Re: [PATCH -next] mm: usercopy: add a debugfs interface to bypass the vmalloc check.

[Kefeng Wang <wangkefeng.wang@huawei.com>]

On 2024/12/3 21:10, zuoze wrote:
> 
> 
> 在 2024/12/3 20:39, Uladzislau Rezki 写道:
>> On Tue, Dec 03, 2024 at 07:23:44PM +0800, zuoze wrote:
>>> We have implemented host-guest communication based on the TUN device
>>> using XSK[1]. The hardware is a Kunpeng 920 machine (ARM architecture),
>>> and the operating system is based on the 6.6 LTS version with kernel
>>> version 6.6. The specific stack for hotspot collection is as follows:
>>>
>>> -  100.00%     0.00%  vhost-12384  [unknown]      [k] 0000000000000000
>>>     - ret_from_fork
>>>        - 99.99% vhost_task_fn
>>>           - 99.98% 0xffffdc59f619876c
>>>              - 98.99% handle_rx_kick
>>>                 - 98.94% handle_rx
>>>                    - 94.92% tun_recvmsg
>>>                       - 94.76% tun_do_read
>>>                          - 94.62% tun_put_user_xdp_zc
>>>                             - 63.53% __check_object_size
>>>                                - 63.49% __check_object_size.part.0
>>>                                     find_vmap_area
>>>                             - 30.02% _copy_to_iter
>>>                                  __arch_copy_to_user
>>>                    - 2.27% get_rx_bufs
>>>                       - 2.12% vhost_get_vq_desc
>>>                            1.49% __arch_copy_from_user
>>>                    - 0.89% peek_head_len
>>>                         0.54% xsk_tx_peek_desc
>>>                    - 0.68% vhost_add_used_and_signal_n
>>>                       - 0.53% eventfd_signal
>>>                            eventfd_signal_mask
>>>              - 0.94% handle_tx_kick
>>>                 - 0.94% handle_tx
>>>                    - handle_tx_copy
>>>                       - 0.59% vhost_tx_batch.constprop.0
>>>                            0.52% tun_sendmsg
>>>
>>> It can be observed that most of the overhead is concentrated in the
>>> find_vmap_area function.
>>>
>> I see. Yes, it is pretty contented, since you run the v6.6 kernel. There
>> was a work that tends to improve it to mitigate a vmap lock contention.
>> See it here: https://lwn.net/Articles/956590/
>>
>> The work was taken in the v6.9 kernel:
>>
>> <snip>
>> commit 38f6b9af04c4b79f81b3c2a0f76d1de94b78d7bc
>> Author: Uladzislau Rezki (Sony) <urezki@gmail.com>
>> Date:   Tue Jan 2 19:46:23 2024 +0100
>>
>>      mm: vmalloc: add va_alloc() helper
>>
>>      Patch series "Mitigate a vmap lock contention", v3.
>>
>>      1. Motivation
>> ...
>> <snip>
>>
>> Could you please try the v6.9 kernel on your setup?
>>
>> How to solve it, probably, it can be back-ported to the v6.6 kernel.
> 
> All the vmalloc-related optimizations have already been merged into 6.6,
> including the set of optimization patches you suggested. Thank you very
> much for your input.
> 

It is unclear, we have backported the vmalloc optimization into our 6.6
kernel before, so the above stack already with those patches and even
with those optimization, the find_vmap_area() is still the hotpots.