OESF | ELSI | pdaXrom | OpenZaurus | Zaurus Themes | Community Links | Ibiblio

IPB

Welcome Guest ( Log In | Register )

 
Reply to this topicStart new topic
> Some News: Kernel 2.6, Arcdisk Support
guylhem
post Jul 12 2005, 12:04 PM
Post #1





Group: Members
Posts: 577
Joined: 17-March 04
Member No.: 2,365



Hello

Thanks to Bruno kind help, I'm working on pdaxrom and debian support using a 40 Gb arcdisk. Ideally, I will follow the release map and offer a big choice of interfaces. [If you guys want a 4 Gb CF support, tell me]

However, I just discovered that for some reason, my kernel had problems with the arcdisk (usb storage too old certainly) - a new release is planned. Moreover, it seems to hang on my good old audiovox at startup for no reason.

Maybe it's time to move to a 2.6 ? I will try and test everything to make sure it will work at least as good. Stay tuned!

Guylhem
Go to the top of the page
 
+Quote Post
adf
post Jul 12 2005, 01:32 PM
Post #2





Group: Members
Posts: 2,821
Joined: 13-September 04
From: Wasilla Ak.
Member No.: 4,572



QUOTE(guylhem @ Jul 12 2005, 08:04 PM)
Hello

Thanks to Bruno kind help, I'm working on pdaxrom and debian support using a 40 Gb arcdisk. Ideally, I will follow the release map and offer a big choice of interfaces. [If you guys want a 4 Gb CF support, tell me]

However, I just discovered that for some reason, my kernel had problems with the arcdisk (usb storage too old certainly) - a new release is planned. Moreover, it seems to hang on my good old audiovox at startup for no reason.

Maybe it's time to move to a 2.6 ? I will try and test everything to make sure it will work at least as good. Stay tuned!

Guylhem
*

Very cool. Both are terrific software, and both need some help with 6000 specific kernels &tc...

BUT..if you are looking at debian support, you may as well include the large cf disk support from the start. It will come up soon enough, I'm sure.
Go to the top of the page
 
+Quote Post
ken
post Jul 12 2005, 02:08 PM
Post #3





Group: Members
Posts: 274
Joined: 17-October 04
Member No.: 5,063



QUOTE(guylhem @ Jul 12 2005, 10:04 AM)
Maybe it's time to move to a 2.6 ? I will try and test everything to make sure it will work at least as good. Stay tuned!

does this mean you're giving up on the fast kernel and guylhem rom?
Go to the top of the page
 
+Quote Post
gfdsa
post Jul 12 2005, 02:17 PM
Post #4





Group: Members
Posts: 55
Joined: 18-January 05
Member No.: 6,256



Glad to see you back!

40Gb is on usb? should be pretty slow...

QUOTE
Maybe it's time to move to a 2.6 ?


do you hold a recent list of non-working stuff there?
Go to the top of the page
 
+Quote Post
guylhem
post Jul 13 2005, 02:50 PM
Post #5





Group: Members
Posts: 577
Joined: 17-March 04
Member No.: 2,365



I'm not dead- just too busy in the ER !

ken, I don't plan to stop the rom - only sample the 2.6 kernel to consider whether it is stable enough for distribution and use, and porting features.

gfdsa - I have no list of known bugs. I have to try it myself. You can test OZ and tell me :-)

"guylhem rom" is going to become more a distribution than everything at once. There are specific things I want to improve: speed, hackability, compatibility - triple binds mutually exclusive - comes what may :-)

Actually I'm very interested in the suspend/resume time and the boottime. I want it to be blazing fast.

To do this, I did cut the kernel messages, and I'll certainly cut the logo. dmesg is still there- messages are just logged to ttyS0 instead of the console by default. I enabled sysv status report (ex: Checking filesystem [OK]) OTOH because it was important to find which part was hanging when a problem occured, even if it display stuff and therefore may add some delay. I am also adding a pretty issue screen, because of the hackability.

See my plans? It's mostly a personal effort- I want it that way because I think it is the right way, but there's no special goal- only to have the best free software distribution for a handheld.

BTW Last week in Dijon I saw some openbsd people working on a OpenBSD over C3100 project, which I promised to give a hand at least on the kernel part (even without the hardware - they had power management problem). I will try to interest them on the boot process, to replace the old bsd or sysv init by something more suitable to handhelds. It could be based on something much more efficent and modern like http://www.atnf.csiro.au/people/rgooch/linux/boot-scripts/

Anyway we'll see... I'll try to make everything fit on the flash, having the various systems on microdrives or 40 Gb usb storage. adf - maybe your pdaxrom system (saw your instructions on the 6000 forum) could be included? All it needs is to work with guylhem rom.
Go to the top of the page
 
+Quote Post
adf
post Jul 13 2005, 06:01 PM
Post #6





Group: Members
Posts: 2,821
Joined: 13-September 04
From: Wasilla Ak.
Member No.: 4,572



By all means.
I'm curious to see how this turns out, and really looking forward to having a better setup on my 6k than my rc10. I really like the pdaxrom... but a better optimised 6k kernel, partition and bettter power an peripheral support would be terrific. Same can be said for qtopia 2.1, too.
Go to the top of the page
 
+Quote Post
ken
post Jul 20 2005, 01:04 PM
Post #7





Group: Members
Posts: 274
Joined: 17-October 04
Member No.: 5,063



QUOTE(guylhem @ Jul 13 2005, 12:50 PM)
ken, I don't plan to stop the rom - only sample the 2.6 kernel to consider whether it is stable enough for distribution and use, and porting features.

ok, that's good to know.

It's understandable that you're busy and all. That being the case, I was wondering if you would consider a few thoughts ...

The guylhem rom is an ambitious but doable project, certainly something that a number of us want to see completed. The distribution however, has many aspects to it, many pieces, many parts that need to fit together.

That being the case, perhaps it would be useful to break them up into smaller bite sized portions?

At the very least, it would help in the troubleshooting process, to see if there is a piece that's inadvertantly affecting other parts? You've already gone far ahead of many of us, and there's little we can do to help figure out what's going on.

Your forums gives hints as to the directions you've taken, but I fear, not enough for someone like myself to get enough of a big picture.

Perhaps starting with something like:

1) tetsu's kernel (for the 6K it's up to 16a). That would provide people with ?. That also may possibly have other inclusions of things you want to do. That could be a starting point. He has for example, included bluetooth. Of course if you feel that using his kernel would have a number of disadvantages (which you'd know and I'd have no clue on), then of course, this would be a bad idea.

2) meanie's patches for usb perhaps?

3) a fast SDL - this is what attracted people to your rom in the first place - the idea that they could add a touch of speed to the 6K.

4) ability to have more than just 25%, 50%, 75% brightness

That being said, I have the feeling that you can already see how each thing would fit together, from just the kernel patches to the full blown guylhem distribution?

At least this way you could take more time to figure out where things are breaking?
QUOTE
There are specific things I want to improve: speed, hackability, compatibility - triple binds mutually exclusive - comes what may :-)

I have no disagreements there. Merely concerned that it may lose its momentum.

QUOTE
Actually I'm very interested in the suspend/resume time and the boottime. I want it to be blazing fast.

boot time is nice, but once it's booted, it's a not a biggie. suspect/resume, yes, I can see that. I suspect that's a long term goal though.
QUOTE
To do this, I did cut the kernel messages, and I'll certainly cut the logo. dmesg is still there- messages are just logged to ttyS0 instead of the console by default. I enabled sysv status report (ex: Checking filesystem [OK]) OTOH because it was important to find which part was hanging when a problem occured, even if it display stuff and therefore may add some delay. I am also adding a pretty issue screen, because of the hackability.

This part can be what I mentioned above. Tiny pieces that provide functionality, little modules that allow you to hack a part of the system, guaranteed to work.

QUOTE
suitable to handhelds. It could be based on something much more efficent and modern like

heh ... it's all good. The downside of course is your time gets spread further apart!

QUOTE
systems on microdrives or 40 Gb usb storage. adf - maybe your pdaxrom system (saw your instructions on the 6000 forum) could be included? All it needs is to work with guylhem rom.

I'm willing to help in any way I can. To that extent, I've been working on a page:

http://www.oesf.org/index.php?title=Applic...6000L/Sharp_ROM

This is geared of course to what we discussed, being that compatibility to the Sharp ROM is important. I'm also afraid that adding in gcc3 might break things, but you'd know better than myself whether it would or not.

Let me know how I can help.

btw, the article here: http://www.externe.net/zaurus/modules.php?...order=0&thold=0

that talks about a website is now located here:

http://home.mchsi.com/~cmisip/zaurus.htm
Go to the top of the page
 
+Quote Post
guylhem
post Jul 24 2005, 10:30 AM
Post #8





Group: Members
Posts: 577
Joined: 17-March 04
Member No.: 2,365



QUOTE(ken @ Jul 20 2005, 01:04 PM)
The guylhem rom is an ambitious but doable project, certainly something that a number of us want to see completed.  The distribution however, has many aspects to it, many pieces, many parts that need to fit together.

That being the case, perhaps it would be useful to break them up into smaller bite sized portions?


If I could be offered some help. For now very little people did help :-(

QUOTE
1) tetsu's kernel (for the 6K it's up to 16a).  That would provide people with ?.  That also may possibly have other inclusions of things you want to do.  That could be a starting point.  He has for example, included bluetooth.  Of course if you feel that using his kernel would have a number of disadvantages (which you'd know and I'd have no clue on), then of course, this would be a bad idea.


It's a testsu that has been forked. Tetsu support for the 6000 is not that good - it stopped updating the source when the 3000 was out.

QUOTE
2) meanie's patches for usb perhaps?


What for? What do they bring?

QUOTE
3) a fast SDL - this is what attracted people to your rom in the first place - the idea that they could add a touch of speed to the 6K.


That's overdue. I should try to do something with SDL and mame soon again. Currently, I'm still thinking about what could be done. Someone from england was IIRC working on the video acceleration. That would help a lot. Without it, all I can do will only bring minor speedups.

QUOTE
4) ability to have more than just 25%, 50%, 75% brightness


already done for the battery. It now reports the battery left percent by percent - 99%, 98% etc.

QUOTE
That being said, I have the feeling that you can already see how each thing would fit together, from just the kernel patches to the full blown guylhem distribution?


I do have a big picture. I should publish it, because I realise it must be very hard to get the whole thing. Chasing too much hares at once as usual :-)

QUOTE
At least this way you could take more time to figure out where things are breaking?


Currently I'm only *fixing* - no new features for the next release unless I consider the major issues fixed.

QUOTE
boot time is nice, but once it's booted, it's a not a biggie.  suspect/resume, yes, I can see that.  I suspect that's a long term goal though.


It's now much better.

QUOTE
I'm willing to help in any way I can.  To that extent, I've been working on a page:

http://www.oesf.org/index.php?title=Applic...6000L/Sharp_ROM

This is geared of course to what we discussed, being that compatibility to the Sharp ROM is important.  I'm also afraid that adding in gcc3 might break things, but you'd know better than myself whether it would or not.


That's why I think it is only needed for statically compiling speed-critical applications such as videogames, video playback etc.

QUOTE
Let me know how I can help.



If I send you a "current" release of the distribution, may you report the bugs? The filesystem problem (causing the file tab not to show the true content, or the ipk installation refusing to use the cf/sd) is still there. I'm working on the power button to suspend bug, but it's not done yet.

Something else I need for that (among other things): the code of the Light and Menu button. I will try to find where I did put my special qkonsole (see http://www.oesf.org/forums/index.php?showt...84&hl=qkonsole) and send you a copy - if you are running a sharp kernel the keycodes would be helpfull. Gab74, if you still have the special qkonsole that'd help if you could do that
Go to the top of the page
 
+Quote Post
ken
post Jul 24 2005, 12:18 PM
Post #9





Group: Members
Posts: 274
Joined: 17-October 04
Member No.: 5,063



QUOTE(guylhem @ Jul 24 2005, 08:30 AM)
If I could be offered some help. For now very little people did help :-(

You have to remember that it's not that people weren't willing to help as much as they have no clue how to help. Their level of knowledge on this is far below what your experience is. That includes me! It's only through constant effort that I'm at the point of where I am, which still is way below where it needs to be.

QUOTE
It's a testsu that has been forked. Tetsu support for the 6000 is not that good - it stopped updating the source when the 3000 was out.

I noticed that 16a is rather recent (2005/7/7). As far as I can interpret it, he's released the source for everything he's done. I've been working on a rough translation of things (the problem with the translators is they mess up on the download links) [http://www.oesf.org/index.php?title=Tetsu_special_kernel_for_the_6000] that is helping me to understand what's going on. It's far from done, but I think you can catch the jist of it. From my take, he's been providing the source of all his changes, including the starting point (2004/4/21) of the 6K kernel.

QUOTE
What for? What do they bring?

The interesting thing is he scripted for auto mounting/unmounting of usb devices. Again, I may be clueless as to the real value of all this.

QUOTE
IIRC working on the video acceleration. That would help a lot. Without it, all I can do will only bring minor speedups.

ok. Didn't know it was minor speedups - I think everyone was excited because of the ability to speed up games. That and the ability to change the clock speed I think.

QUOTE
already done for the battery. It now reports the battery left percent by percent - 99%, 98% etc.

Remember I mentioned about things being completed in pieces? I have no clue where this piece is. It would however, be useful. That being said, let's say that someone wanted to use the part - perhaps if you could provide the instructions on how to get it working (ie, replace this with that, reboot, and so forth)?

QUOTE
I do have a big picture. I should publish it, because I realise it must be very hard to get the whole thing.

Well, I sort of have an idea of where you're going, based on the doc modification I messed around with from your stuff (which I had emailed you?). I think the biggest concerned I had when I tried your modifications, is it broke my SD/CF use and my connectivity. Which led me to think perhaps, maybe it'd be more useful if your rom was in smaller pieces so I could at the very least, use part of it?

QUOTE
That's why I think it is only needed for statically compiling speed-critical applications such as videogames, video playback etc.

This is all great stuff, but stuff that I think many of us are clueless about. Which is where the explanations are useful. I'd be glad to help on documentation (provide I know what I'm doing) or wherever I can be useful. Function is important to me of course, so I'm concerned on not losing what ground I've gained thus far.

QUOTE
If I send you a  "current" release of the distribution, may you report the bugs? The filesystem problem (causing the file tab not to show the true content, or the ipk installation refusing to use the cf/sd) is still there. I'm working on the power button to suspend bug, but it's not done yet.

Of course. I think I did that the last time. As a side thought, how's about this? Would you consider removing all the pieces that are broken, so a "stable" version of your rom could be released? The Sharp ROM for instance, may be slow, inflexible, etc, but it does work, and is very stable. I'm hoping to see a similar thing in your distribution, where people could leave it at certain points, knowing that it's stable, works, and won't give them problems. Then another distribution where little by little things are added and debugged and people can decide to help or not help with those issues?

Hope this makes sense ...
Go to the top of the page
 
+Quote Post

Reply to this topicStart new topic
2 User(s) are reading this topic (2 Guests and 0 Anonymous Users)
0 Members:

 



RSS Lo-Fi Version Time is now: 23rd November 2014 - 09:11 PM