Date: Tue, 27 Oct 1998 23:38:03 -0500 (EST)
Message-Id: <199810280438.XAA10944@indy.delorie.com>
From: DJ Delorie <dj AT delorie DOT com>
To: djgpp AT delorie DOT com
In-reply-to: <000c01be022b$9b0b2240$69ee8081@bhodge>
Subject: Re: DJGPP 2.02 Beta 981027
Reply-To: djgpp AT delorie DOT com


> Coolness, what features can we look forward to in the new release?

Mostly bug fixes, performance, stability, that sort of thing.
Major performance boost for FILE* reads.  LFN is much better,
and now the default. printf() auto-flushes.  New malloc.

djdev202.zip includes, in the Knowledge Base, a "What's Changed in
2.02" section.  Here's a copy of the texinfo source of it:


@node Changes in 2.02, , Changes in 2.01, What Changed
@section Changes in 2.02

Here is a list of changes from DJGPP V2.01 to V2.02

@file{tests/libclink} no longer looks for DJGPP V1.

@file{emu387.dxe} has been rebuilt to reflect fpatan fixes.
@findex emu387

@file{mcount.c} no longer writes to the mono monitor.  This had caused
seg faults when profiling.
@findex mcount.c

You can now ``symlink'' to files with 8 character base names (like
12345678.exe).  Previously, @file{stub.asm} assumed al was zero when it
wasn't always that way.
@pindex stubedit
@findex stub.asm

@code{rawclock()} now actually returns the number of ticks since
midnight, not since program start.
@findex rawclock

@file{libc.tex} now uses ``Master Index'' instead of ``Index'' due
to conflicts with the node for @code{index}.
@findex libc.tex

@code{srand()} is now documented.
@findex srand

@code{memchr()} is now 8-bit clean.
@findex memchr

Documentation for @code{delay()} includes the header file.
@findex delay

@code{__dpmi_simulate_real_mode_procedure_iret()} is now properly
documented as DPMI call 0x0302.
@findex __dpmi_simulate_real_mode_procedure_iret()

@file{edebug} now unassembles @code{scasb} correctly.
@pindex edebug

Documentation for @code{qsort()} has correct prototypes
in the examples.
@findex qsort

@file{info/dir.txi} includes an entry for the FAQ.
@findex info/dir.txi

Functions in @file{conio.h} now support DOS/V's virtual screens.
@findex conio.h

@file{edebug} and other debuggers using the common debugging framework
will properly close files when they terminate the debuggee.
This allows to restart the debuggee without losing handles.
@findex dbgcom.c
@pindex edebug

@code{fnsplit()} now handles files like .emacs properly.
@findex fnsplit

@code{glob()} now handles files like .emacs properly.  This will fix
wildcard expansion on new programs.
@findex glob

@file{go32-v2.exe} properly reverts the interrupt table before running
the child process.
@pindex go32-v2

@code{_lfn_gen_short_name} no longer crashes in rare cases.
@findex _lfn_gen_short_name

@code{s[f]_ceil} and @code{s[f]_floor} functions in @file{libm.a} no
longer crash when emulated.
@findex s_ceil
@findex sf_ceil
@findex s_floor
@findex sf_floor

@code{fixpath} and @code{readdir} support multibyte file names.
@findex fixpath
@findex readdir
@cindex multibyte

You may now @code{set 387=y} to avoid the auto-detection.

@code{opendir("/")} works.
@findex opendir

@code{rename("x", "x/y")} will now fail when "x" is a directory.
@findex rename

@code{sigprocmask} and @code{sigpending} actually do something now.
All of the POSIX signal functions are now documented.
@findex sigprocmask
@findex sigpending
@findex sigaddset
@findex sigdelset
@findex sigemptyset
@findex sigfillset
@findex sigismember

@code{_truename} didn't use @code{putpath} and therefore didn't support
the @code{/dev/} style of standard devices.  @code{stat} also failed on
devices in some cases.  We think this is the cause of @file{bash}
failing on redirection to @file{/dev/null}.
@findex _truename
@findex putpath
@findex stat
@pindex bash

Documentation for FSEXT-enabled functions now mentions that they are
FSEXT-enabled, and how to avoid that.
@findex _close
@findex _creat
@findex _open
@findex _read
@findex _write

Documentation for @code{bioscom} now documents (and uses) the
@code{port} parameter.
@findex bioscom

@code{ffs()} returns the correct value.
@findex ffs

Various fixes and optimizations to @code{_use_lfn()}.
@findex _use_lfn

Return codes of @code{lock()} and @code{unlock()} are correct.
@findex lock
@findex unlock

Added @code{memicmp()}
@findex memicmp

@code{tmpfile()} always opens the temp file in binary mode.
@findex tmpfile

New @code{statfs} returns the actual size of a CD-ROM instead of the
default returned by MSCDEX.
@findex statfs

@file{gxx} won't try to link extra libraries if you specify @code{-c} or
@code{-S} on the command line.
@pindex gxx

@code{_dos_findfirst()} takes a @code{const char *} instead of
@code{char *}.
@findex _dos_findfirst

@code{rename()} handles extra error codes some network redirectors return.
@findex rename

@code{ctime} handles dumped emacs and DOS-style @code{$TZDIR}.
@findex ctime

@code{pathconf()} returns actual values for NAME_MAX and PATH_MAX, rather
than defaults.  Therefore, it can be used to detect whether a given disk
supports long file names.
@findex pathconf

The return values for @code{_go32_dpmi_lock_data} and
@code{_go32_dpmi_lock_code} are now documented.
@findex _go32_dpmi_lock_code
@findex _go32_dpmi_lock_data

The initial stack is aligned to an 8-byte boundary.

@code{getkey} and @code{getxkey} allow you to do @kbd{ALT-keypad-224}.
@findex getkey
@findex getxkey

@code{popen} selects temporary files correctly now.
@findex popen

@code{pclose} returns the status of the program it read from.
@findex pclose

@code{stubedit} won't let you set a transfer buffer bigger than 63k.
If the stub detects a zero there, it assumes 63k.
@pindex stubedit

The stub now returns error codes in the 100..110 range when it has a
startup failure.

@code{system} works if @code{$COMSPEC} isn't set.
@findex system

@code{struct lconv} includes @code{mon_thousands_sep}.
@findex lconv

@code{scanf} works with long double formats like @code{%LG}.
@findex scanf

@code{_osmajor} and @code{_osminor} are now set at startup.
@findex _osmajor
@findex _osminor

@code{write()} now calls FSEXT for ``text'' files.
@findex write

@file{assert.h} causes no more warnings with some gcc options.
@findex assert

@code{access()} now thinks devices are read/write, instead of read-only,
regardless of what the device actually supports.
@findex access

New function @code{stackavail()} in @file{stdlib.h}
@findex stackavail

@file{echo}, @file{split} and @file{merge} are now called @file{djecho},
@file{djsplit} and @file{djmerge} to avoid clashes with GNU's
@file{echo}, @file{split} and @file{merge} programs.
@pindex echo
@pindex split
@pindex merge

@code{tzname[]} bug fixed; @code{tzname} no longer a function stub (!)
@findex tzname

More @code{djasm} improvements.
@pindex djasm

Added @code{_v2_prog_type} to @file{sys/system.h} and extended @code{symlink} with it.
@findex _v2_prog_type
@findex symlink

@code{fread} and @code{fwrite} are documented as returning less than the
requested size, rather than -1 on error, since they never return
negative values.
@code{fread}
@code{fwrite}

@code{stub} forcefully closes file descriptors 18 and 19, so that the
stub and the DPMI server have enough handles to open the swap file and
read the COFF information of the DJGPP executable.
@findex stub

Minor fixes to @code{popen}'s list of pending commands.
@findex popen

@code{tempnam} returns allocated memory, not a pointer to static space,
and does not ignore its arguments @var{tmpdir} and @var{prefix} anymore.
@findex tempnam

@code{difftime} can return negative numbers now.
@findex difftime

@code{gethostname} removes trailing spaces now.
@findex gethostname

@code{insque} works with empty lists now.
@findex insque

@code{sync} also calls @code{_flush_disk_cache()}.
@findex sync
@findex _flush_disk_cache

@code{printf} flushes stdout every time, unless it is redirected to a
file, to reduce programmer confusion.  Note that @code{fprintf} and
@code{vprintf} don't do this.
@findex printf

@code{exit} calls destructors *before* it closes all open files.
@findex exit

The code that reads blocks of data for FILE* streams has a new ``slow
start'' algorithm that's designed to adjust the read size for both
programs that seek a lot reading small amounts of data, and programs
that don't seek but read lots of data.

@code{mktime} handles the DST change hour gracefully
@findex mktime

@code{enable} and @code{disable} don't use DPMI (sti/cli are less buggy!)
@findex enable
@findex disable

The command line parser can handle response files generated by
@samp{find -print0}.

@code{malloc} and @code{free} have been replaced by a more efficient
(and not BSD) algorithm.  The source archive still contains
@file{bsdmallo.c} and a new even faster (but more wasteful) version in
@file{fmalloc.c}.
@findex malloc
@findex free

@code{getmntent} better supports Windows NT and doesn't cause DOS to
return stale data when the disk is changed.  Removable media such as JAZ
drives are no longer reported when the drive is empty.  Disks written by
Windows 9X are no longer reported with bogus names taken from LFN
directory entries (which have their volume label attribute bit set).
@findex getmntent

File names which begin with @file{/dev/} or @file{x:/dev/} (where
@file{x:} is any drive letter) are treated specially by all
file-oriented library functions.  @file{/dev/tty} is converted to
@file{con} and @file{/dev/null} is converted to @file{nul}, to make
porting of Unix programs easier.  Other names get the drive letter and
the @file{/dev/} prefix removed, so e.g. @file{d:/dev/lpt1} is converted
to @file{lpt1}.  This is because DOS handles device names inconsistently
unless they are devoid of the @file{x:/dev/} prefix.  File names of the
form @file{/dev/x/foo} are converted to @file{x:/foo}; this allows to
use Unix-style path names which always begin with a forward slash.

The startup code now masks @strong{all} numeric exceptions when a
floating-point hardware is present.  This means that illegal operations
like @code{sqrt(-1)} will now return @code{NaN} (Not-a-number) rather
than raise the @code{SIGFPE} signal.

@code{fstat} is now more reliable when called from programs under a
debugger, does not report bogus results on NT, reports files on the A:
drive correctly, and correctly reports the write access bit on
Windows9X.
@findex fstat

@code{stat} is now more reliable on NT and with character devices.
@findex stat


@file{djgpp.env} is read before command line arguments are expanded,
so that they'll honor the @code{LFN} setting in @file{djgpp.env}.
@pindex djgpp.env

The @code{sys_siglist[]} array is now available with the names of all
the signals, and the function @code{psignal} can be used to print
messages which include the signal names.
@findex _sys_siglist
@findex psignal

The new functions @code{basename} and @code{dirname} can be used to
extract directory and basename parts from file names.
@findex basename
@findex dirname

_my_* work for selectors with the high bit set

The new function @code{_creatnew} creates a file only if it didn't
already exist.
@findex _creatnew

@code{tmpfile} makes sure it opens a file which is not and will not be
used by any other program.  It does so by repeatedly calling
@code{tmpnam} until @code{_creatnew} succeeds to create a file which
didn't exist before.  The temporary file is opened in DENY_ALL mode, so
that no other process can neither read from nor write to it.  In case of
a failure, @code{tmpfile} does not leak memory anymore.

Previously, there was a small probability that a call to @code{open}
with the @code{O_CREAT} and @code{O_EXCL} bits set would succeed even
though the named file already existed (created by another process).
This window is now closed.
@findex open

@code{mkstemp} now makes sure the file it creates is unique even if
other processes can create files in the same directory, and opens the
file in DENY_ALL mode.
@findex mkstemp

@code{__FSEXT_alloc_fd} now duplicates the initial file handle instead
of reopening the NUL device on each call.  Thus, it is no longer limited
by the value of the @samp{FILES=} parameter on your @file{CONFIG.SYS},
but can allocate up to 254 handles (which is the maximum number allowed
by DOS).
@findex __FSEXT_alloc_fd
@findex tmpfile

Stack dumps will now fit the screen, instead of being limited to 10
frames.

@code{rename} will properly fail if the file doesn't exist.
@findex rename

@code{system} is now compliant with ANSI C Standard when its argument is
a NULL pointer.
@findex system

Previously, at program startup @code{errno} would retain whatever value
it was assigned by the DJGPP startup code.  ANSI C Standard requires
@code{errno} to be zero at the beginning of @code{main}, so it is now
reset to zero before calling @code{main}.
@findex errno

The stack dump printed when a program is aborted or crashes was
improved.  First, it is no longer limited to 10 stack frames.  When the
standard error stream is connected to the console, the maximum number of
frames is computed so that the stack dump will use all the available
screen area, without scrolling off the visible part.  For example, at
50-line display, you can have as much as 36 stack frames printed and
still see the registers and selectors dump, and the exception mnemonic
that preceed it.

When stderr is not the console, the number of printed stack frames is
only limited by the stack size of the running program.

In addition, the top and bottom of both the application code stack and
the locked stack used for processing exceptions, are printed.  This
makes detection of stack overflows and other stack-related atrocities
much easier.
@cindex stack dump, when aborted or crashed

The @code{_doserrno} global variable is now provided.  Whenever a DOS
call fails, this variable gets assigned the value of the error code
returned by the failed DOS function.
@findex _doserrno

The header @file{sys/dxe.h} now works with C++.

@code{fflush} now conforms to ANSI C standard when called with a
@code{NULL} argument.
@findex fflush

@code{__dpmi_yield} doesn't wedge the DOS box on some versions of
Windows 9X anymore.
@findex __dpmi_yield

@code{outports*} takes a const buffer
@findex outportsb
@findex outportsw
@findex outportsl

@code{glob} will initalize @code{*pglob} when appropriate.
@findex glob

The conversion functions @code{ecvtbuf}, @code{ecvt}, @code{fcvtbuf},
@code{fcvt}, and @code{gcvt} are now provided.
@findex fcvtbuf
@findex fcvt
@findex ecvtbuf
@findex ecvt
@findex gcvt

The startup code now recognizes an additional flag
_CRT0_FLAG_KEEP_QUOTES, which forces it to retain quote characters in
command-line arguments.
@vindex _crt0_startup_flags, the _CRT0_FLAG_KEEP_QUOTES flag

@code{redir} invokes the subsidiary program in a way that preserves the
command-line expansion.  Thus, you can now use @code{redir} to invoke
programs whose command lines include quote characters and file name
wildcards, and they will work as expected.

Also, @code{redir} no longer crashes if the subsidiary program was
interrupted or crashed, and reports the run time with millisecond
resolution.
@cindex redir

The exit code now makes sure the timer interrupt is left pointing to its
original handler, even if the program is aborted, or exits without
disabling timers.  @code{__djgpp_exception_toggle} now toggles the timer
interrupt handler as well, if it was hooked by SIGALRM/SIGPROF handler
or by @code{setitimer}.  Thus, timers, periodic alarms, and profiling
will be disabled inside blocks that call @code{__djgpp_exception_toggle}.
@findex __djgpp_exception_toggle
@findex setitimer
@cindex SIGALRM
@cindex SIGPROF
@cindex profiling

The functions of the @code{printf} family now always print a negative
sign for a negative floating-point argument, even when the conversion
generated no significant digits (i.e. only zeros are printed), if the
format specifier requests an explicit sign, like in "%+f".  A special
NaN value will also be printed with a sign in these cases (e.g. try
printing @code{-nanf()}).
@findex printf
@findex fprintf
@findex sprintf
@findex _doprnt
@findex nanf

@code{select} now correctly zeroes out all the bits in the @code{fd_set}
arguments when it returns due to expired timeout.
@findex select