1002:9583 [iMac7,1] X fails to automatically detect ATI radeon RV600 card

Bug #1453052 reported by Raffaele Candeliere
10
This bug affects 1 person
Affects Status Importance Assigned to Milestone
xorg-server (Ubuntu)
Won't Fix
High
Unassigned

Bug Description

Good day.
I'm testing pure EFI boot capability on my iMac 7.1 (graphic card ATI radeon HD2600 pro).
So far, with Ubuntu 14.10, i've been able to boot directly into a perfectly working system using the EFI kernel stub.
No GRUB, no ELILO, no rEFIND, no boot manager whatsoever! And the result was amazing!
Boot up time reduced to a fraction of the time needed by GRUB, for example, Compiz compositing much more quick, smooth and and efficient. I could notice even some graphical detail (e.g. Text shadows in the GDM log-in screen) that were not present if booting via GRUB.
The way to accomplished this is simple:
1) Create a small (40MB or less) HFS+ partition. E.g. #> sudo mkfs.hfsplus /dev/sdaX -v "Ubuntu (EFI-Stub)"
2) Copy to the new partition the files vmlinuz-3.XX.YY-ZZ-generic.efi.signed and initrd.img-3.XX.YY-ZZ-generig. Also rename them to "vmlinuz.efi" and "initrd.img" for sake of simplicity in further command line work. Furthermore, Apple's EFI won't boot anything that doesn't end with the ".efi" extension.
3) Create a boot entry in the NVRAM using efibootmgr like this: #> sudo efibootmgr -d /dev/sdX -p Y -c -L "Ubuntu 14.04 LTS" -l /vmlinuz.efi -u "root=UUID=b0c4efb9-95b0-4e5d-98e1-8beb43f413c2 ro rootflags=subvol=@ quiet splash initrd=\initrd.img"
4) reboot (and enjoy!)

As i've said, in Ubuntu 14 all this is working fine, but after the upgrade to Ubuntu 15, "X" doesn't start anymore. The system starts up correctly but X fails (with error: (EE) No devices detected. (EE) Fatal server error: (EE) no screens found) dropping me to a tty console login.
I'm not shure if it is a X problem or a kernel KMS bug, but i'm more inclined to think of a problem with X because i've tried to boot Ubuntu 15.0 with the old kernel (3.13/16) and the problem is still there.
I've also tried to add "vt.handoff=7" to the command line args for the kernel, like grub does, but the result is even worse. The system freezes completely at the moment of starting X.
Please note also that booting via grub is working. X starts, much more slowly but starts. (Actually i'm reporting this bug from the very same system booted using GRUB. I have a second HFS+ partition where i installed grub and i can access it by holding down the ALT key at power-up)
Thanks for your work. I really hope that booting directly via EFI stubs will grow mature because the performance increase is really big.
Thanks again.

ProblemType: Bug
DistroRelease: Ubuntu 15.04
Package: xorg 1:7.7+7ubuntu4
ProcVersionSignature: Ubuntu 3.19.0-16.16-generic 3.19.3
Uname: Linux 3.19.0-16-generic x86_64
NonfreeKernelModules: wl
.tmp.unity.support.test.0:

ApportVersion: 2.17.2-0ubuntu1
Architecture: amd64
CompizPlugins: [core,composite,opengl,compiztoolbox,decor,vpswitch,snap,mousepoll,resize,place,move,wall,grid,regex,imgpng,session,gnomecompat,animation,fade,staticswitcher,workarounds,scale,expo,ezoom,dbus]
CompositorRunning: compiz
CompositorUnredirectDriverBlacklist: '(nouveau|Intel).*Mesa 8.0'
CompositorUnredirectFSW: true
CurrentDesktop: Unity
Date: Fri May 8 10:20:35 2015
DistUpgraded: 2015-05-06 18:46:56,830 DEBUG enabling apt cron job
DistroCodename: vivid
DistroVariant: ubuntu
DkmsStatus:
 bcmwl, 6.30.223.248+bdcom, 3.19.0-16-generic, x86_64: installed
 bcmwl, 6.30.223.248+bdcom, 3.19.0-16-lowlatency, x86_64: installed
ExtraDebuggingInterest: Yes, including running git bisection searches
GraphicsCard:
 Advanced Micro Devices, Inc. [AMD/ATI] RV630/M76 [Mobility Radeon HD 2600 XT/2700] [1002:9583] (prog-if 00 [VGA controller])
   Subsystem: Apple Inc. iMac 7,1 [106b:0083]
InstallationDate: Installed on 2013-10-27 (557 days ago)
InstallationMedia: Ubuntu 13.10 "Saucy Salamander" - Release amd64 (20131016.1)
MachineType: Apple Inc. iMac7,1
ProcEnviron:
 PATH=(custom, no user)
 XDG_RUNTIME_DIR=<set>
 LANG=it_IT.UTF-8
 SHELL=/bin/bash
ProcKernelCmdLine: BOOT_IMAGE=/vmlinuz-3.19.0-16-generic.efi.signed root=UUID=b0c4efb9-95b0-4e5d-98e1-8beb43f413c2 ro rootflags=subvol=@ quiet splash vt.handoff=7
SourcePackage: xorg
UpgradeStatus: Upgraded to vivid on 2015-05-06 (1 days ago)
dmi.bios.date: 03/05/08
dmi.bios.vendor: Apple Inc.
dmi.bios.version: IM71.88Z.007A.B03.0803051705
dmi.board.asset.tag: Base Board Asset Tag
dmi.board.name: Mac-F4238CC8
dmi.board.vendor: Apple Inc.
dmi.board.version: PVT
dmi.chassis.asset.tag: Asset Tag#
dmi.chassis.type: 13
dmi.chassis.vendor: Apple Inc.
dmi.chassis.version: Mac-F4238CC8
dmi.modalias: dmi:bvnAppleInc.:bvrIM71.88Z.007A.B03.0803051705:bd03/05/08:svnAppleInc.:pniMac7,1:pvr1.0:rvnAppleInc.:rnMac-F4238CC8:rvrPVT:cvnAppleInc.:ct13:cvrMac-F4238CC8:
dmi.product.name: iMac7,1
dmi.product.version: 1.0
dmi.sys.vendor: Apple Inc.
version.compiz: compiz 1:0.9.12.1+15.04.20150410.1-0ubuntu1
version.ia32-libs: ia32-libs N/A
version.libdrm2: libdrm2 2.4.60-2
version.libgl1-mesa-dri: libgl1-mesa-dri 10.5.2-0ubuntu1
version.libgl1-mesa-dri-experimental: libgl1-mesa-dri-experimental N/A
version.libgl1-mesa-glx: libgl1-mesa-glx 10.5.2-0ubuntu1
version.xserver-xorg-core: xserver-xorg-core 2:1.17.1-0ubuntu3
version.xserver-xorg-input-evdev: xserver-xorg-input-evdev 1:2.9.0-1ubuntu2
version.xserver-xorg-video-ati: xserver-xorg-video-ati 1:7.5.0-1ubuntu2
version.xserver-xorg-video-intel: xserver-xorg-video-intel 2:2.99.917-1~exp1ubuntu2build1
version.xserver-xorg-video-nouveau: xserver-xorg-video-nouveau 1:1.0.11-1ubuntu2build1
xserver.bootTime: Fri May 8 09:58:18 2015
xserver.configfile: default
xserver.errors:

xserver.logfile: /var/log/Xorg.0.log
xserver.version: 2:1.17.1-0ubuntu3
xserver.video_driver: radeon

Revision history for this message
Raffaele Candeliere (candeliere) wrote :
Revision history for this message
Raffaele Candeliere (candeliere) wrote :
Revision history for this message
Raffaele Candeliere (candeliere) wrote :
Revision history for this message
Raffaele Candeliere (candeliere) wrote :
Revision history for this message
Raffaele Candeliere (candeliere) wrote :
Revision history for this message
Raffaele Candeliere (candeliere) wrote :
Revision history for this message
Raffaele Candeliere (candeliere) wrote :
Revision history for this message
Raffaele Candeliere (candeliere) wrote :
Revision history for this message
Raffaele Candeliere (candeliere) wrote :
description: updated
Revision history for this message
penalvch (penalvch) wrote :

Raffaele Candeliere, thank you for reporting this and helping make Ubuntu better. Does this issue still occur if you do a new install of Vivid via http://releases.ubuntu.com/vivid/ubuntu-15.04-desktop-amd64.iso ?

tags: added: regression-release
removed: xorg
Changed in xorg (Ubuntu):
importance: Undecided → Medium
status: New → Incomplete
Revision history for this message
Raffaele Candeliere (candeliere) wrote :

Mmhhh... I don't know. Actually, I haven't tried to reinstall the whole system from scratch. I can try, but i cannot promise to be quick because I'll have to backup everything first.

Revision history for this message
Raffaele Candeliere (candeliere) wrote :

By the way, another (I hope) useful hint could be the fact that KMS is initializing the graphic card correctly: during the boot process I can see the magenta (or whatever color it is) screen with the Ubuntu logo and the typical "five flashing dots".
At the moment of starting the DislpayManager the screen turns black and Systemd drops me into a tty console login.

Revision history for this message
RaffaeleC (raffaele-candeliere) wrote :

Ok. Here I am with fresh test results.
I've swapped everything and reinstalled Vivid from scratch, from the amd64.iso.
The bug is still there. Something has changed but the main problem is still there. This time systemd, instead of simply dropping me to a console login, complains about the system running in low graphics mode and asks me what to do next (whether to continue in low graphics mode, to display error logs or fall back to a tty login).
However, strangely enough, I've noticed that in the last update the efi stub has disappeared from the kernel. The latest kernel (3.19.0-18) version doesn't install an efi stub, so I've been forced to use the old 3.19.0-16 version.
Problems with efi stubs? ?field.comment=Ok. Here I am with fresh test results.
I've swapped everything and reinstalled Vivid from scratch, from the amd64.iso.
The bug is still there. Something has changed but the main problem is still there. This time systemd, instead of simply dropping me to a console login, complains about the system running in low graphics mode and asks me what to do next (whether to continue in low graphics mode, to display error logs or fall back to a tty login).
However, strangely enough, I've noticed that in the last update the efi stub has disappeared from the kernel. The latest kernel (3.19.0-18) version doesn't install an efi stub, so I've been forced to use the old 3.19.0-16 version.
Problems with efi stubs?

Revision history for this message
RaffaeleC (raffaele-candeliere) wrote :
Revision history for this message
penalvch (penalvch) wrote :

RaffaeleC, if you install with UEFI/Secure Boot disabled in the BIOS, does that allow you to test the latest kernel?

Revision history for this message
RaffaeleC (raffaele-candeliere) wrote :

(Sorry for the typos but apparently there's no way to edit one's own comments after publishing :-( )

Revision history for this message
Raffaele Candeliere (candeliere) wrote :

Hi, Christopher.
What do you really mean by "testing with the latest kernel"? I try to guess, so please correct me if I'm wrong.
When the 3.19.0-16 kernel was installed, the process ended up with two vmlinuz files: vmlinuz-3-19.0-16-generic and vmlinuz-3-19.0-16-generic.efi.signed.
Now, after the latest update I can find in the /boot folder only the vmlinuz-3-19.0-18-generic file. This can mean only two things: first) efi stub support has been temporarily suspended in this kernel's version (but I don't think so) second) the latest kernel is definitely an efi image that can be booted directly. In this case, ok, I'll try copying this kernel's version to my hfs+ boot partition.
That said, I'm not sure if new macs do have the possibility to tweak the BIOS but my old model for sure doesn't have. In no way I can switch between secure/unsecure boot mode. My guess is that the EFI firmware is booting in UNSAFE mode by default (and actually this is what grub says very early at boot, BEFORE displaying the classical purple screen).
Anyway, I'll try with the latest kernel.

Revision history for this message
Raffaele Candeliere (candeliere) wrote :

Ok. Tried with latest kernel.
Nope.
The whole system boots fine, but X is still failing to detect properly the graphic card.

Revision history for this message
penalvch (penalvch) wrote :

Raffaele Candeliere:

"Ok. Tried with latest kernel."

Which one did you test precisely?

Revision history for this message
Raffaele Candeliere (candeliere) wrote :

3.19.0-18-generic

Revision history for this message
RaffaeleC (raffaele-candeliere) wrote :

Update. Some improvement with the latest update (but also a regression!).
With the last update, I'm able to boot the system using directly the efi stub and now, Xorg CAN detect the graphic card. There's no error in Xorg.0.log. But the problem now is that I get a blank/black screen all the same, even if booting through GRUB.
The system starts, the kernel splash appears and, after a few seconds the screen goes black.
If I hit alt-F1 nothing happens. Hit alt-F7, the screen flashes but gets black again. Now I re-hit alt-F1 and I see a text console saying:
"Welcome to emergency mode. Type sysctl reboot to reboot, sysctl default or ^D to try to start default mode"
If I now hit ^D the screen turns black again and the system freezes.

Revision history for this message
penalvch (penalvch) wrote :

Raffaele Candeliere, any change with the daily via http://cdimage.ubuntu.com/daily-live/current/ ?

Revision history for this message
RaffaeleC (raffaele-candeliere) wrote :

Sorry, false "positive"! I reinstalled all from scratch with the latest daily build and the bug is still there: no graphic card detected.
Actually, I'm not sure now about the "detection" i could see. May be i simply booted into Ubuntu via grub by accident and i didn't realize it.
So, we're at the same point. No efi-stub boot possible.

Revision history for this message
penalvch (penalvch) wrote :

Raffaele Candeliere, the issue you are reporting is an upstream one. Could you please report this problem via https://bugs.freedesktop.org/enter_bug.cgi?product=xorg&component=Driver/Radeon ?

Please provide a direct URL to your bug report once you have made it so that it may be tracked.

Thank you for your understanding.

tags: added: wily
summary: - X fails to automatically detect ATI radeon RV600 card
+ 1002:9583 [iMac7,1] X fails to automatically detect ATI radeon RV600
+ card
Changed in xorg (Ubuntu):
importance: Medium → High
status: Incomplete → Triaged
Revision history for this message
RaffaeleC (raffaele-candeliere) wrote :
Revision history for this message
Daniel van Vugt (vanvugt) wrote :

Thank you for reporting this bug to Ubuntu.
Ubuntu 15.10 (wily) reached end-of-life on July 28, 2016.
Ubuntu 15.04 (vivid) reached end-of-life on February 4, 2016.

See this document for currently supported Ubuntu releases:
https://wiki.ubuntu.com/Releases

We appreciate that this bug may be old and you might not be interested in discussing it any more. But if you are then please upgrade to the latest Ubuntu version and re-test. If you then find the bug is still present in the newer Ubuntu version, please add a comment here telling us which new version it is in and change the bug status to Confirmed.

affects: xorg (Ubuntu) → xorg-server (Ubuntu)
Changed in xorg-server (Ubuntu):
status: Triaged → Won't Fix
To post a comment you must log in.