Re: [RFC PATCH v2] mm: initialize struct pages reserved by ZONE_DEVICE driver.

[David Hildenbrand]

On 09.09.19 07:48, Toshiki Fukasawa wrote:
> On 2019/09/06 19:35, David Hildenbrand wrote:
>> On 06.09.19 12:02, Toshiki Fukasawa wrote:
>>> Thank you for your feedback.
>>>
>>> On 2019/09/06 17:45, David Hildenbrand wrote:
>>>> On 06.09.19 10:09, Toshiki Fukasawa wrote:
>>>>> A kernel panic is observed during reading
>>>>> /proc/kpage{cgroup,count,flags} for first few pfns allocated by
>>>>> pmem namespace:
>>>>>
>>>>> BUG: unable to handle page fault for address: fffffffffffffffe
>>>>> [  114.495280] #PF: supervisor read access in kernel mode
>>>>> [  114.495738] #PF: error_code(0x0000) - not-present page
>>>>> [  114.496203] PGD 17120e067 P4D 17120e067 PUD 171210067 PMD 0
>>>>> [  114.496713] Oops: 0000 [#1] SMP PTI
>>>>> [  114.497037] CPU: 9 PID: 1202 Comm: page-types Not tainted 5.3.0-rc1
>>>>> [  114.497621] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.11.0-0-g63451fca13-prebuilt.qemu-project.org 04/01/2014
>>>>> [  114.498706] RIP: 0010:stable_page_flags+0x27/0x3f0
>>>>> [  114.499142] Code: 82 66 90 66 66 66 66 90 48 85 ff 0f 84 d1 03 00 00 41 54 55 48 89 fd 53 48 8b 57 08 48 8b 1f 48 8d 42 ff 83 e2 01 48 0f 44 c7 <48> 8b 00 f6 c4 02 0f 84 57 03 00 00 45 31 e4 48 8b 55 08 48 89 ef
>>>>> [  114.500788] RSP: 0018:ffffa5e601a0fe60 EFLAGS: 00010202
>>>>> [  114.501373] RAX: fffffffffffffffe RBX: ffffffffffffffff RCX: 0000000000000000
>>>>> [  114.502009] RDX: 0000000000000001 RSI: 00007ffca13a7310 RDI: ffffd07489000000
>>>>> [  114.502637] RBP: ffffd07489000000 R08: 0000000000000001 R09: 0000000000000000
>>>>> [  114.503270] R10: 0000000000000000 R11: 0000000000000000 R12: 0000000000240000
>>>>> [  114.503896] R13: 0000000000080000 R14: 00007ffca13a7310 R15: ffffa5e601a0ff08
>>>>> [  114.504530] FS:  00007f0266c7f540(0000) GS:ffff962dbbac0000(0000) knlGS:0000000000000000
>>>>> [  114.505245] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
>>>>> [  114.505754] CR2: fffffffffffffffe CR3: 000000023a204000 CR4: 00000000000006e0
>>>>> [  114.506401] Call Trace:
>>>>> [  114.506660]  kpageflags_read+0xb1/0x130
>>>>> [  114.507051]  proc_reg_read+0x39/0x60
>>>>> [  114.507387]  vfs_read+0x8a/0x140
>>>>> [  114.507686]  ksys_pread64+0x61/0xa0
>>>>> [  114.508021]  do_syscall_64+0x5f/0x1a0
>>>>> [  114.508372]  entry_SYSCALL_64_after_hwframe+0x44/0xa9
>>>>> [  114.508844] RIP: 0033:0x7f0266ba426b
>>>>>
>>>>> The first few pages of ZONE_DEVICE expressed as the range
>>>>> (altmap->base_pfn) to (altmap->base_pfn + altmap->reserve) are
>>>>> skipped by struct page initialization. Some pfn walkers like
>>>>> /proc/kpage{cgroup, count, flags} can't handle these uninitialized
>>>>> struct pages, which causes the error.
>>>>>
>>>>> In previous discussion, Dan seemed to have concern that the struct
>>>>> page area of some pages indicated by vmem_altmap->reserve may not
>>>>> be allocated. (See https://lore.kernel.org/lkml/CAPcyv4i5FjTOnPbXNcTzvt+e6RQYow0JRQwSFuxaa62LSuvzHQ@xxxxxxxxxxxxxx/)
>>>>> However, arch_add_memory() called by devm_memremap_pages() allocates
>>>>> struct page area for pages containing addresses in the range
>>>>> (res.start) to (res.start + resource_size(res)), which include the
>>>>> pages indicated by vmem_altmap->reserve. If I read correctly, it is
>>>>> allocated as requested at least on x86_64. Also, memmap_init_zone()
>>>>> initializes struct pages in the same range.
>>>>> So I think the struct pages should be initialized.>
>>>>
>>>> For !ZONE_DEVICE memory, the memmap is valid with SECTION_IS_ONLINE -
>>>> for the whole section. For ZONE_DEVICE memory we have no such
>>>> indication. In any section that is !SECTION_IS_ONLINE and
>>>> SECTION_MARKED_PRESENT, we could have any subsections initialized. >
>>>> The only indication I am aware of is pfn_zone_device_reserved() - which
>>>> seems to check exactly what you are trying to skip here.
>>>>
>>>> Can't you somehow use pfn_zone_device_reserved() ? Or if you considered
>>>> that already, why did you decide against it?
>>>
>>> No, in current approach this function is no longer needed.
>>> The reason why we change the approach is that all pfn walkers
>>> have to be aware of the uninitialized struct pages.
>>
>> We should use the same strategy for all pfn walkers then (effectively
>> get rid of pfn_zone_device_reserved() if that's what we want).
> 
> True, but this patch replaces "/proc/kpageflags: do not use uninitialized
> struct pages". If we initialize the uninitialized struct pages, no pfn walker
> will need to be aware of them.

So the function should go.

> 
>>
>>>
>>> As for SECTION_IS_ONLINE, I'm not sure now.
>>> I will look into it next week.
>>
>> SECTION_IS_ONLINE does currently not apply to ZONE_DEVICE and due to
>> sub-section support for ZONE_DEVICE, it cannot easily be reused.
>>
> It seems that SECTION_IS_ONLINE and SECTION_MARKED_PRESENT can be used to
> distinguish uninitialized struct pages if we can apply them to ZONE_DEVICE,
> but that is no longer necessary with this approach.

Let's take a step back here to understand the issues I am aware of. I
think we should solve this for good now:

A PFN walker takes a look at a random PFN at a random point in time. It
finds a PFN with SECTION_MARKED_PRESENT && !SECTION_IS_ONLINE. The
options are:

1. It is buddy memory (add_memory()) that has not been online yet. The
memmap contains garbage. Don't access.

2. It is ZONE_DEVICE memory with a valid memmap. Access it.

3. It is ZONE_DEVICE memory with an invalid memmap, because the section
is only partially present: E.g., device starts at offset 64MB within a
section or the device ends at offset 64MB within a section. Don't access it.

4. It is ZONE_DEVICE memory with an invalid memmap, because the memmap
was not initialized yet. memmap_init_zone_device() did not yet succeed
after dropping the mem_hotplug lock in mm/memremap.c. Don't access it.

5. It is reserved ZONE_DEVICE memory ("pages mapped, but reserved for
driver") with an invalid memmap. Don't access it.

I can see that your patch tries to make #5 vanish by initializing the
memmap, fair enough. #3 and #4 can't be detected. The PFN walker could
still stumble over uninitialized memmaps.

-- 

Thanks,

David / dhildenb