I, too, I have managed to bring up Linux without the need of Minux. I have a 40 MhZ AMD 386 with a 200 Meg clone. A couple of notes: N1) The Minix demo disk which you can get from plains.novak.edu only works for 5.25" disks --- this isn't documented anywhere, but I spent a lot of time trying to get it to boot on my 3.5" A drive. It would print the first message about "loading Minix system", but it would hang before printing the menu. I finally had to give up, recable my drives so that my 5" drive was my A drive to get Minix to come up. N2) The numbering convestion for partition in Minix and Linux are different!!! The way I dealt with it is to use a disk editor to write a message "This is the Linux parittion" in the first sector of the partition, and then use "head -3 /dev/hd[014]" to find the right partition number on both Minix and Linux. N3) I have MS-DOS 5.0, and the mtools worked just fine on my hard disk. All of my partitions are less than 32 MEG, so I'm still using the FAT-16 filesystem. It would be nice if Linux understood the DOS extended partitions, though. This would allow mtools could read my other DOS partitions. In addition, it would mean that I could use one of the DOS extended partitions as Linux partitions, so I could avoid chewing up the 4 primary partitions available on my 200 meg drive. [ Note: whoever decided that IBM hard disks only needed 4 partitions should be condemned to recabling machine room floors; CP/M on Heath/Zenith machines had 8 partitions available, and much more friendly tools to modify said table! ] N4) I have also experienced the panic which Patrick L. McGillan has described. I suspect DOS setting some cruft which isn't getting cleaned up. The panic happens happens right after the "Loading system" and before the system has a chance to print the "Partition tables ok" message. N5) I have managed to build the Linux kernel under Linux, using the 16 bit binaries which Linus provided. One little gotcha is that it is necessary to rename the binaries provided in gccbin.tar.Z to the names which the Makefile is expecting. Other than that, though, things went very smoothly. N6) It wasn't obvious that the "em" program in utilbin.tar.Z was actually MicroEmacs. I was really happy once I found it, though! Perhaps there should be a quick note mentioning this fact somewhere. At the moment, the programs which I seem to be missing the most are 1) more, 2) gdb, and 3) fsck. The first should be a lot easier to get working than the last. :-) And now, for some questions: Q1) On nic.funet.fi, I found sources to shoelace, which seems to be a way to boot Minix without needing a floppy boot disk. Is anyone working on something similar for Linux? The other interesting thing about shoelace is that it came with a file "shoefsck.c" which seems to contain the necessary code filesystem checking for Minix. However, there's no copyright notice on that file, and no email address for the author, either. Does anyone know what the status of that code is? I was considering trying to use that code to make a fsck for Minix. Q2) Similarily, there are sources on nic.funet.fi for a fsck program for Minix. It looks like it was derived from the Minix fsck, but since it is available for anonymous FTP with no notices saying "don't touch this", I wonder if would be consider fair play to base a fsck for Linux on the code, and either distribute the code or patches to the code with a note that you can get the rest of the program from nic.funet.fi. Q3) Linus, can you provide the configuration files for gcc and friends? It would be interesting to see what's necessary to actually compile one's own version of gcc/gas/etc. Thanks to everyone on the list! I don't think I would have managed to get Linux up and running with out a lot of helpful hints which were posted on the list --- and I've only been on the list for a couple of days! - Ted
| Tarkan Erimer | Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3 |
| Greg KH | [GIT PATCH] driver core patches against 2.6.24 |
| Justin Piszcz | exception Emask 0x0 SAct 0x1 / SErr 0x0 action 0x2 frozen |
| Heiko Carstens | Re: -mm merge plans for 2.6.23 -- sys_fallocate |
git: | |
| Gerrit Renker | [PATCH 27/37] dccp: Integration of dynamic feature activation - part 2 (server side) |
| David Miller | [GIT]: Networking |
| Frans Pop | svc: failed to register lockdv1 RPC service (errno 97). |
| Radu Rendec | htb parallelism on multi-core platforms |
