答复: About the system call named "sys_mount".
付新荣
FuXinRong at skyworth.com
Tue Oct 27 00:05:42 EDT 2009
the"c2" register of CP15 isn't changed within the execution of "SWI" instruction.
in a other words, the page table isn't changed.
the reference of the user space data is possible at this time.
but i worry there are some issues when the data is swaped.
-----邮件原件-----
发件人: Russell King - ARM Linux [mailto:linux at arm.linux.org.uk]
发送时间: 2009年10月19日 18:22
收件人: 付新荣
抄送: linux-arm-kernel at lists.infradead.org
主题: Re: About the system call named "sys_mount".
On Mon, Oct 19, 2009 at 05:57:21PM +0800, 付新荣 wrote:
> if there is no potential risk, the speed of "mount" command would be
> improved greatly. Tell me plz if there is problem. thax.
1. You're risking a kernel oops, caused by direct dereferencing of
user pointers
2. The cost of copying parameters in sys_mount has nothing to do with
the speed - the biggest issue will be accessing the media itself.
More information about the linux-arm-kernel
mailing list