[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Re: l4-ka, c++!
From: |
swin |
Subject: |
Re: Re: l4-ka, c++! |
Date: |
Mon, 10 Nov 2003 07:47:20 +0800 |
But do you think with the using of C++, it can make
the object file or executable file bigger while comparing
with c. Is l4-ka so?
======= 2003-11-08 10:40:34 您在来信中写道:=======
>s> I have heard that hurd will discontinue current
>s> hurd-l4, and switch to l4-ka,but l4-ka is written in
>s> c++, maybe, for developing and researching ,it's
>s> convenient, but for a practical OS, do you think it
>s> will degrade efficiency?
>s> Any idea?
>
>First off, baseing an OS on a c++ micro kernel, does not mean that
>you have to write the (full) OS in c++ as well.
>
>Then, what do you mean by efficiency:
>1) Code maintance is more efficient with OO (they say).
>2) Object code execution efficency for OO (in contrast to plain C -
> WYCIWYG) depends on the compiler. If you manage to compile your OS
> on the intel compiler for instance, I think you will get pretty good
> efficiency on a p4 e.g. (gcc has improved as well, afaik)
>3) Driver support and porting efforts will suffer, however, as you
> will have to touch every and any driver (be it from Linux, *BSD or
> wheresoever) and punch it into an object API. Still, once you have
> done the task, you can maintain your drivers and hardware res. more
> cleanly (a well designed API provided).
>
>And last of all, l4 == l4-ka (or am I just missing some information
>here?)
>
>--
>Best regards,
> Max mailto:address@hidden
>
>.
= = = = = = = = = = = = = = = = = = = =
致
礼!
swin
address@hidden
2003-11-10
- l4-ka, c++!, swin, 2003/11/07
- Re: l4-ka, c++!, Max Laier, 2003/11/07
- Re: l4-ka, c++!, Marcus Brinkmann, 2003/11/08
- Re: l4-ka, c++!, Niels Möller, 2003/11/08
- Re: l4-ka, c++!, Jeff Bailey, 2003/11/08