Any recommendations for Tom Farwell's 'AIX Internals' book?
El Toro
Internals, and Performance: Professional Reference Edition" by Tom Farwell
of AIXperts.com (ISBN 0-9714029-0-6).
However, the only detail I can find about the book is a pretty general
"Table of Contents" located at http://www.aixperts.com/images/Table%20of%
20Contents.htm. I can't tell from this whether it discusses Unix inter-
nals in general (such as how CPU and RAM interoperate, what is difference
between 32-bit and 64-bit, what is virtual memory, how TCP/IP works, etc.),
or if it details how these things work in AIX specifically (i.e., the proc
table in AIX looks like this, or SysV IPC resources are dynamically
allocated in AIX like this as opposed to other Unices, etc.).
I would much prefer the latter, as I already have enough books on general
Unix internals, the best of which is Uresh Vahalia's "Unix Internals --
The New Frontiers". I also have a good book that discusses Solaris in-
ternals specifically, and how it differs from others -- "Solaris Internals"
byt Richard McDougall and Jim Mauro. I'm hoping this book will get as
specific with AIX as the 'Solaris Internals' book got specific with SunOS
kernel internals.
So, since I can't get any other info from online bookstores or from the
aixperts.com website itself, I was hoping that someone on this list has
looked it over. Especially if you've taken any kernel internals courses
at IBM, I'd be interested in your take on this book's coverage of the
topic.
Thanks very much in advance for your timely responses.
ppluz...@earthlink.net
> I'm considering purchasing the book entitled "Understanding Hardware, AIX
> Internals, and Performance: Professional Reference Edition" by Tom Farwell
> of AIXperts.com (ISBN 0-9714029-0-6).
>
> However, the only detail I can find about the book is a pretty general
> "Table of Contents" located at http://www.aixperts.com/images/Table%20of%
> 20Contents.htm. I can't tell from this whether it discusses Unix inter-
> nals in general (such as how CPU and RAM interoperate, what is difference
> between 32-bit and 64-bit, what is virtual memory, how TCP/IP works, etc.),
> or if it details how these things work in AIX specifically (i.e., the proc
> table in AIX looks like this, or SysV IPC resources are dynamically
> allocated in AIX like this as opposed to other Unices, etc.).
>
> I would much prefer the latter, as I already have enough books on general
I was also looking for the latter when I got the 'AIX Internals' book.
Unfortunately, it proved to be quite useless ;-( at least to me).
Cheers,
--
In order to understand recursion you must first understand recursion.
El Toro
> Unfortunately, it proved to be quite useless ;-( at least to me).
>
> Cheers,
What did you and/or didn't you like about the book? My hesitation came not
only from the general nature of the subjects listed in the Table of Contents,
but also from the few pages that are dedicated to each subject (as indicated
from the Tabe of Contents). Of course, that is just an impression from the
little info I was able to retrieve off the website. You would better be
able to tell me if this is really so.
There are many books on performance on RISC architectures, and IBM has some
good documentation on performance tuning. Does this book add anything to that,
or amplify the already existing data that exists?
Thanks again, in advance.
ppluz...@earthlink.net
> > I was also looking for the latter when I got the 'AIX Internals' book.
> > Unfortunately, it proved to be quite useless ;-( at least to me).
> >
> What did you and/or didn't you like about the book?
I was particularly interested in the operation of the dynamic
loader. This excerpt from "Program Loading (Dynamic Binding)"
will give you an idea of the level of detail you can expect
to find in the rest of the book:
... p 97 ...
Program Loading (Dynamic Binding)
When a command is typed at the shell prompt, the shell forks a
child process, which then passes the string as an argument to
the exec system call. The exec system call performs a context
switch to the kernel system call handler, which in turn calls
the kernel's loader. Instead of copying the code for the routine
into the application, a small amount of "glue code" is added into
the application.
The glue code provides the name of the symbol (variable or
function), as well as the file name of the library or external
module where the symbol can be found. The loader resolves the
various symbolic references, locates the instruction portion of
the executable, and maps it to the text segment. The process of
mapping associates a part of a disk file to a portion of virtual
memory; no copying of the executable text portion is done.
As you can see, there isn't anything non-obvious or useful here ...
The book is typeset in a large font, the 2 paragraphs above take
half a page.
The section on loading proceeds to explain the concept of demand
paging in the next paragraph, and the fact that text is mapped
into a read-only segment, and that there are 16 segments in the
next paragraph. That concludes "Program Loading" section.
Regards,
Timothy J. Bogart
Nicholas Dronen <ndr...@io.frii.com> wrote:
>ppluz...@earthlink.net wrote:
>
<snip>
>
>Blech! I shall backtrack heartily from my previous statements.
Yes, I wish I had thought of that polite way of getting that
information across.
8-(
--
"I'm a big fan of ignorance based techniques, because humans
have a lot of ignorance, and we want to play our strong suit."
Eric Lander New Scientist
El Toro
> The book is typeset in a large font, the 2 paragraphs above take
> half a page.
Those two paragraphs took half a page!!
> The section on loading proceeds to explain the concept of demand
> paging in the next paragraph, and the fact that text is mapped
> into a read-only segment, and that there are 16 segments in the
> next paragraph. That concludes "Program Loading" section.
Ah. Thanks a bunch. I now have enough info to make my decision.
ppluz...@earthlink.net
> > The book is typeset in a large font, the 2 paragraphs above take
> > half a page.
>
> Those two paragraphs took half a page!!
Yes, they took *exactly* 1/2 of page 97.
Cheers,