FAILED: patch "[PATCH] mm: call cond_resched() from deferred_init_memmap()" failed to apply to 5.7-stable tree

PATCH [email protected] 5 participants 9 messages Jun 18, 2020 – Jun 19, 2020

Subsystems (3)

Stable branch Memory management Page allocator

lore

Share & tools

lore

[email protected]

author maintainer

2020-06-18 15:55 UTC #0 lore raw

The patch below does not apply to the 5.7-stable tree.
If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <[email protected]>.

thanks,

greg k-h

------------------ original commit in Linus's tree ------------------

From da97f2d56bbd880b4138916a7ef96f9881a551b2 Mon Sep 17 00:00:00 2001
From: Pavel Tatashin <[email protected]>
Date: Wed, 3 Jun 2020 15:59:27 -0700
Subject: [PATCH] mm: call cond_resched() from deferred_init_memmap()

Now that deferred pages are initialized with interrupts enabled we can replace touch_nmi_watchdog() with cond_resched(), as it was before 3a2d7fa8a3d5.

For now, we cannot do the same in deferred_grow_zone() as it is still initializes pages with interrupts disabled.

This change fixes RCU problem described in
https://lkml.kernel.org/r/[email protected]

[   60.474005] rcu: INFO: rcu_sched detected stalls on CPUs/tasks:
[   60.475000] rcu:  1-...0: (0 ticks this GP) idle=02a/1/0x4000000000000000 softirq=1/1 fqs=15000
[   60.475000] rcu:  (detected by 0, t=60002 jiffies, g=-1199, q=1)
[   60.475000] Sending NMI from CPU 0 to CPUs 1:
[    1.760091] NMI backtrace for cpu 1
[    1.760091] CPU: 1 PID: 20 Comm: pgdatinit0 Not tainted 4.18.0-147.9.1.el8_1.x86_64 #1
[    1.760091] Hardware name: Red Hat KVM, BIOS 1.13.0-1.module+el8.2.0+5520+4e5817f3 04/01/2014
[    1.760091] RIP: 0010:__init_single_page.isra.65+0x10/0x4f
[    1.760091] Code: 48 83 cf 63 48 89 f8 0f 1f 40 00 48 89 c6 48 89 d7 e8 6b 18 80 ff 66 90 5b c3 31 c0 b9 10 00 00 00 49 89 f8 48 c1 e6 33 f3 ab <b8> 07 00 00 00 48 c1 e2 36 41 c7 40 34 01 00 00 00 48 c1 e0 33 41
[    1.760091] RSP: 0000:ffffba783123be40 EFLAGS: 00000006
[    1.760091] RAX: 0000000000000000 RBX: fffffad34405e300 RCX: 0000000000000000
[    1.760091] RDX: 0000000000000000 RSI: 0010000000000000 RDI: fffffad34405e340
[    1.760091] RBP: 0000000033f3177e R08: fffffad34405e300 R09: 0000000000000002
[    1.760091] R10: 000000000000002b R11: ffff98afb691a500 R12: 0000000000000002
[    1.760091] R13: 0000000000000000 R14: 000000003f03ea00 R15: 000000003e10178c
[    1.760091] FS:  0000000000000000(0000) GS:ffff9c9ebeb00000(0000) knlGS:0000000000000000
[    1.760091] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[    1.760091] CR2: 00000000ffffffff CR3: 000000a1cf20a001 CR4: 00000000003606e0
[    1.760091] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[    1.760091] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
[    1.760091] Call Trace:
[    1.760091]  deferred_init_pages+0x8f/0xbf
[    1.760091]  deferred_init_memmap+0x184/0x29d
[    1.760091]  ? deferred_free_pages.isra.97+0xba/0xba
[    1.760091]  kthread+0x112/0x130
[    1.760091]  ? kthread_flush_work_fn+0x10/0x10
[    1.760091]  ret_from_fork+0x35/0x40
[   89.123011] node 0 initialised, 1055935372 pages in 88650ms

Fixes 3a2d7fa8a3d5 ("mm: disable interrupts while initializing deferred pages")

Reported-by Yiqian Wei <[email protected]>

Signed-off-by Pavel Tatashin <[email protected]>

Signed-off-by Andrew Morton <[email protected]>

Tested-by David Hildenbrand <[email protected]>

Reviewed-by Daniel Jordan <[email protected]>

Reviewed-by David Hildenbrand <[email protected]>

Reviewed-by Pankaj Gupta <[email protected]>

Acked-by Michal Hocko <[email protected]>

Cc Dan Williams <[email protected]>

Cc James Morris <[email protected]>

Cc Kirill Tkhai <[email protected]>

Cc Sasha Levin <[email protected]>

Cc Shile Zhang <[email protected]>

Cc Vlastimil Babka <[email protected]>

Cc <[email protected]> [4.17+]

Link http://lkml.kernel.org/r/[email protected] open here

Signed-off-by Linus Torvalds <[email protected]>

mm/page_alloc.c +1 -1

index c75561a3f144..2dbc571f68de 100644
@@ -1870,7 +1870,7 @@ static int __init deferred_init_memmap(void *data)
 	 */
 	while (spfn < epfn) {
 		nr_pages += deferred_init_maxorder(&i, zone, &spfn, &epfn);
-		touch_nmi_watchdog();
+		cond_resched();
 	}
 zone_empty:
 	/* Sanity check that the next zone really is unpopulated */

Greg KH

author maintainer

2020-06-18 16:26 UTC #1 lore raw

Quoted 5 lines · On Thu, Jun 18, 2020 at 05:55:43PM +0200, [email protected] wrote:

The patch below does not apply to the 5.7-stable tree.
If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <[email protected]>.

Oops, I applied things out of order, I've queued this and the 5.4 version up, but 4.19 doesn't apply as the dependant patch does not apply.

thanks

greg k-h

Sasha Levin

maintainer

2020-06-19 02:28 UTC #2 lore raw

Quoted 10 lines · Greg KH

Quoted 5 lines · On Thu, Jun 18, 2020 at 05:55:43PM +0200, [email protected] wrote:
The patch below does not apply to the 5.7-stable tree.
If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <[email protected]>.
Oops, I applied things out of order, I've queued this and the 5.4 version up, but 4.19 doesn't apply as the dependant patch does not apply.

Something like this should work?

mm/page_alloc.c +0 -0

index 7181dfe76440..b7905a075e79 100644
@@ -1611,11 +1611,13 @@ static int __init deferred_init_memmap(void *data)

		spfn = max_t(unsigned long, first_init_pfn, PFN_UP(spa));
		epfn = min_t(unsigned long, zone_end_pfn(zone), PFN_DOWN(epa));
		nr_pages += deferred_init_pages(nid, zid, spfn, epfn);
+		cond_resched();
	}
	for_each_free_mem_range(i, nid, MEMBLOCK_NONE, &spa, &epa, NULL) {
		spfn = max_t(unsigned long, first_init_pfn, PFN_UP(spa));
		epfn = min_t(unsigned long, zone_end_pfn(zone), PFN_DOWN(epa));
		deferred_free_pages(nid, zid, spfn, epfn);
+		cond_resched();
	}

	/* Sanity check that the next zone really is unpopulated */

Signature · 2 lines

Thanks,
Sasha

Michal Hocko

2020-06-19 09:21 UTC #3 lore raw

Quoted 8 lines · Sasha Levin

Quoted 7 lines · Greg KH
Quoted 5 lines · On Thu, Jun 18, 2020 at 05:55:43PM +0200, [email protected] wrote:
The patch below does not apply to the 5.7-stable tree.
If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <[email protected]>.

Quoted 5 lines

Quoted 3 lines
Oops, I applied things out of order, I've queued this and the 5.4 version up, but 4.19 doesn't apply as the dependant patch does not apply.
Something like this should work?

Nope. Unless I am misreading the old code you are udner pgdat_resize_lock. Or is there any other change queued before this backport to remove the lock? (I didn't get to check more closely but it would be 3d060856adfc5 IIRC).

Quoted 22 lines

mm/page_alloc.c +2 -0

index 7181dfe76440..b7905a075e79 100644
@@ -1611,11 +1611,13 @@ static int __init deferred_init_memmap(void *data)
 		spfn = max_t(unsigned long, first_init_pfn, PFN_UP(spa));
 		epfn = min_t(unsigned long, zone_end_pfn(zone), PFN_DOWN(epa));
 		nr_pages += deferred_init_pages(nid, zid, spfn, epfn);
+		cond_resched();
 	}
 	for_each_free_mem_range(i, nid, MEMBLOCK_NONE, &spa, &epa, NULL) {
 		spfn = max_t(unsigned long, first_init_pfn, PFN_UP(spa));
 		epfn = min_t(unsigned long, zone_end_pfn(zone), PFN_DOWN(epa));
 		deferred_free_pages(nid, zid, spfn, epfn);
+		cond_resched();
 	}
 
 	/* Sanity check that the next zone really is unpopulated */

Signature · 3 lines

-- 
Thanks,
Sasha

Signature · 2 lines

Michal Hocko
SUSE Labs

David Hildenbrand

2020-06-19 09:44 UTC #4 lore raw

Quoted 13 lines · Michal Hocko

Quoted 12 lines · Sasha Levin
Quoted 10 lines · Greg KH
Quoted 5 lines · On Thu, Jun 18, 2020 at 05:55:43PM +0200, [email protected] wrote:
The patch below does not apply to the 5.7-stable tree.
If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <[email protected]>.
Oops, I applied things out of order, I've queued this and the 5.4 version up, but 4.19 doesn't apply as the dependant patch does not apply.

Quoted 6 lines

Quoted 1 line
Something like this should work?
Nope. Unless I am misreading the old code you are udner pgdat_resize_lock. Or is there any other change queued before this backport to remove the lock? (I didn't get to check more closely but it would be 3d060856adfc5 IIRC).

I recently did a similar backport. For pre-5.2, the following commits might be required to backport cleanly

56ec43d8b027 mm: drop meminit_pfn_in_nid as it is redundant 837566e7e08e mm: implement new zone specific memblock iterator 0e56acae4b4d mm: initialize MAX_ORDER_NR_PAGES at a time instead of doing larger sections
b9705d8778e7 mm/page_alloc.c: fix regression with deferred struct page init 117003c32771 mm/pagealloc.c: call touch_nmi_watchdog() on max order boundaries in deferred init
3d060856adfc mm: initialize deferred pages with interrupts enabled da97f2d56bbd mm: call cond_resched() from deferred_init_memmap()

did not verify, though, if anything else is missing (and which commits can be bypassed with less trouble).

Signature · 3 lines

Thanks,

David / dhildenb

Pavel Tatashin

2020-06-19 12:16 UTC #5 lore raw

I will send the backport. I think only three patches are needed that were in my original series:

117003c32771 mm/pagealloc.c: call touch_nmi_watchdog() on max order boundaries in deferred init
3d060856adfc mm: initialize deferred pages with interrupts enabled da97f2d56bbd mm: call cond_resched() from deferred_init_memmap()

Pasha

Quoted 43 lines · David Hildenbrand

Quoted 13 lines · Michal Hocko
Quoted 12 lines · Sasha Levin
Quoted 10 lines · Greg KH
Quoted 5 lines · On Thu, Jun 18, 2020 at 05:55:43PM +0200, [email protected] wrote:
The patch below does not apply to the 5.7-stable tree.
If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <[email protected]>.
Oops, I applied things out of order, I've queued this and the 5.4 version up, but 4.19 doesn't apply as the dependant patch does not apply.
Quoted 6 lines
Quoted 1 line
Something like this should work?
Nope. Unless I am misreading the old code you are udner pgdat_resize_lock. Or is there any other change queued before this backport to remove the lock? (I didn't get to check more closely but it would be 3d060856adfc5 IIRC).
I recently did a similar backport. For pre-5.2, the following commits might be required to backport cleanly
56ec43d8b027 mm: drop meminit_pfn_in_nid as it is redundant 837566e7e08e mm: implement new zone specific memblock iterator 0e56acae4b4d mm: initialize MAX_ORDER_NR_PAGES at a time instead of doing larger sections
b9705d8778e7 mm/page_alloc.c: fix regression with deferred struct page init 117003c32771 mm/pagealloc.c: call touch_nmi_watchdog() on max order boundaries in deferred init
3d060856adfc mm: initialize deferred pages with interrupts enabled da97f2d56bbd mm: call cond_resched() from deferred_init_memmap()
did not verify, though, if anything else is missing (and which commits can be bypassed with less trouble).
--
Thanks,
David / dhildenb

David Hildenbrand

2020-06-19 12:27 UTC #6 lore raw

Quoted 8 lines · Pavel Tatashin

I will send the backport. I think only three patches are needed that were in my original series:
117003c32771 mm/pagealloc.c: call touch_nmi_watchdog() on max order boundaries in deferred init
3d060856adfc mm: initialize deferred pages with interrupts enabled da97f2d56bbd mm: call cond_resched() from deferred_init_memmap()

Just to note, in contrast to the subject, this is about the 4.19 backport IIRC. Is it really only these patches?

Signature · 3 lines

Thanks,

David / dhildenb

Pavel Tatashin

2020-06-19 12:44 UTC #7 lore raw

Quoted 8 lines

Quoted 5 lines
117003c32771 mm/pagealloc.c: call touch_nmi_watchdog() on max order boundaries in deferred init
3d060856adfc mm: initialize deferred pages with interrupts enabled da97f2d56bbd mm: call cond_resched() from deferred_init_memmap()
Just to note, in contrast to the subject, this is about the 4.19 backport IIRC. Is it really only these patches?

Ah, I did not realize it was 4.19 backport.

Yeah, it might be more, I have not checked it.

Thanks,
Pasha

Pavel Tatashin

2020-06-19 13:28 UTC #8 lore raw

Quoted 7 lines · Pavel Tatashin

Quoted 7 lines
Quoted 6 lines
117003c32771 mm/pagealloc.c: call touch_nmi_watchdog() on max order boundaries in deferred init
3d060856adfc mm: initialize deferred pages with interrupts enabled da97f2d56bbd mm: call cond_resched() from deferred_init_memmap()

Quoted 6 lines

Quoted 2 lines
Just to note, in contrast to the subject, this is about the 4.19 backport IIRC. Is it really only these patches?
Ah, I did not realize it was 4.19 backport.
Yeah, it might be more, I have not checked it.

5.7 backport:
https://lore.kernel.org/stable/[email protected]
4.19 backport:
https://lore.kernel.org/stable/[email protected]

Quoted 3 lines

Thanks,
Pasha