I have been mucking about with T2-puppy build tar with dingo alphas. I wish to understand enough to start puppy on different hardware. Like the CBE found in the PS3. I have had it with the horrific big installs, via debian, Suse and YDL to just play around like I do with puppy on my PS3.
The small size and run in RAM mode of puppy would make it a wonderful plaything on the PS3
Am stuck on Perl halting on T2-piglet (code name for puppy). I short cut it [easy to do] but It got me in the end and I broke something else.
So I'm starting this if others want to go for the ride.
1st milestone get T2-piglet to compile all the packages as it is now.
2nd change over to I-586 base,
b) then pre 686 (like VIAs CPUs not quite a full 686),
c) 686
d) AMD
3rd move on to SMP multicore.
--------------
Start additional pack and build scripts to but the RAW files in the right places to make Puppy dance with the most common processors greater than the lowly 486
-------------
4th move to PowerPC 64
5th move to CBE (as modified PowerPC64)
6th move to CBE (as PPC64 & CELL subprocessor)
7th to be determined hopefully the embedded stuff.
T2-puppy for other CPUs
getting around T2-puppy environment with CHROOT
After further checking around I installed the mini- version of T2 and could not read the html documents [provided by BK] needed to play around so I booted puppy 2.16 without the devx and using chroot to the base of the mini T2 install. I was able to use T2 goodies in a terminal window, cut and paste and as I write this its happily compiling away in a terminal.
BASIC:
installed T2 mini in a partition from install CD.
rebooted and setup installed T2.
rebooted again into puppy v2.16 w/o devx just straight puppy
using Pmount mounted T2 partition and un-tarred the 1G T2-puppy package
opened terminal.
#cd /mnt/sda1
#mount -o bind /proc proc
#mount -o bind /dev dev
#chroot . /bin/bash
and I am ready to go.
BASIC:
installed T2 mini in a partition from install CD.
rebooted and setup installed T2.
rebooted again into puppy v2.16 w/o devx just straight puppy
using Pmount mounted T2 partition and un-tarred the 1G T2-puppy package
opened terminal.
#cd /mnt/sda1
#mount -o bind /proc proc
#mount -o bind /dev dev
#chroot . /bin/bash
and I am ready to go.
Last edited by Ted Dog on Sun 13 Jan 2008, 23:25, edited 1 time in total.
T2 build scripts not functional
After trying multiple version of T2, with and without Dingo. I have come to a hard stopping point. The perl and autoconfig compiles do not function. I would be interested in what BK and those at T2 did to get past this road block.
I am surprised that the developer of T2 even released any isos with this defect. The replies geaned from many who tried and failed to use T2 point to this error and no help is provided, no ticket exists even stating this issue.
I'll try LFS to see how they get a working autoconfig with perl.
VERY DISAPPOINTED WITH T2
I am surprised that the developer of T2 even released any isos with this defect. The replies geaned from many who tried and failed to use T2 point to this error and no help is provided, no ticket exists even stating this issue.
I'll try LFS to see how they get a working autoconfig with perl.
VERY DISAPPOINTED WITH T2
-
BarryK
- Puppy Master
- Posts: 9392
- Joined: Mon 09 May 2005, 09:23
- Location: Perth, Western Australia
- Contact:
Ted Dog, you haven't read the documentation. There is a HTML help file bundled with the t2-puppy package that explains the Perl stopping bug and a fix, plus some other stopping points.
These fixes are also online, see this page and read down:
http://www.puppylinux.com/pfs/index.html
These fixes are also online, see this page and read down:
http://www.puppylinux.com/pfs/index.html
[url]https://bkhome.org/news/[/url]
Dang I did skip that section
Ah, good one. 
I did read the html docs but skipped that section. It referred to t2-puppy-4.00.tar.gz which I could not find since puppy 4 is not officially released . I was able to hack past that problem bdb but I did not think to replace it with the original and it killed the compile later on.
That is when I tried to use the last T2 release iso. to set up the environment. I tried multiple T2 build targets and they all broke at a use of PERL So that is my point of frustration.
You do give better directions than anything I could find via Google & T2 website. Thanks to the quick reply I was heading down LFS route but would rather stick with puppy. Looks like the fixes are in the mega tar t2-dingo (production name for the soon to be released t2-puppy-4.00.tar.gz ) so I am off to try it again. See ya all in a day or two.
I would like to try the distcc with puppy T2. In my former life as a C developer the company I worked for, that had an email for everybody to leave the computer on overnight and boot a disc running a distcc client on 15-20 computers. It would make a multi-day full software release compile done in a few hours. I have many underused pcs and a need to heat my house! Remember all watts used by computers are transformed into heat. So instead of wasting electricity to warm air with an electric baseboard, fire up a few computers and give them something useful to do.
I did read the html docs but skipped that section. It referred to t2-puppy-4.00.tar.gz which I could not find since puppy 4 is not officially released . I was able to hack past that problem bdb but I did not think to replace it with the original and it killed the compile later on. That is when I tried to use the last T2 release iso. to set up the environment. I tried multiple T2 build targets and they all broke at a use of PERL So that is my point of frustration.
You do give better directions than anything I could find via Google & T2 website. Thanks to the quick reply I was heading down LFS route but would rather stick with puppy. Looks like the fixes are in the mega tar t2-dingo (production name for the soon to be released t2-puppy-4.00.tar.gz ) so I am off to try it again. See ya all in a day or two.
I would like to try the distcc with puppy T2. In my former life as a C developer the company I worked for, that had an email for everybody to leave the computer on overnight and boot a disc running a distcc client on 15-20 computers. It would make a multi-day full software release compile done in a few hours. I have many underused pcs and a need to heat my house! Remember all watts used by computers are transformed into heat. So instead of wasting electricity to warm air with an electric baseboard, fire up a few computers and give them something useful to do.
-
Pizzasgood
- Posts: 6183
- Joined: Wed 04 May 2005, 20:28
- Location: Knoxville, TN, USA
One of my friends excuses for still using Windows (besides better support for games and required-by-classes software) is that it keeps him warmer. He uses Puppy more during the summer though.
[size=75]Between depriving a man of one hour from his life and depriving him of his life there exists only a difference of degree. --Muad'Dib[/size]
[img]http://www.browserloadofcoolness.com/sig.png[/img]
[img]http://www.browserloadofcoolness.com/sig.png[/img]
distcc compiles fine in DingoA4, Dingo compiler custer..WOOF
The distcc program (very tiny) compiles and installs just fine in DingoA4. I think it could add a bit to devx sfs and be very useful for system wide huge compile projects, like T2.
Still compiling T2 Dingo : just finished linux-header pass one.
found linux.desc for upgrading kernel version.
Still compiling T2 Dingo : just finished linux-header pass one.
found linux.desc for upgrading kernel version.
Compile Day One notes
No hiccups just yet. Currently compiling
pass number [5] :QT4
About 100 more packages in this the five of nine passes, then about 20 more in the rest of the remaining passes.
Waiting to finish this package compile and stopping briefly to copy the partition to storage.
The next package is a font compile that would be a short compile, and a good place to interrupt the T2 compile script. The script handles this interruption nicely, by restarting the interrupting the running package from the beginning of only that package.
Running short on space, had to delete files and folders from earlier attempts. Still not enough room, so I looked hard and found the caches where not being cleaned out.
This message repeats itself:
*> No atime support available, Compiler Cache clean up not possible.
So hoping a Cache is just a Cache. I deleted everything. Nothing stopped so I hope that is OK.
It looks like it will have to regen about 3 packages in the future, that it may not had to otherwise.
pass number [5] :QT4
About 100 more packages in this the five of nine passes, then about 20 more in the rest of the remaining passes.
Waiting to finish this package compile and stopping briefly to copy the partition to storage.
The next package is a font compile that would be a short compile, and a good place to interrupt the T2 compile script. The script handles this interruption nicely, by restarting the interrupting the running package from the beginning of only that package.
Running short on space, had to delete files and folders from earlier attempts. Still not enough room, so I looked hard and found the caches where not being cleaned out.
This message repeats itself:
*> No atime support available, Compiler Cache clean up not possible.
So hoping a Cache is just a Cache. I deleted everything. Nothing stopped so I hope that is OK.
It looks like it will have to regen about 3 packages in the future, that it may not had to otherwise.
first hiccup, not a problem
After 26 hours I hit the first hiccup documented by BarryK.
ibrsvg installation, stage 5, fails
After digging into folder after folder I found the file to open and edit.
quick text search and I made the edit, saving the file (but leaving the text editor open on the change) Less than five minutes it completed that package and I just undid my changes and save and closed the editor.
only 334 Megs of space left on that partition.
This is looking successful so I am moving on to Milestone 2B
compile for VIA C3 not quite a full 686
Onward to a bigger box which is not the VIA C7, more HD space, faster processer, generous swap, and MORE BTUs. My worlds' most efficient laptop just not cutting it in the space-heater requirement of this project [ tang you Everex StepNote laptop!! 12W yah right It feels like only 10W]
Future Milestones: I plan to skip the 686, AMD and SMP muticore parts and move on to PPC64.
Why the VIA C3 should work on my full 686 machines and I can't find my AMD box, it was gutted for sake of others and is currently in hiding.
the family members SMP muticore is being a pain in the vista and I do not want to go near it for now
PPC64 Milestone.
The PPC64 generic package should only require ignoring its Mac'y'ness and use the PS3 booting methods instead, along with a custom PS3 kernel which I will copy from others.
ibrsvg installation, stage 5, fails
After digging into folder after folder I found the file to open and edit.
quick text search and I made the edit, saving the file (but leaving the text editor open on the change) Less than five minutes it completed that package and I just undid my changes and save and closed the editor.
only 334 Megs of space left on that partition.
This is looking successful so I am moving on to Milestone 2B
compile for VIA C3 not quite a full 686
Onward to a bigger box which is not the VIA C7, more HD space, faster processer, generous swap, and MORE BTUs. My worlds' most efficient laptop just not cutting it in the space-heater requirement of this project [ tang you Everex StepNote laptop!! 12W yah right It feels like only 10W]
Future Milestones: I plan to skip the 686, AMD and SMP muticore parts and move on to PPC64.
Why the VIA C3 should work on my full 686 machines and I can't find my AMD box, it was gutted for sake of others and is currently in hiding.
the family members SMP muticore is being a pain in the vista and I do not want to go near it for now
PPC64 Milestone.
The PPC64 generic package should only require ignoring its Mac'y'ness and use the PS3 booting methods instead, along with a custom PS3 kernel which I will copy from others.
Last edited by Ted Dog on Wed 16 Jan 2008, 05:25, edited 1 time in total.
hit second hiccup, no problem
As per BarryK's T2 html. I hit the second hiccup at the 38 hour point of the compile. Less than 50 more packages to remain. on initial compile machine one.
And on machine two, cross compile for VIA C3. Perl failure, same type of workaround as in the bdb section. And it made it passed that point.
The cross compile forced the great majority of packages to compile at level one pass and not level five pass as is the true blue original supplied by Barry. So I made that correction also to the bdb workaround.
and it made it well past the other tricky parts around the kernel compile, currently compiling
And on machine two, cross compile for VIA C3. Perl failure, same type of workaround as in the bdb section. And it made it passed that point.
The cross compile forced the great majority of packages to compile at level one pass and not level five pass as is the true blue original supplied by Barry. So I made that correction also to the bdb workaround.
and it made it well past the other tricky parts around the kernel compile, currently compiling
Code: Select all
shell bashcollection
It's good that your collection of processors is now being put to work.
A really excellent project.
A really excellent project.
Day two compile.
Machine one finished with initial compile. I expected a beep of a congratulatory WOOF WOOF or something. It just returns to the command prompt without errors. It has been safely backup to be evalutated by a mondo diff script to see how close it is to the final Dingo SFS.
The process took a VIA C7 1.5 GHZ laptop 42 hours to compile.
Machine two has come across a cross compile road block on X11 that is not in BarryK's HTML work around so I am on my own.
During this process I have learned how to resize an ext3 partition saved as a file onto a bigger NTFS partition and loop mount the saved ext3 file and run the T2-dingo build scripts from it.
That will be my preferred way of doing this in the future it makes short work in getting back to a known setpoint.
While I try to uncover the X11 cross-compile hiccup I plan to start a loop mounted ext3 T2 compile for the ppc64 also on machine two. And check if running two T2 compiles on the same machine at the same time cause any problems
Back on machine one I am starting to piece together the files need for the PS3 linux kernel based on someone else useful little project more details later.
Machine one:
T2 PPC64 PS3 kernel
Machine two:
stopped T2-cross compile for VIA C3 on hold due to unexpected failure in X11.
Start T2-cross compile for PPC64 straight for the mac & PS3
The process took a VIA C7 1.5 GHZ laptop 42 hours to compile.
Machine two has come across a cross compile road block on X11 that is not in BarryK's HTML work around so I am on my own.
During this process I have learned how to resize an ext3 partition saved as a file onto a bigger NTFS partition and loop mount the saved ext3 file and run the T2-dingo build scripts from it.
That will be my preferred way of doing this in the future it makes short work in getting back to a known setpoint.
While I try to uncover the X11 cross-compile hiccup I plan to start a loop mounted ext3 T2 compile for the ppc64 also on machine two. And check if running two T2 compiles on the same machine at the same time cause any problems
Back on machine one I am starting to piece together the files need for the PS3 linux kernel based on someone else useful little project more details later.
Machine one:
T2 PPC64 PS3 kernel
Machine two:
stopped T2-cross compile for VIA C3 on hold due to unexpected failure in X11.
Start T2-cross compile for PPC64 straight for the mac & PS3
First cross compile for diff arch finished
Well good success, the first cross-compile for a different CPU has finished. Here is what we lost.
Packages which did not compile in a cross arch build [VIA C3] for puppy T2-dingo
cron-3.0pl1-100
ffmpeg-2007-10-07*
font-misc-misc-1.0.0
gdb-6.7
gnome-icon-theme-2.20.0
libcap-1.10
libpcap-0.9.8
mplayer-1.0rc2-071007*
mplayer-fonts-1.0.3
ppp-2.4.4
samba-3.0.26
shadow-4.0.18.1
sudo-1.6.9p6
sysfiles-0000
*workarounds used in orginal T2-dingo.
short list is a good thing.
PPC64 cross compile is still being worked on.
The change from x86 type CPUs to PPC64 has file placement issues.
I will go back and try a x86-64 cross-compile next, it should help clear-up the lib64 and others that are plaguing the PPC64 attempt.
Packages which did not compile in a cross arch build [VIA C3] for puppy T2-dingo
cron-3.0pl1-100
ffmpeg-2007-10-07*
font-misc-misc-1.0.0
gdb-6.7
gnome-icon-theme-2.20.0
libcap-1.10
libpcap-0.9.8
mplayer-1.0rc2-071007*
mplayer-fonts-1.0.3
ppp-2.4.4
samba-3.0.26
shadow-4.0.18.1
sudo-1.6.9p6
sysfiles-0000
*workarounds used in orginal T2-dingo.
short list is a good thing.
PPC64 cross compile is still being worked on.
The change from x86 type CPUs to PPC64 has file placement issues.
I will go back and try a x86-64 cross-compile next, it should help clear-up the lib64 and others that are plaguing the PPC64 attempt.
double post ... heh i even refreshed the page before posting and didn't see it...
Last edited by cb88 on Fri 18 Jan 2008, 17:42, edited 1 time in total.
Taking Puppy Linux to the limit of perfection. meanwhile try "puppy pfix=duct_tape" kernel parem eater.
X86: Sager NP6110 3630QM 16GB ram, Tyan Thunder 2 2x 300Mhz
Sun: SS2 , LX , SS5 , SS10 , SS20 ,Ultra 1, Ultra 10 , T2000
Mac: Platinum Plus, SE/30
X86: Sager NP6110 3630QM 16GB ram, Tyan Thunder 2 2x 300Mhz
Sun: SS2 , LX , SS5 , SS10 , SS20 ,Ultra 1, Ultra 10 , T2000
Mac: Platinum Plus, SE/30
"And finally: We had several feedbacks that PPC was still working (Yes working!). If you are a PPC user, please do keep checking for regressions as we need to be notified ASAP about PPC regressions. "
THat is a quote from the nouveau open source nvidia driver wiki:
probably of interest to you since the PS3 has an Nvidia RXS GPU also i think they are starting to look into shader support and 2D support is improving ie better Xv support
so how hard is a livecd to boot on a PS3?
what about crosscompiling x86 puppy on the PS3!!!
THat is a quote from the nouveau open source nvidia driver wiki:
probably of interest to you since the PS3 has an Nvidia RXS GPU also i think they are starting to look into shader support and 2D support is improving ie better Xv support
so how hard is a livecd to boot on a PS3?
what about crosscompiling x86 puppy on the PS3!!!
Taking Puppy Linux to the limit of perfection. meanwhile try "puppy pfix=duct_tape" kernel parem eater.
X86: Sager NP6110 3630QM 16GB ram, Tyan Thunder 2 2x 300Mhz
Sun: SS2 , LX , SS5 , SS10 , SS20 ,Ultra 1, Ultra 10 , T2000
Mac: Platinum Plus, SE/30
X86: Sager NP6110 3630QM 16GB ram, Tyan Thunder 2 2x 300Mhz
Sun: SS2 , LX , SS5 , SS10 , SS20 ,Ultra 1, Ultra 10 , T2000
Mac: Platinum Plus, SE/30
hmm are you talking about an autobuild T2?
that would be cool with fixes to each architecture so the build goes correctly... would need a target platform though like i586 or i686 or PS3
as far as the kernel detecting the cpu... i think there will allways be limitations good to detect x86 based cpus but an x86 kernel wouldn't even start up on say a sparc or PPC would it....
that would be cool with fixes to each architecture so the build goes correctly... would need a target platform though like i586 or i686 or PS3
as far as the kernel detecting the cpu... i think there will allways be limitations good to detect x86 based cpus but an x86 kernel wouldn't even start up on say a sparc or PPC would it....
Taking Puppy Linux to the limit of perfection. meanwhile try "puppy pfix=duct_tape" kernel parem eater.
X86: Sager NP6110 3630QM 16GB ram, Tyan Thunder 2 2x 300Mhz
Sun: SS2 , LX , SS5 , SS10 , SS20 ,Ultra 1, Ultra 10 , T2000
Mac: Platinum Plus, SE/30
X86: Sager NP6110 3630QM 16GB ram, Tyan Thunder 2 2x 300Mhz
Sun: SS2 , LX , SS5 , SS10 , SS20 ,Ultra 1, Ultra 10 , T2000
Mac: Platinum Plus, SE/30