Date: Tue, 15 Feb 2000 13:46:55 +1100 From: Peter Jeremy <peter.jeremy@alcatel.com.au> To: obrien@NUXI.com Cc: "freebsd-alpha@FreeBSD.ORG" <freebsd-alpha@FreeBSD.ORG> Subject: Re: Alpha kernels Message-ID: <00Feb15.134656est.115251@border.alcanet.com.au> In-Reply-To: <20000214132553.A17797@dragon.nuxi.com>; from obrien@NUXI.com on Tue, Feb 15, 2000 at 08:27:47AM %2B1100 References: <38A85C2C.C66243B8@getrelevant.com> <20000214132553.A17797@dragon.nuxi.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On 2000-Feb-15 08:27:47 +1100, David O'Brien <obrien@NUXI.com> wrote: >Yes this is a RISC vs. CISC thing. Though in my experience, i386 code tends to quite sparse for a CISC architecture. Intel propaganda notwithstanding, the lack of 16-bit offsets seems to make real-world i386 code significantly larger than say the M68K. > Plus the 64bit vs. 32bit ints also contributes some to >the larger size. My guess is that this is the major contributing factor. I don't know how much effort (if any) has been applied to removing padding from kernel data structures when compiled with 64-bit longs/pointers. In any case, the FreeBSD kernel is still small compared to Digital UNIX: # size /vmunix text data bss dec hex 4420416 747008 1201440 6368864 612e60 In the case of kernels, the size is also affected by the complexity of things like interrupt handling, the MMU and device drivers. The SRM gives the Alpha an advantage here. Peter To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?00Feb15.134656est.115251>