasunori Goto <y-goto@jp.fujitsu.com>
Cc: Badari Pulavarty <pbadari@us.ibm.com>
Cc: Yinghai Lu <yhlu.kernel@gmail.com>
Cc: Yasunori Goto <y-goto@jp.fujitsu.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-28 03:13:31 -06:00
|
|
|
unsigned long usemap_size(void)
|
2007-10-16 02:25:56 -06:00
|
|
|
{
|
2017-05-03 15:53:51 -06:00
|
|
|
return BITS_TO_LONGS(SECTION_BLOCKFLAGS_BITS) * sizeof(unsigned long);
|
2007-10-16 02:25:56 -06:00
|
|
|
}
|
|
|
|
|
|
|
|
#ifdef CONFIG_MEMORY_HOTPLUG
|
|
|
|
static unsigned long *__kmalloc_section_usemap(void)
|
|
|
|
{
|
|
|
|
return kmalloc(usemap_size(), GFP_KERNEL);
|
|
|
|
}
|
|
|
|
#endif /* CONFIG_MEMORY_HOTPLUG */
|
|
|
|
|
2008-07-23 22:28:15 -06:00
|
|
|
#ifdef CONFIG_MEMORY_HOTREMOVE
|
|
|
|
static unsigned long * __init
|
2010-02-10 02:20:21 -07:00
|
|
|
sparse_early_usemaps_alloc_pgdat_section(struct pglist_data *pgdat,
|
2012-05-29 16:06:36 -06:00
|
|
|
unsigned long size)
|
2008-07-23 22:28:15 -06:00
|
|
|
{
|
2012-07-11 15:02:53 -06:00
|
|
|
unsigned long goal, limit;
|
|
|
|
unsigned long *p;
|
|
|
|
int nid;
|
2008-07-23 22:28:15 -06:00
|
|
|
/*
|
|
|
|
* A page may contain usemaps for other sections preventing the
|
|
|
|
* page being freed and making a section unremovable while
|
2014-03-31 02:41:58 -06:00
|
|
|
* other sections referencing the usemap remain active. Similarly,
|
2008-07-23 22:28:15 -06:00
|
|
|
* a pgdat can prevent a section being removed. If section A
|
|
|
|
* contains a pgdat and section B contains the usemap, both
|
|
|
|
* sections become inter-dependent. This allocates usemaps
|
|
|
|
* from the same section as the pgdat where possible to avoid
|
|
|
|
* this problem.
|
|
|
|
*/
|
2012-07-11 15:02:51 -06:00
|
|
|
goal = __pa(pgdat) & (PAGE_SECTION_MASK << PAGE_SHIFT);
|
2012-07-11 15:02:53 -06:00
|
|
|
limit = goal + (1UL << PA_SECTION_SHIFT);
|
|
|
|
nid = early_pfn_to_nid(goal >> PAGE_SHIFT);
|
|
|
|
again:
|
2014-01-21 16:50:34 -07:00
|
|
|
p = memblock_virt_alloc_try_nid_nopanic(size,
|
|
|
|
SMP_CACHE_BYTES, goal, limit,
|
|
|
|
nid);
|
2012-07-11 15:02:53 -06:00
|
|
|
if (!p && limit) {
|
|
|
|
limit = 0;
|
|
|
|
goto again;
|
|
|
|
}
|
|
|
|
return p;
|
2008-07-23 22:28:15 -06:00
|
|
|
}
|
|
|
|
|
|
|
|
static void __init check_usemap_section_nr(int nid, unsigned long *usemap)
|
|
|
|
{
|
|
|
|
unsigned long usemap_snr, pgdat_snr;
|
2017-09-29 08:08:16 -06:00
|
|
|
static unsigned long old_usemap_snr;
|
|
|
|
static unsigned long old_pgdat_snr;
|
2008-07-23 22:28:15 -06:00
|
|
|
struct pglist_data *pgdat = NODE_DATA(nid);
|
|
|
|
int usemap_nid;
|
|
|
|
|
2017-09-29 08:08:16 -06:00
|
|
|
/* First call */
|
|
|
|
if (!old_usemap_snr) {
|
|
|
|
old_usemap_snr = NR_MEM_SECTIONS;
|
|
|
|
old_pgdat_snr = NR_MEM_SECTIONS;
|
|
|
|
}
|
|
|
|
|
2008-07-23 22:28:15 -06:00
|
|
|
usemap_snr = pfn_to_section_nr(__pa(usemap) >> PAGE_SHIFT);
|
|
|
|
pgdat_snr = pfn_to_section_nr(__pa(pgdat) >> PAGE_SHIFT);
|
|
|
|
if (usemap_snr == pgdat_snr)
|
|
|
|
return;
|
|
|
|
|
|
|
|
if (old_usemap_snr == usemap_snr && old_pgdat_snr == pgdat_snr)
|
|
|
|
/* skip redundant message */
|
|
|
|
return;
|
|
|
|
|
|
|
|
old_usemap_snr = usemap_snr;
|
|
|
|
old_pgdat_snr = pgdat_snr;
|
|
|
|
|
|
|
|
usemap_nid = sparse_early_nid(__nr_to_section(usemap_snr));
|
|
|
|
if (usemap_nid != nid) {
|
2016-03-17 15:19:50 -06:00
|
|
|
pr_info("node %d must be removed before remove section %ld\n",
|
|
|
|
nid, usemap_snr);
|
2008-07-23 22:28:15 -06:00
|
|
|
return;
|
|
|
|
}
|
|
|
|
/*
|
|
|
|
* There is a circular dependency.
|
|
|
|
* Some platforms allow un-removable section because they will just
|
|
|
|
* gather other removable sections for dynamic partitioning.
|
|
|
|
* Just notify un-removable section's number here.
|
|
|
|
*/
|
2016-03-17 15:19:50 -06:00
|
|
|
pr_info("Section %ld and %ld (node %d) have a circular dependency on usemap and pgdat allocations\n",
|
|
|
|
usemap_snr, pgdat_snr, nid);
|
2008-07-23 22:28:15 -06:00
|
|
|
}
|
|
|
|
#else
|
|
|
|
static unsigned long * __init
|
2010-02-10 02:20:21 -07:00
|
|
|
sparse_early_usemaps_alloc_pgdat_section(struct pglist_data *pgdat,
|
2012-05-29 16:06:36 -06:00
|
|
|
unsigned long size)
|
2008-07-23 22:28:15 -06:00
|
|
|
{
|
2014-01-21 16:50:34 -07:00
|
|
|
return memblock_virt_alloc_node_nopanic(size, pgdat->node_id);
|
2008-07-23 22:28:15 -06:00
|
|
|
}
|
|
|
|
|
|
|
|
static void __init check_usemap_section_nr(int nid, unsigned long *usemap)
|
|
|
|
{
|
|
|
|
}
|
|
|
|
#endif /* CONFIG_MEMORY_HOTREMOVE */
|
|
|
|
|
2013-09-11 15:22:38 -06:00
|
|
|
static void __init sparse_early_usemaps_alloc_node(void *data,
|
2010-02-10 02:20:21 -07:00
|
|
|
unsigned long pnum_begin,
|
|
|
|
unsigned long pnum_end,
|
|
|
|
unsigned long usemap_count, int nodeid)
|
2007-10-16 02:25:56 -06:00
|
|
|
{
|
2010-02-10 02:20:21 -07:00
|
|
|
void *usemap;
|
|
|
|
unsigned long pnum;
|
2013-09-11 15:22:38 -06:00
|
|
|
unsigned long **usemap_map = (unsigned long **)data;
|
2010-02-10 02:20:21 -07:00
|
|
|
int size = usemap_size();
|
2007-10-16 02:25:56 -06:00
|
|
|
|
2010-02-10 02:20:21 -07:00
|
|
|
usemap = sparse_early_usemaps_alloc_pgdat_section(NODE_DATA(nodeid),
|
2012-05-29 16:06:36 -06:00
|
|
|
size * usemap_count);
|
bootmem/sparsemem: remove limit constraint in alloc_bootmem_section
While testing AMS (Active Memory Sharing) / CMO (Cooperative Memory
Overcommit) on powerpc, we tripped the following:
kernel BUG at mm/bootmem.c:483!
cpu 0x0: Vector: 700 (Program Check) at [c000000000c03940]
pc: c000000000a62bd8: .alloc_bootmem_core+0x90/0x39c
lr: c000000000a64bcc: .sparse_early_usemaps_alloc_node+0x84/0x29c
sp: c000000000c03bc0
msr: 8000000000021032
current = 0xc000000000b0cce0
paca = 0xc000000001d80000
pid = 0, comm = swapper
kernel BUG at mm/bootmem.c:483!
enter ? for help
[c000000000c03c80] c000000000a64bcc
.sparse_early_usemaps_alloc_node+0x84/0x29c
[c000000000c03d50] c000000000a64f10 .sparse_init+0x12c/0x28c
[c000000000c03e20] c000000000a474f4 .setup_arch+0x20c/0x294
[c000000000c03ee0] c000000000a4079c .start_kernel+0xb4/0x460
[c000000000c03f90] c000000000009670 .start_here_common+0x1c/0x2c
This is
BUG_ON(limit && goal + size > limit);
and after some debugging, it seems that
goal = 0x7ffff000000
limit = 0x80000000000
and sparse_early_usemaps_alloc_node ->
sparse_early_usemaps_alloc_pgdat_section calls
return alloc_bootmem_section(usemap_size() * count, section_nr);
This is on a system with 8TB available via the AMS pool, and as a quirk
of AMS in firmware, all of that memory shows up in node 0. So, we end
up with an allocation that will fail the goal/limit constraints.
In theory, we could "fall-back" to alloc_bootmem_node() in
sparse_early_usemaps_alloc_node(), but since we actually have HOTREMOVE
defined, we'll BUG_ON() instead. A simple solution appears to be to
unconditionally remove the limit condition in alloc_bootmem_section,
meaning allocations are allowed to cross section boundaries (necessary
for systems of this size).
Johannes Weiner pointed out that if alloc_bootmem_section() no longer
guarantees section-locality, we need check_usemap_section_nr() to print
possible cross-dependencies between node descriptors and the usemaps
allocated through it. That makes the two loops in
sparse_early_usemaps_alloc_node() identical, so re-factor the code a
bit.
[akpm@linux-foundation.org: code simplification]
Signed-off-by: Nishanth Aravamudan <nacc@us.ibm.com>
Cc: Dave Hansen <haveblue@us.ibm.com>
Cc: Anton Blanchard <anton@au1.ibm.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Ben Herrenschmidt <benh@kernel.crashing.org>
Cc: Robert Jennings <rcj@linux.vnet.ibm.com>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: Mel Gorman <mgorman@suse.de>
Cc: <stable@vger.kernel.org> [3.3.1]
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-03-21 17:34:07 -06:00
|
|
|
if (!usemap) {
|
2016-03-17 15:19:50 -06:00
|
|
|
pr_warn("%s: allocation failed\n", __func__);
|
2012-05-29 16:06:36 -06:00
|
|
|
return;
|
2008-07-23 22:28:15 -06:00
|
|
|
}
|
|
|
|
|
bootmem/sparsemem: remove limit constraint in alloc_bootmem_section
While testing AMS (Active Memory Sharing) / CMO (Cooperative Memory
Overcommit) on powerpc, we tripped the following:
kernel BUG at mm/bootmem.c:483!
cpu 0x0: Vector: 700 (Program Check) at [c000000000c03940]
pc: c000000000a62bd8: .alloc_bootmem_core+0x90/0x39c
lr: c000000000a64bcc: .sparse_early_usemaps_alloc_node+0x84/0x29c
sp: c000000000c03bc0
msr: 8000000000021032
current = 0xc000000000b0cce0
paca = 0xc000000001d80000
pid = 0, comm = swapper
kernel BUG at mm/bootmem.c:483!
enter ? for help
[c000000000c03c80] c000000000a64bcc
.sparse_early_usemaps_alloc_node+0x84/0x29c
[c000000000c03d50] c000000000a64f10 .sparse_init+0x12c/0x28c
[c000000000c03e20] c000000000a474f4 .setup_arch+0x20c/0x294
[c000000000c03ee0] c000000000a4079c .start_kernel+0xb4/0x460
[c000000000c03f90] c000000000009670 .start_here_common+0x1c/0x2c
This is
BUG_ON(limit && goal + size > limit);
and after some debugging, it seems that
goal = 0x7ffff000000
limit = 0x80000000000
and sparse_early_usemaps_alloc_node ->
sparse_early_usemaps_alloc_pgdat_section calls
return alloc_bootmem_section(usemap_size() * count, section_nr);
This is on a system with 8TB available via the AMS pool, and as a quirk
of AMS in firmware, all of that memory shows up in node 0. So, we end
up with an allocation that will fail the goal/limit constraints.
In theory, we could "fall-back" to alloc_bootmem_node() in
sparse_early_usemaps_alloc_node(), but since we actually have HOTREMOVE
defined, we'll BUG_ON() instead. A simple solution appears to be to
unconditionally remove the limit condition in alloc_bootmem_section,
meaning allocations are allowed to cross section boundaries (necessary
for systems of this size).
Johannes Weiner pointed out that if alloc_bootmem_section() no longer
guarantees section-locality, we need check_usemap_section_nr() to print
possible cross-dependencies between node descriptors and the usemaps
allocated through it. That makes the two loops in
sparse_early_usemaps_alloc_node() identical, so re-factor the code a
bit.
[akpm@linux-foundation.org: code simplification]
Signed-off-by: Nishanth Aravamudan <nacc@us.ibm.com>
Cc: Dave Hansen <haveblue@us.ibm.com>
Cc: Anton Blanchard <anton@au1.ibm.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Ben Herrenschmidt <benh@kernel.crashing.org>
Cc: Robert Jennings <rcj@linux.vnet.ibm.com>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: Mel Gorman <mgorman@suse.de>
Cc: <stable@vger.kernel.org> [3.3.1]
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-03-21 17:34:07 -06:00
|
|
|
for (pnum = pnum_begin; pnum < pnum_end; pnum++) {
|
|
|
|
if (!present_section_nr(pnum))
|
|
|
|
continue;
|
|
|
|
usemap_map[pnum] = usemap;
|
|
|
|
usemap += size;
|
|
|
|
check_usemap_section_nr(nodeid, usemap_map[pnum]);
|
2010-02-10 02:20:21 -07:00
|
|
|
}
|
2007-10-16 02:25:56 -06:00
|
|
|
}
|
|
|
|
|
2007-10-16 02:24:13 -06:00
|
|
|
#ifndef CONFIG_SPARSEMEM_VMEMMAP
|
2017-12-29 00:53:54 -07:00
|
|
|
struct page __init *sparse_mem_map_populate(unsigned long pnum, int nid,
|
|
|
|
struct vmem_altmap *altmap)
|
2005-06-23 01:08:00 -06:00
|
|
|
{
|
|
|
|
struct page *map;
|
2010-05-24 15:31:57 -06:00
|
|
|
unsigned long size;
|
2005-06-23 01:08:00 -06:00
|
|
|
|
2010-05-24 15:31:57 -06:00
|
|
|
size = PAGE_ALIGN(sizeof(struct page) * PAGES_PER_SECTION);
|
2014-01-21 16:50:34 -07:00
|
|
|
map = memblock_virt_alloc_try_nid(size,
|
|
|
|
PAGE_SIZE, __pa(MAX_DMA_ADDRESS),
|
|
|
|
BOOTMEM_ALLOC_ACCESSIBLE, nid);
|
2007-10-16 02:24:13 -06:00
|
|
|
return map;
|
|
|
|
}
|
2010-02-10 02:20:22 -07:00
|
|
|
void __init sparse_mem_maps_populate_node(struct page **map_map,
|
|
|
|
unsigned long pnum_begin,
|
|
|
|
unsigned long pnum_end,
|
|
|
|
unsigned long map_count, int nodeid)
|
|
|
|
{
|
|
|
|
void *map;
|
|
|
|
unsigned long pnum;
|
|
|
|
unsigned long size = sizeof(struct page) * PAGES_PER_SECTION;
|
|
|
|
|
|
|
|
size = PAGE_ALIGN(size);
|
2017-11-15 18:36:44 -07:00
|
|
|
map = memblock_virt_alloc_try_nid_raw(size * map_count,
|
|
|
|
PAGE_SIZE, __pa(MAX_DMA_ADDRESS),
|
|
|
|
BOOTMEM_ALLOC_ACCESSIBLE, nodeid);
|
2010-02-10 02:20:22 -07:00
|
|
|
if (map) {
|
|
|
|
for (pnum = pnum_begin; pnum < pnum_end; pnum++) {
|
|
|
|
if (!present_section_nr(pnum))
|
|
|
|
continue;
|
|
|
|
map_map[pnum] = map;
|
|
|
|
map += size;
|
|
|
|
}
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* fallback */
|
|
|
|
for (pnum = pnum_begin; pnum < pnum_end; pnum++) {
|
|
|
|
struct mem_section *ms;
|
|
|
|
|
|
|
|
if (!present_section_nr(pnum))
|
|
|
|
continue;
|
2017-12-29 00:53:54 -07:00
|
|
|
map_map[pnum] = sparse_mem_map_populate(pnum, nodeid, NULL);
|
2010-02-10 02:20:22 -07:00
|
|
|
if (map_map[pnum])
|
|
|
|
continue;
|
|
|
|
ms = __nr_to_section(pnum);
|
2016-03-17 15:19:50 -06:00
|
|
|
pr_err("%s: sparsemem memory map backing failed some memory will not be available\n",
|
2016-03-17 15:19:47 -06:00
|
|
|
__func__);
|
2010-02-10 02:20:22 -07:00
|
|
|
ms->section_mem_map = 0;
|
|
|
|
}
|
|
|
|
}
|
2007-10-16 02:24:13 -06:00
|
|
|
#endif /* !CONFIG_SPARSEMEM_VMEMMAP */
|
|
|
|
|
2010-02-27 10:29:38 -07:00
|
|
|
#ifdef CONFIG_SPARSEMEM_ALLOC_MEM_MAP_TOGETHER
|
2013-09-11 15:22:38 -06:00
|
|
|
static void __init sparse_early_mem_maps_alloc_node(void *data,
|
2010-02-10 02:20:22 -07:00
|
|
|
unsigned long pnum_begin,
|
|
|
|
unsigned long pnum_end,
|
|
|
|
unsigned long map_count, int nodeid)
|
|
|
|
{
|
2013-09-11 15:22:38 -06:00
|
|
|
struct page **map_map = (struct page **)data;
|
2010-02-10 02:20:22 -07:00
|
|
|
sparse_mem_maps_populate_node(map_map, pnum_begin, pnum_end,
|
|
|
|
map_count, nodeid);
|
|
|
|
}
|
2010-02-27 10:29:38 -07:00
|
|
|
#else
|
2008-07-25 20:46:22 -06:00
|
|
|
static struct page __init *sparse_early_mem_map_alloc(unsigned long pnum)
|
2007-10-16 02:24:13 -06:00
|
|
|
{
|
|
|
|
struct page *map;
|
|
|
|
struct mem_section *ms = __nr_to_section(pnum);
|
|
|
|
int nid = sparse_early_nid(ms);
|
|
|
|
|
2017-12-29 00:53:54 -07:00
|
|
|
map = sparse_mem_map_populate(pnum, nid, NULL);
|
2005-06-23 01:08:00 -06:00
|
|
|
if (map)
|
|
|
|
return map;
|
|
|
|
|
2016-03-17 15:19:50 -06:00
|
|
|
pr_err("%s: sparsemem memory map backing failed some memory will not be available\n",
|
2016-03-17 15:19:47 -06:00
|
|
|
__func__);
|
2005-09-03 16:54:26 -06:00
|
|
|
ms->section_mem_map = 0;
|
2005-06-23 01:08:00 -06:00
|
|
|
return NULL;
|
|
|
|
}
|
2010-02-10 02:20:22 -07:00
|
|
|
#endif
|
2005-06-23 01:08:00 -06:00
|
|
|
|
2014-04-07 16:37:26 -06:00
|
|
|
void __weak __meminit vmemmap_populate_print_last(void)
|
2008-04-12 02:19:24 -06:00
|
|
|
{
|
|
|
|
}
|
2010-02-10 02:20:21 -07:00
|
|
|
|
2013-09-11 15:22:38 -06:00
|
|
|
/**
|
|
|
|
* alloc_usemap_and_memmap - memory alloction for pageblock flags and vmemmap
|
|
|
|
* @map: usemap_map for pageblock flags or mmap_map for vmemmap
|
|
|
|
*/
|
|
|
|
static void __init alloc_usemap_and_memmap(void (*alloc_func)
|
|
|
|
(void *, unsigned long, unsigned long,
|
|
|
|
unsigned long, int), void *data)
|
|
|
|
{
|
|
|
|
unsigned long pnum;
|
|
|
|
unsigned long map_count;
|
|
|
|
int nodeid_begin = 0;
|
|
|
|
unsigned long pnum_begin = 0;
|
|
|
|
|
mm, sparsemem: break out of loops early
There are a number of times that we loop over NR_MEM_SECTIONS, looking
for section_present() on each section. But, when we have very large
physical address spaces (large MAX_PHYSMEM_BITS), NR_MEM_SECTIONS
becomes very large, making the loops quite long.
With MAX_PHYSMEM_BITS=46 and a section size of 128MB, the current loops
are 512k iterations, which we barely notice on modern hardware. But,
raising MAX_PHYSMEM_BITS higher (like we will see on systems that
support 5-level paging) makes this 64x longer and we start to notice,
especially on slower systems like simulators. A 10-second delay for
512k iterations is annoying. But, a 640- second delay is crippling.
This does not help if we have extremely sparse physical address spaces,
but those are quite rare. We expect that most of the "slow" systems
where this matters will also be quite small and non-sparse.
To fix this, we track the highest section we've ever encountered. This
lets us know when we will *never* see another section_present(), and
lets us break out of the loops earlier.
Doing the whole for_each_present_section_nr() macro is probably
overkill, but it will ensure that any future loop iterations that we
grow are more likely to be correct.
Kirrill said "It shaved almost 40 seconds from boot time in qemu with
5-level paging enabled for me".
Link: http://lkml.kernel.org/r/20170504174434.C45A4735@viggo.jf.intel.com
Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com>
Tested-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2017-07-06 16:36:44 -06:00
|
|
|
for_each_present_section_nr(0, pnum) {
|
2013-09-11 15:22:38 -06:00
|
|
|
struct mem_section *ms;
|
|
|
|
|
|
|
|
ms = __nr_to_section(pnum);
|
|
|
|
nodeid_begin = sparse_early_nid(ms);
|
|
|
|
pnum_begin = pnum;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
map_count = 1;
|
mm, sparsemem: break out of loops early
There are a number of times that we loop over NR_MEM_SECTIONS, looking
for section_present() on each section. But, when we have very large
physical address spaces (large MAX_PHYSMEM_BITS), NR_MEM_SECTIONS
becomes very large, making the loops quite long.
With MAX_PHYSMEM_BITS=46 and a section size of 128MB, the current loops
are 512k iterations, which we barely notice on modern hardware. But,
raising MAX_PHYSMEM_BITS higher (like we will see on systems that
support 5-level paging) makes this 64x longer and we start to notice,
especially on slower systems like simulators. A 10-second delay for
512k iterations is annoying. But, a 640- second delay is crippling.
This does not help if we have extremely sparse physical address spaces,
but those are quite rare. We expect that most of the "slow" systems
where this matters will also be quite small and non-sparse.
To fix this, we track the highest section we've ever encountered. This
lets us know when we will *never* see another section_present(), and
lets us break out of the loops earlier.
Doing the whole for_each_present_section_nr() macro is probably
overkill, but it will ensure that any future loop iterations that we
grow are more likely to be correct.
Kirrill said "It shaved almost 40 seconds from boot time in qemu with
5-level paging enabled for me".
Link: http://lkml.kernel.org/r/20170504174434.C45A4735@viggo.jf.intel.com
Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com>
Tested-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2017-07-06 16:36:44 -06:00
|
|
|
for_each_present_section_nr(pnum_begin + 1, pnum) {
|
2013-09-11 15:22:38 -06:00
|
|
|
struct mem_section *ms;
|
|
|
|
int nodeid;
|
|
|
|
|
|
|
|
ms = __nr_to_section(pnum);
|
|
|
|
nodeid = sparse_early_nid(ms);
|
|
|
|
if (nodeid == nodeid_begin) {
|
|
|
|
map_count++;
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
/* ok, we need to take cake of from pnum_begin to pnum - 1*/
|
|
|
|
alloc_func(data, pnum_begin, pnum,
|
|
|
|
map_count, nodeid_begin);
|
|
|
|
/* new start, update count etc*/
|
|
|
|
nodeid_begin = nodeid;
|
|
|
|
pnum_begin = pnum;
|
|
|
|
map_count = 1;
|
|
|
|
}
|
|
|
|
/* ok, last chunk */
|
2018-06-07 18:06:43 -06:00
|
|
|
alloc_func(data, pnum_begin, __highest_present_section_nr+1,
|
2013-09-11 15:22:38 -06:00
|
|
|
map_count, nodeid_begin);
|
|
|
|
}
|
|
|
|
|
2007-06-08 14:46:51 -06:00
|
|
|
/*
|
|
|
|
* Allocate the accumulated non-linear sections, allocate a mem_map
|
|
|
|
* for each and record the physical to section mapping.
|
|
|
|
*/
|
|
|
|
void __init sparse_init(void)
|
|
|
|
{
|
|
|
|
unsigned long pnum;
|
|
|
|
struct page *map;
|
2007-10-16 02:25:56 -06:00
|
|
|
unsigned long *usemap;
|
2008-04-13 12:51:06 -06:00
|
|
|
unsigned long **usemap_map;
|
2010-02-27 10:29:38 -07:00
|
|
|
int size;
|
|
|
|
#ifdef CONFIG_SPARSEMEM_ALLOC_MEM_MAP_TOGETHER
|
|
|
|
int size2;
|
|
|
|
struct page **map_map;
|
|
|
|
#endif
|
2008-04-13 12:51:06 -06:00
|
|
|
|
2013-07-03 16:04:44 -06:00
|
|
|
/* see include/linux/mmzone.h 'struct mem_section' definition */
|
|
|
|
BUILD_BUG_ON(!is_power_of_2(sizeof(struct mem_section)));
|
|
|
|
|
2012-07-31 17:43:19 -06:00
|
|
|
/* Setup pageblock_order for HUGETLB_PAGE_SIZE_VARIABLE */
|
|
|
|
set_pageblock_order();
|
|
|
|
|
2008-04-13 12:51:06 -06:00
|
|
|
/*
|
|
|
|
* map is using big page (aka 2M in x86 64 bit)
|
|
|
|
* usemap is less one page (aka 24 bytes)
|
|
|
|
* so alloc 2M (with 2M align) and 24 bytes in turn will
|
|
|
|
* make next 2M slip to one more 2M later.
|
|
|
|
* then in big system, the memory will have a lot of holes...
|
2011-03-30 19:57:33 -06:00
|
|
|
* here try to allocate 2M pages continuously.
|
2008-04-13 12:51:06 -06:00
|
|
|
*
|
|
|
|
* powerpc need to call sparse_init_one_section right after each
|
|
|
|
* sparse_early_mem_map_alloc, so allocate usemap_map at first.
|
|
|
|
*/
|
|
|
|
size = sizeof(unsigned long *) * NR_MEM_SECTIONS;
|
2014-01-21 16:50:34 -07:00
|
|
|
usemap_map = memblock_virt_alloc(size, 0);
|
2008-04-13 12:51:06 -06:00
|
|
|
if (!usemap_map)
|
|
|
|
panic("can not allocate usemap_map\n");
|
2013-09-11 15:22:38 -06:00
|
|
|
alloc_usemap_and_memmap(sparse_early_usemaps_alloc_node,
|
|
|
|
(void *)usemap_map);
|
2007-06-08 14:46:51 -06:00
|
|
|
|
2010-02-10 02:20:22 -07:00
|
|
|
#ifdef CONFIG_SPARSEMEM_ALLOC_MEM_MAP_TOGETHER
|
|
|
|
size2 = sizeof(struct page *) * NR_MEM_SECTIONS;
|
2014-01-21 16:50:34 -07:00
|
|
|
map_map = memblock_virt_alloc(size2, 0);
|
2010-02-10 02:20:22 -07:00
|
|
|
if (!map_map)
|
|
|
|
panic("can not allocate map_map\n");
|
2013-09-11 15:22:38 -06:00
|
|
|
alloc_usemap_and_memmap(sparse_early_mem_maps_alloc_node,
|
|
|
|
(void *)map_map);
|
2010-02-10 02:20:22 -07:00
|
|
|
#endif
|
|
|
|
|
mm, sparsemem: break out of loops early
There are a number of times that we loop over NR_MEM_SECTIONS, looking
for section_present() on each section. But, when we have very large
physical address spaces (large MAX_PHYSMEM_BITS), NR_MEM_SECTIONS
becomes very large, making the loops quite long.
With MAX_PHYSMEM_BITS=46 and a section size of 128MB, the current loops
are 512k iterations, which we barely notice on modern hardware. But,
raising MAX_PHYSMEM_BITS higher (like we will see on systems that
support 5-level paging) makes this 64x longer and we start to notice,
especially on slower systems like simulators. A 10-second delay for
512k iterations is annoying. But, a 640- second delay is crippling.
This does not help if we have extremely sparse physical address spaces,
but those are quite rare. We expect that most of the "slow" systems
where this matters will also be quite small and non-sparse.
To fix this, we track the highest section we've ever encountered. This
lets us know when we will *never* see another section_present(), and
lets us break out of the loops earlier.
Doing the whole for_each_present_section_nr() macro is probably
overkill, but it will ensure that any future loop iterations that we
grow are more likely to be correct.
Kirrill said "It shaved almost 40 seconds from boot time in qemu with
5-level paging enabled for me".
Link: http://lkml.kernel.org/r/20170504174434.C45A4735@viggo.jf.intel.com
Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com>
Tested-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2017-07-06 16:36:44 -06:00
|
|
|
for_each_present_section_nr(0, pnum) {
|
2008-04-13 12:51:06 -06:00
|
|
|
usemap = usemap_map[pnum];
|
2007-10-16 02:25:56 -06:00
|
|
|
if (!usemap)
|
|
|
|
continue;
|
|
|
|
|
2010-02-10 02:20:22 -07:00
|
|
|
#ifdef CONFIG_SPARSEMEM_ALLOC_MEM_MAP_TOGETHER
|
|
|
|
map = map_map[pnum];
|
|
|
|
#else
|
2008-04-13 12:51:06 -06:00
|
|
|
map = sparse_early_mem_map_alloc(pnum);
|
2010-02-10 02:20:22 -07:00
|
|
|
#endif
|
2008-04-13 12:51:06 -06:00
|
|
|
if (!map)
|
|
|
|
continue;
|
|
|
|
|
2007-10-16 02:25:56 -06:00
|
|
|
sparse_init_one_section(__nr_to_section(pnum), pnum, map,
|
|
|
|
usemap);
|
2007-06-08 14:46:51 -06:00
|
|
|
}
|
2008-04-13 12:51:06 -06:00
|
|
|
|
2008-04-12 02:19:24 -06:00
|
|
|
vmemmap_populate_print_last();
|
|
|
|
|
2010-02-10 02:20:22 -07:00
|
|
|
#ifdef CONFIG_SPARSEMEM_ALLOC_MEM_MAP_TOGETHER
|
2014-01-21 16:50:34 -07:00
|
|
|
memblock_free_early(__pa(map_map), size2);
|
2010-02-10 02:20:22 -07:00
|
|
|
#endif
|
2014-01-21 16:50:34 -07:00
|
|
|
memblock_free_early(__pa(usemap_map), size);
|
2007-06-08 14:46:51 -06:00
|
|
|
}
|
|
|
|
|
|
|
|
#ifdef CONFIG_MEMORY_HOTPLUG
|
2017-07-06 16:37:56 -06:00
|
|
|
|
|
|
|
/* Mark all memory sections within the pfn range as online */
|
|
|
|
void online_mem_sections(unsigned long start_pfn, unsigned long end_pfn)
|
|
|
|
{
|
|
|
|
unsigned long pfn;
|
|
|
|
|
|
|
|
for (pfn = start_pfn; pfn < end_pfn; pfn += PAGES_PER_SECTION) {
|
2017-09-08 17:13:15 -06:00
|
|
|
unsigned long section_nr = pfn_to_section_nr(pfn);
|
2017-07-06 16:37:56 -06:00
|
|
|
struct mem_section *ms;
|
|
|
|
|
|
|
|
/* onlining code should never touch invalid ranges */
|
|
|
|
if (WARN_ON(!valid_section_nr(section_nr)))
|
|
|
|
continue;
|
|
|
|
|
|
|
|
ms = __nr_to_section(section_nr);
|
|
|
|
ms->section_mem_map |= SECTION_IS_ONLINE;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
#ifdef CONFIG_MEMORY_HOTREMOVE
|
|
|
|
/* Mark all memory sections within the pfn range as online */
|
|
|
|
void offline_mem_sections(unsigned long start_pfn, unsigned long end_pfn)
|
|
|
|
{
|
|
|
|
unsigned long pfn;
|
|
|
|
|
|
|
|
for (pfn = start_pfn; pfn < end_pfn; pfn += PAGES_PER_SECTION) {
|
2018-05-11 17:01:50 -06:00
|
|
|
unsigned long section_nr = pfn_to_section_nr(pfn);
|
2017-07-06 16:37:56 -06:00
|
|
|
struct mem_section *ms;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* TODO this needs some double checking. Offlining code makes
|
|
|
|
* sure to check pfn_valid but those checks might be just bogus
|
|
|
|
*/
|
|
|
|
if (WARN_ON(!valid_section_nr(section_nr)))
|
|
|
|
continue;
|
|
|
|
|
|
|
|
ms = __nr_to_section(section_nr);
|
|
|
|
ms->section_mem_map &= ~SECTION_IS_ONLINE;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
|
2007-10-16 02:26:14 -06:00
|
|
|
#ifdef CONFIG_SPARSEMEM_VMEMMAP
|
2017-12-29 00:53:54 -07:00
|
|
|
static inline struct page *kmalloc_section_memmap(unsigned long pnum, int nid,
|
|
|
|
struct vmem_altmap *altmap)
|
2007-10-16 02:26:14 -06:00
|
|
|
{
|
|
|
|
/* This will make the necessary allocations eventually. */
|
2017-12-29 00:53:54 -07:00
|
|
|
return sparse_mem_map_populate(pnum, nid, altmap);
|
2007-10-16 02:26:14 -06:00
|
|
|
}
|
2017-12-29 00:53:56 -07:00
|
|
|
static void __kfree_section_memmap(struct page *memmap,
|
|
|
|
struct vmem_altmap *altmap)
|
2007-10-16 02:26:14 -06:00
|
|
|
{
|
2013-04-29 16:07:50 -06:00
|
|
|
unsigned long start = (unsigned long)memmap;
|
2013-11-12 16:07:42 -07:00
|
|
|
unsigned long end = (unsigned long)(memmap + PAGES_PER_SECTION);
|
2013-04-29 16:07:50 -06:00
|
|
|
|
2017-12-29 00:53:56 -07:00
|
|
|
vmemmap_free(start, end, altmap);
|
2007-10-16 02:26:14 -06:00
|
|
|
}
|
2013-04-29 16:08:22 -06:00
|
|
|
#ifdef CONFIG_MEMORY_HOTREMOVE
|
2013-11-12 16:07:43 -07:00
|
|
|
static void free_map_bootmem(struct page *memmap)
|
2008-04-28 03:13:34 -06:00
|
|
|
{
|
2013-04-29 16:07:50 -06:00
|
|
|
unsigned long start = (unsigned long)memmap;
|
2013-11-12 16:07:43 -07:00
|
|
|
unsigned long end = (unsigned long)(memmap + PAGES_PER_SECTION);
|
2013-04-29 16:07:50 -06:00
|
|
|
|
2017-12-29 00:53:56 -07:00
|
|
|
vmemmap_free(start, end, NULL);
|
2008-04-28 03:13:34 -06:00
|
|
|
}
|
2013-04-29 16:08:22 -06:00
|
|
|
#endif /* CONFIG_MEMORY_HOTREMOVE */
|
2007-10-16 02:26:14 -06:00
|
|
|
#else
|
2013-11-12 16:07:42 -07:00
|
|
|
static struct page *__kmalloc_section_memmap(void)
|
2005-10-29 19:16:55 -06:00
|
|
|
{
|
|
|
|
struct page *page, *ret;
|
2013-11-12 16:07:42 -07:00
|
|
|
unsigned long memmap_size = sizeof(struct page) * PAGES_PER_SECTION;
|
2005-10-29 19:16:55 -06:00
|
|
|
|
2006-10-28 11:38:32 -06:00
|
|
|
page = alloc_pages(GFP_KERNEL|__GFP_NOWARN, get_order(memmap_size));
|
2005-10-29 19:16:55 -06:00
|
|
|
if (page)
|
|
|
|
goto got_map_page;
|
|
|
|
|
|
|
|
ret = vmalloc(memmap_size);
|
|
|
|
if (ret)
|
|
|
|
goto got_map_ptr;
|
|
|
|
|
|
|
|
return NULL;
|
|
|
|
got_map_page:
|
|
|
|
ret = (struct page *)pfn_to_kaddr(page_to_pfn(page));
|
|
|
|
got_map_ptr:
|
|
|
|
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2017-12-29 00:53:54 -07:00
|
|
|
static inline struct page *kmalloc_section_memmap(unsigned long pnum, int nid,
|
|
|
|
struct vmem_altmap *altmap)
|
2007-10-16 02:26:14 -06:00
|
|
|
{
|
2013-11-12 16:07:42 -07:00
|
|
|
return __kmalloc_section_memmap();
|
2007-10-16 02:26:14 -06:00
|
|
|
}
|
|
|
|
|
2017-12-29 00:53:56 -07:00
|
|
|
static void __kfree_section_memmap(struct page *memmap,
|
|
|
|
struct vmem_altmap *altmap)
|
2005-10-29 19:16:55 -06:00
|
|
|
{
|
2008-02-04 23:28:34 -07:00
|
|
|
if (is_vmalloc_addr(memmap))
|
2005-10-29 19:16:55 -06:00
|
|
|
vfree(memmap);
|
|
|
|
else
|
|
|
|
free_pages((unsigned long)memmap,
|
2013-11-12 16:07:42 -07:00
|
|
|
get_order(sizeof(struct page) * PAGES_PER_SECTION));
|
2005-10-29 19:16:55 -06:00
|
|
|
}
|
2008-04-28 03:13:34 -06:00
|
|
|
|
2013-04-29 16:08:22 -06:00
|
|
|
#ifdef CONFIG_MEMORY_HOTREMOVE
|
2013-11-12 16:07:43 -07:00
|
|
|
static void free_map_bootmem(struct page *memmap)
|
2008-04-28 03:13:34 -06:00
|
|
|
{
|
|
|
|
unsigned long maps_section_nr, removing_section_nr, i;
|
2013-11-12 16:07:43 -07:00
|
|
|
unsigned long magic, nr_pages;
|
2012-11-29 14:54:21 -07:00
|
|
|
struct page *page = virt_to_page(memmap);
|
2008-04-28 03:13:34 -06:00
|
|
|
|
2013-11-12 16:07:43 -07:00
|
|
|
nr_pages = PAGE_ALIGN(PAGES_PER_SECTION * sizeof(struct page))
|
|
|
|
>> PAGE_SHIFT;
|
|
|
|
|
2008-04-28 03:13:34 -06:00
|
|
|
for (i = 0; i < nr_pages; i++, page++) {
|
2017-02-22 16:45:13 -07:00
|
|
|
magic = (unsigned long) page->freelist;
|
2008-04-28 03:13:34 -06:00
|
|
|
|
|
|
|
BUG_ON(magic == NODE_INFO);
|
|
|
|
|
|
|
|
maps_section_nr = pfn_to_section_nr(page_to_pfn(page));
|
2017-02-22 16:45:10 -07:00
|
|
|
removing_section_nr = page_private(page);
|
2008-04-28 03:13:34 -06:00
|
|
|
|
|
|
|
/*
|
|
|
|
* When this function is called, the removing section is
|
|
|
|
* logical offlined state. This means all pages are isolated
|
|
|
|
* from page allocator. If removing section's memmap is placed
|
|
|
|
* on the same section, it must not be freed.
|
|
|
|
* If it is freed, page allocator may allocate it which will
|
|
|
|
* be removed physically soon.
|
|
|
|
*/
|
|
|
|
if (maps_section_nr != removing_section_nr)
|
|
|
|
put_page_bootmem(page);
|
|
|
|
}
|
|
|
|
}
|
2013-04-29 16:08:22 -06:00
|
|
|
#endif /* CONFIG_MEMORY_HOTREMOVE */
|
2007-10-16 02:26:14 -06:00
|
|
|
#endif /* CONFIG_SPARSEMEM_VMEMMAP */
|
2005-10-29 19:16:55 -06:00
|
|
|
|
2005-06-23 01:08:00 -06:00
|
|
|
/*
|
|
|
|
* returns the number of sections whose mem_maps were properly
|
|
|
|
* set. If this is <=0, then that means that the passed-in
|
|
|
|
* map was not consumed and must be freed.
|
|
|
|
*/
|
2017-12-29 00:53:54 -07:00
|
|
|
int __meminit sparse_add_one_section(struct pglist_data *pgdat,
|
|
|
|
unsigned long start_pfn, struct vmem_altmap *altmap)
|
2005-06-23 01:08:00 -06:00
|
|
|
{
|
2005-10-29 19:16:55 -06:00
|
|
|
unsigned long section_nr = pfn_to_section_nr(start_pfn);
|
|
|
|
struct mem_section *ms;
|
|
|
|
struct page *memmap;
|
2007-10-16 02:25:56 -06:00
|
|
|
unsigned long *usemap;
|
2005-10-29 19:16:55 -06:00
|
|
|
unsigned long flags;
|
|
|
|
int ret;
|
2005-06-23 01:08:00 -06:00
|
|
|
|
2005-10-29 19:16:55 -06:00
|
|
|
/*
|
|
|
|
* no locking for this, because it does its own
|
|
|
|
* plus, it does a kmalloc
|
|
|
|
*/
|
2007-12-17 17:19:59 -07:00
|
|
|
ret = sparse_index_init(section_nr, pgdat->node_id);
|
|
|
|
if (ret < 0 && ret != -EEXIST)
|
|
|
|
return ret;
|
2017-12-29 00:53:54 -07:00
|
|
|
memmap = kmalloc_section_memmap(section_nr, pgdat->node_id, altmap);
|
2007-12-17 17:19:59 -07:00
|
|
|
if (!memmap)
|
|
|
|
return -ENOMEM;
|
2007-10-16 02:25:56 -06:00
|
|
|
usemap = __kmalloc_section_usemap();
|
2007-12-17 17:19:59 -07:00
|
|
|
if (!usemap) {
|
2017-12-29 00:53:56 -07:00
|
|
|
__kfree_section_memmap(memmap, altmap);
|
2007-12-17 17:19:59 -07:00
|
|
|
return -ENOMEM;
|
|
|
|
}
|
2005-10-29 19:16:55 -06:00
|
|
|
|
|
|
|
pgdat_resize_lock(pgdat, &flags);
|
2005-06-23 01:08:00 -06:00
|
|
|
|
2005-10-29 19:16:55 -06:00
|
|
|
ms = __pfn_to_section(start_pfn);
|
|
|
|
if (ms->section_mem_map & SECTION_MARKED_PRESENT) {
|
|
|
|
ret = -EEXIST;
|
|
|
|
goto out;
|
|
|
|
}
|
2007-10-16 02:25:56 -06:00
|
|
|
|
2018-04-05 17:23:00 -06:00
|
|
|
#ifdef CONFIG_DEBUG_VM
|
|
|
|
/*
|
|
|
|
* Poison uninitialized struct pages in order to catch invalid flags
|
|
|
|
* combinations.
|
|
|
|
*/
|
|
|
|
memset(memmap, PAGE_POISON_PATTERN, sizeof(struct page) * PAGES_PER_SECTION);
|
|
|
|
#endif
|
2012-12-11 17:00:59 -07:00
|
|
|
|
mm, sparsemem: break out of loops early
There are a number of times that we loop over NR_MEM_SECTIONS, looking
for section_present() on each section. But, when we have very large
physical address spaces (large MAX_PHYSMEM_BITS), NR_MEM_SECTIONS
becomes very large, making the loops quite long.
With MAX_PHYSMEM_BITS=46 and a section size of 128MB, the current loops
are 512k iterations, which we barely notice on modern hardware. But,
raising MAX_PHYSMEM_BITS higher (like we will see on systems that
support 5-level paging) makes this 64x longer and we start to notice,
especially on slower systems like simulators. A 10-second delay for
512k iterations is annoying. But, a 640- second delay is crippling.
This does not help if we have extremely sparse physical address spaces,
but those are quite rare. We expect that most of the "slow" systems
where this matters will also be quite small and non-sparse.
To fix this, we track the highest section we've ever encountered. This
lets us know when we will *never* see another section_present(), and
lets us break out of the loops earlier.
Doing the whole for_each_present_section_nr() macro is probably
overkill, but it will ensure that any future loop iterations that we
grow are more likely to be correct.
Kirrill said "It shaved almost 40 seconds from boot time in qemu with
5-level paging enabled for me".
Link: http://lkml.kernel.org/r/20170504174434.C45A4735@viggo.jf.intel.com
Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com>
Tested-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2017-07-06 16:36:44 -06:00
|
|
|
section_mark_present(ms);
|
2005-06-23 01:08:00 -06:00
|
|
|
|
2007-10-16 02:25:56 -06:00
|
|
|
ret = sparse_init_one_section(ms, section_nr, memmap, usemap);
|
2005-10-29 19:16:55 -06:00
|
|
|
|
|
|
|
out:
|
|
|
|
pgdat_resize_unlock(pgdat, &flags);
|
2007-12-17 17:19:59 -07:00
|
|
|
if (ret <= 0) {
|
|
|
|
kfree(usemap);
|
2017-12-29 00:53:56 -07:00
|
|
|
__kfree_section_memmap(memmap, altmap);
|
2007-12-17 17:19:59 -07:00
|
|
|
}
|
2005-10-29 19:16:55 -06:00
|
|
|
return ret;
|
2005-06-23 01:08:00 -06:00
|
|
|
}
|
2008-04-28 03:12:01 -06:00
|
|
|
|
2013-07-08 17:00:10 -06:00
|
|
|
#ifdef CONFIG_MEMORY_HOTREMOVE
|
2012-12-11 17:00:47 -07:00
|
|
|
#ifdef CONFIG_MEMORY_FAILURE
|
|
|
|
static void clear_hwpoisoned_pages(struct page *memmap, int nr_pages)
|
|
|
|
{
|
|
|
|
int i;
|
|
|
|
|
|
|
|
if (!memmap)
|
|
|
|
return;
|
|
|
|
|
2016-01-15 17:56:22 -07:00
|
|
|
for (i = 0; i < nr_pages; i++) {
|
2012-12-11 17:00:47 -07:00
|
|
|
if (PageHWPoison(&memmap[i])) {
|
2013-02-22 17:34:02 -07:00
|
|
|
atomic_long_sub(1, &num_poisoned_pages);
|
2012-12-11 17:00:47 -07:00
|
|
|
ClearPageHWPoison(&memmap[i]);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
#else
|
|
|
|
static inline void clear_hwpoisoned_pages(struct page *memmap, int nr_pages)
|
|
|
|
{
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
|
2017-12-29 00:53:56 -07:00
|
|
|
static void free_section_usemap(struct page *memmap, unsigned long *usemap,
|
|
|
|
struct vmem_altmap *altmap)
|
2013-04-29 16:08:22 -06:00
|
|
|
{
|
|
|
|
struct page *usemap_page;
|
|
|
|
|
|
|
|
if (!usemap)
|
|
|
|
return;
|
|
|
|
|
|
|
|
usemap_page = virt_to_page(usemap);
|
|
|
|
/*
|
|
|
|
* Check to see if allocation came from hot-plug-add
|
|
|
|
*/
|
|
|
|
if (PageSlab(usemap_page) || PageCompound(usemap_page)) {
|
|
|
|
kfree(usemap);
|
|
|
|
if (memmap)
|
2017-12-29 00:53:56 -07:00
|
|
|
__kfree_section_memmap(memmap, altmap);
|
2013-04-29 16:08:22 -06:00
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* The usemap came from bootmem. This is packed with other usemaps
|
|
|
|
* on the section which has pgdat at boot time. Just keep it as is now.
|
|
|
|
*/
|
|
|
|
|
2013-11-12 16:07:43 -07:00
|
|
|
if (memmap)
|
|
|
|
free_map_bootmem(memmap);
|
2013-04-29 16:08:22 -06:00
|
|
|
}
|
|
|
|
|
2016-01-15 17:56:22 -07:00
|
|
|
void sparse_remove_one_section(struct zone *zone, struct mem_section *ms,
|
2017-12-29 00:53:56 -07:00
|
|
|
unsigned long map_offset, struct vmem_altmap *altmap)
|
2008-04-28 03:12:01 -06:00
|
|
|
{
|
|
|
|
struct page *memmap = NULL;
|
2013-02-22 17:33:02 -07:00
|
|
|
unsigned long *usemap = NULL, flags;
|
|
|
|
struct pglist_data *pgdat = zone->zone_pgdat;
|
2008-04-28 03:12:01 -06:00
|
|
|
|
2013-02-22 17:33:02 -07:00
|
|
|
pgdat_resize_lock(pgdat, &flags);
|
2008-04-28 03:12:01 -06:00
|
|
|
if (ms->section_mem_map) {
|
|
|
|
usemap = ms->pageblock_flags;
|
|
|
|
memmap = sparse_decode_mem_map(ms->section_mem_map,
|
|
|
|
__section_nr(ms));
|
|
|
|
ms->section_mem_map = 0;
|
|
|
|
ms->pageblock_flags = NULL;
|
|
|
|
}
|
2013-02-22 17:33:02 -07:00
|
|
|
pgdat_resize_unlock(pgdat, &flags);
|
2008-04-28 03:12:01 -06:00
|
|
|
|
2016-01-15 17:56:22 -07:00
|
|
|
clear_hwpoisoned_pages(memmap + map_offset,
|
|
|
|
PAGES_PER_SECTION - map_offset);
|
2017-12-29 00:53:56 -07:00
|
|
|
free_section_usemap(memmap, usemap, altmap);
|
2008-04-28 03:12:01 -06:00
|
|
|
}
|
2013-04-29 16:08:22 -06:00
|
|
|
#endif /* CONFIG_MEMORY_HOTREMOVE */
|
|
|
|
#endif /* CONFIG_MEMORY_HOTPLUG */
|