Page 8 of 8

strange fault

Posted: Sat 08 Jan 2011, 09:14
by scsijon
Interesting problem:

one of my boxes has a logitech mk250 keyboard/mouse cordless usb combo,
these devices have an on-off switch so you don't waste power,
i usually leave them on unless I know i'll be away for a day or two as I use rechargable batteries

Only with the latest 432v2, if I power up but leave the keyboard or mouse turned off before starting the linux boot, the system seems to think it's starting it's config routine again.

any ideas? Can we set a config box/marker to stop this somewhere?

regards
scsijon

Posted: Sun 09 Jan 2011, 14:24
by PupGeek
I'm assuming that you would be booting your system up before turning on the keyboard and mouse, and that this would be happening with any Puppy. If that is the case then its probably going through the config process because it did not get any feedback from your input devices. I'm not sure how Puppy goes about detecting keyboards and mice, but you probably should have them turned on before booting puppy. With Ubuntu, it will stick with the previous configuration.

I'm sure the config process is in the init script. You could probably write in an option to use the previous config there (you can probably even write code that will use the old config after a certain amount of time), but that would involve extracting and rebuilding your initrd.gz file. Alternatively, you may want to check your scripts in your /etc/init.d directory for any signs of this process and make any necessary adjustments there (only if you are comfortable with writing or editing shell scripts, that is -- and always make copies of scripts before editing them too.).

Posted: Tue 11 Jan 2011, 00:42
by scsijon
@pupgeek

nope v1 worked ok, setup the first time, wasn't changed later reboots.
v2 is the first one that totally resets

Posted: Mon 07 Feb 2011, 16:20
by vtpup
I'm a bit confused by the versioning of Puppy 4.3.2. I have been running 4.3.2 test v3 on all my computers. Now we have v1 and v2 -- what are these?

I've lost my livecd for 4.3.2 test v3 -- the link at the first post of the older 4.3.2 "official" thread for this version is dead.

I don't want to upgrade to a different version at this point, since I don't want to make changes to what is a stable system at present. I either need the original file, or at least an explanation of what "v1" is, if it is the same, not a step backwards, and v2 if it is a very different "4.3.2".

Posted: Mon 07 Feb 2011, 23:53
by ttuuxxx
vtpup wrote:I'm a bit confused by the versioning of Puppy 4.3.2. I have been running 4.3.2 test v3 on all my computers. Now we have v1 and v2 -- what are these?

I've lost my livecd for 4.3.2 test v3 -- the link at the first post of the older 4.3.2 "official" thread for this version is dead.

I don't want to upgrade to a different version at this point, since I don't want to make changes to what is a stable system at present. I either need the original file, or at least an explanation of what "v1" is, if it is the same, not a step backwards, and v2 if it is a very different "4.3.2".
Hi I started over again at it, So do you want the older v3 version, I could try to find it if so.
ttuuxxx

Posted: Tue 08 Feb 2011, 00:52
by scsijon
try here for the old v3 (about halfway down page)
http://www.lamiaworks.com.au/puppy_isos.html

regards
scsijon
I really must update this page and get it finished!

Not seeing SCSI drives on DELL 1600

Posted: Tue 15 Feb 2011, 18:07
by gcmartin
I'm just getting back to this non-RAID SCSI system.

Question: Is SCSI support available when booting from LiveCD? Or is there some steps that is suppose to occur before reboot and session save to CD so that the drives on the system can be seen by this PUP?

Thanks in advance

Posted: Fri 18 Feb 2011, 23:04
by PupGeek
gcmartin: Like my Dell Poweredge server, it may be using the LSI Fusion MPT controller even if it is not RAID. For some reason, that support was never enabled in puppy.... Even in the SCSI versions. The good news is that it is not very hard to recompile the kernel with that support built in. You can download the kernel sources from http://www.bkhome.org/sources and the devx_xxx.sfs module, and when you make menuconfig, you will find an option for Fusion MPT support. Enable this item with a "y" and then enter it, as it becomes a submenu. In there, you enable the top three items with a "y" so that it is all compiled within the kernel to allow booting from your scsi drives. You do not have to rebuild your initrd image. You can remaster your puppy on a non-scsi machine and place this new kernel in the puppylivecdbuild directory, instead of the one automatically placed there. After that, when you burn the remastered cd, it will contain your recompiled kernel.

Posted: Sat 19 Feb 2011, 03:59
by whatupuppy
Hello Tuxx,

I have an IBM laptop model A22m without a hard drive installed.

The only version on puppy I can get up and running is:
http://distro.ibiblio.org/pub/linux/dis ... modems.iso

I tried to use the version 002 and got an error.
Please see the pic.

Can I continue to manually start puppy at this point?
If yes, can you let me know how or point me in what thread I can find the answer. many thanks.

Posted: Sat 19 Feb 2011, 06:28
by tempestuous
PupGeek wrote:You can download the kernel sources from http://www.bkhome.org/sources and the devx_xxx.sfs module
More accurately, you should obtain the kernel source from the "kernel-2.6.30.5-20100904" directory at that download location.
This kernel source has been configured differently from the standard 4.3.x kernel source for SCSI support.

Posted: Sat 19 Feb 2011, 06:32
by tempestuous
PupGeek wrote:you enable the top three items with a "y" so that it is all compiled within the kernel to allow booting from your scsi drives.
You do not have to rebuild your initrd image.
Yes, you have achieved success here by enabling the MPT support directly within the kernel image.
A more elegant solution would be to rebuild the initrd with the MPT modules included externally, as is standard Puppy (and Linux) practice. But of course, this involves more work.

Posted: Sat 19 Feb 2011, 22:41
by PupGeek
Well, I should have said "that way, you will not have to rebuild your initrd". You are right though tempestuous. I know that more items enabled into the kernel image makes it larger, but this is only a couple of items and not too much larger. I guess that adding much more support would make compiling as modules and rebuilding the initrd much more worthwhile. I thought, however, that support for a specific boot device would have to be built in or or the kernel would not be able to find the initrd image if placed on that drive (or the /lib/modules directory for full install).

@ whatupuppy: The message shown in your attached image typically means that puppy cannot find your <PupletName>_<VersionNumber>.sfs file. It will look on a hard drive first but refer to the cd if not there. This file on your cd may be corrupt or unreadable for some reason. Not sure how old your laptop is, but if it is really old, it may have a scsi cdrom drive that most puppy kernels (or initrd's) will not support.

SCSI on this ISO's LiveCD

Posted: Sun 20 Feb 2011, 19:49
by gcmartin
@PupGeek and @Tempestuous thanks for your explanation of what's needed to talk to the HDDs in my DELL server. There are so many DELL and IBM servers which use SCSI and LSI, that I'm kind of surprised that it may have been overlooked here. I ahve a DELL 1600 with P4s server and an IBM P3s rack mount, each with SCSI, but unable to get this LiveCD to see the HDDs.

All SCSI machines that I have encountered over the past 10 years have more that enough RAM that a LiveCD with this support, RAM should not have been a concern. But, ....

Thanks for your help. If this ISO is extended, this might be something included for use on 32bit SCSI Puppy machines.

64bit SCSI Puppies will probably have this when they surface. SCSI seems to have shown up on ONLY 2 types of systems in my past. Customers who run servers, and customers who do drafting-engineering on workstations. In each of these cases, there is so much RAM available, that Adaptec, LSI will not be a problem in the OS, as, I would think that they would have this included and not stripped-down.

Thanks for helping me thus far. I appreciate it, the explanations, and your understanding.

Posted: Thu 07 Apr 2011, 07:27
by don922
Is the development of 4.3.2-SCSI 2.6.30.5 full older modem support finished?

In another thread which was discussing updating Puppy 4.3.1
Dewbie wrote:I was going to mention 4.3.2 earlier in this thread, but then remembered sending a recent pm to ttuuxxx in which he replied that it was strictly a bugfix version, and that he's no longer supporting it.
responding to this statement
tlchost wrote:Yet another piece of abandonware. Amazing how we turn our backs upon things that work.
http://www.murga-linux.com/puppy/viewto ... 101#511101

Is there going to be an unbroken iso or anything else coming out of this project? The way I read the first entry of this thread V2 is broken and is a lot more than "a bugfix".

I'm not trying to pressure on anyone but I would like to know -- Are we going to get anything useable out of this project?

Edit:
ttuuxxx, today - Fri Apr 08, 2011 wrote:Ok it looks like its a go for this weekend as a start,
I'm not going to use woof, I'm going to manually edit the sfs and build the iso's.
I hope that something will come out of this?

4.3.2 is dead

Posted: Sat 16 Apr 2011, 20:10
by Gurglin
I think the distro-that-was-never-born 4.3.2 is dead, while Barry already replaced it with Wary Puppy 5 (now updated to 5.1.1)

Posted: Mon 12 Dec 2011, 08:57
by tempestuous
UPDATE Dec 12 2011:
Slacko 5.3.1 is the first Puppy version to include support for "Fusion MPT" SCSI interfaces.
Full details here -
http://www.murga-linux.com/puppy/viewto ... 698#588698