[Bug libc/16459] New: endian.h does not define byte-order conversion functions

[Bug libc/16459] New: endian.h does not define byte-order conversion functions

[jaak at ristioja dot ee]

https://sourceware.org/bugzilla/show_bug.cgi?id=16459

            Bug ID: 16459
           Summary: endian.h does not define byte-order conversion
                    functions
           Product: glibc
           Version: 2.17
            Status: NEW
          Severity: normal
          Priority: P2
         Component: libc
          Assignee: unassigned at sourceware dot org
          Reporter: jaak at ristioja dot ee
                CC: drepper.fsp at gmail dot com

"man 3 endian" says it does, but actually it doesn't. If it would, one would be
able to take the address of such a function:

  #define _BSD_SOURCE
  #include <endian.h>

  void g() {
    uint64_t (*f)(uint64_t) = &htole64;
  }

$ gcc -Wimplicit-function-declaration test2.c -c
test.c: In function 'g':
test.c:5:3: warning: implicit declaration of function 'uint64_t'
[-Wimplicit-function-declaration]
test.c:5:14: error: 'f' undeclared (first use in this function)
test.c:5:14: note: each undeclared identifier is reported only once for each
function it appears in
test.c:5:30: error: 'htole64' undeclared (first use in this function)

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

[Bug libc/16459] endian.h does not define byte-order conversion functions

[joseph at codesourcery dot com]

http://sourceware.org/bugzilla/show_bug.cgi?id=16459

--- Comment #1 from joseph at codesourcery dot com <joseph at codesourcery dot com> ---
glibc does not provide manpages, and they are not relevant to what glibc's 
interface should be.  (glibc's code should of course be consistent with 
the manual shipped with glibc; I haven't checked what if anything that 
says about these interfaces.)

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

[Bug libc/16459] endian.h does not define byte-order conversion functions

[jaak at ristioja dot ee]

https://sourceware.org/bugzilla/show_bug.cgi?id=16459

--- Comment #2 from Jaak Ristioja <jaak at ristioja dot ee> ---
(In reply to joseph@codesourcery.com from comment #1)
> glibc does not provide manpages, and they are not relevant to what glibc's 
> interface should be.  (glibc's code should of course be consistent with 
> the manual shipped with glibc; I haven't checked what if anything that 
> says about these interfaces.)

It appears that these interfaces are undocumented in the Texinfo manual.

Why not just add these as functions also, e.g. as done with htons and friends
in <netinet/in.h>:

  extern uint16_t htons (uint16_t __hostshort)
     __THROW __attribute__ ((__const__));

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

Re: [Bug libc/16459] endian.h does not define byte-order conversion functions

[Ondřej Bílka]

There is flexibility that we could use a function-like macro instead
function. If this needs fixing then by changing pattern

 #   define htobe64(x) __bswap_64 (x)
 #   define htole64(x) (x)

to


 #   define htobe64 __bswap_64
 #   define htole64 __bkeep_64

where __bkeep_64 is a identity function like

static uint64_t
__bkeep_64 (uint64_t x)
{
  return x;
}

[Bug libc/16459] endian.h does not define byte-order conversion functions

[neleai at seznam dot cz]

http://sourceware.org/bugzilla/show_bug.cgi?id=16459

--- Comment #3 from Ondrej Bilka <neleai at seznam dot cz> ---
There is flexibility that we could use a function-like macro instead
function. If this needs fixing then by changing pattern

 #   define htobe64(x) __bswap_64 (x)
 #   define htole64(x) (x)

to


 #   define htobe64 __bswap_64
 #   define htole64 __bkeep_64

where __bkeep_64 is a identity function like

static uint64_t
__bkeep_64 (uint64_t x)
{
  return x;
}

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

[Bug libc/16459] endian.h does not define byte-order conversion functions

[fweimer at redhat dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=16459

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.