You are not logged in.
After a testing phase kernel 2.6.24.1 has been moved to Core where it replaces version 2.6.23.14.
Some of the most visible changes for users include:
* more mac80211-based wireless drivers added
* tickless support for x86_64
* uvesafb; support for ATI Xpress 200M and Mac framebuffers
* ACPI usage in libata
* SATA port multipliers support
* NCQ support on MCP51/MCP55/MCP61 chipsets (add 'swncq=1' to kernel boot commandline)
* SDIO support
* ability to use any kernel version to resume a hibernated system
* read-only --bind mounts
* tons of bugfixes and improvements
A detailed human-readable changelog is available at KernelNewbies Wiki.
Attention!
kernel26-2.6.24.1-2 package includes an important security fix (local user privilege escalation).
Everyone is encouraged to upgrade, especially on a system with ssh accounts.
Last edited by ise (2008-02-11 05:45:34)
Offline
I cant find the module padlock-sha during boot? Still it boots just fine. (using luks encrypted root)
And when i try make menuconfig in the sourcedir:
make menuconfig
scripts/kconfig/mconf arch/x86/Kconfig
drivers/crypto/Kconfig:51: can't open file "arch/s390/crypto/Kconfig"
make[1]: *** [menuconfig] Error 1
make: *** [menuconfig] Error 2
Last edited by sveinemann (2008-02-10 20:23:52)
Offline
The important security fix, is it for this exploit: http://it.slashdot.org/article.pl?sid=08/02/10/2011257 ?
Offline
The important security fix, is it for this exploit: http://it.slashdot.org/article.pl?sid=08/02/10/2011257 ?
Yes, it's the local root exploit which was fixed.
Offline
* uvesafb; support for ATI Xpress 200M and Mac framebuffers
Does anyone know whether this is supposed to work on Intel?
I'm trying to get a 1280x800 framebuffer working on my Intel GMA950, but I got a black screen and my computer froze when I tried to use uvesavb (with v86d from aur), no matter if it was inside the initrd or when loading the module after booting. With other framebuffers at least 1024x768 works, which unfortunately freezes with uvesafb, too.
Edit: tryed...ouch
Last edited by wuischke (2008-02-11 11:19:09)
Offline
Do I get it right, that iwlwifi package can now be safely removed?
And, second question - where can I see the list of other drivers included in a readable format? I wonder if they have included the linux-uvc project drivers for webcams, too.
And, finally, I'd like to congratulate all the linux users and especilally the archlinux users!
Last edited by SchwarzeKrause (2008-02-11 02:37:44)
Offline
Do I get it right, that iwlwifi package can now be safely removed?
Yes you got it right. You still need the firmware though : iwlwifi-3945-ucode (or 4965).
But I think the "crash on shutdown" that I mentioned there are in fact caused by the iwlwifi driver in 2.6.24 :
http://www.archlinux.org/pipermail/arch … 16649.html
I should try using the separate driver again (but I will probably get file conflicts now).
pacman roulette : pacman -S $(pacman -Slq | LANG=C sort -R | head -n $((RANDOM % 10)))
Offline
* uvesafb; support for ATI Xpress 200M and Mac framebuffers
Does anyone know whether this is supposed to work on Intel?
I'm trying to get a 1280x800 framebuffer working on my Intel GMA950, but I got a black screen and my computer froze when I tryed to use uvesavb (with v86d from aur), no matter if it was inside the initrd or when loading the module after booting. With other framebuffers at least 1024x768 works, which unfortunately freezes with uvesafb, too.
Did you see the wiki article on uvesafb? Did you recompile klibc against a kernel tree including uvesafb?
Mortuus in anima, curam gero cutis
Offline
I cant find the module padlock-sha during boot? Still it boots just fine. (using luks encrypted root)
And when i try make menuconfig in the sourcedir:
make menuconfig
scripts/kconfig/mconf arch/x86/Kconfig
drivers/crypto/Kconfig:51: can't open file "arch/s390/crypto/Kconfig"
make[1]: *** [menuconfig] Error 1
make: *** [menuconfig] Error 2
Which sourcedir? The archlinux kernel package does not install the linux sources in /usr/src, but only the stuff required to compile external modules. In order to compile your own kernel, follow one of the several guides in the wiki.
Mortuus in anima, curam gero cutis
Offline
Did you see the wiki article on uvesafb? Did you recompile klibc against a kernel tree including uvesafb?
No, I haven't. Thanks a lot for pointing this out. I somehow forgot about the wiki, because I found little information regarding intelfb and instead just checked the official site of uvesafb and installed v86d from aur.
Offline
After upgrading to the kernel 2.6.24.1 my virtual consoles changed their behaviour. In particular, I could not use polish chars. Here is a fragment of my
/etc/rc.conf
LOCALE="pl_PL"
HARDWARECLOCK="localtime"
TIMEZONE="Europe/Warsaw"
KEYMAP="pl"
CONSOLEFONT="lat2a-16"
CONSOLEMAP="8859-2"
USECOLOR="yes"
Fortunately, I have managed to deal with the problem by adding the following
/usr/bin/unicode_stop
to my ~/.bashrc
But I would like to know what is the reason.
Offline
upgrade 2.6.24-2 Gives me problems starting X and powering off in vmware (xp host), gonna see if it's really gotten anything todo with the kernel.
ps. like the logos
EDIT: Seems like it's working ok now, few permission problems and stuff like that did the magic.
Last edited by ibendiben (2008-02-12 15:09:51)
Offline
After upgrading to the kernel 2.6.24.1 my virtual consoles changed their behaviour. In particular, I could not use polish chars. Here is a fragment of my
/etc/rc.confLOCALE="pl_PL" HARDWARECLOCK="localtime" TIMEZONE="Europe/Warsaw" KEYMAP="pl" CONSOLEFONT="lat2a-16" CONSOLEMAP="8859-2" USECOLOR="yes"
Fortunately, I have managed to deal with the problem by adding the following
/usr/bin/unicode_stop
to my ~/.bashrc
But I would like to know what is the reason.
I think this is because Unicode was made the default for terminals in .24 kernel.
I didn't thought this may cause issues.
Please file a bugreport (though it's not quite a bug) about this so it won't get forgotten.
to live is to die
Offline
Please file a bugreport (though it's not quite a bug) about this so it won't get forgotten.
OK, it's already there: http://bugs.archlinux.org/task/9593
to live is to die
Offline