T
Tom St Denis
Knady said:
step 1. "man malloc"
step 2. Pay attention in class when they teach you how to use pointers.
step 3. Don't ask for homework solutions in usenet.
Tom
step 1. "man malloc"
step 2. Pay attention in class when they teach you how to use pointers.
step 3. Don't ask for homework solutions in usenet.
Tom
J
Joona I Palaste
Tom St Denis said:
What makes you so sure Knady is working on a UNIX system?
N
Nils Petter Vaskinn
Do you know how to do it without dynamic memory (for a fixed sice
matrix?), start out with that and modify it to use dynamic memory.
Atleat he was honest about it beeing homework.
Step 4. Try to code something up yourself
Step 5. If you can't get it to work post the code that you have and
someone will probably point out your errors.
matrix?), start out with that and modify it to use dynamic memory.
Atleat he was honest about it beeing homework.
Step 4. Try to code something up yourself
Step 5. If you can't get it to work post the code that you have and
someone will probably point out your errors.
T
Tom St Denis
Joona said:
Who cares if he isn't? Stupid windows users can trap themselves in
ignorance for all I care. As for me and my Gentoo'd box I'll go hug my
manual pages now... ;-)
Tom
K
Knady
Hi,
I have the following problem, I must to do my assignment, but I really do
not know how to use the malloc.
I need create a program that will be used to do some algebrical computation
on the matrix.
How I can create dynamical the matrices with the name in order to be able to
recall them.
Thx
I have the following problem, I must to do my assignment, but I really do
not know how to use the malloc.
I need create a program that will be used to do some algebrical computation
on the matrix.
How I can create dynamical the matrices with the name in order to be able to
recall them.
Thx
M
Malcolm
Knady said:
To create a matrix
double *mtx = malloc(width * height * sizeof(double));
if(mtx == 0)
/* out of memory */
To access an element
double element = mtx[y * width + x];
Once you've finished
free(mtx);
A
Al Bowers
Knady said:
The faq is a good source for help on multidimensional array dynamic
allocations. The link:
http://www.eskimo.com/~scs/C-faq/q6.16.html
C
Chris Fogelklou
Anyone know why engineers get such a bad reputation for being antisocial?
Instead of wasting my time berating you for being a "stupid windows user"
I'll answer the question...
It's actually quite simple... Malloc returns a pointer to a chunk of memory
of a certain size in bytes. You specify the number of bytes in the
parameter you pass to malloc. The prototype for malloc is
void *malloc(int size);
So, you get a void pointer in return... you need to cast it to something
useable, like your matrix, for instance...
typedef int my_matrix_t[20][20];//Hope I did this typedef right!
//Allocate memory for matrix and store in pointer. Cast so compiler doesn't
puke.
my_matrix_t *pmatrix = (my_matrix_t *)malloc(sizeof(my_matrix_t));
Instead of wasting my time berating you for being a "stupid windows user"
I'll answer the question...
It's actually quite simple... Malloc returns a pointer to a chunk of memory
of a certain size in bytes. You specify the number of bytes in the
parameter you pass to malloc. The prototype for malloc is
void *malloc(int size);
So, you get a void pointer in return... you need to cast it to something
useable, like your matrix, for instance...
typedef int my_matrix_t[20][20];//Hope I did this typedef right!
//Allocate memory for matrix and store in pointer. Cast so compiler doesn't
puke.
my_matrix_t *pmatrix = (my_matrix_t *)malloc(sizeof(my_matrix_t));
T
Tom St Denis
Chris said:
And even at that you failed.
That's not the prototype for malloc. It's
void *malloc(size_t size);
Um, no. In C you need not cast void pointers.
This isn't dynamic though [hint: read the OPs question] it's always 20x20
If your compiler pukes at that (without the cast) then you need a new
compiler.
Tom
J
Joona I Palaste
And what about stupid UNIX users? Oh, right. I forgot. To be able to so
much as boot up a UNIX machine you have to be both a hardware and a
software high wizard. How silly of me. =)
Anyway, I hope you are kidding about your attitude towards Windows
users. I don't like Microsoft any more than you do, but that's no
excuse to prejudice against their entire userbase.
T
Tom St Denis
Joona said:
Perhaps but honestly there aren't many "good" reasons to use Windows. For
the most part Windows lowers that standards for users.
But even if the dude is a Windows user he could search MSDN for malloc ;-)
or search for manpages online, etc, etc...The fact that the guy doesn't
know what malloc does shows he has zero research skills.
I mean it's one thing to ask standards questions [e.g. "is this in ISO C",
etc] but honestly I'd expect any half competent developer to at least
memorize the vast majority of C functions and be able to research the rest.
Tom
C
Chris Fogelklou
Anyone know why engineers get such a bad reputation for being
antisocial?
Hmmm... thanks.
It's irrelevant to the answer, and normally size_t is an int anyway. I
could have said it's a size_t, but it might have created more problems than
it solved for the question-asker.
Um, yes you do, if you want truly portable code.
Alright... Do you want to write one up for our embedded system? There have
two compilers, one based on GCC, and neither of them like it if you equate
two pointer types to each other without casting.
malloc. I was trying to solve his confusion relating to malloc itself.
Hopefully with this basic understanding, he can solve the dynamic part of
the problem himself. (I should have also mentioned how to use free, but
someone else did a pretty good job of that.)
The link provided by Al Bowers is an excellent source for info on the
dynamic memory problem.
Cheers,
Chris
antisocial?
Hmmm... thanks.
It's irrelevant to the answer, and normally size_t is an int anyway. I
could have said it's a size_t, but it might have created more problems than
it solved for the question-asker.
Um, yes you do, if you want truly portable code.
Alright... Do you want to write one up for our embedded system? There have
two compilers, one based on GCC, and neither of them like it if you equate
two pointer types to each other without casting.
This is true, but I got the impression that he had no idea how to use
malloc. I was trying to solve his confusion relating to malloc itself.
Hopefully with this basic understanding, he can solve the dynamic part of
the problem himself. (I should have also mentioned how to use free, but
someone else did a pretty good job of that.)
The link provided by Al Bowers is an excellent source for info on the
dynamic memory problem.
Cheers,
Chris
M
Martin Dickopp
Chris Fogelklou said:
No. According to 7.17#2, `size_t' is required to be an unsigned integer
type, so it cannot be `int'.
Chapter and verse, please?
Martin
T
Tom St Denis
Chris said:
You are most certainly welcome.
Um first off, no "size_t" isn't int. I'd think if anything it would be
"unsigned" not just an int. Anyways the type is size_t. Why bother
replying with "seemingly useful facts" if you're just going to mess with
them?
No, no you don't. The C standards [C90, C99] don't require you to cast a
void pointer to any other pointer type. If you're code is supposed to work
in C++ as well then maybe. However, putting the cast in can mask a more
serious problem that you forget to include stdlib.h.
Um dude I've coded for embedded platforms before [from 8085, AVR, 68xx
series to the ARM based boards] and in all cases I've never casted a void
pointer to something.
In particular for the ARM boards you can use GCC and it will accept the
non-cast code just fine.
Me thinks you might want to preface your responses with "I used casts cuz my
non-standard compiler asked me to". This is afterall comp.lang.c and we're
not talking about "vendor lang c"
Point him to the manual pages. Why bother answering his question with code?
How the hell do you think all these "coders" get out there? They take an
ounce of knowledge and spread it so thin to get a job....
If the dude can't learn from a manual page [or googling on the net] then too
F'ing bad.
Tom
C
Chris Fogelklou
whatever...
Martin Dickopp said:
T
Tom St Denis
Chris said:
Yeah, don't let "the facts" stand in the way of an otherwise poor attempt at
a rant.... ;-)
Tom
C
Chris Fogelklou
I used casts cuz my non-standard compiler asked me to 
Anyway... I regret answering now. Please refer to man malloc for more
assistance.
Done.
Anyway... I regret answering now. Please refer to man malloc for more
assistance.
Done.
Tom St Denis said:
T
Thomas Stegen
Chris said:
size_t can be many things. But it is never int. Never.
You could have said that size_t is just an unsigned integer.
Let me quote Tom St Denis on this one:
"Um, no. In C you need not cast void pointers." Before you even attempt
to contradict me go look up the vast amount of discussion on this
topic right here in this newsgroup. If you can come up with an argument
not already proffered please do so.
If your compiler requires a cast from pointer to void to a pointer to
any other object type then it is broken. Why, because it will not
accept all strictly conforming C programs (for stand alone
implementations).
In the following T and S can not be void.
If you have a pointer to S and a pointer to T then you need the cast
to assign one to the other AND you need to know that they are
compatable. A pointer to void can be implicitly converted to any
pointer type, no cast required.
For more information read section 6.3.2.3 of the C99 standard.
C
Chris Fogelklou
Hi Thomas,
Yes, I was wrong about the int. It should definitely an unsigned int. I
wrote my response very quickly and obviously didn't pull up my handy C
library help file when I did it. I just took offense to the "higher than
thou" attitude to a question that was out-and-out stated to be a homework
assignment by a beginner.
I was trying to express how to use malloc. Thinking that malloc takes an
int would probably not affect the application code in any way (and yes I
know that ints have a smaller maximum absolute value do to their signedness,
but think about what would really happen... hint... compiler casting from
positive int to unsigned int... before you catch me on this one, although I
do expect someone to respond in any case.)
As long as he isn't writing his own malloc function, don't you think I still
might have provided some handy, if perhaps a bit imprecise, advice?
That said, apparently the embedded compilers I've been using are broken. I
have a cross platform environment going here, and when building for win32,
casting a void * to anything else is OK. But when I move to the embedded
version, it breaks. I am referring only to my firsthand experience.
Cheers,
Chris
PS. Next time I write anything in this NG, I'll ensure that I have a
complete bibliography appended to the bottom and that I triple check every
fact stated with the latest ANSI standards and their addendums.
Yes, I was wrong about the int. It should definitely an unsigned int. I
wrote my response very quickly and obviously didn't pull up my handy C
library help file when I did it. I just took offense to the "higher than
thou" attitude to a question that was out-and-out stated to be a homework
assignment by a beginner.
I was trying to express how to use malloc. Thinking that malloc takes an
int would probably not affect the application code in any way (and yes I
know that ints have a smaller maximum absolute value do to their signedness,
but think about what would really happen... hint... compiler casting from
positive int to unsigned int... before you catch me on this one, although I
do expect someone to respond in any case.)
As long as he isn't writing his own malloc function, don't you think I still
might have provided some handy, if perhaps a bit imprecise, advice?
That said, apparently the embedded compilers I've been using are broken. I
have a cross platform environment going here, and when building for win32,
casting a void * to anything else is OK. But when I move to the embedded
version, it breaks. I am referring only to my firsthand experience.
Cheers,
Chris
PS. Next time I write anything in this NG, I'll ensure that I have a
complete bibliography appended to the bottom and that I triple check every
fact stated with the latest ANSI standards and their addendums.
M
Martin Dickopp
Chris Fogelklou said:
Actually, the standard only requires `size_t' to be an unsigned integer
type. Sometimes it is `unsigned int', sometimes it is another unsigned
integer type.
When `malloc' works on some systems and doesn't on others, that is often
an indication of incorrect code. A typical error of this kind is to use
`malloc' without a proper declaration in scope (which one usually gets
by including <stdlib.h>). Usually, the compiler would warn you if use
`malloc' and haven't included <stdlib.h>, but by casting the return
value, the warning is suppressed. Despite the lack of a warning, the
code is still incorrect and will break on some systems.
The latter is certainly a good idea. That's what I do unless I'm very
sure. (I still make mistakes sometimes, but fortunately, someone else
will reply and point them out very quickly, so it's not a problem.)
Martin