Re: [Xen-devel] [PATCH v2 00/12] Alternate p2m: support multiple copies of host p2m

[Ed White <edmund.h.white@xxxxxxxxx>]

On 06/25/2015 02:00 AM, Andrew Cooper wrote:
> On 24/06/15 23:55, Ed White wrote:
>> On 06/24/2015 03:45 PM, Lengyel, Tamas wrote:
>>> On Wed, Jun 24, 2015 at 6:02 PM, Ed White <edmund.h.white@xxxxxxxxx> wrote:
>>>
>>>> On 06/24/2015 02:34 PM, Lengyel, Tamas wrote:
>>>>> Hi Ed,
>>>>> I tried the system using memsharing and I collected the following crash
>>>>> log. In this test I ran memsharing on all pages of the domain before
>>>>> activating altp2m and creating the view. Afterwards I used my updated
>>>>> xen-access to create a copy of this p2m with only R/X permissions. The
>>>> idea
>>>>> would be that the altp2m view remains completely shared, while the
>>>> hostp2m
>>>>> would be able to do its CoW propagation as the domain is executing.
>>>>>
>>>>> (XEN) mm locking order violation: 278 > 239
>>>>> (XEN) Xen BUG at mm-locks.h:68
>>>>> (XEN) ----[ Xen-4.6-unstable  x86_64  debug=y  Tainted:    C ]----
>>>>> (XEN) CPU:    2
>>>>> (XEN) RIP:    e008:[<ffff82d0801f8768>]
>>>>> p2m_altp2m_propagate_change+0x85/0x4a9
>>>>> (XEN) RFLAGS: 0000000000010282   CONTEXT: hypervisor (d6v0)
>>>>> (XEN) rax: 0000000000000000   rbx: 0000000000000000   rcx:
>>>> 0000000000000000
>>>>> (XEN) rdx: ffff8302163a8000   rsi: 000000000000000a   rdi:
>>>> ffff82d0802a069c
>>>>> (XEN) rbp: ffff8302163afa68   rsp: ffff8302163af9e8   r8:
>>>> ffff83021c000000
>>>>> (XEN) r9:  0000000000000003   r10: 00000000000000ef   r11:
>>>> 0000000000000003
>>>>> (XEN) r12: ffff83010cc51820   r13: 0000000000000000   r14:
>>>> ffff830158d90000
>>>>> (XEN) r15: 0000000000025697   cr0: 0000000080050033   cr4:
>>>> 00000000001526f0
>>>>> (XEN) cr3: 00000000dbba3000   cr2: 00000000778c9714
>>>>> (XEN) ds: 0000   es: 0000   fs: 0000   gs: 0000   ss: 0000   cs: e008
>>>>> (XEN) Xen stack trace from rsp=ffff8302163af9e8:
>>>>> (XEN)    ffff8302163af9f8 00000000803180f8 000000000000000c
>>>> ffff82d0801892ee
>>>>> (XEN)    ffff82d0801fb4d1 ffff83010cc51de0 000000000008ff49
>>>> ffff82d08012f86a
>>>>> (XEN)    ffff83010cc51820 ffff83010cc51820 0000000000000000
>>>> 0000000000000000
>>>>> (XEN)    ffff83010cc51820 0000000000000000 ffff8300dbb334b8
>>>> ffff8302163afa00
>>>>> (XEN)    ffff8302163afb18 ffff82d0801fd549 0000000500000009
>>>> ffff830200000001
>>>>> (XEN)    0000000000000001 ffff830158d90000 0000000000000002
>>>> 000000000008ff49
>>>>> (XEN)    0000000000025697 000000000000000c ffff8302163afae8
>>>> 80c000008ff49175
>>>>> (XEN)    80c00000d0a97175 01ff83010cc51820 0000000000000097
>>>> ffff8300dbb33000
>>>>> (XEN)    ffff8302163afb78 000000000008ff49 0000000000000000
>>>> 0000000000000001
>>>>> (XEN)    0000000000025697 ffff83010cc51820 ffff8302163afb38
>>>> ffff82d0801fd644
>>>>> (XEN)    ffffffffffffffff 00000000000d0a97 ffff8302163afb98
>>>> ffff82d0801f23c5
>>>>> (XEN)    ffff830158d90000 000000000cc51820 ffff830158d90000
>>>> 000000000000000c
>>>>> (XEN)    000000000008ff49 ffff83010cc51820 0000000000025697
>>>> 00000000000d0a97
>>>>> (XEN)    000000000008ff49 ffff830158d90000 ffff8302163afbd8
>>>> ffff82d0801f45c8
>>>>> (XEN)    ffff83010cc51820 000000000000000c ffff83008fd41170
>>>> 000000000008ff49
>>>>> (XEN)    0000000000025697 ffff82e001a152e0 ffff8302163afc58
>>>> ffff82d080205b51
>>>>> (XEN)    0000000000000009 000000000008ff49 ffff8300d0a97000
>>>> ffff83008fd41160
>>>>> (XEN)    ffff82e001a152f0 ffff82e0011fe920 ffff83010cc51820
>>>> 0000000c00000000
>>>>> (XEN)    0000000000025697 0000000000000003 ffff83010cc51820
>>>> ffff8302163afd34
>>>>> (XEN)    0000000000025697 0000000000000000 ffff8302163afca8
>>>> ffff82d0801f1f7d
>>>>> (XEN) Xen call trace:
>>>>> (XEN)    [<ffff82d0801f8768>] p2m_altp2m_propagate_change+0x85/0x4a9
>>>>> (XEN)    [<ffff82d0801fd549>] ept_set_entry_sve+0x5fa/0x6e6
>>>>> (XEN)    [<ffff82d0801fd644>] ept_set_entry+0xf/0x11
>>>>> (XEN)    [<ffff82d0801f23c5>] p2m_set_entry+0xd4/0x112
>>>>> (XEN)    [<ffff82d0801f45c8>] set_shared_p2m_entry+0x2d0/0x39b
>>>>> (XEN)    [<ffff82d080205b51>] __mem_sharing_unshare_page+0x83f/0xbd6
>>>>> (XEN)    [<ffff82d0801f1f7d>] __get_gfn_type_access+0x224/0x2b0
>>>>> (XEN)    [<ffff82d0801c6df5>] hvm_hap_nested_page_fault+0x21f/0x795
>>>>> (XEN)    [<ffff82d0801e86ae>] vmx_vmexit_handler+0x1764/0x1af3
>>>>> (XEN)    [<ffff82d0801ee891>] vmx_asm_vmexit_handler+0x41/0xc0
>>>> The crash here is because I haven't successfully forced all the shared
>>>> pages in the host p2m to become unshared before copying,
>>>> which is the intended behaviour.
>>>>
>>>> I think I know how that has happened and how to fix it, but what you're
>>>> trying to do won't work by design. By the time a copy from host p2m to
>>>> altp2m occurs, the sharing is supposed to be broken.
>>>>
>>> Hm. If the sharing gets broken before the hostp2m->altp2m copy, maybe doing
>>> sharing after the view has been created is a better route? I guess the
>>> sharing code would need to be adapted to check if altp2m is enabled for
>>> that to work..
>>>
>>>
>>>> You're coming up with some ways of attempting to use altp2m that we
>>>> hadn't thought of. That's a good thing, and just what we want, but
>>>> there are limits to what we can support without more far-reaching
>>>> changes to existing parts of Xen. This isn't going to be do-able for
>>>> 4.6.
>>>>
>>> My main concern is just getting it to work, hitting 4.6 is not a priority.
>>> I understand that my stuff is highly experimental ;) While the gfn
>>> remapping feature is intriguing, in my setup I already have a copy of the
>>> page I would want to present during a singlestep-altp2mswitch - in the
>>> origin domains memory. AFAIU the gfn remapping would work only within the
>>> domains existing p2m space.
>> Understood, but for us hitting 4.6 with the initial version of altp2m
>> is *the* priority. And yes, remapping is restricted to pages from the
>> same host p2m.
> 
> It is fine for experimental features to have known interaction issues. 
> I don't necessarily see this as a blocker to 4.6, although it would
> indeed be better if it could be fixed in time.

I plan to fix the bug, such that unshare will always occur before a copy.
I don't plan to make the altp2m's able to have shared pages.

Ed


_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel