operator precedence differs from C language

operator precedence differs from C language

[Neo Liu]

I found that the operator precedence of "&" and "==" differed from C
language. Take a look at the following statements.

    if ($flags & 07 == 0) { ... }
        I thought this statement would first calculate ($flags & 07)
and then calculate (($flags & 07) == 0), but the it is not the case.
        It first calculates (07 == 0) and then ($flags & (07 == 0)),
so the result is 0 but not the expected 1. As a solution, I wrote the
        statement as:
            if (($flags & 07) == 0) { ... }
        to get the computing order I want

Is this the stap language feature, or should I report this as a bug?

Thanks!

Re: operator precedence differs from C language

[Josh Stone]

On 03/31/2010 07:36 PM, Neo Liu wrote:
> I found that the operator precedence of "&" and "==" differed from C
> language. Take a look at the following statements.

That's not true -- in C and in stap, relational operators bind more
tightly than bitwise operators.  http://tinyurl.com/9q6szp

Here's a program to convince yourself of this:

#include <stdio.h>
#define TEST(e) printf("( %s ) = %d\n", #e, e)
int main(int argc, const char *argv[])
{
    if (argc < 2) return 1;
    int x = atoi(argv[1]);
    TEST(x);
    TEST(x & 07);
    TEST((x & 07) == 0);
    TEST(x & (07 == 0));
    TEST(x & 07 == 0);
    return 0;
}


Josh

Re: operator precedence differs from C language

[Frank Ch. Eigler]

Neo Liu <diabloneo@gmail.com> writes:

> I found that the operator precedence of "&" and "==" differed from C
> language. Take a look at the following statements.
>     if ($flags & 07 == 0) { ... }

Actually, I think we do it the same way C does.  Comparison operators
such as "==" have higher precedence.

% cat foo.c
int main () {
  printf("%d\n", 0xf0 & 0xff == 0);
}
% cat foo.stp
probe begin {printf("%d\n", 0xf0 & 0xff == 0); exit()}

Both these print 0 for == or !=.


- FChE

Re: operator precedence differs from C language

[Josh Stone]

On 04/01/2010 09:29 AM, Josh Stone wrote:
> On 03/31/2010 07:36 PM, Neo Liu wrote:
>> I found that the operator precedence of "&" and "==" differed from C
>> language. Take a look at the following statements.
> 
> That's not true -- in C and in stap, relational operators bind more
> tightly than bitwise operators.  http://tinyurl.com/9q6szp

You might also find this interesting:
http://cm.bell-labs.com/cm/cs/who/dmr/chist.html

"""
Neonatal C

Rapid changes continued after the language had been named, for example
the introduction of the && and || operators. In BCPL and B, the
evaluation of expressions depends on context: within if and other
conditional statements that compare an expression's value with zero,
these languages place a special interpretation on the and (&) and or (|)
operators. In ordinary contexts, they operate bitwise, but in the B
statement

    if (e1 & e2) ...

the compiler must evaluate e1 and if it is non-zero, evaluate e2, and if
it too is non-zero, elaborate the statement dependent on the if. The
requirement descends recursively on & and | operators within e1 and e2.
The short-circuit semantics of the Boolean operators in such
`truth-value' context seemed desirable, but the overloading of the
operators was difficult to explain and use. At the suggestion of Alan
Snyder, I introduced the && and || operators to make the mechanism more
explicit.

Their tardy introduction explains an infelicity of C's precedence rules.
In B one writes

    if (a==b & c) ...

to check whether a equals b and c is non-zero; in such a conditional
expression it is better that & have lower precedence than ==. In
converting from B to C, one wants to replace & by && in such a
statement; to make the conversion less painful, we decided to keep the
precedence of the & operator the same relative to ==, and merely split
the precedence of && slightly from &. Today, it seems that it would have
been preferable to move the relative precedences of & and ==, and
thereby simplify a common C idiom: to test a masked value against
another value, one must write

    if ((a&mask) == b) ...

where the inner parentheses are required but easily forgotten.
"""

Fwd: operator precedence differs from C language

[Neo Liu]

---------- Forwarded message ----------
From: Neo Liu <diabloneo@gmail.com>
Date: Fri, Apr 2, 2010 at 9:49 AM
Subject: Re: operator precedence differs from C language
To: Josh Stone <jistone@redhat.com>


On Fri, Apr 2, 2010 at 1:15 AM, Josh Stone <jistone@redhat.com> wrote:
> On 04/01/2010 09:29 AM, Josh Stone wrote:
>> On 03/31/2010 07:36 PM, Neo Liu wrote:
>>> I found that the operator precedence of "&" and "==" differed from C
>>> language. Take a look at the following statements.
>>
>> That's not true -- in C and in stap, relational operators bind more
>> tightly than bitwise operators.  http://tinyurl.com/9q6szp

I ran your test program and found that my problems due to my misunderstanding
of the C Language.

>
> You might also find this interesting:
> http://cm.bell-labs.com/cm/cs/who/dmr/chist.html
>
> """
> Neonatal C
>
> Rapid changes continued after the language had been named, for example
> the introduction of the && and || operators. In BCPL and B, the
> evaluation of expressions depends on context: within if and other
> conditional statements that compare an expression's value with zero,
> these languages place a special interpretation on the and (&) and or (|)
> operators. In ordinary contexts, they operate bitwise, but in the B
> statement
>
>    if (e1 & e2) ...
>
> the compiler must evaluate e1 and if it is non-zero, evaluate e2, and if
> it too is non-zero, elaborate the statement dependent on the if. The
> requirement descends recursively on & and | operators within e1 and e2.
> The short-circuit semantics of the Boolean operators in such
> `truth-value' context seemed desirable, but the overloading of the
> operators was difficult to explain and use. At the suggestion of Alan
> Snyder, I introduced the && and || operators to make the mechanism more
> explicit.
>
> Their tardy introduction explains an infelicity of C's precedence rules.
> In B one writes
>
>    if (a==b & c) ...
>
> to check whether a equals b and c is non-zero; in such a conditional
> expression it is better that & have lower precedence than ==. In
> converting from B to C, one wants to replace & by && in such a
> statement; to make the conversion less painful, we decided to keep the
> precedence of the & operator the same relative to ==, and merely split
> the precedence of && slightly from &. Today, it seems that it would have
> been preferable to move the relative precedences of & and ==, and
> thereby simplify a common C idiom: to test a masked value against
> another value, one must write
>
>    if ((a&mask) == b) ...
>
> where the inner parentheses are required but easily forgotten.
> """
>

Thank you for your elaboration.