[Bug c/31006] New: long double constant is read as double in i386

[Bug c/31006] New: long double constant is read as double in i386

[chat95 at mac dot com]

when we #define PI 3.141592653589793238462643383279L
and long double a=PI;
PI is substituted in double precision.


-- 
           Summary: long double constant is read as double in i386
           Product: gcc
           Version: 4.2.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: chat95 at mac dot com
 GCC build triplet: i386-portbld-freebsd6.2
  GCC host triplet: i386-portbld-freebsd6.2
GCC target triplet: i386-portbld-freebsd6.2


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=31006

[Bug c/31006] long double constant is read as double in i386

[chat95 at mac dot com]

------- Comment #1 from chat95 at mac dot com  2007-03-01 08:00 -------
Created an attachment (id=13129)
 --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=13129&action=view)
test case


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=31006

[Bug c/31006] long double constant is read as double in i386

[chat95 at mac dot com]

------- Comment #2 from chat95 at mac dot com  2007-03-01 08:01 -------
I used following compiler:
% gcc42 -v
Using built-in specs.
Target: i386-portbld-freebsd6.2
Configured with: ./..//gcc-4.2-20070110/configure --disable-nls
--with-system-zlib --with-libiconv-prefix=/usr/local --with-gmp=/usr/local
--program-suffix=42 --libdir=/usr/local/lib/gcc-4.2.0
--with-gxx-include-dir=/usr/local/lib/gcc-4.2.0/include/c++/
--infodir=/usr/local/info/gcc42 --disable-rpath --prefix=/usr/local
i386-portbld-freebsd6.2
Thread model: posix
gcc version 4.2.0 20070110 (prerelease)


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=31006

[Bug c/31006] long double constant is read as double in i386

[chat95 at mac dot com]

------- Comment #3 from chat95 at mac dot com  2007-03-01 08:06 -------
1. 
//      long double PI * PI + IP 
        ld_a = ld_b = PI;
part becomes:
        movl    $560513024, %eax
        movl    $-921707870, %edx
        movl    $16384, %ecx
This number is equivalent to:
0x4000c90fdaa22168c000 
and we can write down as floating point number as:
2^(16384-16383) * 1.100100100001111 1101101010100010 0010000101101000
11000000000000 (B)

2.//      double PI * PI + IP
        d_a = d_b = PI;
part becomes:
        movl    $1413754136, %eax
        movl    $1074340347, %edx
this number is equivalent to:
0x400921fb54442d18,
and corresponding floating number is:
 2^(1024-1023)* 1.1001001000011111101101010100010001000010110100011000(B)
.

when we compare these significants of these two numbers:
1.100100100001111 1101101010100010 0010000101101000 11000(B)
1.100100100001111 1101101010100010 0010000101101000 11000 000000000 (B)

so - what we can say from here is:
#define PI 3.14...L is read as double for long double number with 0 filled 
lowest 16bits.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=31006

[Bug target/31006] long double constant is read as double in i386

[rguenth at gcc dot gnu dot org]

------- Comment #4 from rguenth at gcc dot gnu dot org  2007-03-01 12:22 -------
What is the expected output of your testcase?


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=31006

[Bug target/31006] long double constant is read as double in i386

[rguenth at gcc dot gnu dot org]

------- Comment #5 from rguenth at gcc dot gnu dot org  2007-03-01 12:27 -------
With gcc 4.1.2 and your testcase fixed to use %f for the double argument printf
I get

rguenther@murzim:/tmp> gcc -o t t.c -m32
rguenther@murzim:/tmp> ./t
long double                 = 3.141592653589793238512808959406
long double - double double = +3.141592653589793115997963468544
rguenther@murzim:/tmp> gcc -o t t.c     
rguenther@murzim:/tmp> ./t
long double                 = 3.141592653589793238512808959406
long double - double double = +3.141592653589793115997963468544

same for 4.2.0.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=31006

[Bug target/31006] long double constant is read as double in i386

[ubizjak at gmail dot com]

------- Comment #6 from ubizjak at gmail dot com  2009-09-17 12:46 -------
Your d_a is defined as double, so gcc truncates constant to fit in the "double"
value range. After that, you pass this double value as an argument to %Lf,
which is wrong. You should write:

        printf("long double - double double = %+5.30Lf\n", (long double) d_a);

As far as your testcse is concerned, there is no error, truncation from long
double to double is what you requested with "double" intermediate register.


-- 

ubizjak at gmail dot com changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |RESOLVED
         Resolution|                            |INVALID


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=31006