[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

[Re: [Fwd: Included Interfaces without documentation update]]



Andrew:

I asked Dave Prosser at SCO about your list. Below are his comments.

Hope this helps some.

Doug

Dave Prosser wrote:
> 
> Doug Beattie wrote:
> > Would you happen to know off the top of your head, - don't spend time
> > looking into this is you don't already know, where these may be
> > documented?
> 
> Almost all of these look like glibc/system header names for various
> internal objects and functions.  I can guess about some of them
> without too much trouble.
> 
> > > _IO_2_1_stderr_
> > > _IO_2_1_stdin_
> > > _IO_2_1_stdout_
> 
> These are likely to be the objects whose address becomes stdin,
> stdout, and stderr.
> 
> > > _IO_feof
> > > _IO_getc
> > > _IO_putc
> > > _IO_puts
> 
> Real function versions of the _IO_-less macros?
> 
> > > _XtCopyFromArg
> > > _XtInherit
> > > _XtInheritTranslations
> > > _XtIsSubclassOf
> 
> X widget routines?
> 
> > > __ctype_b
> > > __ctype_tolower
> > > __ctype_toupper
> 
> The first is likely to be the array indexed for the original <ctype.h>
> macros.  The last two are likely to be arrays to do the upper-/lowercase
> mapping for bytes.
> 
> > > __dcgettext
> 
> Some underpinning for messaging.
> 
> > > __divdi3
> 
> ??
> 
> > > __environ
> 
> The "real" name for environ, the pointer to the array of char *'s
> manipulated by putenv() and getenv().
> 
> > > __libc_start_main
> 
> Possibly some initialization routine for glibc?
> 
> > > __lxstat
> 
> The real function underlying lstat(), just like our _lxstat() in UW.
> 
> > > __mempcpy
> > > __rawmemchr
> > > __stpcpy
> 
> Yet more variations on the similarly named string/memory routines?
> 
> > > _environ
> 
> An alias for __environ, maybe?
> 
> > > _nl_msg_cat_cntr
> 
> Smells like some bad messaging design showing up in a global object,
> but it could always be some supplemental routine for TOG's messaging
> interfaces.
> 
> > > _obstack_begin
> > > _obstack_newchunk
> 
> ??
> 
> > > _sys_errlist
> 
> The array containing the "C" locale strings used by strerror().
> 
> > > _sys_siglist
> 
> A similar array of signal names, I would guess.
> 
> > > adjtime
> 
> UW has an adjtime() routine which is a system call used to slow
> down or speed up the system's clock until enough real seconds
> have passed.  Probably the same in linux.
> 
> > > alphasort64
> 
> The LFS version of alphasort(), I would guess.  See the old BSD
> style of directory scanning.
> 
> > > cfsetspeed
> 
> UW has cf{g,s}{i,o}speed() and tc{g,s}speed().  It's bound to be
> yet another variation on the same.
> 
> > > dcgettext
> 
> An alias for __dcgettext.
> 
> > > dladdr
> 
> An RTLD function like dlopen() which is used to give one the nearest
> symbol and containing library/object for an address.  See UW.
> 
> > > errx
> > > getloadavg
> 
> ??
> 
> > > gettext
> 
> Most likely strongly tied to __dcgettext.
> 
> > > loc2
> 
> See the old (libgen) regexpr routines.  This is a global that points
> just after the matched substring, as I recall.
> 
> > > newscr
> 
> Curses something or other.
> 
> > > rexecoptions
> 
> Related to rexec()?
> 
> > > sem_timedwait
> 
> Could be a TOG/POSIX.1 realtime API, or an extension to the same.
> 
> > > setspent
> 
> Shadow password API, like setpwent().
> 
> > > sigandset
> > > sigisemptyset
> > > sigorset
> 
> Routines that do operations on sigset_t's.
> 
> > > strcasestr
> 
> Probably strstr() but ignoring upper-/lowercase differences.
> 
> > > strerror_r
> 
> A "reentrant" version of strerror().  I would guess you have to
> pass in a buffer and a size for it to copy the string into.
> 
> > > strndup
> 
> strdup() with an upper bound.
> 
> > > strtoq
> > > strtouq
> 
> Like strtol() and strtoul(), but for a "q", whatever that is.
> How about a "quad"?
> 
> > > strverscmp
> 
> Compare version strings???
> 
> > > textdomain
> 
> ??
> 
> > > vasprintf
> > > vdprintf
> 
> The "varargs" version of asprintf() and dprintf(), whatever they are.
> 
> > > verrx
> 
> The varargs version errx() which I am clueless about.  Maybe it's
> some error diagnostic routine to go along with warn() and warnx().
> 
> > > warn
> > > warnx
> 
> See above.
> 
> > > wcstoq
> > > wcstouq
> 
> The wide character string equivalents to strtoq() and strtouq().
> 
> --
> Dave Prosser
-- 
Douglas B. Beattie
------------------
Linux Test Architect - SCO, Inc.
dbb@sco.com



Reply to: