Any known bugs in gcc wrt symbols in gdb and/or var allocation ?
by Kim Lux
I think I've found 2 issues with gcc-3.4.2-6.fc3.
Consider the following code:
<code snippet begins>
void dbug12_stop_reason (enum dbug12_stop *reason, int *sigrc)
{
int replyEnd;
unsigned char reply[254];
replyEnd = 0;
// get the reply
replyEnd = dbug12_get_reply(reply);
printf("Reply is %s\n",reply);
// handle the error condition
// can't pass here with a zero length
if (replyEnd == 0)
...
<code snippet ends>
Problem #1: gdb cannot find replyEnd:
(gdb) whatis replyEnd
No symbol "replyEnd" in current context.
Problem #2:
The code doesn't execute properly if I delete the "replyEnd = 0;" line.
Without "replyEnd = 0;" in the code, it gets an unalterable junk value.
With "replyEnd = 0;" in the code, replyEnd gets set to the return value
of dbug12_get_reply, which it should.
BTW: the definition of dbug12_get_reply is:
static int dbug12_get_reply (unsigned char *packet)
I was also playing around with changing the size of the reply buffer, ie
"reply[255]", reply[253], etc. It didn't look like gcc was changing the
size of the buffer on a clean build.
It looks like gcc isn't allocating something properly.
Has anyone seen anything like this before ?
rpm -q gdb
gdb-6.1post-1.20040607.43
Thanks
--
Kim Lux, Diesel Research Inc.
19 years, 4 months
RAM detection error: 32-bit int issue?
by Scott Saccone
FC2/x86_64 (2.6.5-1.358smp) ran fine with 2GB RAM, but after installing
another 2GB it only detects 3GB. GRUB 0.94 also only detects 3GB. I
noticed in the source code that GRUB is using int to store certain
quantities related to RAM. There is also a bug report at
http://savannah.gnu.org/bugs/?func=detailitem&item_id=9954 related to
this. Could this my problem? If GRUB has a bug, would this affect linux?
Can I fix this by simply updating GRUB?
>From /var/log/dmesg:
BIOS-provided physical RAM map:
BIOS-e820: 0000000000000000 - 000000000009b800 (usable)
BIOS-e820: 000000000009b800 - 00000000000a0000 (reserved)
BIOS-e820: 00000000000e0000 - 0000000000100000 (reserved)
BIOS-e820: 0000000000100000 - 00000000bff60000 (usable)
BIOS-e820: 00000000bff60000 - 00000000bff6f000 (ACPI data)
BIOS-e820: 00000000bff6f000 - 00000000bff80000 (ACPI NVS)
BIOS-e820: 00000000bff80000 - 00000000c0000000 (reserved)
BIOS-e820: 00000000e0000000 - 00000000f0000000 (reserved)
BIOS-e820: 00000000fec00000 - 00000000fec10000 (reserved)
BIOS-e820: 00000000fee00000 - 00000000fee01000 (reserved)
BIOS-e820: 00000000ff800000 - 00000000ffc00000 (reserved)
BIOS-e820: 00000000fffffc00 - 0000000100000000 (reserved)
BIOS-e820: 0000000100000000 - 0000000140000000 (usable)
...
Memory: 3096192k/3145088k available (1879k kernel code,
0k reserved, 1326k data, 176k init)
I processed the above:
start finish start-dec finish-dec dec-width MB
0 9b800 0 636928 636928 0.607 u
9b800 a0000 636928 655360 18432 0.018 r
e0000 100000 917504 1048576 131072 0.125 r
100000 bff60000 1048576 3220570112 3219521536 3070.375 u
bff60000 bff6f000 3220570112 3220631552 61440 0.059 r
bff6f000 bff80000 3220631552 3220701184 69632 0.066 r
bff80000 c0000000 3220701184 3221225472 524288 0.500 r
e0000000 f0000000 3758096384 4026531840 268435456 256.000 r
fec00000 fec10000 4273995776 4274061312 65536 0.062 r
fee00000 fee01000 4276092928 4276097024 4096 0.004 r
ff800000 ffc00000 4286578688 4290772992 4194304 4.000 r
fffffc00 100000000 4294966272 4294967296 1024 0.001 r
100000000 140000000 4294967296 5368709120 1073741824 1024.000 u
Usable=4293900288 (4094.982421875 MB)
Not usable=273505280 (260.8349609375 MB)
Total=4567405568 (4355.8173828125 MB)
The addresses in the last line exceed 2^32. Why is that? Do some of
these addresses refer to something besides my RAM?
SPECS:
CPU: 2 x 3.2GHz Xeon Nocona
MOBO: Tyan Thunder i7525 (S2676)
RAM: 4GB
2 x 1GB Corsair Reg DDR2
4 x 512MB Corsair Reg DDR2
BIOS: S2676 Rev 2.00 (9/28/04)
Linux: 2.6.5-1.358smp
Distro: FC2 x86_64
GRUB: GNU GRUB 0.94
gcc: 3.3.3 20040412 (Red Hat Linux 3.3.3-7)
19 years, 4 months
rawhide report: 20050122 changes
by Build System
New package ksh
The Original ATT Korn Shell
Removed package asp2php
Updated Packages:
Guppi-0.40.3-22
---------------
* Fri Jan 21 2005 Bill Nottingham <notting(a)redhat.com> 0.40.3-22
- fix some random oddness (#114374)
PyQt-3.13-3
-----------
* Fri Jan 21 2005 Than Ngo <than(a)redhat.com> 3.13-3
- Add missing sip dependency
cups-1:1.1.23-5
---------------
* Fri Jan 21 2005 Tim Waugh <twaugh(a)redhat.com> 1.1.23-5
- Use tmpwatch to remove unused files in the spool temporary directory
(bug #110026).
* Thu Jan 20 2005 Tim Waugh <twaugh(a)redhat.com>
- Use gzip's -n flag for the PPDs.
gdb-6.3.0.0-0.10
----------------
* Fri Jan 21 2005 Jeff Johnston <jjohnstn(a)redhat.com> 6.3.0.0-0.10
- Fix to prevent resetting unwind kernel table size to invalid value
when debugging a core file
- Bugzilla 145309
* Fri Jan 21 2005 Andrew Cagney <cagney(a)redhat.com> 6.3.0.0-0.9
- When single stepping handle both back-to-back and nested signals.
- Disable .symbol patch, results in BFD errors.
* Fri Jan 21 2005 Jeff Johnston <jjohnstn(a)redhat.com> 6.3.0.0-0.8
- Support listing both in-charge and not-in-charge dtors when
just the dtor name is given.
- Add new test case for newly added ctor/dtor functionality.
jpilot-0.99.7-5
---------------
* Fri Jan 21 2005 Ivana Varekova <varekova(a)redhat.com> 0.99.7-5
- fix problem with previous patch (problem with cb_cal_changed connection)
- fix problem with add_new_record (problem with cb_cal_changed connection)
* Mon Jan 10 2005 Ivana Varekova <varekova(a)redhat.com> 0.99.7-4
- fix part of bug #142520 - problem with Go to Today
* Mon Nov 22 2004 Ivana Varekova <varekova(a)redhat.com> 0.99.7-3
- fix bug #139377 - problem with x86_64
kernel-2.6.10-1.1107_FC4
------------------------
* Fri Jan 21 2005 Dave Jones <davej(a)redhat.com>
- Rebase to 2.6.11-rc2
* Fri Jan 21 2005 Rik van Riel <riel(a)redhat.com>
- make exec-shield segment limits work inside the xen kernels
lftp-3.0.13-1
-------------
* Fri Jan 21 2005 Jason Vas Dias <jvdias(a)redhat.com> 3.0.13-1
- Upgrade to upstream version 3.0.13 .
lvm2-2.01.02-1.0
----------------
* Fri Jan 21 2005 Alasdair Kergon <agk(a)redhat.com> - 2.01.02-1.0
- Minor fixes.
mailcap-2.1.18-1
----------------
* Fri Jan 21 2005 Bill Nottingham <notting(a)redhat.com> 2.1.18-1
- add iso, img to octet-stream (#142459 <ville.skytta(a)iki.fi>)
memtest86+-1.50-1
-----------------
* Fri Jan 21 2005 Warren Togami <wtogami(a)redhat.com> - 1.50-1
- 1.50
microcode_ctl-1:1.11-1.18
-------------------------
* Fri Jan 21 2005 Dave Jones <davej(a)redhat.com>
- Create/remove the /dev/cpu/microcode dev node as needed.
- Use correct path again for the microcode.dat.
- Remove some no longer needed tests in the init script.
mkinitrd-4.2.0.3-1
------------------
* Fri Jan 21 2005 Peter Jones <pjones(a)redhat.com> - 4.2.0.3-1
- Make nash expand environment variables on command lines (#144474)
- Make nash check pids returned from wait*() (#145660)
* Fri Jan 21 2005 Peter Jones <pjones(a)redhat.com> - 4.2.0.2-1
- Make getArg return NULL on cmd=NULL, fixing #144472 . Based on a
patch from Kasper Dupont.
openoffice.org-1.1.3-2.7
------------------------
* Fri Jan 21 2005 Dan Williams <dcbw(a)redhat.com> - 1.1.3-2
- Update to latest ooo-build.
- NOTE: there may still be some i18n text-entry bugs when using
IIIMF input modules.
* Thu Jan 06 2005 Dan Williams <dcbw(a)redhat.com> - 1.1.3-1
- Update to OpenOffice.org 1.1.3, latest ooo-build
* Sun Dec 12 2004 Dan Williams <dcbw(a)redhat.com> - 1.1.2-18
- #rh137854# Japanese document from Windows is heavily broken
(Pass #2, much better fix this time around)
- #rh129719# Upstream indic translations need to be incorporated into OOo packages
(Work around translation bugs, most Gujarati UI strings should be native now)
parted-1.6.21-1
---------------
* Fri Jan 21 2005 Chris Lumens <clumens(a)redhat.com> 1.6.21-1
- Updated to 1.6.21
pciutils-2.1.99.test8-5
-----------------------
* Fri Jan 21 2005 Bill Nottingham <notting(a)redhat.com> - 2.1.99.test8-5
- fix domain bug (#138722, #144383)
rp-pppoe-3.5-26
---------------
* Sat Jan 22 2005 Than Ngo <than(a)redhat.com> 3.5-26
- rename config files #145255
rpmdb-fedora-1:4-0.20050122
---------------------------
selinux-policy-strict-1.21.2-7
------------------------------
* Thu Jan 20 2005 Dan Walsh <dwalsh(a)redhat.com> 1.21.2-7
- Allow restorecon and setfiles to read default_context_t
* Thu Jan 20 2005 Dan Walsh <dwalsh(a)redhat.com> 1.21.2-6
- Remove crond alias to unconfined_t in targeted
selinux-policy-targeted-1.21.2-6
--------------------------------
* Thu Jan 20 2005 Dan Walsh <dwalsh(a)redhat.com> 1.21.2-6
- Remove crond alias to unconfined_t in targeted
telnet-1:0.17-33
----------------
* Fri Jan 21 2005 Harald Hoyer <harald(a)redhat.com> - 1:0.17-33
- added patch telnetd-0.17-pty_read.patch, which fixes 145636
usbutils-0.11-6.2
-----------------
* Thu Jan 20 2005 David Woodhouse <dwmw2(a)redhat.com> 0.11-6.2
- Don't byteswap parts of device descriptor which kernel already swapped
xloadimage-4.1-33
-----------------
* Fri Jan 21 2005 Bill Nottingham <notting(a)redhat.com> 4.1-33
- fix bad use of format strings (#70867, #78481 <d.binderman(a)virgin.net>)
xorg-x11-6.8.1.902-4
--------------------
* Fri Jan 21 2005 Mike A. Harris <mharris(a)redhat.com> 6.8.1.902-4
- Renamed the xorg-x11-6.8.0-AMD64-enable-i810-driver.patch patch to
xorg-x11-6.8.0-AMD64-override-default-driverlist.patch and added the "vmware"
driver to the default list of drivers to be build on AMD64. (#145588)
* Thu Jan 20 2005 Mike A. Harris <mharris(a)redhat.com> 6.8.1.902-3
- Added xorg-x11-6.8.1.902-xf86pcibus-PCIX-bar-64bit-fix.patch to attempt to
fix bugs (#140601, 143910)
* Mon Jan 17 2005 Mike A. Harris <mharris(a)redhat.com> 6.8.1.902-2
- Added xorg-x11-6.8.1.902-ia64-hp-zx2-support-bug-119364.patch to add support
for new Hewlett Packard IA64 ZX2 chipset. (#119364)
19 years, 4 months
RE: RFC: Optimizing for 386
by Joseph D. Wagner
> If you think there will be a performance gain, then rebuild /one/
> RPM with the optimizations you want and show how much faster it is.
I actually did that once. I recompiled KDE from scratch. 36 hours. No noticeable difference. Then I recompiled the underlying Qt libraries from scratch. 5 hours. A few seconds here, a few seconds there. Then I recompiled XFree86 from scratch. 4 hours. Now we're starting to talk increased responsiveness.
Then XFree86 became Xorg, and both Qt and KDE came out with a new version of there programs, and everything had to be recompiled from scratch all over again.
The problem is EVERYTHING from the ground up has to be optimized to feel the cumulative effect. There is no way you are going to convince me that being optimized for a 386 is OK when I felt the effectiveness of optimizing the graphics programs.
However, because I didn't do my tests in a "scientific" manor with precise measurements, the developers won't even consider my results.
> the developers explain over & over that there won't be
> these performance gains people think there will be
These same people say that theoretically EXT3 with Data Ordered mode should be faster that EXT3 with Data Journaled mode, but reality has proven that this isn't the case. I'd like just one of these developers to take the hours of compiling time necessary to produce an optimized version and test the results. I think the results would be equally surprising.
> In sum, "the huge numbers of people" are asking for something
> that... will hurt others
Who? Who is still using a 386? A 486? Do anyone on this list have anything less than a 586? Who are these victims of my push for better optimizations?
Joseph D. Wagner
19 years, 4 months
Evolution 2.2, OpenOffice 2.0, Gnome 2.10
by Trever L. Adams
I believe I saw these on a list for inclusion in FC4. However, there
have been no signs of it yet.
OO 2 was delayed a bit, so I can see if this was dropped, though I hope
it wasn't.
Evolution 2.1.x should go beta for 2.2 very shortly (may have happened,
but release not done yet).
Gnome 2.10 is on more or less the same schedule as 2.1.x.
When, if, are we going to see these in rawhide?
Is there anything I can do?
Trever
--
"He that demands mercy, and shows none, ruins the bridge over which he
himself is to pass." -- Thomas Adams, 1612-1653
19 years, 4 months
Is there a working swsusp for FC3 2.6.10?
by Jason Haar
Hi there
I've just dropped-kicked SuSE-9.1 off my new HP NX5000 (I've been using
Redhat too long to change, and can't live without yum :-) and put FC3 on
it. I've got WiFi working (madwifi), but one thing SuSe 9.1 had that I
can't live without was the swsusp support. It was great! Hit the power
button, and 10-15 sec later the laptop had written RAM to disk and
turned itself off. Hit the power button again, and 15-20 sec later you
were back where you started. Cool :-)
Anyway, I tried installing vanilla 2.6.10 patched with the current
swsusp, but it fell all over itself while booting. Laptops are so fickle!
However, it likes FC3 well enough (i.e. one of the 1,000 kernel patches
in it must have fixed something to do with my laptop hardware), so has
anyone got a src.rpm or a patch-for-the-swsusp-patch that'll get the
current FC3 kernel playing nicely with swsusp?
Thanks!
--
Cheers
Jason Haar
Information Security Manager, Trimble Navigation Ltd.
Phone: +64 3 9635 377 Fax: +64 3 9635 417
PGP Fingerprint: 7A2E 0407 C9A6 CAF6 2B9F 8422 C063 5EBB FE1D 66D1
19 years, 4 months
RFC: Optimizing for 386
by Joseph D. Wagner
I can count the total number of people in the world who still use a 386 on the fingers of both of my hands. Why are we still catering to this small group of people?
With the exception of the kernel and glibc, all RPM's are optimized for 386 architecture. This is a waste of system resources. Study after study shows that you can achieve a 10% - 40% performance improvement my optimizing code for a specific architecture. Windows XP may only be optimized for a Pentium, but by golly, at least the whole thing is optimized, not just the kernel and the C library.
Just look at X. Is anyone seriously trying to get X to run on a 386? I can understand compiling the text-based programs, like bash, for 386. You can run a text-only box on a 386 just fine. Why do that for X? Why do that for GNOME, KDE, or any graphical program for that matter?
I know that I can recompile of these program from the source code to achieve those optimizations. However, why should everyone who wants to optimize their systems have to go through that, just so a handful of people with 386 machines can run X out-of-the-box? Then, we have recompile all over again with the next RPM release.
Why can't the few people who have a 386 be made to recompile X from sources to get it to run on there machines, so the rest of us can enjoy the performance boost from running optimized binaries?
I think we seriously need to rethink the distribution strategy. At the very least, all graphical programs should be optimized for i686.
Joseph D. Wagner
19 years, 4 months
rawhide report: 20050120 changes
by Build System
Updated Packages:
GConf2-2.8.1-2
--------------
* Wed Jan 19 2005 Mark McLoughlin <markmc(a)redhat.com> 2.8.1-2
- Backport some fixes from upstream CVS
Maelstrom-3.0.6-7
-----------------
* Wed Jan 19 2005 Karsten Hopp <karsten(a)redhat.de> 3.0.6-7
- add patch from Andrew Church
anaconda-10.2.0.12-1
--------------------
* Wed Jan 19 2005 Chris Lumens <clumens(a)redhat.com> 10.2.0.12-1
- Fix partitioning bugs (#101432, #137119)
- Support --bytes-per-inode on a per-partition basis (#57550)
apr-util-0.9.5-3
----------------
* Wed Jan 19 2005 Joe Orton <jorton(a)redhat.com> 0.9.5-3
- restore db-4.3 detection lost in 0.9.5 upgrade
* Wed Jan 19 2005 Joe Orton <jorton(a)redhat.com> 0.9.5-2
- rebuild
cups-1:1.1.23-3
---------------
* Wed Jan 19 2005 Tim Waugh <twaugh(a)redhat.com> 1.1.23-3
- Applied patch to fix CAN-2005-0064.
doxygen-1:1.4.1-1
-----------------
* Wed Jan 19 2005 Than Ngo <than(a)redhat.com> 1:1.4.1-1
- update to 1.4.1
glibc-kernheaders-2.4-9.1.87
----------------------------
* Tue Aug 31 2004 Ajran van de Ven <arjanv(a)redhat.com>
- update input.h
* Thu Jul 29 2004 Arjan van de Ven <arjanv(a)redhat.com>
- update limits.h to reflect NR_GROUPS change
* Thu Jul 22 2004 Jakub Jelinek <jakub(a)redhat.com>
- update syscall numbers from latest kernel on alpha/ia64
gnome-python2-2.6.0-5
---------------------
* Wed Jan 19 2005 Mark McLoughlin <markmc(a)redhat.com> - 2.6.0-5
- Backport wrapping for GConfEngine from upstream CVS
initscripts-8.04-1
------------------
* Wed Jan 19 2005 Bill Nottingham <notting(a)redhat.com> 8.04-1
- split out ifup/ifdown general case to ifup/ifdown-eth;
add ifup/ifdown-bnep (<dwmw2(a)redhat.com>)
- ifup-ipsec: add fwd policies (#145507)
- fix multiple scsi_hostadapter loads (#145432)
- enable syncookies in sysctl.conf (#145201)
kernel-2.6.10-1.1103_FC4
------------------------
* Wed Jan 19 2005 Dave Jones <davej(a)redhat.com>
- Re-add diskdump/netdump based on Jeff Moyers patches.
- Rebase to -bk7
* Tue Jan 18 2005 Jeremy Katz <katzj(a)redhat.com>
- fixup xen0 %post to use new grubby features for multiboot kernels
- conflict with older mkinitrd for kernel-xen0
* Tue Jan 18 2005 Dave Jones <davej(a)redhat.com>
- -bk6
lvm2-2.01.01-1.0
----------------
* Mon Jan 17 2005 Alasdair Kergon <agk(a)redhat.com> - 2.01.01-1.0
- Update vgcreate man page. Preparation for snapshot origin extension fix.
nvi-m17n-1.79-20040401.21
-------------------------
* Thu Jan 20 2005 Akira TAGOH <tagoh(a)redhat.com> - 1.79-20040401.21
- updates m17n patch to 20040401.
- added iso8859-15 support.
octave-6:2.1.57-9
-----------------
* Wed Jan 19 2005 Ivana Varekova <varekova(a)redhat.com> 2.1.57-9
- Fix bug #142440 - change octave.spec: autoconf is BuildPrereq
- Fix bug #142631 - change octave.spec: mkoctfile.1.gz is part of octave-devel not octave
perl-3:5.8.6-2
--------------
* Tue Jan 18 2005 Chip Turner <cturner(a)redhat.com> - 3:5.8.6-2
- bugzilla: 145448, fix invalid utf8 in changelog
* Tue Jan 18 2005 Chip Turner <cturner(a)redhat.com> - 3:5.8.6-1
- bugzilla: 145447, add 5.8.5 to perlmodcompat list
rhpl-0.152-1
------------
* Wed Jan 19 2005 Jeremy Katz <katzj(a)redhat.com> - 0.152-1
- fix build with new glibc-kernheaders (#145442)
rpmdb-fedora-1:4-0.20050120
---------------------------
selinux-policy-strict-1.21.2-2
------------------------------
* Wed Jan 19 2005 Dan Walsh <dwalsh(a)redhat.com> 1.21.2-2
- Fixed policy for telnet and rlogin
* Tue Jan 18 2005 Dan Walsh <dwalsh(a)redhat.com> 1.21.2-1
- Update with latest from NSA
- Add Policy man pages
selinux-policy-targeted-1.21.2-4
--------------------------------
* Wed Jan 19 2005 Nalin Dahyabhai <nalin(a)redhat.com> 1.21.2-4
- Add a default_context entry for the remote_login_t domain, for telnet
* Wed Jan 19 2005 Dan Walsh <dwalsh(a)redhat.com> 1.21.2-2
- Fixed policy for telnet and rlogin
* Tue Jan 18 2005 Dan Walsh <dwalsh(a)redhat.com> 1.21.2-1
- Update with latest from NSA
- Add Policy man pages
squid-7:2.5.STABLE7-2
---------------------
* Tue Jan 18 2005 Jay Fenlason <fenlason(a)redhat.com> 7:2.5.STABLE7-2
- Add a triggerin on samba-common to make /var/cache/samba/winbindd_privileged
accessable so that ntlm_auth will work. It needs to be in this rpm,
because the Samba RPM can't assume the squid user exists.
Note that this will only work if the Samba RPM is recent enough to create
that directory at install time instead of at winbindd startup time.
That should be samba-common-3.0.0-15 or later.
This fixes bugzilla #103726
- Clean up extra whitespace in this spec file.
- Add additional upstream patches. (Now 18 upstream patches).
- patch #112 closes CAN-2005-0096 and CAN-2005-0097, remote DOS security holes.
- patch #113 closes CAN-2005-0094, a remote buffer-overflow DOS security hole.
- patch #114 closes CAN-2005-0095, a remote DOS security hole.
- Remove the -nonbl (replaced by #104) and -close (replaced by #111) patches, since
they're now fixed by upstream patches.
tetex-2.0.2-30
--------------
* Wed Jan 19 2005 Jindrich Novy <jnovy(a)redhat.com> 2.0.2-30
- Fix CAN-2005-0064 xpdf buffer overflow
x3270-3.3.2.p1-10
-----------------
* Wed Jan 12 2005 Tim Waugh <twaugh(a)redhat.com> 3.3.2.p1-10
- Rebuilt for new readline.
* Wed Dec 08 2004 Karsten Hopp <karsten(a)redhat.de> 3.3.2.p1-9
- add icon (#141599, #125577)
- fix variable usage (local variable overwrite) (#116660)
* Wed Dec 08 2004 Karsten Hopp <karsten(a)redhat.de> 3.3.2.p1-8
- rebuild
19 years, 4 months
DM Session cleanup
by Peter Backlund
Hello.
Could /usr/share/xsessions/gnome.desktop be moved to gnome-session,
and /etc/X11/dm/Sessions/gnome.desktop and /etc/X11/gdm/Sessions/GNOME
be removed completely, for consistency?
/Peter
19 years, 4 months
Fedora updates CD-ROM?
by Gérard Milmeister
I recently installed FC3 and had to go through the update procedure and
download several hundreds of megabytes.
I could imagine that once or twice an Updates CD-ROM would be released.
The Fedora installer might ask after an installation for the user to
insert an Updates CD-ROM if available, and perform the required updates.
--
Gérard Milmeister
Langackerstrasse 49
CH-8057 Zürich
19 years, 4 months