[Top] [All Lists]

Re: Almost got to a patchable stage...

To: linux-mips
Subject: Re: Almost got to a patchable stage...
From: "David S. Miller" <>
Date: Fri, 11 Apr 1997 03:37:25 -0400
In-reply-to: <> (
   From: (Paul Antoine)
   Date: Fri, 11 Apr 1997 17:21:52 +1000 (EST)

   I've been working on the tree, as I felt it was probably
   best for me to organise the DECstation patches against the next
   kernel you will release (and a buggy ELF loader isn't an issue for
   me at the moment).  I have found a few things... but the one you
   probably need to look at is a means for setting the mips
   architecture level for compilation in the main arch/mips Makefile
   now that you appear to have done away with the mips1/2/3
   directories, and the CONFIG_CPU_R... variable.

The whole mips[1234] ordeal is the one thing which upset me greatly
about the MIPS processor line, this is the single one thing which
makes a true "all in one" MIPS/Linux kernel damn near impossible.

In fact I did most of the work in my SGI tree to support multiple
cache/mmu architectures in a single kernel, and I have pulled it off
already in the SparcLinux kernel ;-)  Ralf is merging in my code, and
with that it should be possible to do an "all in one" for all R4k and
later MIPS's.

I did ponder for a bit and eventually did come up with a scheme
whereby you could even do a single kernel on the Sparc which supported
both multiple PAGE_SIZE values and 32-bit/64-bit all at once, don't
try this at home.  It would have been extremely bulky and hard to make
run as fast as I would like so I never implemented it...

Yow! 11.26 MB/s remote host TCP bandwidth & ////
199 usec remote TCP latency over 100Mb/s   ////
ethernet.  Beat that!                     ////
-----------------------------------------////__________  o
David S. Miller, /_____________/ / // /_/ ><

<Prev in Thread] Current Thread [Next in Thread>