From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from cthulhu.engr.sgi.com (cthulhu.engr.sgi.com [192.26.80.2]) by neteng.engr.sgi.com (970321.SGI.8.8.5/960327.SGI.AUTOCF) via SMTP id QAA97432 for ; Tue, 7 Oct 1997 16:31:50 -0700 (PDT) Return-Path: Received: (from majordomo-owner@localhost) by cthulhu.engr.sgi.com (950413.SGI.8.6.12/960327.SGI.AUTOCF) id QAA17417 for linux-list; Tue, 7 Oct 1997 16:30:51 -0700 Received: from sgi.sgi.com (sgi.engr.sgi.com [192.26.80.37]) by cthulhu.engr.sgi.com (950413.SGI.8.6.12/960327.SGI.AUTOCF) via ESMTP id QAA17386 for ; Tue, 7 Oct 1997 16:30:49 -0700 Received: from dns.cobaltmicro.com ([209.19.61.1]) by sgi.sgi.com (950413.SGI.8.6.12/970507) via ESMTP id QAA06151 for ; Tue, 7 Oct 1997 16:30:33 -0700 env-from (ralf@mail2.cobaltmicro.com) Received: from dull.cobaltmicro.com (dull.cobaltmicro.com [209.19.61.35]) by dns.cobaltmicro.com (8.8.5/8.8.5) with ESMTP id QAA24302; Tue, 7 Oct 1997 16:30:14 -0700 From: Ralf Baechle Received: (from ralf@localhost) by dull.cobaltmicro.com (8.8.5/8.8.5) id QAA17826; Tue, 7 Oct 1997 16:28:08 -0700 Message-Id: <199710072328.QAA17826@dull.cobaltmicro.com> Subject: Re: More Linux/SGI status To: miguel@nuclecu.unam.mx (Miguel de Icaza) Date: Tue, 7 Oct 1997 16:28:07 -0700 (PDT) Cc: linux@cthulhu.engr.sgi.com, linux-mips@fnet.fr In-Reply-To: <199710072241.RAA01259@athena.nuclecu.unam.mx> from "Miguel de Icaza" at Oct 7, 97 05:41:38 pm Content-Type: text Sender: owner-linux@cthulhu.engr.sgi.com Precedence: bulk Hi, > Ralf's fixes to the dynamic linker fixed the Xt applications. Now, > only the PAM remains to be fixed. The PAM problem is also a dynamic linker problem. It's (I hope) one the last major showstopper class bugs that prevent just building a Linux/MIPS distribution from 95% unchanged source packages - and fullfilling all the promises to those loading hardware to me ... Btw, I stopped the mega load test (Running many compilers and tools in parallel) I posted recently after some more hours because the machine ran out of swap, some processes died therefore and load < 100 isn't that interesting ... System call latencies were pretty lousy bad for MIPS kernel because we were using a syscall handler written in C. I rewrote the entire thing in assembler and now we're as fast as a MMX Pentium accounting for the different clock rates. Pretty nice, because we actually have to save/restore many more registers and deal with a more complex system call convention. Seven nops left to eleminate ... > So, where do we stand now? So, the only bits missing now are: > > 1. me finishing the support for the X server. just a hack > here and there. > > 1.b. Getting the other important bits of the RRM code in the > kernel. > > 2. me fixing the mouse. > > 3. Would it be possible to negotiate with SGI management > the posibility of shipping the IRIX runtime libraries and > the X server as found on IRIX with a Linux disrtibution? Bad point in time, it has been pointed out to me that SGI's stock was doing pretty bad ;-) > 4. A nice, easy-to use install program. Taking the existing > Red Hat/Mustand install program and port it should be > pretty easy. Ralf From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: From: Ralf Baechle Message-ID: <199710072328.QAA17826@dull.cobaltmicro.com> Subject: Re: More Linux/SGI status Date: Tue, 7 Oct 1997 16:28:07 -0700 (PDT) In-Reply-To: <199710072241.RAA01259@athena.nuclecu.unam.mx> from "Miguel de Icaza" at Oct 7, 97 05:41:38 pm Content-Type: text Sender: owner-linux@cthulhu.engr.sgi.com To: Miguel de Icaza Cc: linux@cthulhu.engr.sgi.com, linux-mips@fnet.fr Message-ID: <19971007232807.tudnsE6eaFVgWZgowTmBFvsqBhTFbUYKVczwDZJlqBs@z> Hi, > Ralf's fixes to the dynamic linker fixed the Xt applications. Now, > only the PAM remains to be fixed. The PAM problem is also a dynamic linker problem. It's (I hope) one the last major showstopper class bugs that prevent just building a Linux/MIPS distribution from 95% unchanged source packages - and fullfilling all the promises to those loading hardware to me ... Btw, I stopped the mega load test (Running many compilers and tools in parallel) I posted recently after some more hours because the machine ran out of swap, some processes died therefore and load < 100 isn't that interesting ... System call latencies were pretty lousy bad for MIPS kernel because we were using a syscall handler written in C. I rewrote the entire thing in assembler and now we're as fast as a MMX Pentium accounting for the different clock rates. Pretty nice, because we actually have to save/restore many more registers and deal with a more complex system call convention. Seven nops left to eleminate ... > So, where do we stand now? So, the only bits missing now are: > > 1. me finishing the support for the X server. just a hack > here and there. > > 1.b. Getting the other important bits of the RRM code in the > kernel. > > 2. me fixing the mouse. > > 3. Would it be possible to negotiate with SGI management > the posibility of shipping the IRIX runtime libraries and > the X server as found on IRIX with a Linux disrtibution? Bad point in time, it has been pointed out to me that SGI's stock was doing pretty bad ;-) > 4. A nice, easy-to use install program. Taking the existing > Red Hat/Mustand install program and port it should be > pretty easy. Ralf