LinuxQuestions.org

LinuxQuestions.org (/questions/)
-   Slackware (https://www.linuxquestions.org/questions/slackware-14/)
-   -   Slackware-current upgrades today. (https://www.linuxquestions.org/questions/slackware-14/slackware-current-upgrades-today-4175442453/)

hitest 12-21-2012 09:42 AM

Slackware-current upgrades today.
 
Armageddon is shaping up rather well today. :)

Code:

Fri Dec 21 07:01:05 UTC 2012
Every time the world is about to end, all I want to do is compile kernels.
So, here you go...  the final kernel updates. *
Still testing the new glibc, which removes the gets() function and may or may
not be a safe upgrade yet.  It might be safer after the next gcc comes out.
a/kernel-firmware-20121220git-noarch-1.txz:  Upgraded.
a/kernel-generic-3.7.1-i486-1.txz:  Upgraded.
a/kernel-generic-smp-3.7.1_smp-i686-1.txz:  Upgraded.
a/kernel-huge-3.7.1-i486-1.txz:  Upgraded.
a/kernel-huge-smp-3.7.1_smp-i686-1.txz:  Upgraded.
a/kernel-modules-3.7.1-i486-1.txz:  Upgraded.
a/kernel-modules-smp-3.7.1_smp-i686-1.txz:  Upgraded.
d/binutils-2.23.51.0.6-i486-1.txz:  Upgraded.
d/gcc-4.7.2-i486-1.txz:  Upgraded.
d/gcc-g++-4.7.2-i486-1.txz:  Upgraded.
d/gcc-gfortran-4.7.2-i486-1.txz:  Upgraded.
d/gcc-gnat-4.7.2-i486-1.txz:  Upgraded.
d/gcc-go-4.7.2-i486-1.txz:  Upgraded.
d/gcc-java-4.7.2-i486-1.txz:  Upgraded.
d/gcc-objc-4.7.2-i486-1.txz:  Upgraded.
d/kernel-headers-3.7.1_smp-x86-1.txz:  Upgraded.
d/oprofile-0.9.7-i486-3.txz:  Rebuilt.
k/kernel-source-3.7.1_smp-noarch-1.txz:  Upgraded.
isolinux/initrd.img:  Rebuilt.
kernels/*:  Upgraded.
usb-and-pxe-installers/usbboot.img:  Rebuilt.
* Until the next b'ak'tun.  See you all in 13.0.0.0.0!  :-)


sycamorex 12-21-2012 10:02 AM

Yes, as I wrote in some other thread:

May all the future doomsdays be as good as this one:)

Thank you Pat for the updates.

andrew.46 12-21-2012 04:09 PM

A nice Christmas gift :)

Neil65 12-21-2012 08:27 PM

Okay, so I have a customized boot CD so I just plucked the new kernel and the new /lib/modules/
Everything is good booting up from USB but now my usb stick is /dev/sda and the sata hard-drive is /dev/sdb which is the opposite of before. Has anyone else seen this or is this just a feature of the newer kernels? Needless to say, my installation scripts are pining for the fjords.

TIA and Merry Christmas.

GazL 12-22-2012 06:57 AM

The gcc bugfix release is appealing but I'm not sure I want to jump on the 'current' treadmill just yet. I think I'll just cherry pick these tool-chain updates for now.

sycamorex 12-22-2012 06:58 AM

Quote:

Originally Posted by GazL (Post 4854912)
The gcc bugfix release is appealing but I'm not sure I want to jump on the 'current' treadmill just yet. I think I'll just cherry pick these tool-chain updates for now.

That's my plan for now too.

willysr 12-22-2012 07:49 AM

Yeah, i'm only upgrading the gcc, binutils, and oprofile without touching the kernels-* package for now :)

W4LK 12-22-2012 08:18 AM

Bad MD5 sums
 
I tried to install the upgrades today and it failed with bad MD5 sums. What's the fix?

hitest 12-22-2012 08:30 AM

Quote:

Originally Posted by W4LK (Post 4854944)
I tried to install the upgrades today and it failed with bad MD5 sums. What's the fix?

Try running slackpkg upgrade-all again and re-install any files that had bad md5 sums.

W4LK 12-22-2012 11:40 AM

Bad MD5 sums
 
Tried it again. Slackpkg reports all md5's are bad. Tried another mirror - same results. I downloaded the files manually and installed them manually with no problems. Not quite sure how to upgrade the new initrc.img file. It is not a .tgz format. I was able to make a init.gz for the new kernel which works fine. Thanks & Merry Christmas!

fogpipe 12-22-2012 11:44 AM

Just upgraded and the old nvidia version i was using wouldnt compile because it was looking for a "version.h" in the kernel source that wasnt there.

If you have an nvidia card, save yourself a headache and download the latest nvidia driver for your card before that first reboot.

cwizardone 12-22-2012 12:12 PM

How did he know I wanted a new kernel for Christmas?! :scratch:
:)
Many Thanks!

A Very Merry Christmas and a Happy New Year to You and Yours!
:hattip:

hitest 12-22-2012 12:16 PM

Quote:

Originally Posted by W4LK (Post 4855031)
Tried it again. Slackpkg reports all md5's are bad. Tried another mirror - same results. I downloaded the files manually and installed them manually with no problems. Not quite sure how to upgrade the new initrc.img file. It is not a .tgz format. I was able to make a init.gz for the new kernel which works fine. Thanks & Merry Christmas!

Cool. Happy to hear you got it sorted out! You're very welcome! Merry Christmas to you as well. :)

cwizardone 12-22-2012 01:23 PM

The new kernel is working very well.
If it works well for a near novice like myself, it should work for just about everyone (just about :) ).
Appears to have better support for HDMI audio (off of my Nvidia graphics card). KDE actually automatically found it and used it, but there is still no volume control. I think it woke up everyone in the house.
:)

mlangdn 12-22-2012 01:26 PM

I upgraded all but the kernel. I am already on a custom 3.7.1.

fogpipe 12-22-2012 01:31 PM

Is it just me or does the 3.7 kernel seem faster than the old one?
It might also be that i hadnt rebooted in a couple of months, but things seem snappier generally.

T3slider 12-22-2012 01:34 PM

Quote:

Originally Posted by cwizardone (Post 4855076)
Appears to have better support for HDMI audio (off of my Nvidia graphics card). KDE actually automatically found it and used it, but there is still no volume control. I think it woke up everyone in the house.
:)

I would expect volume control to be confined to the receiver. If you want volume control over HDMI output then you'd need to implement a software volume control. This is not a bug IMO.

cwizardone 12-22-2012 01:41 PM

Quote:

Originally Posted by T3slider (Post 4855086)
I would expect volume control to be confined to the receiver. If you want volume control over HDMI output then you'd need to implement a software volume control. This is not a bug IMO.

Ba-hum-bug, I didn't say it was a bug. :) Perhaps, I should have said the KDE mixer doesn't control it, and, yes, I'm aware you have to use software control with various applications. Regardless, I went back to the built-in audio card on the motherboard. Easier to control with the KDE or Xfce mixer.

willysr 12-22-2012 06:06 PM

Quote:

Originally Posted by fogpipe (Post 4855033)
Just upgraded and the old nvidia version i was using wouldnt compile because it was looking for a "version.h" in the kernel source that wasnt there.

If you have an nvidia card, save yourself a headache and download the latest nvidia driver for your card before that first reboot.

Try this one
Code:

ln -s /usr/src/linux-3.7/include/generated/uapi/linux/version.h /usr/src/linux-3.7/include/linux/version.h
If you used older NVidia driver, you would need a patch described on my Blog

tux_dude 12-23-2012 12:04 AM

Quote:

Originally Posted by Neil65 (Post 4854725)
Okay, so I have a customized boot CD so I just plucked the new kernel and the new /lib/modules/
Everything is good booting up from USB but now my usb stick is /dev/sda and the sata hard-drive is /dev/sdb which is the opposite of before. Has anyone else seen this or is this just a feature of the newer kernels? ...

What is the official word on this? I notice a similar issue with my RC7 kernel on my usb drive which I use for resizing the volumes > 16TB. My OS drive that is on the sata_0 on the motherboard has always been sda. With RC7 it was no longer sda. I have not dug into it, but it seems the HBA drives are assigned before the onboard sata dirves. I'll be updating my usb drive, but cannot test until I have another resize to perform.

fogpipe 12-23-2012 06:53 AM

Quote:

Originally Posted by willysr (Post 4855194)
Try this one
Code:

ln -s /usr/src/linux-3.7/include/generated/uapi/linux/version.h /usr/src/linux-3.7/include/linux/version.h
If you used older NVidia driver, you would need a patch described on my Blog

Actually upgrading the driver worked fine, thats good to know tho.

audriusk 12-23-2012 07:01 AM

Quote:

Originally Posted by tux_dude (Post 4855288)
What is the official word on this? I notice a similar issue with my RC7 kernel on my usb drive which I use for resizing the volumes > 16TB. My OS drive that is on the sata_0 on the motherboard has always been sda. With RC7 it was no longer sda. I have not dug into it, but it seems the HBA drives are assigned before the onboard sata dirves. I'll be updating my usb drive, but cannot test until I have another resize to perform.

AFAIK this behavior is normal for recent kernels (since the switch to SATA subsystem for all hard drives?), it's also unavoidable. If you want to be sure that the right device is mounted on the right mount point, use labels or UUIDs. Look them up using
Code:

lsblk -f
and in your /etc/fstab first column write LABEL=actual-label or UUID=actual-uuid instead of path to device.

lkraemer 12-23-2012 07:24 AM

I'm new enough I don't understand the fix for this issue.
Quote:

Originally Posted by W4LK:
I tried to install the upgrades today and it failed with bad MD5 sums. What's the fix?
Quote:

Originally Posted by hitest:
Try running slackpkg upgrade-all again and re-install any files that had bad md5 sums.
Quote:

Originally Posted by W4LK:
Slackpkg reports all md5's are bad....I downloaded the files manually and installed them manually with no problems.
Not quite sure how to upgrade the new initrc.img file. It is not a .tgz format. I was able to make a init.gz for the
new kernel which works fine.
I still have the following warnings:
Code:

==============================================================================
WARNING!        WARNING!        WARNING!        WARNING!        WARNING!
==============================================================================
One or more errors occurred while slackpkg was running:                     

binutils-2.23.51.0.6-i486-1.txz:        md5sum
gcc-4.7.2-i486-1.txz:        md5sum
gcc-g++-4.7.2-i486-1.txz:        md5sum
gcc-gfortran-4.7.2-i486-1.txz:        md5sum
gcc-gnat-4.7.2-i486-1.txz:        md5sum
gcc-go-4.7.2-i486-1.txz:        md5sum
gcc-java-4.7.2-i486-1.txz:        md5sum
gcc-objc-4.7.2-i486-1.txz:        md5sum
kernel-headers-3.7.1_smp-x86-1.txz:        md5sum
kernel-modules-3.7.1-i486-1.txz:        md5sum
kernel-modules-smp-3.7.1_smp-i686-1.txz:        md5sum
kernel-source-3.7.1_smp-noarch-1.txz:        md5sum
oprofile-0.9.7-i486-3.txz:        md5sum

==============================================================================

And since I'm still a novice without the following knowledge:
1. Where do I download the files with bad MD5SUM's?
2. How do I manually install them?
3. How do I make a new init.gz file for the kernel?

Somehow I've missing this information...


Thanks.

Larry

escaflown 12-23-2012 02:21 PM

Quote:

Originally Posted by lkraemer (Post 4855439)
I'm new enough I don't understand the fix for this issue.

And since I'm still a novice without the following knowledge:
1. Where do I download the files with bad MD5SUM's?
2. How do I manually install them?
3. How do I make a new init.gz file for the kernel?

Somehow I've missing this information...


Thanks.

Larry

Try another mirror in the slackpkg mirror file (/etc/slackpkg/mirrors). I had the same issue and I ended up using the kernel.org mirror ( ftp://mirrors.us.kernel.org/slackwar...kware-current/), which solved the issue.

stormtracknole 12-23-2012 02:54 PM

Quote:

Originally Posted by fogpipe (Post 4855083)
Is it just me or does the 3.7 kernel seem faster than the old one?
It might also be that i hadnt rebooted in a couple of months, but things seem snappier generally.

I have noticed that my netbook runs much cooler under this kernel.

lkraemer 12-23-2012 03:33 PM

escaflown,
THANKS for the TIP. I had a situation where I couldn't boot from the Hard Drive. I was beginning to Panic.
I finally got it working. For others with the same condition, here is what I did.

Assuming you have Ethernet Cable connection, I did the following:
1. Booted form my Slackware 14 DVD
2. Logged in as root
3. chroot into my system
Code:

    mkdir /mnt/centon
    mount /dev/sda1 /mnt/centon
    mount --bind /proc /mnt/centon/proc
    mount --bind /sys /mnt/centon/sys
    mount --bind /dev /mnt/centon/dev
    chroot /mnt/centon
    nano /etc/slackpkg/mirrors
## scrolled down to find the mirror I had enabled, then replaced it with:
## ftp://mirrors.us.kernel.org/slackware/slackware-current/
## Saved the file with CNTL O
    lilo
    exit
    reboot

Now I'm happy,happy,happy.

THANKS.

Larry

Neil65 12-23-2012 03:37 PM

Quote:

Originally Posted by audriusk (Post 4855430)
AFAIK this behavior is normal for recent kernels (since the switch to SATA subsystem for all hard drives?), it's also unavoidable. If you want to be sure that the right device is mounted on the right mount point, use labels or UUIDs. Look them up using
Code:

lsblk -f
and in your /etc/fstab first column write LABEL=actual-label or UUID=actual-uuid instead of path to device.

No, this is definitely new behavior since 3.2.x, and I got rid of the IDE native drivers long before then. Besides, its USB drives that are jumping to the head of the queue and they have always been in the scsi namespace. I've googled for the cause and haven't found anything, though I think Slackware is ahead of the curve when it comes to new OSs on bootable USB images. Just waiting for more poo to hit some bigger fans.

I can fix the symptom but I'd rather not have too. This is similar to randomized NICs have been a source of joy for a while. Though I suppose more checks before an fdisk isn't all bad.


All times are GMT -5. The time now is 11:44 PM.