peculiar problem with expression evaluation

[sarab.nsit]

Hi All,

I observed something surprising:
Here is code snippet

*****************
int i = 5;
i = i++ + ((((8 << 1) << (5 << 1)) * 3 * 3)) + ++i;
printf ( "i = %d \n " ,i);
*****************
output
i = 147469
But if i just assign the middle expression i.e. ((((8 << 1) << (5 <<
1)) * 3 * 3)) to an integer k my output changes

**************
int i = 5;
int k = 0;
k = ((((8 << 1) << (5 << 1)) * 3 * 3));
i = i++ + k + ++i;
printf ( "i = %d \n " ,i);
***************
output
i = 147468

As far as i know postfix increment should evaluate only at the end of
expression as in case 2 . but in case 1 it is getting incremented
during evaluation .

I am not able to figure out why i am getting this difference in
output.

Thanks in advance
Sarab

[Ian Collins]

sa********@gmail.com wrote:

Hi All,

I observed something surprising:
Here is code snippet

*****************
int i = 5;
i = i++ + ((((8 << 1) << (5 << 1)) * 3 * 3)) + ++i;
printf ( "i = %d \n " ,i);
*****************

Please see the scores of previous postings about this form of undefined
behavior.

--
Ian Collins.

[sarab.nsit]

Forgot to tell my compiler version . i am using gcc compiler i.e.
gcc (GCC) 3.2.3 20030502 (Red Hat Linux 3.2.3-47)

[Mark Bluemel]

sa********@gmail.com wrote:

Forgot to tell my compiler version . i am using gcc compiler i.e.
gcc (GCC) 3.2.3 20030502 (Red Hat Linux 3.2.3-47)

We don't care about that. What version of the C standard did you
consult? What version of the C-FAQ did you neglect to read?

Some suggestions:-

* Read the FAQ at c-faq.com

* Search this group in Google for "sequence point"

* Don't waste our time and network bandwidth

[Richard]

Ian Collins <ia******@hotmail.comwrites:

sa********@gmail.com wrote:

>Hi All,

I observed something surprising:
Here is code snippet

*****************
int i = 5;
i = i++ + ((((8 << 1) << (5 << 1)) * 3 * 3)) + ++i;
printf ( "i = %d \n " ,i);
*****************

Please see the scores of previous postings about this form of undefined
behavior.

By this rational, you should post this reply to 99% of posts which
cover material already covered and archived. If *you* are tired of
answering these questions I suggest you have one of two options:

a) Leave
b) Leave it to someone else to answer.

A noob might have no experience to spot the issue. It IS a common
issue. It is not necessarily easy to find. No one should be expected to
memorise a FAQ before posting to a help news group.

[Charlton Wilbur]

>>>>"R" == Richard <rg****@gmail.comwrites:

RA noob might have no experience to spot the issue. It IS a
Rcommon issue. It is not necessarily easy to find. No one should
Rbe expected to memorise a FAQ before posting to a help news
Rgroup.

No, but everyone *is* expected to *consult* a FAQ before posting to a
news group. Doing so makes memorizing the FAQ unnecessary.

Charlton
--
Charlton Wilbur
cw*****@chromatico.net

[Ian Collins]

Richard wrote:

Ian Collins <ia******@hotmail.comwrites:

>sa********@gmail.com wrote:

>>Hi All,

I observed something surprising:
Here is code snippet

*****************
int i = 5;
i = i++ + ((((8 << 1) << (5 << 1)) * 3 * 3)) + ++i;
printf ( "i = %d \n " ,i);
*****************

Please see the scores of previous postings about this form of undefined
behavior.

A noob might have no experience to spot the issue. It IS a common
issue. It is not necessarily easy to find. No one should be expected to
memorise a FAQ before posting to a help news group.

Bollocks code like that common? Good to know there'll be plenty of work
fixing it...

--
Ian Collins.

[Martin Ambuhl]

Richard wrote:

Ian Collins <ia******@hotmail.comwrites:

>sa********@gmail.com wrote:

>>Hi All,

I observed something surprising:
Here is code snippet

*****************
int i = 5;
i = i++ + ((((8 << 1) << (5 << 1)) * 3 * 3)) + ++i;
printf ( "i = %d \n " ,i);
*****************

Please see the scores of previous postings about this form of undefined
behavior.

By this rational, you should post this reply to 99% of posts which
cover material already covered and archived. If *you* are tired of
answering these questions I suggest you have one of two options:

a) Leave
b) Leave it to someone else to answer.

I suggest you be condemned to spend the rest of your life answering the
same questions over and over. The fact is that (b) is covered by the
FAQ, but not by you, who as usual; carps without contributing anything.
You will note that you did *nothing* toward answering the origianl
question. Your behavior is nothing but destructive.

A noob might have no experience to spot the issue. It IS a common
issue. It is not necessarily easy to find. No one should be expected to
memorise a FAQ before posting to a help news group.

And what, exactly, did *you* contribute to answering the question?
Nothing. You have only negativism to offer. Twit.

[CBFalconer]

"sa********@gmail.com" wrote:

>
Forgot to tell my compiler version . i am using gcc compiler i.e.
gcc (GCC) 3.2.3 20030502 (Red Hat Linux 3.2.3-47)

So what? If you have a problem, spit it out. In this group the
actual compiler doesn't matter.

--
Chuck F (cbfalconer at maineline dot net)
Available for consulting/temporary embedded and systems.
<http://cbfalconer.home.att.net>

--
Posted via a free Usenet account from http://www.teranews.com

[Keith Thompson]

Richard <rg****@gmail.comwrites:
[snip]

And I frequently post on topic comments. I just don't feel the need to
jump in and berate nOObs like some, too frequently, do. I also don't feel
the need to repeat an answer that someone else has already
addressed. And just by hanging around I learn a lot here too.

Bottom line : If you can't be nice, say nothing at all.

[...]

Time for me to exhibit some more foolish optimism ...

Richard, I've occasionally seen you post topical comments, but it
seems to me that the vast majority of your posts (at least recently)
are complaints about other people's complaints. If you followed your
own advice, "If you can't be nice, say nothing at all", then you would
post far less than you do (and your S/N ratio would rise
considerably). I fail to see how berating the regulars, as you do, is
so much better than berating newbies. You ask us to ignore posts that
we think are inappropriate, yet you spend a great deal of time
responding to posts that you think are inappropriate.

Please give it some thought.

--
Keith Thompson (The_Other_Keith) ks***@mib.org <http://www.ghoti.net/~kst>
San Diego Supercomputer Center <* <http://users.sdsc.edu/~kst>
"We must do something. This is something. Therefore, we must do this."
-- Antony Jay and Jonathan Lynn, "Yes Minister"

[Army1987]

On Tue, 21 Aug 2007 17:31:27 +0200, Richard wrote:

Ian Collins <ia******@hotmail.comwrites:

>sa********@gmail.com wrote:

>>Hi All,

I observed something surprising:
Here is code snippet

*****************
int i = 5;
i = i++ + ((((8 << 1) << (5 << 1)) * 3 * 3)) + ++i;
printf ( "i = %d \n " ,i);
*****************

Please see the scores of previous postings about this form of undefined
behavior.

By this rational, you should post this reply to 99% of posts which
cover material already covered and archived. If *you* are tired of
answering these questions I suggest you have one of two options:

a) Leave
b) Leave it to someone else to answer.

A noob might have no experience to spot the issue. It IS a common
issue. It is not necessarily easy to find. No one should be expected to
memorise a FAQ before posting to a help news group.

Yeah, everybody tries to assign three different values to the same
variable at once and are surprised to see that the result is not
what they expected... (And everybody tries to stuff 9 * 2**40 in
an int...)

--
Army1987 (Replace "NOSPAM" with "email")
No-one ever won a game by resigning. -- S. Tartakower

[Default User]

Keith Thompson wrote:

Richard <rg****@gmail.comwrites:
[snip]

Bottom line : If you can't be nice, say nothing at all.

[...]

Time for me to exhibit some more foolish optimism ...

Richard, I've occasionally seen you post topical comments, but it
seems to me that the vast majority of your posts (at least recently)
are complaints about other people's complaints.

Seems obvious to me that he's been trolling for some time now. That's
why I plonked him quite a while back.


Brian