From ryan@ravage.labs.gmu.edu Wed Apr 2 01:07:59 1997 Received-Date: Wed, 2 Apr 1997 01:07:58 +0200 (MET DST) Received: from ravage.labs.gmu.edu ([129.174.166.55]) by Utopia.EUnet.fr (5.65c8d/EUNET-4.2.3) via EUnet-France id AA05282; Wed, 2 Apr 1997 01:05:14 +0200 (MET) Received: from localhost (ryan@localhost) by ravage.labs.gmu.edu (8.8.3/8.8.3) with SMTP id SAA11027 for ; Mon, 1 Apr 1996 18:05:29 -0500 Date: Mon, 1 Apr 1996 18:05:29 -0500 (EST) From: Ryan Rafferty To: linux-mips Subject: r10000 boards In-Reply-To: <199704011711.JAA00570@scruz.net> Message-Id: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Content-Length: 654 Lines: 15 After seeing the MIPS-based boards currently being produced, it appears that all the manufacturers making MIPS boards are using a processor based on the R4000 architecture. Is there anyone other than SGI making R10000 boards? Another stupid question--why has secondary cache been overlooked in the MIPS world? My own Magnum is of the 4000PC variety--that is, no secondary cache; it seems counterintuitive to me for manufacturers to build a solid motherboard based on the MIPS architecture but then skip out on a relatively cheap component like cache which goes such a long way to increase overall system performance. Is mise le meas, Ryan Rafferty From ryan@ravage.labs.gmu.edu Wed Apr 2 02:52:28 1997 Received-Date: Wed, 2 Apr 1997 02:52:28 +0200 (MET DST) Received: from ravage.labs.gmu.edu ([129.174.166.55]) by Utopia.EUnet.fr (5.65c8d/EUNET-4.2.3) via EUnet-France id AA07649; Wed, 2 Apr 1997 02:49:40 +0200 (MET) Received: from localhost (ryan@localhost) by ravage.labs.gmu.edu (8.8.3/8.8.3) with SMTP id TAA11598 for ; Mon, 1 Apr 1996 19:49:54 -0500 Date: Mon, 1 Apr 1996 19:49:54 -0500 (EST) From: Ryan Rafferty To: linux-mips Subject: Re: r10000 boards In-Reply-To: <199704020000.CAA05217@informatik.uni-koblenz.de> Message-Id: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Content-Length: 695 Lines: 20 On Wed, 2 Apr 1997, Systemkennung Linux wrote: > A good RAM interface is a much more effective mean to accelerate a system than > caches especially when you have an application that has a working set that > exceeds the cache size. The Magnum's RAM interface may be slow by today's > standards but it was very fast at it's time. Ok, that's cool. But then why do computer manufacturers still incorporate cache into today's machines instead of using blazing memory interfaces? For example, the fastest Alpha computers today have 8 megs of static RAM cache. If a faster memory interface would have eliminated the need for it, why didn't they use it? Thanks for your reply, Ryan > Ralf > From davem@caip.rutgers.edu Thu Apr 4 11:57:12 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id LAA00937 for ; Thu, 4 Apr 1996 11:57:11 +0200 (MET DST) Received: from caipfs.rutgers.edu by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA14946; Thu, 4 Apr 1996 05:45:16 +0200 (MET) Received: from huahaga.rutgers.edu (huahaga.rutgers.edu [128.6.155.53]) by caipfs.rutgers.edu (8.6.9+bestmx+oldruq+newsunq+grosshack/8.6.9) with ESMTP id WAA08449; Wed, 3 Apr 1996 22:40:44 -0500 Received: (davem@localhost) by huahaga.rutgers.edu (8.6.9+bestmx+oldruq+newsunq+grosshack/8.6.9) id WAA24535; Wed, 3 Apr 1996 22:40:44 -0500 Date: Wed, 3 Apr 1996 22:40:44 -0500 Message-Id: <199604040340.WAA24535@huahaga.rutgers.edu> From: "David S. Miller" To: miguel@roxanne.nuclecu.unam.mx Cc: sparclinux-cvs@caipfs.rutgers.edu, linux-kernel@vger.rutgers.edu, sparclinux@vger.rutgers.edu, hackers@cafe.anu.edu.au, linux-arm@vger.rutgers.edu, linux-alpha@vger.rutgers.edu, linux-mips@fnet.fr In-Reply-To: <199604040253.UAA23275@roxanne.nuclecu.unam.mx> (message from Miguel de Icaza on Wed, 3 Apr 1996 20:53:20 -0600) Subject: Re: CVS Update@vger.rutgers.edu: linux Content-Length: 5225 Lines: 130 Date: Wed, 3 Apr 1996 20:53:20 -0600 From: Miguel de Icaza I took a quick glance at the patch, and I think that a bunch of people would benefit if you tell us (with a cc to linux-kernel) what is the general idea behind the patch (it's difficult to see the exact context in some places of what is being done). Right you are. Here is what is going on in the 1.3.83 changes I made. First of all, the name invalidate() itself doesn't mean very much. And if it means anyone to someone, it only makes sense if your tlb and cache architectures are coherent with each other, the i386 and alpha (???) are like this so the traditional naming scheme works ok. But on most RISC based cache/mmu architectures it makes more sense to split up the operations in two (and as you will see later on, there is a third operation needed). flush_cache_*(); change_page_tables(); flush_tlb_*(); Where '*' describes the extent to which page tables are changing. We keep the old granularity levels for the new scheme: a) all b) mm c) range d) page See, on many machines, the cache and the tlb are seperate entities and have different rules for getting rid of old stale data. As an example I will show how this all functions on the HyperSparc MBUS module. The HyperSparc has a VIPT (virtually indexed, physically tagged) level 2 cache, it also has an on-chip ICACHE (instruction cache). The level 2 cache can be either 128k or 256k in size, the ICACHE is 8k in size. Flushing the HyperSparc cache is a process where one must be careful, in order to flush properly the chip must make sure that it doesn't flush a cache line that supervisor software isn't asking to be flushed. Therefore, a real page translation can be initiated for the check for a tag match during a cache flush (the tags a physical remember). So if we did something like: change_page_tables(); flush_cache_*(); whoope... we just got rid of the mapping that the chip needs to check for a flush cache-tag match and the processor will take a fault due to the missed translation. This is undesirable. There are two more elements to fully support the vast array of cache architecture out there. I've only partially added one of them at this time. It is for the benefit of copy-back style caches (I use the hypersparc in copy-back mode for SMP kernels, the sun4 cache architecture is also copy-back in nature). Consider the following section of code during a copy-on-write fault: memory.c:do_wp_page() copy_page(old_page,new_page); set_pte(page_table, pte_mkwrite(pte_mkdirty(mk_pte(new_page, vma->vm_page_prot)))); free_page(old_page); invalidate_page(vma, address); whoops, we have a lot of problems going on here, first the aforementioned necessity to flush the cache before changing the page tables, let's fix that first: copy_page(old_page,new_page); flush_cache_page(vma, address); set_pte(page_table, pte_mkwrite(pte_mkdirty(mk_pte(new_page, vma->vm_page_prot)))); free_page(old_page); flush_tlb_page(vma, address); Ok, much better. There is still a problem here with copy-back caches. Note that what the kernel is doing here is copying the kernel space copies of a page to make sure the faulting process gets an exclusive copy of the page which it can then write to. This is fine, but after this stream of code the cache is in a very insistant state. Watch what happens: 1) user COW faults at address 0x10000 2) Kernel copies exclusive page to the user using kernel addresses old_page = 0xf0ef0000 new_page = 0xf0184000 (addresses are for illustration purposes only to show the case we are trying to avoid) At this point old_page and new_page are in the level 2 copy-back cache, they have not reached real memory yet. The cache only knows about the identity of this page based upon the virtual and physical address of the page, HyperSparc is VIPT remember. 3) Kernel flushes the user page 4) Kernel makes the page table entry for the user 5) Kernel flushes the tlb 6) User reads/writes to page 0x10000 but this misses the cache and goes to real memory Oops... the cache doesn't know to go to the kernel aliases for the pages which were used during the copy, the index and tag wouldn't match. We need to validate main memory when the kernel does stuff like this to keep copy-back caches happy. Here is the resultant code: copy_page(old_page,new_page); flush_page_to_ram(old_page); flush_page_to_ram(new_page); flush_cache_page(vma, address); set_pte(page_table, pte_mkwrite(pte_mkdirty(mk_pte(new_page, vma->vm_page_prot)))); free_page(old_page); flush_tlb_page(vma, address); Much better. flush_page_to_ram() flushes a kernel address such that it reaches main memory (and therefore when the user misses the cache later on it will get the right data from main memory). The last and final thing necessary for true multi-architure support for the mm code has to do with SMP. I won't discuss this one until I work out the fine points with Linus and Alan as to how it should really be used. I hope this helps people understand my changes much better. Later, David S. Miller davem@caip.rutgers.edu From tor@spacetec.no Thu Apr 4 14:23:05 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id OAA04208 for ; Thu, 4 Apr 1996 14:23:04 +0200 (MET DST) Received: from pallas.spacetec.no by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA24565; Thu, 4 Apr 1996 14:22:55 +0200 (MET) Received: (from tor@localhost) by pallas.spacetec.no (8.7.1/8.7.1) id OAA06068 for linux-mips@fnet.fr; Thu, 4 Apr 1996 14:24:01 +0200 Message-Id: <199604041224.OAA06068@pallas.spacetec.no> From: tor@spacetec.no (Tor Arntsen) Date: Thu, 4 Apr 1996 14:23:59 +0200 X-Mailer: Mail User's Shell (7.2.5 10/14/92) To: linux-mips@fnet.fr Subject: sgi crossdev Content-Length: 2619 Lines: 48 Hi all, If there's any interest I could prepare and upload SGI/IRIX cross-dev tools (gcc, binutils). But first I need to know a couple of things: - Should I bother? (I will have to re-build the tools because I don't currently have them in /usr/local.) Or could we just assume that anyone interested in Linux/MIPS who intends to do cross-compiling on the nearest SGI box can easily enough whip up the tools him/herself without too much trouble? (And, unlike most of those using a Linux/x86 for cross development, many SGI users certainly wouldn't want a /usr/local setup (say, no root access), and would need to build their tools for their own setup anyway. Of course the same could be said about the SunOS/Solaris tools that already exist on fnet.. - If I do, what would be the preferred layout/content? Should I do it e.g. the (ftp.fnet.fr) crossdev/i486-linux style, with mipsel-linux, mipsel-linuxelf directories/tools, or should I just make one big honkin' tar file with ./usr/local/mipsel-linuxelf, ./usr/local/mipsel-linux[aout], ./usr/local/lib/gcc-lib/[..], ./usr/local/bin/[..] etc. (I'm not sure if I would want to bother with the bigendian versions at the moment, as I won't be able to test them.) Is there any chance that in the near future we will have a kernel that can be compiled ELF so I could leave out the aout version? (however it doesn't add much effort to make it as soon as I set up the environment, it's mostly a question of how big the tar file gets..) If there's interest I'll be happy to do it. To do the build would be a matter of minutes (make -j and a multi-cpu Challenge helps a lot), but it would still take some days though: - I have to get fresh sources. My gcc 2.7.2 sources (which I use everywhere here) have been modified so that the (IMO) stupid mis-feature of accepting '//' comments is backed out. If people start using such comments here at our site (and be sure some of them do!) it crashes with all the other compilers we have here so.. But there is at least one file in the current Linux/MIPS sources with such a comment and I guess it's better to use the 'standard' 2.7.2 for least- surprice reasons. - However, I would like to keep the 'fstrength-reduce' bug fix, now quite common I believe. - I have to set up some hardware (eg. a fresh /usr/local) for this purpose. The easter days (more like a week actually, in Norway) would be perfect for this, so please somebody let me know if I should go ahead or if I would be wasting my time (it's probably better for me to go skiing!) Cheers, Tor From linux@mailhost.uni-koblenz.de Thu Apr 4 14:58:31 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id OAA04943 for ; Thu, 4 Apr 1996 14:58:30 +0200 (MET DST) Received: from informatik.uni-koblenz.de (mailhost.uni-koblenz.de) by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA25352; Thu, 4 Apr 1996 14:56:59 +0200 (MET) Received: from flake.uni-koblenz.de (linux@flake.uni-koblenz.de [141.26.4.37]) by informatik.uni-koblenz.de (8.7.4/8.6.9) with SMTP id OAA11931 for ; Thu, 4 Apr 1996 14:53:25 +0200 (MET DST) From: Systemkennung Linux Message-Id: <199604041253.OAA11931@informatik.uni-koblenz.de> Received: by flake.uni-koblenz.de (4.1/KO-2.0) id AA19429; Thu, 4 Apr 96 14:52:27 +0200 Subject: Re: sgi crossdev To: linux-mips@fnet.fr Date: Thu, 4 Apr 1996 14:52:26 +0200 (MET DST) In-Reply-To: <199604041224.OAA06068@pallas.spacetec.no> from "Tor Arntsen" at Apr 4, 96 02:23:59 pm Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Length: 3777 Lines: 78 Hi all, > If there's any interest I could prepare and upload SGI/IRIX cross-dev tools > (gcc, binutils). But first I need to know a couple of things: > > - Should I bother? (I will have to re-build the tools because I don't > currently have them in /usr/local.) > Or could we just assume that anyone interested in Linux/MIPS who > intends to do cross-compiling on the nearest SGI box can easily enough > whip up the tools him/herself without too much trouble? (And, unlike most > of those using a Linux/x86 for cross development, many SGI users certainly > wouldn't want a /usr/local setup (say, no root access), and would need to > build their tools for their own setup anyway. Of course the same could be > said about the SunOS/Solaris tools that already exist on fnet.. /usr/local/ or some package in /opt sound reasonable to me. The no-root access is probably less important since most of the people that do kernel hacking are sysadmins anyway ... > - If I do, what would be the preferred layout/content? Should I do it e.g. > the (ftp.fnet.fr) crossdev/i486-linux style, with mipsel-linux, > mipsel-linuxelf directories/tools, or should I just make one big honkin' > tar file with ./usr/local/mipsel-linuxelf, ./usr/local/mipsel-linux[aout], > ./usr/local/lib/gcc-lib/[..], ./usr/local/bin/[..] etc. > (I'm not sure if I would want to bother with the bigendian versions at the > moment, as I won't be able to test them.) The bigendian stuff builds but it as long as noone is actually really working on with it I think this is just wasted time and disk space. > Is there any chance that in the near future we will have a kernel that > can be compiled ELF so I could leave out the aout version? (however it > doesn't add much effort to make it as soon as I set up the environment, > it's mostly a question of how big the tar file gets..) Finally eleminating the complete a.out stuff is one of the things on my todo list. When I do this systemflavours mips*-*-linux and mips*-*-linuxelf will be the same. > If there's interest I'll be happy to do it. To do the build would be a matter > of minutes (make -j and a multi-cpu Challenge helps a lot), but it would > still take some days though: Would you mind to borrow me your Challenge ;-) Well, now that Dave has working SMP I think I should try to get ... > - I have to get fresh sources. My gcc 2.7.2 sources (which I use everywhere > here) have been modified so that the (IMO) stupid mis-feature of accepting > '//' comments is backed out. If people start using such comments here at > our site (and be sure some of them do!) it crashes with all the other > compilers we have here so.. Well, I like the // feature alot. I tried not to have them in the released sources but expect these sooner or later to creep in. That // might break the semantics of some braindead code is IMHO no problem of interest in real live. > But there is at least one file in the current Linux/MIPS sources with such > a comment and I guess it's better to use the 'standard' 2.7.2 for least- > surprice reasons. > - However, I would like to keep the 'fstrength-reduce' bug fix, now quite > common I believe. My published sources already contain a fix for this bug. > - I have to set up some hardware (eg. a fresh /usr/local) for this purpose. Not really necessary for this purpose. Try make install prefix=/tmp/installdir/usr/local for creating such binary package. > The easter days (more like a week actually, in Norway) would be perfect for > this, so please somebody let me know if I should go ahead or if I would be > wasting my time (it's probably better for me to go skiing!) Too bad - not enough snow left for skiing here ... Ralf From tor@spacetec.no Thu Apr 4 18:20:17 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id SAA09573 for ; Thu, 4 Apr 1996 18:20:16 +0200 (MET DST) Received: from pallas.spacetec.no by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA01866; Thu, 4 Apr 1996 18:20:07 +0200 (MET) Received: (from tor@localhost) by pallas.spacetec.no (8.7.1/8.7.1) id SAA06206 for linux-mips@fnet.fr; Thu, 4 Apr 1996 18:21:16 +0200 Message-Id: <199604041621.SAA06206@pallas.spacetec.no> From: tor@spacetec.no (Tor Arntsen) Date: Thu, 4 Apr 1996 18:21:13 +0200 In-Reply-To: Systemkennung Linux "Re: sgi crossdev" (Apr 4, 13:59) X-Mailer: Mail User's Shell (7.2.5 10/14/92) To: linux-mips@fnet.fr Subject: Re: sgi crossdev Content-Length: 3125 Lines: 85 On Apr 4, 13:59, Systemkennung Linux wrote: >/usr/local/ or some package in /opt sound reasonable to me. The no-root >access is probably less important since most of the people that do kernel >hacking are sysadmins anyway ... Ok, building for /usr/local.. >The bigendian stuff builds but it as long as noone is actually really working >on with it I think this is just wasted time and disk space. I'll leave that out then. >> - However, I would like to keep the 'fstrength-reduce' bug fix, now quite >> common I believe. > >My published sources already contain a fix for this bug. Hm, I didn't find your sources.. I only found the diffs, which does not include the fix. I'm using the following: - gcc2.7.2 from prep.ai.mit.edu - plus gcc-2.7.2-1.diffs from ftp.fnet.fr - plus this patch to loop.c (from amylaar@meolyon.hanse.DE, posted to gnu.gcc.bug in September, line numbers slightly offset for 2.7.2): --- loop.c.orig Tue Nov 14 14:59:59 1995 +++ loop.c Wed Feb 14 19:05:30 1996 @@ -6190,6 +6190,11 @@ { if (invariant_p (arg) == 1) { +#if 0 /* Linear transformations of the compared values + are incorrect unless it can be verified that + overflow behaviour stays the same for all possible + values of the loop bound, for all iterations. */ + /* Look for giv with constant positive mult_val and nonconst add_val. Insert insns to compute new compare value. */ @@ -6216,6 +6221,7 @@ if (apply_change_group ()) return 1; } +#endif } /* This code has problems. Basically, you can't know when - binutils-2.6 - plus binutils-2.6-1.diffs from ftp.fnet.fr - plus float.h from ftp.fnet.fr in usr/local/lib/gcc-lib/*/2.7.2/include/ > make install prefix=/tmp/installdir/usr/local Right, that saves a lot of work. I have it done now. Questions: I have the following: ls usr/local/ bin/ info/ man/ mipsel-linuxelf/ include/ lib/ mipsel-linux/ I have just tar'ed everything into one big tar file (compressed size is almost 9MB), which includes binutils-2.6 mipsel-linux and mipsel-linuxelf, gcc 2.7.2 mipsel-linux and mipsel-linuxelf. (I'm not sure if I it would be enough with two versions of binutils and just one GCC, however I'm using two GCCs for my own setup and I know this works so..) - Should I split up this big mips-sgi-irix5.3.tar.gz file into binutils, gcc, mipsel-linux, mipsel-linuxelf, or could I just keep it as it is -- a big, 'cd /;zcat file|tar xvf - ' one-go installation thing? - Where should I ftp the file/the files? I've understood from the mail archive that ftp.fnet.fr may not be the right place to ftp to, although things will/should end up there after a while.. >> The easter days (more like a week actually, in Norway) would be perfect for >> this, so please somebody let me know if I should go ahead or if I would be >> wasting my time (it's probably better for me to go skiing!) > >Too bad - not enough snow left for skiing here ... Got literally speaking meters of it here! Cheers, Tor From linux@mailhost.uni-koblenz.de Thu Apr 4 22:03:39 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id WAA17571 for ; Thu, 4 Apr 1996 22:03:38 +0200 (MET DST) Received: from informatik.uni-koblenz.de ([141.26.4.1]) by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA08354; Thu, 4 Apr 1996 22:03:19 +0200 (MET) Received: from flake.uni-koblenz.de (linux@flake.uni-koblenz.de [141.26.4.37]) by informatik.uni-koblenz.de (8.7.4/8.6.9) with SMTP id WAA01641 for ; Thu, 4 Apr 1996 22:02:06 +0200 (MET DST) From: Systemkennung Linux Message-Id: <199604042002.WAA01641@informatik.uni-koblenz.de> Received: by flake.uni-koblenz.de (4.1/KO-2.0) id AA21417; Thu, 4 Apr 96 22:01:06 +0200 Subject: Re: sgi crossdev To: linux-mips@fnet.fr Date: Thu, 4 Apr 1996 22:01:06 +0200 (MET DST) In-Reply-To: <199604041621.SAA06206@pallas.spacetec.no> from "Tor Arntsen" at Apr 4, 96 06:21:13 pm Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Length: 4088 Lines: 100 Hi, > >> - However, I would like to keep the 'fstrength-reduce' bug fix, now quite > >> common I believe. I tried to reproduce the bug for MIPS but apprently the MIPS port of GCC is pretty immune to this bug. > >My published sources already contain a fix for this bug. > > Hm, I didn't find your sources.. I only found the diffs, which does not > include the fix. I'm using the following: Ooops, I was talking about the patch, of course. > - gcc2.7.2 from prep.ai.mit.edu > - plus gcc-2.7.2-1.diffs from ftp.fnet.fr > - plus this patch to loop.c (from amylaar@meolyon.hanse.DE, posted to > gnu.gcc.bug in September, line numbers slightly offset for 2.7.2): > --- loop.c.orig Tue Nov 14 14:59:59 1995 > +++ loop.c Wed Feb 14 19:05:30 1996 > @@ -6190,6 +6190,11 @@ > { > if (invariant_p (arg) == 1) > { > +#if 0 /* Linear transformations of the compared values > + are incorrect unless it can be verified that > + overflow behaviour stays the same for all possible > + values of the loop bound, for all iterations. */ > + > /* Look for giv with constant positive mult_val and nonconst > add_val. Insert insns to compute new compare value. */ > > @@ -6216,6 +6221,7 @@ > if (apply_change_group ()) > return 1; > } > +#endif > } > > /* This code has problems. Basically, you can't know when > > - binutils-2.6 > - plus binutils-2.6-1.diffs from ftp.fnet.fr > - plus float.h from ftp.fnet.fr in usr/local/lib/gcc-lib/*/2.7.2/include/ Btw - the float.h was created on a R3000 RISC/os 4.52 machine. It's 100% identical to the one generated on Linux/MIPS. > > make install prefix=/tmp/installdir/usr/local > > Right, that saves a lot of work. I have it done now. Questions: > I have the following: > > ls usr/local/ > bin/ info/ man/ mipsel-linuxelf/ > include/ lib/ mipsel-linux/ I recommend to delete all .a libraries and include/ generated by the binutils on installation and to cleanup the manpages and info files. There is lots of trash in which many GNU programs install but noone uses. Similar for GCC. Not installing libbfd.a and header might even prevent some problems - there might already be another bfd lib and header files for it installed in the same lib directory if you also installed gcc with the same --prefix for another target (which was probably your native compiler). > I have just tar'ed everything into one big tar file (compressed size is > almost 9MB), which includes binutils-2.6 mipsel-linux and mipsel-linuxelf, > gcc 2.7.2 mipsel-linux and mipsel-linuxelf. (I'm not sure if I it would be > enough with two versions of binutils and just one GCC, however I'm using > two GCCs for my own setup and I know this works so..) For now we need two versions of both GCC and binutils. With the elemination of a.out only ELF will be left over and therefore half that disk waste ... > - Should I split up this big mips-sgi-irix5.3.tar.gz file into binutils, gcc, > mipsel-linux, mipsel-linuxelf, or could I just keep it as it is -- a big, > 'cd /;zcat file|tar xvf - ' one-go installation thing? I think it's best to keep one tool and one target flavour in one binary archive. > - Where should I ftp the file/the files? I've understood from the mail > archive that ftp.fnet.fr may not be the right place to ftp to, although > things will/should end up there after a while.. You can either talk to to Luc.Beurton@fnet.fr about making you an account for uploads on lena.fnet.fr or upload into some incomfing directory on ftp.fnet.fr and sending a mail to Luc and Stoned.Elipot@univ-evry.fr. > >> The easter days (more like a week actually, in Norway) would be perfect for > >> this, so please somebody let me know if I should go ahead or if I would be > >> wasting my time (it's probably better for me to go skiing!) > > > >Too bad - not enough snow left for skiing here ... > > Got literally speaking meters of it here! (Just getting envious ...) Ralf From kaos@audio.apana.org.au Sat Apr 6 08:00:51 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id IAA01886 for ; Sat, 6 Apr 1996 08:00:45 +0200 (MET DST) From: kaos@audio.apana.org.au Received: from audio.apana.org.au by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA08912; Sat, 6 Apr 1996 08:00:26 +0200 (MET) Received: (from kaos@localhost) by audio.apana.org.au (8.6.11/8.6.9) id QAA11123; Sat, 6 Apr 1996 16:00:06 +1000 Date: Sat, 6 Apr 1996 16:00:06 +1000 Message-Id: <199604060600.QAA11123@audio.apana.org.au> To: linux-mips@fnet.fr X-Mailer: Lynx, Version 2.3 BETA X-Personal_Name: Keith Owens Subject: RC3300 Linux? Content-Length: 719 Lines: 20 First, there is a bad link on your MIPS WWW page. File that you are currently viewing Linkname: The Linux/MIPS FAQ URL: http://lena.fnet.fr/#top Owner(s): None size: 290 lines mode: normal Link that you currently have selected Linkname: MIPS Filename: http://www.mips.com/HTMLs/Mips_Chip_Rm.html The "MIPS" link does not exist. Secondly, I have been asked if there is anything we can do with an old Prime 7333 system which seems to be a rebadged RC3300. Knowing nothing about MIPS but a fair bit about Linux (currently 1.3.84 ELF), I'd like to know if it is worth trying to install Linux on a RC3300 and where I would start. Sorry about the lack of word wrap, old text browser. From r0j6245@tam2000.tamu.edu Mon Apr 8 03:48:39 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id DAA26589 for ; Mon, 8 Apr 1996 03:48:34 +0200 (MET DST) Received: from tam2000.tamu.edu by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA05152; Mon, 8 Apr 1996 03:48:23 +0200 (MET) Received: from localhost (r0j6245@localhost) by tam2000.tamu.edu (8.7.3/8.7.3) with SMTP id UAA20929 for ; Sun, 7 Apr 1996 20:47:12 -0500 (CDT) Date: Sun, 7 Apr 1996 20:47:12 -0500 (CDT) From: Roji John To: linux-mips@fnet.fr Subject: sgi indy? Message-Id: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Content-Length: 656 Lines: 15 Hi there. I am working on a Silicon Graphics Indy workstation. It is using an R4400 cpu. SGI's decision to withhold a compiler (unless you want to dish out extra money) has made me consider moving from IRIX to now-popular Linux. I'm a little lost on where to get started. Perhaps someone can answer the following questions for me : 1> Is there a version of Linux for my system? 2> If so, where can I get it? (possible options are CD or ftp sites) 3> Will it run emulation-style or will it completely replace the current OS I am using? Any further information is always helpful, even if it us just pointing me in the right direction. Thank you! From lm@neteng.engr.sgi.com Mon Apr 8 04:35:29 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id EAA27575 for ; Mon, 8 Apr 1996 04:35:29 +0200 (MET DST) Received: from sgi.sgi.com (SGI.COM) by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA05584; Mon, 8 Apr 1996 04:35:19 +0200 (MET) Received: from cthulhu.engr.sgi.com by sgi.sgi.com via ESMTP (950405.SGI.8.6.12/910110.SGI) for <@sgi.engr.sgi.com:linux-mips@fnet.fr> id TAA08165; Sun, 7 Apr 1996 19:35:24 -0700 Received: from neteng.engr.sgi.com by cthulhu.engr.sgi.com via ESMTP (950511.SGI.8.6.12.PATCH526/911001.SGI) for <@cthulhu.engr.sgi.com:linux-mips@fnet.fr> id TAA29376; Sun, 7 Apr 1996 19:35:23 -0700 Received: from localhost (lm@localhost) by neteng.engr.sgi.com (950413.SGI.8.6.12/960327.SGI.AUTOCF) via SMTP id TAA01728 for ; Sun, 7 Apr 1996 19:35:23 -0700 Message-Id: <199604080235.TAA01728@neteng.engr.sgi.com> To: linux-mips@fnet.fr From: lm@gate1-neteng.engr.sgi.com (Larry McVoy) Subject: Re: sgi indy? Date: Sun, 07 Apr 1996 19:35:23 -0700 Sender: lm@neteng.engr.sgi.com Content-Length: 485 Lines: 11 : Hi there. I am working on a Silicon Graphics Indy workstation. It is : using an R4400 cpu. SGI's decision to withhold a compiler (unless you : want to dish out extra money) has made me consider moving from IRIX to : now-popular Linux. Linux isn't ready yet for SGI systems. It will be in about 4 months. In the mean time, you should bug your sales rep for a freeware CD that includes precompiled binaries for much of the GNU stuff, including gcc, unless I'm mistaken. --lm From haua@taec.com Wed Apr 10 18:50:53 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id SAA07352 for ; Wed, 10 Apr 1996 18:50:50 +0200 (MET DST) Received: from godzilla.taec.com by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA04452; Wed, 10 Apr 1996 18:50:47 +0200 (MET) Received: from taec.com by godzilla.taec.com (8.6.12/Toshiba-AEC-RELAY.1) id JAA12660; Wed, 10 Apr 1996 09:48:58 -0700 Received: from cobh.sanjose by taec.com (4.1/SMI-4.1) id AA21299; Wed, 10 Apr 96 09:50:18 PDT Date: Wed, 10 Apr 96 09:50:18 PDT From: haua@taec.com (Anne Hau) Message-Id: <9604101650.AA21299@taec.com> To: linux-mips@fnet.fr Subject: MILO Cc: haua@taec.com Content-Length: 927 Lines: 33 Hi there, My name is Anne Hau, currently working on getting Linux to run on MIPS rPC44 Deskstation. I do not have problem installing Linux on PC. Many Linux distributions available make installing Linux on PC achievable. With MIPS-linux, unfortunately, information is not widely available. I was able to compile mips-linux 1.3.62 kernel and MILO. I compiled using mips-gcc on the PC host, then copied these images to floppies. Now the next step is to get Linux kernel installed. Upon reset rPC44, I got into ARCBIOS set up screen. From this screen, it gives the option of running a program which I put:a:\milo. It then complained : ERROR Load: Bad magic value in optional header: 00000107 != 00000000 Am I doing the right thing ? Did I miss any thing here ? This machine has not been powered up for two years. It currenly running windowNT 3.1. Any help would be greatly appreciated. Best regards, Anne From imp@rover.village.org Wed Apr 10 21:05:39 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id VAA10145 for ; Wed, 10 Apr 1996 21:05:36 +0200 (MET DST) Received: from rover.village.org by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA06888; Wed, 10 Apr 1996 21:05:32 +0200 (MET) Received: from localhost (localhost [127.0.0.1]) by rover.village.org (8.6.12/8.6.6) with SMTP id NAA04479; Wed, 10 Apr 1996 13:05:23 -0600 Message-Id: <199604101905.NAA04479@rover.village.org> To: linux-mips@fnet.fr Subject: Re: MILO Cc: haua@taec.com In-Reply-To: Your message of Wed, 10 Apr 1996 09:50:18 PDT Date: Wed, 10 Apr 1996 13:05:23 -0600 From: Warner Losh Content-Length: 2763 Lines: 61 : My name is Anne Hau, currently working on getting Linux to run on MIPS rPC44 : Deskstation. Yow! I didn't even know there were more than one of these ever made :-) : I do not have problem installing Linux on PC. Many Linux distributions : available make installing Linux on PC achievable. With MIPS-linux, : unfortunately, information is not widely available. True. However, don't expect a road as simple as Red Hat for the Linux/MIPS stuff. : I was able to compile mips-linux 1.3.62 kernel and MILO. I compiled using : mips-gcc on the PC host, then copied these images to floppies. : : Now the next step is to get Linux kernel installed. : : Upon reset rPC44, I got into ARCBIOS set up screen. From this screen, : it gives the option of running a program which I put:a:\milo. : It then complained : : ERROR : Load: Bad magic value in optional header: 00000107 != 00000000 Hmmm. I've never seen this particular error message. You need to be careful to not strip the milo that you build, but this bug may be fixed. Also, milo needs to be a ECOFF image, not an ELF one, which may be what you are seeing. Finally, for the rPC44, you'll need to compile milo to load at a different address than the stock MILO that comes with the system. I've not been able to get any released version of milo to work with an ELF kernel either, so you may want to compile a normal ecoff (aka a.out) kernel as well. Also, the MILO that comes with the milo distributions won't work on the rpc44. You have to hack the makefile to load at the right place, but the comments might be clear. I'm at work right this instant and can't take but a minute or two to look things up. Once you do all of that, you may be in for a let down. So far I've had problems mounting the root file system and having it read files off of it correctly. Some minor DMA problem is precluding it from getting data twice. I've been off working on taxes and some final (again) details on a consulting contract that I've had for far too long, so I've not investigated this problem too closely. It may be possible to have a NFS mounted root file system and that might work for you. Up until very very recently, I was unable to do *ANY* nfs on my network for reasons of policy rather than technical. I also think that you'll have problems with the kernel unless it is configured only for the Deskstation. I've not completed the integration of machine independent stuff that Ralf is working towards yet. Please do let me know how things go. This is the first time I've seen anybody else with another Deskstation, let alone another rPC44! As far as I know, there are no install tools, or install packages avaialble (yet) for the mips stuff, so you may be in for a bumpy ride. Warner From imp@rover.village.org Wed Apr 10 21:09:25 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id VAA10233 for ; Wed, 10 Apr 1996 21:09:25 +0200 (MET DST) Received: from rover.village.org by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA06934; Wed, 10 Apr 1996 21:09:21 +0200 (MET) Received: from localhost (localhost [127.0.0.1]) by rover.village.org (8.6.12/8.6.6) with SMTP id NAA04542; Wed, 10 Apr 1996 13:09:19 -0600 Message-Id: <199604101909.NAA04542@rover.village.org> To: linux-mips@fnet.fr Subject: Re: MILO Cc: haua@taec.com In-Reply-To: Your message of Wed, 10 Apr 1996 09:50:18 PDT Date: Wed, 10 Apr 1996 13:09:19 -0600 From: Warner Losh Content-Length: 198 Lines: 7 : unfortunately, information is not widely available. I almost forgot, the FAQ for linux mips can be found at http://www.fnet.fr/linux-mips. This is most up to date doc that there is :-(. Warner From gwesp@cosy.sbg.ac.at Thu Apr 11 12:14:18 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id MAA29474; Thu, 11 Apr 1996 12:14:09 +0200 (MET DST) Received: from loewe.cosy.sbg.ac.at by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA21169; Thu, 11 Apr 1996 12:14:02 +0200 (MET) Received: from dodo.cosy.sbg.ac.at (dodo [141.201.2.47]) by loewe.cosy.sbg.ac.at (8.7.4/8.7.3) with ESMTP id MAA09183; Thu, 11 Apr 1996 12:14:13 +0200 (MET DST) From: Gerhard Wesp Received: (from gwesp@localhost) by dodo.cosy.sbg.ac.at (8.7.1/8.7.1) id MAA05663; Thu, 11 Apr 1996 12:13:36 +0200 (MET DST) Message-Id: <199604111013.MAA05663@dodo.cosy.sbg.ac.at> Subject: DECstation 5000/33 To: linux-mips@fnet.fr Date: Thu, 11 Apr 1996 12:13:35 +0200 (MET DST) Cc: linux-mips-ftp@fnet.fr X-Mailer: ELM [version 2.4 PL24 ME8b] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 315 Lines: 10 Hi! I have this R3000 based decstation and would like to try linux on it. I already tried the ``experimental kernel'' from softway.au. I'd like to get in contact with the developers (if there are any ;-) Do you by chance have an email address of anybody involved? What about mailing lists? Greetings, -Gerhard From paul@suite.sw.oz.au Thu Apr 11 14:13:39 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id OAA02003 for ; Thu, 11 Apr 1996 14:13:39 +0200 (MET DST) Received: from staff.cs.su.OZ.AU ([129.78.8.1]) by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA23776; Thu, 11 Apr 1996 14:13:32 +0200 (MET) Received: from suite.sw.oz.au by swallow.sw.oz.au with SMTP id NAA27637; Thu, 11 Apr 1996 13:17:32 GMT (8.6.10/Unixware) (from paul@suite.sw.oz.au for ) Received: by suite.sw.oz.au id AA17331; Thu, 11 Apr 1996 22:13:11 +1000 (5.65c/1.34) (from paul@suite.sw.oz.au for linux-mips@fnet.fr) From: paul@suite.sw.oz.au (Paul Antoine) Message-Id: <199604111213.AA17331@suite.sw.oz.au> Subject: Re: DECstation 5000/33 To: linux-mips@fnet.fr Date: Thu, 11 Apr 1996 22:13:10 +1000 (EST) In-Reply-To: <199604111013.MAA05663@dodo.cosy.sbg.ac.at> from "Gerhard Wesp" at Apr 11, 96 12:13:35 pm Organization: Softway Pty Ltd X-Face: U)Kb/c%d9`JI>Y>s'evmMaR`z*WE$<,["^2)I8o?jkx/d^=dMa$GD(s6[p|npE}r0?)qQ7H *3S2vbK66POj?=%pCn?Bp}D/3A{`-.v0=%L9emS+|Pc91$EP0-;i\ON*p;Mwp$tt3wO~g1Z W.mIz'n~$U&uIM9ix=\1=~""=W?t$3b)>`foC8+3g$m2l;^D$?i}WI(Uxc@THJ9OBbk3g6 X-Mailer: ELM [version 2.4 PL24 PGP3 *ALPHA*] Content-Type: text Content-Length: 1654 Lines: 40 Gerhard wrote: > I have this R3000 based decstation and would like to try linux on it. > I already tried the ``experimental kernel'' from softway.au. I'd like > to get in contact with the developers (if there are any ;-) > Do you by chance have an email address of anybody involved? What about > mailing lists? Well, you've reached the right mailing list. Have you looked at the DEC WWW pages at: http://www.softway.com.au/people/paul/dec ??? This has the latest info. and email addresses for those who've said they're contributing (I'm the de facto co-ordinator of the DEC and R3000 portions of the Linux/MIPS project). Not much has been done in the last couple of months, as I gather most of us have been a bit busy. The test kernel was based on 1.2.11 sources, and the current general Linux/MIPS source release of 1.3.6x has all of the code from the earlier release re-integrated. Much more remains to be done - please read the WWW pages if you can. Glad to have you on board! Regards, Paul P.S. I'm off to Brussels, Geneva and Paris in the next few days. Any of our European members who'd like to meet over a beer should email me within the next day or so. (Big hint to Stoned & Luc!) _______________________________________________________________________________ Paul M. Antoine, Net: paul@sw.oz.au Softway Pty Ltd WWW: www.softway.com.au PO Box 305, Strawberry Hills, NSW 2012, Australia Tel: +61 2 698 2322 Level 2, 79 Myrtle St, Chippendale, NSW 2008, Australia Fax: +61 2 699 9174 "It is the lack of acceptance of diversity which threatens to destroy society, NOT the free expression of it." - Me. From lorenz@tinman.dfki.uni-sb.de Thu Apr 11 22:10:49 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id WAA05796 for ; Thu, 11 Apr 1996 22:10:43 +0200 (MET DST) Received: from tinman.dfki.uni-sb.de by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA03923; Thu, 11 Apr 1996 22:10:34 +0200 (MET) Received: (from lorenz@localhost) by tinman.dfki.uni-sb.de (8.7.5/8.7.3) id WAA14013 for linux-mips@fnet.fr; Thu, 11 Apr 1996 22:10:00 +0200 (MET DST) From: Benjamin Lorenz Message-Id: <199604112010.WAA14013@tinman.dfki.uni-sb.de> Subject: Linux on a DEC station 3100 with R2000A (fwd) To: linux-mips@fnet.fr Date: Thu, 11 Apr 1996 22:10:00 +0200 (MET DST) X-Mailer: ELM [version 2.4 PL23] Content-Type: text Content-Length: 1732 Lines: 56 Hi, I already sent this mail to Ralf personally, but now I know this list address and think it might be a good idea to remail this to the whole list. --Benni PS: Can I subscribe to this list or is it for "hackers only"? Forwarded message: > From lorenz Thu Apr 11 21:39:03 1996 > Subject: Linux on a DEC station 3100 with R2000A > To: ralf@mailhost.uni-koblenz.de > Date: Thu, 11 Apr 1996 21:39:03 +0200 (MET DST) > X-Mailer: ELM [version 2.4 PL23] > Content-Type: text > Content-Length: 1267 > > Huhu Ralf, > > wir haben hier im DFKI ein paar alte MIPS Kisten rumstehen und > ich hatte die gloreiche Idee, da mal Linux drauf zu installieren... > > Kannst Du mir kurz bestaetigen, ob Dein Linux z.B. auf folgenden > Kisten laeuft: > > DEC station 3100 mit R2000A ("Baujahr" 1988) > > DEC station 5000 mit folgender Boot-Info: > swayambhu,ULTRIX V4.4 (Rev. 69) System #3: Wed Apr 13 19:13:20 MET DST 1994 > real mem = 33554432 > avail mem = 25743360 > using 819 buffers containing 3354624 bytes of memory > DECstation 5000 Model 125 - system rev 48 > 64Kb Instruction Cache, 64Kb Data Cache > cpu0 ( version 3.0, implementation 2 ) > fpu0 ( version 3.0, implementation 3 ) > HBA 'PMAZ-BA ' and DME 'PMAZ-BA ' attached on Controller 0. > asc0 at ibus3 > rz0 at asc0 slave 0 (DEC RZ57 (C) DEC 5000) > rz2 at asc0 slave 2 (DEC RZ24 (C) DEC 211B) > rz3 at asc0 slave 3 (DEC RZ24 (C) DEC 211B) > rz4 at asc0 slave 4 (DEC RRD42 (C) DEC 4.3d) > ln0 at ibus3 > ln0: DEC LANCE Ethernet Interface, hardware address: 08:00:2b:24:83:44 > scc0 at ibus3 > px0 at ibus0 > pa0 (5x1 8+8+0+0) > px1 at ibus1 > pa1 (5x1 8+8+0+0) > > Schonmal danke, ...und gruesse mir Chris Blum! :-) > > Ciao, > Benni From hannon@Crissy.Stanford.EDU Sat Apr 13 02:56:17 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id CAA12753 for ; Sat, 13 Apr 1996 02:56:16 +0200 (MET DST) Received: from relay2.EUnet.fr by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA00676; Sat, 13 Apr 1996 02:56:12 +0200 (MET) Received: from crissy.Stanford.EDU by relay2.eunet.fr (5.65c8d/96.04.04) via EUnet-France id AA03984; Sat, 13 Apr 1996 02:56:08 +0200 (MET) Received: by Crissy.Stanford.EDU (8.6.8/25-LAGUNITA-eef) id PAA09437; Thu, 11 Apr 1996 15:23:09 -0700 From: Steph Hannon Message-Id: <199604112223.PAA09437@Crissy.Stanford.EDU> Subject: Looking for big endian root disk image To: linux-mips@fnet.fr Date: Thu, 11 Apr 96 15:23:08 PDT Cc: hannon@Crissy.Stanford.EDU (Steph Hannon), munira@Crissy.Stanford.EDU (Munira Rahemtulla) X-Mailer: ELM [version 2.3 PL11] Content-Length: 612 Lines: 15 We are looking for a root-disk image for Linux/MIPS that is big endian. We undertaking a project to port Linux/MIPS to a hardware simulation environment that only supports big endian. We would prefer not to rewrite the hardware simulator to support little endian if possible. Any information would be greatly appreciated. Could someone also tell us how to subscribe to this mailing list. We would very much like to be involved and share information about our development status. Munira Rahemtulla Stephanie Hannon munira@crissy.stanford.edu hannon@crissy.stanford.edu SimOS Project Stanford University From ryan@ravage.labs.gmu.edu Mon Apr 14 20:45:57 1997 Received-Date: Mon, 14 Apr 1997 20:45:56 +0200 (MET DST) Received: from ravage.labs.gmu.edu ([129.174.166.55]) by Utopia.EUnet.fr (5.65c8d/EUNET-4.2.3) via EUnet-France id AA13263; Mon, 14 Apr 1997 20:42:57 +0200 (MET) Received: from localhost (ryan@localhost) by ravage.labs.gmu.edu (8.8.3/8.8.3) with SMTP id CAA14403 for ; Sun, 14 Apr 1996 02:48:10 -0400 Date: Sun, 14 Apr 1996 02:48:09 -0400 (EDT) From: Ryan Rafferty To: linux-mips Subject: NT for MIPS In-Reply-To: <9704012314.AA26413@hilbert.numerik.math.uni-siegen.de> Message-Id: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Content-Length: 676 Lines: 14 This is probably the wrong place to ask questions regarding NT, I know =). But I have had no luck getting Linux to boot on my (now) decked-out Magnum 4000PC. I have a 230meg SCSI disk, a 4x SCSI CD-ROM, the standard g364 frame buffer video, and 16 megs of RAM. But when I load the Linux kernel I built under Linux/Intel using the cross-compiler, it halts. Arrgh. Anyway, until I iron out the bug I have with Linux, can anyone point me toward any resources for Windows NT/MIPS? I have the latest Windows Internet Explorer for MIPS, but I can't find a single other piece of software out there. (Hmm, could this be the reason NT/MIPS was canned? =) Thanks, Ryan Rafferty From ryan@ravage.labs.gmu.edu Wed Apr 16 23:12:42 1997 Received-Date: Wed, 16 Apr 1997 23:12:41 +0200 (MET DST) Received: from ravage.labs.gmu.edu ([129.174.166.55]) by Utopia.EUnet.fr (5.65c8d/EUNET-4.2.3) via EUnet-France id AA11666; Wed, 16 Apr 1997 23:09:38 +0200 (MET) Received: from localhost (ryan@localhost) by ravage.labs.gmu.edu (8.8.3/8.8.3) with SMTP id FAA06848 for ; Tue, 16 Apr 1996 05:14:13 -0400 Date: Tue, 16 Apr 1996 05:14:13 -0400 (EDT) From: Ryan Rafferty To: Linux Mips Subject: Console problems In-Reply-To: <1350939722-72211321@viavision.com> Message-Id: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Content-Length: 1076 Lines: 28 I tried again today to get Linux running on my poor underutilized Magnum. I compiled the kernel with the cross-compiler and AFAIK, I said 'yes' to all the right options during 'make config'. Then I put the precompiled pandora and milo on my SCSI disk along with the kernel. Next, I restarted the box and chose 'run a program' at the ARC console, and pointed to pandora, which loaded fine. Then I typed > boot C:\zImage (I defined C: to point to my FAT-formatted SCSI disk under ARC) from which point it appeared to load the kernel. After a few seconds, the screen turned red-orange for an instant, then the bottom half of the screen went blue and the top was black with what appeared to be scrambled text with the scan lines all messed up and unreadable. When I hit the 'enter' key, the lines of garbled text scrolled down a few lines and then nothing would happen. Any suggestions as to what I'm doing wrong? I have the g364 framebuffer video card installed and an NE2000 in an ISA slot. Other than that, it's a standard Magnum 4000PC... Thanks. Ryan Rafferty From ryan@ravage.labs.gmu.edu Thu Apr 17 00:32:48 1997 Received-Date: Thu, 17 Apr 1997 00:32:47 +0200 (MET DST) Received: from ravage.labs.gmu.edu ([129.174.166.55]) by Utopia.EUnet.fr (5.65c8d/EUNET-4.2.3) via EUnet-France id AA13869; Thu, 17 Apr 1997 00:29:46 +0200 (MET) Received: from localhost (ryan@localhost) by ravage.labs.gmu.edu (8.8.3/8.8.3) with SMTP id GAA07213 for ; Tue, 16 Apr 1996 06:34:13 -0400 Date: Tue, 16 Apr 1996 06:34:13 -0400 (EDT) From: Ryan Rafferty To: linux-mips Subject: Re: Console problems In-Reply-To: <199704162153.XAA27495@informatik.uni-koblenz.de> Message-Id: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Content-Length: 1156 Lines: 29 On Wed, 16 Apr 1997, Systemkennung Linux wrote: > Are you using the Magnum in Hires mode (1280x1024)? Only 1024x768 is > working. It also looks as if you've launched the kernel directly without > milo which won't work. Ah yes. I started changing the resolutions right after I sent that e-mail and now the console comes up OK. This is of course a major improvement =). There is still junk at the bottom of the screen, but the text is there. I'm gonna look at the g364 code and try to see how hard it would be to get the console working in 1280x1024, since that's the most convenient mode for dual-booting with NT. A few questions--what is the latest development kernel you're using for Linux/MIPS? I've got 2.0.21 which may be a little old. Also, where can I find the specs for the JAZZ stuff that you used to port to the Magnum? > The onboard Sonic Ethernet is supported, so no need for the NE2000 - > unless you wish to use something worse than the Sonic ... The Sonic may be better, but the fact that it only has an AUI connector is a real pain. Our network only has BNC and UTP, and transceivers are expensive. Thanks, Ryan > Ralf > From linux@mailhost.uni-koblenz.de Tue Apr 16 15:25:35 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id PAA00440 for ; Tue, 16 Apr 1996 15:25:34 +0200 (MET DST) Received: from informatik.uni-koblenz.de (mailhost.uni-koblenz.de) by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA06501; Tue, 16 Apr 1996 14:04:26 +0200 (MET) Received: from ozzy (ozzy.uni-koblenz.de [141.26.5.8]) by informatik.uni-koblenz.de (8.7.4/8.6.9) with SMTP id OAA09304 for ; Tue, 16 Apr 1996 14:05:29 +0200 (MET DST) From: Systemkennung Linux Message-Id: <199604161205.OAA09304@informatik.uni-koblenz.de> Received: by ozzy (SMI-8.6/KO-2.0) id OAA12110; Tue, 16 Apr 1996 14:02:39 +0200 Subject: New kernel patches To: linux-mips@fnet.fr Date: Tue, 16 Apr 1996 14:02:37 +0200 (MET DST) Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Length: 1631 Lines: 34 Hi all, it's been quite ago that I've uploaded the last kernel patches. Now I'm trying to compensate for this sin ... As usual you can find the new kernel patches on ftp.fnet.fr:/linux-mips/src/kernel/v1.3/. Please forgive that I didn't upload a full kernel archive - the european Internet connections are currently slower than 0.1 kb/s. Some of the noteworthy MIPS specific changes in these kernels are: - 64 bit operations can now be used withhin the kernel and the kernel is being compiled as MIPS III code. - The CPU model specific code has been cleaned up and moved into four subdirectories arch/mips/mips[1-4]. As noone of the MIPS hackers has MIPS II and MIPS IV machine the code for these CPUs is just a goody I supply in the hope it will be usefull for somebody. The MIPS I code is currently under development. - All Jazz family machines (Acer PICA, Mips Magnum 4000, Oily M700-10) now use the same interrupt/exception handler code. Have fun, Ralf 89b403e6656ebe21fdfb012befef65af linux-1.3.62-1.3.63.diffs.gz 01d6c22f75930c53bc52a66bdeab20fb linux-1.3.63-1.3.64.diffs.gz 22969600f804d75690a3b3cc3105256f linux-1.3.64-1.3.65.diffs.gz 79f8f2404f360b4342b0a16730510ea3 linux-1.3.65-1.3.66.diffs.gz e4811920c04182c118190f2b2413f730 linux-1.3.66-1.3.67.diffs.gz 121928766aeaec5c0b831e687af8690f linux-1.3.67-1.3.68.diffs.gz 9631311f0eb3838b682d5f27a2299ac5 linux-1.3.68-1.3.69.diffs.gz 61c75a1d8bfee4a9b3f688f7babb0492 linux-1.3.69-1.3.70.diffs.gz 6ae7fa0c38cd1aa9b0a18e1ca54a8595 linux-1.3.70-1.3.71.diffs.gz c1d1aeefeb08ab95ac6b8ea963f0e7b6 linux-1.3.71-1.3.72.diffs.gz From ryan@ravage.labs.gmu.edu Thu Apr 17 02:28:40 1997 Received-Date: Thu, 17 Apr 1997 02:28:39 +0200 (MET DST) Received: from ravage.labs.gmu.edu ([129.174.166.55]) by Utopia.EUnet.fr (5.65c8d/EUNET-4.2.3) via EUnet-France id AA16554; Thu, 17 Apr 1997 02:25:38 +0200 (MET) Received: from localhost (ryan@localhost) by ravage.labs.gmu.edu (8.8.3/8.8.3) with SMTP id IAA07882 for ; Tue, 16 Apr 1996 08:30:19 -0400 Date: Tue, 16 Apr 1996 08:30:19 -0400 (EDT) From: Ryan Rafferty To: linux-mips Subject: Re: Console problems In-Reply-To: <199704162343.BAA07330@informatik.uni-koblenz.de> Message-Id: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Content-Length: 764 Lines: 16 On Thu, 17 Apr 1997, Systemkennung Linux wrote: > I think what you observe is that the screen jump up by a random amount > and then stays there. The space at the bottom is filled with random > junk. Bad because the top lines are no longer readable ... > > This is a bug in Milo for which I once posted a fix. Since that time > a newer version of Milo never has been released, so ... Bummer. This does make it harder to see things that are going on. In your past posts you said that at this point it is actually impossible to recompile milo because of a break in the tool chain, right? Is anyone working on correcting this situation, or is it ripe for someone to take over? =) Without the ability to recompile the bootloader, it can be kinda limiting... From linux@mailhost.uni-koblenz.de Tue Apr 16 21:03:05 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id VAA07273 for ; Tue, 16 Apr 1996 21:03:05 +0200 (MET DST) Received: from informatik.uni-koblenz.de (mailhost.uni-koblenz.de) by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA13989; Tue, 16 Apr 1996 21:02:48 +0200 (MET) Received: from ozzy (ozzy.uni-koblenz.de [141.26.5.8]) by informatik.uni-koblenz.de (8.7.4/8.6.9) with SMTP id VAA03390 for ; Tue, 16 Apr 1996 21:02:50 +0200 (MET DST) From: Systemkennung Linux Message-Id: <199604161902.VAA03390@informatik.uni-koblenz.de> Received: by ozzy (SMI-8.6/KO-2.0) id VAA17596; Tue, 16 Apr 1996 21:01:13 +0200 Subject: G364 video trouble To: linux-mips@fnet.fr Date: Tue, 16 Apr 1996 21:01:12 +0200 (MET DST) Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Length: 2956 Lines: 63 -----BEGIN PGP SIGNED MESSAGE----- Hi all, I just tracked down the problems with the messed up display that seemed to affect certain Olivetti M700s. Below is a fix for Milo 0.26. Ralf (Stoned, could you please put this into your Milo sources? Thanks!) begin 644 milo-0.26.diffs.gz M'XL(".SL9P M@(11$G+-;BX!.ZO5*$*-W4`K/I#=S(!F\]^WV@9L@TW,OFT"G;B.S^VORM7= M9;+Q&)2Y]P0VLUQ%+5?ULNNQ2>61O-$QLV@DWX@D15&RS(]N/0:_$P>J.FA: MNZ&V:Z>@M5JZ),MR"M;18$[A`H7%Q)< M]HZ..@@R\ZFE6,R9+Q3B2?!PO2.V3`D>+_^X"11X*TGNW]\\/(C+$?&G$GS" M#UR[SF\:L,5P1:DPQ7U+-:L+LGMC,G\(C,4S7(2)J M4*Q(\B?F&-;\%W!\0UI*T^_(,\;'?&,"G[3I:_6(YY;@N;7A>0_7EX[I4>+#U=SWZ4$, M8KHY9L6DF./9F@GU4I4SPJ<.L:F_E]-)3:\G^`P%:5R&FFT>Z^U&,XW'E76L M?E1543^JL5S5M)(N"@C^J0<\1L\P-7\&$$.3CH/'DV#N^&SB4!-?=PYW0C<0 M+_#WZNOWT]?.+TFI%`%^J0NL1^JBJN*@J6)LC'"HZ3CHP;56)^\E#((26K>$ M=75C7:=BB%GK]??2VI8(VV:`(73Z:&W;#&T)>9?D2O'H@%G(1[DG$9KFF0.\ MGR%C"OY"I2A"V74=[KD6].B$^9QZ/EP"RO'^$9O]O[X.NO?/M[?#;X.[DQ\N M,PL2_!*/GF#>X)XU)"5BFMZ94(I_.B%"=]![&%X&TM"J1=-OL//TOR)&UP.F_)0Z8CTT<)D5ZR&\$`\K M39!7858I&]XFE`]-YL\LLCQ)WG912EXO"Y(K('0$'^)=WP0KB'40K0V7?BP$`85D%*>Z1]$@D=)[B@C2%3'>AB^N9TQ3H+2#H-8!VL6I'H33G^+6 M<`]:[2"T;U<6<=Y28.J'P63/IW$0T`-SZ(#9-`5(?ST@DT4:I\#W;OHW`\S; MC!*3HP`4Q!*;MQ#)L40_J!2MWX12:2-)[D(R4A_HY"D!9(A8/>]\ON* M%$(LRG9 M%"D1+.A`6J'Z"",J3^D86@Z,560`,N91S8&1J$DI&+4<&&$E@LQYU/-@)!YF M%Z.1`V-=@;(P]-?,C(5$:N(ZF'GP829U.!LO$X>?2)AV`(JT*8?):MHA*.81 M/TRJ[4:K7:O%#D)JT$A1HV/[`S.HX],R]"D-V@]!LZ?[_/+WUZ<[T?`0,IO@ M/R;SJ,%=;RGZ&GPJFB*8C>P'NK@>V*X7'O!-R@FS_/)!!]!0)]HISF1U"HN= MY'V[,G)=SIRQN_>,B>?ZM'Y57)Q&=UR_33B>(_4TPA,^2Z*-_K&8C.7 MPWEX60XO+\8.Y>6Q]^7C1)3W)&)V&FZI1.`\.@FZ($K.S`X;*(;K^*Y%/^R\ M[(WDQ)K3G2B&PJP(AMJ4Z#6SHK?RV(JV&]UHO1$-:B=EIH@XZB'_.QN&(_M1>G8 M7HHU!L!GP62&/[#`NT/#M7Q MU:M:O7;&9+D@*4?1WL=RG4DX)/=$&_EZQ>U=/LH,MT:XJZLG?\5ZF-@LX5[M )7Z2Z8/M2&``` ` end -----BEGIN PGP SIGNATURE----- Version: 2.6.2 iQCVAgUBMXPudkckbl6vezDBAQE7QwQAlm7bwFTg+ec7djxqzkVyU8pj0GRCTzuW 8WEYYH8ZqWrNqcaKQb/xcP76ronBEv5zalir3pjOeAaORc5kkBxffNTKhaTyihaN JS7GHR97TYGr+yL0cbS5MOkA8PnY+vqSa1MwqkB6RJtjSSL8dz9nkQLPI9vpbb2G sO6FgnZJ9M8= =hRtZ -----END PGP SIGNATURE----- From Stoned.Elipot@univ-evry.fr Wed Apr 17 10:29:56 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id KAA24665 for ; Wed, 17 Apr 1996 10:29:55 +0200 (MET DST) Received: from amiante.univ-evry.fr by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA26893; Wed, 17 Apr 1996 10:29:43 +0200 (MET) Received: from bandsept.univ-evry.fr (bandsept.univ-evry.fr [192.93.214.19]) by amiante.univ-evry.fr with ESMTP (8.6.12/96.04.09/louis-blanc); id KAA19614; Wed, 17 Apr 1996 10:29:48 +0200 Received: from localhost.univ-evry.fr (localhost.univ-evry.fr [127.0.0.1]) by bandsept.univ-evry.fr with SMTP (8.6.12/94.01.26); id KAA28832; Wed, 17 Apr 1996 10:29:09 +0200 Message-Id: <199604170829.KAA28832@bandsept.univ-evry.fr> To: Systemkennung Linux Cc: luc.beurton@fnet.fr (Luc Beurton), linux-mips@fnet.fr (The List) Cc: +sent@univ-evry.fr Subject: Re: G364 stuff in Milo In-Reply-To: Your message of "Tue, 16 Apr 1996 20:26:17 +0200." <199604161827.UAA02015@informatik.uni-koblenz.de> X-It: bandsept (wrong OS, but right CPU) X-Mailer: MH 6.8.3, MH-E 5.0.2 Date: Wed, 17 Apr 1996 10:29:02 +0200 From: Stoned Elipot Content-Length: 1257 Lines: 28 Hi Ralf, I post that to list also, as it can be of general interest for Magnuj/Oily owners. >>>>> "Ralf" == Systemkennung Linux writes: Ralf> Hi, I've played a bit more with the G364 stuff in Milo. The Ralf> problem seems to be that G364_get_display() calls Ralf> G364_SWITCHOFF_VTG() and lateron G364_SWITCHON_VTG(). I dunno Ralf> why this procedure is required. Additionally doing Ralf> *(volatile unsigned int*)G364_TopOfScreen = 0; Ralf> after G364_SWITCHON_VTG() fixes the problem. It only looks as Ralf> if the monitor is resynchonising again when G364_SWITCHON_VTG() Ralf> is getting called. Not having a manual for the G364 I'm Ralf> interested in a comment from you what this all means ... Luc can surely answer you more correctly as he had done this G364 stuff, but what I had understood is that when various G364 registers are read when the Video Timing Generator is on you just get non significant values, that's the reason why the VTG has to be turned off before reading them. This means that you monitor just lost its synchronisation input so seems to crash, but it recover pretty quickly. And well for the TopOfScreen stuff we'll look at it, we got the patches you posted, thanks. Cheers Stoned. From beurton@nil.fnet.fr Wed Apr 17 11:44:50 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id LAA26164 for ; Wed, 17 Apr 1996 11:44:49 +0200 (MET DST) Received: from nil.fnet.fr by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA28792; Wed, 17 Apr 1996 11:44:39 +0200 (MET) Received: from nil.fnet.fr (localhost.fnet.fr [127.0.0.1]) by nil.fnet.fr (8.7.5/8.7.3) with ESMTP id LAA23994 for ; Wed, 17 Apr 1996 11:45:00 +0200 (METDST) Message-Id: <199604170945.LAA23994@nil.fnet.fr> To: linux-mips@fnet.fr Subject: Test. X-Mailer: MH 6.8.3 Date: Wed, 17 Apr 1996 11:45:00 +0200 From: Luc Beurton Content-Length: 12 Lines: 3 Test. Luc. From beurton@nil.fnet.fr Wed Apr 17 12:00:33 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id MAA00496 for ; Wed, 17 Apr 1996 12:00:32 +0200 (MET DST) Received: from nil.fnet.fr by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA29112; Wed, 17 Apr 1996 12:00:22 +0200 (MET) Received: from nil.fnet.fr (localhost.fnet.fr [127.0.0.1]) by nil.fnet.fr (8.7.5/8.7.3) with ESMTP id MAA24049 for ; Wed, 17 Apr 1996 12:00:42 +0200 (METDST) Message-Id: <199604171000.MAA24049@nil.fnet.fr> To: linux-mips@fnet.fr Subject: Test X-Mailer: MH 6.8.3 Date: Wed, 17 Apr 1996 12:00:42 +0200 From: Luc Beurton Content-Length: 11 Lines: 2 Test2 Luc. From beurton@nil.fnet.fr Wed Apr 17 12:13:40 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id MAA00967 for ; Wed, 17 Apr 1996 12:13:40 +0200 (MET DST) Received: from nil.fnet.fr by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA29370; Wed, 17 Apr 1996 12:13:30 +0200 (MET) Received: from nil.fnet.fr (localhost.fnet.fr [127.0.0.1]) by nil.fnet.fr (8.7.5/8.7.3) with ESMTP id MAA24086 for ; Wed, 17 Apr 1996 12:13:52 +0200 (METDST) Message-Id: <199604171013.MAA24086@nil.fnet.fr> To: linux-mips@fnet.fr Subject: test 4 X-Mailer: MH 6.8.3 Date: Wed, 17 Apr 1996 12:13:52 +0200 From: Luc Beurton Content-Length: 14 Lines: 3 Test 4 Luc. From beurton@nil.fnet.fr Wed Apr 17 12:51:52 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id MAA02105 for ; Wed, 17 Apr 1996 12:51:52 +0200 (MET DST) Received: from nil.fnet.fr by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA00269; Wed, 17 Apr 1996 12:51:43 +0200 (MET) Received: from nil.fnet.fr (localhost.fnet.fr [127.0.0.1]) by nil.fnet.fr (8.7.5/8.7.3) with ESMTP id MAA24280 for ; Wed, 17 Apr 1996 12:52:05 +0200 (METDST) Message-Id: <199604171052.MAA24280@nil.fnet.fr> To: linux-mips@fnet.fr Subject: Some problems. X-Mailer: MH 6.8.3 Date: Wed, 17 Apr 1996 12:52:04 +0200 From: Luc Beurton Content-Length: 93 Lines: 7 Hi all, There were some problems on the list yesterday. I sorry. I hope it's ok now. Luc. From Joe@widind.com Thu Apr 18 05:18:42 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id FAA13150 for ; Thu, 18 Apr 1996 05:18:41 +0200 (MET DST) Received: from mail2.jump.net (news.jump.net) by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA16455; Thu, 18 Apr 1996 05:18:27 +0200 (MET) Received: from widind.jump.net by mail2.jump.net (8.7.3/BERK-6.8.11) id VAA06111; Wed, 17 Apr 1996 21:29:48 -0500 (CDT) Date: Wed, 17 Apr 1996 21:29:48 -0500 (CDT) Message-Id: <199604180229.VAA06111@mail2.jump.net> X-Sender: widind@mail.jump.net X-Mailer: Windows Eudora Light Version 1.5.2 Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" To: linux-mips@fnet.fr From: Joe Subject: The Linux/MIPS HOWTO Content-Length: 698 Lines: 15 Greetings, I am trying to install Linux on a MIPS Magnum 4000. I have tried to follow the instructions in this HOWTO but am getting nowhere. I downloaded severl versions of MILO but am doing something wrong. If I just use gzip -d milo26-0.gz to extract MILO, it converts it to s single 500K file and when I copy that to a blank formatted floppy the Magnum returns "invalid image file" The HOWTO refers to a "vmlinux" and I can not find that anywhere or refered to anyplace else. I am a complete newbe to RISC machines, to UNIX and of course to Linux. Considerable experience with DOS/Windows. Can you help steer me towards some help on this?? Thanks and Best Reguards---- Joe Westenhaver From ryan@ravage.labs.gmu.edu Fri Apr 18 19:14:39 1997 Received-Date: Fri, 18 Apr 1997 19:14:38 +0200 (MET DST) Received: from ravage.labs.gmu.edu ([129.174.166.55]) by Utopia.EUnet.fr (5.65c8d/EUNET-4.2.3) via EUnet-France id AA25911; Fri, 18 Apr 1997 19:11:35 +0200 (MET) Received: from localhost (ryan@localhost) by ravage.labs.gmu.edu (8.8.3/8.8.3) with SMTP id BAA13264 for ; Thu, 18 Apr 1996 01:15:53 -0400 Date: Thu, 18 Apr 1996 01:15:53 -0400 (EDT) From: Ryan Rafferty To: linux-mips Subject: Re: Milo, the next generation In-Reply-To: <199704181128.NAA06053@informatik.uni-koblenz.de> Message-Id: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Content-Length: 358 Lines: 14 On Fri, 18 Apr 1997, Systemkennung Linux wrote: > Urghhh also. Tested 0.27.1 some minutes ago on the Oily sitting on a desk > here at the university. Crashes like a charm after reading the kernel > image from disk :-( That's a shame... It works great on my Magnum. Totally gets rid of that annoying screen-shift! Muchismas gracias. Ryan > Ralf > From Joe@widind.com Thu Apr 18 10:17:23 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id KAA19282 for ; Thu, 18 Apr 1996 10:17:23 +0200 (MET DST) Received: from gatekeeper.waldorf-gmbh.de by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA00530; Thu, 18 Apr 1996 10:17:12 +0200 (MET) Received: from relay.xlink.net (relay.xlink.net [193.141.40.5]) by gatekeeper.waldorf-gmbh.de (8.6.12/8.6.9) with SMTP id KAA15529 for ; Thu, 18 Apr 1996 10:20:01 +0200 Received: from news.jump.net by relay.xlink.net id <55750-0@relay.xlink.net>; Thu, 18 Apr 1996 08:26:03 +0000 Received: from widind.jump.net by mail2.jump.net (8.7.3/BERK-6.8.11) id BAA09611; Thu, 18 Apr 1996 01:07:08 -0500 (CDT) Date: Thu, 18 Apr 1996 01:07:08 -0500 (CDT) Message-Id: <199604180607.BAA09611@mail2.jump.net> X-Sender: widind@mail.jump.net X-Mailer: Windows Eudora Light Version 1.5.2 Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" To: linux@waldorf-gmbh.de From: Joe Subject: The Linux/MIPS HOWTO Content-Length: 1007 Lines: 19 Can you give me a little steering help? I,ve read your HOWTO and would like to try the Slackware distribution on my machine. I am trying to install Linux on a MIPS Magnum 4000. I have tried to follow the instructions in the The Linux/MIPS HOWTO but am getting nowhere. I downloaded several versions of MILO but am doing something wrong. If I just use gzip -d milo26-0.gz to extract MILO, it converts it to s single 500K file and when I copy that to a blank formatted floppy the Magnum returns "invalid image file" The HOWTO refers to a "vmlinux" and I can not find that anywhere or refered to anyplace else. I am a complete newbe to RISC machines, to UNIX and of course to Linux. Considerable experience with DOS/Windows. Can you help steer me towards some help on this?? This machine is "new" in the sense it has never been activated before or had any os or software installed. I have a configuration floppy and the PROM initialization seems fine. Thanks and Best Reguards---- Joe Westenhaver From hodgen@mailhost.uni-koblenz.de Thu Apr 18 14:34:44 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id OAA01520 for ; Thu, 18 Apr 1996 14:34:38 +0200 (MET DST) Received: from informatik.uni-koblenz.de (mailhost.uni-koblenz.de) by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA06011; Thu, 18 Apr 1996 14:34:15 +0200 (MET) Received: from ozzy (ozzy.uni-koblenz.de [141.26.5.8]) by informatik.uni-koblenz.de (8.7.4/8.6.9) with SMTP id OAA29644 for ; Thu, 18 Apr 1996 14:32:31 +0200 (MET DST) From: Wayne Hodgen Message-Id: <199604181232.OAA29644@informatik.uni-koblenz.de> Received: by ozzy (SMI-8.6/KO-2.0) id OAA27261; Thu, 18 Apr 1996 14:30:54 +0200 Subject: M700 partition problem To: linux-mips@fnet.fr Date: Thu, 18 Apr 1996 14:30:53 +0200 (MET DST) Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Length: 1901 Lines: 45 Hi all, I finally got started working on Olli stuff. I put a half G ide disk in the Olli so I can avoid our (overloaded) net. There were of course a set of DOS partitions on it already. This causes an adel in genhd.c-> extended_partition when it tries to do a partition check on the contents of the extended partition. I traced the problem to the line if (!NR_SECTS(p) || is_extended_partition(p)) (line 163 in drivers/block/genhd.c). The crash comes at NR_SECTS(p), this is defined at the top of genhd as follows: #ifdef __alpha__ /* * On the Alpha, we get unaligned access exceptions on * p->nr_sects and p->start_sect, when the partition table * is not on a 4-byte boundary, which is frequently the case. * This code uses unaligned load instructions to prevent * such exceptions. */ #include #define NR_SECTS(p) ldl_u(&p->nr_sects) #define START_SECT(p) ldl_u(&p->start_sect) #else /* __alpha__ */ #define NR_SECTS(p) p->nr_sects #define START_SECT(p) p->start_sect #endif /* __alpha__ */ I can only assume that the MIPS needs similar handling to the alpha (to avoid unaligned access) but someone (Ralf? Michael?) borrowed my MIPS bible some time ago and never brought it back... Has anyone got the book on hand to tell me the nessecary code for MIPS? I seem to remember it was a construct not a single instruction but it's been months since I read the book. I wouldn't have bothered you all with this except that Ralf is away for a day or so and I'd like to get the disk installed so I can start on that disgusting console... Ciao -- Wayne Hodgen | hodgen@informatik.uni-koblenz.de | #include From pefo@enea.se Thu Apr 18 15:47:40 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id PAA02959 for ; Thu, 18 Apr 1996 15:47:40 +0200 (MET DST) Received: from outside.enea.se by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA07451; Thu, 18 Apr 1996 15:47:36 +0200 (MET) Received: by outside.enea.se (8.6.9/8.6.9) id QAA24253 for ; Thu, 18 Apr 1996 16:48:54 +0200 Received: by outside via smap (V1.3) id sma024251; Thu Apr 18 16:48:45 1996 Received: (from pefo@localhost) by enea.se (8.7.1/8.7.1) id PAA09802 for linux-mips@fnet.fr; Thu, 18 Apr 1996 15:47:20 +0200 (MET DST) From: Per Fogelstrom Message-Id: <199604181347.PAA09802@gordon.enea.se> Subject: Re: M700 partition problem To: linux-mips@fnet.fr Date: Thu, 18 Apr 1996 15:47:20 +0200 (MET DST) In-Reply-To: <199604181232.OAA29644@informatik.uni-koblenz.de> from "Wayne Hodgen" at Apr 18, 96 02:30:53 pm X-Mailer: ELM [version 2.4 PL23] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 1310 Lines: 36 > > Hi all, > > I finally got started working on Olli stuff. I put a half G ide disk in > the Olli so I can avoid our (overloaded) net. There were of course a > set of DOS partitions on it already. This causes an adel in genhd.c-> > extended_partition when it tries to do a partition check on the contents > of the extended partition. I traced the problem to the line > > if (!NR_SECTS(p) || is_extended_partition(p)) > > (line 163 in drivers/block/genhd.c). The crash comes at NR_SECTS(p), this > is defined at the top of genhd as follows: > > #ifdef __alpha__ > /* > * On the Alpha, we get unaligned access exceptions on > * p->nr_sects and p->start_sect, when the partition table > * is not on a 4-byte boundary, which is frequently the case. > * This code uses unaligned load instructions to prevent > * such exceptions. > */ > #include > #define NR_SECTS(p) ldl_u(&p->nr_sects) > #define START_SECT(p) ldl_u(&p->start_sect) > #else /* __alpha__ */ > #define NR_SECTS(p) p->nr_sects > #define START_SECT(p) p->start_sect > #endif /* __alpha__ */ > You need to do the same as alphas do. Mipses can't load unaligned without using special instructions or load a byte at a time and merge them. I ran into this when porting XFree86 to my 4000 running NetBSD. Per From hodgen@mailhost.uni-koblenz.de Thu Apr 18 18:07:16 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id SAA05917 for ; Thu, 18 Apr 1996 18:07:16 +0200 (MET DST) Received: from relay2.EUnet.fr by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA00558; Thu, 18 Apr 1996 18:07:13 +0200 (MET) Received: from informatik.uni-koblenz.de (mailhost.uni-koblenz.de) by relay2.eunet.fr (5.65c8d/96.04.04) via EUnet-France id AA13986; Thu, 18 Apr 1996 16:47:21 +0200 (MET) Received: from ozzy (ozzy.uni-koblenz.de [141.26.5.8]) by informatik.uni-koblenz.de (8.7.4/8.6.9) with SMTP id QAA07524 for ; Thu, 18 Apr 1996 16:45:31 +0200 (MET DST) From: Wayne Hodgen Message-Id: <199604181445.QAA07524@informatik.uni-koblenz.de> Received: by ozzy (SMI-8.6/KO-2.0) id QAA27431; Thu, 18 Apr 1996 16:43:53 +0200 Subject: Re: M700 partition problem To: linux-mips@fnet.fr Date: Thu, 18 Apr 1996 16:43:53 +0200 (MET DST) In-Reply-To: <199604181347.PAA09802@gordon.enea.se> from "Per Fogelstrom" at Apr 18, 96 03:47:20 pm Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Length: 830 Lines: 18 Hi, > You need to do the same as alphas do. Mipses can't load unaligned without > using special instructions or load a byte at a time and merge them. > > I ran into this when porting XFree86 to my 4000 running NetBSD. Yeah, I already worked that out, but as I wrote at the bottom of the first mail, I can't do MIPS assembler off the top of my head (yet ;) and my book is missing. Maybe I phrased it wrong, can you tell me the assembler intruction to do the unaligned load? Thats what I need to know. Ciao -- Wayne Hodgen | hodgen@informatik.uni-koblenz.de | #include From pefo@enea.se Thu Apr 18 19:04:01 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id TAA07142 for ; Thu, 18 Apr 1996 19:04:00 +0200 (MET DST) Received: from outside.enea.se (gateway.enea.se) by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA02096; Thu, 18 Apr 1996 19:03:58 +0200 (MET) Received: by outside.enea.se (8.6.9/8.6.9) id UAA10175 for ; Thu, 18 Apr 1996 20:05:17 +0200 Received: by outside via smap (V1.3) id sma010173; Thu Apr 18 20:05:03 1996 Received: (from pefo@localhost) by enea.se (8.7.1/8.7.1) id TAA18541 for linux-mips@fnet.fr; Thu, 18 Apr 1996 19:03:38 +0200 (MET DST) From: Per Fogelstrom Message-Id: <199604181703.TAA18541@gordon.enea.se> Subject: Re: M700 partition problem To: linux-mips@fnet.fr Date: Thu, 18 Apr 1996 19:03:37 +0200 (MET DST) In-Reply-To: <199604181445.QAA07524@informatik.uni-koblenz.de> from "Wayne Hodgen" at Apr 18, 96 04:43:53 pm X-Mailer: ELM [version 2.4 PL23] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 898 Lines: 28 > > Hi, > > > You need to do the same as alphas do. Mipses can't load unaligned without > > using special instructions or load a byte at a time and merge them. > > > > I ran into this when porting XFree86 to my 4000 running NetBSD. > > Yeah, I already worked that out, but as I wrote at the bottom of the first > mail, I can't do MIPS assembler off the top of my head (yet ;) and my > book is missing. Maybe I phrased it wrong, can you tell me the assembler > intruction to do the unaligned load? Thats what I need to know. > Ok! #define ldunaligned(x) \ ({ int __result; \ __asm__("lwr %0,0(%1);lwl %0,3(%1)" \ : "&=r" (__result) : "r" (x)); __result; }) int nn = ldunaligned(*arbitrary address*); You have to swap "lwr" and "lwl" if you run in big endian mode. And the order too perhaps, if the last one ends up as a nop if "x" is aligned. Per From linux@mailhost.uni-koblenz.de Thu Apr 18 19:52:40 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id TAA07989 for ; Thu, 18 Apr 1996 19:52:40 +0200 (MET DST) Received: from informatik.uni-koblenz.de (mailhost.uni-koblenz.de) by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA03162; Thu, 18 Apr 1996 19:52:32 +0200 (MET) Received: from ozzy (ozzy.uni-koblenz.de [141.26.5.8]) by informatik.uni-koblenz.de (8.7.4/8.6.9) with SMTP id TAA15894 for ; Thu, 18 Apr 1996 19:52:25 +0200 (MET DST) From: Systemkennung Linux Message-Id: <199604181752.TAA15894@informatik.uni-koblenz.de> Received: by ozzy (SMI-8.6/KO-2.0) id TAA29367; Thu, 18 Apr 1996 19:50:46 +0200 Subject: Re: M700 partition problem To: linux-mips@fnet.fr Date: Thu, 18 Apr 1996 19:50:45 +0200 (MET DST) In-Reply-To: <199604181347.PAA09802@gordon.enea.se> from "Per Fogelstrom" at Apr 18, 96 03:47:20 pm Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Length: 512 Lines: 13 Hi all, > You need to do the same as alphas do. Mipses can't load unaligned without > using special instructions or load a byte at a time and merge them. > > I ran into this when porting XFree86 to my 4000 running NetBSD. The IDE/partition table stuff in the Linux kernel is full of these alignment problems. I've already fixed several of these bugs. It's just been luck that I didn't run into this particular problem. Anyway, I now have Wayne's M700 sitting on my table and will fix this thing. Ralf From linux@mailhost.uni-koblenz.de Thu Apr 18 21:31:50 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id VAA10020 for ; Thu, 18 Apr 1996 21:31:49 +0200 (MET DST) Received: from informatik.uni-koblenz.de (mailhost.uni-koblenz.de) by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA04651; Thu, 18 Apr 1996 21:31:44 +0200 (MET) Received: from ozzy (ozzy.uni-koblenz.de [141.26.5.8]) by informatik.uni-koblenz.de (8.7.4/8.6.9) with SMTP id VAA19075 for ; Thu, 18 Apr 1996 21:31:23 +0200 (MET DST) From: Systemkennung Linux Message-Id: <199604181931.VAA19075@informatik.uni-koblenz.de> Received: by ozzy (SMI-8.6/KO-2.0) id VAA29632; Thu, 18 Apr 1996 21:29:41 +0200 Subject: Re: M700 partition problem To: linux-mips@fnet.fr Date: Thu, 18 Apr 1996 21:29:39 +0200 (MET DST) In-Reply-To: <199604181703.TAA18541@gordon.enea.se> from "Per Fogelstrom" at Apr 18, 96 07:03:37 pm Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Length: 625 Lines: 20 Hi, > Ok! > > #define ldunaligned(x) \ > ({ int __result; \ > __asm__("lwr %0,0(%1);lwl %0,3(%1)" \ > : "&=r" (__result) : "r" (x)); __result; }) > > int nn = ldunaligned(*arbitrary address*); > > You have to swap "lwr" and "lwl" if you run in big endian mode. > And the order too perhaps, if the last one ends up as a nop if "x" is > aligned. You can also use the macro instructions uld, ulw, ulh, ulhu, usd, usw, ush, ushu provided by the assembler. That way you won't have to worry about byte order. Btw - GAS generates bad code for these under some circumstances. Ralf From linux@mailhost.uni-koblenz.de Thu Apr 18 23:32:47 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id XAA12402 for ; Thu, 18 Apr 1996 23:32:46 +0200 (MET DST) Received: from informatik.uni-koblenz.de (mailhost.uni-koblenz.de) by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA06439; Thu, 18 Apr 1996 23:32:32 +0200 (MET) Received: from ozzy (ozzy.uni-koblenz.de [141.26.5.8]) by informatik.uni-koblenz.de (8.7.4/8.6.9) with SMTP id XAA23426 for ; Thu, 18 Apr 1996 23:32:30 +0200 (MET DST) From: Systemkennung Linux Message-Id: <199604182132.XAA23426@informatik.uni-koblenz.de> Received: by ozzy (SMI-8.6/KO-2.0) id XAA02056; Thu, 18 Apr 1996 23:30:51 +0200 Subject: Fix for M700 partition problem To: linux-mips@fnet.fr Date: Thu, 18 Apr 1996 23:30:50 +0200 (MET DST) Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Length: 2876 Lines: 60 -----BEGIN PGP SIGNED MESSAGE----- Hi all, below are the diffs for the alignment problem reported by Wayne. There were several more of these buglets but I think I got them all :-) Ralf begin 644 linux-1.3.77.diffs.gz M'XL("+&Q=C$"`VQI;G5X+3$N,RXW-RYD:69FR0YEK__NM9,!`"#"]M!_B9+6[ MVF?U[&K5D/;[8*>R`XSR]*OMUO9JS69-2#JH^S*(ZD,:J_K=]?GU@L+26M&V M[8T."G MGH+M-IK50ZCHSQ&]@134V_=2*0LY+\D9K&'L2>4*5!I'`N9 M@,\A9:,4AGX@Q:_%BK%!@ZH6PX!P(OT$[?P0`A$2Z`N9[1*3@/H,`E\1&$6$ M0P_CPD`&5"5$8ER2&&^0J?O#&3Z=+<)LDT;"^O84;TA5X,O0]AF#W=I?OO@, M/4G\!Y7Y#P4G(#CH970YR1[E*D$)"6'`:"\`T?L'\Z7,N5"=$C]!HV"V^ZIB=**P%BX18J?(,ZU;J M%OY$,(9#37#V6VZS==!\AGRK[==R<,^I-ERHZ$_#<+`(KR@/6(IG^\97P[H: MX_D-:]$[7+)?T7Y(^N!Y/HLCW_.*%90`BBC'(*V9O`S?OLV+]1&CM`CUW:(- MNW#-#0O:6KL*(X+$2B#E/J,#;>$'`5$*R->`3%@M.)+'V!D;I&EHCFY;8T!C MB.UW7'JZ?I1Q@'\C661B1-6,53JL&&54FT+B]QC)C)$27"2&<+!O]\:Z"$3* M0U^.M2D-(JW2E^1+2GC"QL:3KH::L;^+<-543*J0A7FX3&`I:=;*-,C"Q080 M2_*(7K*=58J^[[L[NN%9<++&1>:NW,I:&< M*W;OVK=W1G=1-<]0&<^?,"SO^F[.`!U*92;>R(`L\A6QS07U7$P+P>A8.);F MJF"F\NVB*4*6>]W#1OH7-_=76">MLW`AH2.]C0":8Z6[6" MJ?)\+3L'+6P'>P=;]8)5#K`9[#DM=_Y"<@^K3;R0S$N7_MXY7\#SA(3MT/"0#XZB9XZ"<^GBH6 M:B_MZW/-%H._G,\H[IY]//=NKMJ?VO?GE]=NX^0%(L<23KY@.5D6$WQ0AB@1 MG('%>KZ'8>,EI'\SU"[#FS>PUT!/F9*5KU2+4+!2KJ9=@P]@MPP[)O+&Y_*) MS@%6H4E`PSG8TPF8?*>C18[R]GW[W#L[GT?O:KF%U_D0;TB,HE%^*>BZD^20 MRU7(85">K$:!AKM6MFK-5C'`.3\G6Q4D)TE=L2AXKACS]0V%F"L6/D@*'T@/ MX!@MY^QH[=(]UJ-E'QUH`"&*H((3U`798V,=#+7Q_D?`**ADAF6CL M]7V5>`$FPZ)!7(7)ELBM'M?U3)QDM(X)M9&*K]@0,]N."X+:2JHU\M3>Z MR1GKXI7AY(QUJ@XVVNJQJTE0P?%.CVF7G.GKNY_R?*0)Q9,9C>`L@^K&Q,Q' M?6[W[ M3OOJ\J+S_MS['=%BS4 M5]"2S-PBNB##4):+E7^+E<(*#<^31)WH,`JHJH:>9Y7P2?AW\MJI6J_=<@D7 M"JW2VQU9TN6*RN5,DOUMG)TF2U$29>?R^C,WL]P/8V#PV`$ MA]85S$^P?A&4#KJ4JF7RZ7&_H__?0Z4,7WU3;!,";H^- L;<2V@H[KH(U^%K0)%;>'-MH(;24QUX&+7A#<_`-P5=_4WOX#F"8A6R44``"- ` end -----BEGIN PGP SIGNATURE----- Version: 2.6.2 iQCVAgUBMXa0h0ckbl6vezDBAQEc8wP+KqeGkq/zWLAStjOXnrhWsbRyX+S0TZq4 ySd3TCWSy5zh0fj1A+sR3d6waWAFSeYPQOqPd2itAY5Paq8H5Kqrb4J/8UDAHbY3 QXu0ZKKC1sEc25FbU7kqdYtMUgwWGjPflXsqVkpG5FsylHTmko6ZCG33me8PKvBL NcUlBB5oiSE= =QBK4 -----END PGP SIGNATURE----- From ralf@Julia.DE Fri Apr 19 20:51:03 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id UAA09086 for ; Fri, 19 Apr 1996 20:51:02 +0200 (MET DST) Received: from alles.intern.julia.de ([194.221.49.2]) by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA27374; Fri, 19 Apr 1996 20:50:39 +0200 (MET) Received: from pris.intern.julia.de (pris.intern.julia.de [194.221.49.37]) by alles.intern.julia.de (8.7.5/8.7.3) with ESMTP id UAA23528 for ; Fri, 19 Apr 1996 20:50:06 +0200 From: Ralf Baechle Received: (from ralf@localhost) by pris.intern.julia.de (8.7.5/8.7.3) id UAA01390 for linux-mips@fnet.fr; Fri, 19 Apr 1996 20:49:56 +0200 (MET DST) Message-Id: <199604191849.UAA01390@pris.intern.julia.de> Subject: More kernel patches To: linux-mips@fnet.fr Date: Fri, 19 Apr 1996 20:49:55 +0200 (MET DST) Reply-To: linux@uni-koblenz.de Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Length: 1203 Lines: 35 -----BEGIN PGP SIGNED MESSAGE----- Hi all, as you've noticed I'm trying to close the gap between Linus' kernel releases and my MIPS port. So here are the next patches. As usual they're available on ftp.fnet.fr and it's mirror. The full kernel source archive isn't yet online, but the "inner circle" get it as usual from private/ Please note that 1.3.77 doesn't yet include the kernel patch for the partition table bug that I've posted yesterday. It is however in 1.3.78 which I'll release rsn. Happy hacking, Ralf 714b4a5f3758c9d5e345a22c3d6121ad linux-1.3.72-1.3.73.diffs.gz bfa14aca2761e8d64afcd10e8c9eb564 linux-1.3.73-1.3.74.diffs.gz 38513936d62513576bbba66733c370d2 linux-1.3.74-1.3.75.diffs.gz 73452a47bc1c26f6e2af183cce735651 linux-1.3.75-1.3.76.diffs.gz e14488ba278a2f321ac5a821a7d44dd4 linux-1.3.76-1.3.77.diffs.gz 696f8939f1277a19ddd59170c6bb2e06 linux-1.3.77.tar.gz -----BEGIN PGP SIGNATURE----- Version: 2.6.3i Charset: latin1 iQCVAwUBMXfgUkckbl6vezDBAQFe4QP/QjQGwvx6lk9UTCLEBU182rZkg0SxFL0q WJXwrSo8l9e84bA7n73wzhSEwFtgbB3l83NaobAHcxwxg+7hOuTo62bUqNyu7UIX ssxYrXiWxOeWEcH5A65kqxOzVLJs3EpQT/gtJIdeENSLpBZ/8qu5ZKsrNOYUNAXT U+q703EfgWQ= =uzdI -----END PGP SIGNATURE----- From ryan@ravage.labs.gmu.edu Mon Apr 21 05:18:58 1997 Received-Date: Mon, 21 Apr 1997 05:18:57 +0200 (MET DST) Received: from ravage.labs.gmu.edu ([129.174.166.55]) by Utopia.EUnet.fr (5.65c8d/EUNET-4.2.3) via EUnet-France id AA15320; Mon, 21 Apr 1997 05:15:51 +0200 (MET) Received: from localhost (ryan@localhost) by ravage.labs.gmu.edu (8.8.3/8.8.3) with SMTP id LAA02348 for ; Sat, 20 Apr 1996 11:20:24 -0400 Date: Sat, 20 Apr 1996 11:20:23 -0400 (EDT) From: Ryan Rafferty To: linux-mips Subject: Magnums In-Reply-To: <199704210237.CAA01057@suede.sw.oz.au> Message-Id: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Content-Length: 1128 Lines: 24 I managed to get bootp on my Linux/Intel box and nfsroot on my Magnum working OK, which is great. But after that point, things go downhill... I untarred the Linux/MIPS root filesystem image and that mounts fine via NFS on the Magnum. When I boot the 2.0.22 kernel I compiled on my Linux/Intel machine, it gets the bootp signal correctly, then mounts the root filesystem OK, then dies when it gets to the part of /etc/init.d/boot where it tries to turn on swap, etc. Of course, it complains about illegal instructions all over the place, but things seem to progress OK up til that point. If I comment out the swapon and "mnt -a nfs" statements in /etc/init.d/boot, the Magnum gets all the way to where it says "INIT: Entering Runlevel: 2" and then after about a minute it complains: INIT: Id "1" respawning too fast: disabled for 5 minutes (repeats for Id 1 through 5) INIT: No more processes left in runlevel: 2 and then it hangs. I assume this is because /sbin/getty keeps dying for some reason. Any help? Could I be running a kernel that's too old, or a root filesystem image that's too old? Thanks, Ryan Rafferty From ralf@Julia.DE Sat Apr 20 19:20:34 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id TAA07444 for ; Sat, 20 Apr 1996 19:20:33 +0200 (MET DST) Received: from alles.intern.julia.de ([194.221.49.2]) by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA12731; Sat, 20 Apr 1996 19:20:25 +0200 (MET) Received: from pris.intern.julia.de (pris.intern.julia.de [194.221.49.37]) by alles.intern.julia.de (8.7.5/8.7.3) with ESMTP id TAA24872 for ; Sat, 20 Apr 1996 19:20:07 +0200 From: Ralf Baechle Received: (from ralf@localhost) by pris.intern.julia.de (8.7.5/8.7.3) id TAA02019 for linux-mips@fnet.fr; Sat, 20 Apr 1996 19:19:45 +0200 (MET DST) Message-Id: <199604201719.TAA02019@pris.intern.julia.de> Subject: Even more kernelpatches To: linux-mips@fnet.fr Date: Sat, 20 Apr 1996 19:19:44 +0200 (MET DST) Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Length: 940 Lines: 29 -----BEGIN PGP SIGNED MESSAGE----- Hi all, once again a bunch of kernel patches. I think I'll catch up with Linus kernel releases during this weakend or shortly after ... As usual you can get these patches from ftp.fnet.fr and it's mirrors. Until someone moved them online you can get from private/. Ralf 3e404f43bb3b4345d38488ceb649efce b linux-1.3.77-1.3.78.diffs.gz de50ba86a67fce4939142a039e404c4b b linux-1.3.78-1.3.79.diffs.gz 360f098d98d8821dd9e611ca6d2df83b b linux-1.3.79-1.3.80.diffs.gz 2ad8a460856a2cdd68539de1ff9cecc1 b linux-1.3.80-1.3.81.diffs.gz 4a01bd757fa0f3c7a207aa5bbac4987e b linux-1.3.81-1.3.82.diffs.gz -----BEGIN PGP SIGNATURE----- Version: 2.6.3i Charset: latin1 iQCVAwUBMXkcr0ckbl6vezDBAQG8pAQArhGNn8rQp5sqBf5z1VtQN+l1agbV28FJ 2vKDpDOeeN2MyDGqTeXBD8IHVlXF60ZMg2Aw55FTRuhvtnEQA270/rLEf8hD25q+ BnYQt4DGiRWIeJBEwNhQFyXGiXIk4tv39AhIEkCvMfQ7SqPphOLTlgkqLiwVEZCl o3ZLaci8Hk4= =NOYW -----END PGP SIGNATURE----- From bwadswor@cs.tamu.edu Sat Apr 20 21:09:05 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id VAA09694 for ; Sat, 20 Apr 1996 21:09:04 +0200 (MET DST) Received: from cs.tamu.edu (clavin.cs.tamu.edu) by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA13982; Sat, 20 Apr 1996 21:08:58 +0200 (MET) Received: from dasher.cs.tamu.edu (2336@dasher.cs.tamu.edu [128.194.135.90]) by cs.tamu.edu (8.6.10/8.6.4) with ESMTP id OAA17187 for ; Sat, 20 Apr 1996 14:08:24 -0500 From: Brandon S Wadsworth Received: (bwadswor@localhost) by dasher.cs.tamu.edu (8.6.12/8.6.4) id OAA29277 for linux-mips@fnet.fr; Sat, 20 Apr 1996 14:07:48 -0500 Date: Sat, 20 Apr 1996 14:07:48 -0500 Message-Id: <199604201907.OAA29277@dasher.cs.tamu.edu> To: linux-mips@fnet.fr Subject: help with mips magnum r4000. Content-Length: 132 Lines: 4 i need help with the instal of linux on a mips magnum r4000 sc-50.. anyhelp of anykind would be much apprecatiated... thanks. scott From linux@mailhost.uni-koblenz.de Sun Apr 21 00:58:10 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id AAA14429 for ; Sun, 21 Apr 1996 00:58:09 +0200 (MET DST) Received: from informatik.uni-koblenz.de (mailhost.uni-koblenz.de) by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA16830; Sun, 21 Apr 1996 00:57:59 +0200 (MET) Received: from ozzy (ozzy.uni-koblenz.de [141.26.5.8]) by informatik.uni-koblenz.de (8.7.4/8.6.9) with SMTP id AAA10763 for ; Sun, 21 Apr 1996 00:58:00 +0200 (MET DST) From: Systemkennung Linux Message-Id: <199604202258.AAA10763@informatik.uni-koblenz.de> Received: by ozzy (SMI-8.6/KO-2.0) id AAA04799; Sun, 21 Apr 1996 00:56:20 +0200 Subject: Memory managment & DMA To: linux-mips@fnet.fr Date: Sun, 21 Apr 1996 00:56:19 +0200 (MET DST) Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Length: 1058 Lines: 24 Hi all, I'm currently collecting data about the memory architecture of various targets for Linux/MIPS. Background is tht the current Linux/MIPS memory handling is basicly only suited for MIPS systems with one continuous chunk of memory; it also makes some assumptions about the address of the available memory. These assumptions fail eg. for DECstation, rPC44 and supposedly SGI - in short everything that isn't similar to a Magnum. Could people which such machines please send me a description of their machine's memory layout? Another problem is the DMA. The NCR53C94 is a very polular SCSI chip among the various MIPS machines; Sparcs also use chips from that family. As large as the number of machines that use these NCR chips is the number of different DMA engines. We therefore need a generic NCR driver that can be efficiently be used with all these DMA engines. I know that there are some pretty wiered of these DMA engines out there. So again here the request to post descriptions or pointers to such describing the DMA engine. Ralf From imp@rover.village.org Sun Apr 21 01:32:29 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id BAA15504 for ; Sun, 21 Apr 1996 01:32:28 +0200 (MET DST) Received: from rover.village.org by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA17229; Sun, 21 Apr 1996 01:32:21 +0200 (MET) Received: from localhost (localhost [127.0.0.1]) by rover.village.org (8.7.5/8.6.6) with SMTP id RAA22177 for ; Sat, 20 Apr 1996 17:31:52 -0600 (MDT) Message-Id: <199604202331.RAA22177@rover.village.org> To: linux-mips@fnet.fr Subject: Re: Memory managment & DMA In-Reply-To: Your message of Sun, 21 Apr 1996 00:56:19 +0200 Date: Sat, 20 Apr 1996 17:31:52 -0600 From: Warner Losh Content-Length: 816 Lines: 17 : Could people which such machines please send me a description of their : machine's memory layout? On the rPC44, you have all the memory starting at 0x80000000. The ISA bus creates a hole in the memory from 0x800a0000 to 0x80100000. DMA, as far as I know, is bog standard PC DMA. I suspect that the DMA engine that is in 1.3.62 is non-functional for my machine. The address space is repeased at 0xaxxxxxxx for uncached access, just like all the other MIPSen. The rPC44 doesn't have any builtin hardware, as far as I can tell. Well, that's not entirely true. There appears to be connectors on the mb for things like IDE, Floppy, Serial, etc. However, I'm not sure how they are enabled as my machine has a multi-function card in it for such things. If you need more details, please feel free to ask. Warner From ryan@ravage.labs.gmu.edu Mon Apr 21 20:01:55 1997 Received-Date: Mon, 21 Apr 1997 20:01:54 +0200 (MET DST) Received: from ravage.labs.gmu.edu ([129.174.166.55]) by Utopia.EUnet.fr (5.65c8d/EUNET-4.2.3) via EUnet-France id AA12713; Mon, 21 Apr 1997 19:58:48 +0200 (MET) Received: from localhost (ryan@localhost) by ravage.labs.gmu.edu (8.8.3/8.8.3) with SMTP id CAA04634 for ; Sun, 21 Apr 1996 02:03:20 -0400 Date: Sun, 21 Apr 1996 02:03:19 -0400 (EDT) From: Ryan Rafferty To: linux-mips@guadalquivir.fnet.fr Subject: Re: Magnums In-Reply-To: <199704210613.IAA09971@informatik.uni-koblenz.de> Message-Id: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Content-Length: 1302 Lines: 31 On Mon, 21 Apr 1997, Systemkennung Linux wrote: > Hi, > > and then it hangs. I assume this is because /sbin/getty keeps dying for > > some reason. Any help? Could I be running a kernel that's too old, or a > > root filesystem image that's too old? > > The root filesystem is too old. Try to put all the other packages > that are available on FNet on top. I tried updating as many packages as I could. Unfortunately, the old root filesystem seems to have been based on a Debian installation but many of the updated packages on Fnet are in RPM format--since I am using Debian on my Intel box that's serving the Magnum, I have no way of expanding some of the updated files. In any case, I tried to boot the Magnum again using the updated files (including the latest glibc-121296.tgz) and this time I got even nastier errors than before =). It complains about illegal instructions and now memory address errors, too. arrrgh. My largest concern is whether or not I even updated the correct packages. Ralf, could you tell me which packages are vital to update? I'm really anxious to get to a shell prompt on this machine =). My ultimate goal is to achieve a self-hosting system so I can begin work on a driver for the NCR 53C90 chipset that comes on the Jazz boxes. Thanks, Ryan > Ralf > From ryan@ravage.labs.gmu.edu Mon Apr 21 23:21:14 1997 Received-Date: Mon, 21 Apr 1997 23:21:13 +0200 (MET DST) Received: from ravage.labs.gmu.edu ([129.174.166.55]) by Utopia.EUnet.fr (5.65c8d/EUNET-4.2.3) via EUnet-France id AA17589; Mon, 21 Apr 1997 23:18:05 +0200 (MET) Received: from localhost (ryan@localhost) by ravage.labs.gmu.edu (8.8.3/8.8.3) with SMTP id FAA06458 for ; Sun, 21 Apr 1996 05:22:46 -0400 Date: Sun, 21 Apr 1996 05:22:46 -0400 (EDT) From: Ryan Rafferty To: linux-mips@guadalquivir.fnet.fr Subject: GGI on Alpha In-Reply-To: <199704171005.MAA20584@informatik.uni-koblenz.de> Message-Id: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Content-Length: 346 Lines: 8 I got my Alpha system running the 2.0.22 kernel and it's a great Linux box. I've been following the advance of the GGI project for a while, and I've got a question--is there any underlying reason why the GGI kernel modifications wouldn't compile on non-Intel versions of Linux, or is it just a case of not having been tried yet? Ryan Rafferty From ryan@ravage.labs.gmu.edu Tue Apr 22 18:31:07 1997 Received-Date: Tue, 22 Apr 1997 18:31:07 +0200 (MET DST) Received: from ravage.labs.gmu.edu ([129.174.166.55]) by Utopia.EUnet.fr (5.65c8d/EUNET-4.2.3) via EUnet-France id AA24747; Tue, 22 Apr 1997 18:27:58 +0200 (MET) Received: from localhost (ryan@localhost) by ravage.labs.gmu.edu (8.8.3/8.8.3) with SMTP id AAA10166 for ; Mon, 22 Apr 1996 00:32:35 -0400 Date: Mon, 22 Apr 1996 00:32:35 -0400 (EDT) From: Ryan Rafferty To: linux-mips@guadalquivir.fnet.fr Subject: Re: GGI on Alpha In-Reply-To: <199704220657.IAA21699@informatik.uni-koblenz.de> Message-Id: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Content-Length: 1041 Lines: 24 On Tue, 22 Apr 1997, Systemkennung Linux wrote: > The real problem is probably more that GGI has been discussed very > religiously with one of the core arguments against GGI being that > it was implying a lot of kernel bloat. GGI has been cured of at least > this main desease. I'm not shure, but I think for the non-Intel world > the core GGI problem might now be that we can't use the card's BIOSes. Again, I must apologize for sending that e-mail to the wrong mailing list--but along the same lines, it has been my hope that GGI would be easily portable to the Linux architectures, including of course Linux/MIPS. I realize that this isn't the right forum to discuss GGI matters, but I'd like to clarify that GGI currently has no plans to rely on VGA BIOSes and such. In fact, it should be quite easy to port GGI kernel stuff to Linux/MIPS and the graphics options availible for MIPS machines, like g364 etc., once the project is more stable (GGI, I mean). Anyway, enough off-topic rambling for me =). Ryan Rafferty > Ralf > From linux@mailhost.uni-koblenz.de Mon Apr 22 12:27:48 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id MAA00445 for ; Mon, 22 Apr 1996 12:27:47 +0200 (MET DST) Received: from informatik.uni-koblenz.de (mailhost.uni-koblenz.de) by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA14364; Mon, 22 Apr 1996 12:27:21 +0200 (MET) Received: from ozzy (ozzy.uni-koblenz.de [141.26.5.8]) by informatik.uni-koblenz.de (8.7.4/8.6.9) with SMTP id MAA18910 for ; Mon, 22 Apr 1996 12:27:18 +0200 (MET DST) From: Systemkennung Linux Message-Id: <199604221027.MAA18910@informatik.uni-koblenz.de> Received: by ozzy (SMI-8.6/KO-2.0) id MAA07581; Mon, 22 Apr 1996 12:25:39 +0200 Subject: Exceptions To: linux-mips@fnet.fr Date: Mon, 22 Apr 1996 12:25:38 +0200 (MET DST) Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Length: 691 Lines: 15 Hi, not enough with the recently reported address error exceptions. There is also a whole bunch of userland stuff that throws ADEL/ADES exceptions. I'm fed up of fixing this "Intel desease". 1.3.84 has a new exception handler for ADEL/ADES exceptions that fixes this problem for usermode stuff. For the kernel the policy is still to panic() on address error. As in other MIPS operating systems with this feature the exception handler can be switched on/off as per process option. For now the default is on until we've fixed all the user programs. I'll also add the option to the kernel to print warnings additionally. These can be logged to syslog and later on be analysed. Ralf From linux@mailhost.uni-koblenz.de Mon Apr 22 16:02:42 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id QAA04871 for ; Mon, 22 Apr 1996 16:02:41 +0200 (MET DST) Received: from informatik.uni-koblenz.de (mailhost.uni-koblenz.de) by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA19563; Mon, 22 Apr 1996 16:02:01 +0200 (MET) Received: from ozzy (ozzy.uni-koblenz.de [141.26.5.8]) by informatik.uni-koblenz.de (8.7.4/8.6.9) with SMTP id QAA00918 for ; Mon, 22 Apr 1996 16:01:10 +0200 (MET DST) From: Systemkennung Linux Message-Id: <199604221401.QAA00918@informatik.uni-koblenz.de> Received: by ozzy (SMI-8.6/KO-2.0) id PAA08074; Mon, 22 Apr 1996 15:59:30 +0200 Subject: Magnum 4000 configuration disk To: linux-mips@fnet.fr Date: Mon, 22 Apr 1996 15:59:30 +0200 (MET DST) Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Length: 359 Lines: 11 Hi all, during the last time I found that one quite common problem among Magnum users with Linux/MIPS is that they don't have the configuration disk that is required to reconfigure the machine to little endian byte order anymore. Could therefore someone please send me a disk image of this disk so that I can help these people? Thanks in advance, Ralf From ralf@Julia.DE Tue Apr 23 21:11:37 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id VAA27065 for ; Tue, 23 Apr 1996 21:11:13 +0200 (MET DST) Received: from alles.intern.julia.de ([194.221.49.2]) by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA20129; Tue, 23 Apr 1996 21:10:47 +0200 (MET) Received: from pris.intern.julia.de (pris.intern.julia.de [194.221.49.37]) by alles.intern.julia.de (8.7.5/8.7.3) with ESMTP id VAA30399 for ; Tue, 23 Apr 1996 21:11:00 +0200 From: Ralf Baechle Received: (from ralf@localhost) by pris.intern.julia.de (8.7.5/8.7.3) id VAA04355 for linux-mips@fnet.fr; Tue, 23 Apr 1996 21:10:02 +0200 (MET DST) Message-Id: <199604231910.VAA04355@pris.intern.julia.de> Subject: Announce Linux/MIPS 1.3.84 To: linux-mips@fnet.fr Date: Tue, 23 Apr 1996 21:10:02 +0200 (MET DST) Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Length: 983 Lines: 28 -----BEGIN PGP SIGNED MESSAGE----- Hi all, again some more patches. Linux/MIPS 1.3.84 contains the address error handler to fix adel/ades exceptions. Have fun, Ralf 3e404f43bb3b4345d38488ceb649efce b linux-1.3.77-1.3.78.diffs.gz de50ba86a67fce4939142a039e404c4b b linux-1.3.78-1.3.79.diffs.gz 360f098d98d8821dd9e611ca6d2df83b b linux-1.3.79-1.3.80.diffs.gz 2ad8a460856a2cdd68539de1ff9cecc1 b linux-1.3.80-1.3.81.diffs.gz 4a01bd757fa0f3c7a207aa5bbac4987e b linux-1.3.81-1.3.82.diffs.gz 4bfc188b5f1a7f652b344e0a76edfed8 b linux-1.3.82-1.3.83.diffs.gz e238ed9e3004af54a9d39cd17c1087b5 b linux-1.3.83-1.3.84.diffs.gz fbcedb2520540d3eb5fb7e82f16facae b linux-1.3.84.tar.gz -----BEGIN PGP SIGNATURE----- Version: 2.6.3i Charset: latin1 iQCVAwUBMX0rCUckbl6vezDBAQF1eAP/YlpEN41p/7QIKkNajdi55y9COp1ARhCA eT9LtKauUIOZdbtbtgeIWiMGlN9K9YUCRJafGVOb94HvzcrSXdk9ypp0skqEb771 RDss0wuMYSDsUIUhW6gdJZhV20oHDiv1nmRJETDxDGXqrPI/U8cs3vw2ObGwWgz+ pARu4iFlk5g= =IqSB -----END PGP SIGNATURE----- From imp@village.org Wed Apr 24 01:15:54 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id BAA02057 for ; Wed, 24 Apr 1996 01:15:54 +0200 (MET DST) Received: from rover.village.org by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA23499; Wed, 24 Apr 1996 01:15:41 +0200 (MET) Received: (from imp@localhost) by rover.village.org (8.7.5/8.6.6) id RAA10768 for linux-mips@fnet.fr; Tue, 23 Apr 1996 17:15:44 -0600 (MDT) Date: Tue, 23 Apr 1996 17:15:44 -0600 (MDT) From: Warner Losh Message-Id: <199604232315.RAA10768@rover.village.org> To: linux-mips@fnet.fr Subject: FYI: Can't disable /proc Content-Length: 175 Lines: 4 You can't disable /proc in 1.3.82. There are references to it in mips.o and scsi.o. Time to go off and fix them. Warner "Why is it so hard to make a 640k MIPS kernel" Losh From linux@mailhost.uni-koblenz.de Wed Apr 24 15:22:16 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id PAA19882 for ; Wed, 24 Apr 1996 15:22:14 +0200 (MET DST) Received: from informatik.uni-koblenz.de (mailhost.uni-koblenz.de) by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA07317; Wed, 24 Apr 1996 15:21:57 +0200 (MET) Received: from ozzy (ozzy.uni-koblenz.de [141.26.5.8]) by informatik.uni-koblenz.de (8.7.4/8.6.9) with SMTP id PAA14228 for ; Wed, 24 Apr 1996 15:21:33 +0200 (MET DST) From: Systemkennung Linux Message-Id: <199604241321.PAA14228@informatik.uni-koblenz.de> Received: by ozzy (SMI-8.6/KO-2.0) id PAA14248; Wed, 24 Apr 1996 15:19:50 +0200 Subject: Re: FYI: Can't disable /proc To: linux-mips@fnet.fr Date: Wed, 24 Apr 1996 15:19:49 +0200 (MET DST) In-Reply-To: <199604232315.RAA10768@rover.village.org> from "Warner Losh" at Apr 23, 96 05:15:44 pm Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Length: 749 Lines: 18 Hi, > You can't disable /proc in 1.3.82. There are references to it in mips.o > and scsi.o. Time to go off and fix them. > > Warner "Why is it so hard to make a 640k MIPS kernel" Losh I ussually calculate a size factor between 1.4 and 1.7 between the size of Intel and MIPS code. So a kernel of just 640k can contain just enough to be almost worthless. I'll add a hack for the rPC44 that will link the kernel to address KSEG0 + 1mb, copy the exception handlers down to KSEG0, and then again declare the memory from KSEG0 + PAGE_SIZE to KSEG0 + 640 free again. All you need to do is to change Milo to copy the kernel to KSEG0 + 1mb, not KSEG0. Sounds easy, doesn't it :-) Ralf "Always trying to support hardware he's never seen" Baechle From linux@mailhost.uni-koblenz.de Wed Apr 24 22:43:01 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id WAA28445 for ; Wed, 24 Apr 1996 22:43:00 +0200 (MET DST) Received: from informatik.uni-koblenz.de (mailhost.uni-koblenz.de) by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA15460; Wed, 24 Apr 1996 22:42:42 +0200 (MET) Received: from ozzy (ozzy.uni-koblenz.de [141.26.5.8]) by informatik.uni-koblenz.de (8.7.4/8.6.9) with SMTP id WAA28934 for ; Wed, 24 Apr 1996 22:42:31 +0200 (MET DST) From: Systemkennung Linux Message-Id: <199604242042.WAA28934@informatik.uni-koblenz.de> Received: by ozzy (SMI-8.6/KO-2.0) id WAA17401; Wed, 24 Apr 1996 22:40:48 +0200 Subject: rPC44 patches To: linux-mips@fnet.fr Date: Wed, 24 Apr 1996 22:40:47 +0200 (MET DST) Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Length: 3384 Lines: 71 -----BEGIN PGP SIGNED MESSAGE----- Hi all, The patch below implements the kernel part of the support for >640k kernel on the rPC44 as I suggested this afternoon. As goodie this patch also removes an ancient and old hack in free_pages() that I once had added to get the system stable. It's no longer need; actually it's useless since several kernel patches. Happy patching and (Linus emulator on) as always try this one; it's good for you, Ralf begin 644 linux-1.3.85.diffs.gz M'XL("$J/?C$"`VQI;G5X+3$N,RXX-2YD:69FMP'G3 MOYH`?;H]^$=E\Z@*YA2V>T=^77HI3[*ZSSP=6"29;NP9GGPL+#N?[@;MMFZ\ M9#]C3QE83V>6;=%'-Y19>+6:M5:+?1Z@4_E%AX$2U32E.BCTWI@)3YCNORKA M#RR-6=0(F>O7)\_EHR3T8NI+TIO\8]JM5M<^[;:;+^:_;&&O"#K=]O&V"-I6 MS;;`H$M'U<`;S!2/&8QN+X9._^:Z/]&L)\MB%BZMXV8TCG0#CF`:,LB6`N1J M?B\B"?=LQF.'/7DLR<"-?0KT^G;NI@^0H4+JQC-4"]T,N`1/))SYREHFX.-D M>&6!B$%F;IHMDCHM-'1C^-O=[:=I9==^50=-JU.&=6@<:3HH?]`B_LG$]1@$ M(E4;\CAC:;I`'T+T*6*IK.MF[OT*"Q>Q*CPWPR"A1X\5:JTVZLL6NI1;!%TS&[S*8L0Q\GC(OBU:X@>NC_77(\OBA3AX;+N+BJ$HKC_%& MN`[YT3/:"8.$>R]DALE35E*FG'?OD<0\X;,Z1;9!P<:DXS<%[N87%$."B6HP M_.T&>CVPNM5A M^6]ZHT+*\'K2=^X7$EVWR4$@+!H$Q8%(5H=)3H&7,,^S=Q*0TI&#?"8S_$'\ ML&;'W$Q!/C*G95\0WY2XAD1Y6E"2W*42)+#5\[*)[242^E]168>CJK)9*R>DMLY;5:5R)1U5GD&TD.&^8B%9@U\&&,PA M::".XBUT<,*PHV`,`G<1(2^KDI5JB>B-XE'A/>L]/^^U[/?<,%3HD:.<+7&A M1H77BA`[R.AU&$!3"WGY MJZKPD@45Q=9]979Z,,MS(2E'/DNPIB6(0&UU+]P4\9\WJ`,/M\N-@(O=HJ-I#8-%_GPCV.[A<-[]_CXQ2:QU=[M#Z?= MIMUMM[;]H6E1?\#OO#_LMX;[Z"&G\\W,5R;V>R&02@)Q@/._TS[\N?M,4RE6 ME&=8)N3:R7'-+L:(94CO5)5LGL`YA'P6.D@=(EU55:,@'D$J37X?]>^<\2<2 MJ_Y17Z,32=.B$M)(V^C!7?]JZ$RN_STD/OE:@MAE,3&FQ+RP9.\0N1LS^92F MMD7^$S@T>2'Q7\K8#I'FZT$JYMN-B"[5S>WEY60X!0-G2)=>2,"=N3S>0O3Y M/O)M7\"3&_F9>KO&Z5"E<7;Q\[>XC.XV9/:Q0R],#5Q,'H&,_N:.W@AL'2Z]L[;,%ZI@R69S%F=8P3BTJM::MS8G$PVJA4#NS5\*0CP3.,=]=V@RB]@ZCJH+ MAQX\AS5S@U4B1XA31RYQK*(CR@($Q?,`:W;W.9VO@VAK@]%9'W#[)AAQF56> MJEB:>;$8>$/S^;>^0?$AGE[_)EA1KP;VR-(53@38/!0\5),@]R6BZ&*A9A5Z MMGYA\M04,2M9DBB.*%Q(S)2:-99<-Q8X'.! M.]-C5K*5SRHQ6^(W:=9W5QO;F[V1`SO%7_A_L9XP-E'7,#1.G,+;'@Y`\H$H KIQQJ!)`-Y^<@4I^E[[>LINZ1U\:?G-%PY-Q<3Z83TR:$_@WH+G"7=!(``($H ` end -----BEGIN PGP SIGNATURE----- Version: 2.6.2 iQCVAgUBMX6RzUckbl6vezDBAQHcNwP/ehccOM1XY4zhOfIxzCHkROJL1f09LT11 +R6In3QYATXumSZXkjznng955visJ29DboEI4l+5wjHasW3UmIZ/pIssCotk9tKf 4Prztvk1CF+Y2/nG5Tq8stZ0LlIF6qYquk7UP6rZiWW/5UOvo4gYEaiCUZA6o/G+ IqjBh4wJ2nk= =3JR4 -----END PGP SIGNATURE----- From hannon@Crissy.Stanford.EDU Thu Apr 25 10:20:40 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id KAA13858 for ; Thu, 25 Apr 1996 10:20:39 +0200 (MET DST) Received: from crissy.Stanford.EDU by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA26274; Thu, 25 Apr 1996 10:20:27 +0200 (MET) Received: by Crissy.Stanford.EDU (8.6.8/25-LAGUNITA-eef) id BAA25020; Thu, 25 Apr 1996 01:21:13 -0700 From: Steph Hannon Message-Id: <199604250821.BAA25020@Crissy.Stanford.EDU> Subject: Linux boot question To: linux-mips@fnet.fr Date: Thu, 25 Apr 96 1:21:13 PDT In-Reply-To: <199604191849.UAA01390@pris.intern.julia.de>; from "Ralf Baechle" at Apr 19, 96 8:49 pm X-Mailer: ELM [version 2.3 PL11] Content-Length: 495 Lines: 11 Hi, I am trying to determine the exact sequence of events that occurs when Linux is booted on a MIPS machine. I am working on a port of Linux to a hardware simulation environment and I need to figure out where Linux is polling the hardware for details, so we can make sure the simulator reponds correctly. I have been searching through the Linux Bible without a lot of luck. Do you have any suggestions about where I might look for more documentation on the boot sequence? Stephanie Hannon From imp@rover.village.org Thu Apr 25 16:29:33 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id QAA21044 for ; Thu, 25 Apr 1996 16:29:32 +0200 (MET DST) Received: from rover.village.org by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA04726; Thu, 25 Apr 1996 16:29:17 +0200 (MET) Received: from localhost (localhost [127.0.0.1]) by rover.village.org (8.7.5/8.6.6) with SMTP id IAA19253 for ; Thu, 25 Apr 1996 08:29:23 -0600 (MDT) Message-Id: <199604251429.IAA19253@rover.village.org> To: linux-mips@fnet.fr Subject: Re: Linux boot question In-Reply-To: Your message of Thu, 25 Apr 1996 01:21:13 PDT Date: Thu, 25 Apr 1996 08:29:22 -0600 From: Warner Losh Content-Length: 389 Lines: 11 : I have been searching through the Linux Bible without a lot of luck. Do you : have any suggestions about where I might look for more documentation on the : boot sequence? I'd look at milo :-) Right now the Linux/MIPS port gets loaded by some BIOS on the machine, then starts running. arch/mips/kernel/head.S is where the whole process starts. You might start looking there. Warner From Stoned.Elipot@univ-evry.fr Thu Apr 25 17:27:34 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id RAA22216 for ; Thu, 25 Apr 1996 17:27:33 +0200 (MET DST) Received: from amiante.univ-evry.fr by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA06116; Thu, 25 Apr 1996 17:27:22 +0200 (MET) Received: from bandsept.univ-evry.fr (bandsept.univ-evry.fr [192.93.214.19]) by amiante.univ-evry.fr with ESMTP (8.6.12/96.04.09/louis-blanc); id RAA15520; Thu, 25 Apr 1996 17:27:30 +0200 Received: from localhost.univ-evry.fr (localhost.univ-evry.fr [127.0.0.1]) by bandsept.univ-evry.fr with SMTP (8.6.12/94.01.26); id RAA11812; Thu, 25 Apr 1996 17:26:47 +0200 Message-Id: <199604251526.RAA11812@bandsept.univ-evry.fr> To: linux-mips@fnet.fr Subject: Re: Linux boot question In-Reply-To: Your message of "Thu, 25 Apr 1996 01:21:13 PDT." <199604250821.BAA25020@Crissy.Stanford.EDU> X-It: bandsept (wrong OS, but right CPU) X-Mailer: MH 6.8.3, MH-E 5.0.2 Date: Thu, 25 Apr 1996 17:26:40 +0200 From: Stoned Elipot Content-Length: 1276 Lines: 27 Hi all (I'm finally back from holidays:) >>>>> "Steph" == Steph Hannon writes: Steph> Hi, I am trying to determine the exact sequence of events that Steph> occurs when Linux is booted on a MIPS machine. I am working on Warner already answered this but I want to add some precisions. Milo for the ARC compliant boxes figure out various hardware parameters of the box and pass them to the kernel. As the last versions of Milo/kernel, this is done by putting this infos at the top of memory (we call them 'tags'). The kernel functions responsible for snarfing this infos are in arch/mips/kernel/setup.c, they are called bi_TagFind() and bi_EarlySnarf() and are for now called in head.S:kernel_entry(), and setup.c:setup_arch(). For non-ARC compliant box the kernel itself should do this hardware poking, create the tags at the end of memory. As an example, for DECs - Paul is working on that - it should be done in decstation.S:dec_entry() before jumping to head.S:kernel_entry() from where tags decoding/use can begin. In Milo, tags creation take place the functions you can find in bitags.c Hopefully all this would be a little bit more commented in the future Milo 0.27 which is a kind of rewritre of the current Milo. Cheers, Stoned. From linux@mailhost.uni-koblenz.de Fri Apr 26 00:04:46 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id AAA00235 for ; Fri, 26 Apr 1996 00:04:46 +0200 (MET DST) Received: from informatik.uni-koblenz.de (mailhost.uni-koblenz.de) by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA12476; Fri, 26 Apr 1996 00:04:22 +0200 (MET) Received: from ozzy (ozzy.uni-koblenz.de [141.26.5.8]) by informatik.uni-koblenz.de (8.7.4/8.6.9) with SMTP id AAA00466 for ; Fri, 26 Apr 1996 00:04:21 +0200 (MET DST) From: Systemkennung Linux Message-Id: <199604252204.AAA00466@informatik.uni-koblenz.de> Received: by ozzy (SMI-8.6/KO-2.0) id AAA19125; Fri, 26 Apr 1996 00:02:38 +0200 Subject: Re: Linux boot question To: linux-mips@fnet.fr Date: Fri, 26 Apr 1996 00:02:37 +0200 (MET DST) In-Reply-To: <199604251526.RAA11812@bandsept.univ-evry.fr> from "Stoned Elipot" at Apr 25, 96 05:26:40 pm Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Length: 776 Lines: 20 Hi all, > Hi all (I'm finally back from holidays:) Hope you had a lot fun :-) > For non-ARC compliant box the kernel itself should do this hardware > poking, create the tags at the end of memory. As an example, for DECs > - Paul is working on that - it should be done in > decstation.S:dec_entry() before jumping to head.S:kernel_entry() from > where tags decoding/use can begin. Just some additional words about head.S - with more and more platforms that Linux/MIPS gets ported to (ranging from evalutation board over RISC PCs to workstations (If only SNI'd deliver what they promise ...) and a whole bunch of processor types that differ a lot I intend to eleminate most of the startup code in head.S and place the exception handler code into multiple files. Ralf From lm@neteng.engr.sgi.com Fri Apr 26 00:16:30 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id AAA00479 for ; Fri, 26 Apr 1996 00:16:29 +0200 (MET DST) Received: from sgi.sgi.com (SGI.COM) by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA12841; Fri, 26 Apr 1996 00:16:15 +0200 (MET) Received: from cthulhu.engr.sgi.com by sgi.sgi.com via ESMTP (950405.SGI.8.6.12/910110.SGI) id PAA22819; Thu, 25 Apr 1996 15:16:00 -0700 Received: from neteng.engr.sgi.com by cthulhu.engr.sgi.com via ESMTP (950511.SGI.8.6.12.PATCH526/911001.SGI) id PAA29597; Thu, 25 Apr 1996 15:15:50 -0700 Received: from localhost (lm@localhost) by neteng.engr.sgi.com (950413.SGI.8.6.12/960327.SGI.AUTOCF) via SMTP id PAA11845; Thu, 25 Apr 1996 15:15:50 -0700 Message-Id: <199604252215.PAA11845@neteng.engr.sgi.com> To: linux-mips@fnet.fr Cc: davem@caip.rutgers.edu From: lm@gate1-neteng.engr.sgi.com (Larry McVoy) Subject: Linux/SGI coming Date: Thu, 25 Apr 1996 15:15:50 -0700 Sender: lm@neteng.engr.sgi.com Content-Length: 925 Lines: 21 Hi folks, some reinforcements are showing up soon. SGI has hired David Miller, the guy that did most of the work on sparc linux, to come in as a summer intern and try and do the same for SGI machines. SGI is being pretty cooperative with this, in terms of supplying information and equipment for David to work on. *All* results will be GPLed (SGI will retain ownership so we can use it under a different copyright if we want) and all results will go out to the net. We will want to cooperate and coordinate with the existing Linux/MIPS team, seeing as you have accomplished so much already. I'm not sure how much overlap there is, since the SGI devices are probably different than the MIPS pc style devices, and the byteorder is different, but I'll bet there is a lot. Please add David Miller, davem@caip.rutgers.edu, to this alias so he can be made aware of your efforts. Thanks mucho and happy Linux to ya, --lm From imp@rover.village.org Fri Apr 26 04:48:07 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id EAA07188 for ; Fri, 26 Apr 1996 04:48:06 +0200 (MET DST) Received: from rover.village.org by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA15935; Fri, 26 Apr 1996 04:47:51 +0200 (MET) Received: from localhost (localhost [127.0.0.1]) by rover.village.org (8.7.5/8.6.6) with SMTP id UAA22258; Thu, 25 Apr 1996 20:47:59 -0600 (MDT) Message-Id: <199604260247.UAA22258@rover.village.org> To: linux-mips@fnet.fr Subject: Re: Linux/SGI coming Cc: davem@caip.rutgers.edu In-Reply-To: Your message of Thu, 25 Apr 1996 15:15:50 PDT Date: Thu, 25 Apr 1996 20:47:59 -0600 From: Warner Losh Content-Length: 1421 Lines: 28 : some reinforcements are showing up soon. SGI has hired David : Miller, the guy that did most of the work on sparc linux, to come in : as a summer intern and try and do the same for SGI machines. SGI is : being pretty cooperative with this, in terms of supplying information : and equipment for David to work on. *All* results will be GPLed (SGI : will retain ownership so we can use it under a different copyright if : we want) and all results will go out to the net. Cool! Glad to see more developers tossed onto the Linux/MIPS project. It is getting very close to being at the critical mass stage. The kernel seems to be stable for some people, but not for others. More people actively working on it can't hurt! : We will want to cooperate and coordinate with the existing : Linux/MIPS team, seeing as you have accomplished so much already. : I'm not sure how much overlap there is, since the SGI devices are probably : different than the MIPS pc style devices, and the byteorder is different, : but I'll bet there is a lot. Still, it is good to have people helping out. If nothing else, SGI might be a good place to host at least a read-only CVS archive of what we're doing. Or maybe not, it isn't required (and I'm not sure if the other Linux/MIPS people are keen on the idea). I know the *BSD people do well in this area.... Welcome on board David! Good work on the Linux/Sparc port, btw. Warner From linux@mailhost.uni-koblenz.de Fri Apr 26 12:04:43 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id MAA00462 for ; Fri, 26 Apr 1996 12:04:40 +0200 (MET DST) Received: from informatik.uni-koblenz.de (mailhost.uni-koblenz.de) by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA23586; Fri, 26 Apr 1996 12:04:16 +0200 (MET) Received: from ozzy (ozzy.uni-koblenz.de [141.26.5.8]) by informatik.uni-koblenz.de (8.7.4/8.6.9) with SMTP id MAA00683 for ; Fri, 26 Apr 1996 12:04:11 +0200 (MET DST) From: Systemkennung Linux Message-Id: <199604261004.MAA00683@informatik.uni-koblenz.de> Received: by ozzy (SMI-8.6/KO-2.0) id MAA19679; Fri, 26 Apr 1996 12:02:28 +0200 Subject: Re: Linux/SGI coming To: linux-mips@fnet.fr Date: Fri, 26 Apr 1996 12:02:27 +0200 (MET DST) In-Reply-To: <199604252215.PAA11845@neteng.engr.sgi.com> from "Larry McVoy" at Apr 25, 96 03:15:50 pm Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Length: 1133 Lines: 27 Hi all, > some reinforcements are showing up soon. A dream comes true ... No more me who writes almost all that code ... SGI has hired David > Miller, the guy that did most of the work on sparc linux, to come in > as a summer intern and try and do the same for SGI machines. SGI is > being pretty cooperative with this, in terms of supplying information > and equipment for David to work on. *All* results will be GPLed (SGI > will retain ownership so we can use it under a different copyright if > we want) and all results will go out to the net. > > We will want to cooperate and coordinate with the existing > Linux/MIPS team, seeing as you have accomplished so much already. > I'm not sure how much overlap there is, since the SGI devices are probably > different than the MIPS pc style devices, and the byteorder is different, > but I'll bet there is a lot. > > Please add David Miller, davem@caip.rutgers.edu, to this alias so he > can be made aware of your efforts. Dave and some of his Sparc collegues are subscribed since quite some time to this list. Ralf From linux@mailhost.uni-koblenz.de Fri Apr 26 12:56:30 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id MAA00786 for ; Fri, 26 Apr 1996 12:56:30 +0200 (MET DST) Received: from informatik.uni-koblenz.de (mailhost.uni-koblenz.de) by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA23843; Fri, 26 Apr 1996 12:15:37 +0200 (MET) Received: from ozzy (ozzy.uni-koblenz.de [141.26.5.8]) by informatik.uni-koblenz.de (8.7.4/8.6.9) with SMTP id MAA01093 for ; Fri, 26 Apr 1996 12:15:42 +0200 (MET DST) From: Systemkennung Linux Message-Id: <199604261015.MAA01093@informatik.uni-koblenz.de> Received: by ozzy (SMI-8.6/KO-2.0) id MAA19705; Fri, 26 Apr 1996 12:13:58 +0200 Subject: Re: Linux/SGI coming To: linux-mips@fnet.fr Date: Fri, 26 Apr 1996 12:13:58 +0200 (MET DST) In-Reply-To: <199604260247.UAA22258@rover.village.org> from "Warner Losh" at Apr 25, 96 08:47:59 pm Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Length: 1094 Lines: 27 Hi all, > Cool! Glad to see more developers tossed onto the Linux/MIPS > project. It is getting very close to being at the critical mass > stage. The kernel seems to be stable for some people, but not for > others. I think the main problem is just the Magnum console - then it will be stable for the other people. The console sometimes overwrites CPU memory with data that actually should go into video memory ... > Still, it is good to have people helping out. If nothing else, SGI > might be a good place to host at least a read-only CVS archive of what > we're doing. Or maybe not, it isn't required (and I'm not sure if the > other Linux/MIPS people are keen on the idea). Well, I'v been the person that dislikes CVS most on this list. So you're probably talking about me .... This is only because I don't have the resources to use/run a CVS archive is a reasonable way. Otherwise I think this is a good thing. > Welcome on board David! Good work on the Linux/Sparc port, btw. Yes, indeed. We still have trouble booting it at all here but when it runs it runs ... Ralf From linux@mailhost.uni-koblenz.de Fri Apr 26 12:26:41 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id MAA00266 for ; Fri, 26 Apr 1996 12:26:41 +0200 (MET DST) Received: from informatik.uni-koblenz.de (mailhost.uni-koblenz.de) by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA24142; Fri, 26 Apr 1996 12:25:27 +0200 (MET) Received: from ozzy (ozzy.uni-koblenz.de [141.26.5.8]) by informatik.uni-koblenz.de (8.7.4/8.6.9) with SMTP id MAA01476; Fri, 26 Apr 1996 12:25:27 +0200 (MET DST) From: Systemkennung Linux Message-Id: <199604261025.MAA01476@informatik.uni-koblenz.de> Received: by ozzy (SMI-8.6/KO-2.0) id MAA19731; Fri, 26 Apr 1996 12:23:44 +0200 Subject: Re: Linux boot question To: zaitcev@lab.ipmce.su (Pete A. Zaitcev) Date: Fri, 26 Apr 1996 12:23:43 +0200 (MET DST) Cc: linux-mips@fnet.fr In-Reply-To: <199604260459.IAA00258@mellorn.lab.sun.mcst.ru> from "Pete A. Zaitcev" at Apr 26, 96 08:59:03 am Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Length: 942 Lines: 24 Hi Pete, > >to workstations (If only SNI'd deliver what they promise ...> > > Hello, Ralf, > > BTW, how is it going? RM series are produced by license in Moscow now, > that's why I'm asking. They are shown on any major exibition. Well, we still didn't receive the machine promised by SNI. So I continue porting Linux/MIPS to dead hardware (Acer Pica). Well, I consider a machine not available for sale in Europe/US dead hardware. This however doesn't mean that the project is dead - it just hasn't really started yet. A contract about SNI loaning a machine to Soft & Hard, Andy's company has been signed. SNI just hasn't fullfilled it's part yet. People at SNI seem to be pretty interested in Linux/MIPS for their machines, though. Apparently one of them has asked for a port to a RM400 which is a *SMP* machine. Well, I'd cut off at least a dozen fingers for such a beast :-) Probably SGI will score better than SNI ... Ralf From pab@cisco.com Fri Apr 26 12:56:50 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id MAA00810 for ; Fri, 26 Apr 1996 12:56:49 +0200 (MET DST) Received: from malcolm.rmnet.it by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA24919; Fri, 26 Apr 1996 12:56:31 +0200 (MET) Received: from [192.190.227.51] (paris-macip12.cisco.com [192.190.227.51]) by malcolm.rmnet.it (8.7.3/8.6.9) with SMTP id MAA17463 for ; Fri, 26 Apr 1996 12:56:07 +0200 X-Sender: pab@malcolm.rmnet.it Message-Id: Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Fri, 26 Apr 1996 12:57:02 +0200 To: linux-mips@fnet.fr From: pab@cisco.com (Paolo Bevilacqua) Subject: Re: Linux boot question Content-Length: 325 Lines: 14 Hi Stoned, >Hi all (I'm finally back from holidays:) good! (or bad :), and i'm in Paris until tomorrow, my phone number is 45.35.38.32, gimme a call as i don't have your number! byebye /pab -- Paolo Bevilacqua | | Tel: +39 6 50281230 cisco Systems Engineer ..:|||:..:|||:.. Fax: +39 6 5020016 From pab@cisco.com Fri Apr 26 13:02:31 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id NAA01131 for ; Fri, 26 Apr 1996 13:02:30 +0200 (MET DST) Received: from malcolm.rmnet.it by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA25269; Fri, 26 Apr 1996 13:02:15 +0200 (MET) Received: from [192.190.227.51] (paris-macip12.cisco.com [192.190.227.51]) by malcolm.rmnet.it (8.7.3/8.6.9) with SMTP id NAA17494 for ; Fri, 26 Apr 1996 13:01:53 +0200 X-Sender: pab@malcolm.rmnet.it Message-Id: Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Fri, 26 Apr 1996 13:02:49 +0200 To: linux-mips@fnet.fr From: pab@cisco.com (Paolo Bevilacqua) Subject: Re: Linux boot question Content-Length: 418 Lines: 16 I wrote: > >good! (or bad :), and i'm in Paris until tomorrow, my phone number is >45.35.38.32, gimme a call as i don't have your number! > Please excuse, this was supposed to be private! I'm into a kind of regression phase about the capacity of using common tools like mail :( /pab -- Paolo Bevilacqua | | Tel: +39 6 50281230 cisco Systems Engineer ..:|||:..:|||:.. Fax: +39 6 5020016 From andy@soft-n-hard.de Fri Apr 26 13:09:11 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id NAA01303 for ; Fri, 26 Apr 1996 13:09:11 +0200 (MET DST) Received: from aw.bnc.net by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA25426; Fri, 26 Apr 1996 13:08:55 +0200 (MET) Received: from newton.soft-n-hard.de (andy@newton.soft-n-hard.de [194.163.66.33]) by aw.bnc.net (8.6.12/8.6.12) with ESMTP id NAA28857 for ; Fri, 26 Apr 1996 13:11:05 +0200 Received: (from andy@localhost) by newton.soft-n-hard.de (8.6.12/8.6.9) id NAA00887; Fri, 26 Apr 1996 13:06:58 +0200 Date: Fri, 26 Apr 1996 13:06:58 +0200 From: Andreas Busse Message-Id: <199604261106.NAA00887@newton.soft-n-hard.de> To: linux-mips@fnet.fr Subject: Re: Linux boot question In-Reply-To: <199604261025.MAA01476@informatik.uni-koblenz.de> X-Mailer: [XMailTool v3.1.2b] Content-Length: 2055 Lines: 48 Hi all, (yeah, I know, I was *very* silent in the past months -- sorry!) > Well, we still didn't receive the machine promised by SNI. So I continue > porting Linux/MIPS to dead hardware (Acer Pica). Well, I consider a machine > not available for sale in Europe/US dead hardware. Mr. Schweisser @ SNI Munich called me this morning and promised me to send the box off next week. They recently ran out of machines, but now I'm at the top of their priority list, he said. Well, let's see... > This however doesn't mean that the project is dead - it just hasn't really > started yet. A contract about SNI loaning a machine to Soft & Hard, > Andy's company has been signed. SNI just hasn't fullfilled it's part yet. > People at SNI seem to be pretty interested in Linux/MIPS for their machines, > though. Apparently one of them has asked for a port to a RM400 which is > a *SMP* machine. Well, I'd cut off at least a dozen fingers for such a > beast :-) Let's start the race, but be careful with your fingers! You need at least three to hit CTRL-ALT-DEL ;-) BTW, just a dumb and not Mips-related question, certainly to Michael Neuffer who will take care of the SNI baby: I have trouble installing *any* of the newer Linux/i386 releases on systems equipped with DPT SCSI controllers. Is this a known problem, and is there a workaround? I tried four releases on PM2024 controllers of different age. This is the result: - Slackware-2.3/Linux-1.2.8/aout (works) - Slackware-3.0/Linux-1.2.13/elf (fails) - Debian 0.93R6/Linux-1.2.13/aout (works) - Caldera 1.0/Linux-1.2.13/elf (fails) Looks like that all ELF kernels don't work with the DPT controller. I always get "HBA at 0x**** didn't react on INQUIRY. Sorry." Any ideas? Cheers, Andy ----------------------------------------------------------- Andreas Busse | andy@soft-n-hard.de Soft N Hard GbR | Phone: +49 2636-970105 Im Hufen Boden 16, D-53498 Waldorf | Fax: +49 2636-970106 ----------------------------------------------------------- From linux@mailhost.uni-koblenz.de Fri Apr 26 13:42:37 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id NAA01895 for ; Fri, 26 Apr 1996 13:42:36 +0200 (MET DST) Received: from informatik.uni-koblenz.de (mailhost.uni-koblenz.de) by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA25893; Fri, 26 Apr 1996 13:41:03 +0200 (MET) Received: from ozzy (ozzy.uni-koblenz.de [141.26.5.8]) by informatik.uni-koblenz.de (8.7.4/8.6.9) with SMTP id NAA04055 for ; Fri, 26 Apr 1996 13:40:51 +0200 (MET DST) From: Systemkennung Linux Message-Id: <199604261140.NAA04055@informatik.uni-koblenz.de> Received: by ozzy (SMI-8.6/KO-2.0) id NAA20052; Fri, 26 Apr 1996 13:39:08 +0200 Subject: Re: Linux boot question To: linux-mips@fnet.fr Date: Fri, 26 Apr 1996 13:39:07 +0200 (MET DST) In-Reply-To: <199604261106.NAA00887@newton.soft-n-hard.de> from "Andreas Busse" at Apr 26, 96 01:06:58 pm Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Length: 2418 Lines: 54 Hi all, > (yeah, I know, I was *very* silent in the past months -- sorry!) The death return to life :-) > Mr. Schweisser @ SNI Munich called me this morning and promised me > to send the box off next week. They recently ran out of machines, but > now I'm at the top of their priority list, he said. Well, let's > see... > > > This however doesn't mean that the project is dead - it just hasn't really > > started yet. A contract about SNI loaning a machine to Soft & Hard, > > Andy's company has been signed. SNI just hasn't fullfilled it's part yet. > > People at SNI seem to be pretty interested in Linux/MIPS for their machines, > > though. Apparently one of them has asked for a port to a RM400 which is > > a *SMP* machine. Well, I'd cut off at least a dozen fingers for such a > > beast :-) > > Let's start the race, but be careful with your fingers! You need at > least three to hit CTRL-ALT-DEL ;-) I'd like a machine with the reset button on a two meter cable much more. One can even use this without fingers... (Hi Gero (who luckily isn't reading this mailing list ...)). > BTW, just a dumb and not Mips-related question, certainly to Michael Neuffer > who will take care of the SNI baby: I have trouble installing *any* of the > newer Linux/i386 releases on systems equipped with DPT SCSI controllers. > Is this a known problem, and is there a workaround? I tried four releases > on PM2024 controllers of different age. This is the result: > > - Slackware-2.3/Linux-1.2.8/aout (works) > - Slackware-3.0/Linux-1.2.13/elf (fails) > - Debian 0.93R6/Linux-1.2.13/aout (works) > - Caldera 1.0/Linux-1.2.13/elf (fails) > > Looks like that all ELF kernels don't work with the DPT controller. > I always get "HBA at 0x**** didn't react on INQUIRY. Sorry." Any ideas? The bug is completly unrelated to ELF kernels. Linux/1.2.13 has a bug in the kernel the breaks the DPT driver. Mike tried to convince Linus to make a 1.2.14 but since this kernel was never released ... Some distributions featuring the 1.2.13 kernel have modified kernels with a DPT bug fix. That's why Debian is running. Anyway, Mike has promised to give me his ISA DPT and I'll make it work with Linux/MIPS also. I assume the most problems they I'll have are rather generic ones and I'll therefore be able to make many types of cards work. Ralf (Hey Mike ... This is now your catchword to advertise for DPT ;-) From andy@soft-n-hard.de Fri Apr 26 13:58:35 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id NAA02175 for ; Fri, 26 Apr 1996 13:58:35 +0200 (MET DST) Received: from aw.bnc.net by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA26168; Fri, 26 Apr 1996 13:58:20 +0200 (MET) Received: from newton.soft-n-hard.de (andy@newton.soft-n-hard.de [194.163.66.33]) by aw.bnc.net (8.6.12/8.6.12) with ESMTP id OAA30043 for ; Fri, 26 Apr 1996 14:00:30 +0200 Received: (from andy@localhost) by newton.soft-n-hard.de (8.6.12/8.6.9) id NAA01402; Fri, 26 Apr 1996 13:56:24 +0200 Date: Fri, 26 Apr 1996 13:56:24 +0200 From: Andreas Busse Message-Id: <199604261156.NAA01402@newton.soft-n-hard.de> To: linux-mips@fnet.fr Subject: Re: Linux boot question In-Reply-To: <199604261140.NAA04055@informatik.uni-koblenz.de> X-Mailer: [XMailTool v3.1.2b] Content-Length: 1454 Lines: 41 Hi again, > > The death return to life :-) Well, I wasn't actually dead but with zillions tons of workload on my shoulders I really, really wasn't able to respond to any of the postings. But I tried hard to read them all ;-) > > I'd like a machine with the reset button on a two meter cable much more. > One can even use this without fingers... (Hi Gero (who luckily isn't reading > this mailing list ...)). Yeah, with your nose. Or some other "device" ;-) > The bug is completly unrelated to ELF kernels. Linux/1.2.13 has a bug > in the kernel the breaks the DPT driver. Mike tried to convince Linus > to make a 1.2.14 but since this kernel was never released ... Some > distributions featuring the 1.2.13 kernel have modified kernels with > a DPT bug fix. That's why Debian is running. Ah, I see. Hmmm. I need to figure out if it's possible to make a Caldera bootdisk with a Debian kernel. Or shall I completely use the Debian distribution? > (Hey Mike ... This is now your catchword to advertise for DPT ;-) I have four of them and I prefer them over any other SCSI controllers for PC's. Enough advertising? Cheers, Andy ----------------------------------------------------------- Andreas Busse | andy@soft-n-hard.de Soft N Hard GbR | Phone: +49 2636-970105 Im Hufen Boden 16, D-53498 Waldorf | Fax: +49 2636-970106 ----------------------------------------------------------- From linux@mailhost.uni-koblenz.de Fri Apr 26 14:51:24 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id OAA03298 for ; Fri, 26 Apr 1996 14:51:24 +0200 (MET DST) Received: from informatik.uni-koblenz.de (mailhost.uni-koblenz.de) by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA27078; Fri, 26 Apr 1996 14:50:58 +0200 (MET) Received: from ozzy (ozzy.uni-koblenz.de [141.26.5.8]) by informatik.uni-koblenz.de (8.7.4/8.6.9) with SMTP id OAA06804 for ; Fri, 26 Apr 1996 14:50:57 +0200 (MET DST) From: Systemkennung Linux Message-Id: <199604261250.OAA06804@informatik.uni-koblenz.de> Received: by ozzy (SMI-8.6/KO-2.0) id OAA21984; Fri, 26 Apr 1996 14:49:13 +0200 Subject: More kernels ... To: linux-mips@fnet.fr Date: Fri, 26 Apr 1996 14:49:12 +0200 (MET DST) Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Length: 1831 Lines: 43 -----BEGIN PGP SIGNED MESSAGE----- Hi all, here again some kernel patches. Noteworthy MIPS changes: - Improved handling of caches. In one of the usual attempts to support hardware I don't have I fixed sys_cacheflush(2) to support the R10000. - Added support for machines with a PC style RAM layout like the rPC44. For these machines the kernel will be loaded to 0x80100000 and the low 640kb memory freed later on. This still need some special support in Milo which should be easy to add. - Improved handling for overflow and floating point exceptions. It is now possible to compile GCC native without giving some special magic options to the compiler. I'll also upload a little endian binary package of fileutils 3.12. The old binaries contained in root-0.01.tar.gz were compiled with a ancient version of libc. This bug results in mknod and other programs using mknod(2) creating the device file correctly but locking up after this. As I'm writing this I'm still uploading (at 0.067 Kbytes/sec ...), so it might still take a while for my uploads to appear on ftp.fnet.fr. Sorry, but the EBone is deadslow currently ... Ralf b9b3a3d6324398d44d945633ceff9c4f fileutils-3.12.tar.gz bd8d73c8d388a756aa2562e78dfa1bc1 linux-1.3.85-1.3.86.diffs.gz c524e2c581b54b5bf9e9762e05cd2afb linux-1.3.86-1.3.87.diffs.gz 5f086956431b613f03b26e4bfbf66297 linux-1.3.87-1.3.88.diffs.gz 159dbba52a7cf60c3a684b7decf5a328 linux-1.3.88-1.3.89.diffs.gz b18bfaf203cf0ca22496f156a3ac09f0 linux-1.3.89-1.3.90.diffs.gz -----BEGIN PGP SIGNATURE----- Version: 2.6.2 iQCVAgUBMYDGRUckbl6vezDBAQGf4gP+MnOvZUXBo61uLIKkn+pPYWCoOI3Jwi95 KGgYd+vfNl4LUqfw+XMAPlQhDzqBdFl8lrvqw7In9pVlJY+KlwSviGMLUrCpNMCX GJqkxRz7UQB1rrokxVS229utn0cspG3AYn7fw/EdUPsNPdWl4aMLMgIrwhBylk2j kSsP3vSAJns= =Wtfw -----END PGP SIGNATURE----- From imp@rover.village.org Fri Apr 26 16:05:42 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id QAA05025 for ; Fri, 26 Apr 1996 16:05:41 +0200 (MET DST) Received: from rover.village.org by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA17535; Fri, 26 Apr 1996 16:05:23 +0200 (MET) Received: from localhost (localhost [127.0.0.1]) by rover.village.org (8.7.5/8.6.6) with SMTP id IAA24542 for ; Fri, 26 Apr 1996 08:05:30 -0600 (MDT) Message-Id: <199604261405.IAA24542@rover.village.org> To: linux-mips@fnet.fr Subject: Re: More kernels ... In-Reply-To: Your message of Fri, 26 Apr 1996 14:49:12 +0200 Date: Fri, 26 Apr 1996 08:05:30 -0600 From: Warner Losh Content-Length: 649 Lines: 13 : b9b3a3d6324398d44d945633ceff9c4f fileutils-3.12.tar.gz : bd8d73c8d388a756aa2562e78dfa1bc1 linux-1.3.85-1.3.86.diffs.gz : c524e2c581b54b5bf9e9762e05cd2afb linux-1.3.86-1.3.87.diffs.gz : 5f086956431b613f03b26e4bfbf66297 linux-1.3.87-1.3.88.diffs.gz : 159dbba52a7cf60c3a684b7decf5a328 linux-1.3.88-1.3.89.diffs.gz : b18bfaf203cf0ca22496f156a3ac09f0 linux-1.3.89-1.3.90.diffs.gz Got a 1.3.84-1.3.85 patch md5 signature? Also, if someone could take a look at the src/kernel/v1.3 directory and make sure that each step is in place. I seem to recall having to get one of the 70's from the private/Uploads (or was that private/Incoming). Warner From beurton@nil.fnet.fr Fri Apr 26 16:26:07 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id QAA05407 for ; Fri, 26 Apr 1996 16:26:07 +0200 (MET DST) Received: from nil.fnet.fr by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA23143; Fri, 26 Apr 1996 16:25:53 +0200 (MET) Received: from nil.fnet.fr (localhost.fnet.fr [127.0.0.1]) by nil.fnet.fr (8.7.5/8.7.3) with ESMTP id RAA02084 for ; Fri, 26 Apr 1996 17:25:30 +0200 (METDST) Message-Id: <199604261525.RAA02084@nil.fnet.fr> To: linux-mips@fnet.fr Subject: Re: More kernels ... In-Reply-To: Your message of "Fri, 26 Apr 1996 08:05:30 METDST." <199604261405.IAA24542@rover.village.org> X-Mailer: MH 6.8.3 Date: Fri, 26 Apr 1996 17:25:30 +0200 From: Luc Beurton Content-Length: 1014 Lines: 29 ### ### : b9b3a3d6324398d44d945633ceff9c4f fileutils-3.12.tar.gz ### : bd8d73c8d388a756aa2562e78dfa1bc1 linux-1.3.85-1.3.86.diffs.gz ### : c524e2c581b54b5bf9e9762e05cd2afb linux-1.3.86-1.3.87.diffs.gz ### : 5f086956431b613f03b26e4bfbf66297 linux-1.3.87-1.3.88.diffs.gz ### : 159dbba52a7cf60c3a684b7decf5a328 linux-1.3.88-1.3.89.diffs.gz ### : b18bfaf203cf0ca22496f156a3ac09f0 linux-1.3.89-1.3.90.diffs.gz ### ### Got a 1.3.84-1.3.85 patch md5 signature? Also, if someone could take ### a look at the src/kernel/v1.3 directory and make sure that each step ### is in place. I seem to recall having to get one of the 70's from the ### private/Uploads (or was that private/Incoming). ### ### Warner Well, linux-1.3.86-1.3.87.diffs.gz linux-1.3.87-1.3.88.diffs.gz linux-1.3.88-1.3.89.diffs.gz linux-1.3.89-1.3.90.diffs.gz Are arrived in src/kernel/v1.3 fileutils-3.12.tar.gz and linux-1.3.85-1.3.86.diffs.gz are not yet arrived ... There's a file .sig with all the md5 signature in src.kernel/v1.3 Luc. From imp@rover.village.org Fri Apr 26 18:16:32 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id SAA02013 for ; Fri, 26 Apr 1996 18:16:29 +0200 (MET DST) Received: from rover.village.org by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA16915; Fri, 26 Apr 1996 18:16:09 +0200 (MET) Received: from localhost (localhost [127.0.0.1]) by rover.village.org (8.7.5/8.6.6) with SMTP id KAA25341; Fri, 26 Apr 1996 10:16:14 -0600 (MDT) Message-Id: <199604261616.KAA25341@rover.village.org> To: linux-mips@fnet.fr Cc: Ralf Baechle Subject: Re: More kernels ... In-Reply-To: Your message of Fri, 26 Apr 1996 17:25:30 +0200 Date: Fri, 26 Apr 1996 10:16:14 -0600 From: Warner Losh Content-Length: 542 Lines: 15 : fileutils-3.12.tar.gz and linux-1.3.85-1.3.86.diffs.gz : are not yet arrived ... Looks like fileutils-3.12.tar.gz is in private/Incoming right now. Neither its size nor its date has changed in the last 10 minutes. However, I'm more keen on getting the 1.3.85->86 patch because I want to try out Ralf's latest changes for the rPC44 stuff.... Hey Ralf, can you upload that again? Also, the file src/kernel/v1.3/LATEST-IS-1.3.85 looks a little out of date, or does that get generated automatically based on what patches are there? Warner From ralf@Julia.DE Fri Apr 26 18:36:04 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id SAA02428 for ; Fri, 26 Apr 1996 18:36:01 +0200 (MET DST) Received: from alles.intern.julia.de ([194.221.49.2]) by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA18585; Fri, 26 Apr 1996 18:34:31 +0200 (MET) Received: from pris.intern.julia.de (pris.intern.julia.de [194.221.49.37]) by alles.intern.julia.de (8.7.5/8.7.3) with ESMTP id RAA01024 for ; Fri, 26 Apr 1996 17:31:31 +0200 From: Ralf Baechle Received: (from ralf@localhost) by pris.intern.julia.de (8.7.5/8.7.3) id SAA06315 for linux-mips@fnet.fr; Fri, 26 Apr 1996 18:33:47 +0200 (MET DST) Message-Id: <199604261633.SAA06315@pris.intern.julia.de> Subject: Oh no, not again ... To: linux-mips@fnet.fr Date: Fri, 26 Apr 1996 18:33:46 +0200 (MET DST) Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Length: 452 Lines: 20 -----BEGIN PGP SIGNED MESSAGE----- a kernel patch ... Have fun, Ralf 0b0c23bd443b8f686658b8e7f5769f79 linux-1.3.90-linux-1.3.91.diffs.gz -----BEGIN PGP SIGNATURE----- Version: 2.6.3i Charset: latin1 iQCVAwUBMYD66Uckbl6vezDBAQGf8wQAo9XN3n94qN1MocAt03dmqhTkbRkUAQp/ 4pUoXB8d+OO2CkNXor27Ck/VEPGhlzuo4dIweWFnQ8sKNwvzIVgE4+zlB65uQ5va iVvHKnX1PeWL/XWdiLd2Dyu7z2/IqzIjSg/KlFwpNcRzCFS0F1fflMCdf9C/gHOs uzv+4vRjsuU= =mxB3 -----END PGP SIGNATURE----- From Stoned.Elipot@univ-evry.fr Fri Apr 26 18:38:46 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id SAA02491 for ; Fri, 26 Apr 1996 18:38:46 +0200 (MET DST) Received: from amiante.univ-evry.fr by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA18680; Fri, 26 Apr 1996 18:38:30 +0200 (MET) Received: from bandsept.univ-evry.fr (bandsept.univ-evry.fr [192.93.214.19]) by amiante.univ-evry.fr with ESMTP (8.6.12/96.04.09/louis-blanc); id SAA19208; Fri, 26 Apr 1996 18:38:38 +0200 Received: from localhost.univ-evry.fr (localhost.univ-evry.fr [127.0.0.1]) by bandsept.univ-evry.fr with SMTP (8.6.12/94.01.26); id SAA14111; Fri, 26 Apr 1996 18:37:56 +0200 Message-Id: <199604261637.SAA14111@bandsept.univ-evry.fr> To: linux-mips@fnet.fr Subject: Re: More kernels ... In-Reply-To: Your message of "Fri, 26 Apr 1996 10:16:14 MDT." <199604261616.KAA25341@rover.village.org> X-It: bandsept (wrong OS, but right CPU) X-Mailer: MH 6.8.3, MH-E 5.0.2 Date: Fri, 26 Apr 1996 18:37:48 +0200 From: Stoned Elipot Content-Length: 342 Lines: 10 Hi all, >>>>> "Warner" == Warner Losh writes: Warner> Also, the file src/kernel/v1.3/LATEST-IS-1.3.85 looks a little Warner> out of date, or does that get generated automatically based on Warner> what patches are there? When the guy who put the patches in the right place don't forget it, it's just a 'mv ...' :) Stoned. From imp@rover.village.org Fri Apr 26 18:45:51 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id SAA02694 for ; Fri, 26 Apr 1996 18:45:51 +0200 (MET DST) Received: from rover.village.org by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA18823; Fri, 26 Apr 1996 18:45:34 +0200 (MET) Received: from localhost (localhost [127.0.0.1]) by rover.village.org (8.7.5/8.6.6) with SMTP id KAA25618 for ; Fri, 26 Apr 1996 10:45:41 -0600 (MDT) Message-Id: <199604261645.KAA25618@rover.village.org> To: linux-mips@fnet.fr Subject: fileutiles-3.12.tar.gz corrupt??? Date: Fri, 26 Apr 1996 10:45:40 -0600 From: Warner Losh Content-Length: 683 Lines: 17 The fileutiles-3.12.tar.gz in private/Incoming seems to be corrupt. I transferred it in binary mode and I get: gnutar tvfz fileutils-3.12.tar.gz drwxr-xr-x root/root 0 Apr 24 18:12 1996 bin/ -rwxr-xr-x root/root 204325 Apr 24 18:05 1996 bin/chgrp gzip: stdin: invalid compressed data--format violated gnutar: Unexpected EOF on archive file when I try to untar it. I don't have a md5 checksum routine to check this against the mail on this machine (well, on the machine at work). Might want to hold off moving it until its integrety can be verified. Ralf, just what you need to do with your "fast" net connect: pump more copies of large things through it :-(. Warner From lm@neteng.engr.sgi.com Fri Apr 26 18:59:35 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id SAA02996 for ; Fri, 26 Apr 1996 18:59:34 +0200 (MET DST) Received: from sgi.sgi.com (SGI.COM) by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA19030; Fri, 26 Apr 1996 18:59:14 +0200 (MET) Received: from cthulhu.engr.sgi.com by sgi.sgi.com via ESMTP (950405.SGI.8.6.12/910110.SGI) for <@sgi.engr.sgi.com:linux-mips@fnet.fr> id JAA20279; Fri, 26 Apr 1996 09:59:07 -0700 Received: from neteng.engr.sgi.com by cthulhu.engr.sgi.com via ESMTP (950511.SGI.8.6.12.PATCH526/911001.SGI) for <@cthulhu.engr.sgi.com:linux-mips@fnet.fr> id JAA17911; Fri, 26 Apr 1996 09:58:59 -0700 Received: (from lm@localhost) by neteng.engr.sgi.com (950413.SGI.8.6.12/960327.SGI.AUTOCF) id JAA25779 for linux-mips@fnet.fr; Fri, 26 Apr 1996 09:58:59 -0700 Date: Fri, 26 Apr 1996 09:58:59 -0700 From: lm@neteng.engr.sgi.com (Larry McVoy) Message-Id: <199604261658.JAA25779@neteng.engr.sgi.com> To: linux-mips@fnet.fr Subject: Re: Linux/SGI coming Content-Length: 772 Lines: 16 So, can someone provide a quick summary of the current status? All the usual stuff, like what compilers are used, what linker, what are you doing for userland, etc. Also, the handhold on information has loosened quite a bit. If there is specific information you are looking for, let me know. In particular, I have access to all the chip errata, something that our internal people use a lot. --lm --- Larry McVoy lm@sgi.com http://reality.sgi.com/lm (415) 933-1804 Copyright 1996, all rights reserved. Microsoft Network is prohibited from redistributing this work in any form, in whole or in part without license. License to distribute this work is available to Microsoft at $500. Transmission without permission constitutes an agreement to these terms. From lm@neteng.engr.sgi.com Sat Apr 27 17:31:55 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id RAA01607 for ; Sat, 27 Apr 1996 17:31:54 +0200 (MET DST) Received: from sgi.sgi.com (SGI.COM) by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA19904; Sat, 27 Apr 1996 17:31:37 +0200 (MET) Received: from cthulhu.engr.sgi.com by sgi.sgi.com via ESMTP (950405.SGI.8.6.12/910110.SGI) for <@sgi.engr.sgi.com:linux-mips@fnet.fr> id IAA26251; Sat, 27 Apr 1996 08:31:45 -0700 Received: from neteng.engr.sgi.com by cthulhu.engr.sgi.com via ESMTP (950511.SGI.8.6.12.PATCH526/911001.SGI) for <@cthulhu.engr.sgi.com:linux-mips@fnet.fr> id KAA18406; Fri, 26 Apr 1996 10:02:38 -0700 Received: (from lm@localhost) by neteng.engr.sgi.com (950413.SGI.8.6.12/960327.SGI.AUTOCF) id KAA26060 for linux-mips@fnet.fr; Fri, 26 Apr 1996 10:02:38 -0700 Date: Fri, 26 Apr 1996 10:02:38 -0700 From: lm@neteng.engr.sgi.com (Larry McVoy) Message-Id: <199604261702.KAA26060@neteng.engr.sgi.com> To: linux-mips@fnet.fr Subject: Re: Linux/SGI coming Content-Length: 846 Lines: 15 I think I can probably find some resources to put a linux-cvs.sgi.com machine on the net. We have some spare ports on our internet ethernet. If I put a machine there with disk space, and provide logins, can you guys do the rest? Note that I might be putting a linux pc there and hacking it to limit the rate at which the data can move. Has anyone done this? That's SGI's primary concern, they would like to "reserve" some bandwidth for other stuff on that T1 line. --lm --- Larry McVoy lm@sgi.com http://reality.sgi.com/lm (415) 933-1804 Copyright 1996, all rights reserved. Microsoft Network is prohibited from redistributing this work in any form, in whole or in part without license. License to distribute this work is available to Microsoft at $500. Transmission without permission constitutes an agreement to these terms. From ralf@Julia.DE Fri Apr 26 19:17:30 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id TAA03574 for ; Fri, 26 Apr 1996 19:17:29 +0200 (MET DST) Received: from alles.intern.julia.de ([194.221.49.2]) by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA19370; Fri, 26 Apr 1996 19:17:04 +0200 (MET) Received: from pris.intern.julia.de (pris.intern.julia.de [194.221.49.37]) by alles.intern.julia.de (8.7.5/8.7.3) with ESMTP id SAA01135 for ; Fri, 26 Apr 1996 18:14:14 +0200 From: Ralf Baechle Received: (from ralf@localhost) by pris.intern.julia.de (8.7.5/8.7.3) id TAA06349 for linux-mips@fnet.fr; Fri, 26 Apr 1996 19:16:29 +0200 (MET DST) Message-Id: <199604261716.TAA06349@pris.intern.julia.de> Subject: Re: More kernels ... To: linux-mips@fnet.fr Date: Fri, 26 Apr 1996 19:16:29 +0200 (MET DST) In-Reply-To: <199604261750.TAA03730@nil.fnet.fr> from "Luc Beurton" at Apr 26, 96 07:50:20 pm Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Length: 1621 Lines: 41 Hi, > ### : fileutils-3.12.tar.gz and linux-1.3.85-1.3.86.diffs.gz > ### : are not yet arrived ... > ### > ### Looks like fileutils-3.12.tar.gz is in private/Incoming right now. > ### Neither its size nor its date has changed in the last 10 minutes. > > Yep! every half hour ftp.fnet.fr ask to lena.fnet.fr what's new in > Ralf's home > > ### However, I'm more keen on getting the 1.3.85->86 patch because I want > ### to try out Ralf's latest changes for the rPC44 stuff.... Hey Ralf, > ### can you upload that again? > > The upload of fileutils-3.12.tar.gz is still running but I'm afraid that > Ralf forget the linux-1.3.85-1.3.86.diffs.gz No, but the connection just died while I was away from the console :-( There is however the usefull invention of the restart command and by now the upload is complete :-> > ### Also, the file src/kernel/v1.3/LATEST-IS-1.3.85 looks a little out of > ### date, or does that get generated automatically based on what patches > ### are there? > > No, I move LATEST-IS-1.3.XX to LATEST-IS-1.3.YY when all the patchs are in the > src/kernel/v1.3, You're right a little script could be done :-)) > > ### When the guy who put the patches in the right place don't forget it, > ### it's just a 'mv ...' :) > > Ralf have just put linux-1.3.90-linux-1.3.91.diffs.gz > I don't know where he wants to stop That's an easy question: just do ftp ftp.cs.helsinki.fi, cd /pub/Software/Linux/Kernel/v1.3, dir LAT* and you have the answer. At least as I write this, the answer isn't 42 but 1.3.95. Oh, my machine at home should have been ready with '92 half an hour ago ... Ralf From beurton@nil.fnet.fr Fri Apr 26 18:50:59 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id SAA02837 for ; Fri, 26 Apr 1996 18:50:58 +0200 (MET DST) Received: from nil.fnet.fr by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA18956; Fri, 26 Apr 1996 18:50:44 +0200 (MET) Received: from nil.fnet.fr (localhost.fnet.fr [127.0.0.1]) by nil.fnet.fr (8.7.5/8.7.3) with ESMTP id TAA03730 for ; Fri, 26 Apr 1996 19:50:21 +0200 (METDST) Message-Id: <199604261750.TAA03730@nil.fnet.fr> To: linux-mips@fnet.fr Subject: Re: More kernels ... In-Reply-To: Your message of "Fri, 26 Apr 1996 10:16:14 METDST." <199604261616.KAA25341@rover.village.org> X-Mailer: MH 6.8.3 Date: Fri, 26 Apr 1996 19:50:20 +0200 From: Luc Beurton Content-Length: 1537 Lines: 38 ### : fileutils-3.12.tar.gz and linux-1.3.85-1.3.86.diffs.gz ### : are not yet arrived ... ### ### Looks like fileutils-3.12.tar.gz is in private/Incoming right now. ### Neither its size nor its date has changed in the last 10 minutes. Yep! every half hour ftp.fnet.fr ask to lena.fnet.fr what's new in Ralf's home ### However, I'm more keen on getting the 1.3.85->86 patch because I want ### to try out Ralf's latest changes for the rPC44 stuff.... Hey Ralf, ### can you upload that again? The upload of fileutils-3.12.tar.gz is still running but I'm afraid that Ralf forget the linux-1.3.85-1.3.86.diffs.gz ### Also, the file src/kernel/v1.3/LATEST-IS-1.3.85 looks a little out of ### date, or does that get generated automatically based on what patches ### are there? No, I move LATEST-IS-1.3.XX to LATEST-IS-1.3.YY when all the patchs are in the src/kernel/v1.3, You're right a little script could be done :-)) ### When the guy who put the patches in the right place don't forget it, ### it's just a 'mv ...' :) Ralf have just put linux-1.3.90-linux-1.3.91.diffs.gz I don't know where he wants to stop ### when I try to untar it. I don't have a md5 checksum routine to check ### this against the mail on this machine (well, on the machine at work). ### Might want to hold off moving it until its integrety can be verified. ### Ralf, just what you need to do with your "fast" net connect: pump more ### copies of large things through it :-(. Yep! you're right, the transfert seems finish and the file is corrupted :( From ralf@Julia.DE Fri Apr 26 20:15:47 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id UAA04813 for ; Fri, 26 Apr 1996 20:15:46 +0200 (MET DST) Received: from alles.intern.julia.de ([194.221.49.2]) by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA20164; Fri, 26 Apr 1996 20:15:03 +0200 (MET) Received: from pris.intern.julia.de (pris.intern.julia.de [194.221.49.37]) by alles.intern.julia.de (8.7.5/8.7.3) with ESMTP id TAA01219; Fri, 26 Apr 1996 19:12:08 +0200 From: Ralf Baechle Received: (from ralf@localhost) by pris.intern.julia.de (8.7.5/8.7.3) id UAA06380; Fri, 26 Apr 1996 20:14:24 +0200 (MET DST) Message-Id: <199604261814.UAA06380@pris.intern.julia.de> Subject: Re: Linux/SGI coming To: lm@neteng.engr.sgi.com, davem@caip.rutgers.edu, linux-mips@fnet.fr Date: Fri, 26 Apr 1996 20:14:24 +0200 (MET DST) In-Reply-To: <199604261658.JAA25779@neteng.engr.sgi.com> from "Larry McVoy" at Apr 26, 96 09:58:59 am Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Length: 4153 Lines: 81 Hi all, > So, can someone provide a quick summary of the current status? All the > usual stuff, like what compilers are used, what linker, what are you > doing for userland, etc. OK. There is a bootloader available for ARC conformant systems. This implies little endian byte order. Some time ago we redesigned the interface to pass bootinformation to the kernel. The new interface uses a tagged list and can also be used to pass big information like ramdisks or fonts from the bootloader to the kernel. Due to the byteorder Milo isn't really of interest. Furthermore the Milo distribution is crippled due to (C) problems. Since you, Larry, probably have signed an NDA with SGI ;-) you can however get the complete sources, it you wish. Otherwise the crippled distribution is available via ftp.fnet.fr. What is missing are the lowlevel interfaces between the pseudo libc implementation that Milo uses and the ARC BIOS. This was required due to M$ holding the (C) for the information Milo is based on. The crosscompiler environment is based on GCC 2.7.2 and Binutils 2.6. The required patches that fix lots of bugs and add Linux/MIPS support for a.out, ELF little/big endian byte order are available on ftp.fnet.fr. Their installation isn't easy but Dave shouldn't have to much trouble with this. There are some binary packages on ftp.fnet.fr for Linux/i386 host. I'd like to have packages for more hosts, so if someone sends me binaries we'll put the online. (I could create packages for Stun's OSes if someone wants them) The libc implementation is based on the GNU libc which is available only as ELF. My home sources are based on snapshot 960210; upgrading to the newest libc will be the next thing for me to do when I sent my kernel patches to Linus. There is an older Linux libc port available - don't use it! I changed the low level kernel interfaces for 1.3 so that they should now look much more like IRIX 5.3 / RISC/os 5.01 which breaks this old (a.out ...) port completly. The distribution is still available via ftp to help building the a.out crosscompiler. The quality of the libc port to Linux is good enough to build most user stuff. GNU libc is however very aggresive to bad code; quite some packages need to be patched to compile. I'll put patches online as soon as I can. Most of the userland stuff available via ftp is based on the GNU or Debian source distributions. The kernel itself runs pretty solid for me. It currently supports the PC versions of R4000 and R4400 CPUs but probably doesn't work for SC and MC versions. I tried to support R2000/R3000/R6000/R8000/R10000 as good as possible without having hardware, so a port shouldn't be too hard. For R4000 and better many of the lowlevel functions have been written to be SMP proof but again without SMP hardware and at this point of the project SMP is nothing really invest my time in. One of the kernel features that are still missing is the special kernel code to handle denormalized numbers - not of interest for you - FPU emulation, branch delay emulation. Also there is no support for ptrace(2), therefore no debugger. The supported machines are the Mips Magnum 4000, Olivetti M700-10 (a OEM Magnum version) and Acer Pica 61, which is a kind of a successor type for the Magnum built by Acer. All these machines have a similar DMA engine, Floppy, NCR53C94 SCSI and Sonic Ethernet onboard. This SCSI/Ethernet hardware is currently not supported. I thought it would be better to use (E)ISA hardware where drivers are very easy to port and to put my time on the other parts of the system. Paul Antoine is working on DECstation/R3000 support. > Also, the handhold on information has loosened quite a bit. If there is > specific information you are looking for, let me know. In particular, > I have access to all the chip errata, something that our internal people > use a lot. I think it would be great to have all this stuff on our ftp server as it could especially help peoples with ancient CPUs. Andy's R4000 2.2 was driving me crazy ... This should answer many of your questions but probably not all, so feel free to mail me. Ralf From imp@rover.village.org Fri Apr 26 21:53:41 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id VAA06715 for ; Fri, 26 Apr 1996 21:53:40 +0200 (MET DST) Received: from rover.village.org by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA21438; Fri, 26 Apr 1996 21:53:12 +0200 (MET) Received: from localhost (localhost [127.0.0.1]) by rover.village.org (8.7.5/8.6.6) with SMTP id NAA26233; Fri, 26 Apr 1996 13:53:19 -0600 (MDT) Message-Id: <199604261953.NAA26233@rover.village.org> To: linux-mips@fnet.fr Subject: Re: Linux/SGI coming Cc: lm@neteng.engr.sgi.com, davem@caip.rutgers.edu In-Reply-To: Your message of Fri, 26 Apr 1996 20:14:24 +0200 Date: Fri, 26 Apr 1996 13:53:19 -0600 From: Warner Losh Content-Length: 3362 Lines: 67 : MILO... It should be relatively easy for SGI to build a MILO clone for their bootroms. It basically loads the kernel into memory, then jumps to it. The ARC BIOS stsuff is how it loads it in and sets some things up. You'd have to follow the tagging stuff that MILO scribbles into high memory somehow so that the kernel can read it out. Whether this is done in a separate boot loader or in your code is a grey area right now. The ARC BIOS is a crude DOS like environment that has been stripped down to allow files to be read off disk for the pruposes of booting and configuration. Some versions allow for TFTP file booting as well. SGI will likely also have to provide some way to read their file system if it isn't ext2fs. This can range from full kernel support to an mtools like program for writing something to a efs partition (I recall that's what sgi uses, and it isn't ufs, feel free to correct me if I'm wrong). Even if that was just used for the boot loader. Also, some disk layout issues are likely there because SGI doesn't use the standard DOS partition scheme. These may be answered by the Alpha/Sparc ports, so it might not be a big issue. : The libc implementation is based on the GNU libc which is available only : as ELF. My home sources are based on snapshot 960210; upgrading to the : newest libc will How does one joint this development effort? :-) : R4000 and better many of the lowlevel functions have been written to be : SMP proof but again without SMP hardware and at this point of the project : SMP is nothing really invest my time in. SMP would be a fun project. All you need is an SMP MIPS box :-(. : The supported machines are the Mips Magnum 4000, Olivetti M700-10 (a OEM : Magnum version) and Acer Pica 61, which is a kind of a successor type for : the Magnum built by Acer. All these machines have a similar DMA engine, : Floppy, NCR53C94 SCSI and Sonic Ethernet onboard. This SCSI/Ethernet : hardware is currently not supported. I thought it would be better to : use (E)ISA hardware where drivers are very easy to port and to put my : time on the other parts of the system. Sniff Sniff Sniff. Nothing about teh Deskstation rPC44? :-) It is up and limping at this time, but I need to upgrade to the latest kernel to see if some of the nastier problems have been fixed, or if I need to fix them still. I've not tried the SCSI drivers yet, but the floppies in 1.3.62 gave me fits. : I think it would be great to have all this stuff on our ftp server as it : could especially help peoples with ancient CPUs. Andy's R4000 2.2 was : driving me crazy ... I know that MIPS has published a bunch of errata on the web, including ones for the R4000. Don't know if the 2.2 is one in the list or not, but will check my printed copies when I get home. None the less, it would be good to have more complete versions of these items if they were available. However, I just tried the address I have for this data, and I can't find it now. The cross compiler stuff is fairly easy to generate. For the SGI platform, it should go well. There is a howto generate cross compilers web page hanging off the Linux-MIPS howto page that should give you what you need to know to build the goofy things, should you require assistance. There are a numnber of minor gotchas, so you might want to glance there first. Warner From ralf@Julia.DE Fri Apr 26 22:18:10 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id WAA07337 for ; Fri, 26 Apr 1996 22:18:09 +0200 (MET DST) Received: from alles.intern.julia.de ([194.221.49.2]) by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA21721; Fri, 26 Apr 1996 22:17:45 +0200 (MET) Received: from pris.intern.julia.de (pris.intern.julia.de [194.221.49.37]) by alles.intern.julia.de (8.7.5/8.7.3) with ESMTP id VAA01416 for ; Fri, 26 Apr 1996 21:14:42 +0200 From: Ralf Baechle Received: (from ralf@localhost) by pris.intern.julia.de (8.7.5/8.7.3) id WAA06479 for linux-mips@fnet.fr; Fri, 26 Apr 1996 22:16:56 +0200 (MET DST) Message-Id: <199604262016.WAA06479@pris.intern.julia.de> Subject: Re: Linux/SGI coming To: linux-mips@fnet.fr Date: Fri, 26 Apr 1996 22:16:56 +0200 (MET DST) In-Reply-To: <199604261953.NAA26233@rover.village.org> from "Warner Losh" at Apr 26, 96 01:53:19 pm Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Length: 3536 Lines: 74 Hi all, > It should be relatively easy for SGI to build a MILO clone for their > bootroms. It basically loads the kernel into memory, then jumps to > it. The ARC BIOS stsuff is how it loads it in and sets some things > up. You'd have to follow the tagging stuff that MILO scribbles into > high memory somehow so that the kernel can read it out. Whether this > is done in a separate boot loader or in your code is a grey area right > now. The ARC BIOS is a crude DOS like environment that has been > stripped down to allow files to be read off disk for the pruposes of > booting and configuration. Some versions allow for TFTP file booting > as well. One of the design ideas in Milo was that there is a lowlevel library containing the interface to the os. Milo itself uses the library interface. The library interface again exports functions that look somewhat like unix systemcalls. Stoned's version 0.27 of Milo will make this difference much clearer. It might be a good idea to have one common bootloader package for ARC, SGI and as many others as possible. > SGI will likely also have to provide some way to read their file > system if it isn't ext2fs. This can range from full kernel support to > an mtools like program for writing something to a efs partition (I > recall that's what sgi uses, and it isn't ufs, feel free to correct me > if I'm wrong). Even if that was just used for the boot loader. Also, > some disk layout issues are likely there because SGI doesn't use the > standard DOS partition scheme. These may be answered by the > Alpha/Sparc ports, so it might not be a big issue. Btw - Linux 1.3.95 has a read-only implementation of UFS mainly for use by Linux/Sparc. There are many variations of UFS in use. I dunno if the Linux version is what SGI uses but anyway SGI's main filesystem are efs and xfs in the future. > : The libc implementation is based on the GNU libc which is available only > : as ELF. My home sources are based on snapshot 960210; upgrading to the > : newest libc will > > How does one joint this development effort? :-) Aeh... I'll release this stuff rsn. I really need to rework this as my libc port does no longer compile with newer kernel headers. There is a mailing list for the port of GNU libc to Linux. If someone wants to join just tell me. > : R4000 and better many of the lowlevel functions have been written to be > : SMP proof but again without SMP hardware and at this point of the project > : SMP is nothing really invest my time in. > > SMP would be a fun project. All you need is an SMP MIPS box :-(. Yes, and this is what I don't have :-( > I know that MIPS has published a bunch of errata on the web, including > ones for the R4000. Don't know if the 2.2 is one in the list or not, > but will check my printed copies when I get home. None the less, it > would be good to have more complete versions of these items if they > were available. However, I just tried the address I have for this > data, and I can't find it now. I think the information isn't really complete. > The cross compiler stuff is fairly easy to generate. For the SGI > platform, it should go well. There is a howto generate cross > compilers web page hanging off the Linux-MIPS howto page that should > give you what you need to know to build the goofy things, should you > require assistance. There are a numnber of minor gotchas, so you > might want to glance there first. Gimme an account on an SGI and you'll have the crossdevelopment a bit later :-) Ralf From lm@neteng.engr.sgi.com Fri Apr 26 23:58:23 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id XAA09208 for ; Fri, 26 Apr 1996 23:58:22 +0200 (MET DST) Received: from sgi.sgi.com (SGI.COM) by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA22953; Fri, 26 Apr 1996 23:58:04 +0200 (MET) Received: from cthulhu.engr.sgi.com by sgi.sgi.com via ESMTP (950405.SGI.8.6.12/910110.SGI) id OAA28503; Fri, 26 Apr 1996 14:58:02 -0700 Received: from neteng.engr.sgi.com by cthulhu.engr.sgi.com via ESMTP (950511.SGI.8.6.12.PATCH526/911001.SGI) id OAA02127; Fri, 26 Apr 1996 14:58:01 -0700 Received: from localhost (lm@localhost) by neteng.engr.sgi.com (950413.SGI.8.6.12/960327.SGI.AUTOCF) via SMTP id OAA02458; Fri, 26 Apr 1996 14:57:58 -0700 Message-Id: <199604262157.OAA02458@neteng.engr.sgi.com> To: Warner Losh From: lm@gate1-neteng.engr.sgi.com (Larry McVoy) Cc: linux-mips@fnet.fr, lm@neteng.engr.sgi.com, davem@caip.rutgers.edu Subject: Re: Linux/SGI coming Date: Fri, 26 Apr 1996 14:57:58 -0700 Sender: lm@neteng.engr.sgi.com Content-Length: 1623 Lines: 34 : It should be relatively easy for SGI to build a MILO clone for their : bootroms. We've decided to simply give away our stand alone shell. It has all the basics in the shell, knows how to read our two file systems (we may remove XFS support if it turns out to be an issue, XFS is pretty valuable), we'll add ext2fs support. We're also looking at giving away our debugging environment. It's pretty cool, a low level catch-the-keyboard and do a few things, and a higher level interface that knows how to call kernel functions. So most of the debugging is compiled into the kernel, only the basics are in the debugger. I'm not sure if we really want this or if we want to make Linux support the first tty as a console and do remote gdb. I'd actually prefer the gdb style, but the SGI style allows you to debug with only one machine. : SMP would be a fun project. All you need is an SMP MIPS box :-(. Are you interested in R3K boxes? I may be able to find those as time progresses. **No promises**, demand exceeds supply. : : The supported machines are the Mips Magnum 4000, Olivetti M700-10 (a OEM : : Magnum version) and Acer Pica 61, which is a kind of a successor type for : : the Magnum built by Acer. All these machines have a similar DMA engine, : : Floppy, NCR53C94 SCSI and Sonic Ethernet onboard. This SCSI/Ethernet : : hardware is currently not supported. I thought it would be better to : : use (E)ISA hardware where drivers are very easy to port and to put my : : time on the other parts of the system. We'll be helping out with driver sources from an internal, non-AT&T related project. --lm From lm@neteng.engr.sgi.com Sat Apr 27 00:07:14 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id AAA09656 for ; Sat, 27 Apr 1996 00:07:14 +0200 (MET DST) Received: from sgi.sgi.com (SGI.COM) by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA23241; Sat, 27 Apr 1996 00:06:57 +0200 (MET) Received: from cthulhu.engr.sgi.com by sgi.sgi.com via ESMTP (950405.SGI.8.6.12/910110.SGI) for <@sgi.engr.sgi.com:linux-mips@fnet.fr> id PAA00016; Fri, 26 Apr 1996 15:07:01 -0700 Received: from neteng.engr.sgi.com by cthulhu.engr.sgi.com via ESMTP (950511.SGI.8.6.12.PATCH526/911001.SGI) for <@cthulhu.engr.sgi.com:linux-mips@fnet.fr> id PAA03373; Fri, 26 Apr 1996 15:06:59 -0700 Received: from localhost (lm@localhost) by neteng.engr.sgi.com (950413.SGI.8.6.12/960327.SGI.AUTOCF) via SMTP id PAA02565 for ; Fri, 26 Apr 1996 15:06:53 -0700 Message-Id: <199604262206.PAA02565@neteng.engr.sgi.com> To: linux-mips@fnet.fr From: lm@gate1-neteng.engr.sgi.com (Larry McVoy) Subject: Re: Linux/SGI coming Date: Fri, 26 Apr 1996 15:06:52 -0700 Sender: lm@neteng.engr.sgi.com Content-Length: 2933 Lines: 56 : One of the design ideas in Milo was that there is a lowlevel library : containing the interface to the os. Milo itself uses the library : interface. The library interface again exports functions that look somewhat : like unix systemcalls. Stoned's version 0.27 of Milo will make this : difference much clearer. It might be a good idea to have one common : bootloader package for ARC, SGI and as many others as possible. SGI's stand alone shell takes the exact same approach. There is a library which is basically the SGI BIOS. I'll go check and see if I can give you guys access to all of that. It would be cool to merge this stuff. By the way, I have no great love for SGI's stuff, using it would be done only if it's the easiest path. Having MILO work on SGI hardware would be cool. : Btw - Linux 1.3.95 has a read-only implementation of UFS mainly for use : by Linux/Sparc. There are many variations of UFS in use. I dunno if the : Linux version is what SGI uses but anyway SGI's main filesystem are efs : and xfs in the future. We don't have a UFS file system. : > I know that MIPS has published a bunch of errata on the web, including : > ones for the R4000. Don't know if the 2.2 is one in the list or not, : > but will check my printed copies when I get home. None the less, it : > would be good to have more complete versions of these items if they : > were available. However, I just tried the address I have for this : > data, and I can't find it now. : : I think the information isn't really complete. Yes, I believe you are correct. The current set of info is at the copy center for David & and getting 2 spare sets. This stuff is all SGI NDA material, unfortunately. However, David isn't under NDA and we are giving him a copy. I'm pretty sure that if you will agree not to publish or redistribute **the documents**, I can get you copies and you can redistribute the code that you write using those documents. I think the SGI concern is that if too many people code to our internal interfaces we'll be stuck with 'em. It doesn't make sense to me either, that's just what I was told. Whatever, if you are willing to send a letter swearing you won't copy or redistribute the docs, I'll work on getting you a copy. : > The cross compiler stuff is fairly easy to generate. For the SGI : > platform, it should go well. There is a howto generate cross : > compilers web page hanging off the Linux-MIPS howto page that should : > give you what you need to know to build the goofy things, should you : > require assistance. There are a numnber of minor gotchas, so you : > might want to glance there first. : : Gimme an account on an SGI and you'll have the crossdevelopment a bit : later :-) Really? Let me look into this. I think I could get a Challenge/S outside the firewall as a CVS machine that you could hack on. Try pinging sgi.com and tell me what the response time is like. --lm From drepper%myware.rz.uni-karlsruhe.de@ibm3090.rz.uni-karlsruhe.de Sat Apr 27 02:17:44 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id CAA12455 for ; Sat, 27 Apr 1996 02:17:39 +0200 (MET DST) Received: from ibm3090.rz.uni-karlsruhe.de by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA24604; Sat, 27 Apr 1996 02:17:25 +0200 (MET) Received: from ipd.info.uni-karlsruhe.de by ibm3090.rz.uni-karlsruhe.de (IBM MVS SMTP V2R2.1) with TCP; Sat, 27 Apr 96 02:19:37 MET Received: from sl-kb05.rz.uni-karlsruhe.de (root@sl-kb05.rz.uni-karlsruhe.de [129.13.223.49]) by ipd.info.uni-karlsruhe.de (8.6.4/8.6.4) with ESMTP id CAA18252; Sat, 27 Apr 1996 02:17:20 +0200 Received: from myware (drepper@localhost) by sl-kb05.rz.uni-karlsruhe.de (8.7.5/8.7.5/ud-960406) with ESMTP id BAA27245; Sat, 27 Apr 1996 01:43:12 +0200 Message-Id: <199604262343.BAA27245@sl-kb05.rz.uni-karlsruhe.de> To: linux-mips@fnet.fr Cc: lm@neteng.engr.sgi.com, davem@caip.rutgers.edu Subject: Re: Linux/SGI coming Reply-To: drepper@ipd.info.uni-karlsruhe.de (Ulrich Drepper) In-Reply-To: Your message of "Fri, 26 Apr 1996 13:53:19 -0600" References: <199604261953.NAA26233@rover.village.org> X-Mailer: Mew version 1.05+ on Emacs 19.13 XEmacs Lucid Mime-Version: 1.0 Content-Type: Text/Plain; charset=us-ascii Date: Sat, 27 Apr 1996 01:43:11 +0200 From: Ulrich Drepper Content-Length: 786 Lines: 20 From: Warner Losh Subject: Re: Linux/SGI coming Date: Fri, 26 Apr 1996 13:53:19 -0600 > : The libc implementation is based on the GNU libc which is available only > : as ELF. My home sources are based on snapshot 960210; upgrading to the > : newest libc will > > How does one joint this development effort? :-) Tell me if you are *really* interested in glibc development. Just trying to compile it is not useful because it changes a lot. The status is that glibc really seems to run on i386 and Alpha, the Sparc people also had some successes. -- Uli --------------. drepper@cygnus.com ,-. Rubensstrasse 5 Ulrich Drepper \ ,--------------------' \ 76149 Karlsruhe/Germany Cygnus Support `--' drepper@gnu.ai.mit.edu `------------------------ From imp@rover.village.org Sat Apr 27 17:57:02 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id RAA02048 for ; Sat, 27 Apr 1996 17:57:00 +0200 (MET DST) Received: from rover.village.org by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA24566; Sat, 27 Apr 1996 17:56:43 +0200 (MET) Received: from localhost (localhost [127.0.0.1]) by rover.village.org (8.7.5/8.6.6) with SMTP id JAA02239 for ; Sat, 27 Apr 1996 09:56:45 -0600 (MDT) Message-Id: <199604271556.JAA02239@rover.village.org> To: linux-mips@fnet.fr Subject: Re: Linux/SGI coming In-Reply-To: Your message of Fri, 26 Apr 1996 10:02:38 PDT Date: Sat, 27 Apr 1996 09:56:45 -0600 From: Warner Losh Content-Length: 2304 Lines: 44 : I think I can probably find some resources to put a linux-cvs.sgi.com : machine on the net. We have some spare ports on our internet ethernet. : If I put a machine there with disk space, and provide logins, can you : guys do the rest? With sufficient privs, yes. It will be done. : Note that I might be putting a linux pc there and : hacking it to limit the rate at which the data can move. Has anyone : done this? That's SGI's primary concern, they would like to "reserve" : some bandwidth for other stuff on that T1 line. Hmmm, ordinarily you'd just rate limit FTP. However, since this will be used for CVS operations, etc, etc, that will likely not be enough. It would have to be done in the driver for the ethernet card. I've not done this before on a linux machine, but I think it would be relatively easy to pretend the packets take a really long time to transmit and somehow tie into the clock interrupt (likely via a timeout that gets armed in the transmit complete interrupt that most ethernet cards give). I'd imagine that would be the easiest thing to do. I'd not worry about what packets want to go out, just do it for all of them. That should give the least pathological behavior. I don't think this project would generate that much of a load on the T1 to warrant effort to implement the throttling. Therea aren't that many of us, last time I checked. Hey FTP.FNET.FR guys! How much bandwidth (file transfers, etc) is used now? If only the developers of Linux/MIPS had access to this machine, then it would be much less. The linux kernel at T1 speeds represents about 30 seconds of T1 time. Generally development activity will push accross patches that are much smaller than that. There are about 10-15 linux mips developers right now that are active. However, it is your machine and your bandwidth, so do what you think is necessary for this to be allowed to happen... Actually another idea springs to mind. It may be a simple matter to hack the default TOS to be really horrible. That is if your gateway routers do the right thing with TOS (given that you are SGI, and who I know has worked on the TCP/IP for your SGI boxes, the chances are good that this will do the right thing). That way the packets from the linux box will always be last in the queue. Warner From tor@spacetec.no Sat Apr 27 18:44:24 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id SAA03045 for ; Sat, 27 Apr 1996 18:44:22 +0200 (MET DST) Received: from pallas.spacetec.no by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA02400; Sat, 27 Apr 1996 18:44:04 +0200 (MET) Received: (from tor@localhost) by pallas.spacetec.no (8.7.1/8.7.1) id SAA22773 for linux-mips@fnet.fr; Sat, 27 Apr 1996 18:48:57 +0200 Message-Id: <199604271648.SAA22773@pallas.spacetec.no> From: tor@spacetec.no (Tor Arntsen) Date: Sat, 27 Apr 1996 18:48:57 +0200 In-Reply-To: Ralf Baechle "Re: Linux/SGI coming" (Apr 26, 19:15) X-Mailer: Mail User's Shell (7.2.5 10/14/92) To: linux-mips@fnet.fr Subject: Re: Linux/SGI coming Content-Length: 1476 Lines: 30 On Apr 26, 19:15, Ralf Baechle wrote: >The crosscompiler environment is based on GCC 2.7.2 and Binutils 2.6. The >required patches that fix lots of bugs and add Linux/MIPS support for >a.out, ELF little/big endian byte order are available on ftp.fnet.fr. >Their installation isn't easy but Dave shouldn't have to much trouble >with this. > >There are some binary packages on ftp.fnet.fr for Linux/i386 host. I'd like >to have packages for more hosts, so if someone sends me binaries we'll put >the online. (I could create packages for Stun's OSes if someone wants them) I (finally) have the SGI Linux/MIPS cross-compiling tools packed up ready for installation in the crossdev directory (I made them once during Easter, but due to an unfortunate reboot of my Challenge-L I lost it all as I kept it in /tmp! :-/ -- and then I didn't have more time for a while). (It's not on ftp.fnet.fr yet as I didn't find any writable incoming/ to ftp it to. I've emailed Luc and Stoned for advise about where to put it.) However, these tools are for little-endian Linux. I have an Olivetti M700-10 so I don't need big endian (I'm just using the SGI for cross-compiling). I'm sure Dave & co can easily (and may even prefer to) build the big-endian versions themselves, but as I have Ralf's required patches + fix for the GCC strength-reduce bug already installed in my sources I could produce an eb version quickly I guess -- if someone wants it. Just let me know. Cheers, Tor From lm@neteng.engr.sgi.com Sat Apr 27 19:27:22 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id TAA03949 for ; Sat, 27 Apr 1996 19:27:20 +0200 (MET DST) Received: from sgi.sgi.com (SGI.COM) by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA09327; Sat, 27 Apr 1996 19:27:05 +0200 (MET) Received: from cthulhu.engr.sgi.com by sgi.sgi.com via ESMTP (950405.SGI.8.6.12/910110.SGI) for <@sgi.engr.sgi.com:linux-mips@fnet.fr> id KAA01028; Sat, 27 Apr 1996 10:27:13 -0700 Received: from neteng.engr.sgi.com by cthulhu.engr.sgi.com via ESMTP (950511.SGI.8.6.12.PATCH526/911001.SGI) for <@cthulhu.engr.sgi.com:linux-mips@fnet.fr> id KAA11014; Sat, 27 Apr 1996 10:27:01 -0700 Received: (from lm@localhost) by neteng.engr.sgi.com (950413.SGI.8.6.12/960327.SGI.AUTOCF) id KAA23439 for linux-mips@fnet.fr; Sat, 27 Apr 1996 10:25:51 -0700 Date: Sat, 27 Apr 1996 10:25:51 -0700 From: lm@neteng.engr.sgi.com (Larry McVoy) Message-Id: <199604271725.KAA23439@neteng.engr.sgi.com> To: linux-mips@fnet.fr Subject: Re: Linux/SGI coming Content-Length: 1159 Lines: 22 I think we'll wait to see if the machine becomes a problem before hacking the interface to go slow. It will probably be fine. --lm P.S. I'll try and have something set up next week. People that wish development account, please send me your passwd entry and a short description of why you want the account. The account policy is that everyone who asks gets one, but you get bumped if you become a pain in the ass (which isn't going to happen because we all love everyone :-) Actually, if all you are going to need is ftp access to the tree, then don't ask for an account, you are the pressure I'll need to make sure that tree is available. Another thing: is there someone out there that is a developer that wants to setup the stuff we need to mirror the current linux/MIPs tree? --- Larry McVoy lm@sgi.com http://reality.sgi.com/lm (415) 933-1804 Copyright 1996, all rights reserved. Microsoft Network is prohibited from redistributing this work in any form, in whole or in part without license. License to distribute this work is available to Microsoft at $500. Transmission without permission constitutes an agreement to these terms. From lm@neteng.engr.sgi.com Sat Apr 27 19:30:02 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id TAA04027 for ; Sat, 27 Apr 1996 19:30:00 +0200 (MET DST) Received: from sgi.sgi.com (SGI.COM) by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA09796; Sat, 27 Apr 1996 19:29:43 +0200 (MET) Received: from cthulhu.engr.sgi.com by sgi.sgi.com via ESMTP (950405.SGI.8.6.12/910110.SGI) for <@sgi.engr.sgi.com:linux-mips@fnet.fr> id KAA01129; Sat, 27 Apr 1996 10:29:51 -0700 Received: from neteng.engr.sgi.com by cthulhu.engr.sgi.com via ESMTP (950511.SGI.8.6.12.PATCH526/911001.SGI) for <@cthulhu.engr.sgi.com:linux-mips@fnet.fr> id KAA11062; Sat, 27 Apr 1996 10:29:43 -0700 Received: (from lm@localhost) by neteng.engr.sgi.com (950413.SGI.8.6.12/960327.SGI.AUTOCF) id KAA23545 for linux-mips@fnet.fr; Sat, 27 Apr 1996 10:28:38 -0700 Date: Sat, 27 Apr 1996 10:28:38 -0700 From: lm@neteng.engr.sgi.com (Larry McVoy) Message-Id: <199604271728.KAA23545@neteng.engr.sgi.com> To: linux-mips@fnet.fr Subject: Re: Linux/SGI coming Content-Length: 1626 Lines: 32 From: tor@spacetec.no (Tor Arntsen) $ I (finally) have the SGI Linux/MIPS cross-compiling tools packed up ready $ for installation in the crossdev directory (I made them once during Easter, $ but due to an unfortunate reboot of my Challenge-L I lost it all as I kept $ it in /tmp! :-/ -- and then I didn't have more time for a while). $ $ (It's not on ftp.fnet.fr yet as I didn't find any writable incoming/ to ftp $ it to. I've emailed Luc and Stoned for advise about where to put it.) I'll want to suck these down onto linux.sgi.com next week. Do you have both binary and source distributions? $ However, these tools are for little-endian Linux. I have an Olivetti M700-10 $ so I don't need big endian (I'm just using the SGI for cross-compiling). $ I'm sure Dave & co can easily (and may even prefer to) build the big-endian $ versions themselves, but as I have Ralf's required patches + fix for the $ GCC strength-reduce bug already installed in my sources I could produce an $ eb version quickly I guess -- if someone wants it. Just let me know. If you want to build the other endian so we are all set up, I'll snarf up that stuff and try and build an SGI stand alone shell using it. Interesting little experiment, huh? Lemme know, --lm --- Larry McVoy lm@sgi.com http://reality.sgi.com/lm (415) 933-1804 Copyright 1996, all rights reserved. Microsoft Network is prohibited from redistributing this work in any form, in whole or in part without license. License to distribute this work is available to Microsoft at $500. Transmission without permission constitutes an agreement to these terms. From tor@spacetec.no Sat Apr 27 19:44:07 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id TAA04453 for ; Sat, 27 Apr 1996 19:44:06 +0200 (MET DST) Received: from pallas.spacetec.no by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA11266; Sat, 27 Apr 1996 19:43:50 +0200 (MET) Received: (from tor@localhost) by pallas.spacetec.no (8.7.1/8.7.1) id TAA22812 for linux-mips@fnet.fr; Sat, 27 Apr 1996 19:48:48 +0200 Message-Id: <199604271748.TAA22812@pallas.spacetec.no> From: tor@spacetec.no (Tor Arntsen) Date: Sat, 27 Apr 1996 19:48:48 +0200 In-Reply-To: lm@neteng.engr.sgi.com (Larry McVoy) "Re: Linux/SGI coming" (Apr 27, 18:29) X-Mailer: Mail User's Shell (7.2.5 10/14/92) To: linux-mips@fnet.fr Subject: Re: Linux/SGI coming Content-Length: 867 Lines: 19 On Apr 27, 18:29, Larry McVoy wrote: >I'll want to suck these down onto linux.sgi.com next week. Do you have >both binary and source distributions? I can make source distributions too, no problem (that would just be the binutils and gcc sources with the patches I mentioned applied). >If you want to build the other endian so we are all set up, I'll snarf >up that stuff and try and build an SGI stand alone shell using it. >Interesting little experiment, huh? Yep.. I have big endian binutils made now, but I just ran into one little snag with the gcc build. I think I ran into the same one when I made my own little endian version some months ago, I should hopefully have it sorted out quite soon. But I have to pop over to my house for food now, so I'll probably finish this tomorrow. Should at least have it ready for next week if all goes well. Tor From davem@caip.rutgers.edu Sat Apr 27 23:58:16 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id XAA09628 for ; Sat, 27 Apr 1996 23:58:15 +0200 (MET DST) Received: from caipfs.rutgers.edu by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA25553; Sat, 27 Apr 1996 23:58:00 +0200 (MET) Received: from huahaga.rutgers.edu (huahaga.rutgers.edu [128.6.155.53]) by caipfs.rutgers.edu (8.6.9+bestmx+oldruq+newsunq+grosshack/8.6.9) with ESMTP id RAA24819 for ; Sat, 27 Apr 1996 17:58:10 -0400 Received: (davem@localhost) by huahaga.rutgers.edu (8.6.9+bestmx+oldruq+newsunq+grosshack/8.6.9) id RAA07259; Sat, 27 Apr 1996 17:58:05 -0400 Date: Sat, 27 Apr 1996 17:58:05 -0400 Message-Id: <199604272158.RAA07259@huahaga.rutgers.edu> From: "David S. Miller" To: linux-mips@fnet.fr Cc: linux-mips@fnet.fr In-Reply-To: <199604261702.KAA26060@neteng.engr.sgi.com> (lm@neteng.engr.sgi.com) Subject: Re: Linux/SGI coming Content-Length: 494 Lines: 14 Date: Fri, 26 Apr 1996 10:02:38 -0700 From: lm@neteng.engr.sgi.com (Larry McVoy) Note that I might be putting a linux pc there and hacking it to limit the rate at which the data can move. Has anyone done this? That's SGI's primary concern, they would like to "reserve" some bandwidth for other stuff on that T1 line. Besides the "OSPF routing CISCO penguin", I thought cisco (or any) router could provide this functionality. Later, David S. Miller davem@caip.rutgers.edu From tor@spacetec.no Sun Apr 28 01:23:41 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id BAA11641 for ; Sun, 28 Apr 1996 01:23:40 +0200 (MET DST) Received: from pallas.spacetec.no by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA29364; Sun, 28 Apr 1996 01:23:24 +0200 (MET) Received: (from tor@localhost) by pallas.spacetec.no (8.7.1/8.7.1) id BAA23007 for linux-mips@fnet.fr; Sun, 28 Apr 1996 01:28:24 +0200 Message-Id: <199604272328.BAA23007@pallas.spacetec.no> From: tor@spacetec.no (Tor Arntsen) Date: Sun, 28 Apr 1996 01:28:23 +0200 In-Reply-To: Ralf Baechle "Re: Linux/SGI coming" (Apr 26, 21:17) X-Mailer: Mail User's Shell (7.2.5 10/14/92) To: linux-mips@fnet.fr Subject: Re: Linux/SGI coming Content-Length: 1411 Lines: 47 Hi Ralf, I found a problem when building gcc 2.7.2 (sources + your patches) as cross-compiler on mips (sgi) for big-endian elf. Patch follows: Tor --- config/mips/linux.h.old Thu Apr 4 15:45:33 1996 +++ config/mips/linux.h Sun Apr 28 00:56:51 1996 @@ -41,16 +41,6 @@ #define MIPS_DEFAULT_GVALUE 0 /* - * Specify predefined symbols in preprocessor. - */ -#undef CPP_PREDEFINES -#ifdef WITH_SHARED_LIBS -#define CPP_PREDEFINES "-D__ELF__ -D__PIC__ -D__pic__ -Dunix -Dmips -DR3000 -DMIPSEB -Dlinux -Asystem(unix) -Asystem(posix) -Acpu(mips) -Amachine(mips)" -#else -#define CPP_PREDEFINES "-D__ELF__ -Dunix -Dmips -DR3000 -DMIPSEB -Dlinux -Asystem(unix) -Asystem(posix) -Acpu(mips) -Amachine(mips)" -#endif - -/* * MIPS uses $ but . seems standard for Sys V */ #undef LOCAL_LABEL_PREFIX @@ -106,9 +96,19 @@ } while (0) /* - * This also includes + * This also includes (which undefines CPP_PREDEFINES) */ #include + +/* + * Specify predefined symbols in preprocessor. + */ +#undef CPP_PREDEFINES +#ifdef WITH_SHARED_LIBS +#define CPP_PREDEFINES "-D__ELF__ -D__PIC__ -D__pic__ -Dunix -Dmips -DR3000 -DMIPSEB -Dlinux -Asystem(unix) -Asystem(posix) -Acpu(mips) -Amachine(mips)" +#else +#define CPP_PREDEFINES "-D__ELF__ -Dunix -Dmips -DR3000 -DMIPSEB -Dlinux -Asystem(unix) -Asystem(posix) -Acpu(mips) -Amachine(mips)" +#endif #undef CPP_SPEC #define CPP_SPEC "\ From ralf@Julia.DE Sun Apr 28 02:06:22 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id CAA12592 for ; Sun, 28 Apr 1996 02:06:20 +0200 (MET DST) Received: from alles.intern.julia.de ([194.221.49.2]) by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA29949; Sun, 28 Apr 1996 02:05:59 +0200 (MET) Received: from pris.intern.julia.de (pris.intern.julia.de [194.221.49.37]) by alles.intern.julia.de (8.7.5/8.7.3) with ESMTP id BAA02397 for ; Sun, 28 Apr 1996 01:03:18 +0200 From: Ralf Baechle Received: (from ralf@localhost) by pris.intern.julia.de (8.7.5/8.7.3) id CAA07264 for linux-mips@fnet.fr; Sun, 28 Apr 1996 02:05:19 +0200 (MET DST) Message-Id: <199604280005.CAA07264@pris.intern.julia.de> Subject: Re: Linux/SGI coming To: linux-mips@fnet.fr Date: Sun, 28 Apr 1996 02:05:18 +0200 (MET DST) In-Reply-To: <199604272158.RAA07259@huahaga.rutgers.edu> from "David S. Miller" at Apr 27, 96 05:58:05 pm Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Length: 1529 Lines: 34 Hi, > Date: Fri, 26 Apr 1996 10:02:38 -0700 > From: lm@neteng.engr.sgi.com (Larry McVoy) > > Note that I might be putting a linux pc there and > hacking it to limit the rate at which the data can move. Has anyone > done this? That's SGI's primary concern, they would like to "reserve" > some bandwidth for other stuff on that T1 line. > > Besides the "OSPF routing CISCO penguin", I thought cisco (or any) > router could provide this functionality. Hmmm... "My" domain Julia.DE is bandwidth limited by a serial line between the providers Cisco's 7000 and the terminal adapter connected to our leased line. So I suppose just limiting the bandwidth at the router doesn't work as expected. One might implement this by routing the Internet traffic via a SLIP/PPP serial line set to the maximum rate that SGI is willing to give. This is just what Larry suggested. What however is definately is working is priority routing with Cisco. Just set the machine's bulk transfer services on low priority. This is probably preferable as high bandwidth is available during low traffic times. Aside of this I belive that it will take quite some time until the Linux traffic from that site reaches a level that makes bandwidth limiting neccesary. I hope that I'll have the neccesary bandwidth available in the near future so that I can take over the Internet services, if required. (Our provider has > 8mbit/s unused bandwith available :-) Ralf (Hey Paolo, your hour has come. Tell us about your routers :-) From ralf@Julia.DE Sun Apr 28 02:23:53 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id CAA13219 for ; Sun, 28 Apr 1996 02:23:51 +0200 (MET DST) Received: from alles.intern.julia.de ([194.221.49.2]) by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA00346; Sun, 28 Apr 1996 02:23:23 +0200 (MET) Received: from pris.intern.julia.de (pris.intern.julia.de [194.221.49.37]) by alles.intern.julia.de (8.7.5/8.7.3) with ESMTP id BAA02423 for ; Sun, 28 Apr 1996 01:20:48 +0200 From: Ralf Baechle Received: (from ralf@localhost) by pris.intern.julia.de (8.7.5/8.7.3) id CAA07284 for linux-mips@fnet.fr; Sun, 28 Apr 1996 02:22:48 +0200 (MET DST) Message-Id: <199604280022.CAA07284@pris.intern.julia.de> Subject: The goal ... To: linux-mips@fnet.fr Date: Sun, 28 Apr 1996 02:22:48 +0200 (MET DST) Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Length: 604 Lines: 15 which there is named Linux 1.3.96 is reached. My Acer is now running 1.3.96 since several hours. The system is a bit less stable as I removed some of the (evil ...) reliability hacks and probably also caught some buglets during my fastupgrade. If you consider that I use a 486/33 with 8mb you can easily see that I didn't have too much time for testing ... The bad message is that my university's fileserver (Blowlaris ...) had to be reformated once again (The *absolutly* indestructible logging filesystem ...) and so you'll have to wait a bit until I upload the stuff. Ok, happy hacking, Ralf From imp@rover.village.org Sun Apr 28 05:33:46 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id FAA18055 for ; Sun, 28 Apr 1996 05:33:44 +0200 (MET DST) Received: from rover.village.org by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA02637; Sun, 28 Apr 1996 05:33:08 +0200 (MET) Received: from localhost (localhost [127.0.0.1]) by rover.village.org (8.7.5/8.6.6) with SMTP id VAA05905 for ; Sat, 27 Apr 1996 21:33:00 -0600 (MDT) Message-Id: <199604280333.VAA05905@rover.village.org> To: linux-mips@fnet.fr Subject: Re: The goal ... In-Reply-To: Your message of Sun, 28 Apr 1996 02:22:48 +0200 Date: Sat, 27 Apr 1996 21:33:00 -0600 From: Warner Losh Content-Length: 1393 Lines: 32 : so you'll have to wait a bit until I upload the stuff. Ralf, you are such a tease :-) I've manage to compile the latest kernel (1.3.91 is the latest that I grabbed) for the Deskstation. I did it w/o procfs, and found and fixed a couple of bugs in that support. One of the things I noticed was the load address for my machine is now 0x80100000. Cool! It also looks like the interrupt handlers are copied down. Isn't it dangerous to do that inplace copy? Shouldn't the code have a if (begin_except != KSEG0) protecting the copy? Anyway, I'm not having much luck. For reasons unknown I find that I'm unable to get past paging_init() in the boot sequence (at least I get a 'b' in the upper left hand corner of the screen and I've hacked start_kernel to look like: ... *(unsigned char *)0xa00a0000 = 'b'; memory_start = paging_init(memory_start,memory_end); *(unsigned char *)0xa00a0000 = 'c'; ... which is a pretty strong smoking gun). I recall 1.3.64 hanging on boot for reasons unknown, or was that 1.3.82. It didn't happen on 1.3.62. I'm off to find out why this is happening to me. I wish there was some easy way to get early panics to always print something... Ah, a problem to chase when I have time... Anyway, I thought I'd let people know about this, since I had to hassle Ralf about telling us about the cool patches and then telling us we'd just have to wait. Warner From imp@rover.village.org Sun Apr 28 07:48:18 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id HAA20680 for ; Sun, 28 Apr 1996 07:48:18 +0200 (MET DST) Received: from rover.village.org by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA03750; Sun, 28 Apr 1996 07:48:02 +0200 (MET) Received: from localhost (localhost [127.0.0.1]) by rover.village.org (8.7.5/8.6.6) with SMTP id XAA07258 for ; Sat, 27 Apr 1996 23:48:13 -0600 (MDT) Message-Id: <199604280548.XAA07258@rover.village.org> To: linux-mips@fnet.fr Subject: Deskstation ... No Go Date: Sat, 27 Apr 1996 23:48:13 -0600 From: Warner Losh Content-Length: 580 Lines: 16 Looks like it is dying in free_area_init. Sometimes in the do ... while loop, other times in the for loop, depending on the debug hacks that I add: *(unsigned char *)0xa00a0006 = ' '; or (*(unsigned char *)0xa00a0006)++; When it makes it to the for loop, it executes about 100ish times and then hangs. I have 32M of memory, 31 of which can be used by the kernel due to the load address. Anyway, I'm punting for tonight. If anybody has chased a bug in this code before and can offer some words of advise, I'd be happy to listen :-). Otherwise, I start here tomorrow. Warner From Stoned.Elipot@univ-evry.fr Sun Apr 28 19:04:40 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id TAA04126 for ; Sun, 28 Apr 1996 19:04:39 +0200 (MET DST) Received: from amiante.univ-evry.fr by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA11058; Sun, 28 Apr 1996 19:04:23 +0200 (MET) Received: from bandsept.univ-evry.fr (bandsept.univ-evry.fr [192.93.214.19]) by amiante.univ-evry.fr with ESMTP (8.6.12/96.04.09/louis-blanc); id TAA21462; Sun, 28 Apr 1996 19:04:36 +0200 Received: from localhost.univ-evry.fr (localhost.univ-evry.fr [127.0.0.1]) by bandsept.univ-evry.fr with SMTP (8.6.12/94.01.26); id TAA16841; Sun, 28 Apr 1996 19:03:58 +0200 Message-Id: <199604281703.TAA16841@bandsept.univ-evry.fr> To: linux-mips@fnet.fr (The List) Subject: Milo Arc/SGI, question about 1.3.91 and ftp.fnet.fr bandwidth X-It: bandsept (wrong OS, but right CPU) X-Mailer: MH 6.8.3, MH-E 5.0.2 Date: Sun, 28 Apr 1996 19:03:52 +0200 From: Stoned Elipot Content-Length: 1231 Lines: 24 Hi all, In any case I'm trying to make Milo sources to serve as a framework for a future SGI bootstrap loader, but is it really usefull ? I end up using GNUism in Milo's Makefile is it an inconvient ? I'd like to ear your comments on this. My idea is to make the whole package ready to compile or cross-compile. So I guess using relying on GNU make for 'compilation' aka under Linux/MIPS is not a problem but for 'cross-compilation' ? I'll also try to make it a 'configure;make' package. Regarding kernel 1.3.91, I'm not able to launch a /bin/sh. This /bin/sh is ash - from sources tuned by Ralf available on ftp.fnet.fr - compiled in ELF against GNU libc 951212 - the lastest available Ralf's work. But the kernel_thread(do_shell......) in init/main.c:init() keeps getting launched and it says: 'child xxx died with code 000b' The do_execve() call in arch/mips/kernel/syscall.c:sys_execve() seems good anyway. I had also tried the ash binary from Ralf -> same thing... :( Any hints ? And about the bandwidth used by linux/MIPS on ftp.fnet.fr we should wait for Luc coming back from weekend but I recall that it was more important than the other stuff on this FTP server - which is in fact EUnet France one's. Cheers, Stoned. From tor@spacetec.no Sun Apr 28 19:16:56 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id TAA04519 for ; Sun, 28 Apr 1996 19:16:55 +0200 (MET DST) Received: from pallas.spacetec.no by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA11193; Sun, 28 Apr 1996 19:16:39 +0200 (MET) Received: (from tor@localhost) by pallas.spacetec.no (8.7.1/8.7.1) id TAA23469 for linux-mips@fnet.fr; Sun, 28 Apr 1996 19:21:46 +0200 Message-Id: <199604281721.TAA23469@pallas.spacetec.no> From: tor@spacetec.no (Tor Arntsen) Date: Sun, 28 Apr 1996 19:21:46 +0200 In-Reply-To: Stoned Elipot "Milo Arc/SGI, question about 1.3.91 and ftp.fnet.fr bandwidth" (Apr 28, 18:04) X-Mailer: Mail User's Shell (7.2.5 10/14/92) To: linux-mips@fnet.fr Subject: Re: Milo Arc/SGI, question about 1.3.91 and ftp.fnet.fr bandwidth Content-Length: 940 Lines: 18 On Apr 28, 18:04, Stoned Elipot wrote: >In any case I'm trying to make Milo sources to serve as a framework >for a future SGI bootstrap loader, but is it really usefull ? >I end up using GNUism in Milo's Makefile is it an inconvient ? I'd like >to ear your comments on this. My idea is to make the whole package ready >to compile or cross-compile. So I guess using relying on GNU make for >'compilation' aka under Linux/MIPS is not a problem but for >'cross-compilation' ? I'll also try to make it a 'configure;make' package. IMHO it's perfectly fine to rely on GNU make in any case. The company I'm working for makes software for different customers and the only way to stay sane and get real work done is to use GNU make on all platforms. As far as SGI is concerned I'm also using gawk instead of SGI awk (which fails) for the Linux cross-compile. But then again SGI awk (on 5.3) is really oawk, and I didn't try SGI's nawk. Tor From ralf@Julia.DE Sun Apr 28 20:01:20 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id UAA05624 for ; Sun, 28 Apr 1996 20:01:19 +0200 (MET DST) Received: from alles.intern.julia.de ([194.221.49.2]) by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA11692; Sun, 28 Apr 1996 20:00:57 +0200 (MET) Received: from pris.intern.julia.de (pris.intern.julia.de [194.221.49.37]) by alles.intern.julia.de (8.7.5/8.7.3) with ESMTP id SAA03197 for ; Sun, 28 Apr 1996 18:58:27 +0200 From: Ralf Baechle Received: (from ralf@localhost) by pris.intern.julia.de (8.7.5/8.7.3) id UAA07660 for linux-mips@fnet.fr; Sun, 28 Apr 1996 20:00:19 +0200 (MET DST) Message-Id: <199604281800.UAA07660@pris.intern.julia.de> Subject: Re: The goal ... To: linux-mips@fnet.fr Date: Sun, 28 Apr 1996 20:00:19 +0200 (MET DST) In-Reply-To: <199604280333.VAA05905@rover.village.org> from "Warner Losh" at Apr 27, 96 09:33:00 pm Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Length: 2124 Lines: 49 Hi, > : so you'll have to wait a bit until I upload the stuff. > > Ralf, you are such a tease :-) Oh, I just thought I should keep you informed :-> > I've manage to compile the latest kernel (1.3.91 is the latest that I > grabbed) for the Deskstation. I did it w/o procfs, and found and > fixed a couple of bugs in that support. One of the things I noticed > was the load address for my machine is now 0x80100000. Cool! It also > looks like the interrupt handlers are copied down. Isn't it dangerous > to do that inplace copy? Shouldn't the code have a if (begin_except > != KSEG0) protecting the copy? Why should an inplace copy be dangerous? Ok, for Jazz machines the kernel needs several *microseconds* longer to startup. A long time if you're in hurry :-) > Anyway, I'm not having much luck. For reasons unknown I find that I'm > unable to get past paging_init() in the boot sequence (at least I get > a 'b' in the upper left hand corner of the screen and I've hacked > start_kernel to look like: > ... > *(unsigned char *)0xa00a0000 = 'b'; > memory_start = paging_init(memory_start,memory_end); > *(unsigned char *)0xa00a0000 = 'c'; > ... > which is a pretty strong smoking gun). I definately bugged mem_init() for your machine which frees the complete low 640k including the exception handlers ... As paging_init() gets called before mem_init() this cannot be your problem. Hmmm... I recall 1.3.64 hanging on > boot for reasons unknown, or was that 1.3.82. It didn't happen on > 1.3.62. I'm off to find out why this is happening to me. I wish > there was some easy way to get early panics to always print > something... Ah, a problem to chase when I have time... >From my Tyne hackery I've got still a tiny console in arch/mips/lib/tinycon.c. Adjust the videoparams and you have a debugging console. Usage should be clear after a short look into the code. With print_string() and sprintf() in combination you can the same things as with printk(). tinycon.c is also easily useable for other machines with a PC style character based display. Ralf From ralf@Julia.DE Sun Apr 28 20:24:22 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id UAA06196 for ; Sun, 28 Apr 1996 20:24:21 +0200 (MET DST) Received: from alles.intern.julia.de ([194.221.49.2]) by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA11926; Sun, 28 Apr 1996 20:23:58 +0200 (MET) Received: from pris.intern.julia.de (pris.intern.julia.de [194.221.49.37]) by alles.intern.julia.de (8.7.5/8.7.3) with ESMTP id TAA03220 for ; Sun, 28 Apr 1996 19:21:38 +0200 From: Ralf Baechle Received: (from ralf@localhost) by pris.intern.julia.de (8.7.5/8.7.3) id UAA07677 for linux-mips@fnet.fr; Sun, 28 Apr 1996 20:23:29 +0200 (MET DST) Message-Id: <199604281823.UAA07677@pris.intern.julia.de> Subject: Re: Milo Arc/SGI, question about 1.3.91 and ftp.fnet.fr bandwidth To: linux-mips@fnet.fr Date: Sun, 28 Apr 1996 20:23:29 +0200 (MET DST) In-Reply-To: <199604281703.TAA16841@bandsept.univ-evry.fr> from "Stoned Elipot" at Apr 28, 96 07:03:52 pm Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Length: 1958 Lines: 47 Hi all, > In any case I'm trying to make Milo sources to serve as a framework > for a future SGI bootstrap loader, but is it really usefull ? I think we shouldn't answer this question before we've sawn SGI's standalone shell. I remember some people on this list saying that they'd like to have a standalone shell for their ARC machines. > I end up using GNUism in Milo's Makefile is it an inconvient ? I'd like > to ear your comments on this. My idea is to make the whole package ready > to compile or cross-compile. So I guess using relying on GNU make for > 'compilation' aka under Linux/MIPS is not a problem but for > 'cross-compilation' ? I'll also try to make it a 'configure;make' package. If you use autoconf you can live without gmake much easier. While beliving that it's good not to rely on the GNU tools this is not really necessary. After all they're available for free :-) Hmm... if you want to configure the Milo-package (possibly supporting multiple host environments in the future) we'll need a way for Milo to identify which bootloader exactly is required. How about adding an entry boot_method : ARC or boot_method : SGI to /proc/cpuinfo so that Milo can autoconfigure completly for native compiles? > Regarding kernel 1.3.91, I'm not able to launch a /bin/sh. This /bin/sh > is ash - from sources tuned by Ralf available on ftp.fnet.fr - compiled in > ELF against GNU libc 951212 - the lastest available Ralf's work. But > the kernel_thread(do_shell......) in init/main.c:init() keeps getting launched > and it says: 'child xxx died with code 000b' > The do_execve() call in arch/mips/kernel/syscall.c:sys_execve() seems good > anyway. > I had also tried the ash binary from Ralf -> same thing... :( > Any hints ? Sorry, no. I've been running the latest kernels on Wayne's M700 with success also. So it can't be a M700 specific problem. Could you send me a more detailed description of your setup? Ralf From ralf@Julia.DE Sun Apr 28 20:53:51 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id UAA06872 for ; Sun, 28 Apr 1996 20:53:50 +0200 (MET DST) Received: from alles.intern.julia.de ([194.221.49.2]) by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA12243; Sun, 28 Apr 1996 20:53:21 +0200 (MET) Received: from pris.intern.julia.de (pris.intern.julia.de [194.221.49.37]) by alles.intern.julia.de (8.7.5/8.7.3) with ESMTP id TAA03255 for ; Sun, 28 Apr 1996 19:50:55 +0200 From: Ralf Baechle Received: (from ralf@localhost) by pris.intern.julia.de (8.7.5/8.7.3) id UAA07728 for linux-mips@fnet.fr; Sun, 28 Apr 1996 20:52:46 +0200 (MET DST) Message-Id: <199604281852.UAA07728@pris.intern.julia.de> Subject: Re: Linux/SGI coming To: linux-mips@fnet.fr Date: Sun, 28 Apr 1996 20:52:46 +0200 (MET DST) In-Reply-To: <199604272328.BAA23007@pallas.spacetec.no> from "Tor Arntsen" at Apr 28, 96 01:28:23 am Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Length: 569 Lines: 13 Hi all, > I found a problem when building gcc 2.7.2 (sources + your patches) as > cross-compiler on mips (sgi) for big-endian elf. Patch follows: Thanks, I already knew this one; I just thought I'd already have released gcc-2.7.2. I'mcurrently doing a little bit rework on my GCC port to make it look more similar to the SGI/MIPS ABI port. Also the __builtin_return() function is cmpletly broken thus making the GNU libc profiling code useless. I however will put gcc-2.7.2-2.diffs.gz on ftp.fnet.fr including everything I've yet fixed/changed tomorrow. Ralf From imp@rover.village.org Sun Apr 28 23:20:27 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id XAA10025 for ; Sun, 28 Apr 1996 23:20:25 +0200 (MET DST) Received: from rover.village.org by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA14337; Sun, 28 Apr 1996 23:20:08 +0200 (MET) Received: from localhost (localhost [127.0.0.1]) by rover.village.org (8.7.5/8.6.6) with SMTP id PAA11276 for ; Sun, 28 Apr 1996 15:20:19 -0600 (MDT) Message-Id: <199604282120.PAA11276@rover.village.org> To: linux-mips@fnet.fr Subject: Re: Milo Arc/SGI, question about 1.3.91 and ftp.fnet.fr bandwidth In-Reply-To: Your message of Sun, 28 Apr 1996 19:21:46 +0200 Date: Sun, 28 Apr 1996 15:20:18 -0600 From: Warner Losh Content-Length: 499 Lines: 11 : IMHO it's perfectly fine to rely on GNU make in any case. The company I'm : working for makes software for different customers and the only way to stay : sane and get real work done is to use GNU make on all platforms. I agree. The other option is to use imake, which has enough difficulties outside of the X world to make it less convenient to use than gnu make. Given that gnu make has been ported to a lot of places, and is the Linux standard make, I'd say that gnuisms are fine. Warner From imp@rover.village.org Sun Apr 28 23:24:09 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id XAA10115 for ; Sun, 28 Apr 1996 23:24:09 +0200 (MET DST) Received: from rover.village.org by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA14362; Sun, 28 Apr 1996 23:23:41 +0200 (MET) Received: from localhost (localhost [127.0.0.1]) by rover.village.org (8.7.5/8.6.6) with SMTP id PAA11302 for ; Sun, 28 Apr 1996 15:23:43 -0600 (MDT) Message-Id: <199604282123.PAA11302@rover.village.org> To: linux-mips@fnet.fr Subject: Re: The goal ... In-Reply-To: Your message of Sun, 28 Apr 1996 20:00:19 +0200 Date: Sun, 28 Apr 1996 15:23:43 -0600 From: Warner Losh Content-Length: 1146 Lines: 26 : Why should an inplace copy be dangerous? Ok, for Jazz machines the kernel : needs several *microseconds* longer to startup. A long time if you're : in hurry :-) In the past I've been bit by having inplace copy routines destroy what was being copied. : I definately bugged mem_init() for your machine which frees the complete : low 640k including the exception handlers ... As paging_init() gets : called before mem_init() this cannot be your problem. Hmmm... I'll have to take a look at that. The last kernel I built is dying in free_area_init, so I don't think your freeing the entire lower part of memory is the problem either... : Adjust the videoparams and you have a debugging console. Usage should : be clear after a short look into the code. With print_string() and : sprintf() in combination you can the same things as with printk(). : tinycon.c is also easily useable for other machines with a PC style : character based display. I'll have to whip that into shape, since I need to get the values of things like the end of memory, etc. I'd rather not do the gdb debugging yet, since I think that would be harder... Warner From imp@rover.village.org Sun Apr 28 23:27:26 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id XAA10189 for ; Sun, 28 Apr 1996 23:27:26 +0200 (MET DST) Received: from rover.village.org by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA14389; Sun, 28 Apr 1996 23:27:09 +0200 (MET) Received: from localhost (localhost [127.0.0.1]) by rover.village.org (8.7.5/8.6.6) with SMTP id PAA11335 for ; Sun, 28 Apr 1996 15:27:21 -0600 (MDT) Message-Id: <199604282127.PAA11335@rover.village.org> To: linux-mips@fnet.fr Subject: Re: Milo Arc/SGI, question about 1.3.91 and ftp.fnet.fr bandwidth In-Reply-To: Your message of Sun, 28 Apr 1996 20:23:29 +0200 Date: Sun, 28 Apr 1996 15:27:20 -0600 From: Warner Losh Content-Length: 838 Lines: 17 : > In any case I'm trying to make Milo sources to serve as a framework : > for a future SGI bootstrap loader, but is it really usefull ? : : I think we shouldn't answer this question before we've sawn SGI's : standalone shell. I remember some people on this list saying that they'd : like to have a standalone shell for their ARC machines. Agreed. One of the things I'd like to see is a MASH or something like that for my machine. It would make the bootstrapping procedure easier. Also, I'd like to have some way to partition and newfs disk areas from the ARC bios (even if it only works from "MOST" of the implementations). I now have a disk with 32M for kernels (DOS partition) and 280M for hacking (soon to be ext2fs). Ralf, will that be enough? Also, are disk roots supported yet. setup.c leads me to believe not. Warner From ralf@Julia.DE Mon Apr 29 01:38:44 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id BAA13414 for ; Mon, 29 Apr 1996 01:38:43 +0200 (MET DST) Received: from alles.intern.julia.de ([194.221.49.2]) by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA16122; Mon, 29 Apr 1996 01:38:22 +0200 (MET) Received: from pris.intern.julia.de (pris.intern.julia.de [194.221.49.37]) by alles.intern.julia.de (8.7.5/8.7.3) with ESMTP id AAA03459 for ; Mon, 29 Apr 1996 00:36:04 +0200 From: Ralf Baechle Received: (from ralf@localhost) by pris.intern.julia.de (8.7.5/8.7.3) id BAA07897 for linux-mips@fnet.fr; Mon, 29 Apr 1996 01:37:52 +0200 (MET DST) Message-Id: <199604282337.BAA07897@pris.intern.julia.de> Subject: Re: Milo Arc/SGI, question about 1.3.91 and ftp.fnet.fr bandwidth To: linux-mips@fnet.fr Date: Mon, 29 Apr 1996 01:37:52 +0200 (MET DST) In-Reply-To: <199604282120.PAA11276@rover.village.org> from "Warner Losh" at Apr 28, 96 03:20:18 pm Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Length: 679 Lines: 16 Hi, > I agree. The other option is to use imake, which has enough > difficulties outside of the X world to make it less convenient to use > than gnu make. Given that gnu make has been ported to a lot of > places, and is the Linux standard make, I'd say that gnuisms are > fine. In my opinion imake is broken by design especially for an operating system having such an inhomogen consistence like Linux with it's dozens different library versions, kernel versions, directory structures and more. Autoconfig takes the right approach in that direction; it tests for features. Oh, and Stoned likes m4 which is being used by Autoconfig and of course Milo's manpages ... Ralf From ralf@Julia.DE Mon Apr 29 01:46:47 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id BAA13572 for ; Mon, 29 Apr 1996 01:46:46 +0200 (MET DST) Received: from alles.intern.julia.de ([194.221.49.2]) by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA16224; Mon, 29 Apr 1996 01:46:27 +0200 (MET) Received: from pris.intern.julia.de (pris.intern.julia.de [194.221.49.37]) by alles.intern.julia.de (8.7.5/8.7.3) with ESMTP id AAA03475 for ; Mon, 29 Apr 1996 00:44:10 +0200 From: Ralf Baechle Received: (from ralf@localhost) by pris.intern.julia.de (8.7.5/8.7.3) id BAA07908 for linux-mips@fnet.fr; Mon, 29 Apr 1996 01:45:59 +0200 (MET DST) Message-Id: <199604282345.BAA07908@pris.intern.julia.de> Subject: Re: The goal ... To: linux-mips@fnet.fr Date: Mon, 29 Apr 1996 01:45:59 +0200 (MET DST) In-Reply-To: <199604282123.PAA11302@rover.village.org> from "Warner Losh" at Apr 28, 96 03:23:43 pm Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Length: 1180 Lines: 28 Hi, > : Why should an inplace copy be dangerous? Ok, for Jazz machines the kernel > : needs several *microseconds* longer to startup. A long time if you're > : in hurry :-) > > In the past I've been bit by having inplace copy routines destroy what > was being copied. You're talking about overlapping copying but not inplace copies where source == dest? This should not happen in case of the exception handler since strcpy(9) copies from low to high addresses. > : Adjust the videoparams and you have a debugging console. Usage should > : be clear after a short look into the code. With print_string() and > : sprintf() in combination you can the same things as with printk(). > : tinycon.c is also easily useable for other machines with a PC style > : character based display. > > I'll have to whip that into shape, since I need to get the values of > things like the end of memory, etc. I'd rather not do the gdb > debugging yet, since I think that would be harder... Grin, I've *once* tried to use the GDB support in the Kernel introduced by Andy. When ld had wasted more than 5 minutes forcing my harddisk to create yet unheard sounds I stopped that ... Ralf From ralf@Julia.DE Mon Apr 29 01:56:45 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id BAA13766 for ; Mon, 29 Apr 1996 01:56:44 +0200 (MET DST) Received: from alles.intern.julia.de ([194.221.49.2]) by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA16294; Mon, 29 Apr 1996 01:56:25 +0200 (MET) Received: from pris.intern.julia.de (pris.intern.julia.de [194.221.49.37]) by alles.intern.julia.de (8.7.5/8.7.3) with ESMTP id AAA03486 for ; Mon, 29 Apr 1996 00:54:06 +0200 From: Ralf Baechle Received: (from ralf@localhost) by pris.intern.julia.de (8.7.5/8.7.3) id BAA07918 for linux-mips@fnet.fr; Mon, 29 Apr 1996 01:55:54 +0200 (MET DST) Message-Id: <199604282355.BAA07918@pris.intern.julia.de> Subject: Re: Milo Arc/SGI, question about 1.3.91 and ftp.fnet.fr bandwidth To: linux-mips@fnet.fr Date: Mon, 29 Apr 1996 01:55:54 +0200 (MET DST) In-Reply-To: <199604282127.PAA11335@rover.village.org> from "Warner Losh" at Apr 28, 96 03:27:20 pm Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Length: 1176 Lines: 30 Hi, > Agreed. One of the things I'd like to see is a MASH or something like > that for my machine. It would make the bootstrapping procedure > easier. Also, I'd like to have some way to partition and newfs disk > areas from the ARC bios (even if it only works from "MOST" of the > implementations). I think it should work on *all* implementations even though we have to use an undocumented feature. I suppose the ARC installer for WNT uses this same interface that we'll be using for an ARC based partitioning program. I believe ramdisk based Installation disk are preferable for this purpose, though. I now have a disk with 32M for kernels (DOS > partition) and 280M for hacking (soon to be ext2fs). Ralf, will that > be enough? Also, are disk roots supported yet. setup.c leads me to > believe not. Aeehhh... Disk roots do work. What is broken is the passing of the command line from Milo to the kernel. That's why the boot device is hardcoded in setup.c. For now just hack it to fir your needs. Stoned has already fixed this bug; the fix will be included in release 0.27 of Milo. Stoned, any idea when 0.27 will be available? Ralf From imp@rover.village.org Mon Apr 29 07:57:46 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id HAA22460 for ; Mon, 29 Apr 1996 07:57:44 +0200 (MET DST) Received: from rover.village.org by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA22559; Mon, 29 Apr 1996 07:57:24 +0200 (MET) Received: from localhost (localhost [127.0.0.1]) by rover.village.org (8.7.5/8.6.6) with SMTP id XAA13045 for ; Sun, 28 Apr 1996 23:57:30 -0600 (MDT) Message-Id: <199604290557.XAA13045@rover.village.org> To: linux-mips@fnet.fr Subject: FYI Date: Sun, 28 Apr 1996 23:57:29 -0600 From: Warner Losh Content-Length: 2098 Lines: 52 OK. I've found the problem. Seems like some fool[*] put the following code into setup.c: if (mips_machtype == MACH_DESKSTATION_RPC44) memory_start = (unsigned long) KSEG0 + (1 << 20); else memory_start = (unsigned long) &_end; without a sanity check to make sure that &_end was < KSEG0 + 1M. *BLUSH* Fixing this, I'm able to make some progress. I now get some strange results, likely due to memory being freed or something. I don't have the first part of this, but the end goes like: keyboard error [ repeat 24 times ] keyboard error keyboard: unrecognized scan code (68) - ignored While configuring BusLogic Host Adapater at I/O Address 0x330: HOST ADAPTER INTERRUPT TEST FAILED - DETACHING Interrupts are not getting through from the Host Adapter to the BusLogic Driver Interrupt Handler. The most likely cause is that either the Host Adapter or Motherboard is configured incorrectly. Please check the Host Adapter configuration with AutoSCSI or by examining any dip switch and jumper settings on the Host Adapter, and verify that no other device is attempting to use the same IRQ channel. For PCI Host Adapters, it may also be necessary to investigate an dmanually set the PCI interrupt assignments and edge/level interrpt type selection in the BIOS Setup PRogram or with Motherboard jumpers. scsi : 0 hosts. scsi : detected total. VFS: Insert root floppy and press ENTER ------------------ I'm not getting keyboard interrupts, so that's what I'll chase next. The numlock light test is failing. And I might have a config problem with the BusLogic board. Hey Ralf, what was that routine that you said you likely broke on me? I got carried away with the D key in my mailer. However, it is trying to mount a root file system again, despite the problems it is having. This is good news. It is also getting close to time to generate patches. I've been busy :-). This time I've also been prostrating myself before the church of the daily backup, so I can go longer w/o sending diffs :-) Warner [*] Fool is known to be one Mr. "xxx imp fixme" per the comment in the code :-) From Stoned.Elipot@univ-evry.fr Mon Apr 29 11:56:35 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id LAA00660 for ; Mon, 29 Apr 1996 11:56:34 +0200 (MET DST) Received: from amiante.univ-evry.fr by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA27366; Mon, 29 Apr 1996 10:32:20 +0200 (MET) Received: from bandsept.univ-evry.fr (bandsept.univ-evry.fr [192.93.214.19]) by amiante.univ-evry.fr with ESMTP (8.6.12/96.04.09/louis-blanc); id KAA22345; Mon, 29 Apr 1996 10:32:27 +0200 Received: from localhost.univ-evry.fr (localhost.univ-evry.fr [127.0.0.1]) by bandsept.univ-evry.fr with SMTP (8.6.12/94.01.26); id KAA17917; Mon, 29 Apr 1996 10:31:48 +0200 Message-Id: <199604290831.KAA17917@bandsept.univ-evry.fr> To: linux-mips@fnet.fr Subject: Re: Milo Arc/SGI, question about 1.3.91 and ftp.fnet.fr bandwidth In-Reply-To: Your message of "Mon, 29 Apr 1996 01:55:54 +0200." <199604282355.BAA07918@pris.intern.julia.de> X-It: bandsept (wrong OS, but right CPU) X-Mailer: MH 6.8.3, MH-E 5.0.2 Date: Mon, 29 Apr 1996 10:31:41 +0200 From: Stoned Elipot Content-Length: 329 Lines: 10 Hi all, >>>>> "Ralf" == Ralf Baechle writes: [SNIP] Ralf> Stoned, any idea when 0.27 will be available? I'll deadly try to make it available before the end of the week. I had a little kernel patch which will be required as I had ripped off the last memcopy() stuff in Milo... Tags more tags :) Cheers, Stoned. From imp@rover.village.org Mon Apr 29 16:56:55 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id QAA06073 for ; Mon, 29 Apr 1996 16:56:54 +0200 (MET DST) Received: from rover.village.org by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA04897; Mon, 29 Apr 1996 16:42:32 +0200 (MET) Received: from localhost (localhost [127.0.0.1]) by rover.village.org (8.7.5/8.6.6) with SMTP id IAA14203 for ; Mon, 29 Apr 1996 08:42:43 -0600 (MDT) Message-Id: <199604291442.IAA14203@rover.village.org> To: linux-mips@fnet.fr Subject: Re: The goal ... In-Reply-To: Your message of Mon, 29 Apr 1996 01:45:59 +0200 Date: Mon, 29 Apr 1996 08:42:42 -0600 From: Warner Losh Content-Length: 1041 Lines: 23 : > In the past I've been bit by having inplace copy routines destroy what : > was being copied. : : You're talking about overlapping copying but not inplace copies where : source == dest? This should not happen in case of the exception handler : since strcpy(9) copies from low to high addresses. I know that overlapping copies are undefined, even in place ones, by the C standard. It is possible that our current code will work, and most future code will work. I've been bit in the past by a PC compiler that had a bug that bit people on inplace copies, so I'm a little gun shy when I see code like that... : Grin, I've *once* tried to use the GDB support in the Kernel introduced : by Andy. When ld had wasted more than 5 minutes forcing my harddisk : to create yet unheard sounds I stopped that ... I think you can just build one or two modules -g, and not the whole thing. I know in userland gdb can cope with this case. I'll keep that in mind because I'm limited to 1.44M in kernel size at the moment, wonder why :-) Warner From imp@rover.village.org Mon Apr 29 16:45:26 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id QAA05957 for ; Mon, 29 Apr 1996 16:45:24 +0200 (MET DST) Received: from rover.village.org by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA04944; Mon, 29 Apr 1996 16:45:04 +0200 (MET) Received: from localhost (localhost [127.0.0.1]) by rover.village.org (8.7.5/8.6.6) with SMTP id IAA14218 for ; Mon, 29 Apr 1996 08:45:16 -0600 (MDT) Message-Id: <199604291445.IAA14218@rover.village.org> To: linux-mips@fnet.fr Subject: Re: Milo Arc/SGI, question about 1.3.91 and ftp.fnet.fr bandwidth In-Reply-To: Your message of Mon, 29 Apr 1996 01:55:54 +0200 Date: Mon, 29 Apr 1996 08:45:16 -0600 From: Warner Losh Content-Length: 797 Lines: 17 : I think it should work on *all* implementations even though we have to : use an undocumented feature. I suppose the ARC installer for WNT uses : this same interface that we'll be using for an ARC based partitioning : program. : : I believe ramdisk based Installation disk are preferable for this purpose, : though. Hmmm. You may be right. We have more control in a Ramdisk based env. However, the number of kernels is going to be confusing to people, so this might be nighmarish. Can we find one place for MILO to load, and then have a special install version of MILO that knows to load vmlinux.rpc for my machine and vmlinux.pic for Ralf's machine, and vmlinux.jaz for others? Just a thought. It can also warn if the load address doesn't match a load address in its table too. Warner From linux@mailhost.uni-koblenz.de Mon Apr 29 20:37:18 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id UAA08783 for ; Mon, 29 Apr 1996 20:37:18 +0200 (MET DST) Received: from informatik.uni-koblenz.de (mailhost.uni-koblenz.de) by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA09704; Mon, 29 Apr 1996 20:36:50 +0200 (MET) Received: from ozzy (ozzy.uni-koblenz.de [141.26.5.8]) by informatik.uni-koblenz.de (8.7.4/8.6.9) with SMTP id UAA06601 for ; Mon, 29 Apr 1996 20:36:55 +0200 (MET DST) From: Systemkennung Linux Message-Id: <199604291836.UAA06601@informatik.uni-koblenz.de> Received: by ozzy (SMI-8.6/KO-2.0) id UAA05049; Mon, 29 Apr 1996 20:35:08 +0200 Subject: Announce To: linux-mips@fnet.fr Date: Mon, 29 Apr 1996 20:35:07 +0200 (MET DST) Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Length: 2099 Lines: 56 Hi all, once again some patches & binaries. Noteworthy MIPS changes/comments: - Cleaned up signal handlers a bit - Signal handlers now save/restore the fp context also. - mmap(2) is broken upto and including 1.3.96. Stupid bug. The tar binaries that I'm also uploading assume a working mmap. I'm appending the patch below. - There is a strange problem with IP checksums that slows down NFS traffic a lot and causes NFS server timeouts. I tend to believe it's a GCC bug related to inline functions. - New GCC patches. Improvements over gcc-2.7.2-1.diffs.gz are: - The buggy strength-reduce has been disabled. - GCC now builds for big endian ELF targets. - Some other minor changes. - Little endian binaries of cpio 2.5.2 and GNU tar 1.11.8. As mentioned above the tar binaries depend on a working mmap(). The binaries have been recompiled and fix the "lockup on mknod" bug. Due to the horribly slow connection to FNET the uploads may still take several hours. Enjoy just two of the FTP messages ;-) > 74631 bytes sent in 1.55e+03 secs (0.047 Kbytes/sec) > 426 Data Connection: Connection reset by peer. Happy Mipsomania, Ralf 59cc9f4f381142634b8618b6cd0318e6 cpio-2.4.2.tar.gz b0447f4894a419a2df164f1ae6ca3032 gcc-2.7.2-2.diffs.gz 9ae7a5cc169ccc9208ff4ac94dadf45b linux-1.3.90-1.3.91.diffs.gz d844ea80ba8b31b9ebdf7098a3be6948 linux-1.3.91-1.3.92.diffs.gz e242f27019e8b4641f931b4aaefc293f linux-1.3.92-1.3.93.diffs.gz ebdb34d8ad3a71a20d693c6efcee1a03 linux-1.3.93-1.3.94.diffs.gz 7bc26cb86a15e578100408fca60cbb51 linux-1.3.94-1.3.95.diffs.gz 9473c33d57fe007eafd47cd95e644a1c linux-1.3.95-1.3.96.diffs.gz f1071fd38f400ec649e0f79aa1f11f89 linux-1.3.96.tar.gz f6758e8189b3b35b082ed2cec8874326 tar-1.11.8.tar.gz --- arch/mips/kernel/syscall.c.orig Mon Apr 29 18:10:03 1996 +++ arch/mips/kernel/syscall.c Mon Apr 29 18:10:19 1996 @@ -42,7 +42,7 @@ { struct file * file = NULL; - if (flags & MAP_RENAME) { + if (!(flags & MAP_ANONYMOUS)) { if (fd >= NR_OPEN || !(file = current->files->fd[fd])) return -EBADF; } From tor@spacetec.no Tue Apr 30 12:53:01 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id MAA21155 for ; Tue, 30 Apr 1996 12:52:59 +0200 (MET DST) Received: from pallas.spacetec.no by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA24535; Tue, 30 Apr 1996 12:52:38 +0200 (MET) Received: (from tor@localhost) by pallas.spacetec.no (8.7.1/8.7.1) id MAA24747 for linux-mips@fnet.fr; Tue, 30 Apr 1996 12:58:03 +0200 Message-Id: <199604301058.MAA24747@pallas.spacetec.no> From: tor@spacetec.no (Tor Arntsen) Date: Tue, 30 Apr 1996 12:58:02 +0200 X-Mailer: Mail User's Shell (7.2.5 10/14/92) To: linux-mips@fnet.fr Subject: Announcing SGI crossdev tools Content-Length: 1070 Lines: 25 Hi all, Cross development tools precompiled for SGI are now available on ftp.fnet.fr in the linux-mips/crossdev/mips-sgi-irix5/ directory. Big endian and little endian versions of binutils-2.6 and gcc 2.7.2 are provided, in both AOUT and ELF versions. The documentation (man/info pages) is provided as two separate tar files, one for gcc another one for binutils. If you already have e.g. a native gcc and/or binutils installed somewhere you probably don't need this. I hope there are no gotchas left (there were a couple to take care of even *after* a successful make/make install), however I could do only minimal testing of the big endian versions (only built libgcc.a with them). The big-endian versions are the mips- files, the little-endian are the mipsel- files, the aout versions are the -linux files, the elf versions are the -linuxelf files. I believe this is the same convention as used for e.g. the i486-linux tools. README files are provided. If someone tries the tools I would appreciate feedback on how well (or not) they work.. Cheers, Tor From nicholas@computone.com Tue Apr 30 21:38:25 1996 Received: from ns.fnet.fr (ns.fnet.fr [192.134.192.2]) by guadalquivir.fnet.fr (8.7.4/8.7.3) with SMTP id VAA03988 for ; Tue, 30 Apr 1996 21:38:24 +0200 (MET DST) Received: from cton.computone.com by ns.fnet.fr (5.65c8d/AFUU-4.2.3) via EUnet-France id AA17649; Tue, 30 Apr 1996 21:38:03 +0200 (MET) Received: from mom.computone.com by cton.computone.com with smtp (Smail3.1.28.1 #2) id m0uEKNo-0006p6C; Tue, 30 Apr 96 14:42 EDT Received: by mom.computone.com (Smail3.1.28.1 #12) id m0uELTf-001hyqC; Tue, 30 Apr 96 15:52 EDT Message-Id: From: nicholas@computone.com (Nick Hennenfent) Subject: MIPS R3000 Workstations To: linux-mips@fnet.fr Date: Tue, 30 Apr 1996 15:52:50 -0400 (EDT) X-Mailer: ELM [version 2.4 PL22] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 224 Lines: 9 Anyone know where I can get a MIPS R3000 workstation running 4.52 MIPS/OS??? Nicholas Hennenfent - Computone Corporation 1100 Northmeadow Parkway, Suite 150, Roswell, GA 30076 voice: 770.475.2725 x362 fax: 770.343.9735