[Bug libc/13958] New: printf(3) format spec "%La" generates odd output (at least on x86)

[Bug libc/13958] New: printf(3) format spec "%La" generates odd output (at least on x86)

[cloos at jhcloos dot com]

http://sourceware.org/bugzilla/show_bug.cgi?id=13958

             Bug #: 13958
           Summary: printf(3) format spec "%La" generates odd output (at
                    least on x86)
           Product: glibc
           Version: 2.14
            Status: NEW
          Severity: normal
          Priority: P2
         Component: libc
        AssignedTo: unassigned@sourceware.org
        ReportedBy: cloos@jhcloos.com
                CC: drepper.fsp@gmail.com
    Classification: Unclassified


Some examples:

double: 1 (0x1p+0)
long double: 1 (0x8p-3)

double: 1.5 (0x1.8p+0)
long double: 1.5 (0xcp-3)

I presume this is related to the 80-bit representation’s lack of an implied
bit?

I noticed this while looking at bug #13940; this is based on Andreas’ code in
comment #1 there:


#include <stdlib.h>
#include <stdio.h>
int
main (int c, char *v[])
{
  float f;
  double d;
  long double ld;

  ld = -1.0L;
  if (c > 1)
        ld = strtold (v[1], NULL);

  f = (float)ld;
  d = (double)ld;

  printf ("float %.8g (%a)\n", f,f);
  printf ("double %.16g (%a)\n", d,d);
  printf ("long double %.20Lg (%La)\n", ld, ld);

  return 0;
}

-- 
Configure bugmail: http://sourceware.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

[Bug libc/13958] printf(3) format spec "%La" generates odd output (at least on x86)

[schwab@linux-m68k.org]

http://sourceware.org/bugzilla/show_bug.cgi?id=13958

Andreas Schwab <schwab@linux-m68k.org> changed:

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

--- Comment #1 from Andreas Schwab <schwab@linux-m68k.org> 2012-04-06 20:08:15 UTC ---
There is nothing wrong with it.  The %a format only requires a single (nonzero
when normalized) hexadecimal digit before the decimal point, but does not
restrict it any further.  The standard explicitly acknowleges that the leading
digit can be chosen freely for alignment purpose.

-- 
Configure bugmail: http://sourceware.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

[Bug libc/13958] printf(3) format spec "%La" generates odd output (at least on x86)

[fweimer at redhat dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=13958

Florian Weimer <fweimer at redhat dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
              Flags|                            |security-

-- 
You are receiving this mail because:
You are on the CC list for the bug.