cvs-src summary for October 11-18

From: Mark Johnston <mjohnston_at_skyweb.ca>
Date: Mon, 18 Oct 2004 14:47:37 -0500
FreeBSD cvs-src summary for 11/10/04 to 18/10/04
++++++++++++++++++++++++++++++++++++++++++++++++
This is a regular weekly summary of FreeBSD's cutting-edge development.
It is intended to help the FreeBSD community keep up with the fast-paced
work going on in FreeBSD-CURRENT by distilling the deluge of data from
the CVS mailing list into a (hopefully) easy-to-read newsletter.  This
newsletter is marked up in reStructuredText_, so any odd punctuation
that you see is likely intended for the reST parser.

.. _reStructuredText: http://docutils.sourceforge.net/rst.html

You can get old summaries, and an HTML version of this one, at
http://www.xl0.org/FreeBSD/.  Please send any comments to Mark Johnston
(mark at xl0.org).

If you would like to get the summary without subscribing to current_at_,
please send mail to freebsd-cvs-summary-subscribe_at_lists.enderunix.org.
Thanks to Omer Faruk Sen and EnderUNIX for hosting this list.

For Lukasz Dudek and Szymon Roczniak's Polish translations of these
summaries, which may lag the English ones slightly, please see
http://mocart.pinco.pl/FreeBSD/.

.. contents::

============
New features
============
Simplified Chinese translation of release notes and other release 
documentation
-------------------------------------------------------------------------------
Xin Li (delphij) added a Simplified Chinese translation of the 5.3 release
notes, errata, hardware notes, installation guide, readme, and migration
guide.  The readme and installation guide were done by Chinese translation
project committers known as gavin and zhangluo; the rest was done by Xin.

http://www.freebsd.org/cgi/mid.cgi?200410150643.i9F6hnIY068137

GNU Readline updated to 5.0
---------------------------
Andrey A. Chernov (ache) updated GNU Readline, a command-line editing
library, to 5.0 plus official patches 001-005.  This replaces the old
version 4.3.  A changelog_ is available.

.. _changelog: 
http://www.freebsd.org/cgi/cvsweb.cgi/~checkout~/src/contrib/libreadline/CHANGES?rev=1.1.1.6&content-type=text/plain

http://www.freebsd.org/cgi/mid.cgi?200410180702.i9I72iLV056456

===============
Notable changes
===============
RELENG_5_3 moves into RC, RELENG_5 becomes -STABLE
--------------------------------------------------
Scott Long (scottl) created the RELENG_5_3 branch, moving 5.3 from beta to
Release Candidate status.  At the same time, he made the RELENG_5 branch
into the new -STABLE.

http://www.freebsd.org/cgi/mid.cgi?200410161926.i9GJQEQA026728

Serial port control files and devices renamed
---------------------------------------------
Poul-Henning Kamp (phk) updated sio, the most common serial port driver,
to use new serial code.  This also introduces a new naming scheme for
some related control files and the cua devices.  The device /dev/ttyid?
has been replaced by /dev/ttyd?.init, and /dev/ttyld? has been replaced
by /dev/ttyd?.lock.  /dev/cuaa? has been replaced with /dev/cuad?,
and /dev/cuaia? and /dev/cuala? have changed to /dev/cuad?.init and
/dev/cuad?.lock, respectively, like the tty* change.

http://www.freebsd.org/cgi/mid.cgi?200410130827.i9D8RKBM082371

New time zone data imported
---------------------------
Garrett Wollman (wollman) imported new timezone data, updating numerous
current and historical timestamps.  The new data creates new regions in
Argentina and Brazil, so if you live there, you should run tzsetup to make
sure that times in the future will be interpreted properly.

http://www.freebsd.org/cgi/mid.cgi?200410180003.i9I03qiF030812

=================
Discussion topics
=================
Support for real 386 chips
--------------------------
David O'Brien (obrien) updated some networking code with new assembler,
saying, "Sync with sys/i386/include/endian.h: use the single instruction
'bswap'."  The code tested the variable I386_CPU to see whether bswap
should be used or not.

Ruslan Ermilov (ru) responded, saying, "Ahem, but this makes this code
unsuitable for i386.  I386_CPU is only defined if kernel config file has
'cpu I386_CPU', and this has little relation to libc [where the change was
made] if at all."

David replied, "Yes. [ . . . ] For 6-CURRENT and RELENG_5 running on a
real I386 is not supported."  To the kernel vs. libc comment, he said,
"I know.  Give me some other symbol then we can publish that one should
add to /etc/make.conf CFLAGS to try to build a userland for an I386.
'I386_CPU' looks as good a spelling as any and I didn't see any need to
spell it differently than in the kernel."

Ruslan answered, "I remember the plan was to decomission the i386
support only in 6.0-CURRENT [ . . .]. And [the hardware list] says i386
(except 386SX) is still supported.  If you are burning 6.0-CURRENT
bridges already, and this change is the step in the I386_CPU deprecation
direction, why not just unifdef(1) this libc code, so that the 'xchg'
is always used?"  In response to the "other symbol" request, he said,
"You can check for the CPUTYPE="i386" for ${MACHINE_ARCH} == "i386" in
libc/Makefile, and set I386_CPU in CFLAGS in this case."

David responded, "I'm not decommisioning it -- I left a code path in there
for someone who cares about running on real I386 CPU's. [ . . . ] As noted
[on -current] we haven't built a 5-CURRENT I386 runable kernel since 2001.
Do you really think things haven't attrophied such that one must do some
work to run on an I386 at this point?"

Kris Kennaway (kris) replied, "Don't be too quick with the assertions;
someone already reported successfully running 5.2.1 on an i386 without
significant hassles."

Erik Trulsson responded to Ruslan's earlier post about the 386SX being
unsupported, saying, "[That] doesn't make much sense since (AFAIK) there
is no difference between 386DX and 386SX from a programmers point of view.
(The difference between them is that the 386sx has a 16-bit databus and a
24-bit address bus, while the 386dx has both 32-bit, but this only affects
the motherboard designers, not the people who write software.)"

David replied to Ruslan's post about setting CPUTYPE, saying,
"[S]pecifically what patch do you suggest that we used in cases like this
rather than #ifdef I386_CPU."

Ruslan said, "In src/lib/Makefile, add this:", gave some code, and
continued, "Or maybe even add -DI386_CPU to _CPUCFLAGS globally in
bsd.cpu.mk, so it can be used anywhere.  Note: this is completely
untested!"

Erik replied to David's earlier post saying the I386 is not supported,
saying, "I have certainly not seen any message stating that on -current.
I know that the GENERIC kernel does not work on a i386, so that you need
a custom-built kernel to install on a real i386. [ . . . ] I think you
people need to decide to either drop support for 80386 completely (in
which case all the '#ifdef I386_CPU' and similar can go away) or decide
to keep support for it, in which case most things should just work 'out
of the box' on a 80386 (without recompilation needed more than absolutely
necessary.)"

David responded, "I see nothing wrong with having to build a new userland
before it will run on I386 since we already have to do that with the
kernel."  To answer the comment about dropping support or working
out-of-the-box, he said, "That is a whole nother issue, and I welcome you
to bring it up in freebsd-arch_at_ (vs. as a follow up to this commit) where
all can debate it."

http://www.freebsd.org/cgi/mid.cgi?200410181719.i9IHJa9l097436

=================
Committer changes
=================
Dan Moschuk (dan) has returned his commit bit for safekeeping.

http://www.freebsd.org/cgi/mid.cgi?200410151229.i9FCTGhI086956

===============
Other bug fixes
===============
Brian Feldman (green) fixed a bug that would cause RAM disks to fail to be
set up properly when the system was booting.

http://www.freebsd.org/cgi/mid.cgi?200410120444.i9C4isW3059321

Yar Tikhiy (yar) fixed a bug in the FTP server that could result in an
infinite loop if a file being downloaded is overwritten.  Parts of the fix
were submitted by Oleg Koreshkov.

http://www.freebsd.org/cgi/mid.cgi?200410150931.i9F9V9g5080846

Robert Watson (rwatson) fixed a bug in the code supporting raw sockets in
jails that would cause processes that created raw sockets then lowered
their own privilege level to be unable to modify the IP socket options.
This would break multicast routing daemons, among other tools.

http://www.freebsd.org/cgi/mid.cgi?200410121647.i9CGlPBw027133
Received on Mon Oct 18 2004 - 17:47:38 UTC

This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:38:18 UTC