kernel 3.18

just a quick heads-up - I installed the 3.18 kernel from:

using kde4 & 5 tumbelweed and it works fine - a faster boot I noticed

I did the same using 13.2 and Gnome 3.14 and noticed high cpu and Xorg errors tied to Intel 965GM graphics.
Seems something linked to pageflip kicking or sort of.
Killing and restarting Xorg all seems good, but avoid hibernate since the same problem kicks in on resume.

sorry i should have added that i am using a ati graphic card and the fglrx driver

On Sat, 13 Dec 2014 07:26:01 GMT
dth2 <> wrote:

> sorry i should have added that i am using a ati graphic card and the
> fglrx driver

I doubt that it has anything to do with 3.18 as I’m having the same
trouble with fglrx in Tumbleweed with 3.17.x and in 13.2 with 3.16.6.

On the same machine, I’ve had no trouble with 13.1 and 3.18.0 using the
default driver. Current Xorg memory-use is 38 MiB whereas, after
running for the same amount of time, it had been climbing to near 3 GiB.
As the machine has only 4 GiB, it was grinding to a halt.

I’ll have to switch over to Tumbleweed now and see how that goes after
disabling fglrx.

Graham Davis [Retired Fortran programmer - now a mere computer user]
openSUSE 13.2 (64-bit); KDE 4.14.3; AMD Phenom II X2 550 Processor;
Kernel: 3.17.2; Video: nVidia GeForce 210 (using nVidia driver);
Sound: ATI SBx00 Azalia (Intel HDA)

Still problems with Kernel 3.18.1, on fresh install of Tumbleweed 20141230 snapshot, with 965GM integrated Intel graphics.
This is ironic, since the updated 3.16.7-7 kernel on 13.2 runs smoothly.
I don’t know enough about the Tumbleweed development model to judge if this is a regression or a new problem.
I understand that 3.16.6 had problem with the new Intel graphics stack, but those were solved in 3.16.7-7 with some 3.17 backports, among other fixes.
These fixes got then lost going to 3.18?

If anyone is interested and can understand it, I enclose what my journal says…

gen 02 10:25:49 kernel: ------------ cut here ]------------
gen 02 10:25:49 kernel: WARNING: CPU: 1 PID: 0 at ../drivers/gpu/drm/i915/intel_display.c:9916 intel_check_page_flip+0xbb/0xd0 [i915]()
gen 02 10:25:49 kernel: Kicking stuck page flip: queued at 205, now 206
gen 02 10:25:49 kernel: Modules linked in: af_packet bnep tpm_infineon fuse arc4 snd_hda_codec_hdmi hp_wmi iTCO_wdt sparse_keymap iTCO_vendor_support snd_hda_codec_analog snd_hda_codec_generic ppdev dm_mod snd_hda_intel snd_hda_controller snd_hda_codec snd_hwdep snd_pcm coretemp kvm iwl3945 pcspkr joydev iwlegacy serio_raw snd_timer snd lpc_ich mfd_core mac80211 tg3 ptp pps_core libphy cfg80211 btusb bluetooth rfkill shpchp soundcore acpi_cpufreq fan thermal wmi battery 8250_fintek parport_pc parport tpm_tis tpm ac hp_accel lis3lv02d input_polldev processor sr_mod cdrom ata_generic pcmcia ata_piix i915 i2c_algo_bit yenta_socket pcmcia_rsrc pcmcia_core firewire_ohci drm_kms_helper firewire_core crc_itu_t drm video button sg
gen 02 10:25:49 kernel: CPU: 1 PID: 0 Comm: swapper/1 Not tainted 3.18.1-1-desktop #1
gen 02 10:25:49 kernel: Hardware name: Hewlett-Packard HP Compaq 6510b (GB867ET#ABZ)/30C0, BIOS 68DDU Ver. F.20 12/01/2011
gen 02 10:25:49 kernel:  0000000000000000 0000000000000009 ffffffff81658d3e ffff88007f503d88
gen 02 10:25:49 kernel:  ffffffff81060827 ffff8800798b5000 ffff880037727800 ffff8800377279a8
gen 02 10:25:49 kernel:  0000000000000000 0000000000000086 ffffffff8106088a ffffffffa019cb40
gen 02 10:25:49 kernel: Call Trace:
gen 02 10:25:49 kernel:  <ffffffff8100541e>] dump_trace+0x8e/0x350
gen 02 10:25:49 kernel:  <ffffffff81005786>] show_stack_log_lvl+0xa6/0x190
gen 02 10:25:49 kernel:  <ffffffff81006f01>] show_stack+0x21/0x50
gen 02 10:25:49 kernel:  <ffffffff81658d3e>] dump_stack+0x49/0x6a
gen 02 10:25:49 kernel:  <ffffffff81060827>] warn_slowpath_common+0x77/0x90
gen 02 10:25:49 kernel:  <ffffffff8106088a>] warn_slowpath_fmt+0x4a/0x50
gen 02 10:25:49 kernel:  <ffffffffa014ab1b>] intel_check_page_flip+0xbb/0xd0 [i915]
gen 02 10:25:49 kernel:  <ffffffffa0118c7d>] i915_handle_vblank+0x4d/0xa0 [i915]
gen 02 10:25:49 kernel:  <ffffffffa011d822>] i965_irq_handler+0x282/0x350 [i915]
gen 02 10:25:49 kernel:  <ffffffff810b5a15>] handle_irq_event_percpu+0x35/0x1d0
gen 02 10:25:49 kernel:  <ffffffff810b5bee>] handle_irq_event+0x3e/0x60
gen 02 10:25:49 kernel:  <ffffffff810b8a9d>] handle_edge_irq+0x8d/0x140
gen 02 10:25:49 kernel:  <ffffffff8100537d>] handle_irq+0x1d/0x30
gen 02 10:25:49 kernel:  <ffffffff816624f9>] do_IRQ+0x49/0xe0
gen 02 10:25:49 kernel:  <ffffffff816604ad>] common_interrupt+0x6d/0x6d
gen 02 10:25:49 kernel:  <ffffffff8150f862>] cpuidle_enter_state+0x62/0x1b0
gen 02 10:25:49 kernel:  <ffffffff8109eff1>] cpu_startup_entry+0x361/0x3f0
gen 02 10:25:49 kernel:  <ffffffff81035c3f>] start_secondary+0x1af/0x1e0
gen 02 10:25:49 kernel: --- end trace a33f0ff6b4481e55 ]---

Sounds like this upstream bug:
Bug is marked as fixed in development branch. Please file openSUSE bug report and point to above upstream bug.

if anyone wants to know,
I am running kernel 3.18 on GNOME 3.14 with proprietary NVIDIA drivers for over a week now, and without any hiccups so far…
My version of inkscape keeps crashing when I use my WACOM pen with it… although that’s Inkscape 0.91 which is not the recommended release… I might go back to the other one.
also, I hate nautilus (for lack of features even with extensions) and am using Nemo as the file manager.
On Tumbleweed, I just use Yast to freeze updating the kernel if I’m unsure about a new kernel being supported by NVIDIA… then even doing a zypper dup won’t install it.
I’ve been using Tumbleweed for over a year, and must admit that it’s only getting better. Stablility is good, so I think it’s the best rolling release distro out there… less hassle than any other. Even, when something gets missed (which is really not often) it gets fixed within a short period of time by updates.
Congratulations to the team who maintains this.

Thanks for pointing out that report, seems exactly the same problem.

Here is the openSUSE bug.

Currently testing kernel 3.18.1-2.g161f750-desktop from Kernel:stable and all seems good :slight_smile:

Hi, please be aware that 3.18 brings a new scheduler to block devices (or better “no” scheduler?):

bruno@linux:~> uname -r
bruno@linux:~> cat /sys/block/sda/queue/scheduler

versus the old:

bruno@linux:~> uname -r
bruno@linux:~> cat /sys/block/sda/queue/scheduler
noop [deadline] cfq

See this thread and this report. No more tinkering with SSD tuning?