Re: Fwd: [leaf:bering-uclibc] 2 new commits to Bering-uClibc

kp kirchdoerfer <[email protected]>
Newsgroups gmane.linux.leaf.devel
Organization none
Message-ID <2468079.FLVjXGSRYG@stalker>
Hi Erich;

Am Montag, 17. August 2015, 10:32:29 schrieb Erich Titl:
> Hi KP
> 
> Am 16.08.2015 um 18:40 schrieb kp kirchdoerfer:
> > Am Sonntag, 16. August 2015, 19:23:07 schrieb Andrew:
> >> If no option UPDATE_MODDB is in leaf.cfg nor in kernel cmdline -
> >> behavior is same as earlier (modules that were probed from squash aren't
> >> stored in moddb.
> > 
> > Earlier behaviour has been to save to moddb. We should preserve that.
> > 
> > Anyway, last change to leaf.cfg has been made three yrs ago - so it
> > doesn't
> > sound like an unstable configuration file.
> 
> No, but changes in these files are critical.
> 
> > Any others idea Eric how to solve the issue?
> 
> Yes, behaviour
> 
> - If the user requests a save to moddb, no flag needed, it is a user
> decision.
> - If on boot there is a moddb use it, if not use modules.sqfs

point taken...

The whole pb started here AFAIK:
"Modules that are copied manually, should be saved. Modules that are 
loaded from squashfs aren't added to moddb.tgz to reduce it's size."


Now we can argue that a user who has enough space for leaving modules.sqfs on 
his boot media, will have enuogh left for a bigger moddb.lrp than necessary 
(except those booting from iso image, who can't remove modules.sqfs).
 
There is a case to remove modules.sqfs even beyond space, to speed up boot 
time, so autodetection stops immediately due to "missing modules.sqfs".

So if we go back and never write anything to /var/lib/lrpkg/mdoules.sqfs, we 
can get rid of the changes to update moddb properly, addinga config option to 
leaf.cfg  etc...
Andrew,  will there be any other issues related to such a change?

We may then think about adding variable to enable/disable using modules.sqfs, 
but leaving the remaining stuff untouched....???

I'll intend to provide a 5.2-rc2 soon (as-is), but we can always change this 
until final version if we agree on a better solution.

> Now the ony flag that would come to mind is one that _forces_ the use of
> modules.sqfs at boot even though moddb is present. There may be cases in
> a development environment where such behaviour is wanted, but for day to
> day use cases it is not necessary IMHO
> 
> Any other use case?

kp


------------------------------------------------------------------------------
lmpx.com only provides a reader for public news (NNTP) servers. It is not affiliated with the servers or forums shown here and is not responsible for the content of articles, which is written by their respective authors.