+++ This bug was initially created as a clone of Bug 15734 comment 39 +++
I tried to install Mageia 5 on my new laptop (UEFI) an ASUS N551J with two graphical cards (intel and nvidia). I have tried every option mentioned in the errata (hybrid graphics). I have tried the Live KDE 64 DVD and the classical 64 DVD. Nothing works. I am now 'doomed' to use Mint, that works at once, but I like Mageia much better (if it is installable).
Sometimes at boottime (LiveDVD) I get many nouveau errors and after a while I get a KDE desktop (intel detected?). I can than install it and after reboot I sometimes can configure a user and can log in into KDE. After a few seconds everything freezes. Poweroff and poweron and than it freezes already at boottime (blank screen or with nouveau errors with sometimes I see kernel panic).
Sometimes at boottime (LiveDVD) I get less nouveau errors and it freezes already (no KDE).
I tried boot options as suggested 'optimus' and/or 'rd.blacklist=nouveau rd.blacklist=nvidia xdriver=intel' â> No effect. Still nouveau errors at boottime and/or freezes.
Installation works until after reboot. Same problems as with the LiveDVD.
I tried the boot options, no effect. At the point of the installer to configure the graphics, it suggested 'Intel 810 and later' and not 'use intel only' as told in the errata. Even with Intel selected I get nouveau errors at boottime, with or without boot options.
It seems a problem like mentioned in the errata with the hybrid graphics (Intel+Nvidia), but the suggested options does not work.
Linux Mint 17.1:
No problem, works at once. No errors.
Please also refer to:
attachment 6786 [details] boot screen LiveKDE64DVD on ASUS N551J nouveau errors (part1)
attachment 6787 [details] boot screen LiveKDE64DVD on ASUS N551J nouveau errors (part2)
attachment 6788 [details] Xorg log of LiveKDE64DVD on ASUS N551J (if I am lucky to get so far)
Same issues on my new laptop:
MSI GE70 2PC
4th Gen Core Processor Integrated Graphics Controller
GeForce GTX 850M
Went back to Mageia 4, no issues there, as there never were.
A quick search found this:
My recollection is that using rd.blacklist didn't work for me, because the nouveau driver got loaded later in the boot sequence. You could try adding
to /etc/modprobe.conf on the installed system to see if that helps.
I see in Peter's log that he is also experiencing bug 16033, although that is probably not important at this stage.
Just done some experiments on my laptop, and I need both the boot option *and* the entry in /etc/modprobe.conf to prevent the nouveau module being loaded.
Furthermore, the required boot option is "rdblacklist=nouveau", not "rd.blacklist=nouveau".
It is working now :-)
I use the boot options 'rdblacklist=nouveau nouveau.runpm=0 xdriver=intel'
nouveau.runpm=0 is a tip from doktor500, with this the LiveDVD/USB gives nouveau errors but I get at least at every boot a KDE desktop.
The same boot options I use for the installed system.
Right after install with the first boot I selected 'Safe Mode' and added 'blacklist nouveau' to /etc/modprobe.conf.
After this I get no errors and get a beautiful Mageia KDE desktop.
Thanks for all the help.
Installed Mageia 5 in a free partition, so i could leave Mageia 4 in place, in case of.
But, so far the abovementioned workaround worked.
Acer aspire 5750g installed on external usb3.0 disk, same problem here
black listed nouveau workarounded here also
I have the model from Asus called N550JK and I have the exact same problem.
My fix was to add blacklist=nouveau the kernel boot line.
Is there an ETA on this fix?
Before the release of Mageia 6?
Created attachment 7386 [details]
Here's my Xorg.log
Nouveau errors at boot time and complete desktop freeze on ASUS N551J with optimus hybrid graphics =>
Nouveau errors at boot time and complete desktop freeze on several systems with optimus hybrid graphics
*** Bug 18808 has been marked as a duplicate of this bug. ***
*** Bug 21438 has been marked as a duplicate of this bug. ***
Same problem for me.
Adding kernel option nouveau.runpm=0 worked for me (seen here : https://bugzilla.redhat.com/show_bug.cgi?id=1525213)
This bug is also discussed on launchpad:
They recommand kernel option nouveau.modeset=0.
It also works.
Also discussed at bumblebee:
They recommand kernel option acpi_rev_override=5.
It also works... Don't know which option is the best.
Did not yet tested the alternatives proposed in this thread of this forum, I will and report here the result as soon as possible.
However, I think it is important to report that the new mageia 7 beta3 iso does not works without tweaks that only power users know about.
Mageia 7 beta 2 live did work only when I did
rd.driver.blacklist=nouveau driver.blacklist=nouveau xorg.blacklist=nouveau
at booting the live mageia.
Now this trick no longer works for MGA 7 beta3.
I have two laptops with more than one video card. The old one has optimus technology and Mageia 7 beta3 goes fine without any trick. The new one is a hybrid graphics laptop, with a nvidia dedicated video card and an intel integrated video card and Mageia 7 beta 3 live does not go.
I am also seeing same problem on my Lenovo Yoga 720 15" inch laptop that I have updated from mga6 to mga7.
I have these kernel boot options
$ cat /proc/cmdline
BOOT_IMAGE=/boot/vmlinuz-5.1.3-desktop-6.mga7 root=UUID=73f3f3b4-e763-40b6-99f7-600c14605175 ro splash quiet noiswmd resume=/dev/nvme0n1p2 audit=0
and lspci shows following info from my graphic cards
$ lspci -v | grep -i nvi
01:00.0 3D controller: NVIDIA Corporation GP107M [GeForce GTX 1050 Mobile] (rev a1)
Kernel modules: nvidiafb, nouveau
$ lspci -v | grep -i vga
00:02.0 VGA compatible controller: Intel Corporation HD Graphics 630 (rev 04) (prog-if 00 [VGA controller])
In xfdrake I have selected "Intel 810" or newer for my graphic card.
I can always get in to gdm login screen but then the system will most of the time freeze while logging in to Wayland gnome session. (not always)
Now I am logged in to X11 Gnome and after some strugling (looked almost like freezing until recovered) I was able to login.
But it seems that any graphic intensive tasks like dmesg on terminal will try to freeze the system and I think this is because it try and fails to use the nouveau graphics card and fails.
I have checked from the kernel driver and I am pretty sure it is in stuck on this type of places where the graphic is tried to be flushed and which has 2000 msec timetout before giving up.
if (nvkm_rd32(device, 0x100c80) & 0x00008000)
[ 266.099158] nouveau 0000:01:00.0: timeout
[ 266.099212] WARNING: CPU: 1 PID: 1 at drivers/gpu/drm/nouveau/nvkm/subdev/mmu/vmmgf100.c:224 gf100_vmm_invalidate+0x1ba/0x1d0 [nouveau]
[ 266.099213] Modules linked in: fuse ib_core xt_recent xt_comment ip6t_REJECT nf_reject_ipv6 xt_hashlimit ip6table_raw nf_log_ipv6 xt_mark xt_CT iptable_raw xt_multiport nfnetlink_log xt_NFLOG nf_log_ipv4 nf_log_common xt_LOG nf_nat_tftp nf_nat_snmp_basic nf_conntrack_snmp nf_nat_sip nf_nat_pptp nf_nat_irc nf_nat_h323 nf_nat_ftp nf_nat_amanda ts_kmp nf_conntrack_amanda nf_conntrack_sane nf_conntrack_tftp nf_conntrack_sip nf_conntrack_pptp nf_conntrack_netbios_ns nf_conntrack_broadcast nf_conntrack_irc nf_conntrack_h323 nf_conntrack_ftp ccm cmac rfcomm xt_CHECKSUM af_packet ipt_REJECT nf_reject_ipv4 ipt_MASQUERADE nf_conntrack_netlink xt_tcpudp xfrm_user xfrm_algo xt_addrtype xt_conntrack tun br_netfilter bridge stp llc ip6table_mangle ip6table_nat aufs iptable_mangle iptable_nat nf_nat nf_conntrack nf_defrag_ipv6 nf_defrag_ipv4 nf_tables nfnetlink ip6table_filter ip6_tables iptable_filter bnep msr sunrpc nls_iso8859_1 nls_cp437 vfat fat arc4 joydev hid_sensor_accel_3d hid_sensor_als
[ 266.099223] hid_sensor_trigger industrialio_triggered_buffer kfifo_buf hid_sensor_iio_common industrialio wacom hid_multitouch 8250_dw usbhid ath10k_pci hid_sensor_hub ath10k_core hid_generic iTCO_wdt ath mei_hdcp iTCO_vendor_support intel_rapl mac80211 x86_pkg_temp_thermal intel_powerclamp coretemp snd_hda_codec_hdmi kvm_intel uvcvideo videobuf2_vmalloc videobuf2_memops snd_hda_codec_realtek videobuf2_v4l2 kvm snd_hda_codec_generic videobuf2_common ledtrig_audio btusb videodev btbcm btrtl snd_hda_intel btintel bluetooth media irqbypass intel_cstate snd_hda_codec cfg80211 intel_uncore intel_rapl_perf snd_hda_core snd_hwdep input_leds ecdh_generic snd_pcm snd_timer idma64 virt_dma i2c_i801 snd mei_me processor_thermal_device mei intel_pch_thermal intel_soc_dts_iosf soundcore intel_lpss_pci wmi_bmof intel_lpss intel_wmi_thunderbolt ideapad_laptop thermal fan sparse_keymap rfkill battery int3403_thermal int340x_thermal_zone tpm_crb ac tpm_tis tpm_tis_core tpm pinctrl_sunrisepoint
[ 266.099234] pinctrl_intel int3400_thermal acpi_thermal_rel evdev acpi_pad sch_fq_codel input_polldev nvram efivarfs ip_tables x_tables ipv6 crc_ccitt autofs4 algif_skcipher af_alg dm_crypt ipmi_msghandler nouveau crc32_pclmul crc32c_intel ghash_clmulni_intel xhci_pci xhci_hcd aesni_intel usbcore aes_x86_64 mxm_wmi crypto_simd cryptd glue_helper ttm serio_raw usb_common i915 i2c_hid hid i2c_algo_bit drm_kms_helper wmi video button drm dm_mirror dm_region_hash dm_log dm_mod
[ 266.099242] CPU: 1 PID: 1 Comm: systemd Tainted: P W O 5.1.3-desktop-6.mga7 #1
[ 266.099242] Hardware name: LENOVO 80X7/Lenovo YOGA 720-15IKB, BIOS 4MCN19WW(V1.05) 03/01/2017
[ 266.099269] RIP: 0010:gf100_vmm_invalidate+0x1ba/0x1d0 [nouveau]
[ 266.099270] Code: 40 10 48 8b 78 10 48 8b 5f 50 48 85 db 75 04 48 8b 5f 10 e8 98 57 01 cd 48 89 da 48 c7 c7 e0 82 7c c0 48 89 c6 e8 20 11 a2 cc <0f> 0b e9 5a ff ff ff e8 0a 0e a2 cc 66 2e 0f 1f 84 00 00 00 00 00
[ 266.099271] RSP: 0018:ffffb14e0190fa48 EFLAGS: 00010286
[ 266.099271] RAX: 0000000000000000 RBX: ffff8eacabc6bed0 RCX: 0000000000000006
[ 266.099272] RDX: 0000000000000007 RSI: 0000000000000096 RDI: ffff8eacaea564d0
[ 266.099272] RBP: ffff8eaca71d5620 R08: 0000000000000ddf R09: 0000000000000004
[ 266.099272] R10: 0000000000000000 R11: 0000000000000001 R12: 0000000000000005
[ 266.099273] R13: ffff8eaca71d7400 R14: ffff8eaca79fd480 R15: 0003ffffffffffff
[ 266.099274] FS: 00007f9549d2c880(0000) GS:ffff8eacaea40000(0000) knlGS:0000000000000000
[ 266.099274] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[ 266.099274] CR2: 000000000140b000 CR3: 00000004ae42e005 CR4: 00000000003606e0
[ 266.099275] Call Trace:
[ 266.099301] nvkm_vmm_iter.constprop.12+0x364/0x820 [nouveau]
[ 266.099326] ? gf100_vmm_invalidate_pdb+0x30/0x30 [nouveau]
[ 266.099349] ? nvkm_vmm_map_choose+0xa0/0xa0 [nouveau]
[ 266.099373] ? gf100_vmm_invalidate_pdb+0x30/0x30 [nouveau]
[ 266.099397] nvkm_vmm_ptes_unmap_put+0x2c/0x40 [nouveau]
[ 266.099420] ? nvkm_vmm_map_choose+0xa0/0xa0 [nouveau]
[ 266.099443] ? gf100_vmm_invalidate_pdb+0x30/0x30 [nouveau]
[ 266.099466] nvkm_vmm_put_locked+0x203/0x240 [nouveau]
[ 266.099489] nvkm_vmm_put+0x30/0x50 [nouveau]
[ 266.099510] nv50_instobj_dtor+0x9b/0xe0 [nouveau]
[ 266.099524] nvkm_memory_unref+0x3f/0x60 [nouveau]
[ 266.099546] nvkm_mmu_ptc_put+0xfb/0x160 [nouveau]
[ 266.099599] nvkm_vmm_del+0x124/0x180 [nouveau]
[ 266.099624] nvkm_vmm_unref+0x30/0x40 [nouveau]
[ 266.099682] nvkm_uvmm_dtor+0x12/0x20 [nouveau]
[ 266.099696] nvkm_object_dtor+0xad/0x100 [nouveau]
[ 266.099710] nvkm_object_del+0x20/0x80 [nouveau]
[ 266.099723] nvkm_ioctl_del+0x42/0x50 [nouveau]
[ 266.099736] nvkm_ioctl+0xd8/0x170 [nouveau]
[ 266.099749] nvif_object_fini+0x59/0x80 [nouveau]
[ 266.099776] nouveau_vmm_fini+0x12/0x20 [nouveau]
[ 266.099803] nouveau_cli_fini+0x48/0x90 [nouveau]
[ 266.099831] nouveau_drm_postclose+0xa6/0xe0 [nouveau]
[ 266.099839] drm_file_free.part.7+0x206/0x2e0 [drm]
[ 266.099845] drm_release+0x4b/0x80 [drm]
[ 266.099846] __fput+0xaf/0x1f0
[ 266.099848] task_work_run+0x84/0xa0
[ 266.099849] exit_to_usermode_loop+0xeb/0xf0
[ 266.099850] do_syscall_64+0xe8/0x110
[ 266.099852] entry_SYSCALL_64_after_hwframe+0x44/0xa9
[ 266.099853] RIP: 0033:0x7f954ab46347
[ 266.099853] Code: 00 00 0f 05 48 3d 00 f0 ff ff 77 3f c3 66 0f 1f 44 00 00 53 89 fb 48 83 ec 10 e8 54 fc ff ff 89 df 89 c2 b8 03 00 00 00 0f 05 <48> 3d 00 f0 ff ff 77 34 89 d7 89 44 24 0c e8 96 fc ff ff 8b 44 24
[ 266.099854] RSP: 002b:00007ffdd7023cf0 EFLAGS: 00000293 ORIG_RAX: 0000000000000003
[ 266.099855] RAX: 0000000000000000 RBX: 0000000000000078 RCX: 00007f954ab46347
[ 266.099855] RDX: 0000000000000000 RSI: 000000000206b010 RDI: 0000000000000078
[ 266.099856] RBP: 00007f9549d2c800 R08: 00000000020b76c0 R09: 0000000000000006
[ 266.099856] R10: 0000000000000000 R11: 0000000000000293 R12: 0000000000000000
[ 266.099856] R13: 0000000000000000 R14: 0000000000000000 R15: 000000000210a958
[ 266.099857] ---[ end trace 6905e7a7704896dd ]---
[lamikr@linux ~]$ rpm -qa | grep gdm
Just try this one required boot option is "rdblacklist=nouveau", not "rd.blacklist=nouveau"
Seems there are some valid fixes: by adding keyword to kernel cmdline on
/etc/default/grub and also editing it on GRUB.
(In reply to Sébastien GUERIN from comment #12)
> Same problem for me.
> Adding kernel option nouveau.runpm=0 worked for me (seen here :
(In reply to Sébastien GUERIN from comment #13)
> This bug is also discussed on launchpad:
> They recommand kernel option nouveau.modeset=0.
> It also works.
(In reply to Sébastien GUERIN from comment #14)
> Also discussed at bumblebee:
> They recommand kernel option acpi_rev_override=5.
> It also works... Don't know which option is the best.
(In reply to Adelson Oliveira from comment #15)
> Did not yet tested the alternatives proposed in this thread of this forum, I
> will and report here the result as soon as possible.
> However, I think it is important to report that the new mageia 7 beta3 iso
> does not works without tweaks that only power users know about.
> Mageia 7 beta 2 live did work only when I did
> rd.driver.blacklist=nouveau driver.blacklist=nouveau xorg.blacklist=nouveau
> at booting the live mageia.
> Now this trick no longer works for MGA 7 beta3.
So closing this fixed.