Incorrect timestamp logging in gssftpd (krb5-workstation)
by Philip Prindeville
I found a situation (fairly common, actually) under which gssftpd
generates incorrect timestamps as a result of chroot()ing and losing
access to the /etc/localtime file (it surprises me that localtime(3)
doesn't cache the file's contents in glibc... oh, well).
I came up with a work-around for this which is fairly simple.
Can someone else code review it? And if the Kerberos folks
(krbdev) don't want to include the fix (I sent it to them, but didn't
hear back) what's the criteria for it being included in the Redhat/
Fedora packaging instead?
Here's the bug report (fix included):
https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=216356
-Philip
17 years, 5 months
Re: Static linking considered harmful
by Jakub Jelinek
On Wed, Nov 22, 2006 at 09:11:07AM -0800, John Reiser wrote:
> It is true. Dynamic linking to libstdc++.so is a recipe for disaster in
> a program that intends to be portable, because there have been bugs in the
> backwards compatibility, and there is no intention of forwards compatibility.
> Even some pieces of 'new' and 'delete' do not work across different
> g++/libstdc++ versions because the name mangling changed. This is
> particularly true for programs that [attempt to] support dynamically-loaded
> runtime plugin modules. In practice this almost certainly fails if the plugin
> uses a different version of g++/libstdc++ than the base program, because
> simultaneous use of two (or more!) different versions of libstdc++
> by the same process breaks. Each of the various libstdc++.so thinks
> that it is the only one, and so does not coordinate with the others.
Plugins are a separate issue, by definition those are dynamically linked
and -Bstatic -lstdc++ -Bdynamic plus symver script to hide the stuff
isn't the right answer. One of the reasons is that e.g. libstdc++.a isn't
-fpic, so linking it into a plugin is highly undesirable if at all possible.
For plugins the simplest answer is "don't write plugins in C++ unless it
is a plugin for a C++ project where everything already uses some one
libstdc++", or libstdc++so7's versioned namespaces.
> Speaking of glibc, even that has had compatibility problems.
> For instance, in history the character classification and translation
> facilities islower, isupper, tolower, toupper have had incompatible usage
> of external symbols with "ctype" in the name. For example, see this thread:
> http://groups.google.com/group/comp.os.linux.development.apps/browse_thre...
> Also, there have been libc.so compatibility bugs with symbol visibility of the
> high precision timers used by ld-linux to measure its own performance.
ctype wasn't in any way a compatibility problem. glibc is even in that
regard backwards compatible (though not forwards, as with dozens of other
symbols) and never guarantees ABI compatibility if you compile against
one glibc and link against a different one - only binaries and shared
libraries can count on backwards compatibility.
> > Care to share the details? libgcc_s is symbol versioned and is backwards
> > compatible. In FC we try to backport libgcc_s additions even from newer
> > gcc's than the one shipped, so e.g. current GCC 4.1.x-RH libgcc_s
> > has even all of GCC_4.2.0 stuff in it (and will add GCC_4.3.0 stuff soon).
>
> The unwinding code for DWARF2 is not compatible between libgcc_s 3.2.x
> and libgcc_s 4.y. The bugs are different, so user code that calls the
> unwinders, and insists on working despite the bugs, must adapt according
> to which version is being used. True binary compatibility sometimes
> means bug-for-bug compatibility, because sometimes the runtime language
> support has not caught up to the user code.
I'd argue that apps shouldn't work around unwinder bugs, instead they
should be simply reported and the vendors should fix them.
Certainly that's far better than e.g. having two or more unwinders within
one process.
Jakub
17 years, 5 months
Re: Static linking considered harmful
by Benjamin De Kosnik
(sorry about the messed up references: field. Jakub just pointed me at
this discussion and I realized that I was dropped/not-readded
from this list.)
> Far simpler, cleaner and more friendly will be to dump the static
> libraries into libfoo-devel-static packages which are not in the
> default install. That ensures people who need to can do static links
> and the default behaviour is that they are not there. The -static
> packages could even get dropped off the CD easily enough once the
> mechanisms for handling one srpm spitting out binary components for
> extras and core are sorted
yay! I'm glad somebody suggested this. This is my preference as well.
I think it provides a better transition plan then the ignominious
beheading of all *.a's.
-benjamin
17 years, 5 months
After Rawhide update yesterday xfs crashes
by Jochen Schlick
immediately. Xorg.log says something about not finding
the fixed font (which is because xfs isn't running)
Is this a known problem ??
best regards
--
------------------------------------------------------------
Jochen Schlick
17 years, 5 months
gnome-appliation/firefox crashed with Xfce
by A S Alam
Hi
during login to Xfce environment, firefox and gnome applications are
crashing
with message as following:
** Error ** Could not locate registry
aborting....
Bonobo accessibility support initialized
GTK accessibility Module initialized
**(gnome_segv2:3345): CRITICAL **AT_SPI_REGISTRY was not started at
session startup.
** (gnome_seg3:3345): WARNING ** IOR not set.
I have latest rawhide. Can some check th
--
A S Alam
timezone: GMT+5:30
join us at #fedora-l10n (freenode)
"Either find a way or make one"
17 years, 5 months
rawhide report: 20061122 changes
by Build System
Updated Packages:
anaconda-11.2.0.1-1
-------------------
* Tue Nov 21 2006 Chris Lumens <clumens(a)redhat.com> - 11.2.0.1
- Use .discinfo files to determine if a CD tree exists instead of a set
limit (pjones, #214787).
- Allow fixing/retrying when unable to grab a ksfile (#216446).
- Fix typos (pnasrat, #216410, #215232).
- Set the RAID minor number in text installs (#215231).
- Be smarter about detecting if iscsi is available (katzj, #216128).
- Kernel naming fix (katzj, #215746).
- Activate/login/logout of all iscsi devices (pjones).
- Depsolve on optional/non-grouped packages (pnasrat, #214848).
- Update kickstart documentation.
- Don't always write out xconfig and monitor in anaconda-ks.cfg (#211977).
- Follow drive order specified in kickstart file (#214881).
- Unmount source on image installs before %post is run (#214677).
- Check return value of getBiosDisk (pjones, #214653).
- splittree shouldn't fail with non-rpms in the directory (jkeating).
- Order bind mounts correctly on upgrades (#212270).
- Always skip networking on kickstart installs (#214584).
- Handle ipv6= command line option (dcantrell).
- Split up ipv4 and ipv6 options, add radvd support (dcantrell, #213108,
- Log exceptions when activating raid (pjones).
- Update install method documentation (#214159).
- Netconfig UI fixes (katzj, #213356).
- Avoid traceback on unused PReP partitions (#211098).
- Fix no free space traceback (ddearauj AT us.ibm.com, #213616(.
- Fix implantisomd5 output formatting (#214046).
- Remove virt group hacks (katzj).
- Default to text install if a KVM lies to us about the monitor.
- Split media FTP/HTTP loopback mount install fixes (pnasrat, #212014).
autoconf-2.61-2
---------------
* Tue Nov 21 2006 Karsten Hopp <karsten(a)redhat.com> 2.61-2
- drop obsolete linkX11 patch
* Tue Nov 21 2006 Karsten Hopp <karsten(a)redhat.com> 2.61-1
- autoconf-2.61
* Thu Nov 09 2006 Karsten Hopp <karsten(a)redhat.com> 2.60-4
- autoconf-2.60
automake-1.10-2
---------------
* Tue Nov 21 2006 Karsten Hopp <karsten(a)redhat.com> 1.10-2
- rebuild
* Fri Nov 10 2006 Karsten Hopp <karsten(a)redhat.de> 1.10-1
- automake 1.10
* Wed Jul 12 2006 Jesse Keating <jkeating(a)redhat.com> - 1.9.6-2.1
- rebuild
cpuspeed-1:1.2.1-1.43.fc7
-------------------------
* Tue Nov 21 2006 Dave Jones <davej(a)redhat.com>
- Merge one more patch from Michal Jaegermann <michal(a)harddata.com> (#216816)
cups-1:1.2.7-3.fc7
------------------
* Tue Nov 21 2006 Tim Waugh <twaugh(a)redhat.com> 1:1.2.7-3
- Run the serial backend as root (bug #212577).
curl-7.16.0-3.fc7
-----------------
* Thu Nov 16 2006 Jindrich Novy <jnovy(a)redhat.com> -7.16.0-3
- prevent curl from dlopen()ing missing ldap libraries so that
ldap:// requests work (#215928)
elinks-0.11.1-5.1
-----------------
* Tue Nov 21 2006 Karel Zak <kzak(a)redhat.com> 0.11.1-5.1
- fix #215734: CVE-2006-5925 elinks smb protocol arbitrary file access
fonts-indic-2.0.9-1.fc7
-----------------------
* Tue Nov 21 2006 Parag Nemade <pnemade(a)redhat.com> - 2.0.9-1.fc7
- Fixed Bugs from Parag Nemade
- Bug 216060: [pa_IN]Lohit Punjabi: (U+0A03) Gurmukhi Sign Visarga right margin is too wide.
* Tue Nov 21 2006 Parag Nemade <pnemade(a)redhat.com> - 2.0.8-1.fc7
- Fixed Bugs from Parag Nemade
- Bug 216629 : [te_IN] GPOS position should be at middle instead of left hand side - Priority B
- Bug 216631 : [or_IN] one GSUB Rule is wrongly defined - Priority A
- Bug 216628 : [ml] characters are not shown bold during selecting BOLD font
- Bug 216624 : [pa_IN] 0a01/0a03/0964/0965 is missing from font
- Bug 216634 : Glyphs for two combinations are wrong in Gujarati (gu_IN)
- Bug 197216 : [bn_IN]Incorrect glyph for conjunct
- Bug 215894 : Relative height of 0x0901 (and 0x0902) on 0x0915 is different than
other devnagari characters (hi_IN, mr_IN)
- Fixed Bugs from LingNing Zhang
- Bug 216626 : [as_IN] Pango - A particular char when Conjuncts, creates Cursor Nevigation Problem
- Bug 216627 : [ml_IN] The glyph of 0x25CC is not existing in lohit_ml.ttf
- Bug 216639 : [kn_IN] OTF rules to be fixed - Priority A
gaim-2:2.0.0-0.22.beta5.fc7
---------------------------
* Tue Nov 21 2006 Warren Togami <wtogami(a)redhat.com> - 2:2.0.0-0.22.beta4
- 2.0.0 beta5
- Debian patches
02_gnthistory-in-gtk
03_gconf-gstreamer
04_blist-memleak
05_url-handler-xmpp
06_jabber-registration-srv
07_msn-custom-smiley-crash
- SILC Account Edit Crash
* Tue Nov 21 2006 Warren Togami <wtogami(a)redhat.com> - 2:2.0.0-0.21.beta4
- #212817 Jabber needs cyrus-sasl plugins for authentication
gtk2-2.10.6-5.fc7
-----------------
* Tue Nov 21 2006 Matthias Clasen <mclasen(a)redhat.com> - 2.10.6-5
- Change the search patch to check for beagle first
k3b-0:0.12.17-1
---------------
* Thu Oct 26 2006 Harald Hoyer <harald(a)redhat.com> - 0:0.12.17-1
- - new version 0.12.17
libao-0.8.6-4
-------------
* Tue Nov 21 2006 Behdad Esfahbod <besfahbo(a)redhat.com> - 0.8.6-4
- Only export namespaced symbols. (bug 216108)
liboil-0.3.10-1.fc7
-------------------
* Tue Nov 21 2006 Behdad Esfahbod <besfahbo(a)redhat.com> - 0.3.10-1
- Update to 0.3.10
libsoup-2.2.98-1.fc7
--------------------
* Tue Nov 21 2006 Matthew Barnes <mbarnes(a)redhat.com> - 2.2.98-1
- Update to 2.2.98
- Remove patch for RH bug #215919 (fixed upstream).
lvm2-2.02.15-3.fc7
------------------
* Tue Nov 21 2006 Alasdair Kergon <agk(a)redhat.com> - 2.02.15-3
- Fix clvmd init script line truncation.
* Tue Nov 21 2006 Alasdair Kergon <agk(a)redhat.com> - 2.02.15-2
- Fix lvm.conf segfault.
* Mon Nov 20 2006 Alasdair Kergon <agk(a)redhat.com> - 2.02.15-1
- New upstream - see WHATS_NEW.
mod_python-3.2.10-3
-------------------
* Tue Nov 21 2006 Joe Orton <jorton(a)redhat.com> 3.2.10-3
- update to 3.2.10
nspr-4.6.4-1
------------
* Tue Nov 21 2006 Kai Engert <kengert(a)redhat.com> - 4.6.4-1
- Update to 4.6.4
nss-3.11.4-1
------------
* Tue Nov 21 2006 Kai Engert <kengert(a)redhat.com> - 3.11.4-1
- Update to 3.11.4
ntp-4.2.2p4-1.fc7
-----------------
* Tue Nov 21 2006 Miroslav Lichvar <mlichvar(a)redhat.com> 4.2.2p4-1
- update to 4.2.2p4
- fix buffer overflow in WWV Audio driver (#216309)
- don't mark init script as config
openoffice.org-1:2.1.0-4.2
--------------------------
* Tue Nov 21 2006 Caolan McNamara <caolanm(a)redhat.com> - 1:2.1.0-4.2
- Resolves: rhbz#216347 openoffice.org-2.1.0.ooo71815.bridges.x86_64.patch
- Resolves: rhbz#216662 stick to a single .desktop launcher name
privoxy-3.0.6-1
---------------
* Tue Nov 21 2006 Karsten Hopp <karsten(a)redhat.com> 3.0.6-1
- privoxy-3.0.6
rhythmbox-0.9.6-3.fc7
---------------------
* Tue Nov 21 2006 Ray Strode <rstrode(a)redhat.com> - 0.9.6-3
- drop keybinding patch
scim-pinyin-0.5.91-14.fc7
-------------------------
* Wed Nov 22 2006 Shawn Huang <phuang(a)redhat.com> - 0.5.91-14
- Translate refined online help to Chinese (#200702).
- Update some msgids in po/zh_CN.po to fix Some tooltips were not
translated (#216831).
* Wed Oct 11 2006 Caius Chance <cchance(a)redhat.com> - 0.5.91-13.fc6
- bz#200702 - refine online help of pinyin helper.
sysstat-7.0.2-3.fc7
-------------------
* Tue Nov 21 2006 Ivana Varekova <varekova(a)redhat.com> - 7.0.2-3
- update NFS mount statistic patch
system-config-bind-4.0.2-3.fc7
------------------------------
* Tue Nov 21 2006 Martin Stransky <stransky(a)redhat.com> - 4.0.2-3
- added a second patch for issue #216584
* Tue Nov 21 2006 Martin Stransky <stransky(a)redhat.com> - 4.0.2-2
- added version patch (for #216584)
* Tue Oct 17 2006 Martin Stransky <stransky(a)redhat.com> - 4.0.2-1
- updated translations
system-config-display-1.0.46-1.fc7
----------------------------------
* Tue Nov 21 2006 Matthias Clasen <mclasen(a)redhat.com> - 1.0.46-1
- Fix mnemonic for the color depth combo on the last tab
Resolves: rhbz#216391
system-config-httpd-5:1.4.2-1.fc7
---------------------------------
* Mon Nov 20 2006 Phil Knirsch <pknirsch(a)redhat.com> 1.4.2-1.fc7
- Fixed description.
system-config-keyboard-1.2.11-1.fc7
-----------------------------------
* Tue Nov 21 2006 Paul Nasrat <pnasrat(a)redhat.com> - 1.2.11-1
- Update translations
system-config-printer-0.7.39-1.fc7
----------------------------------
* Tue Nov 21 2006 Tim Waugh <twaugh(a)redhat.com> 0.7.39-1
- 0.7.39:
- Busy cursor while loading foomatic and PPD list (bug #215527).
- Make PPD NickName selectable.
- Added SMB hint label on device screen (bug #212759).
system-config-users-1.2.48-1.fc7
--------------------------------
* Sat Oct 14 2006 Nils Philippsen <nphilipp(a)redhat.com> - 1.2.48
- pick up updated translations
traceroute-3:2.0.2-1.fc7
------------------------
* Tue Nov 21 2006 Martin Bacovsky <mbacovsk(a)redhat.com> - 3:2.0.2-1.fc7
- new source
- more accurate check_expired() routine.
- some minor fixes.
vim-2:7.0.164-2
---------------
* Tue Nov 21 2006 Karsten Hopp <karsten(a)redhat.com> 7.0.164-2
- patchlevel 164
Broken deps for s390
----------------------------------------------------------
systemtap - 0.5.10-1.fc7.s390 requires kernel >= 0:2.6.9-11
systemtap-runtime - 0.5.10-1.fc7.s390 requires kernel >= 0:2.6.9-11
17 years, 5 months
Re: Static linking considered harmful
by William W. Austin
On 2006-11-22 05:37:43, Patrice Dumas <pertusus(a)free.fr> wrote:
> On Wed, Nov 22, 2006 at 10:29:44AM +0000, Andrew Haley wrote:
> > >
> > > Not only. There are cases when all those issues are moot,
> > > a prominent one
> > > being for numerical models. Compiling models statically makes
> > > it possible
> > > to run them on any other linux (including different fedora
> > > version) box
> > > without recompiling. So all the libraries that can be used for
> > > numerical
> > > computations should have static libraries kept.
> >
> > This seems like a non sequitur. What's special about numerical
> > computations?
>
> I already explained it in my mail? To be more precise
>
> * security considerations are moot
> * nss, iconv are not needed
>
> Other arguments for dynamic linking are not compulsory.
> And numerical models are programs we would like to be able to run
> on other linux box without recompiling. Imagine that compilation
> takes 30 minutes and that you want to start 20 runs in paralell
> from nfs mounted filesystem in an heterogeneous linux environment
> with some redhat 6.2, 7.x, centos, debian, mandrake and fedora
> core (real world example, of course).
>
> --
> Pat
I have to agree with Patrice Dumas. I regularly do extensive numerical
analysis and modeling using such a network, and the compilation time
for the central program is VERY long, taking over 40 minutes on the
fastest machine I have available. While I don't think we have anything
older than RH8 in the pool of runtime machines, many of them cannot be
upgraded for other reasons (I don't own them - I just get to use them
when they are available, so the pool of machines changes probably every
time I visit phase space), and if I had to do a compile for every
different O/S+release combination I'm using, it would be a nightmare.
While my run time is longer than Pat's example (a little over 4 hours
using around 30 machines), if I had to recompile for each of the
different variations I use, it would be prohibitively time-consuming
and would be a reason to stop upgrading several machines here.
I don't know whether eliminating static libraries would be a problem
for other types of apps, for what I'm doing its elimination would be a
disaster.
While I don't see a problem with eliminating static linking of system
utilities, IMHO eliminating it for user-written utilities is not a
"universal benefit."
--
william w. austin waustin(a)speakeasy.net
"life is just another phase i'm going through. this time, anyway ..."
17 years, 5 months
Re: FC6: some impressions
by Otto Rey
Evolution hang when press send/recive with M$ Exchange account. I am tired of Evolution bugs. One update, work, next update, broken, next update work, next update broken and so on. Do you know some alternative for Exchange accounts?
I am the Developer of JPVMailer email client (only winshit, now deprecated) and im starting development of JPVMailerX (GPL licence of course) completely written in Java (Sun GPL java :) ).
More information http://www.ottorey.com.ar or the same http://www.jpvmailer.com.ar (Sorry, only spanish by now)
Im planning to setup CVS repository and translate the site to English. Next i will call for voluntaries. (umm... looks like sourceforge...:) )
Well... for now just ideas.... and a bit of work...
----- Original Message ----
From: Richard Hughes <hughsient(a)gmail.com>
To: Development discussions related to Fedora Core <fedora-devel-list(a)redhat.com>
Sent: Tuesday, November 21, 2006 8:22:35 PM
Subject: Re: FC6: some impressions
On Tue, 2006-11-21 at 17:30 -0500, Dimi Paun wrote:
>
> 2. Evolution hangs when sending email
> https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=208227
> https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=208724
> This renders Evolution almost entirely useless,
> making this a critical bug.
Happening to me too - evolutions hangs when it tries to connect to
smtp.gmail.com and then requires a manual kill -9 to close it.
Richard.
--
fedora-devel-list mailing list
fedora-devel-list(a)redhat.com
https://www.redhat.com/mailman/listinfo/fedora-devel-list
__________________________________________________
Correo Yahoo!
Espacio para todos tus mensajes, antivirus y antispam ¡gratis!
¡Abrí tu cuenta ya! - http://correo.yahoo.com.ar
17 years, 5 months
iSCSI Boot and FC6
by Narendra Hadke
Hi,
I am exploring iSCSI Install to target LUN support in FC6 and came across few issues while trying out the same.
1. Found that if network is not configured before graphical install, Advance Storage
Option dialog doesnt prompt for the target ip address. It prompts for network
config but doesn't go to "target ip dialog". For some reason "linux asknetwork"
did not work as expected.
2. I used option linux updates to configure network manually from the
command prompt. Fedora 6 was installed(default) completely on the target LUN.
After booting, kernel panic while mounting root filesystem. I beleive reason being
install image didn't have "initrd" which can does discovery for boot LUN.
3. Before rebooting the installation, created "initrd" and modified grub setting.
Discovery happened and root filesystem got mounted correctly.
4. Booting got struck during network install as interface on which initiator connected
was brought down by network script. I bypassed this commenting this script( not
elegant way of doing though) in sysinit. I was able to boot with FC6 from target
LUN.
Question I have is, if this feature fully supported without the workarrounds in FC6
I have done or I am missing some thing?
Thanks,
Narendra Hadke
---------------------------------
Sponsored Link
Rates near 39yr lows. $510,000 Loan for $1698/mo - Calculate new house payment
17 years, 5 months
CVS Restored, Help Needed in Verification
by Warren Togami
Sometime Early Friday, November 17th we suffered some kind of
catastrophic multi-disk or SCSI backplane failure accompanied with
filesystem corruption in cvs.fedora.redhat.com. Thanks to volunteer
efforts of Fedora Infrastructure and tremendous help from Red Hat IS/IT,
we are now nearing restoration of cvs.fedora.redhat.com from a backup on
a powerful new server graciously donated by Dell.
We are restoring from a backup that began November 16th, 2006 at 8:01PM
MST (Friday, November 17, 2006 at 03:01) and finished 100 minutes later.
Because this backup is not an exact snapshot, it is possible that
checkins that happened after 8:01PM are already in the restored data.
https://www.redhat.com/archives/fedora-extras-commits/2006-November/date....
We require your help to verify that all checkins are properly in
/cvs/extras and everything generally is working as expected. Please pay
careful attention to everything that changed in November 16th and 17th
according to the mailing list archive.
My personal records indicate that no CVS imports happened after this
backup, so we are fine in that regard.
PACKAGE OWNERS - especially kevin, corsepiu, remi and qspencer are to
verify and fix their own CVS modules, or explicitly identify someone
else to do it on their behalf.
EVERYONE ELSE - please keep an eye on CVS. SCREAM LOUDLY on
fedora-extras-list if you see anything that is wrong.
Sorry about any inconvenience or disruption that this unexpected problem
may have caused. Everyone owes a beer to the Fedora Infrastructure and
Red Hat IS/IT teams. =)
Thank you for supporting the Fedora Project.
Warren Togami
wtogami(a)redhat.com
17 years, 5 months