af6f39063b
1997-10-26 18:12 Ulrich Drepper <drepper@cygnus.com> * libio/genops.c: Partial undo of last patch. * libio/stdfiles.c: Likewise. * libio/iofdopen.c: Use _IO_FILE_complete, not _IO_file_plus. * libio/iopopen.c: Likewise. * libio/iovdprintf.c: Likewise. * libio/libio.h: Remove duplicated `;'. * libio/stdio.c: Remove misleading comment. * libio/stdio.h: Declare standard streams as variables. * login/Makefile (distribute): Add README.utmpd. * login/README.utmpd: New file. Provided by Mark M. Kettenis <kettenis@phys.uva.nl>. * manual/job.texi: Document tcgetsid. * manual/pattern.texi: Document globfree. * manual/terminal.texi: Document B38400 ... B460800. * posix/confstr.c: Print "-D_FILE_OFFSET_SIZE=64" for _CS_LFS_CFLAGS. * posix/unistd.h: Add explanation of _POSIX_* constants. * posix/unists.h: Add prototypes for __pread, __pread64, __pwrite and __pwrite64. * sysdeps/generic/pread.c: Define as __pread and make pread weak alias. * sysdeps/generic/pread64.c: Likewise. * sysdeps/generic/pwrite.c: Likewise. * sysdeps/generic/pwrite64.c: Likewise. * sysdeps/posix/pread.c: Likewise. * sysdeps/posix/pwrite.c: Likewise. * sysdeps/posix/pread64.c: New file. * sysdeps/posix/pwrite64.c: Likewise. * sysdeps/unix/sysv/linux/Makefile [$(subdir)=posix] (sysdep_routines): Add s_pread64 and s_pwrite64. * sysdeps/unix/sysv/linux/pread.c: New file. * sysdeps/unix/sysv/linux/pread64.c: New file. * sysdeps/unix/sysv/linux/pwrite.c: New file. * sysdeps/unix/sysv/linux/pwrite64.c: New file. * sysdeps/unix/sysv/linux/s_pread64.c: New file. * sysdeps/unix/sysv/linux/s_pwrite64.c: New file. * sysdeps/unix/sysv/linux/syscalls.list: Add pread and pwrite. * sysdeps/unix/sysv/linux/alpha/pread64.c: New (empty) file. * sysdeps/unix/sysv/linux/alpha/pwrite64.c: New (empty) file. * sysdeps/unix/sysv/linux/sparc/sparc64/pread64.c: New (empty) file. * sysdeps/unix/sysv/linux/sparc/sparc64/pwrite64.c: New (empty) file. * sysdeps/unix/sysv/linux/alpha/syscalls.list: Add pread and pwrite with weak aliases for *64 functions. * sysdeps/unix/sysv/linux/sparc/sparc64/syscalls.list: Likewise. * string/bits/string2.h: Add casts to allow void * arguments. * sysdeps/i386/i486/bits/string.h: Define index and rindex only if __USE_BSD or __USE_XOPEN_EXTENDED. * sysdeps/unix/sysv/linux/bits/socket.h: Add SCM_RIGHTS and other SCM_* constants from kernel header. * termios/termios.h: Add prototype for tcgetsid. 1997-10-26 13:26 Thorsten Kukuk <kukuk@vt.uni-paderborn.de> * sunrpc/clnt_perr.c: Add trailing '\0' to strings. * sunrpc/get_myaddr.c: Include rpc/clnt.h for prototypes. * sunrpc/pmap_clnt.c: Use get_myaddress from header file. 1997-10-26 05:26 Ulrich Drepper <drepper@cygnus.com> * configure.in: Punt if any directory mentioned in the enable-add-on parameter does not exist. 1997-10-25 19:25 Ulrich Drepper <drepper@cygnus.com> * termios/Makefile (routines): Add tcgetsid. * termios/tcgetsid.c: New file. Provided by Mark M. Kettenis <kettenis@phys.uva.nl>. 1997-10-25 18:56 Ulrich Drepper <drepper@cygnus.com> * stdlib/stdlib.h: Remove mblen optimization. * stdlib/mblen.c: Rewrite to make sure global state is not changed. Reported by anderson@metrolink.com. 1997-10-19 21:51 Wolfram Gloger <wg@wolfram.dent.med.uni-muenchen.de> * malloc/thread-m.h [_LIBC]: Use new __libc_internal_tsd_{set,get} interface for thread-specific data. 1997-10-25 06:51 Ulrich Drepper <drepper@cygnus.com> * elf/dl-addr.c: Use braces for correct logical grouping. Patch by Wolfram Gloger <wmglo@dent.med.uni-muenchen.de>. 1997-10-18 09:15 Geoff Keating <geoffk@ozemail.com.au> * io/ftwtest-sh: Sometimes /tmp is a symlink to somewhere more convenient; that caused this test to break. * sysdeps/powerpc/dl-machine.h: Fix typo. * sysdeps/powerpc/bits/fenv.h: Don't use floating-point registers when -msoft-float is in effect, because this causes compilation to stop. * sysdeps/powerpc/bits/mathinlines.h: Likewise. * rpm/template: Add description, use RPM flags rather than the ones used to build the spec. Build in a temporary directory, not /. * elf/dl-lookup.c: Don't include _itoa.h, it's not used. * elf/dl-minimal.c: Use _itoa_word rather than _itoa. It seems that _itoa is the only routine that ld.so uses that requires something from libgcc.a on powerpc, so it would be best to avoid it in ld.so. * elf/rtld.c: Likewise. * sysdeps/generic/_strerror.c: Likewise. * stdio-common/_itoa.c: Split out digits strings. * stdio-common/itoa-digits.c: New file. * stdio-common/Makefile: Add itoa-digits. 1997-10-21 Andreas Jaeger <aj@arthur.rhein-neckar.de> * manual/filesys.texi (Scanning Directory Content): Document error case more. * dirent/scandir.c (scandir): Ignore errors from select function. Suggested by urbanw@cs.umu.se (closes PR libc/316). 1997-10-25 06:18 Ulrich Drepper <drepper@cygnus.com> * sysdeps/unix/sysv/linux/sparc/sparc32/socket.S: Corrections. Patch by Erik Troan <ewt@redhat.com>. 1997-10-25 04:00 Ulrich Drepper <drepper@cygnus.com> * sysdeps/generic/dl-cache.c (_dl_load_cache_lookup): Favour exact matching of version function if both the general (1) and glibc-specific (3) entry are present. 1997-10-22 18:47 Thorsten Kukuk <kukuk@vt.uni-paderborn.de> * sunrpc/rpc/clnt.h: Add get_myaddress prototype. * nis/libnsl.map: Fix typo. * nis/nis_call.c: Fix memory leak. 1997-10-22 19:29 Ulrich Drepper <drepper@cygnus.com> * sysdeps/generic/memcmp.c: Define __P if not defined before. Patch by Jim Meyering <meyering@eng.ascend.com>. 1997-10-21 22:09 Ulrich Drepper <drepper@cygnus.com> * sysdeps/unix/sysv/linux/sys/prctl.h: New file by Richard Gooch <rgooch@atnf.csiro.au>. 1997-10-21 21:50 Ulrich Drepper <drepper@cygnus.com> * misc/syslog.c (vsyslog): Open console with O_NOCTTY. Patch by Zack Weinberg <zack@rabi.phys.columbia.edu>. 1997-10-21 18:07 Ulrich Drepper <drepper@cygnus.com> * posix/wordexp.c: Improve handling of $... expressions. Patch by Tim Waugh <tim@cyberelk.demon.co.uk>. 1997-10-21 16:12 Ulrich Drepper <drepper@cygnus.com> * manual/string.texi: Correct return values of bcopy and bzero. Patch by Matthew Wilcox <willy@odie.barnet.ac.uk>. 1997-10-18 15:03 Philip Blundell <Philip.Blundell@pobox.com> * sysdeps/unix/sysv/linux/bits/socket.h: Correct types of some elements in struct msghdr and struct cmsghdr, to keep in step with the kernel. 1997-10-17 22:29 Ulrich Drepper <drepper@cygnus.com> * sysdeps/unix/sysv/linux/sparc/sparc32/init-first.h: Fix another bug in startup code. Patch by Eric Delaunay <delaunay@lix.polytechnique.fr>. 1997-10-16 20:17 Richard Henderson <rth@cygnus.com> * sysdeps/unix/sysv/linux/sparc/sparc32/socket.S: Dump args to the stack and give the kernel a pointer. Use the sysdep.h macros. 1997-10-17 04:07 Ulrich Drepper <drepper@cygnus.com> * sysdeps/sparc/sparc32/elf/start.S: Calculate argv correctly. Patch by Eric Delaunay <delaunay@lix.polytechnique.fr>. 1997-10-16 Andreas Jaeger <aj@arthur.rhein-neckar.de> * sysdeps/libm-ieee754/s_nextafterxf.c [!__STDC__]: Correct typo. 1997-10-16 14:50 Ulrich Drepper <drepper@cygnus.com> * manual/pattern.texi: Document globfree. 1997-10-15 21:11 Philip Blundell <Philip.Blundell@pobox.com> * sysdeps/unix/sysv/linux/net/if_packet.h: New file. * sysdeps/unix/sysv/linux/Makefile (sysdep_headers): Add net/if_packet.h. * sysdeps/unix/sysv/linux/net/if_arp.h (ARPHRD_ASH): New type, for 64Mbps ASH. (ARPHRD_ETHER): This is used for 100Mbps networks too. 1997-10-15 Andreas Jaeger <aj@arthur.rhein-neckar.de> * Makerules (install): Use full pathnames for linker script. This is to work around a limitation in `ld' while no better solution is possible. 1997-10-15 Andreas Jaeger <aj@arthur.rhein-neckar.de> * malloc/malloc.c (mmap_chunk): Put inline before static in function definition to avoid compiler warning. (malloc_extend): Likewise. * sysdeps/generic/des_impl.c: Include "des.h" to avoid warning. 1997-10-15 Andreas Jaeger <aj@arthur.rhein-neckar.de> * NEWS: Fix @gnu.ai.mit.edu -> @gnu.org. * README.template: Likewise. * db/makedb.c: Likewise. * elf/ldd.bash.in: Likewise. * elf/ldd.sh.in: Likewise. * intl/locale.alias: Likewise. * login/programs/utmpd.c: Likewise. * libio/stdfiles.c [!_IO_MTSAFE] (DEF_STDFILE): Fix parameter list. 1997-10-14 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * Rules: Remove all empty.* files. (shared-only-routines): Correct implementation. 1997-10-14 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/libm-ieee754/s_lrintl.c: Make compilable. * sysdeps/libm-ieee754/s_llrintl.c: Likewise. Optimized. 1997-10-14 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * elf/ldd.bash.in: Only prepend ./ if the file contains no slash at all. * elf/ldd.sh.in: Likewise. 1997-10-14 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/m68k/sys/ucontext.h: New file. 1997-10-13 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/m68k/fpu/s_scalbln.c: New (empty) file. * sysdeps/m68k/fpu/s_scalblnf.c: New (empty) file. * sysdeps/m68k/fpu/s_scalblnl.c: New (empty) file. * sysdeps/m68k/fpu/s_scalbn.c: Add scalbln alias. * sysdeps/m68k/fpu/s_scalbnf.c: Adapted. * sysdeps/m68k/fpu/s_scalbnl.c: Adapted. * sysdeps/m68k/fpu/s_lrint.c: Add standard skeleton stuff. * sysdeps/m68k/fpu/s_lrintf.c: New file. * sysdeps/m68k/fpu/s_lrintl.c: New file. * sysdeps/m68k/fpu/bits/mathinline.h: Add fma and scalbln. Update lrint and scalbn. (__m81_inline) [__cplusplus]: Define to __inline. * math/bits/mathcalls.h: Remove whitespace before second argument of __MATHDECL. Add note explaining this. 1997-10-13 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * manual/arith.texi (Absolute Value): Spelling fix. 1997-10-13 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * malloc/obstack.h (obstack_empty_p) [!__GNUC__]: Properly parenthesize the macro parameter. * Rules: Remove rules to magically install <subdir>.h headers.
286 lines
12 KiB
Plaintext
286 lines
12 KiB
Plaintext
GNU libc SNAPSHOT SYSTEM
|
|
(general info)
|
|
Updated 1997-9-26
|
|
|
|
WHAT ARE GNU libc SNAPSHOTS
|
|
---------------------------
|
|
|
|
Snapshots are an "image" of the main glibc development tree, captured at a
|
|
particular random instant in time. When you use the snapshots, you should be
|
|
able to maintain a local copy of libc that is no more than one day older than
|
|
the official source tree used by the libc maintainers.
|
|
|
|
The primary purpose of providing snapshots is to widen the group of motivated
|
|
developers that would like to help test, debug, and enhance glibc, by providing
|
|
you with access to the "latest and greatest" source. This has several
|
|
advantages, and several disadvantages.
|
|
|
|
First the advantages:
|
|
|
|
o Once we have a large base of motivated testers using the snapshots,
|
|
this should provide good coverage across all currently supported
|
|
glibc hosts and targets. If a new bug is introduced in glibc due to
|
|
fixing another bug or ongoing development, it should become
|
|
obvious much more quickly and get fixed before the next general
|
|
net release. This should help to reduce the chances of glibc being
|
|
released to the general public with a major bug that went unnoticed
|
|
during the release cycle testing because they are machine dependent.
|
|
We hope to greatly improve glibc's stability and reliability by
|
|
involving more people and more execution environments in the
|
|
prerelease testing.
|
|
|
|
o With access to the latest source, any diffs that you send to fix
|
|
bugs or add new features should be much easier for the glibc team
|
|
to merge into the official source base (after suitable review
|
|
of course). This encourages us to merge your changes quicker,
|
|
while they are still "fresh".
|
|
|
|
o Once your diffs are merged, you can obtain a new copy of glibc
|
|
containing your changes almost immediately. Thus you do not
|
|
have to maintain local copies of your changes for any longer
|
|
than it takes to get them merged into the official source base.
|
|
This encourages you to send in changes quicker.
|
|
|
|
And the disadvantages:
|
|
|
|
o The snapshot you get will be largely untested and of unknown quality.
|
|
It may fail to configure or compile. It may have serious bugs.
|
|
You should always keep a copy of the last known working version
|
|
before updating to the current snapshot, or at least be able to
|
|
regenerate a working version if the latest snapshot is unusable
|
|
in your environment for some reason.
|
|
|
|
If a production version of glibc has a bug and a snapshot has the fix,
|
|
and you care about stability, you should put only the fix for that
|
|
particular problem into your production version. Of course, if you
|
|
are eager to test glibc, you can use the snapshot versions in your
|
|
daily work, but users who have not been consulted about whether they
|
|
feel like testing glibc should generally have something which is at
|
|
least as bug free as the last released version.
|
|
|
|
o Providing timely response to your questions, bug reports, and
|
|
submitted patches will require the glibc development team to allocate
|
|
time from an already thin time budget. Please try to help us make
|
|
this time as productive as possible. See the section below about
|
|
how to submit changes.
|
|
|
|
|
|
WHO SHOULD TRY THE SNAPSHOTS
|
|
----------------------------
|
|
|
|
Remember, these are snapshots not tested versions. So if you use
|
|
these versions you should be able to
|
|
|
|
o make sure your system stays usable
|
|
|
|
o locate and hopefully fix problems
|
|
|
|
o to port glibc to a new target yourself
|
|
|
|
You should not use the snapshots if
|
|
|
|
o your system is needed in a production environment which needs
|
|
stability
|
|
|
|
o you expect us to fix your problems since you somehow depend on them.
|
|
You must be willing to fix the problems yourself, we don't want to
|
|
see "I have problems, fix this" messages.
|
|
|
|
|
|
HOW TO GET THE SNAPSHOTS
|
|
------------------------
|
|
|
|
At the moment we provide a full snapshot weekly (every sunday), so
|
|
that users getting a snapshot for the first time, or updating after
|
|
a long period of not updating, can get the latest version in a single
|
|
operation. Along with the full snapshot, we will provide incremental
|
|
diffs on a nearly daily basis (whenever code changes). Each daily
|
|
diff will be relative to the source tree after applying all previous
|
|
daily diffs. The daily diffs are for people who have relatively low
|
|
bandwidth ftp or uucp connections.
|
|
|
|
The files will be available via anonymous ftp from alpha.gnu.ai.mit.edu, in
|
|
directory /gnu/libc and on linux.kernel.org in /pub/software/libs/glibc. The
|
|
directories should look something like:
|
|
|
|
libc-970921.tar.gz
|
|
libc-970917-970922.diff.gz
|
|
libc-970922-970925.diff.gz
|
|
.
|
|
.
|
|
.
|
|
|
|
Please note that the snapshots on alpha.gnu.ai.mit.edu and on
|
|
linux.kernel.org are not always in sync. Patches to some files might
|
|
appear a day a diff earlier or later on alpha than on kernel.
|
|
Use always alpha or always kernel but don't mix them.
|
|
|
|
There are sometimes additionally test releases of the add-ons available in
|
|
these directories. If a new version of an add-on is available it is normally
|
|
required for the corresponding snapshot so always pay attention for these.
|
|
|
|
Note that we provide GNU gzip compressed files only. You can ftp gzip
|
|
from prep.ai.mit.edu in directory pub/gnu.
|
|
|
|
In some cases the dates for diffs and snapshots do not match like in the
|
|
example above. The full release is for 970921 but the patch is for
|
|
970917-970922. This only means that nothing changed between 970917 and 970922
|
|
and that you have to use this patch on top of the 970921 snapshot since the
|
|
patch is made on 970922.
|
|
|
|
Also, as the gcc developers did with their gcc snapshot system, even though we
|
|
will make the snapshots available on a publically accessible ftp area, we ask
|
|
that recipients not widely publicise their availability. The motivation for
|
|
this request is not to hoard them, but to avoid the situation where the
|
|
general glibc user base naively attempts to use the snapshots, has trouble with
|
|
them, complains publically, and the reputation of glibc declines because of a
|
|
perception of instability or lack of quality control.
|
|
|
|
|
|
GLIBC TEST SUITE
|
|
----------------
|
|
|
|
A test suite is distributed as an integral part of the snapshots. A simple
|
|
"make check" in your build directory is sufficient to run the tests. glibc
|
|
should pass all tests and if any fails, please report it. A failure might not
|
|
originate from a bug in glibc but also from bugs in the tools, e.g. with gcc
|
|
2.7.2.x the math tests fail some of the tests because of compiler bugs.
|
|
|
|
Note that the test suite is still in its infancy. The tests themselves only
|
|
cover a small portion of libc features, and where tests do exist for a feature
|
|
they are not exhaustive. New tests are welcome.
|
|
|
|
|
|
GETTING HELP, GLIBC DISCUSSIONS, etc
|
|
------------------------------------
|
|
|
|
People who want to help with glibc and who test out snapshots regularly should
|
|
get on the libc-alpha@gnu.ai.mit.edu mailing list by sending an email to
|
|
libc-alpha-request@gnu.ai.mit.edu. This list is meant (as the name suggests)
|
|
for the discussion of test releases and also reports for them. People who are
|
|
on this list are welcome to post questions of general interest.
|
|
|
|
People who are not only willing to test the snapshots but instead really want
|
|
to help developing glibc should contact libc-hacker-request@gnu.ai.mit.edu to
|
|
be put on the developers mailing list. This list is really only meant for
|
|
developers. No questions about installation problems or other simple topics
|
|
are wanted nor will they be answered.
|
|
|
|
Do *not* send any questions about the snapshots or patches specific to the
|
|
snapshots to bug-glibc@prep.ai.mit.edu. Nobody there will have any idea what
|
|
you are talking about and it will just cause confusion.
|
|
|
|
|
|
BUG REPORTS
|
|
-----------
|
|
|
|
Send bug reports directly to Ulrich Drepper <drepper@gnu.ai.mit.edu>. Please
|
|
do *not* use the glibcbug script for reporting bugs in the snapshots.
|
|
glibcbug should only be used for problems with the official released versions.
|
|
We don't like bug reports in the bug database because otherwise the impression
|
|
of instability or lack of quality control of glibc as a whole might manifest
|
|
in people's mind.
|
|
|
|
Note that since no testing is done on the snapshots, and snapshots may even be
|
|
made when glibc is in an inconsistent state, it may not be unusual for an
|
|
occasional snapshot to have a very obvious bug, such as failure to compile on
|
|
*any* machine. It is likely that such bugs will be fixed by the next
|
|
snapshot, so it really isn't necessary to report them unless they persist for
|
|
a couple of days.
|
|
|
|
Missing files should always be reported, since they usually mean there is a
|
|
problem with the snapshot-generating process and we won't know about them
|
|
unless someone tells us.
|
|
|
|
Bugs which are non-obvious, such as failure to compile on only a specific
|
|
machine, a new machine dependent or obscure bug (particularly one not detected
|
|
by the testsuite), etc should be reported when you discover them, or have a
|
|
suggested patch to fix them.
|
|
|
|
|
|
FORMAT FOR PATCHES
|
|
------------------
|
|
|
|
If you have a fix for a bug, or an enhancement to submit, send your patch to
|
|
Ulrich Drepper <drepper@gnu.ai.mit.edu>. Here are some simple guidelines for
|
|
submitting patches:
|
|
|
|
o Use "unified diffs" for patches. A typical command for generating
|
|
context diffs is "diff -ru glibc-old glibc-patched".
|
|
|
|
o Use the "minimalist approach" for patches. That is, each patch
|
|
should address only one particular bug, new feature, etc. Do not
|
|
save up many unrelated changes and submit them all in one big
|
|
patch, since in general, the larger the patch the more difficult
|
|
it is for us to decide if the patch is either correct or
|
|
desirable. And if we find something about the patch that needs
|
|
to be corrected before it can be installed, we would have to reject
|
|
the entire patch, which might contain changes which otherwise would
|
|
be accepted if submitted separately.
|
|
|
|
o Submit a sample ChangeLog entry with your patch. See the existing
|
|
glibc ChangeLog for examples of what a ChangeLog entry should look
|
|
like. The emacs command ^X4A will create a ChangeLog entry header
|
|
for you.
|
|
|
|
|
|
BUILDING SNAPSHOTS
|
|
------------------
|
|
|
|
The `best' way to build glibc is to use an extra directory, e.g.:
|
|
tar xzf libc-970921.tar.gz
|
|
mkdir build-glibc
|
|
cd build-glibc
|
|
../libc-970921/configure ...
|
|
|
|
In this way you can easily clean up (since `make clean' doesn't work at
|
|
the moment) and rebuild glibc.
|
|
|
|
|
|
NECESSARY TOOLS
|
|
---------------
|
|
|
|
For the recommended versions of gcc, binutils, make, texinfo, gettext,
|
|
autoconf and other tools which might be especially needed when using patches,
|
|
please read the file INSTALL.
|
|
|
|
|
|
HOW CAN YOU HELP
|
|
----------------
|
|
|
|
It helps already a lot if you just install glibc on your system and try to
|
|
solve any problems. You might want to look at the file `PROJECTS' and help
|
|
with one of those projects, fix some bugs (see `BUGS' or the bug database),
|
|
port to an unsupported platform, ...
|
|
|
|
|
|
FURTHER DOCUMENTATION
|
|
---------------------
|
|
|
|
A lot of questions are answered in the FAQ. The files `INSTALL', `README' and
|
|
`NOTES' contain the most important documentation. Furthermore glibc has its
|
|
own 700+ pages info documentation, ...
|
|
|
|
|
|
|
|
And finally a word of caution: The libc is one of the most fundamental parts
|
|
of your system - and these snapshots are untested and come without any
|
|
guarantee or warranty. You might be lucky and everything works or you might
|
|
crash your system. If you install a glibc snapshot as primary library, you
|
|
should have a backup somewhere.
|
|
|
|
On many systems it is also a problem to replace the libc while the system is
|
|
running. In the worst case on broken OSes some systems crash. On better
|
|
systems you can move the old libc aside but removing it will cause problems
|
|
since there are still processes using this libc image and so you might have to
|
|
check the filesystem to get rid of the libc data. One good alternative (which
|
|
is also safer) is to use a chroot'ed environment.
|
|
|
|
Thanks for your help and support.
|
|
|
|
Thanks to Fred Fish from Cygnus for the original version of this text
|
|
(for GDB).
|
|
|
|
|
|
Ulrich Drepper
|