Home • About Us •A-Z Index• Search •

1. Internationalised System Calls and Libraries Extended

Product Information

Environment Specification

Temporary Waivers

1.1 General Attributes

1.1.1 Feature Groups

Question 1: Which of the following Feature Groups are supported by the implementation?

Response

POSIX.2 C-language Binding	Yes
Shared Memory	Yes
Encryption	Yes
Enhanced Internationalisation	Yes
X/Open UNIX Extension	Yes

The POSIX.2 C-language Binding, Shared Memory, Enhanced Internationalisation and X/Open UNIX Extension Feature Groups are mandatory for Internationalised System Calls and Libraries Extended conformance.

Support for a Feature Group can only be claimed if all interfaces in that group behave according to the relevant descriptions in System Interfaces and Headers, Issue 4, Version 2.

The interfaces in the Encryption Feature Group must exist, whether or not the Feature Group is supported, and each interface must either behave according to the description in System Interfaces and Headers, Issue 4, Version 2, or indicate an error, with errno set to [ENOSYS].

Rationale

System Interfaces and Headers, Issue 4, Version 2 states that the system may provide one or more of the Feature Groups listed. XPG4 Components, version 2 states that the POSIX.2 C-language Binding, Shared Memory, Enhanced Internationalisation and X/Open UNIX Extension Feature Groups are mandatory for compliance to the Internationalised System Calls and Libraries Extended component.

1.1.2 POSIX.1 Supported Features

Question 2: Which of the following options, specified in the <unistd.h> header file, are available on the system?

Response

Macro Name	Meaning	Provided
_POSIX_CHOWN_RESTRICTED	The use of chown() is restricted to a process with appropriate privileges, and to changing the group ID of a file only to the effective group ID of the process or one of its supplementary group IDs.	Yes
_POSIX_NO_TRUNC	Pathname components longer than {NAME_MAX) generate an error.	Yes
_POSIX_VDISABLE	Terminal special characters defined in <termios.h> can be disabled using this character value.	Yes
_POSIX_SAVED_IDS	Each process has a saved set-user-ID and a saved set-group-ID.	Yes
_POSIX_JOB_CONTROL	Implementation supports job control.	Yes

Rationale

For a conformant implementation, all of these POSIX features must be provided. In some cases the feature need not be provided for all files or devices supported by the implementation.

1.1.3 Float, Stdio and Limit Values

Question 3: What are the values associated with the following constants specified in the <float.h> header file?

Response

Macro Name	Meaning	Value
FLT_RADIX	Radix of the exponent representation.	2
FLT_MANT_DIG	Number of base-FLT_RADIX digits in the float significand.	24
DBL_MANT_DIG	Number of base-FLT_RADIX digits in the double significand.	53
LDBL_MANT_DIG	Number of base-FLT_RADIX digits in the long double significand.	64
FLT_DIG	Number of decimal digits, q, such that any floating point number with q digits can be rounded into a float representation and back again without change to the q digits.	6
DBL_DIG	Number of decimal digits, q, such that any floating point number with q digits can be rounded into a double representation and back again without change to the q digits.	15
LDBL_DIG	Number of decimal digits, q, such that any floating point number with q digits can be rounded into a long double representation and back again without change to the q digits.	18
FLT_MIN_EXP	Minimum negative integer such that FLT_RADIX raised to that power minus 1 is a normalised float.	-125
DBL_MIN_EXP	Minimum negative integer such that FLT_RADIX raised to that power minus 1 is a normalised double.	-1021
LDBL_MIN_EXP	Minimum negative integer such that FLT_RADIX raised to that power minus 1 is a normalised long double.	-16381
FLT_MIN_10_EXP	Minimum negative integer such that 10 raised to that power is in the range of normalised floats.	-37
DBL_MIN_10_EXP	Minimum negative integer such that 10 raised to that power is in the range of normalised doubles.	-307
LDBL_MIN_10_EXP	Minimum negative integer such that 10 raised to that power is in the range of normalised long doubles.	-4931
FLT_MAX_EXP	Maximum integer such that FLT_RADIX raised to that power minus 1 is a representable finite float.	128
DBL_MAX_EXP	Maximum integer such that FLT_RADIX raised to that power minus 1 is a representable finite double.	1024
LDBL_MAX_EXP	Maximum integer such that FLT_RADIX raised to that power minus 1 is a representable finite long double.	16384
FLT_MAX_10_EXP	Maximum integer such that 10 raised to that power is in the range of representable finite floats.	38
DBL_MAX_10_EXP	Maximum integer such that 10 raised to that power is in the range of representable finite doubles.	308
LDBL_MAX_10_EXP	Maximum integer such that 10 raised to that power is in the range of representable finite long doubles.	4932
FLT_MAX	Maximum representable finite float.	0x1.fffffep+127F
DBL_MAX	Maximum representable finite double.	0x1.fffffffffffffp+1023
LDBL_MAX	Maximum representable finite long double.	0xf.fffffffffffffffp+16380L
FLT_EPSILON	Difference between 1.0 and the least value greater than 1.0 that is representable as a float.	0x1.p-23F
DBL_EPSILON	Difference between 1.0 and the least value greater than 1.0 that is representable as a double.	0x1.p-52
LDBL_EPSILON	Difference between 1.0 and the least value greater than 1.0 that is representable as a long double.	0x8.p-66L
FLT_MIN	Minimum normalised positive float.	0x1.p-126F
DBL_MIN	Minimum normalised positive double.	0x1.p-1022
LDBL_MIN	Minimum normalised positive long double.	0x8.p-16385L

Rationale

This set of constants provides useful information regarding the underlying architecture of the implementation.

Question 4: What are the values associated with the following constants (optionally specified in the <limits.h> header file)?

Response

Macro Name	Meaning	Value
ARG_MAX	Maximum length of argument to the exec functions including the environment data.	10240
ATEXIT_MAX	Maximum number of functions that may be registered with atexit().	32
CHILD_MAX	Maximum number of processes per user ID.	25
IOV_MAX	Maximum number of iovec structures that one process has available for use with readv() or writev().	16
LINK_MAX	Maximum number of links to a single file.	1000
MAX_CANON	Maximum number of bytes in a terminal canonical input line.	256
MAX_INPUT	Maximum number of bytes for which space will be available in a terminal input queue.	512
NAME_MAX	Maximum number of bytes in a filename (not including terminating null).	255
OPEN_MAX	Maximum number of open files that one process can have open at any one time.	60
PAGESIZE	Size of a page in bytes.	4096
PAGE_SIZE	Same as PAGESIZE. If either PAGESIZE or PAGE_SIZE is defined, the other will be defined with the same value.	4096
PATH_MAX	Maximum number of bytes in a pathname (including the terminating null).	1024
PIPE_BUF	Maximum number of bytes that is guaranteed to be atomic when writing to a pipe.	5120
STREAM_MAX	Number of streams that one process can have open at one time.	64
TZNAME_MAX	Maximum number of bytes supported for the name of a time zone.	8

Rationale

Each of these limits can vary within bounds set by System Interfaces and Headers, Issue 4, Version 2. The minimum permitted value is specified in Chapter 4, <limits.h>.

Question 5: What are the values associated with the following constants specified in the <limits.h> header file?

Response

For each line in the table below, add the minimum and maximum values for your implementation. These values may be stated to be "Unlimited" if your implementation does not impose a limit. The minimum should be the smallest value that is returned from sysconf(), pathconf(), or fpathconf(), or as defined in <limits.h>. The maximum value should be the largest value that is returned from sysconf(), pathconf(), or fpathconf().

Macro Name	Meaning	Minimum	Maximum
BC_BASE_MAX	Maximum ibase and obase values allowed by the bc utility.	99	INT_MAX
BC_DIM_MAX	Maximum number of elements permitted in an array by the bc utility.	2048	2048
BC_SCALE_MAX	Maximum scale value allowed by the bc utility.	99	99
BC_STRING_MAX	Maximum length of a string constant accepted by the bc utility.	1000	1000
COLL_WEIGHTS_MAX	Maximum number of weights that can be assigned to an entry of the LC_COLLATE order keyword in the locale definition file.	2	14
EXPR_NEST_MAX	Maximum number of expressions that can be nested within parentheses by the expr utility.	32	32
LINE_MAX	Maximum length in bytes including the trailing newline of a utility's input line when the utility is described as processing text files.	2048	2048
NGROUPS_MAX	Maximum number of simultaneous supplementary group IDs per process.	16	16
RE_DUP_MAX	Maximum number of repeated occurrences of a regular expression permitted when using interval notation.	255	5100

Rationale

Each of these limits can vary within bounds set by System Interfaces and Headers, Issue 4, Version 2. The minimum value that a limit can take on any conforming system is given in the corresponding _POSIX_ or _POSIX2_ value. A specific conforming implementation may provide a higher minimum value than this and the maximum value that it provides can differ from the minimum. Some conforming implementations may provide a potentially infinite value as the maximum, in which case the value is considered to be indeterminate. The minimum value must always be definitive since the _POSIX_ or _POSIX2_ value provides a known lower bound for the range of possible values.

Question 6: What are the values associated with the following numerical constants specified in the <limits.h> header file?

Response

Macro Name	Meaning	Value
CHAR_BIT	Number of bits in a char.	8
CHAR_MAX	Maximum value of a char.	127
INT_MAX	Maximum value of an int.	2147483647
LONG_BIT	Number of bits in a long int.	32
LONG_MAX	Maximum value of a long int.	2147483647
MB_LEN_MAX	Maximum number of bytes in a character, for any supported locale.	5
SCHAR_MAX	Maximum value of a signed char.	127
SHRT_MAX	Maximum value of a short.	327767
SSIZE_MAX	Maximum value of an object of type ssize_t.	2147483647
UCHAR_MAX	Maximum value of an unsigned char.	255
UINT_MAX	Maximum value of an unsigned int.	4294967295
ULONG_MAX	Maximum value of an unsigned long int.	4294967295
USHRT_MAX	Maximum value of an unsigned short int.	65535
WORD_BIT	Number of bits in a word or int.	32

Rationale

This set of constants provides useful information regarding the underlying architecture of the implementation.

Question 7: What are the values associated with the following numerical constants specified in the <stdio.h> header file?

Response

Macro Name	Meaning	Value
FILENAME_MAX	Maximum size in bytes of the longest filename string that the implementation guarantees can be opened.	1024
FOPEN_MAX	Number of streams which the implementation guarantees can be open simultaneously.	60
L_ctermid	Maximum size of character array to hold ctermid() output.	9
L_tmpnam	Maximum size of character array to hold tmpnam() output.	25
TMP_MAX	Minimum number of unique filenames generated by tmpnam(), which is the maximum number of times an application can call tmpnam() reliably.	17576

Rationale

This set of constants provide useful information about the implementation.

1.1.4 Error Conditions

Question 8: Which of the following option errors listed in System Interfaces and Headers, Issue 4, Version 2 are detected in the circumstances specified?

Response

Function	Error	Detected
access()	EINVAL	Yes
	ENAMETOOLONG	Yes
	ETXTBSY	No
acos()	EDOM	Yes
acosh()	EDOM	Yes
asin()	EDOM	Yes
	ERANGE	No
asinh()	EDOM	Yes
atan()	EDOM	Yes
	ERANGE	No
atanh()	EDOM	Yes
atan2()	EDOM	Yes
	ERANGE	No
brk()	EAGAIN	Yes
	ENOMEM	Yes
catclose()	EBADF	No
	EINTR	No
catgets()	EBADF	No
	EINTR	No
	EINVAL	No
	ENOMSG	No
catopen()	EACCES	No
	EMFILE	No
	ENAMETOOLONG	No
	ENFILE	No
	ENOENT	No
	ENOMEM	No
	ENOTDIR	No
cbrt()	EDOM	Yes
ceil()	EDOM	Yes
cfsetispeed()	EINVAL	Yes
cfsetospeed()	EINVAL	Yes
chdir()	ENAMETOOLONG	Yes
chmod()	EINTR	Yes
	EINVAL	No
	ENAMETOOLONG	Yes
chown()	EINTR	Yes
	EINVAL	No
	EIO	Yes
	ENAMETOOLONG	Yes
chroot()	ENAMETOOLONG	Yes
close()	EIO	Yes
closedir()	EBADF	Yes
	EINTR	Yes
cos()	EDOM	Yes
	ERANGE	No
cosh()	EDOM	Yes
endpwent()	EIO	No
erf()	EDOM	Yes
	ERANGE	No
erfc()	EDOM	Yes
	ERANGE	No
exec	ENAMETOOLONG	Yes
	ENOMEM	Yes
	ETXTBSY	No
exp()	EDOM	Yes
	ERANGE	No
expm1()	EDOM	Yes
	ERANGE	Yes
fabs()	EDOM	Yes
	ERANGE	No
fattach()	EINVAL	Yes
	ENAMETOOLONG	Yes
fchdir()	EINTR	Yes
	EIO	Yes
fchmod()	EINTR	Yes
	EINVAL	No
fchown()	EINVAL	Yes
	EIO	Yes
	EINTR	Yes
fclose()	ENXIO	Yes
fcntl()	EDEADLK	Yes
fdetach()	ENAMETOOLONG	Yes
fdopen()	EBADF	Yes
	EINVAL	Yes
	EMFILE	No
	ENOMEM	Yes
fflush()	ENXIO	Yes
fgetc()	ENOMEM	Yes
	ENXIO	No
fgetpos()	EBADF	Yes
	ESPIPE	Yes
fgetwc()	ENOMEM	Yes
	ENXIO	No
	EILSEQ	Yes
fileno()	EBADF	Yes
floor()	EDOM	Yes
fmod()	EDOM	Yes
	ERANGE	No
fopen()	EINVAL	Yes
	EMFILE	Yes
	ENAMETOOLONG	Yes
	ENOMEN	No
	ETXTBSY	No
fork()	ENOMEM	Yes
fpathconf()	EBADF	Yes
	EINVAL	Yes
fprintf()	EINVAL	No
	EILSEQ	No
	ENOMEM	Yes
fputc()	ENOMEM	Yes
	ENXIO	Yes
fputwc()	ENOMEM	Yes
	ENXIO	Yes
	EILSEQ	Yes
freopen()	EINVAL	Yes
	ENAMETOOLONG	Yes
	ENOMEM	Yes
	ENXIO	No
	ETXTBSY	No
frexp()	EDOM	Yes
fscanf()	EILSEQ	Yes
	EINVAL	No
	ENOMEM	Yes
	ENXIO	No
fsetpos()	EBADF	Yes
	ESPIPE	Yes
fstat()	EOVERFLOW	Yes
ftok()	ENAMETOOLONG	Yes
ftw()	EINVAL	No
	ENAMETOOLONG	Yes
getcwd()	EACCES	Yes
	ENOMEM	No
getgrent()	EINTR	No
	EIO	No
	EMFILE	Yes
	ENFILE	Yes
getgrgid()	EIO	No
	EINTR	No
	EMFILE	No
	ENFILE	No
getgrnam()	EIO	No
	EINTR	No
	EMFILE	No
	ENFILE	No
getitimer()	EINVAL	Yes
getlogin()	EMFILE	Yes
	ENFILE	Yes
	ENXIO	No
getpass()	EINTR	No
	EIO	No
	EMFILE	Yes
	ENFILE	Yes
	ENXIO	No
getpgid()	EINVAL	No
getpwent()	EIO	No
	EMFILE	Yes
	ENFILE	Yes
getpwnam()	EIO	No
	EINTR	No
	EMFILE	No
	ENFILE	No
getpwuid()	EIO	No
	EINTR	No
	EMFILE	No
	ENFILE	Yes
grantpt()	EBADF	No
	EINVAL	No
	EACCES	No
hcreate()	ENOMEM	Yes
hsearch()	ENOMEM	No
hypot()	EDOM	Yes
	ERANGE	No
iconv()	EBADF	No
iconv_close()	EBADF	Yes
iconv_open()	EMFILE	Yes
	ENFILE	Yes
	ENOMEM	No
	EINVAL	Yes
isatty()	EBADF	Yes
	ENOTTY	No
j0()	EDOM	Yes
	ERANGE	No
j1()	EDOM	Yes
	ERANGE	No
jn()	EDOM	Yes
	ERANGE	No
lchown()	EIO	Yes
	EINTR	Yes
	ENAMETOOLONG	Yes
ldexp()	EDOM	Yes
	ERANGE	No
lgamma()	EDOM	Yes
	ERANGE	No
link()	ENAMETOOLONG	Yes
lockf()	EAGAIN	Yes
	EDEADLK	Yes
	EINVAL	Yes
	ENOLCK	No
	EOPNOTSUPP	No
log()	EDOM	Yes
	ERANGE	No
log10()	EDOM	Yes
	ERANGE	No
log1p()	EDOM	Yes
	ERANGE	Yes
logb()	EDOM	Yes
lstat()	ENAMETOOLONG	Yes
	EOVERFLOW	No
mblen()	EILSEQ	No
mbstowcs()	EILSEQ	No
mbtowc()	EILSEQ	Yes
mkdir()	ENAMETOOLONG	Yes
mkfifo()	ENAMETOOLONG	Yes
mknod()	ENAMETOOLONG	Yes
mmap()	EINVAL	Yes
modf()	EDOM	Yes
	ERANGE	No
mprotect()	EAGAIN	Yes
msync()	EBUSY	No
nextafter()	EDOM	Yes
nftw()	ELOOP	Yes
	EMFILE	Yes
	ENAMETOOLONG	Yes
	ENFILE	Yes
open()	EAGAIN	Yes
	EINVAL	No
	ENAMETOOLONG	Yes
	ENOMEM	Yes
	ETXTBSY	No
opendir()	EMFILE	Yes
	ENAMETOOLONG	Yes
	ENFILE	Yes
pathconf()	EACCES	Yes
	EINVAL	Yes
	ENAMETOOLONG	Yes
	ENOENT	Yes
	ENOTDIR	Yes
popen()	EMFILE	Yes
	EINVAL	No
pow()	EDOM	Yes
	ERANGE	No
printf()	ENOMEM	Yes
putenv()	ENOMEM	Yes
pututxline()	EPERM	No
read()	ENXIO	No
readdir()	EBADF	Yes
	ENOENT	Yes
readlink()	EACCES	Yes
	ENAMETOOLONG	Yes
readv()	EINVAL	Yes
	ENXIO	No
realpath()	ENAMETOOLONG	Yes
	ENOMEM	No
regcmp()	ENOMEM	Yes
remainder()	EDOM	Yes
rename()	ENAMETOOLONG	Yes
	ETXTBSY	No
rint()	EDOM	Yes
rmdir()	ENAMETOOLONG	Yes
sbrk()	EAGAIN	Yes
	ENOMEM	Yes
scalb()	EDOM	Yes
setitimer()	EINVAL	Yes
setpriority()	EPERM	Yes
	EACCES	No
setpwent()	EIO	No
	EMFILE	Yes
	ENFILE	Yes
setrlimit()	EINVAL	Yes
setvbuf()	EBADF	Yes
sigaction()	EINVAL	Yes
sigaddset()	EINVAL	Yes
sigdelset()	EINVAL	Yes
sigismember()	EINVAL	Yes
signal()	EINVAL	Yes
sin()	EDOM	Yes
	ERANGE	No
sinh()	EDOM	Yes
	ERANGE	No
sqrt()	EDOM	Yes
stat()	ENAMETOOLONG	Yes
	EOVERFLOW	Yes
statvfs()	ENAMETOOLONG	Yes
strcoll()	EINVAL	No
strdup()	ENOMEM	Yes
strerror()	EINVAL	No
strtod()	EINVAL	Yes
strtol()	EINVAL	Yes
strtoul()	EINVAL	Yes
strxfrm()	EINVAL	No
symlink()	ENAMETOOLONG	Yes
system()	ECHILD	No
tan()	EDOM	Yes
	ERANGE	No
tanh()	EDOM	Yes
	ERANGE	No
tcdrain()	EIO	Yes
tcflow()	EIO	Yes
tcflush()	EIO	Yes
tcsendbreak()	EIO	Yes
tcsetattr()	EIO	Yes
tmpfile()	EMFILE	Yes
	ENOMEM	Yes
truncate()	ENAMETOOLONG	Yes
ttyname()	EBADF	Yes
	ENOTTY	No
ungetwc()	EILSEQ	Yes
unlink()	ENAMETOOLONG	Yes
	ETXTBSY	No
unlockpt()	EBADF	Yes
	EINVAL	Yes
usleep()	EINVAL	No
utime()	ENAMETOOLONG	Yes
utimes()	ENAMETOOLONG	Yes
wait3()	ECHILD	Yes
wcscoll()	EINVAL	No
wcstod()	EINVAL	Yes
wcstol()	EINVAL	Yes
wcstombs()	EILSEQ	No
wcstoul()	EINVAL	Yes
wcsxfrm()	EINVAL	No
write()	EINVAL	Yes
	ENXIO	Yes
writev()	EINVAL	Yes
	ENXIO	Yes
y0()	EDOM	Yes
	ERANGE	No
y1()	EDOM	Yes
	ERANGE	No
yn()	EDOM	Yes
	ERANGE	No

Rationale

Each of the above error conditions is marked as optional in System Interfaces and Headers, Issue 4, Version 2 and an implementation may return this error in the circumstances specified or may not provide the error indication.

1.1.5 Mathematical Interfaces

Question 9: What format of floating-point numbers is supported by this implementation?

Response

IEEE floating point format.

Rationale

Most implementations support IEEE floating point format either in hardware or software. Some implementations support other formats with different exponent and mantissa accuracy. These differences need to be defined.

1.1.6 Data Encryption

Question 10: Are the optional data encryption interfaces provided?

Response

Function	Provided
crypt()	Yes
encrypt()	Yes
setkey()	Yes

Export Restrictions:

Due to export restrictions on the encoding algorithm, the version of encrypt() shipped outside the USA has restricted capabilities in that it performs encoding but no decoding. Encrypt() will fail when called to perform decoding in an international environment by setting errno to ENOSYS.

Rationale

Normally, an implementation will either provide all three of these routines or will provide none of them at all. If the routines are not provided, then the implementation must provide a dummy interface which always raises an ENOSYS error condition.

It is also possible that the implementation of the encrypt() function may be affected by export restrictions, in which case, the restrictions should be documented here.

For example, historical implementations have supplied all three of these routines outside the U.S.A., but due to export restrictions on the decoding algorithm, a dummy version of encrypt() is provided that does encoding but no decoding.

1.2 Process Handling

1.2.1 Process Generation

Question 11: Which file types (regular, directory, FIFO, special, and so on) are considered to be executable?

Response

Only regular file types may be executed.

Rationale

The [EACCES] error associated with exec functions occurs in circumstances when the implementation does not support execution of files of the type specified. A list of these file types needs to be provided.

1.3 File Handling

1.3.1 Access Control

Question 12: What file access control mechanisms does the implementation provide?

Response

Standard access control is provided.

Rationale

System Interfaces and Headers, Issue 4, Version 2 notes that implementations may provide additional or alternate file access control mechanisms, or both.

1.3.2 Files and Directories

Question 13: Are any additional or alternate file access control mechanisms implemented that could cause fstat() or stat() to fail?

Response

No

Rationale

System Interfaces and Headers, Issue 4, Version 2 notes that there could be an interaction between additional and alternate access controls and the success of fstat() and stat(). This would suggest that an implementation can allow access to a file but not allow the process to gain information about the status of the file.

1.3.3 Formatting Interfaces

Question 14: Does the printf() function produce character string representations for Infinity and NaN to represent the respective values?

Response

Yes

Rationale

This behaviour is often provided on systems with mathematical functions that produce these results.

1.4 Internationalised System Interfaces

1.4.1 Coded Character Sets

Question 15: What coded character sets are supported by the implementation?

Response

The implementation uses the Extended UNIX Code (EUC) mechanism for codeset support. This enables the implementation to support up to four single or multibyte code sets concurrently at both process and file levels. This always includes the portable character set. This coding scheme conforms to the International Standard ISO 2022. We also support UTF-8 encoded Unicode (ISO 10646).

Rationale

System Interface Definitions, Issue 4, Version 2 states that conforming implementations support one or more coded character sets, and that each of these includes the portable character set.

Question 16: What is the implementation's underlying internal codeset?

Response

ISO 8859-1:1987

Rationale

It is useful to be aware of the underlying codeset of the implementation.

1.5 STREAMS

Question 17: What networking services or other character-based I/O device types are implemented using STREAMS?

Response

Terminals, Pseudo-Terminals, Sockets, FIFOs

Rationale

System Interfaces and Headers, Issue 4, Version 2, defines that STREAMS provide a uniform mechanism for implementing networking services and other character-based I/O. However, the specification does not mandate which device or file types should be STREAMS-based. Although applications are discouraged from making assumptions in this area, it may be that certain applications are sensitive to whether interfaces such as getmsg() and putmsg(), for example, are supported on specific device or file types.

1.6 Pseudo-Terminals

1.6.1 Master close

Question 18: Does closing the master side of a pseudo-terminal flush all queued input and output?

Response

No

Rationale

The behaviour of a conforming implementation in this area is not mandated in the specification and needs to be defined.

1.6.2 Slave close

Question 19: Does closing the slave side of a pseudo-terminal cause a zero-length message to be sent to the master?

Response

Yes

Rationale

The behaviour of a conforming implementation in this area is not mandated in the specification and needs to be defined.

1.6.3 Naming Conventions

Question 20: What naming conventions are associated with the master side of pseudo-terminal devices?

Response

/dev/ptmx, this is a STREAMS multiplexor device.

Rationale

This information is not specified in System and Interfaces, Issue 4, Version 2, and needs to be be defined.

1.7 Polling

Question 21: What types of file can be polled?

Response

Regular files, Terminals, Pseudo-Terminals, STREAMS, Sockets, FIFOs, Pipes

Rationale

Conformance requires that the poll() function supports regular files, terminals, pseudo-terminals, STREAMS, sockets, FIFOs and pipes. The behaviour of poll() with regards to other file types needs to be defined.

1.8 Alternate Stacks

Question 22: What allocation routine(s) is provided for creating alternate stack areas?

Response

calloc(), malloc(), valloc(), memalign(), realloc()

Rationale

Conformance requires that an implementation supports alternate signal stacks. The APPLICATION USAGE section of the sigaltstack() entry describes one method using malloc() to perform this function. However, the specification does not guarantee that malloc'ed space can be used in this way, nor does it define a specific alternate stack allocation routine.

1.9 Signal Codes

Question 23: Which of the following si_code values may be generated?

Response

Signal	Code	Generated?
SIGILL	ILL_ILLOPC	Yes
	ILL_ILLOPN	Yes
	ILLL_ILLADR	No
	ILL_ILLTRP	No
	ILL_PRVOPC	No
	ILL_PRVREG	No
	ILL_COPROC	No
	ILL_BADSTK	No
SIGFPE	FPE_INTDIV	Yes
	FPE_INTOVF	Yes
	FPE_FLTDIV	Yes
	FPE_FLTOVF	Yes
	FPE_FLTUND	Yes
	FPE_FLTRES	Yes
	FPE_FLTINV	Yes
	FPE_FLTSUB	Yes
SIGSEGV	SEGV_MAPERR	Yes
	SEGV_ACCERR	Yes
SIGBUS	BUS_ADRALN	Yes
	BUS_ADRERR	No
	BUS_OBJERR	Yes
SIGTRAP	TRAP_BRKPT	Yes
	TRAP_TRACE	Yes
SIGCHLD	CLD_EXITED	Yes
	CLD_KILLED	Yes
	CLD_DUMPED	Yes
	CLD_TRAPPED	Yes
	CLD_STOPPED	Yes
	CLD_CONTINUED	Yes
SIGPOLL	POLL_IN	Yes
	POLL_OUT	Yes
	POLL_MSG	Yes
	POLL_ERR	Yes
	POLL_PRI	Yes
	POLL_HUP	Yes

Rationale

An Internationalised System Calls and Libraries Extended conformant system may contain limitations that prevent some of the above values from being generated.

1.10 Set Process Group ID

Question 24: Does the setpgrp() function create a new session?

Response

Yes

Rationale

It is unspecified whether or not a successful call to the setpgrp() function will cause a new session to be created.

Question 25: Does the implementation provide a signal, when delivered to a process, that generates a core file?

Response

Yes

Rationale

Implementation-dependent abnormal termination actions, such as creation of a core file, may occur.

Copyright © 2003 Caldera International d/b/a The SCO Group
All rights reserved.

2. Change History

---

 Copyright © 1998-2011 The Open Group. All Rights Reserved.

 The Open Group and Boundaryless Information Flow are trademarks and UNIX is a registered trademark of The Open Group in the United States and other countries. All other trademarks are the property of their respective owners.

---

[ Home ] [ The Open Brand ] [ Conformance Statement Library ] [ Search Conformance Statements ] [ Send Feedback ]