Current is moving again YIPPEEEEEEEEEEEE!

njb · 10-14-2010, 12:16 PM

I have current.
I want to update for the new kernel.
I'm not sure everything else will be OK .
I'm not sure I must rebuilt other packages.

NjB

Intel_ · 10-14-2010, 12:35 PM

Hello, upgraded successfully, but the proccess gtk-update-icon very slow booting proccess. Before it was not started, cif I kill it everything freezes. Can I stop it someway?

@njb just don't forget to create initrd, edit lilo.conf and rerun lilo.

ponce · 10-14-2010, 12:36 PM

FYI, just updated to sysklogd latest git, seem to have fixed the issue.
maybe their git commits got some hints but I'm not a C coder so it's hard to spot which are.
just guessing: maybe the last is involved, but surely also some others.

not of much help

linus72 · 10-14-2010, 12:40 PM

njb I upgraded all my installs to new kernel no problems
but if you upgrade gcc and such some SlackBuilds may not compile

linus72 · 10-14-2010, 12:40 PM

njb I upgraded all my installs to new kernel no problems
but if you upgrade gcc and such some SlackBuilds may not compile

vdemuth · 10-14-2010, 01:20 PM

Upgraded here without to much drama. Just have to wait now for the ATI driver to catch up.

rworkman · 10-14-2010, 01:32 PM

Quote:

Originally Posted by ponce

FYI, just updated to sysklogd latest git, seem to have fixed the issue.
maybe their git commits got some hints but I'm not a C coder so it's hard to spot which are.
just guessing: maybe the last is involved, but surely also some others.

Yep, you were correct - I found that over lunch. The one line change to replace strcpy() with memmove() in that latest git commit fixes it right up.

Edit: I attached the patch to this post.

ponce · 10-14-2010, 01:48 PM

lucky shot

rg3 · 10-14-2010, 02:00 PM

Quote:

Originally Posted by rworkman

Yep, you were correct - I found that over lunch. The one line change to replace strcpy() with memmove() in that latest git commit fixes it right up.

Edit: I attached the patch to this post.

I'm the one who reported that yesterday on ##slackware. Glad to know that the problem has been located. In the mean time, the workaround is not to use split lines (i.e. the ones with backslashes). If you use them, syslogd is unable to parse the config file and spits those error messages out. This is the reason it doesn't process the rest of the file and appears not to log anything, because nothing would really be configured.

sahko · 10-14-2010, 02:01 PM

Quote:

Originally Posted by ponce

lucky shot

Luckily the git repository isnt that active and commit title is quite descriptive

ponce · 10-14-2010, 02:06 PM

to say the truth tried extracting the patch from git but it seem it didn't work on 1.4.1 here so I thought other was involved too

looking at robby's patch it seems he also updated to 1.5?

Lucifer4o · 10-14-2010, 02:10 PM

Can you give me a quick hads up how to apply the patch, so I can find why my user has no X?

Lucifer4o · 10-14-2010, 02:16 PM

Ok guys ...

Quick workaround:

Open /etc/syslogd.conf

and delete the tabulations in front of every authpriv ....

It just started ... I mean everything ... KDM, Xorg for user ... Even ksyslogd.

rg3 · 10-14-2010, 02:23 PM

Quote:

Originally Posted by ponce

to say the truth tried extracting the patch from git but it seem it didn't work on 1.4.1 here so I thought other was involved too

looking at robby's patch it seems he also updated to 1.5?

Try this one, but I had a little trouble compiling the old sources in Slackware64-current (I had to pass -I/usr/src/linux/include and make a link from asm to asm-generic for it to compile), so I don't know if it's worth rebuilding 1.4.1.

Edit: attaching instead of pasting.

rworkman · 10-14-2010, 03:17 PM

Ooh, yeah, I also had bumped to 1.5 locally - see http://connie.slackware.com/~rworkman/sysklogd/ (this link will break once the package gets into -current).