Yadd broken since 25/04/06 ?

Beitrag von **jubbler** » Samstag 29. April 2006, 05:01

Hi Guys,

I can seen the kernel-cdk name has changed to kernel-yadd since 26/04 and I have allowed for this in the dchpd.conf file but it doesn't seem to boot up since this date ?

I notice that the dietmarw file sizes for yadds have dropped since 25/04 at about 25Mb down to about 18Mb - could this be related please ? What is going on here ?

Thanks.

Beitrag von **jubbler** » Samstag 29. April 2006, 23:26

34 views and no one knows ?

Beitrag von **Houdini** » Samstag 29. April 2006, 23:36

maybe dietmarw changed to newmake
newmake can generate a dhcpd.conf file (as far as i remember)
what is your problem?

Beitrag von **jubbler** » Samstag 29. April 2006, 23:47

Houdini hat geschrieben:maybe dietmarw changed to newmake
newmake can generate a dhcpd.conf file (as far as i remember)
what is your problem?

Hi Houdini, thanks for coming back to me...,

I thought Dietmarw had always used newmake - certainly for the last month or 2 since I've been playing about.

Problem is that compiling my own images with newmake (like Dietmarw ?) that yadds no longer boot up after 25/4 and I noticed by chance the file size decrease on dietmarw's so thought a connection somewhere. Is there a problem in the newmake since 25/4 then ? as I also notice that ppl saying use a rcS file from head for the kernel_v2.6 yadd, and I tried that too but no luck.

What is the difference in the rcS files and is this the only thing at play here ? and why the sudden decrease in yadd size for newmake since 25/4, has someone been tidying up but deleted something important by mistake maybe ?

Just confused at the moment

(

Beitrag von **Houdini** » Sonntag 30. April 2006, 00:08

what is the error massage?
what is not working?
I need more info

Beitrag von **jubbler** » Sonntag 30. April 2006, 00:23

Houdini hat geschrieben:what is the error massage?
what is not working?
I need more info

No error message, it wont boot. It boots the kernel, shows a grey screen for a few seconds then shuts down. It all started on 25/4, it was fine before and I've not changed anything.

I'm going to try (old)make as its clear to me that "something" has changed in newmake since 25.4.

Beitrag von **Houdini** » Sonntag 30. April 2006, 00:50

can u post the console log when the box boots?

Beitrag von **jubbler** » Sonntag 30. April 2006, 01:22

Houdini hat geschrieben:can u post the console log when the box boots?

No console log, box miles away from Pc buried in tv cabinet and no com cbl long enough. Not helping myself much I know, but surely someone knows if anythings changed since this date and why the file sizes are so much smaller ? I cant be the only person that boots newmake yadds can I ?

I wouldn't bother with newmake, if (old)make had a jffs2 target as maybe thats more "stable", I dont know, I'm just confused !

But I thought newmake was better supported and more up to date so even more confused why no-one knows whats happened since 25.04

Beitrag von **racker** » Sonntag 30. April 2006, 02:37

The lib libstdc++.so.6.0.3 is missing (and some other files).
May be dietmarw has stripped a little bit to much

Beitrag von **jubbler** » Sonntag 30. April 2006, 03:24

racker hat geschrieben:The lib libstdc++.so.6.0.3 is missing (and some other files).
May be dietmarw has stripped a little bit to much

Does this "stripping" apply just to Dietmarw cvs or to all newmake builds ?

Are oldmake builds "safe" from the strip ?

Nice to have an explanation - thanks !

Cheers.

Beitrag von **dietmarw** » Sonntag 30. April 2006, 09:22

stripping made by hand (my script)
and yes.. i ed changes something at stripping since this time (i el take a look) edit: try tomorrow

but if you make your own newmake yadd this should ent take effect??
newmake doesnt strip himself

Beitrag von **jubbler** » Sonntag 30. April 2006, 18:24

dietmarw hat geschrieben:stripping made by hand (my script)
and yes.. i ed changes something at stripping since this time (i el take a look) edit: try tomorrow

but if you make your own newmake yadd this should ent take effect??
newmake doesnt strip himself

OK cool, thanks for the info. Was not suggesting newmake stripped itself, was trying to see if it was the root-cause of the changes that has caused my problems since 25.04, and if anyone else had them - e.g. had the newmake process been revised so as to affect the build in this way ? Seems not, from what ppl have said here, so thanks for the responses. I'll treat it as a local (affecting me only) problem and go from there.

Cheers.

Beitrag von **Barf** » Sonntag 30. April 2006, 23:36

Please allow me to make a few statements here.

newmake, as well as HEAD make, knows how to make "CDK" (a file system suitable for dBox NFS-mounting as root file system + some TFTP-able files), but not how to make YADDs (a distributable tarred filesystem, for the recipient to be unpacked, and mounted from the dBox). How exactly dietmarw makes YADD out of CDK only he knows (probably he picks a subset, possibly stripping something, renaming kernel-cdk -> kernel-yadd,...).

so even more confused why no-one knows whats happened since 25.04

Let me suggest that you browse the CVS http://cvs.tuxbox.org/ for questions like this. Don't take anyone elses word! In particular not mine.

Beitrag von **jubbler** » Montag 1. Mai 2006, 06:27

Thanks for you help barf

Perhaps my inablity to boot beyond the kernel (on yadd, as above) was related to Dunryc's post in the English section regarding the busybox and his inability to boot past the kernel on an image ? It looks like same issue I had. Presumably your advice to update a file from cvs to cure his issues, is equally applicable to mine and I assume that that file has now been changed/fixed, hence that particualr advice ?

Cheers.