F23 4.3 rebase plan
by Josh Boyer
Hi All,
Really no different from any of the last major version rebases we've
done. We'll likely rebase F23 to the 4.3.y kernel around the time
4.3.1 or 4.3.2 is released upstream. Until then, the stabilization
branch is tracking what should eventually be merged into F23.
F22 will follow sometime thereafter. F21 will not be rebased to 4.2
or 4.3 and will go EOL with 4.1.13. Please migrate off of F21 as soon
as possible.
If you have questions, please let us know.
josh
7 years, 4 months
Building dell_rbu module
by Major Hayden
Hey folks,
I've rebuilt quite a few Fedora kernels to enable the dell_rbu module for firmware updates on various Dell servers, workstations, and laptops that are running Fedora 21 and 22. Would it be possible to have it enabled in the normal kernel builds?
I found a bugzilla ticket[1] referenced in the kernel config[2] that points to some installation issues in Fedora 20. I can't tell if the bug was within dell_rbu or perhaps in systemd-udevd from the ticket. However, I've had success loading the module on various Dell equipment over the last 18 months or so.
If we enable it, could we ensure that it's not enabled by default? I wondered if a kernel subpackage would be helpful but that seems a bit like overkill.
Thanks!
[1] https://bugzilla.redhat.com/show_bug.cgi?id=997149
[2] http://pkgs.fedoraproject.org/cgit/kernel.git/tree/config-x86-generic?h=f22
--
Major Hayden
7 years, 5 months
Building Fedora kernel RPM in mock
by Ian Pilcher
I am hitting this pesign issue when trying to (locally) build a custom
kernel RPM with mock:
https://bugzilla.redhat.com/show_bug.cgi?id=1283475
Is there any way to work around this before the pesign update makes it
into stable?
--
========================================================================
Ian Pilcher arequipeno(a)gmail.com
-------- "I grew up before Mark Zuckerberg invented friendship" --------
========================================================================
7 years, 6 months
Rawhide: Disabling early microcode, because kernel does not support it
by Reindl Harald
why? AFAIR that was the solution for Haswell/TSX troubles disabled with
newer microcode
[root@rawhide ~]# dracut -f
Disabling early microcode, because kernel does not support it.
CONFIG_MICROCODE_[AMD|INTEL]_EARLY!=y
[root@rawhide ~]# uname -r
4.4.0-0.rc0.git7.2.fc24.x86_64
7 years, 6 months
Kernel panic: NULL pointer dereference
by glandvador
Hi all,
I got a kernel panic with the trace:
[ 0.370825] smpboot: CPU0: AMD G-T40E Processor (fam: 14, model: 02,
stepping: 00)
[ 0.378560] Performance Events: AMD PMU driver.
[ 0.383187] ... version: 0
[ 0.387204] ... bit width: 48
[ 0.391313] ... generic registers: 4
[ 0.395333] ... value mask: 0000ffffffffffff
[ 0.400656] ... max period: 00007fffffffffff
[ 0.405977] ... fixed-purpose events: 0
[ 0.409999] ... event mask: 000000000000000f
[ 0.417579] NMI watchdog: enabled on all CPUs, permanently consumes one
hw-PMU counter.
[ 0.426082] x86: Booting SMP configuration:
[ 0.430330] .... node #0, CPUs: #1
[ 0.447454] x86: Booted up 1 node, 2 CPUs
[ 0.451681] smpboot: Total of 2 processors activated (3999.93 BogoMIPS)
[ 0.458314] BUG: unable to handle kernel NULL pointer dereference at
(null)
[ 0.466195] IP: [< (null)>] (null)
[ 0.471276] PGD 0
[ 0.473322] Oops: 0010 [#1] SMP
[ 0.476605] Modules linked in:
[ 0.479694] CPU: 0 PID: 1 Comm: swapper/0 Not tainted
4.2.5-300.fc23.x86_64 #1
[ 0.486919] Hardware name: PC Engines APU, BIOS
SageBios_PCEngines_APU-45 04/05/2014
[ 0.494659] task: ffff880119b80000 ti: ffff880119adc000 task.ti:
ffff880119adc000
[ 0.502139] RIP: 0010:[<0000000000000000>] [< (null)>]
(null)
[ 0.509644] RSP: 0018:ffff880119adfe40 EFLAGS: 00010246
[ 0.514955] RAX: ffff88011a005438 RBX: 0000000000000000 RCX:
ffffffff81d2ece0
[ 0.522090] RDX: 0000000000000000 RSI: ffff88011a005438 RDI:
ffff88011a02e808
[ 0.529221] RBP: ffff880119adfe68 R08: 0000000000000000 R09:
ffff880119c00000
[ 0.536354] R10: 0000000000000000 R11: 00000000000000bf R12:
0000000000000002
[ 0.543484] R13: ffffffff81f0cb00 R14: ffff88011a02e808 R15:
0000000000000000
[ 0.550622] FS: 0000000000000000(0000) GS:ffff88011ec00000(0000)
knlGS:0000000000000000
[ 0.558713] CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b
[ 0.564460] CR2: 0000000000000000 CR3: 0000000001c0b000 CR4:
00000000000006f0
[ 0.571592] Stack:
[ 0.573612] ffffffff81d6e53d 00000000001e8460 000000000000a040
ffffffff81d30200
[ 0.581107] 0000000000000000 ffff880119adfe88 ffffffff81d6a8a2
0000000000000002
[ 0.588602] 0000000000000002 ffff880119adfeb8 ffffffff81d7cc1f
0000000000000000
[ 0.596099] Call Trace:
[ 0.598565] [<ffffffff81d6e53d>] ? setup_ioapic_dest+0xc3/0xeb
[ 0.604495] [<ffffffff81d6a8a2>] native_smp_cpus_done+0xfb/0x105
[ 0.610600] [<ffffffff81d7cc1f>] smp_init+0x62/0x82
[ 0.615575] [<ffffffff81d57153>] kernel_init_freeable+0x105/0x228
[ 0.621760] [<ffffffff81768f50>] ? rest_init+0x80/0x80
[ 0.626992] [<ffffffff81768f5e>] kernel_init+0xe/0xe0
[ 0.632145] [<ffffffff817795df>] ret_from_fork+0x3f/0x70
[ 0.637548] [<ffffffff81768f50>] ? rest_init+0x80/0x80
[ 0.642779] Code: Bad RIP value.
[ 0.646150] RIP [< (null)>] (null)
[ 0.651317] RSP <ffff880119adfe40>
[ 0.654809] CR2: 0000000000000000
[ 0.658144] ---[ end trace 830735da38a70b70 ]---
[ 0.662806] Kernel panic - not syncing: Attempted to kill init!
exitcode=0x00000009
[ 0.662806]
[ 0.671998] ---[ end Kernel panic - not syncing: Attempted to kill
init! exitcode=0x00000009
This is the x86_64 kernel.4.2.5-300.fc23.
[ 0.000000] Linux version 4.2.5-300.fc23.x86_64
(mockbuild(a)bkernel02.phx2.fedoraproject.org) (gcc version 5.1.1 20150618
(Red Hat 5.1.1-4) (GCC) ) #1 SMP Tue Oct 27 04:29:56 UTC 2015
[ 0.000000] Command line: BOOT_IMAGE=/vmlinuz-4.2.5-300.fc23.x86_64
root=UUID=589968d6-2283-4903-8699-1b23f5c341a5 ro console=ttyS0,115200n8
This machine was yum/dnf updated from f21 to f22 and then f23. The same
kernel on the f22 had the same behavior.
The kernel x86_64 4.1.10-100.fc21 is working.
The following log is from working kernel:
[ 0.452476] smpboot: Total of 2 processors activated (4000.29 BogoMIPS)
[ 0.460310] devtmpfs: initialized
[ 0.477700] clocksource jiffies: mask: 0xffffffff max_cycles:
0xffffffff, max_idle_ns: 1911260446275000 ns
I reinstalled the kernel, regenerate the initramfs. This same initramfs is
booting fine on another (different hw) f23 machine.
I don't know where to look further. I can open a bugzilla bug report, but
I am not fully convinced that it is only a kernel bug.
7 years, 6 months
Can't build kernel RPM with local (non-git) patch
by Ian Pilcher
I just installed F23 (coming from F21) and encountered an irritating
new behavior when trying to build a custom kernel RPM.
Patch does not have a valid e-mail address.
Well no kidding. It's a locally created (with diff) patch, not coming
from a git repository.
Is there a way to disable this check?
--
========================================================================
Ian Pilcher arequipeno(a)gmail.com
-------- "I grew up before Mark Zuckerberg invented friendship" --------
========================================================================
7 years, 6 months
kernel-tests master: Add introduction with mailing list info to documentation (eb972f3)
by Justin Forbes
Repository : http://git.fedorahosted.org/cgit/kernel-tests.git
On branch : master
>---------------------------------------------------------------
commit eb972f3cb8b4ffda31cfc53d260be01413c055d9
Author: Justin M. Forbes <jforbes(a)redhat.com>
Date: Thu Nov 5 08:27:36 2015 -0600
Add introduction with mailing list info to documentation
>---------------------------------------------------------------
documentation.txt | 11 +++++++++++
1 files changed, 11 insertions(+), 0 deletions(-)
diff --git a/documentation.txt b/documentation.txt
index 3d6fb9a..367d639 100644
--- a/documentation.txt
+++ b/documentation.txt
@@ -1,3 +1,14 @@
+=== Introduction ===
+This test suite is designed to look for regressions in the running kernel. It
+is maintained by Justin Forbes and the Fedora kernel team. Any issues or
+enhancements should be sent to the Fedora kernel list:
+kernel(a)lists.fedoraproject.org
+While Fedora runs this test suite on every kernel it builds, hardware can
+vary significantly. Anyone running this suite and submitting results will
+only give us better test coverage. You can set up basic result submission
+and secure boot signature checking in '.config'. The 'config.example' file
+is a great place to start.
+
=== Running Tests ===
In the simplest of terms, simply run the runtests.sh script with no arguments.
7 years, 6 months
%posttrans in rawhide fails
by Reindl Harald
result is that the package is installed but nothing in /boot
________________________
/bin/kernel-install: Zeile 95: Syntaxfehler beim unerwarteten Wort `done'
/bin/kernel-install: Zeile 95: ` done'
Warnung: %posttrans(kernel-core-4.4.0-0.rc0.git1.1.fc24.x86_64)
Scriptlet fehlgeschlagen, Beenden-Status 2
Non-fatal POSTTRANS scriptlet failure in rpm package kernel-core
Non-fatal POSTTRANS scriptlet failure in rpm package kernel-core
________________________
[root@rawhide ~]# ls /boot/
insgesamt 19M
drwxr-xr-x. 6 root root 1,0K 2015-11-03 20:00 grub2
drwx------. 2 root root 12K 2012-05-27 17:46 lost+found
-rw------- 1 root root 8,7M 2015-11-03 19:56
initramfs-4.3.0-0.rc7.git2.1.fc24.x86_64.img
-rw-r--r-- 1 root root 159K 2015-10-28 18:22
config-4.3.0-0.rc7.git2.1.fc24.x86_64
-rw------- 1 root root 3,3M 2015-10-28 18:22
System.map-4.3.0-0.rc7.git2.1.fc24.x86_64
-rwxr-xr-x 1 root root 6,4M 2015-10-28 18:22
vmlinuz-4.3.0-0.rc7.git2.1.fc24.x86_64
7 years, 6 months
kdbus module being removed from Rawhide
by Josh Boyer
Hi All,
We will be removing the kdbus driver from Rawhide kernels before the
4.3 final release upstream. Realistically, this means kdbus will be
gone from Fedora by Monday November 2nd at the latest. If you have a
setup using kdbus, please adjust it accordingly.
The upstream developers asked me to remove the module from Fedora
while they rethink some of the approach they are taking with kdbus.
josh
7 years, 7 months
