The patch titled
Subject: mm: make deferred init's max threads arch-specific
has been added to the -mm tree. Its filename is
mm-make-deferred-inits-max-threads-arch-specific.patch
This patch should soon appear at
http://ozlabs.org/~akpm/mmots/broken-out/mm-make-deferred-inits-max-threads-arch-specific.patch
and later at
http://ozlabs.org/~akpm/mmotm/broken-out/mm-make-deferred-inits-max-threads-arch-specific.patch
Before you just go and hit "reply", please:
a) Consider who else should be cc'ed
b) Prefer to cc a suitable mailing list as well
c) Ideally: find the original patch on the mailing list and do a
reply-to-all to that, adding suitable additional cc's
*** Remember to use Documentation/process/submit-checklist.rst when testing your code ***
The -mm tree is included into linux-next and is updated
there every 3-4 working days
------------------------------------------------------
From: Daniel Jordan <daniel.m.jordan@xxxxxxxxxx>
Subject: mm: make deferred init's max threads arch-specific
Using padata during deferred init has only been tested on x86, so for now
limit it to this architecture.
If another arch wants this, it can find the max thread limit that's best
for it and override deferred_page_init_max_threads().
Link: http://lkml.kernel.org/r/20200520182645.1658949-7-daniel.m.jordan@xxxxxxxxxx
Signed-off-by: Daniel Jordan <daniel.m.jordan@xxxxxxxxxx>
Cc: Alexander Duyck <alexander.h.duyck@xxxxxxxxxxxxxxx>
Cc: Alex Williamson <alex.williamson@xxxxxxxxxx>
Cc: Dan Williams <dan.j.williams@xxxxxxxxx>
Cc: Dave Hansen <dave.hansen@xxxxxxxxxxxxxxx>
Cc: David Hildenbrand <david@xxxxxxxxxx>
Cc: Herbert Xu <herbert@xxxxxxxxxxxxxxxxxxx>
Cc: Jason Gunthorpe <jgg@xxxxxxxx>
Cc: Jonathan Corbet <corbet@xxxxxxx>
Cc: Josh Triplett <josh@xxxxxxxxxxxxxxxx>
Cc: Kirill Tkhai <ktkhai@xxxxxxxxxxxxx>
Cc: Michal Hocko <mhocko@xxxxxxxxxx>
Cc: Pavel Machek <pavel@xxxxxx>
Cc: Pavel Tatashin <pasha.tatashin@xxxxxxxxxx>
Cc: Peter Zijlstra <peterz@xxxxxxxxxxxxx>
Cc: Randy Dunlap <rdunlap@xxxxxxxxxxxxx>
Cc: Robert Elliott <elliott@xxxxxxx>
Cc: Shile Zhang <shile.zhang@xxxxxxxxxxxxxxxxx>
Cc: Steffen Klassert <steffen.klassert@xxxxxxxxxxx>
Cc: Steven Sistare <steven.sistare@xxxxxxxxxx>
Cc: Tejun Heo <tj@xxxxxxxxxx>
Cc: Zi Yan <ziy@xxxxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
arch/x86/mm/init_64.c | 12 ++++++++++++
include/linux/memblock.h | 3 +++
mm/page_alloc.c | 13 ++++++++-----
3 files changed, 23 insertions(+), 5 deletions(-)
--- a/arch/x86/mm/init_64.c~mm-make-deferred-inits-max-threads-arch-specific
+++ a/arch/x86/mm/init_64.c
@@ -1265,6 +1265,18 @@ void __init mem_init(void)
mem_init_print_info(NULL);
}
+#ifdef CONFIG_DEFERRED_STRUCT_PAGE_INIT
+int __init deferred_page_init_max_threads(const struct cpumask *node_cpumask)
+{
+ /*
+ * More CPUs always led to greater speedups on tested systems, up to
+ * all the nodes' CPUs. Use all since the system is otherwise idle
+ * now.
+ */
+ return max_t(int, cpumask_weight(node_cpumask), 1);
+}
+#endif
+
int kernel_set_to_readonly;
void mark_rodata_ro(void)
--- a/include/linux/memblock.h~mm-make-deferred-inits-max-threads-arch-specific
+++ a/include/linux/memblock.h
@@ -273,6 +273,9 @@ void __next_mem_pfn_range_in_zone(u64 *i
#define for_each_free_mem_pfn_range_in_zone_from(i, zone, p_start, p_end) \
for (; i != U64_MAX; \
__next_mem_pfn_range_in_zone(&i, zone, p_start, p_end))
+
+int __init deferred_page_init_max_threads(const struct cpumask *node_cpumask);
+
#endif /* CONFIG_DEFERRED_STRUCT_PAGE_INIT */
/**
--- a/mm/page_alloc.c~mm-make-deferred-inits-max-threads-arch-specific
+++ a/mm/page_alloc.c
@@ -1844,6 +1844,13 @@ deferred_init_memmap_chunk(unsigned long
atomic_long_add(nr_pages, &args->nr_pages);
}
+/* An arch may override for more concurrency. */
+__weak int __init
+deferred_page_init_max_threads(const struct cpumask *node_cpumask)
+{
+ return 1;
+}
+
/* Initialise remaining memory on a node */
static int __init deferred_init_memmap(void *data)
{
@@ -1892,11 +1899,7 @@ static int __init deferred_init_memmap(v
first_init_pfn))
goto zone_empty;
- /*
- * More CPUs always led to greater speedups on tested systems, up to
- * all the nodes' CPUs. Use all since the system is otherwise idle now.
- */
- max_threads = max(cpumask_weight(cpumask), 1u);
+ max_threads = deferred_page_init_max_threads(cpumask);
while (spfn < epfn) {
epfn_align = ALIGN_DOWN(epfn, PAGES_PER_SECTION);
_
Patches currently in -mm which might be from daniel.m.jordan@xxxxxxxxxx are
mm-call-touch_nmi_watchdog-on-max-order-boundaries-in-deferred-init.patch
padata-remove-exit-routine.patch
padata-initialize-earlier.patch
padata-allocate-work-structures-for-parallel-jobs-from-a-pool.patch
padata-add-basic-support-for-multithreaded-jobs.patch
mm-parallelize-deferred_init_memmap.patch
mm-make-deferred-inits-max-threads-arch-specific.patch
padata-document-multithreaded-jobs.patch
