WelcomeWelcome | FAQFAQ | DownloadsDownloads | WikiWiki

Author Topic: i915 firmware bug  (Read 2326 times)

Offline yacc1

  • Newbie
  • *
  • Posts: 2
i915 firmware bug
« on: May 09, 2020, 05:44:45 AM »
Hello
Trying to start TC11 (USB stick) on Lenovo T440, it fails with a fuzzy display.
Looking in dmesg report, there are lines indicating that i915 firmware initialization failed and that a bug report should be sent :

[drm:0xfe79b77c] *ERROR* Failed to map the ggtt page table
i915 0000:00:02.0: Device initialization failed (-12)
i915 0000:00:02.0: Please file a bug at https://bugs.freedesktop.org/enter_bug.cgi?product=DRI against DRM/Intel providing the dmesg log by booting with drm.debug=0xf


I understand that it is not really a TCE bug but instead a Linux one.
But before escalating to Linux development people, I would like a confimation from TC team (I'm really newbie with Linux)
The attached dmesg file has been produced with the requested boot option drm_debug=0xf

Thank you
« Last Edit: May 09, 2020, 05:46:39 AM by yacc1 »

Offline Juanito

  • Administrator
  • Hero Member
  • *****
  • Posts: 14851
Re: i915 firmware bug
« Reply #1 on: May 09, 2020, 08:47:05 AM »
Do you need the firmware-i915 extension "onboot"?

Online Rich

  • Administrator
  • Hero Member
  • *****
  • Posts: 11702
Re: i915 firmware bug
« Reply #2 on: May 09, 2020, 08:57:29 AM »
Hi yacc1
Welcome to the forum. See if adding this boot code helps:
Code: [Select]
intel_iommu=off

Offline curaga

  • Administrator
  • Hero Member
  • *****
  • Posts: 11049
Re: i915 firmware bug
« Reply #3 on: May 09, 2020, 02:27:31 PM »
Quote
vmap allocation for size 2101248 failed: use vmalloc=<size> to increase size
Try
Code: [Select]
vmalloc=256Mtoo.
The only barriers that can stop you are the ones you create yourself.

Offline yacc1

  • Newbie
  • *
  • Posts: 2
Re: i915 firmware bug
« Reply #4 on: May 09, 2020, 03:53:41 PM »
Hello
Thank you for the answers and happy not to have raised a bug report!
@Rich : intel_iommu=off alone, still an error in dmesg (code -5) but display Ok, low resolution
@Curage : vmalloc=256M alone, no more error, display OK high resolution
vmalloc=256M + intel_iommu=off, no difference with vmalloc=256M alone

@Juanito : to my (limited) understanding, firmware execution is part of the boot process, so the required extensions shall be loaded 'on boot', no?

Anyway, everything seems fine now.
Thanks again for your prompt return.

EDIT : @Juanito : actually you are right, this works fine without loading the firmware-i915.tcz on boot. Mystery for me, this means that Linux already holds the firmware in the kernel?
« Last Edit: May 09, 2020, 04:17:27 PM by yacc1 »