linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Pavel Tatashin <pasha.tatashin@oracle.com>
To: Michal Hocko <mhocko@kernel.org>
Cc: Naoya Horiguchi <n-horiguchi@ah.jp.nec.com>,
	"linux-mm@kvack.org" <linux-mm@kvack.org>,
	Oscar Salvador <osalvador@techadventures.net>,
	Oscar Salvador <osalvador@suse.de>,
	Steven Sistare <steven.sistare@oracle.com>,
	Daniel Jordan <daniel.m.jordan@oracle.com>,
	Matthew Wilcox <willy@infradead.org>,
	"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
	Andrew Morton <akpm@linux-foundation.org>,
	"mingo@kernel.org" <mingo@kernel.org>,
	"dan.j.williams@intel.com" <dan.j.williams@intel.com>,
	Huang Ying <ying.huang@intel.com>
Subject: Re: [PATCH v3] x86/e820: put !E820_TYPE_RAM regions into memblock.reserved
Date: Fri, 15 Jun 2018 10:00:00 -0400	[thread overview]
Message-ID: <20180615140000.44tht4f3ek3lh2u2@xakep.localdomain> (raw)
In-Reply-To: <20180615084142.GE24039@dhcp22.suse.cz>

On 18-06-15 10:41:42, Michal Hocko wrote:
> On Fri 15-06-18 07:29:48, Naoya Horiguchi wrote:
> [...]
> > From: Naoya Horiguchi <n-horiguchi@ah.jp.nec.com>
> > Date: Thu, 14 Jun 2018 16:04:36 +0900
> > Subject: [PATCH] x86/e820: put !E820_TYPE_RAM regions into memblock.reserved
> > 
> > There is a kernel panic that is triggered when reading /proc/kpageflags
> > on the kernel booted with kernel parameter 'memmap=nn[KMG]!ss[KMG]':
> > 
> >   BUG: unable to handle kernel paging request at fffffffffffffffe
> >   PGD 9b20e067 P4D 9b20e067 PUD 9b210067 PMD 0
> >   Oops: 0000 [#1] SMP PTI
> >   CPU: 2 PID: 1728 Comm: page-types Not tainted 4.17.0-rc6-mm1-v4.17-rc6-180605-0816-00236-g2dfb086ef02c+ #160
> >   Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.11.0-2.fc28 04/01/2014
> >   RIP: 0010:stable_page_flags+0x27/0x3c0
> >   Code: 00 00 00 0f 1f 44 00 00 48 85 ff 0f 84 a0 03 00 00 41 54 55 49 89 fc 53 48 8b 57 08 48 8b 2f 48 8d 42 ff 83 e2 01 48 0f 44 c7 <48> 8b 00 f6 c4 01 0f 84 10 03 00 00 31 db 49 8b 54 24 08 4c 89 e7
> >   RSP: 0018:ffffbbd44111fde0 EFLAGS: 00010202
> >   RAX: fffffffffffffffe RBX: 00007fffffffeff9 RCX: 0000000000000000
> >   RDX: 0000000000000001 RSI: 0000000000000202 RDI: ffffed1182fff5c0
> >   RBP: ffffffffffffffff R08: 0000000000000001 R09: 0000000000000001
> >   R10: ffffbbd44111fed8 R11: 0000000000000000 R12: ffffed1182fff5c0
> >   R13: 00000000000bffd7 R14: 0000000002fff5c0 R15: ffffbbd44111ff10
> >   FS:  00007efc4335a500(0000) GS:ffff93a5bfc00000(0000) knlGS:0000000000000000
> >   CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> >   CR2: fffffffffffffffe CR3: 00000000b2a58000 CR4: 00000000001406e0
> >   Call Trace:
> >    kpageflags_read+0xc7/0x120
> >    proc_reg_read+0x3c/0x60
> >    __vfs_read+0x36/0x170
> >    vfs_read+0x89/0x130
> >    ksys_pread64+0x71/0x90
> >    do_syscall_64+0x5b/0x160
> >    entry_SYSCALL_64_after_hwframe+0x44/0xa9
> >   RIP: 0033:0x7efc42e75e23
> >   Code: 09 00 ba 9f 01 00 00 e8 ab 81 f4 ff 66 2e 0f 1f 84 00 00 00 00 00 90 83 3d 29 0a 2d 00 00 75 13 49 89 ca b8 11 00 00 00 0f 05 <48> 3d 01 f0 ff ff 73 34 c3 48 83 ec 08 e8 db d3 01 00 48 89 04 24
> > 
> > According to kernel bisection, this problem became visible due to commit
> > f7f99100d8d9 which changes how struct pages are initialized.
> > 
> > Memblock layout affects the pfn ranges covered by node/zone. Consider
> > that we have a VM with 2 NUMA nodes and each node has 4GB memory, and
> > the default (no memmap= given) memblock layout is like below:
> > 
> >   MEMBLOCK configuration:
> >    memory size = 0x00000001fff75c00 reserved size = 0x000000000300c000
> >    memory.cnt  = 0x4
> >    memory[0x0]     [0x0000000000001000-0x000000000009efff], 0x000000000009e000 bytes on node 0 flags: 0x0
> >    memory[0x1]     [0x0000000000100000-0x00000000bffd6fff], 0x00000000bfed7000 bytes on node 0 flags: 0x0
> >    memory[0x2]     [0x0000000100000000-0x000000013fffffff], 0x0000000040000000 bytes on node 0 flags: 0x0
> >    memory[0x3]     [0x0000000140000000-0x000000023fffffff], 0x0000000100000000 bytes on node 1 flags: 0x0
> >    ...
> > 
> > If you give memmap=1G!4G (so it just covers memory[0x2]),
> > the range [0x100000000-0x13fffffff] is gone:
> > 
> >   MEMBLOCK configuration:
> >    memory size = 0x00000001bff75c00 reserved size = 0x000000000300c000
> >    memory.cnt  = 0x3
> >    memory[0x0]     [0x0000000000001000-0x000000000009efff], 0x000000000009e000 bytes on node 0 flags: 0x0
> >    memory[0x1]     [0x0000000000100000-0x00000000bffd6fff], 0x00000000bfed7000 bytes on node 0 flags: 0x0
> >    memory[0x2]     [0x0000000140000000-0x000000023fffffff], 0x0000000100000000 bytes on node 1 flags: 0x0
> >    ...
> > 
> > This causes shrinking node 0's pfn range because it is calculated by
> > the address range of memblock.memory. So some of struct pages in the
> > gap range are left uninitialized.
> > 
> > We have a function zero_resv_unavail() which does zeroing the struct
> > pages within the reserved unavailable range (i.e. memblock.memory &&
> > !memblock.reserved). This patch utilizes it to cover all unavailable
> > ranges by putting them into memblock.reserved.
> > 
> > Fixes: f7f99100d8d9 ("mm: stop zeroing memory during allocation in vmemmap")
> > Signed-off-by: Naoya Horiguchi <n-horiguchi@ah.jp.nec.com>
> > Tested-by: Oscar Salvador <osalvador@suse.de>
> 
> OK, this makes sense to me. It is definitely much better than the
> original attempt.
> 
> Unless I am missing something this should be correct
> Acked-by: Michal Hocko <mhocko@suse.com>

First of all thank you Naoya for finding and root causing this issue.

So, with this fix we reserve any hole and !E820_TYPE_RAM or
!E820_TYPE_RESERVED_KERN in e820.  I think, this will work because we
do pfn_valid() check in zero_resv_unavail(), so the ranges that do not have
backing struct pages will be skipped. But, I am worried on the performance
implications of when the holes of invalid memory are rather large. We would
have to loop through it in zero_resv_unavail() one pfn at a time.

Therefore, we might also need to optimize zero_resv_unavail() a little like
this:

6407			if (!pfn_valid(ALIGN_DOWN(pfn, pageblock_nr_pages)))
6408				continue;

Add before "continue":
	pfn = ALIGN_DOWN(pfn, pageblock_nr_pages) + pageblock_nr_pageas - 1.
At least, this way, we would skip a section of invalid memory at a time.

For the patch above:
Reviewed-by: Pavel Tatashin <pasha.tatashin@oracle.com>

But, I think the 2nd patch with the optimization above should go along this
this fix.

Thank you,
Pasha

> 
> > ---
> >  arch/x86/kernel/e820.c | 15 ++++++++++++---
> >  1 file changed, 12 insertions(+), 3 deletions(-)
> > 
> > diff --git a/arch/x86/kernel/e820.c b/arch/x86/kernel/e820.c
> > index d1f25c831447..c88c23c658c1 100644
> > --- a/arch/x86/kernel/e820.c
> > +++ b/arch/x86/kernel/e820.c
> > @@ -1248,6 +1248,7 @@ void __init e820__memblock_setup(void)
> >  {
> >  	int i;
> >  	u64 end;
> > +	u64 addr = 0;
> >  
> >  	/*
> >  	 * The bootstrap memblock region count maximum is 128 entries
> > @@ -1264,13 +1265,21 @@ void __init e820__memblock_setup(void)
> >  		struct e820_entry *entry = &e820_table->entries[i];
> >  
> >  		end = entry->addr + entry->size;
> > +		if (addr < entry->addr)
> > +			memblock_reserve(addr, entry->addr - addr);
> > +		addr = end;
> >  		if (end != (resource_size_t)end)
> >  			continue;
> >  
> > +		/*
> > +		 * all !E820_TYPE_RAM ranges (including gap ranges) are put
> > +		 * into memblock.reserved to make sure that struct pages in
> > +		 * such regions are not left uninitialized after bootup.
> > +		 */
> >  		if (entry->type != E820_TYPE_RAM && entry->type != E820_TYPE_RESERVED_KERN)
> > -			continue;
> > -
> > -		memblock_add(entry->addr, entry->size);
> > +			memblock_reserve(entry->addr, entry->size);
> > +		else
> > +			memblock_add(entry->addr, entry->size);
> >  	}
> >  
> >  	/* Throw away partial pages: */
> > -- 
> > 2.7.4
> 
> -- 
> Michal Hocko
> SUSE Labs
> 

  reply	other threads:[~2018-06-15 14:02 UTC|newest]

Thread overview: 39+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-06-05  0:54 kernel panic in reading /proc/kpageflags when enabling RAM-simulated PMEM Naoya Horiguchi
2018-06-05  1:18 ` Matthew Wilcox
2018-06-05  7:35   ` Naoya Horiguchi
2018-06-06  5:16     ` Naoya Horiguchi
2018-06-06  8:04       ` Oscar Salvador
2018-06-06  8:53         ` Oscar Salvador
2018-06-06  9:06           ` Naoya Horiguchi
2018-06-06  9:24             ` Naoya Horiguchi
2018-06-07  6:22               ` Naoya Horiguchi
2018-06-07  6:59                 ` Oscar Salvador
2018-06-07  9:49                   ` Oscar Salvador
2018-06-07 10:02                     ` Naoya Horiguchi
2018-06-11  9:05                       ` Naoya Horiguchi
2018-06-13  5:41                       ` [PATCH v1] mm: zero remaining unavailable struct pages (Re: kernel panic in reading /proc/kpageflags when enabling RAM-simulated PMEM) Naoya Horiguchi
2018-06-13  8:40                         ` Oscar Salvador
2018-06-14  4:56                           ` Naoya Horiguchi
2018-06-13  9:07                         ` Michal Hocko
2018-06-14  5:16                           ` Naoya Horiguchi
2018-06-14  5:38                             ` Oscar Salvador
2018-06-14  6:34                               ` [PATCH v2] x86/e820: put !E820_TYPE_RAM regions into memblock.reserved Naoya Horiguchi
2018-06-14  7:21                                 ` Oscar Salvador
2018-06-14 11:24                                   ` Oscar Salvador
2018-06-15  0:58                                     ` Naoya Horiguchi
2018-06-14 21:30                                 ` Oscar Salvador
2018-06-15  1:09                                   ` Naoya Horiguchi
2018-06-15  7:29                                     ` [PATCH v3] " Naoya Horiguchi
2018-06-15  8:41                                       ` Michal Hocko
2018-06-15 14:00                                         ` Pavel Tatashin [this message]
2018-06-15 14:10                                           ` Michal Hocko
2018-06-15 14:33                                           ` Oscar Salvador
2018-06-15 16:02                                             ` Pavel Tatashin
2018-06-18 23:36                                           ` Andrew Morton
2018-06-19  0:49                                             ` Pavel Tatashin
2018-07-02 20:05                                             ` Pavel Tatashin
2018-07-02 20:28                                               ` Andrew Morton
2018-07-02 20:31                                                 ` Pavel Tatashin
2018-06-14  7:00                             ` [PATCH v1] mm: zero remaining unavailable struct pages (Re: kernel panic in reading /proc/kpageflags when enabling RAM-simulated PMEM) Michal Hocko
2018-06-15  1:07                               ` Naoya Horiguchi
2018-06-15  8:39                                 ` Michal Hocko

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20180615140000.44tht4f3ek3lh2u2@xakep.localdomain \
    --to=pasha.tatashin@oracle.com \
    --cc=akpm@linux-foundation.org \
    --cc=dan.j.williams@intel.com \
    --cc=daniel.m.jordan@oracle.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=mhocko@kernel.org \
    --cc=mingo@kernel.org \
    --cc=n-horiguchi@ah.jp.nec.com \
    --cc=osalvador@suse.de \
    --cc=osalvador@techadventures.net \
    --cc=steven.sistare@oracle.com \
    --cc=willy@infradead.org \
    --cc=ying.huang@intel.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).