Created attachment 833450 [details] dmesg output that lockup messages are contained Description of problem: Invoking a spice server(qemu-system-x86) which is emulator of Windows XP, Booting Windows i stopped by "BUG: soft lockup" message. Version-Release number of selected component (if applicable): linux-3.13.0-0.rc2.git5.1.fc21.i686 How reproducible: always Steps to Reproduce: 1. invoke a spice server (qemu-system-x86) 2. then, qemu is crashed by the bug. 3. Actual results: Booting Windows process is completed. Expected results: qemu-system-x86 is crashed. Additional info:
Are you still seeing this with 3.13-rc4?
(In reply to Josh Boyer from comment #1) > Are you still seeing this with 3.13-rc4? Wait a moment. I will try it.
I tried. But same result as follows. --------------------------------------------- [ 256.107005] BUG: soft lockup - CPU#1 stuck for 22s! [qemu-system-x86:2267] [ 256.107005] Modules linked in: fuse ip6table_filter ip6_tables ebtable_nat ebtables ipt_MASQUERADE iptable_nat nf_nat_ipv4 nf_nat nf_conntrack_ipv4 nf_defrag_ipv4 xt_conntrack nf_conntrack xt_CHECKSUM iptable_mangle tun bridge stp llc arc4 md4 nls_utf8 cifs dns_resolver snd_pcm_oss snd_mixer_oss fscache kvm_amd dm_service_time kvm snd_hda_codec_idt joydev snd_hda_intel snd_hda_codec snd_hwdep snd_seq snd_seq_device snd_pcm snd_page_alloc ppdev tg3 snd_timer snd soundcore dcdbas nv_tco i2c_nforce2 k8temp serio_raw ptp pps_core parport_pc parport binfmt_misc uinput dm_multipath usb_storage nouveau ata_generic pata_acpi video mxm_wmi wmi i2c_algo_bit drm_kms_helper ttm drm i2c_core [ 256.107005] CPU: 1 PID: 2267 Comm: qemu-system-x86 Not tainted 3.13.0-rc4.git0.1.fc21.i686 #1 [ 256.107005] Hardware name: Dell Inc. OptiPlex 740 Enhanced/0YP693, BIOS 2.2.2 04/01/2009 [ 256.107005] task: dd3b4100 ti: d6e5a000 task.ti: d6e5a000 [ 256.107005] EIP: 0060:[<f84b5e12>] EFLAGS: 00000246 CPU: 1 [ 256.107005] EIP is at svm_complete_interrupts+0x32/0x1a0 [kvm_amd] [ 256.107005] EAX: dbaef000 EBX: d6f40000 ECX: 00000000 EDX: dbaef000 [ 256.107005] ESI: 80000008 EDI: 00000000 EBP: d6e5be0c ESP: d6e5be00 [ 256.107005] DS: 007b ES: 007b FS: 00d8 GS: 00e0 SS: 0068 [ 256.107005] CR0: 8005003b CR2: 00000000 CR3: 16e1b000 CR4: 000007f0 [ 256.107005] Stack: [ 256.107005] d6f40000 00000001 00000000 d6e5be14 f84b5fb5 d6e5beb0 f846b739 00000000 [ 256.107005] d6e5be54 002c4020 df4af898 00000000 c0dcbf3c 00000286 00000000 df4af898 [ 256.107005] 00000000 4e5ae5e7 00000035 00000000 00e388a4 f4f2e024 7ffbfeff fffffffe [ 256.107005] Call Trace: [ 256.107005] [<f84b5fb5>] svm_cancel_injection+0x35/0x40 [kvm_amd] [ 256.107005] [<f846b739>] kvm_arch_vcpu_ioctl_run+0x289/0x11a0 [kvm] [ 256.107005] [<f8468474>] ? kvm_arch_vcpu_load+0x194/0x1f0 [kvm] [ 256.107005] [<f84b20d6>] ? svm_vcpu_put+0x36/0x50 [kvm_amd] [ 256.107005] [<f8457d28>] ? vcpu_load+0x58/0xa0 [kvm] [ 256.107005] [<f84581ce>] kvm_vcpu_ioctl+0x40e/0x4d0 [kvm] [ 256.107005] [<c04c14d4>] ? do_futex+0xf4/0xa10 [ 256.107005] [<f8457dc0>] ? vcpu_put+0x50/0x50 [kvm] [ 256.107005] [<c058c4c2>] do_vfs_ioctl+0x2e2/0x4d0 [ 256.107005] [<c0a0737b>] ? __do_page_fault+0x1eb/0x560 [ 256.107005] [<c04c1e7c>] ? SyS_futex+0x8c/0x140 [ 256.107005] [<c0594eaf>] ? fget_light+0x6f/0xc0 [ 256.107005] [<c058c710>] SyS_ioctl+0x60/0x80 [ 256.107005] [<c0a0ac0d>] sysenter_do_call+0x12/0x28 [ 256.107005] Code: 8d 74 26 00 89 c3 8b 80 80 34 00 00 f6 83 68 01 00 00 10 8b bb 30 35 00 00 8b b0 88 00 00 00 c7 83 30 35 00 00 00 00 00 00 75 1e <85> f6 c6 83 64 1c 00 00 00 c6 83 48 05 00 00 00 c6 83 50 05 00 ---------------------------------------------------------------------------
linux-3.12.8-200.fc19.i686 is good. There is no soft lockup.
I have tested linux-3.14.0-0.rc0.git12.1.fc21.i686 . First, I edited arch/x86/kvm/x86.c as follows. static int __vcpu_run(struct kvm_vcpu *vcpu) if (need_resched()) { srcu_read_unlock(&kvm->srcu, vcpu->srcu_idx); //cond_resched(); schedule(); vcpu->srcu_idx = srcu_read_lock(&kvm->srcu); } replacing cond_resched() with schedule(), and build a kernel. Then, run qemu-kvm. The above soft lockup is disappered. Kernel works. I don't know why.
Masao, Can you try reverting https://git.kernel.org/cgit/virt/kvm/kvm.git/patch/?id=01b71917b55d28c09ade9fb8c683cf0d2aad1858
(In reply to Marcelo Tosatti from comment #6) > Masao, > > Can you try reverting > > https://git.kernel.org/cgit/virt/kvm/kvm.git/patch/ > ?id=01b71917b55d28c09ade9fb8c683cf0d2aad1858 I have tried. But, the soft lockup is occurred.
(In reply to Masao Takahashi from comment #7) > (In reply to Marcelo Tosatti from comment #6) > > Masao, > > > > Can you try reverting > > > > https://git.kernel.org/cgit/virt/kvm/kvm.git/patch/ > > ?id=01b71917b55d28c09ade9fb8c683cf0d2aad1858 > > I have tried. > But, the soft lockup is occurred. Please try patch at http://marc.info/?l=linux-kernel&m=139038631607917&q=raw
(In reply to Marcelo Tosatti from comment #8) > (In reply to Masao Takahashi from comment #7) > > (In reply to Marcelo Tosatti from comment #6) > > > Masao, > > > > > > Can you try reverting > > > > > > https://git.kernel.org/cgit/virt/kvm/kvm.git/patch/ > > > ?id=01b71917b55d28c09ade9fb8c683cf0d2aad1858 > > > > I have tried. > > But, the soft lockup is occurred. > > Please try patch at > > http://marc.info/?l=linux-kernel&m=139038631607917&q=raw I have tried on linux-3.14.0-0.rc0.git15.1.fc21.i686 which is applied an above patch. Soft lockup is occurred.
I have resolved this issue. But, there is a question as below. should_resched() procudure is defined in two different files. One : arch/x86/include/asm/preempt.h static __always_inline bool should_resched(void) { return unlikely(!__this_cpu_read_4(__preempt_count)); /* here */ } It is called as should-1 Another: include/asm-generic/preempt.h static __always_inline bool should_resched(void) { return unlikely(!preempt_count() && tif_need_resched()); } This is called as should-2 Then, I replace shoud-1 with should-2 and regenerated a kernel. So, the above softlockup is disappered. What is wrong?
Just to confirm.. I have the same problem during power down. Suspend is OK. Current kernel: 3.12.9-201.fc19.x86_64 The problem came up after I changed the mother board few years! ago. I thought it was HW related. Did'n bother to investgate because of a 7/24 operation. Since I want to poweroff the system remotely it got nasty. I fiddled around with acpi and friends but did not find a working combination. In case it matters cat /proc/cpuinfo processor : 0 vendor_id : AuthenticAMD cpu family : 16 model : 4 model name : AMD Phenom(tm) II X4 965 Processor stepping : 3 microcode : 0x10000c8 cpu MHz : 800.000 ... Plan to upgrade to F20. But not tomorrow...
Got hit by same issue. OS: Fedora 20 with latest updates kernel 3.12.9-301.fc20.i686 => kvm virtual machine starts kernel >= 3.13.x => same report as topic kernel-3.14.0-0.rc5.git2.1.fc21.i686 => same report as topic Is there any newer kernel around to test? System: Dell Latitude D620 CPU: $ cat /proc/cpuinfo processor : 0 vendor_id : GenuineIntel cpu family : 6 model : 15 model name : Intel(R) Core(TM)2 CPU T7400 @ 2.16GHz stepping : 6 microcode : 0xd1 cpu MHz : 2167.000 cache size : 4096 KB ... processor : 1 vendor_id : GenuineIntel cpu family : 6 model : 15 model name : Intel(R) Core(TM)2 CPU T7400 @ 2.16GHz stepping : 6 microcode : 0xd1 cpu MHz : 1000.000 cache size : 4096 KB ..
I am seeing a similar problem this morning on the newest kernel. Fedora (3.12.7-300.fc20.i686+PAE) 20 (Heisenbug) -KVM seems ok -this is what I fell back to and the guests are fine now Fedora (3.13.5-202.fc20.i686+PAE) 20 (Heisenbug) -KVM breaks https://www.dropbox.com/s/0qvej1hem2z1z3c/2014-03-10%2011.17.29.jpg https://www.dropbox.com/s/q4opmf4vd08ax89/2014-03-10%2011.26.00.jpg a couple more shots: https://www.dropbox.com/s/rfu73iqyhxj2zbt/2014-03-10%2010.38.04.jpg https://www.dropbox.com/s/lmkyxe78hdaa2t2/2014-03-10%2010.38.23.jpg Fedora (3.12.10-300.fc20.i686+PAE) 20 (Heisenbug) -KVM also breaks Fedora 3.12.7-300.fc20.i686+PAE also showed: vcpu0 unhandled rdmsr: 0xc0010001 vcpu1 unhandled rdmsr: 0x3a https://www.dropbox.com/s/db2wa9d7go592iw/2014-03-10%2010.13.13.jpg /var/log/messages from the tests: https://www.dropbox.com/s/tiafantivfpos51/messages From 3am today: https://www.dropbox.com/s/x6s7mq0e4cqecxz/messages-20140310 This something that started happening maybe once a month. At first I thought it was the swap size, but right now there is 16GB RAM and swapon reports Size 17825788. Ironcially the newer kernels are worse. The box: Gigabyte GA-78LMT AMD FX(tm)-6100 Six-Core Processor (bogomips 6629.62 per core) 16GB two 3TB SATA drives, mdraid mirrored with LVM on top of the md's 32bit PAE kernel (longish story but a new 64bit install wasn't booting, but installed fine... I fell back to grub2/32bit with PAE... which also matches the kernel in our other systems including blackboxes in customers' closets) guests are 32bit Fedora and XPee Pro two NIC's: one public for VM's access only, 2nd for private local net I wonder if this ticket is indirectly related to the lockups in https://bugzilla.redhat.com/show_bug.cgi?id=1061885 [I have done ~30ish years of Unix, so if you have things you like like me to try/see, please let me know.] Thanks, -Jeff
Fedora-19, 32bit, PAE After upgrade kernel to 3.13.x it is not possible not run guest Fedora-20 32 bit. Even update kernel to vmlinuz-3.13.5-103.fc19.i686.PAE do not help. No issue with guest if host kernel is 3.12.11-201.fc19.i686.PAE --- Mar 10 20:04:27 homedesk kernel: [ 228.040001] BUG: soft lockup - CPU#1 stuck for 22s! [qemu-system-x86:9462] Mar 10 20:04:27 homedesk kernel: [ 228.040001] Modules linked in: vhost_net vhost macvtap macvlan bnep bluetooth fuse ip6table_fil ter ip6_tables ebtable_nat ebtables ipt_MASQUERADE iptable_nat nf_nat_ipv4 nf_nat nf_conntrack_ipv4 nf_defrag_ipv4 xt_conntrack nf_ conntrack xt_CHECKSUM iptable_mangle tun bridge stp llc adt7475 hwmon_vid arc4 raid0 gpio_ich iTCO_wdt iTCO_vendor_support snd_emu1 0k1 coretemp kvm_intel snd_util_mem snd_hwdep snd_rawmidi kvm snd_ac97_codec ac97_bus snd_seq snd_seq_device microcode snd_pcm nvid ia(POF) serio_raw rtl8187 i2c_i801 eeprom_93cx6 uvcvideo mac80211 videobuf2_vmalloc videobuf2_memops videobuf2_core videodev media snd_page_alloc cfg80211 rfkill snd_timer snd soundcore lpc_ich drm r8169 mii emu10k1_gp gameport mfd_core sky2 i2c_core asus_atk011 0 acpi_cpufreq uinput binfmt_misc Mar 10 20:04:27 homedesk kernel: [ 228.040001] CPU: 1 PID: 9462 Comm: qemu-system-x86 Tainted: PF O 3.13.5-103.fc19.i686. PAE #1 Mar 10 20:04:27 homedesk kernel: [ 228.040001] Hardware name: System manufacturer P5K Deluxe/P5K Deluxe, BIOS 0902 06/19/2008 Mar 10 20:04:27 homedesk kernel: [ 228.040001] task: ea7f9680 ti: e70f8000 task.ti: e70f8000 Mar 10 20:04:27 homedesk kernel: [ 228.040001] EIP: 0060:[<f82e90cd>] EFLAGS: 00000202 CPU: 1 Mar 10 20:04:27 homedesk kernel: [ 228.040001] EIP is at kvm_arch_vcpu_ioctl_run+0x24d/0x10c0 [kvm] Mar 10 20:04:27 homedesk kernel: [ 228.040001] EAX: 00000088 EBX: e6d08000 ECX: 0000401c EDX: 00000001 Mar 10 20:04:27 homedesk kernel: [ 228.040001] ESI: 00000000 EDI: 00000000 EBP: e70f9eb0 ESP: e70f9e18 Mar 10 20:04:27 homedesk kernel: [ 228.040001] DS: 007b ES: 007b FS: 00d8 GS: 00e0 SS: 0068 Mar 10 20:04:27 homedesk kernel: [ 228.040001] CR0: 80050033 CR2: 098dc478 CR3: 26c52000 CR4: 000027f0 Mar 10 20:04:27 homedesk kernel: [ 228.040001] Stack: Mar 10 20:04:27 homedesk kernel: [ 228.040001] 00000007 00000000 00000002 0000000c 00000000 ea7f9680 ea7f9680 0000000c Mar 10 20:04:27 homedesk kernel: [ 228.040001] 00000001 f82d9360 00000001 0000000c 00000000 0000000e 0048fce0 ea54e024 Mar 10 20:04:27 homedesk kernel: [ 228.040001] 7ffbfeff fffffffe c047f301 ea7fde80 00000286 e70f9e94 c0480e6e e70f8000 Mar 10 20:04:27 homedesk kernel: [ 228.040001] Call Trace: Mar 10 20:04:27 homedesk kernel: [ 228.040001] [<f82d9360>] ? kvm_set_ioapic_irq+0x30/0x30 [kvm] Mar 10 20:04:27 homedesk kernel: [ 228.040001] [<c047f301>] ? ttwu_do_activate.constprop.87+0x51/0x60 Mar 10 20:04:27 homedesk kernel: [ 228.040001] [<c0480e6e>] ? try_to_wake_up+0x12e/0x210 Mar 10 20:04:27 homedesk kernel: [ 228.040001] [<f82e5d7f>] ? kvm_arch_vcpu_load+0x4f/0x200 [kvm] Mar 10 20:04:27 homedesk kernel: [ 228.040001] [<f82d5d04>] ? vcpu_load+0x44/0x70 [kvm] Mar 10 20:04:27 homedesk kernel: [ 228.040001] [<f82d616e>] kvm_vcpu_ioctl+0x40e/0x4b0 [kvm] Mar 10 20:04:27 homedesk kernel: [ 228.040001] [<c04b7ac4>] ? do_futex+0xf4/0xa90 Mar 10 20:04:27 homedesk kernel: [ 228.040001] [<c040f908>] ? __switch_to+0xb8/0x350 Mar 10 20:04:27 homedesk kernel: [ 228.040001] [<f82d5d60>] ? vcpu_put+0x30/0x30 [kvm] Mar 10 20:04:27 homedesk kernel: [ 228.040001] [<c0578212>] do_vfs_ioctl+0x2e2/0x4c0 Mar 10 20:04:27 homedesk kernel: [ 228.040001] [<c06429bf>] ? file_has_perm+0x7f/0x90 Mar 10 20:04:27 homedesk kernel: [ 228.040001] [<c06433ac>] ? selinux_file_ioctl+0x4c/0xf0 Mar 10 20:04:27 homedesk kernel: [ 228.040001] [<c0578450>] SyS_ioctl+0x60/0x80 Mar 10 20:04:27 homedesk kernel: [ 228.040001] [<c09cffcd>] sysenter_do_call+0x12/0x28 Mar 10 20:04:27 homedesk kernel: [ 228.040001] [<c09c0000>] ? wait_noreap_copyout.isra.7+0x96/0xb1 Mar 10 20:04:27 homedesk kernel: [ 228.040001] Code: 00 83 c0 24 e8 d5 d8 1b c8 fa 66 66 90 66 90 83 7b 1c 02 74 0b 8b 43 20 85 c0 0f 84 86 08 00 00 c7 43 1c 00 00 00 00 fb 66 66 90 <66> 90 8b 03 be 01 00 00 00 83 c0 24 e8 52 d8 1b c8 89 43 18 8b
Same here with 3.13.6-200.fc20.i686
Same problem kernel 3.13.6-100.fc19.i686.PAE, Mate-compis Mate 1.6.2 Nvidia driver 331.49 (GeForce GT610) Core2 Quad ver 6.7.10 kmod-nvidia-3.13.6-100.fc19.i686.PAE-331.49-2.fc19.i686 this is intermittent and not always for the same conditions qemu-system-x86-1.4.2-15.fc19.i686 qemu-kvm-1.4.2-15.fc19.i686 qemu-guest-agent-1.4.2-15.fc19.i686 ipxe-roms-qemu-20130517-2.gitc4bce43.fc19.noarch qemu-kvm-tools-1.4.2-15.fc19.i686 qemu-common-1.4.2-15.fc19.i686 qemu-img-1.4.2-15.fc19.i686 other bug related with kernel update: glibc crashed,ld-2.17.so killed by SIGSEGV,glibc-2.17-20.fc19 syslog for kernel CPU stuck: Message from syslogd@localhost at Mar 26 10:16:57 ... kernel:[174484.048001] BUG: soft lockup - CPU#2 stuck for 22s! [qemu-system-x86:26935] Message from syslogd@localhost at Mar 26 10:16:57 ... kernel:[174484.048001] CPU: 2 PID: 26935 Comm: qemu-system-x86 Tainted: PF O 3.13.6-100.fc19.i686.PAE #1 Message from syslogd@localhost at Mar 26 10:16:57 ... kernel:[174484.048001] Hardware name: /DG35EC, BIOS ECG3510M.86A.0112.2009.0203.1136 02/03/2009 Message from syslogd@localhost at Mar 26 10:16:57 ... kernel:[174484.048001] task: e5aee780 ti: f6cb0000 task.ti: f6cb0000 Message from syslogd@localhost at Mar 26 10:16:57 ... kernel:[174484.048001] Stack: Message from syslogd@localhost at Mar 26 10:16:57 ... kernel:[174484.048001] Call Trace: Message from syslogd@localhost at Mar 26 10:16:57 ... kernel:[174484.048001] Code: 00 83 c0 24 e8 15 c9 29 c8 fa 66 66 90 66 90 83 7b 1c 02 74 0b 8b 43 20 85 c0 0f 84 86 08 00 00 c7 43 1c 00 00 00 00 fb 66 66 90 <66> 90 8b 03 be 01 00 00 00 83 c0 24 e8 92 c8 29 c8 89 43 18 8b
Same with 3.13.7-200.fc20.i686 processor : 0 vendor_id : GenuineIntel cpu family : 6 model : 15 model name : Intel(R) Core(TM)2 Duo CPU T7100 @ 1.80GHz stepping : 13 microcode : 0xa4 cpu MHz : 1800.000 cache size : 2048 KB physical id : 0 siblings : 2 core id : 0 cpu cores : 2 apicid : 0 initial apicid : 0 fdiv_bug : no f00f_bug : no coma_bug : no fpu : yes fpu_exception : yes cpuid level : 10 wp : yes flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe lm constant_tsc arch_perfmon pebs bts aperfmperf pni dtes64 monitor ds_cpl vmx est tm2 ssse3 cx16 xtpr pdcm lahf_lm ida dtherm tpr_shadow vnmi flexpriority bogomips : 3591.20 clflush size : 64 cache_alignment : 64 address sizes : 36 bits physical, 48 bits virtual power management: processor : 1 vendor_id : GenuineIntel cpu family : 6 model : 15 model name : Intel(R) Core(TM)2 Duo CPU T7100 @ 1.80GHz stepping : 13 microcode : 0xa4 cpu MHz : 800.000 cache size : 2048 KB physical id : 0 siblings : 2 core id : 1 cpu cores : 2 apicid : 1 initial apicid : 1 fdiv_bug : no f00f_bug : no coma_bug : no fpu : yes fpu_exception : yes cpuid level : 10 wp : yes flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe lm constant_tsc arch_perfmon pebs bts aperfmperf pni dtes64 monitor ds_cpl vmx est tm2 ssse3 cx16 xtpr pdcm lahf_lm ida dtherm tpr_shadow vnmi flexpriority bogomips : 3591.20 clflush size : 64 cache_alignment : 64 address sizes : 36 bits physical, 48 bits virtual power management:
*** Bug 1082268 has been marked as a duplicate of this bug. ***
*** Bug 1082272 has been marked as a duplicate of this bug. ***
Hi, Affecting me also. I'm stuck back on 3.12.11-201 to keep my VMs operational. Checked after last kernel update and it is still broken on Kernel 3.13.7-200 with the same dmesg and stdout messages as detailed in other reports. CPU in question is: processor : 0 vendor_id : GenuineIntel cpu family : 6 model : 23 model name : Intel(R) Core(TM)2 Duo CPU E8400 @ 3.00GHz stepping : 10 microcode : 0xa07 Regards Dave
Quick update: Still in Kernel 3.13.9-200 Regards, Dave Harley.
it looks like only 32bit,PAE is affected. After migration to F20 64bit issue is gone
Shortly after upgrade to F20 64Bit there is still the same problem. Sometimes CPU#3 gets stuck. But to dig deeper into it I first have to fix the "Oh no! ..... " problem.
I've noticed that lockup occurs more likely with higher disk usage (system upgrade for example).
Seems to be solved with 3.13.10-200.fc20.x86_64. I have powered off from local and remote session several times. It is always a clean poweroff - no excessive waits, no complaining messages. Thank you.
(In reply to H.-P. Sorge from comment #25) > Seems to be solved with 3.13.10-200.fc20.x86_64. > > I have powered off from local and remote session several times. > > It is always a clean poweroff - no excessive waits, no complaining messages. > > Thank you. I testd as below. But the problem still exists. 1. kernel linux-3.13.11-200.fc19.i686 2. cpuinfo processor : 0 vendor_id : AuthenticAMD cpu family : 15 model : 107 model name : AMD Athlon(tm) 64 X2 Dual Core Processor 5600+ stepping : 2 cpu MHz : 1000.000 cache size : 512 KB physical id : 0 siblings : 2 core id : 0 cpu cores : 2 apicid : 0 initial apicid : 0 fdiv_bug : no f00f_bug : no coma_bug : no fpu : yes fpu_exception : yes cpuid level : 1 wp : yes flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt rdtscp lm 3dnowext 3dnow extd_apicid pni cx16 lahf_lm cmp_legacy svm extapic cr8_legacy 3dnowprefetch lbrv bogomips : 2004.11 clflush size : 64 cache_alignment : 64 address sizes : 40 bits physical, 48 bits virtual power management: ts fid vid ttp tm stc 100mhzsteps processor : 1 vendor_id : AuthenticAMD cpu family : 15 model : 107 model name : AMD Athlon(tm) 64 X2 Dual Core Processor 5600+ stepping : 2 cpu MHz : 1000.000 cache size : 512 KB physical id : 0 siblings : 2 core id : 1 cpu cores : 2 apicid : 1 initial apicid : 1 fdiv_bug : no f00f_bug : no coma_bug : no fpu : yes fpu_exception : yes cpuid level : 1 wp : yes flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt rdtscp lm 3dnowext 3dnow extd_apicid pni cx16 lahf_lm cmp_legacy svm extapic cr8_legacy 3dnowprefetch lbrv bogomips : 2004.11 clflush size : 64 cache_alignment : 64 address sizes : 40 bits physical, 48 bits virtual power management: ts fid vid ttp tm stc 100mhzsteps
I also confirmed the 32bit PAE host kernel-PAE-3.13.10-200.fc20.i686 still fails when running a 32bit guest kernel. XPee Pro SP3 guest seems fine. kernel-PAE-3.12.7-300.fc20.i686 is mostly stable and only ~monthly will crash all the guests and run slow until rebooted. 3.12.7 seems to only do this under heavy guest I/O, but not that different of a pattern from other loads at least at a high level (Munin graphs pre-crash, etc.)
I have tried a new kernel as below. Then softlockup problem is disappered. 1. kernel linux-3.15.0-0.rc5.git2.8.fc21.i686 2. Patch applied kvm/linux-next here is a link. https://git.kernel.org/cgit/linux/kernel/git/next/linux-next.git/commit/?id=0091d63ea508b831b63dbb3e23e204da51ce1521 Please any one try this.
I've tried the 3.14.5-200.fc20.i686 and it seems to be resolved. Thanks.
Seems fixed by: [jwboyer@vader linux]$ git log v3.14.4..v3.14.5 arch/x86/kvm/ commit 8c02c2a4f89a4eda43b4679e8f0e170edeebc85f Author: Marcelo Tosatti <mtosatti> Date: Thu Apr 10 18:19:12 2014 -0300 KVM: x86: remove WARN_ON from get_kernel_ns() commit b351c39cc9e0151cee9b8d52a1e714928faabb38 upstream. Function and callers can be preempted. https://bugzilla.kernel.org/show_bug.cgi?id=73721 Signed-off-by: Marcelo Tosatti <mtosatti> Reviewed-by: Paolo Bonzini <pbonzini> Signed-off-by: Greg Kroah-Hartman <gregkh>
(In reply to Josh Boyer from comment #30) > Seems fixed by: > > [jwboyer@vader linux]$ git log v3.14.4..v3.14.5 arch/x86/kvm/ > commit 8c02c2a4f89a4eda43b4679e8f0e170edeebc85f > Author: Marcelo Tosatti <mtosatti> > Date: Thu Apr 10 18:19:12 2014 -0300 > > KVM: x86: remove WARN_ON from get_kernel_ns() > > commit b351c39cc9e0151cee9b8d52a1e714928faabb38 upstream. > > Function and callers can be preempted. > > https://bugzilla.kernel.org/show_bug.cgi?id=73721 > > Signed-off-by: Marcelo Tosatti <mtosatti> > Reviewed-by: Paolo Bonzini <pbonzini> > Signed-off-by: Greg Kroah-Hartman <gregkh> OK, I have tested linux-3.14.6-200.fc20.i686. This issue is resolved.