[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Minios-devel] [PATCH v3 20/43] mini-os: implement the memmove/memchr
- To: Huang Shijie <shijie.huang@xxxxxxx>
- From: Julien Grall <julien.grall@xxxxxxx>
- Date: Thu, 26 Apr 2018 09:49:20 +0100
- Cc: jgross@xxxxxxxx, wei.liu2@xxxxxxxxxx, wei.chen@xxxxxxx, steve.capper@xxxxxxx, vlad.babchuk@xxxxxxxxx, minios-devel@xxxxxxxxxxxxxxxxxxxx, kaly.xin@xxxxxxx, samuel.thibault@xxxxxxxxxxxx, baozich@xxxxxxxxx, nd@xxxxxxx
- Delivery-date: Thu, 26 Apr 2018 08:49:29 +0000
- List-id: Mini-os development list <minios-devel.lists.xenproject.org>
On 26/04/18 09:42, Huang Shijie wrote:
On Wed, Apr 25, 2018 at 11:09:43AM +0100, Julien Grall wrote:
On 25/04/18 10:37, Huang Shijie wrote:
On Wed, Apr 25, 2018 at 10:10:58AM +0100, Julien Grall wrote:
On 04/16/2018 07:32 AM, Huang Shijie wrote:
This patch is split from the Chen Baozi's old patch:
"This activates the ARM code added in the previous patches. On
ARM, Mini-OS will boot and display some output on the console. Tested with:"
Note: The code is copied from FreeBSD code:
freebsd/contrib/ldns/compat/memmove.c
That's only for memmove. How about memchr?
The memchr is also from the FreeBSD code...
You specify the file for memmove.c. But not memchr.
Also, this seems to come from a userspace application and not the kernel.
Why not importing the assembly version (memmove.S and memcpy.S) that has
been optimized for aarch64?
okay. I can use the assembly version..
I was asking the rationale behind using userspace application memcpy
over the assembly one.
Cheers,
--
Julien Grall
_______________________________________________
Minios-devel mailing list
Minios-devel@xxxxxxxxxxxxxxxxxxxx
https://lists.xenproject.org/mailman/listinfo/minios-devel
|