Re: Is AMD-64 mature enough to start using? (possibly on K8V-MX moptherboard)
Am Mittwoch, den 28.12.2005, 09:05 + schrieb Matthew Robinson: > I was under the impression apple used 'fat binaries' to make their code work > with the G3 and their previous line of processors Ancient Apple machines have an Motorola cpu like 68030. These are not PPC machines and need "68K" code. However, I haven't seen such a thing for years, except in a don't-drop-oldtimers-section in my company. However, we keep them, but we don't use them. When Apple changed to PPC, the G3 was not the first PPC-cpu. I still use some pre-G3-PPC-machines under linux for networking tasks. "Fat Binaries" include splitted 68K code and PPC code. The do not contain splittet code for PPC=G3. Apple has used that technology multiple times since then. Later there was i.e. a Version of Photoshop, that run under OS 9 and OS X, the same executable. And in the future there will be PPC/Intel combinations. Bye, Ratti -- -o) fontlinge | Fontmanagement for Linux | Schriftenverwaltung in Linux /\\ http://freshmeat.net/projects/fontlinge/ _\_V http://www.gesindel.de https://sourceforge.net/projects/fontlinge/ signature.asc Description: This is a digitally signed message part
Re: lockup at boot with kernel 2.6.14...
Am Freitag, den 11.11.2005, 16:37 -0200 schrieb Eduardo M KALINOWSKI: > I have an Asus motherboard (K8U-X, and it seems nobody else has this > brand) and it has the "EZ Flash" feature in which you can update the > bios by copying the image to a floppy and pressing a key (Alt-F2) during > boot. Perhaps your motherboard has this feature too. My Asus K8V Deluxe has this feature also, however, not every ASUS board can do this. I have updated my Bios several times that way and it worked perfect (except that you have to reconfigure the complete bios... grmbl...) However, I had ASUS boards before that could not upgrade that way (i.e. my old A7V), but there has always been a different way. Asus provides DOS-Tools (without warranty), and I used a free DOS (Freedos?) from the Web. This worked also. However, be careful loading bootdisks from the web. Some years ago I just took a bootdisk from bootdisk.com, and the disk contained a virus that erases the bios. I had to buy a new chip. Unbelievable - I ran a M$ OS once in five years and i catch malware. =%-) Bye, Ratti P.S.: I really can recommend Asus. Even they don't care Linux officially, their "inoffical" support always worked perfect for me. No, I don not work for them nor do I own parts of that company. :-) -- -o) fontlinge | Fontmanagement for Linux | Schriftenverwaltung in Linux /\\ http://freshmeat.net/projects/fontlinge/ _\_V http://www.gesindel.de https://sourceforge.net/projects/fontlinge/ signature.asc Description: This is a digitally signed message part
Re: lockup at boot with kernel 2.6.14...
Hi, Am Donnerstag, den 10.11.2005, 12:50 +0100 schrieb Giacomo Mulas: > Hello, I am experiencing a very strange behaviour on my Asus A6k > Turion64 laptop running sid. It used to work ok with a custom-compiled (from > the debian-packaged source) 2.6.12 kernel. When the debian-packaged source > for 2.6.14 became available, I copied over my old .config, did a make > oldconfig answering the relevant questions, built a new kernel without any > errors, installed it. > Then I experienced this weird situation: the system boots, loads the > ramdisk, does its initial setup, mounts the root filesystem, switches > root... all ok until it gets to starting udev. Then udev's hotplug kicks in > and starts loading modules and... in an erratic, unpredictable way, it hangs > while loading the usb hcd module, without any error message. Sometimes it > does, sometimes it doesn't and the boot procedure then goes smooth and the > computer runs without a hitch thereafter. It does not happen at all with > 2.6.12, therefore it must be something which changed between 2.6.12 and > 2.6.14 (a shipload of things, unfortunately...). I was unable to relate the > hangs to anything, they seem to be just random. The only thing which is > reproducible is that the system always hangs (when it does) while loading > the ehci_hcd module. I did not even file a bug report yet, since I don't > quite understand whether this is a kernel or a udev issue. Same here on sid/x86. Please scroll up using shift-PgUp. It looked the same here: The machine bootet into a kind of "emergency shell" with limited features. So I get a prompt. But then(!) the fresh loaded USB-modules produce a lot of output, so the real problem ist described four pages up. The last messages you can see on screen is that output. I accidently pressed return and was very surprised to get a prompt. The problem for me was that kernel2.6.14/udec/initramfd/whoever changes my drives order while booting. Instead of sda1 sdb1 sdc1,2,3 ...I suddenly had: sda1,2,3 sdb1 sdc1 I managed to get my machine kind-of booting by building a new kernel, where I included lots of modules for board, ide, sata, fs,... Bye, Ratti -- -o) fontlinge | Fontmanagement for Linux | Schriftenverwaltung in Linux /\\ http://freshmeat.net/projects/fontlinge/ _\_V http://www.gesindel.de https://sourceforge.net/projects/fontlinge/ signature.asc Description: This is a digitally signed message part
Re: nvidia driver overrides gdm.conf AutomaticLogin !?? SOLVED!
Am Freitag, den 07.10.2005, 15:57 -0700 schrieb Ali H. Caliskan: > Mike was right about modules-assist, since the new > nvidia-kernel that I compiled manually(make-kpkg > modules_image) works fine with gnome login manager! Wow. I had the problem for some weeks under x86 and nobody could help. Great not to be alone. :-) I've discovered funny stuff: 1. When I kill gdm and restart it as root from the terminal - it works. 2. Sometimes, when I just want to finish my session, I am immediately automatically re-login'ed. So, the mechanism isn't totally broken. It works sometimes. I've found a very simple solution: Don't use the normal autologin feature. Use the one that logs in a user after X seconds. I use 2 seconds and it works. Since I have no second user this is acceptable and less work as reinstalling this and that. Bye, Ratti -- -o) fontlinge | Fontmanagement for Linux | Schriftenverwaltung in Linux /\\ http://freshmeat.net/projects/fontlinge/ _\_V http://www.gesindel.de https://sourceforge.net/projects/fontlinge/ signature.asc Description: This is a digitally signed message part
Re: Upgrading to current udev is disastrous if not on kernel 2.6.12
Hi, Am Montag, den 18.07.2005, 14:19 +0200 schrieb GOMBAS Gabor: > On Sat, Jul 16, 2005 at 12:33:12PM +0200, Joerg Rossdeutscher wrote: > > > Packages should not depend on any kernel, since many people run their > > own. However, I just don't understand why the package has been published > > at all, since even in experimental there is no 2.6.12. > > Because people already using 2.6.12 need it, since the previous udev > (more precisely, libsysfs that udev uses) had a bug which was exposed by > the 2.6.12 kernel. If they use a non-distribution-kernel, then they can also build a non-distribution-udev. It's not useful to break the distribution for that reason. Bye, ratti -- -o) fontlinge | Fontmanagement for Linux | Schriftenverwaltung in Linux /\\ http://freshmeat.net/projects/fontlinge/ _\_V http://www.gesindel.de https://sourceforge.net/projects/fontlinge/ signature.asc Description: This is a digitally signed message part
Re: Upgrading to current udev is disastrous if not on kernel 2.6.12
Am Samstag, den 16.07.2005, 17:22 +1000 schrieb Hamish Moffatt: > I can't imagine how it's a good idea to upload a package which depends > on a kernel not available for Debian yet. Packages should not depend on any kernel, since many people run their own. However, I just don't understand why the package has been published at all, since even in experimental there is no 2.6.12. Bye, Ratti -- -o) fontlinge | Fontmanagement for Linux | Schriftenverwaltung in Linux /\\ http://freshmeat.net/projects/fontlinge/ _\_V http://www.gesindel.de https://sourceforge.net/projects/fontlinge/ signature.asc Description: This is a digitally signed message part
Re: Filesystem stability?
Hi, Am Montag, den 11.07.2005, 18:18 -0700 schrieb Mark Ferlatte: > I've heard rumours that some of the Linux filesystems aren't as stable > as they should be on AMD64; in particular, I've heard of bad things > happening with JFS and XFS. > > That being said, I can't find anything even approaching authoritative, > so I thought I'd ask: > > What filesystems are you guys using, and anyone had any bad experiences? I always used XFS on 32bit and was happy with it. Some month ago I installed an Ubuntu/64bit with XFS as IMAP/POP3 server. After some weeks the machine crashed, since I had misconfigured a spamfilter that ate up all ressources. The filesystem was damaged, and I repaired it from a rescue CD. Since some binaries appeared in /lost +found, I --reinstall'ed all installed packages, and everything was fine. My faults until to here. Two weeks later the machine crashed again. This time I found a message in syslog that said XFS-Data had been corrupted in RAM and the filesystem would be remounted read-only. That broke some processes, so I could not login anymore. I shutdown by sysrequest-keys. Again, the filesystem was damaged. I ran a Ramtest from a CD. No problems. Two weeks later the machine crashed again. Same message in syslog. Filesystem damaged. I decided to run the machine without fam (since I had no better idea) The machine then worked for another two weeks, then I did a kernel-update and rebooted. So, the countdown is reset to start... I do not want to blame XFS, but that looks suspicious to me. Bye, Ratti -- -o) fontlinge | Fontmanagement for Linux | Schriftenverwaltung in Linux /\\ http://freshmeat.net/projects/fontlinge/ _\_V http://www.gesindel.de https://sourceforge.net/projects/fontlinge/ signature.asc Description: This is a digitally signed message part
Re: Asus A8V Deluxe, Xfree display problems with BIOS 1011
Hi, I know, this thread is 6 weeks old, but... solved. Finally. On Thu, 2005-05-05 at 15:08 +0200, Joerg Rossdeutscher wrote: > Am Mittwoch, den 04.05.2005, 15:41 + schrieb Jean-Luc Coulon > > > I've an ASUS A8V Deluxe with an ASUS A9250T/D video card. > > I've just updated my BIOS from 1009 to 1011. > > > > After this update, the machine boot fine and the display works in > > cosole mode but while swtiching to X when gdm starts, I get a black > > screen with only the cursor image in white. Me2 with an K8V Deluxe. Today a beta(!)-Bios was released that seems to fix the problem. I can see my desktop again, and glxgears runs fine. ftp://ftp.asuscom.de/pub/ASUSCOM/BIOS/Socket_754/VIA_Chipset/K8T800/K8V_Deluxe/ 1011d_02.zip Bye, Ratti -- -o) fontlinge | Fontmanagement for Linux | Schriftenverwaltung in Linux /\\ http://freshmeat.net/projects/fontlinge/ _\_V http://www.gesindel.de https://sourceforge.net/projects/fontlinge/ signature.asc Description: This is a digitally signed message part
Re: Asus A8V Deluxe, Xfree display problems with BIOS 1011
Am Montag, den 09.05.2005, 19:12 + schrieb Jean-Luc Coulon (f5ibh):
> Le 09.05.2005 20:55:17, Joerg Rossdeutscher a écrit :
> > > And if your bios chipset is corrupt ,the main boot block of bios is
> > > damaged ,
> > Crazy.
> >
> > The problem is: Their bios breaks the nvidia drivers. And these guy
> > tells me something about broken eeproms and flashing the wrong bios.
>
> Yes really crazy.
> In your case, probably you got the answer for the question of an other
> guy...
No, he quoted my mail, I just didn't post that very long part into the
list. It just would make it even more ridiculous ("I have an Asus K8V
deluxe..." "What board do you have?")
Bye, Ratti
--
-o) fontlinge | Fontmanagement for Linux | Schriftenverwaltung in Linux
/\\ http://freshmeat.net/projects/fontlinge/
_\_V http://www.gesindel.de https://sourceforge.net/projects/fontlinge/
signature.asc
Description: This is a digitally signed message part
Re: Asus A8V Deluxe, Xfree display problems with BIOS 1011
Hi, Am Mittwoch, den 04.05.2005, 15:41 + schrieb Jean-Luc Coulon (f5ibh): > I've asked Asus support, the reply (In French) is that "they have no > UNIX/linux support for the moment, sorry for the inconvenient". I've also contacted Asus. The answer I got today is... bizzare: > Dear Friend : > Thank you for contacting ASUS Customer Service. > My name is and I would be assisting you today > > Sir, What's the motherboard do you have ? A8V deluxe ? K8V deluxe ? or > K8V ? do you have use other motherboard bios (eg , k8v deluxe) to > update A8V deluxe . Thanks a lot for your information . by the way , > you don't use other motherboard bios to update this motherboard . Or > it will kill your bios chipset . > > And if your bios chipset is corrupt ,the main boot block of bios is > damaged , > > so you can not re-flash it back because the system won't boot . > > if you have a eeprom programmer you can use it to write bios chip > back ; > > If you have another ASUS board( same type and same PCB version) ,you > can use "hot-swap" method to write bios back . > > > if above 2 ways are all impossible ,you need contact your mother board > dealer to send this faulty board to! repaired . > > > > Best regards ! Crazy. The problem is: Their bios breaks the nvidia drivers. And these guy tells me something about broken eeproms and flashing the wrong bios. Rated zero. Bye, Ratti -- -o) fontlinge | Fontmanagement for Linux | Schriftenverwaltung in Linux /\\ http://freshmeat.net/projects/fontlinge/ _\_V http://www.gesindel.de https://sourceforge.net/projects/fontlinge/ signature.asc Description: This is a digitally signed message part
Re: Asus A8V Deluxe, Xfree display problems with BIOS 1011
Hi,
Am Freitag, den 06.05.2005, 09:34 -0400 schrieb Lennart Sorensen:
> Upgrading is for fixing problems.
I do not agree with you. That may have been that way in the 90s, when a
Bios was "a program that boots your OS". Nowadays technology is moving
very fast, and I want to use modern features like powermanagement,
Firewire, Instant-On, ... This machine is not a server, it's my
fun-at-home-computer. If the bios breaks - it's not a big deal.
Asus-bords can flash a new Bios at boot-time directly from a 3,5"-disk
without needing any OS or software, and they can fallback on an
emergency-Bios in case of problems. So I install any Bios-Upgrade
immediately.
When I boot into linux, some of the modules ouptput "unhappy" messages
about the hardware ("...implementation of blabla is broken, tell your
vendor to fix, get an update..."), and I think this is also meant
seriously by the authors.
Bye,
Ratti
--
-o) fontlinge | Fontmanagement for Linux | Schriftenverwaltung in Linux
/\\ http://freshmeat.net/projects/fontlinge/
_\_V http://www.gesindel.de https://sourceforge.net/projects/fontlinge/
signature.asc
Description: This is a digitally signed message part
Re: Asus A8V Deluxe, Xfree display problems with BIOS 1011
Am Mittwoch, den 04.05.2005, 15:41 + schrieb Jean-Luc Coulon > I've an ASUS A8V Deluxe with an ASUS A9250T/D video card. > I've just updated my BIOS from 1009 to 1011. > > After this update, the machine boot fine and the display works in > cosole mode but while swtiching to X when gdm starts, I get a black > screen with only the cursor image in white. > > The machine is still alive and I can access it via ssh. But I cannot > even swith to a console and try a "blind reboot" with > and then log in. I cannot kill the X server with . Me2. Same board, but I use sid-32bit. It seems to be a mess, but works: I use the "nvidia" module (lsmod), but for X11 I use "nv" (XF86Config-4). When I configure X11 to use "nvidia" instead of "nv", the machine crashes like you described. "nv" is working. Some weeks ago I tried Ubuntu unstable, but didn't use it anymore since then. When this problem occured, I booted into Ubuntu just to see it crashing like you described, even it was working before Bios-Update and I didn't reconfigure anything. That means: X.org is hit by the problem also, exactly like XFree. Maybe using nv instead of nvidia made you believe Ubuntu would run? And up/downgrading the kernel may have broken the nvidia module, so nv is used instead, so that made you believe the kernel-up/downgrade fixed it? I've to agree: Asus broke our machines. Bye, Ratti -- -o) fontlinge | Fontmanagement for Linux | Schriftenverwaltung in Linux /\\ http://freshmeat.net/projects/fontlinge/ _\_V http://www.gesindel.de https://sourceforge.net/projects/fontlinge/ signature.asc Description: This is a digitally signed message part
Re: Re: evolution addressbook problem
Am Freitag, den 31.12.2004, 14:36 +0100 schrieb Christian Thalinger: > On Fri, 2004-12-31 at 14:20 +0100, Kurt Roeckx wrote: > > Check for any processes left running that might be related to it? > > This was the first thing come to my mind. But it seems this is not the > problem: > > [EMAIL PROTECTED]:~$ ps ax|grep evo > 8107 pts/1S+ 0:00 grep evo evolution --force-shutdown Bye, Ratti -- -o) fontlinge | Fontmanagement for Linux | Schriftenverwaltung in Linux /\\ http://freshmeat.net/projects/fontlinge/ _\_V http://www.gesindel.de https://sourceforge.net/projects/fontlinge/ signature.asc Description: Dies ist ein digital signierter Nachrichtenteil
Re: gnoem background not setting on startup
Hi, Am Mittwoch, den 22.09.2004, 08:46 +0800 schrieb Nicholas Hemsley: > small problem with the background not loading in gnome. If I run the > gnome-background-properties, it will set the background I choose, just > not from startup. I had a similar problem with x86. After starting nautilus the first time, the background appeared and stayed forever. Bye, Ratti -- -o) fontlinge | Fontmanagement for Linux | Schriftenverwaltung in Linux /\\ http://freshmeat.net/projects/fontlinge/ _\_V http://www.gesindel.de https://sourceforge.net/projects/fontlinge/ signature.asc Description: Dies ist ein digital signierter Nachrichtenteil
Re: Nautilus fix
Am Fr, den 30.07.2004 um 23:47 Uhr +0100 schrieb Matt Kay:
> I did a bit of Googling and found a temporary fix to the bug in Nautilus
> where it crashes on startup,
That one is really annoying.
I did some googling, too, and found bugreports to orbit and orbit2
causing that bug. I downloaded two patches and they work. Unfortunately
I forgot where I got them (since I closed my session to check if the
problem is solved), but I attach both. One patch is against orbit, the
other one against orbit2. Looking at the dependencies it might be the
orbit2-patch is enough.
I'm not a coder, so dont's ask me. :-) I read about some type-problems
expecting a return value being 32 Bit, which is not true on all !32bit
platforms.
Hope it helps.
Bye, Ratti
--
-o) fontlinge | Fontmanagement for Linux | Schriftenverwaltung in Linux
/\\ http://freshmeat.net/projects/fontlinge/
_\_V http://www.gesindel.de https://sourceforge.net/projects/fontlinge/
--- ORBit-0.5.17/src/orb/allocators.c.alignment 2003-04-01 18:51:46.0 -0500
+++ ORBit-0.5.17/src/orb/allocators.c 2003-04-02 05:28:36.0 -0500
@@ -185,9 +185,11 @@
case CORBA_tk_except:
case CORBA_tk_struct:
mem = ALIGN_ADDRESS (mem, ORBit_find_alignment (tc));
- for (i = 0; i < tc->sub_parts; i++)
+ for (i = 0; i < tc->sub_parts; i++) {
+ mem = ALIGN_ADDRESS (mem, ORBit_find_alignment (tc->subtypes[i]));
mem = ORBit_free_via_TypeCode (
mem, &tc->subtypes[i], CORBA_TRUE);
+ }
retval = mem;
break;
case CORBA_tk_union: {
--- ORBit-0.5.17/src/orb/corba_any.c.alignment 2002-06-06 07:07:39.0 -0400
+++ ORBit-0.5.17/src/orb/corba_any.c 2003-04-02 06:35:53.0 -0500
@@ -169,6 +169,7 @@
case CORBA_tk_struct:
*val = ALIGN_ADDRESS(*val, ORBit_find_alignment(tc));
for(i = 0; i < tc->sub_parts; i++) {
+ *val = ALIGN_ADDRESS(*val, ORBit_find_alignment(tc->subtypes[i]));
ORBit_marshal_value(buf, val, tc->subtypes[i], mi);
}
break;
@@ -197,9 +198,10 @@
}
break;
case CORBA_tk_wstring:
+ *val = ALIGN_ADDRESS(*val, ALIGNOF_CORBA_POINTER);
+
ulval = strlen(*(char **)*val) + 1;
- *val = ALIGN_ADDRESS(*val, ALIGNOF_CORBA_POINTER);
giop_send_buffer_append_mem_indirect_a(buf,
&ulval,
sizeof(CORBA_unsigned_long));
@@ -208,10 +210,10 @@
*val = ((guchar *)*val) + sizeof(char *);
break;
case CORBA_tk_string:
- ulval = strlen(*(char **)*val) + 1;
-
*val = ALIGN_ADDRESS(*val, ALIGNOF_CORBA_POINTER);
+ ulval = strlen(*(char **)*val) + 1;
+
giop_send_buffer_append_mem_indirect_a(buf,
&ulval,
sizeof(CORBA_unsigned_long));
@@ -595,6 +597,7 @@
case CORBA_tk_struct:
*val = ALIGN_ADDRESS(*val, ORBit_find_alignment(tc));
for(i = 0; i < tc->sub_parts; i++) {
+ *val = ALIGN_ADDRESS(*val, ORBit_find_alignment(tc->subtypes[i]));
ORBit_demarshal_value(buf, val, tc->subtypes[i], dup_strings, orb);
}
break;
@@ -820,6 +823,8 @@
*val = ALIGN_ADDRESS(*val, ORBit_find_alignment(tc));
*newval = ALIGN_ADDRESS(*newval, ORBit_find_alignment(tc));
for(i = 0; i < tc->sub_parts; i++) {
+ *val = ALIGN_ADDRESS(*val, ORBit_find_alignment(tc->subtypes[i]));
+ *newval = ALIGN_ADDRESS(*newval, ORBit_find_alignment(tc->subtypes[i]));
_ORBit_copy_value(val, newval, tc->subtypes[i]);
}
break;
--- ORBit-0.5.17/src/orb/orbit.c.alignment 2003-04-01 19:02:40.0 -0500
+++ ORBit-0.5.17/src/orb/orbit.c 2003-04-02 06:33:28.0 -0500
@@ -230,9 +230,12 @@
*a = ALIGN_ADDRESS (*a, ORBit_find_alignment (tc));
*b = ALIGN_ADDRESS (*b, ORBit_find_alignment (tc));
- for (i = 0; i < tc->sub_parts; i++)
+ for (i = 0; i < tc->sub_parts; i++) {
+ *a = ALIGN_ADDRESS (*a, ORBit_find_alignment (tc->subtypes[i]));
+ *b = ALIGN_ADDRESS (*b, ORBit_find_alignment (tc->subtypes[i]));
if (!ORBit_value_equivalent (a, b, tc->subtypes [i], ev))
return FALSE;
+ }
return TRUE;
}
diff -Naur --exclude=CVS ORBit2/src/orb/orb-core/allocators.c ORBit2.patched/src/orb/orb-core/allocators.c
--- ORBit2/src/orb/orb-core/allocators.c 2003-07-07 09:11:31.0 -0400
+++ ORBit2.patched/src/orb/orb-core/allocators.c 2004-06-17 13:13:07.563883105 -0400
@@ -59,6 +59,7 @@
}
case CORBA_tk_except:
case CORBA_tk_struct:
+ mem = ALIGN_ADDRESS (mem, tc->c_align);
for (i = 0; i < tc->sub_parts; i++) {
subtc = tc->subtypes [i];
mem = ALIGN_ADDRESS (mem, subtc->c_align);
diff -Naur --exclude=CVS ORBit2/src/orb/orb-core/corba-any.c ORBit2.patched/src/orb/orb-core/corba-any.c
--- ORBit2/src/orb/orb-core/corba-any.c 2004-01-16 08:56:12.0 -0500
+++ ORBit2.patched/src/orb/orb-core/corba-any.c 2004-06-17 13:13:07.565882999 -0400
@@ -102,6 +102,7 @@
CORBA_TypeCode tc)
{
CORBA_unsigned_long i, ulval;
+ CORBA_TypeCode subtc;
gconstpointer subval;
SKIP_ALIAS (tc);
@@ -168,12 +169,14 @@
case CORBA_tk_except:
case CORBA_tk_struct:
*val = ALIGN_ADDRESS (*val, tc->c_align);
- for (i = 0; i < tc->
