| From | Subject | Date |
|---|---|---|
| Robert Blum | Diversification of development
Hello Linuxers!
Sorry for my last mail being very short, but writing mail at 300 baud *IS*
a pain. Now, there are a couple of things to be adressed.
1) Linux is undergoing a process of being spread across several countrys/
FTP-servers. I recently read in a posting: If you want the blah binaries,
just connect to the foo-server, etc.
To keep a common environment for Linux, especially now where it is still
say a bit more than beta, I'd suggest using a common set of tools. It
is a p...
| Nov 11, 6:56 pm 1991 |
| Theodore Ts'o | Weekend hacking and other randomness...
Over the weekend, I've managed to put together some stuff which other
people might find useful. First of all, I've managed to build GNU diff,
GNU sed, Larry Wall's patch and tr. They pretty much built without
needing to make any changes --- the only real problem which I
encountered while trying to build them is that the configuration progam
which you need to run to build sed itself wants to uses sed to build the
Makefile. Oops. :-)
In any cases, if anyone wants to just snarf the binaries, they ...
| Nov 11, 5:08 pm 1991 |
| Lars Wirzenius | Header structure for mixed Classic/ANSI
Wouldn't it be easier to make two header trees:
/usr/include/ansi
/usr/include/classic
and either use the -I option to select which tree is used, or put
files like this:
#if __STDC__ == 1 /* not: not just ifdef; see comp.lang.c FAQ */
# include <ansi/foo.h>
#else
# include <classic/foo.h>
#endif
into /usr/include.
This would IMHO be more neat than your proposed solution.
| Nov 11, 3:45 pm 1991 |
| Theodore Ts'o | Re: rename, file system errors
Date: Mon, 11 Nov 1991 15:53:55 +0200
From: Linus Benedict Torvalds <torvalds@cc.helsinki.fi>
gets symbolic links. I think the easiest way is to move downwards from
the "to"-directory via ".." until one hits root (or a mount-point) or
the "from"-directory. It shouldn't be hard, it's just SMOP.
Well, the problem is that another rename() could be happening while you are
traversing up the tree (towards the root) checking to see if you hit the
"from" directory. The simplest cas...
| Nov 11, 3:12 pm 1991 |
| Peter MacDonald | shoelace patch#3: put environment ptr where linux can find it.
I know, I know. Three patches in 24 hours... I plead end of term
instability. The following one line patch puts the environment
offset relative to the shoelace corpse which is located at segment 0x80000.
Thus, an offset of 0x1234 would put the address of the env at 0x81234.
This offset is stored at location 0x90506: ie. in the word before the
root device word (508) in the bootsect.s corpse.
By including the shoelace header file shoeconf.h, the linux kernel
could now read in any of the environm...
| Nov 11, 2:58 pm 1991 |
| Robert Blum | C-Compilers on a 2Meg-Machine
Hi Linuxers!
I recently ported CvW's C386 to Linux. I can do development on a 2Meg-System
too. (BTW Linus, that caused the gcc-errors..)
But C386 is non-ANSI. Now are we
a) Giving it an ANSI front end ? (any gurus out there?)
b) put an #ifdef _STDC_ in all includes
c) Forget about the 2Meg-machines (NO, PLEASE NOT:-()
What do you thinnk about it ?
Robert Blum
| Nov 11, 1:54 pm 1991 |
| Linus Benedict Torvalds | Re: C-Compilers on a 2Meg-Machine
a) Not very easy (mild understatement).
b) I *HATE* unclear include-files
c) No way.
I'd like to propose a following change to /usr/include/*:
Add a directory "/usr/include/non-ansi", and to each file (example
ctype.h, cdiff type adding):
#ifndef _CTYPE_H
#define _CTYPE_H
+ #if !defined(__STDC__) && !defined(__GCC__)
+ #include <non-ansi/ctype.h>
+ #else
... old ctype.h unchanged (except for the bug-fix, see below)
+ #endif /* __STDC__ */
#endif /* _CTYPE_H */
...
| Nov 11, 2:49 pm 1991 |
| Linus Benedict Torvalds | rename, file system errors
Yeah, it'll be implemented. Hopefully not even by me :-), as
darken-christian showed some interest. I'll do it if nobody else does,
but it will take some time. Currently you can always use "tar" and
"rm -rf" to move a directory (but look out for a full filesystem).
There are quite a few problems with rename, most of which have been
mentioned here (and that's why I left rename unimplemented :-). I
wouldn't say checking the filename is a good idea: it breaks when linux
gets symbolic links. I ...
| Nov 11, 9:53 am 1991 |
| Patrick L. McGillan | more on floppies
Hi all,
I also note that a known dos disk with files on it gives back an error
with the usual stuff and a last line of mdir: Cannot initialize 'A:'
--
Patrick L. McGillan
Computer Systems Specialist
University Of Wisconsin Ph: (715) 394-8191
Superior, Wisconsin pmcgilla@uwsuper.edu
| Nov 11, 9:52 am 1991 |
| Patrick L. McGillan | floppies
Hi all,
Yes, Linus, I did format the disks under dos. In fact I made a bootable
and could not corrupt it with any kind of write action. Also tried
360k and 1.44 meg disks.
--
Patrick L. McGillan
Computer Systems Specialist
University Of Wisconsin Ph: (715) 394-8191
Superior, Wisconsin pmcgilla@uwsuper.edu
| Nov 11, 9:37 am 1991 |
| Kurt Wachmann | About mvdir and linking directories
In reply to tytso@ATHENA.MIT.EDU:
You can let the system call rename() do the trick for you, and
forbid linking of directories - correct. If you forbid linking
of directories, you see if one directoy is an ancestor of the
other *by looking at the names*. That will take away your
race condition for free.
We do however still have to avoid concurrent directory moves
from the same source directory - and into the old tree.
open()'s etc. are ok (I think), they must be atomic, and will
fail or suc...
| Nov 11, 6:35 am 1991 |
| Kurt Wachmann | About mvdir and linking directories
In reply to your discussion about keeping track of parent
directories in the kernel; the answer is: DON'T.
The filesystem should work like this: Directories may be lin--
ked and unlinked by the superuser only. The filesystem code in
the kernel should not check the correctness of these opera--
tions, except for the root priviledge. The superuser is suppo--
sed to know what (s)he is doing.
The file "." is a link to the directory in which it resides,
and ".." is a link to the parent ...
| Nov 11, 5:04 am 1991 |
| Marc CORSINI | help installing linux on hd
This is it, there is no hdX with X in 5..9, there is at 0 and 1
ps 0 and 1, tty, and some pat devices ??
aarrrggghhhh :-(
it seems that i have to keep on "playing" with linux OR <any idea
wellcomed except the "do a partition on ur big hd1">
[mmc]
La vie est un long fleuve tranquille ou il ne fait pas bon ramer
| Nov 11, 4:07 am 1991 |
| Linus Benedict Torvalds | help installing linux on hd
Ok, here's one possibility: mount the /linux/ root-floppy from the
minix-demo, and use the devices on that.
mount /dev/fd0 /dir
mkfs /dir/dev/hd6 partition-size-in-K
^___ 6 or 7 (minix numbering)
This should work. (I hope). You should still check which partition is
which (unless you don't care).
Linus
| Nov 11, 4:24 am 1991 |
| Marc CORSINI | help installing linux on hd
Hi,
In fact the subject says it all
I've been able to get the minix demo disk from plains
also the different Image for linux
I was able to rawrite them on floppy quite easily. In fact the problems arised
when i wanted to put linux on my hd:
I used a 386Sx with 2Mo Ram, and 2 HD
the first one is a 120Meg, and the second one a 40Meg
The big one is a single partionned disk under DRDOS 5.0
The other is a 3 partitionned one
Using Minix I've tried mkfs /dev/hdX with Xgreater than 5
the only reponse i...
| Nov 11, 1:23 am 1991 |
| Peter MacDonald | shoelace patch #2
There is a big difference between floppy the device numbers on linux and
on minix. In particular linux has fd0 as 0x208 while minix uses 0x200.
ie. linux /dev/fd0 == minix /dev/fd0-dshd5in
The following patch will keep linux from using 0x200, which it doesn't
know about and instead use 0x208 and 0x21c.
Apply this after all other patches.
pmacdona@sol.uvic.ca
*** shoe.c.bad Thu Nov 7 08:43:30 1991
--- shoe.c Thu Nov 7 08:52:27 1991
***************
*** 638,646 ****
longjmp(...
| Nov 10, 8:57 pm 1991 |
| Peter MacDonald | patch to use shoelace with linux + fsck and mkfs
Following are patches to shoelace to allow it to boot the Linux
image (sorry, no building by parts cause there are no parts).
This also means that the fsck function of shoelace is available,
including fsck -m, or mkfs, which I have not tested. Please note
that by using strip on the shoelace executable, I was able to fsck
a 20Meg partition. 64Meg partitions, I doubt it can handle.
In this way, both minix and linux can be run from the same partition,
with a bit of mv'ing of directories like bin.
...
| Nov 10, 8:21 pm 1991 |
| Theodore Ts'o | Re: More answers
Date: Sun, 10 Nov 1991 18:09:31 +0200
From: Linus Benedict Torvalds <torvalds@cc.helsinki.fi>
darken.@cs.yale.edu:
> 1) mv doesn't seem to move directories properly. It looks for
> something called mvdir and dies. I want to fix this right away,
> so I'm goin to start working on it unless one of you drops me a
> line and explains what stupid thing I've done.
This is a possibly deficiency in the linux kernel. I don't allow
links/unlinks to directori...
| Nov 11, 5:21 am 1991 |
| previous day | today | next day |
|---|---|---|
| November 10, 1991 | November 11, 1991 | November 12, 1991 |
| Sunil Naidu | Re: Linux 2.6.20-rc6 |
| david | Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3 |
| Jan Engelhardt | intel iommu (Re: -mm merge plans for 2.6.23) |
| Linus Torvalds | Re: init's children list is long and slows reaping children. |
| David Miller | Re: [GIT]: Networking |
| Jarek Poplawski | [PATCH] pkt_sched: Destroy gen estimators under rtnl_lock(). |
| Gerrit Renker | [PATCH 27/37] dccp: Integration of dynamic feature activation - part 2 (server side) |
| Andrew Morton | Re: [BUG] New Kernel Bugs |
git: | |
