[PATCH, RFC] um: remove set_fs

[Anton Ivanov]

On 23/08/2021 08:04, Christoph Hellwig wrote:
> Richard, Anton - any comments?
> 
> On Fri, Jul 09, 2021 at 04:47:34PM +0200, Johannes Berg wrote:
>> On Fri, 2021-07-09 at 15:38 +0200, Christoph Hellwig wrote:
>>> Remove address space overrides using set_fs() for user mode Linux.
>>>
>>> Compile tested only.
>>
>> Seems to survive my normal (wireless related) tests :)
>>
>> But I don't really understand what any of it does, so ...
>>
>> In particular, why no pagefaults for kernel addresses?
>>
>> johannes
> ---end quoted text---
> 
> _______________________________________________
> linux-um mailing list
> linux-um at lists.infradead.org
> http://lists.infradead.org/mailman/listinfo/linux-um
> 

In first read - you do not want to do that.

buffer_op is slow (even with all the tweaks we have done to that). It is one of the reasons UML userspace is slower than it should be.

The primary reason for that is that it never copies more than a page at a time and pages in/out a page at a time.

We retain reasonable kernel speed because we bypass it for kernel - "if (uaccess_kernel())".

Unless I am missing something, this change will use the slow path currently used for userspace for the kernel.

-- 
Anton R. Ivanov
Cambridgegreys Limited. Registered in England. Company Number 10273661
https://www.cambridgegreys.com/