Home > Archive > Unix Programming > July 2004 > design of the unix operating system - maurice bach
You are viewing an archived Text-only version of the thread.
To view this thread in it's original format and/or if you want to reply to
this thread please [click here]
| Author |
design of the unix operating system - maurice bach
|
|
| sean larsson 2004-07-19, 8:57 pm |
| what (free)unix most closely resembles what is described in this book? i realize it was written more than a bit ago, but i'd imagine alot of the key concepts and algorithms are similar.
--
-sean
| |
| Rich Teer 2004-07-19, 8:57 pm |
| On Mon, 19 Jul 2004, sean larsson wrote:
> what (free)unix most closely resembles what is described in this
> book? i realize it was written more than a bit ago, but i'd imagine
> alot of the key concepts and algorithms are similar.
Bach's book describes SVR2 (with some SVR3 bits), so the closest to that
would be any SVR4-derived UNIX. Solaris is "free" for home users,
so I guess that would qualify it for you. But if you're interested in
contemporary UNIX kernel internals, you should read any or all of
The Magic Garden, Solaris Internals, and UNIX Internals, and perhaps
The Design and Implementation of the BSD4.4 OS.
HTH,
--
Rich Teer, SCNA, SCSA
President,
Rite Online Inc.
Voice: +1 (250) 979-1638
URL: http://www.rite-online.net
| |
| Andrew Gabriel 2004-07-20, 8:58 am |
| In article <Pine.SOL.4.58.0407191608420.638@zaphod.rite-group.com>,
Rich Teer <rich.teer@rite-group.com> writes:
> On Mon, 19 Jul 2004, sean larsson wrote:
>
>
> Bach's book describes SVR2 (with some SVR3 bits), so the closest to that
> would be any SVR4-derived UNIX.
That's really quite different from SVR3, which has no vnodes, no
paging, etc. The nearest surviving unix from that era would be
SCO Openserver 5, although even that is quite hacked away from its
SVR3.2 roots. If you can find an old copy of SCO SVR3.2 (also went
under the name of OpenServer 3 if it included optional X Windows)
or Interactive UNIX, these are more of the correct vintage. For
these older OS's, you're going to have to find a 15 year old 486
to run them on too;-)
> Solaris is "free" for home users,
> so I guess that would qualify it for you. But if you're interested in
> contemporary UNIX kernel internals, you should read any or all of
> The Magic Garden, Solaris Internals, and UNIX Internals, and perhaps
> The Design and Implementation of the BSD4.4 OS.
--
Andrew Gabriel
Consultant Software Engineer
| |
| sean larsson 2004-07-20, 3:59 pm |
| ok thanks for the suggestions. i bought this book b/c everyone said it is 'the classic' book on unix, but perhaps i should have went with one of the newer ones instead. ah well, so far it is a good book.
--
-sean
| |
| Dragan Cvetkovic 2004-07-20, 3:59 pm |
| sean larsson <infamous42md@ERASEMEhotpop.com> writes:
> ok thanks for the suggestions. i bought this book b/c everyone said it
> is 'the classic' book on unix, but perhaps i should have went with one of
> the newer ones instead. ah well, so far it is a good book.
Well, it _is_ a good book, and it is _the classic_ book on unix, so people
were not lying to you. I would assume that some concepts would be much
easier to understand reading Bach's book than some newer ones that describe
more complex UNIX systems...
Dragan
--
Dragan Cvetkovic,
To be or not to be is true. G. Boole No it isn't. L. E. J. Brouwer
!!! Sender/From address is bogus. Use reply-to one !!!
| |
| sean larsson 2004-07-21, 3:57 am |
| On Tue, 20 Jul 2004 13:52:28 -0400
Dragan Cvetkovic <me@privacy.net> wrote:
> sean larsson <infamous42md@ERASEMEhotpop.com> writes:
>
>
> Well, it _is_ a good book, and it is _the classic_ book on unix, so people
> were not lying to you. I would assume that some concepts would be much
> easier to understand reading Bach's book than some newer ones that describe
> more complex UNIX systems...
>
> Dragan
>
> --
> Dragan Cvetkovic,
>
> To be or not to be is true. G. Boole No it isn't. L. E. J. Brouwer
>
> !!! Sender/From address is bogus. Use reply-to one !!!
well, regarding the buffer cache for example. i knew what it was from reading 'linux device drivers' and the floppy driver on linux, but i hadn't fully understood some of the source i was reading that dealt with block and file system layers. then from r
eading the chapter in the book about buffer cache and reading the source in linux it was much clearer what was going with it. however i notice problems everywhere with locking. well it's not a problem on those systems, but would be on SMP. i dont reall
y understand why someone decided that kernels shouldn't be pre-emptive, or didn't plan on running an OS on multi processor systems.
--
-sean
| |
| Barry Margolin 2004-07-21, 3:58 pm |
| In article <20040720234453.194101ee.infamous42md@ERASEMEhotpop.com>,
sean larsson <infamous42md@ERASEMEhotpop.com> wrote:
> i dont really understand why someone decided that kernels shouldn't be
> pre-emptive, or didn't plan on running an OS on multi processor systems.
It's harder to design such an OS, and hard to implement it correctly.
--
Barry Margolin, barmar@alum.mit.edu
Arlington, MA
*** PLEASE post questions in newsgroups, not directly to me ***
| |
| Lew Pitcher 2004-07-21, 3:58 pm |
| -----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Barry Margolin wrote:
> In article <20040720234453.194101ee.infamous42md@ERASEMEhotpop.com>,
> sean larsson <infamous42md@ERASEMEhotpop.com> wrote:
>
>
>
>
> It's harder to design such an OS, and hard to implement it correctly.
And, when Unix was originally designed, there were no multi-processor
systems, and a pre-emptive kernel wasn't considered a necessity.
IIRC, there wasn't even hardware memory protection in the processors
early Unix was developed for.
- --
Lew Pitcher, IT Consultant, Enterprise Application Architecture
Enterprise Technology Solutions, TD Bank Financial Group
(Opinions expressed here are my own, not my employer's)
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (MingW32)
iD8DBQFA/ nj6agVFX4UWr64RAowuAKDj6TnCz0eocm3KaHWzp
0aq4ZL7/wCeM5G4
6QFDOibTJF8YlW4BV+TUwiE=
=itkK
-----END PGP SIGNATURE-----
| |
| Barry Margolin 2004-07-21, 3:58 pm |
| In article <1OuLc.20338$Gf7.808574@news20.bellglobal.com>,
Lew Pitcher <Lew.Pitcher@td.com> wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> Barry Margolin wrote:
>
>
> And, when Unix was originally designed, there were no multi-processor
> systems, and a pre-emptive kernel wasn't considered a necessity.
Sure there were. The designers came out of the Multics project, which
was a multi-processor system.
I think they knowingly ignored it, to keep their design simple.
> IIRC, there wasn't even hardware memory protection in the processors
> early Unix was developed for.
I believe they had memory protection, but not demand-paged virtual
memory.
--
Barry Margolin, barmar@alum.mit.edu
Arlington, MA
*** PLEASE post questions in newsgroups, not directly to me ***
| |
| Lew Pitcher 2004-07-21, 3:58 pm |
| -----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Lew Pitcher wrote:
[snip]
> IIRC, there wasn't even hardware memory protection in the processors
> early Unix was developed for.
See http://cm.bell-labs.com/cm/cs/who/dmr/odd.html
- --
Lew Pitcher, IT Consultant, Enterprise Application Architecture
Enterprise Technology Solutions, TD Bank Financial Group
(Opinions expressed here are my own, not my employer's)
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (MingW32)
iD8DBQFA/ oRHagVFX4UWr64RAsvbAJ9EXji5pLHg3IYsm1dq3
QZDGzVe/ACgwTlf
d8LHBRbfHCn5vVdDil3Z7Mg=
=KYfd
-----END PGP SIGNATURE-----
| |
| Lew Pitcher 2004-07-22, 3:58 pm |
| -----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Barry Margolin wrote:
> In article <1OuLc.20338$Gf7.808574@news20.bellglobal.com>,
> Lew Pitcher <Lew.Pitcher@td.com> wrote:
[snip]
>
>
> I believe they had memory protection, but not demand-paged virtual
> memory.
>
- From the http://cm.bell-labs.com/cm/cs/who/dmr/odd.html link...
A hardware story
Back around 1970-71, Unix on the PDP-11/20 ran on hardware that not only did
not support virtual memory, but didn't support any kind of hardware memory
>. . . . . . . . . . . . . . . . ========================================
==
mapping or protection, for example against writing over the kernel. This was a
>=====================
pain, because we were using the machine for multiple users. When anyone was
working on a program, it was considered a courtesy to yell "A.OUT?" before
trying it, to warn others to save whatever they were editing.
- --
Lew Pitcher
IT Consultant, Enterprise Application Architecture,
Enterprise Technology Solutions, TD Bank Financial Group
(Opinions expressed are my own, not my employers')
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (MingW32)
iD8DBQFA/ 7oVagVFX4UWr64RAn2qAJsFkc4+EH3TaiZ9uzs4X
SlAabnSkwCgk+Sc
xcPcXYCHAZMWdcz7eZMsJXI=
=3JYC
-----END PGP SIGNATURE-----
| |
| junky_fellow 2004-07-23, 8:57 am |
| Rich Teer <rich.teer@rite-group.com> wrote in message news:<Pine.SOL.4.58.0407191608420.638@zaphod.rite-group.com>...
> On Mon, 19 Jul 2004, sean larsson wrote:
>
>
> Bach's book describes SVR2 (with some SVR3 bits), so the closest to that
> would be any SVR4-derived UNIX. Solaris is "free" for home users,
> so I guess that would qualify it for you. But if you're interested in
> contemporary UNIX kernel internals, you should read any or all of
> The Magic Garden, Solaris Internals, and UNIX Internals, and perhaps
> The Design and Implementation of the BSD4.4 OS.
>
> HTH,
There's one more good book by "Uresh Vahalia" on unix internals
( Pearson Education Inc)
| |
| Rich Teer 2004-07-23, 3:57 pm |
| On Fri, 23 Jul 2004, junky_fellow wrote:
>
> There's one more good book by "Uresh Vahalia" on unix internals
> ( Pearson Education Inc)
That's the third one on my list!
--
Rich Teer, SCNA, SCSA
President,
Rite Online Inc.
Voice: +1 (250) 979-1638
URL: http://www.rite-online.net
|
|
|
|
|