Can Somebody Explain this message please

OldTango · l33t Joined: 21 Feb 2004 Posts: 718

I am getting a message during booting that I don't understand how to fix.

Right after
*Starting syslog-ng.....
seq: invaild floating point argument:
Try 'seq --help' for more information.
*Setting DNS domainname to....yadayada

Now I tried the help for seq but that really wasn,t much help as I have no clue what to allpy it to.

Can somebody point me in the right direction please.......

Thanks

irf2003 · Veteran Joined: 10 Sep 2003 Posts: 1078

I am not familiar with syslog-ng, try another one, like
metalog, or msyslog, metalog is fine. e.g:

bengi · Posted: Mon Mar 08, 2004 4:41 pm Post subject:

Hi,

I'm having exactly the same problem,

OldTango · l33t Joined: 21 Feb 2004 Posts: 718

It wasn't caused by syslog-ng. After I decided to dig into it deeper, and reading the last dmesg I noticed a couple of other errrors. Which were basicly as follows.

unable to register PCM OSS
unable to register some other OSS device.
RamDisk can,t find bootsplash image at :0

Now I am using a monolithic kernel. The gentoo-dev-sources-2.6.3-r2. I have all the sound drivers and Alsa as well as OSS built into the kernel. I thought maybe the OSS part was causing the problem and it very well may have been doing it. However after rebuilding the kernel and removing the OSS stuff, my sound no longer functioned., and the message was still there.

Also I wanted to have a nice bootslpash screen so I followed the little how to in the forums on how to get it working. And was able to sort of make it work right. However I noticed that a boot@ link or some kind of link appeared in my boot partition. It made me think that I had done something installing the bootsplash that caused the segmentation fault.

Well after almost hosing my whole system being a total noob I deleted this boot@ link from my boot partition and the default link folder in my /etc/bootsplash, (i think it was there) WARNING DO NOT MAKE THIS MISTAKE unless you are perpared to work out how to recover from it. Because after doing this little manuevor my system would no longer boot to any O/S.

After the recovery process I removed the link to /usr/src/linux, emerge sync, world update, re-emerged my kernel sources and made sure it was patched in the process. I rebuilt the kernel as before, all sound built into the kernel including OSS. Installed the new kernel build and rebooted the system. My system was back in business with errors gone and sound working perfectly. I do not yet have a working bootsplash and untill I have a lot more info that is up to date I won't have one either......... :cry:

The only other thing of note is that I have tried on a few occasions to install gnucash without success. It constantly fails and after it is emerged I get emerge errors. A waring that three errors exsit and to do a emerge --verbose to view the errors. It is painfuly clear that this program will not run on a 2.6 kernel as an ebuild. Also I bootstraped this system to optimize it to my system hardware and to make sure it had the lastest tools necessary for future use. Every time I have tried to emerge gnucash it has emerged severl other programs that are major downgrades to the ones already on my system, and it never installs sucessfully. I think the powers that be should remove this program as a ebuild.

So I belive this error was caused by a bad bootsplash install (my kernel is currantly booting fine with out that boot@ link in the boot partition) or a corupted gnucash install. or maybe just a corupt kernel build. I would like to here from others on this matter just to see if there is any thing simular to their builds and maybe we can track down the exact cause of this error.

OldTango · l33t Joined: 21 Feb 2004 Posts: 718

benji

Did you write that make.conf or was it written by the system during the install. I am just wondering because mine is like 3 (+ my USE flags) lines long and has been that way sense my bootstrap process. It seems to me that I need to work on that config to better support my system.

bengi · Posted: Mon Mar 08, 2004 9:03 pm Post subject:

The text above is not my make.conf, it is the result of typing "emerge info". With my make.conf, I copied the make.conf.example file and modified it to my needs. I can post it here if you like. If you post your "emerge info" here we could compare setups. I also have noticed that the error only occurs when the bootsplash daemon is started or stopped, if I prevent it from starting at boot by removing it from the default runlevel it is gone. I would be interested to know what kernel sources (which is shown in emerge info) you are running and what version of the bootsplash software you are using also.

OldTango · l33t Joined: 21 Feb 2004 Posts: 718

Well that was some nice information and here is my emerge info

bengi · Posted: Mon Mar 08, 2004 11:46 pm Post subject:

If you enter "emerge -p bootsplash" that should say the version it would emerge and the version you are currently running in square brackets at the end of the line.

OldTango · l33t Joined: 21 Feb 2004 Posts: 718

I am running bootsplash-0.6-r9, at least I think I am. The emerge -p just shows 1 ver and it has an R in the ebuild brackets.

I am also curious if you also followed the gentoo forum tutorial for the bootsplash install.

bengi · Posted: Tue Mar 09, 2004 12:45 am Post subject:

Yep, I also followed that. I am running bootsplash version 0.6-r11 - which is the latest version in portage. I also notice slight screen corruptions if I have bootsplash added to the default runlevel when I press F2 to switch from silent to verbose mode.

OldTango · l33t Joined: 21 Feb 2004 Posts: 718

Hi

I comfirmed the above mentioned installed version of the bootsplash I have, using.

tredman2 · Posted: Fri Mar 12, 2004 1:32 pm Post subject: seq error in bootup

Quick fix here.

The problem is in the bootsplash.conf file in /etc/conf.d. There's a setting called BOOTSPLASH_TTYS that is commented out. By default, the value is the same as RC_TTY_NUMBER. However, for some reason, on my machine RC_TTY_NUMBER isn't defined, or isn't defined correctly. Therefore, when /etc/init.d/bootsplash issues the command:

bengi · Posted: Fri Mar 12, 2004 7:04 pm Post subject:

Nice one, cheers Tim

tredman2 · Posted: Sat Mar 13, 2004 2:05 am Post subject: seq error in bootup

I actually have another machine, also running bootsplash but with an older bootsplash.conf file. In the older file, there's no such mention of RC_TTY_NUMBER. So the only thing I can figure is that it's a new addition to the package. After doing an upgrade of the software, it looks like it appeared somewhere between 0.6-r9 and 0.6-r11, so it appears to be fairly recent.

However, I still haven't found out where RC_TTY_NUMBER comes from. Just out of curiosity, and after finding nothing at all on Google about it, I started a system-wide grep for the term. Seeing as how it's a 300Mhz PII with about 6G of used drive space, I'll wait a bit and let you know what I find.

Tim
_________________
Tim
"I drank what?" - Socrates, 399 BCE

OldTango · l33t Joined: 21 Feb 2004 Posts: 718

Thanks for your imput Tim. It is very helpful. Well after a few days of digging into this and tweaking this and that I discovered where I was having problems. The first problem was that I had complied into my kernel both the logo and the splash option..... :oops:

The second was in fact a bad bootsplash.conf. At some point I had upgraded from r9 to r11 and during the conf updates the conf was hosed some how, most likely buy me. I also discoverd for myself, (as I have'nt seen this info posted anyplace on the forums except by me in another thread), is that you can not use a color depth greater than 64k.......ie use 0x317 instead of 0x318 for a 1024x768 screen. My splash was working just not displaying the image because of me setting it to 16m. Being the noob that I am it never dawned of me I couldn't use this setting. The seq faults were being cause by a bad bootsplash install and nothing more. After I reset my color depth to the correct setting I downgraded to r9 again and simply over the old conf's, replaced my initrd file in /boot and rebooted. Everything is working great now. To confirm r9 does not have that RC_TTY_NUMBER, line in the config, I checked on that. The only other thing to note is that when making the initrd link to your /boot do it from / and not from boot. Though I haven't comfirmed this I think if you do it from /boot it creates a boot@ link in /boot, that is a link to /boot itself and than require you to add initrd=/boot/initrd-????x????. I created my link from / and all I required on that grub line was initrd=/initrd-1024x768. If I placed /boot in the line gurb would not boot the kernel and I had to chroot back in to fix the problem.

Sorry to be so long winded.......................Thanks All

tredman2 · Posted: Sun Mar 14, 2004 6:33 pm Post subject: Bootsplash

Well, as far as I can tell, the powers that be have backed up bootsplash to -r9. It may be that there were more issues than just the RC_TTY_NUMBER, and that would certainly jive with some of the issues that you were having.

Tim
_________________
Tim
"I drank what?" - Socrates, 399 BCE

soroh6 · Apprentice Joined: 07 Nov 2002 Posts: 232

Hm, I updated to most recent in portage a couple days ago (0.6-r11) and had the seq error, merely uncommented the BOOTSPLASH_TTYS and it works fine now. (thanks, tredman)
_________________
:: soroh -*~

Abraxas · l33t Joined: 25 May 2003 Posts: 814

tredman2 · Posted: Thu Mar 25, 2004 1:28 pm Post subject: Bootsplash

Well, I haven't really looked at any details, but -r11 was re-keyword-masked, so my system downgraded back to -r9. However, just this week, -r11 was unmasked and seems to be working fine without the line commented out, though I really haven't paid too much attention to it.

Tim
_________________
Tim
"I drank what?" - Socrates, 399 BCE

JesseP · n00b Joined: 23 Feb 2003 Posts: 11

RC_TTY_NUMBER is part of baselayout-1.8.6.13-r1, which is currently unstable. See https://bugs.gentoo.org/show_bug.cgi?id=45924 for details about the problem and how it has been addressed.