Hello there!

On Saturday 29 December 2007 10:13:48 Rod Whitby wrote:
> I often get asked to recommend a distribution for the NSLU2.
[--SNIP--]
> Unslung
[--SNIP--]
> SlugOS
[--SNIP--]
> Angstrom
[--SNIP--]
> Debian
[--SNIP--]
> OpenWrt
[--SNIP--]

Thank you Rod for these explanations! Very nice to read, and straight to the
point.

I'll add a sixth one: DIY (Do It Yourself).

DIY follows a different route than other distributions. With DIY you build
your firmware from source, adding and building the packages you need,
configuring them the way you want. DIY is targeted at very advanced geeks
that know their way through JTAG, kernel, building packages, etc...

The way to go for DIY is as follows:
  - confirm JTAG access to the slug
  - confirm serial access to the slug
  - grab a toolchain (or better yet, build one from source)
  - grab APEX sources, configure and build APEX
  - flash APEX, using JTAG, overwriting existing RedBoot
  - confirm APEX is working (serial console)
  - grab latest kernel sources, configure and build kernel
  - transfer kernel to slug (serial, or ethernet with latest APEX)
  - confirm kernel is booting (down to panic while mounting root)
  - grab busybox, configure with only a shell, and build
  - build a file system (JFFS2, ext2, as you wish) with busybox
  - transfer the file system to slug (again, via serial or ethernet)
  - boot kernel with init=/bin/sh on the comand line
  - confirm shell is working (type some commands)
  - go on with more complex stuff, increasing complexity one step at a time.

This is how I've gone to setup my toy-slug. Merely a proof a concept though.

Regards,
Yann E. MORIN.

--
.-----------------.--------------------.------------------.--------------------.
|  Yann E. MORIN  | Real-Time Embedded | /"\ ASCII RIBBON | Erics' conspiracy: |
| +0/33 662376056 | Software  Designer | \ / CAMPAIGN     |   ^                |
| --==< °_° >==-- °------------.-------:  X  AGAINST      |  /e\  There is no  |
| http://ymorin.is-a-geek.org/ | (*_*) | / \ HTML MAIL    |  """  conspiracy.  |
°------------------------------°-------°------------------°--------------------°

It's a uNSLUng v8 beta test. I installed again in another hard disk,
and I copied all the files from the small partition (sda2) from the
new install into the old hard disk sda2, and it worked just fine
afterwards, since I changed very little from the original install. I
basically only use it to store files.
The Linux box is a FC6. This can be a simple solution for many
problems relating to "HDD not formated" scenarios.
I have a feeling that the entire issue started as I did the fsck.ext3
and something malfunctioned in the hardware as the check run.

I have revised the SMART status on the HDD, and all shows OK, not even
relocated sectors. I think I will be purchasing a shielded SATA cable
very soon.

so far so good.





--- In nslu2-linux@yahoogroups.com, "luis_perez25" <luis_perez25@...>
wrote:
>
> This is the last thing I did before I could no longer see the hdd
>
> # umount -f /dev/sda2
> # /sbin/fsck.ext3 -fyv /dev/sda2
> e2fsck 1.34 (25-Jul-2003)
> Couldn't find ext2 superblock, trying backup blocks...
> Superblock has a bad ext3 journal (inode 8).
> Clear? yes
>
> *** ext3 journal has been deleted - filesystem is now ext2 only ***
>
> Pass 1: Checking inodes, blocks, and sizes
> Bad block inode has illegal block(s).  Clear? yes
>
> ------------------------------------ the last lines where
> /dev/sda2: ***** FILE SYSTEM WAS MODIFIED *****
>
>       12 inodes used (0%)
>        0 non-contiguous inodes (0.0%)
>          # of inodes with ind/dind/tind blocks: 6/6/6
>     3828 blocks used (3%)
>        2 bad blocks
>        0 large files
>
>        0 regular files
>        1 directory
>        0 character device files
>        0 block device files
>        0 fifos
>       -1 links
>        0 symbolic links (0 fast symbolic links)
>        0 sockets
> --------
>
> afterthat, I tried to use tunefs from another linux box, but all the
> files in sda2 are lost.
>
> before I was able to see sda1 fine now I get:
> ================================================
> # /sbin/fsck.ext3 -fv /dev/sda1
> e2fsck 1.34 (25-Jul-2003)
> Pass 1: Checking inodes, blocks, and sizes
> Error allocating block bitmap (1): Memory allocation failed
> e2fsck: aborted
>
>   216048 inodes used (0%)
>        0 non-contiguous inodes (0.0%)
>          # of inodes with ind/dind/tind blocks: 0/0/0
> 66592282 blocks used (68%)
>        0 bad blocks
>        0 large files
>
>        0 regular files
>        0 directories
>        0 character device files
>        0 block device files
>        0 fifos
>        1 link
>        0 symbolic links (0 fast symbolic links)
>        0 sockets
> --------
>        1 file
> Segmentation fault
> =========================================
> as of now, I am taking out all the data using another Linux box, any
> comments or is this the only recourse for me now?
>

Yann E. MORIN wrote:
> I'll add a sixth one: DIY (Do It Yourself).
>
> DIY follows a different route than other distributions. With DIY you build
> your firmware from source, adding and building the packages you need,
> configuring them the way you want. DIY is targeted at very advanced geeks
> that know their way through JTAG, kernel, building packages, etc...
>
> The way to go for DIY is as follows:
>  - confirm JTAG access to the slug
>  - confirm serial access to the slug
>  - grab a toolchain (or better yet, build one from source)
>  - grab APEX sources, configure and build APEX
>  - flash APEX, using JTAG, overwriting existing RedBoot
>  - confirm APEX is working (serial console)
>  - grab latest kernel sources, configure and build kernel
>  - transfer kernel to slug (serial, or ethernet with latest APEX)
>  - confirm kernel is booting (down to panic while mounting root)
>  - grab busybox, configure with only a shell, and build
>  - build a file system (JFFS2, ext2, as you wish) with busybox
>  - transfer the file system to slug (again, via serial or ethernet)
>  - boot kernel with init=/bin/sh on the comand line
>  - confirm shell is working (type some commands)
>  - go on with more complex stuff, increasing complexity one step at a time.
>
> This is how I've gone to setup my toy-slug. Merely a proof a concept though.

Once you've done that, you're ready to be part of the nslu2-linux
firmware development team :-)

-- Rod

In the last five days, we have made five new NSLU2 firmware releases:

2007-12-31 - Unslung 6.10-beta Release
2007-12-30 - SlugOS 4.8-beta Release
2007-12-29 - OpenWrt/NSLU2 Kamikaze 7.09 Release
2007-12-28 - Angstrom/NSLU2 2007.12 Release
2007-12-27 - Debian/NSLU2 Stable 4.0r2 Release

All of these new releases are available at

http://www.slug-firmware.net/

See http://tech.groups.yahoo.com/group/nslu2-linux/message/20476 for
an explanation of the pros and cons of each different firmware
distribution, and the installable packages available for each.

Thanks to everyone in the NSLU2-Linux, OpenWrt, Angstrom, OpenEmbedded
and Debian projects who contributed to these releases.

Remember, if you find any of the firmware or packages that the
NLSU2-Linux project provides useful, feel free to make a donation to
the project at

http://www.nslu2-linux.org/wiki/Main/Donate

We are currently in need of about $500 to buy a RAID controller card
and some disks for our autobuild machine to support all this new
firmware with up-to-date package feeds ...

-- Rod Whitby
-- NSLU2-Linux Project Lead

Here: http://suspend.tv/NSLU2/unslung.htm is unslung site with my
custom made logo. In folder: http://suspend.tv/NSLU2/unslung/ You may
find zipped one - bigger, in psd format. Font used is an "Blockhead
Alphabet Dark" owned by emigre.com, used for demo only.

Opinions welcome!

BTW: How to add an icon in taskbar for regular uNSLUng'ed website?

I try to configure bash for user: "myuser" as it was described in
tutoriall: http://www.nslu2-linux.org/wiki/Optware/Bash.

Unfortunally, I cannot configure connection between user and bash.

"The /etc/passwd file

You can configure the default shell for a normal user: - Open the
/etc/passwd file - Change the line of the normal user (myuser):
myuser:*crypted*:2000:501:Normal
User:/share/hdd/data/myuser:/opt/bin/bash"

In current version of unslung there no are path to "myuser" as
above, rather than /myuser in root. Shell I change path description
in "passwd" file or so?

Generally: I would like to configure environment for
"NativelyCompileUnslungPackages". There will be other diffrences in
"HowTo" tutoriall?

Thanks in advance!

On Mon, Dec 31, 2007 at 01:52:12AM -0000, nsluv2 wrote:
> In current version of unslung there no are path to "myuser" as
> above, rather than /myuser in root. Shell I change path description
> in "passwd" file or so?

Yes, you can change myuser's home directory in /etc/passwd to any
location that you find convenient.

However, there are rumors that by hand-editing such files you might
confuse some of Linksys' proprietary code.  (Unlike the other slug
distributions, Unslung is only partly Linux; the rest of it is
Linksys' secret sauce.)  So, for Unslung, some report that it is
best to make as many of the user/group changes via the web
interface as possible.

--- In nslu2-linux@yahoogroups.com, Rod Whitby <rod@...> wrote:
>
> vivekv wrote:
> >   Thanks for the info.  would be be possible to run a poll on nslu2-
> > linux.org to see who uses which firmware?  That would be an
> > interesting data to know
>
> The download statisic information (which is recorded automatically)
> gives us a much better indication compared to an optional poll.
>
> Even though some people will be downloading twice, or trying out
> multiple firmware before settling on one, it's the best indication that
> we have, and will still be far more accurate than a poll.
>
> And really, you should be basing your choice on your own requirements,
> not on some popularity figure :-)
>
> -- Rod
>

Rod,
   Agreed - I am looking at Debian so I can put on a USB Hub and hang
some interesting devices off of it.  From what I am reading from
various posts on the Wiki, I dont think that is a sure shot thing with
the Unslung firmware.

Regards
Vivek

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

apex> memscan -u 0+256m
  0x0 0x02000000 (32 MiB)
  0xa000000 0x02000000 (32 MiB)

Hi there .. I just fattened my slug (by adding 2 piggy-backed chips to the
original design)

So .. the memory is found, but for some reason the 7th hex is 8+2 instead of 2
only .. where's my soldering-fault?

Christian
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.7 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFHeO2mYMBp7Qf+VpkRAnQrAJ42RKlTcPTcCkngEx/isE86kv76MQCcDSrh
0MQ7PGK2y6hMSoGfgdDSR1o=
=HzKd
-----END PGP SIGNATURE-----

* vince_reynolds2002 <djvincereynolds@...> [2007-12-29 14:15]:
> At the end of this process the slug restarted as expected, but never
> rebooted properly i.e. it remains in a state where the Ready/Status
> light is solid amber.

I would have said try replacing 'allow-hotplug eth0' with 'auto eth0'
in /etc/network/interfaces on the slug.  Apparently some people need
this...

> When in this state the slug does not show up on my router, however
> with a direct connection I can ping 192.168.0.2 and get a response.

... but so you're saying that you can ping the slug when Debian has
started, but you cannot connect with SSH?  Is this correct?
--
Martin Michlmayr
http://www.cyrius.com/

I suggest you run sdram-init to see if it finds 2 banks and configures
the RAM controller. Then run memscan. The results you have suggest two
noncontiguous regions and I have never seen that before.

My understanding of the commands is that sdram-init is required to
configure the chip selects and memscan is required to determine the
usable memory.

Steve



--- In nslu2-linux@yahoogroups.com, Christian Rothländer
<mailinglisten@...> wrote:
>
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> apex> memscan -u 0+256m
>  0x0 0x02000000 (32 MiB)
>  0xa000000 0x02000000 (32 MiB)
>
> Hi there .. I just fattened my slug (by adding 2 piggy-backed chips
to the original design)
>
> So .. the memory is found, but for some reason the 7th hex is 8+2
instead of 2 only .. where's my soldering-fault?
>
> Christian
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v2.0.7 (GNU/Linux)
> Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org
>
> iD8DBQFHeO2mYMBp7Qf+VpkRAnQrAJ42RKlTcPTcCkngEx/isE86kv76MQCcDSrh
> 0MQ7PGK2y6hMSoGfgdDSR1o=
> =HzKd
> -----END PGP SIGNATURE-----
>

So, in tutoriall I was found instructions as follow:

"Open the /etc/passwd file - Change the line of the normal user (myuser):

myuser:*crypted*:2000:501:Normal
User:/share/hdd/data/myuser:/opt/bin/bash"

but "myuser" line in "passwd" file looks like:

myuser:scDu05Ttl4YCs:2000:501:::/dev/null

also "myuser" folder created by website, is located in root catalog.

and finally login as "myuser" give the info:

No directory, logging in with HOME=/
...
Cannot execute /dev/null: Permission denied

What can I do? ;)

> Yes, you can change myuser's home directory in /etc/passwd to any
> location that you find convenient.
>
> However, there are rumors that by hand-editing such files you might
> confuse some of Linksys' proprietary code.  (Unlike the other slug
> distributions, Unslung is only partly Linux; the rest of it is
> Linksys' secret sauce.)  So, for Unslung, some report that it is
> best to make as many of the user/group changes via the web
> interface as possible.
>

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hi Steve

did what you wrote to get the output I posted :-) ..
On my first try yesterday I had more or less contiguous regions, so I desoldered
everything and started again, Now the
regions are stable, but noncontiguous. I bet I changed something by accident.
Seems as if the "8" is activated
simultaniously with the "2" on the address register, so if the chip requests
something abobe 0x2000000 its redirected to
0xa000000 but I don't know where to search on the board to find that part.

Christian

sdm485 wrote:
> I suggest you run sdram-init to see if it finds 2 banks and configures
> the RAM controller. Then run memscan. The results you have suggest two
> noncontiguous regions and I have never seen that before.
>
> My understanding of the commands is that sdram-init is required to
> configure the chip selects and memscan is required to determine the
> usable memory.
>
> Steve
>
>
>
> --- In nslu2-linux@yahoogroups.com, Christian Rothländer
> <mailinglisten@...> wrote:
> apex> memscan -u 0+256m
>  0x0 0x02000000 (32 MiB)
>  0xa000000 0x02000000 (32 MiB)
>
> Hi there .. I just fattened my slug (by adding 2 piggy-backed chips
>> to the original design)
> So .. the memory is found, but for some reason the 7th hex is 8+2
>> instead of 2 only .. where's my soldering-fault?
> Christian
>>

> Yahoo! Groups Links





-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.7 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFHePz5YMBp7Qf+VpkRAhSpAJ9Wq1AOrckNpBBw4M8amUG6SfrFugCggdck
pvQtPxLBo5YaX+46mbBAfFs=
=+HEu
-----END PGP SIGNATURE-----

Hello Rod and all!

On Sunday 30 December 2007 22:57:29 Rod Whitby wrote:
> Yann E. MORIN wrote:
> > I'll add a sixth one: DIY (Do It Yourself).
[--SNIP--]
> Once you've done that,

Already done (and not only on the slug!).
Bah, this is part of my day-time job building-up linux-based devices.

> you're ready to be part of the nslu2-linux firmware development team :-)

Unfortunately, I haven't much time to take active part in such a project. I'm
but a lurker in this group. Too many other project I'm already involved in.
If I ever lower my involved time in another for any reason, then why not
coming here... But until then...

From what I've seen and used, you guys already have done a wonderfull job!

Regards,
Yann E. MORIN.

--
.-----------------.--------------------.------------------.--------------------.
|  Yann E. MORIN  | Real-Time Embedded | /"\ ASCII RIBBON | Erics' conspiracy: |
| +0/33 662376056 | Software  Designer | \ / CAMPAIGN     |   ^                |
| --==< °_° >==-- °------------.-------:  X  AGAINST      |  /e\  There is no  |
| http://ymorin.is-a-geek.org/ | (*_*) | / \ HTML MAIL    |  """  conspiracy.  |
°------------------------------°-------°------------------°--------------------°

--- In nslu2-linux@yahoogroups.com, "Nelson Neves"
<nelson.s.neves@...> wrote:
>
> --- In nslu2-linux@yahoogroups.com, Martin Michlmayr <tbm@> wrote:
> >
> > * sevensins777 <nelson.s.neves@> [2007-12-29 01:46]:
> > > -Unable to install the selected kernel
> > > An error was returned when trying to install the kernel into the
> > > target system.
> > > Kernel package: "linux-image-2.6-ixp4xx"
> > > Check /var/log/syslog or see virtual console 4 for the details.
> > >
> > > I have then selected to install the base system again and now it
shows
> > > 4 options:
> > >
> > > -Install the base system.
> > > The list shows the available kernels. Please choose one of them in
> > > order to make the system bootable from the hard drive.
> > > Kernel to install:
> > > - linux-image-ixp4xx
> > > - linux-image-2.6.18-5-ixp4xx
> >
> > linux-image-2.6.18-5-ixp4xx, but the question is why it still fails to
> > install linux-image-2.6-ixp4xx for you.  This should definitely work
> > now.  Maybe you were using a mirror that's out of date.  Did you check
> > /var/log/syslog to see what the error was installing
> > linux-image-2.6-ixp4xx?
> > --
> > Martin Michlmayr
> > http://www.cyrius.com/
> >
>
> You where right! I`ve checked the /var/log/syslog and the mirror was
> missing some files! I`ve started all over and selected a new mirror:
> debian.uchicago.edu
>
> Now it runs almost to the end :(
> It stops with message:
> -An installation step failed. You can try to run the failing item
> again from menu, or skip it and choose something else. The failing
> step is:
> Configure flash memory to boot the system.
>
> I then retried the action from the menu but still the same error!
>
> (I had a problem on previous installation where my slug had a freeze
> and needed to remove power :( )
>
> Thank you for all your help! Can you just please advise what else to
> do? I have no ideas left ...
>

... consulting /var/log/syslog I get the following error:
apt-install E:
apt-install:
apt-install: Couldn't find package flash-kernel
flash-kernel-installer: error:apt- install flash-kernel failed

How can I manually install flash-kernel? Once again, thanks for all
your help!

The second bank is showing up at the 160M (128M+32M) boundary and that
suggests that the memory mapper is correct but an upper address line
of the stacked chips is linked to the lower bank chip select (a guess)
so that when the lower bank is not selected, the upper address bit of
the upper bank is high. This could also just be a floating pin I
suppose; they can do the strangest things.

The next suggestion is that you very carefully check continuity
between the upper and lower chips. I say carefully because just the
act of pushing the probe against the upper pin can close the joint. I
had this happen and it took a while to straighten out my flaky slug.

Next thing would be to take a step back and confirm that 32M works
properly. memtester is an installable program that can exercise RAM.

It is also worth checking that the upper RAM in not just a copy of the
lower RAM by writing some values at 0x0b000000 and seeing if they show
up at 0x01000000 (for example).

If that does not help, and you feel like it, take a picture of the
chips and send it to me. I will compare it to mine.

Steve






--- In nslu2-linux@yahoogroups.com, Christian Rothländer
<mailinglisten@...> wrote:
>
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> Hi Steve
>
> did what you wrote to get the output I posted :-) ..
> On my first try yesterday I had more or less contiguous regions, so
I desoldered everything and started again, Now the
> regions are stable, but noncontiguous. I bet I changed something by
accident. Seems as if the "8" is activated
> simultaniously with the "2" on the address register, so if the chip
requests something abobe 0x2000000 its redirected to
> 0xa000000 but I don't know where to search on the board to find that
part.
>
> Christian
>
> sdm485 wrote:
> > I suggest you run sdram-init to see if it finds 2 banks and configures
> > the RAM controller. Then run memscan. The results you have suggest two
> > noncontiguous regions and I have never seen that before.
> >
> > My understanding of the commands is that sdram-init is required to
> > configure the chip selects and memscan is required to determine the
> > usable memory.
> >
> > Steve
> >
> >
> >
> > --- In nslu2-linux@yahoogroups.com, Christian Rothländer
> > <mailinglisten@> wrote:
> > apex> memscan -u 0+256m
> >  0x0 0x02000000 (32 MiB)
> >  0xa000000 0x02000000 (32 MiB)
> >
> > Hi there .. I just fattened my slug (by adding 2 piggy-backed chips
> >> to the original design)
> > So .. the memory is found, but for some reason the 7th hex is 8+2
> >> instead of 2 only .. where's my soldering-fault?
> > Christian
> >>
>
> > Yahoo! Groups Links
>
>
>
>
>
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v2.0.7 (GNU/Linux)
> Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org
>
> iD8DBQFHePz5YMBp7Qf+VpkRAhSpAJ9Wq1AOrckNpBBw4M8amUG6SfrFugCggdck
> pvQtPxLBo5YaX+46mbBAfFs=
> =+HEu
> -----END PGP SIGNATURE-----
>

* Nelson Neves <nelson.s.neves@...> [2007-12-31 15:03]:
> ... consulting /var/log/syslog I get the following error:
> apt-install E:
> apt-install:
> apt-install: Couldn't find package flash-kernel
> flash-kernel-installer: error:apt- install flash-kernel failed
>
> How can I manually install flash-kernel? Once again, thanks for all
> your help!

The mirror you're using is either very broken or (more likely) there
was a problem with your network.  I'd try another installation.
--
Martin Michlmayr
http://www.cyrius.com/

Can anyone suggest a program I can install that will thoroughly thrash
a fatslug? I am particularly interested in something that will cause a
lot of swapping in order to test the USB subsystem as requested
sometime ago.
Steve

I am using Unslung V2.3R63-uNSLUng-6.8-beta with a EXT3 formatted
500Gig USB SATA drive attached to the Disk 1 port and a 512MB Sandisk
mini Cruzer attached to Disk 2. I use it primarily
for streaming music to a music bridge attached to my home
entertainment system.

I am using WinAmp 5.5 as the agent to stream the music to a Linksys
WMB54G Music Bridge.

My question is, what is the best format (EXT3, FAT32 or NTSF) to use
on the drive used to store the music files? I currently use EXT3 but
it seems really slow. I have another USB SATA drive attached to my PC,
formatted as NTFS with the same selection of songs and that streams
the songs much faster than the drive attached to the NSLU2. I am not
sure if the issue is related to the drives format or the ability of
the NSLU2's ability to process the files.
Can someone offer me any advice or suggestions?

Thanks in advance.

First of all remember that your hard drive will ALWAYS be a LOT
faster than the USB wire it is talking down, so more important than
the speed of your hard drive is the speed of the USB.  Then, of
course, you transfer the data over a 100Mbit network ...which is
about 20% the speed of USB.

So long as your hard drive can read data a little faster than
ethernet can send it, hard drive speed [or rather, data transfer
thereof] will NOT matter.

I personally use FAT32 on all my external drives ...FAT32 is really
crap for: (a) MAASIVE (>4GB) files and (b) milllions of little files
and probably a whole bunch of other reasons as well (no journalling,
etc) ...BUT, FAT32 is great for cross-platform compatability ...I
like that I can unplug my hard drive and connect it to any
(linux/mac/windoze/BeOS/etc) computer (even my home DVD player) and
it will "just work"(tm)

In my experience SaMBa is the culprit for sharing speed ...On my xbox
it is PAINFUL (although I think this is something to do with XBMC
too) ...if you can, try a different protocol ...if streaming to XBMC,
ccxstream is a good option :)

It sounds like you stream from:
HDD -usb-> SLUG -e/net-> PC + WinAmp -wireless-> WMB54G
Maybe you can find software for your slug that will stream direct to
the linksys?

BC

At 16:31 31/12/2007, you wrote:
>I am using Unslung V2.3R63-uNSLUng-6.8-beta with a EXT3 formatted
>500Gig USB SATA drive attached to the Disk 1 port and a 512MB Sandisk
>mini Cruzer attached to Disk 2. I use it primarily
>for streaming music to a music bridge attached to my home
>entertainment system.
>
>I am using WinAmp 5.5 as the agent to stream the music to a Linksys
>WMB54G Music Bridge.
>
>My question is, what is the best format (EXT3, FAT32 or NTSF) to use
>on the drive used to store the music files? I currently use EXT3 but
>it seems really slow. I have another USB SATA drive attached to my PC,
>formatted as NTFS with the same selection of songs and that streams
>the songs much faster than the drive attached to the NSLU2. I am not
>sure if the issue is related to the drives format or the ability of
>the NSLU2's ability to process the files.
>Can someone offer me any advice or suggestions?
>
>Thanks in advance.
>
>
>
>
>Yahoo! Groups Links
>
>
>

mpd on the NSLU2 works great. But, I just turned on the replay gain
processing option in mpd and everything fell apart (CPU pegged at
100%). I suspect the replay gain code is using float.  Someine fixed
mpd to use int for mp3 decoding, is there a similar library available
for integer replay gain processing?

--
Jon Smirl
jonsmirl@...

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hi Steve

I will recheck in the new year. Currently the Machine is running and recording
TV, at least it's running again :-).

> The second bank is showing up at the 160M (128M+32M) boundary and that
> suggests that the memory mapper is correct but an upper address line
> of the stacked chips is linked to the lower bank chip select (a guess)
> so that when the lower bank is not selected, the upper address bit of
> the upper bank is high. This could also just be a floating pin I
> suppose; they can do the strangest things.

Sounds like the thing I suspected. There's only the big question which of the
112 Pins is responsible for something. :-(
Always those weird problems with floating pins, the world has to invent
something against that at one time ...

> The next suggestion is that you very carefully check continuity
> between the upper and lower chips. I say carefully because just the
> act of pushing the probe against the upper pin can close the joint. I
> had this happen and it took a while to straighten out my flaky slug.

How I know that .. I think I checked 20 times and always I found a new problem.
My first setup brought up different
memory sizes with ever run of memscan, but this time it became stable, so I
think most of them should be up and running now.

> Next thing would be to take a step back and confirm that 32M works
> properly. memtester is an installable program that can exercise RAM.

Already installed, but it crashes quite quickly at 32mb for some reason (debian)

> It is also worth checking that the upper RAM in not just a copy of the
> lower RAM by writing some values at 0x0b000000 and seeing if they show
> up at 0x01000000 (for example).

Has Apex a function to do that? Otherwise I have to wait until the 64mb are up
and running (including kernel)

> If that does not help, and you feel like it, take a picture of the
> chips and send it to me. I will compare it to mine.

Will definetly do that if I don't get it running.

so I wish you and every reader a good new year, we're off for a party now..

Cya & thanks

Christian



-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.7 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFHeSxmYMBp7Qf+VpkRAg7RAJsFWjD9/Nwl8kTivs75tVXvaeZ8jwCfTCsS
R32xPzyZj7WC69XOBH0gECw=
=w4c0
-----END PGP SIGNATURE-----

On Mon, Dec 31, 2007 at 02:25:33PM -0000, nsluv2 wrote:
> myuser:scDu05Ttl4YCs:2000:501:::/dev/null
> also "myuser" folder created by website, is located in root catalog.
> and finally login as "myuser" give the info:
> No directory, logging in with HOME=/
> Cannot execute /dev/null: Permission denied
> What can I do? ;)

First, read man 5 passwd so that you know what the fields in
/etc/passwd are supposed to mean.

Then, try (assuming that you have already installed bash):

myuser:scDu05Ttl4YCs:2000:501:Normal User:/myuser:/bin/bash

--- In nslu2-linux@yahoogroups.com, Bluechip <csbluechip@...> wrote:
>
> First of all remember that your hard drive will ALWAYS be a LOT
> faster than the USB wire it is talking down, so more important than
> the speed of your hard drive is the speed of the USB.  Then, of
> course, you transfer the data over a 100Mbit network ...which is
> about 20% the speed of USB.
>
> So long as your hard drive can read data a little faster than
> ethernet can send it, hard drive speed [or rather, data transfer
> thereof] will NOT matter.

Do you know why when I run hdparm on my attached PATA drive in a
USB2.0 enclosure the speed from hdparm on the slug is 8.5 MBps while
from my desktop it's closer to 30 MBps ?

This is with unslung firmware: V2.3R63-uNSLUng-6.8-beta.

best,
Alasdair

On 12/31/07, ragacycle <alcoheca@...> wrote:
> --- In nslu2-linux@yahoogroups.com, Bluechip <csbluechip@...> wrote:
> >
> > First of all remember that your hard drive will ALWAYS be a LOT
> > faster than the USB wire it is talking down, so more important than
> > the speed of your hard drive is the speed of the USB.  Then, of
> > course, you transfer the data over a 100Mbit network ...which is
> > about 20% the speed of USB.
> >
> > So long as your hard drive can read data a little faster than
> > ethernet can send it, hard drive speed [or rather, data transfer
> > thereof] will NOT matter.
>
> Do you know why when I run hdparm on my attached PATA drive in a
> USB2.0 enclosure the speed from hdparm on the slug is 8.5 MBps while
> from my desktop it's closer to 30 MBps ?
>
> This is with unslung firmware: V2.3R63-uNSLUng-6.8-beta.


I just tried this with my own desktop/NSLU2 pair and there is quite a
difference. This isn't a problem with the file system on the device.

I get 34.70MB/sec for cached reads (278Mb/sec). The link is capable of
480Mb/sec. So that is a 50% loss somewhere. Some is lost in protocol
overhead but not 50%.

For buffered reads I see 9.51MB/sec. I know the disk is capable of 40MB/sec.

Most likely explanation for the drop from 40MB to 9.51MB is the tiny
microcontroller in the USB disk enclosure. It may not be fast enough.
Another likely culprit is that the NSLU2 is too slow to keep up with
the revolutions of the disk and is being forced to wait for another
revolution before the sector reappears. I don't know how hdparm
requests its IO, is it one big request or lots of little ones?

Anyway, all of this is way, way faster than MP3's need by 100x. If you
are having trouble playing MP3's (100% CPU) most likely your player
software is trying to use floating point and you need to switch to one
that is implemented using fixed point. The NSLU2 does not have an FPU.
mpd has a fixed point implementation.


>
> best,
> Alasdair
>
>
>
>
> Yahoo! Groups Links
>
>
>
>


--
Jon Smirl
jonsmirl@...

Should you upgrade from Unslung 6.8-beta to 6.10-beta?

The answer is "probably not, unless you're having specific problems with
Unslung 6.8-beta".

Anyone installing Unslung from scratch should always use the latest
version (in this case, Unslung 6.10-beta).  Anyone running Unslung
5.5-beta (or an earlier version) is encouraged to upgrade, since we
don't support that version any more.

If you already have Unslung 6.8-beta installed, unslung to an external
hard disk, and running happily, then there is no real reason for you to
go through the hassle of upgrading the firmware on the internal flash.

The changes between Unslung 6.8-beta and Unslung 6.10-beta (which are
listed at http://www.nslu2-linux.org/wiki/Unslung/ChangeLog6x10) are
mostly about improving the installation and recovery experience.

There are no new normal-run-time user-visible features.

The kernel has not changed, so there is no new hardware support.  If a
device didn't work on Unslung 6.8-beta, then it still won't work on
Unslung 6.10-beta (and it's very likely that it will *never* work on the
Unslung firmware, since the kernel is effectively frozen).

The Linksys binaries have not changed, and there are no changes to the
web interface (except the removal of the non-working web upgrade page).

There are no performance improvements.  The same optware packages will
run on either 6.8-beta to 6.10-beta, and will run with exactly the same
features and performance.  Optware will continue to support 6.8-beta and
6.10-beta equally well into the future.

However, if you have not been able to get Unslung 6.8-beta to be able to
boot from an external hard disk, then you might want to give 6.10-beta a
try, as we have increased the disk wait time in early bootup to support
large disks that take longer to initialize.  In fact, if you haven't
been able to get 6.8-beta to work, then we recommend trying again with
6.10-beta and then reporting the failure on the mailing list.

-- Rod Whitby
-- NSLU2-Linux Project Lead

Hi,
   I'm hoping to use a NSLU2 to replace a dedicated x86 Linux box
running Ubuntu 7.10 to serve as a SLIP bridge to my silly homebrew
computer project (http://www.homebrewcpu.com).  Because I need
hardware handshaking, I plan on using a USB to Serial converter rather
than the internal serial port.
    My NSLU2 arrived in the mail today, and so far I've unslung it
(with 6.10) and installed net tools, busybox and the USBSerial
modules.  I also verified the basic operation of the USBSerial
converter (by ssh'ng into the NSLU2 and then using Minicom to log into
my HomebrewCPU machine).
    I'm running into a bit of trouble, though, with slattach.  On my
Ubuntu box, I can use a USBSerial converter as a SLIP bridge as follows:

     slattach -s 38400 -p slip /dev/ttyUSB0 2> /dev/null &
     ifconfig sl0 192.168.0.89 pointopoint 192.168.0.76 mtu 576 up
     route add -host 192.168.0.76 sl0
     arp -Ds 192.168.0.76 eth1 pub
     echo 1 > /proc/sys/net/ipv4/ip_forward

However, my first attempt at repeating this sequence on my NSLU2
results in slattach throwing the following error message:

"SLIP_set_disc(1): Invalid argument"

Has anyone out there successfully used an NSLU2 in a similar fashion?
  Any other sugggestions?

Thanks much,
...Bill Buzbee
http://www.homebrewcpu.com

Hello!
(Hello Bill.)
Whose USB to Serial dongle are you using for the NSLU2 end? Or for that
matter on your Linux box? Some of the devices being sold don't fully enable
the usual serial functions.

For example, one made by Belkin, the F5U109 definitely does not. I believe
the ones from Keyspan do. Certainly according to Parallax the ones made
using the FTDICHIP parts do.

I believe there is a list somewhere on their site.

  They are recommended by them for the microcontroller modules the firm
sells.

I should also commend you for your efforts regarding using such a device
with an NSLU2 to act as a connection host for your further efforts and
activities. It's, ah, it's an idea I've long pursued.
--
Gregg C Levine hansolofalcon@...
"The Force will be with you always." Obi-Wan Kenobi
 


> -----Original Message-----
> From: nslu2-linux@yahoogroups.com [mailto:nslu2-linux@yahoogroups.com] On
Behalf
> Of Bill Buzbee
> Sent: Monday, December 31, 2007 10:30 PM
> To: nslu2-linux@yahoogroups.com
> Subject: [nslu2-linux] SLIP over USBSerial?
>
> Hi,
>   I'm hoping to use a NSLU2 to replace a dedicated x86 Linux box
> running Ubuntu 7.10 to serve as a SLIP bridge to my silly homebrew
> computer project (http://www.homebrewcpu.com).  Because I need
> hardware handshaking, I plan on using a USB to Serial converter rather
> than the internal serial port.
>    My NSLU2 arrived in the mail today, and so far I've unslung it
> (with 6.10) and installed net tools, busybox and the USBSerial
> modules.  I also verified the basic operation of the USBSerial
> converter (by ssh'ng into the NSLU2 and then using Minicom to log into
> my HomebrewCPU machine).
>    I'm running into a bit of trouble, though, with slattach.  On my
> Ubuntu box, I can use a USBSerial converter as a SLIP bridge as follows:
>
>     slattach -s 38400 -p slip /dev/ttyUSB0 2> /dev/null &
>     ifconfig sl0 192.168.0.89 pointopoint 192.168.0.76 mtu 576 up
>     route add -host 192.168.0.76 sl0
>     arp -Ds 192.168.0.76 eth1 pub
>     echo 1 > /proc/sys/net/ipv4/ip_forward
>
> However, my first attempt at repeating this sequence on my NSLU2
> results in slattach throwing the following error message:
>
> "SLIP_set_disc(1): Invalid argument"
>
> Has anyone out there successfully used an NSLU2 in a similar fashion?
>  Any other sugggestions?
>
> Thanks much,
> ...Bill Buzbee
> http://www.homebrewcpu.com
>
>
>
>
>
> Yahoo! Groups Links
>
>
>

"Bill Buzbee" <bill@...> writes:
> Hi,
>   I'm hoping to use a NSLU2 to replace a dedicated x86 Linux box
> running Ubuntu 7.10 to serve as a SLIP bridge to my silly homebrew
> computer project (http://www.homebrewcpu.com).  Because I need
> hardware handshaking, I plan on using a USB to Serial converter rather
> than the internal serial port.
>    My NSLU2 arrived in the mail today, and so far I've unslung it
> (with 6.10) and installed net tools, busybox and the USBSerial
> modules.  I also verified the basic operation of the USBSerial
> converter (by ssh'ng into the NSLU2 and then using Minicom to log into
> my HomebrewCPU machine).
>    I'm running into a bit of trouble, though, with slattach.  On my
> Ubuntu box, I can use a USBSerial converter as a SLIP bridge as follows:
>
>     slattach -s 38400 -p slip /dev/ttyUSB0 2> /dev/null &
>     ifconfig sl0 192.168.0.89 pointopoint 192.168.0.76 mtu 576 up
>     route add -host 192.168.0.76 sl0
>     arp -Ds 192.168.0.76 eth1 pub
>     echo 1 > /proc/sys/net/ipv4/ip_forward
>
> However, my first attempt at repeating this sequence on my NSLU2
> results in slattach throwing the following error message:
>
> "SLIP_set_disc(1): Invalid argument"
>
> Has anyone out there successfully used an NSLU2 in a similar fashion?
>  Any other sugggestions?

I'd give SlugOS a try - I'm betting that SlugOS is probably a lot closer to
Ubuntu than is Unslung, especially when it comes to the drivers (the kernel
on Unslung is rather ancient).

Mike (mwester)

--- In nslu2-linux@yahoogroups.com, "Gregg C Levine"
<hansolofalcon@...> wrote:
>
> Whose USB to Serial dongle are you using for the NSLU2 end? Or for that
> matter on your Linux box?

I've got several - but the one I'm using right now is a development
kit board with the FTDI chipset.  I'm confident that the basic
USB-to-Serial functionality works.  I'm currently able to use minicom
on the NSLU2 to connect to my machine over the serial port.

I've done a bit more Googling since my first message, and my current
suspicion is that the unslung kernel wasn't compiled with SLIP support.

> I should also commend you for your efforts regarding using such a device
> with an NSLU2 to act as a connection host for your further efforts and
> activities. It's, ah, it's an idea I've long pursued.

I'm pretty jazzed about getting this working - it looks to be an easy
and cheap way to bridge various of my old serial devices to the net.