AMD Catalyst 8.01 Linux Driver

[HighHo]

Just seen 8.01 for linux is out.
Download: https://a248.e.akamai.net/f/674/9206/0/www2.ati.com/drivers/linux/ati-driver-installer-8-01-x86.x86_64.run
Release Notes: https://a248.e.akamai.net/f/674/9206/0/www2.ati.com/drivers/linux/catalyst_81_linux.html

[lucky_]

On workstation hardware 3D applications will be corrupted if the screen width is not an integer multiple of 64 pixels, for example with a 1680x1050 wide screen display. Further details can be found in topic number 737-31720
Crap, this the reason why I couldn't get any 3d app.
And it isn't even fixed. (was already the case for 7.12)

[lucky_]

Crap, this the reason why I couldn't get any 3d app.
And it isn't even fixed. (was already the case for 7.12)
What is so hard/wrong with the firegl ? It seems that I didn't see a correct driver since several months now ? (read more than 6)

[yoshi314]

# Corruption will no longer be noticed in the lower right corner of the display or on the mouse pointer after the system is running for a long period of time
# Connecting a display device that supports 1680x1050 to a system running Linux will no longer result in a maximum display resolution of 1280x1024 only being available
# Custom mode lines in xorg.conf will no longer be ignored by the fglrx driver
# Suspending to RAM or DISK on kernel version 2.6.23 or later no longer fails
i was expecting a bit more :/

looks like this release caught phoronix pants-down :]

[cagljevic]

Once again I am let down... Maybe next month the window tearing and no XV support will get some attention. Good to see the 1680x1050 bug is resolved, that was rediculous.

[gmplague]

this sucks... i have dual 1440x900 monitors, (which is, i think, a far more common setup than 1680x1050), and was broken with the last release despite no mention in the bug reports. 1440/32 = 22.5, so I guess I'm out of luck. If anyone wants to buy a Radeon 2600 XT HD from me, so that i can buy nvidia, PM me.

[MagicTempest]

Well, I'm going to try this release. I've been without far too long now. I hope I'll be able to play movies now :-)

[Bigon]

3D accel doesn't even work here

(EE) fglrx(0): Failed to initialize CMMQS driver.

Ubuntu hardy (xorg 7.3)

[d2kx]

this sucks... i have dual 1440x900 monitors, (which is, i think, a far more common setup than 1680x1050), and was broken with the last release despite no mention in the bug reports. 1440/32 = 22.5, so I guess I'm out of luck. If anyone wants to buy a Radeon 2600 XT HD from me, so that i can buy nvidia, PM me.

Give 8-01 a try.

[lucky_]

I was ok to wait until the firegl were supported (i.e 3 months after the release of the new fglrx).
But now that it clains to support them and the firegl do not have xv nor opengl support?
WTF, is it what we call useless ?
Correct me if I'am wrong but I should care of this driver right ?
I am so disappointed, for sure the next laptop won't have any ati in it.
For the next desktop I wouldn't mind to buy ATI (radeonhd) since it's easier to swap if it is not satisfying.
Well good luck every one with this driver, as I am once again out of range.

[Kano]

"Custom mode lines in xorg.conf will no longer be ignored by the fglrx driver"

Extra funny, because as soon as I use one or more I get:

(II) fglrx(0): Ranges: V min: 50 V max: 160 Hz, H min: 30 H max: 95 kHz,
(II) fglrx(0): End of Display1 EDID data --------------------
(WW) fglrx(0): Only one display is connnected,so single mode is enabled
(II) fglrx(0): Primary Controller - CRT on secondary DAC
(II) fglrx(0): Internal Desktop Setting: 0x00000001
(II) fglrx(0): POWERplay version 3. 1 power state available:
(II) fglrx(0): 1. 398/250MHz @ 0Hz [enable load balancing]

Backtrace:
0: /usr/bin/X(xf86SigHandler+0x84) [0x80c43e4]
1: [0xb7f82420]
2: /usr/lib/xorg/modules/drivers/fglrx_drv.so [0xb7a275cb]
3: /usr/lib/xorg/modules/drivers/fglrx_drv.so(DALCWDDE+0x1a2) [0xb7a25d52]
4: /usr/lib/xorg/modules/drivers/fglrx_drv.so(swlDalHelperAddCustomizeMode+0x1ef) [0xb79807df]
5: /usr/lib/xorg/modules/drivers/fglrx_drv.so(atiddxPreInit+0xab9) [0xb79584e9]
6: /usr/bin/X(InitOutput+0x9e4) [0x80a0594]
7: /usr/bin/X(main+0x276) [0x806e4a6]
8: /lib/tls/libc.so.6(__libc_start_main+0xc8) [0xb7d85ea8]
9: /usr/bin/X(FontFileCompleteXLFD+0xa5) [0x806d9f1]

Fatal server error:
Caught signal 11. Server aborting

Xorg 7.1.1 from Debian Etch (or Kanotix)

[phoronix]

Phoronix: AMD Catalyst 8.01 Linux Driver

AMD has today released the 8.01 Linux Catalyst package, which contains the fglrx 8.45 driver. New in this release are updated packaging scripts and a few fixes, which many have been hopeful for in this release. Compared to our usual Linux driver articles, this one is a bit shorter due to the KDE 4.0 release event going on at the Googleplex, which we are in the process of covering.

http://www.phoronix.com/vr.php?view=11704

[chikazuku]

Damage report:

- Starting X with the kernel module, results in X11 failing to start because of a segfault. So I have no 3D, but I can live with that, because I didn't use that anyway. Just to bad for AIGLX.

- Starting KDE4.0 fails, I'm stuck with a black screen. Killing the X-server is possible however.

[Compxpert]

Anyone got a gentoo ebuild yet? As for maintence release this one is a lot better than 8.43 i think that was 7.11 idk it was the version right after 8.42 which had a few very useless fixes. Least this one has meaningful ones. lol

[lucky_]

I don't mind to wait for features or bug fixes, but I want to know what to expect.
I want to be able to decide now what to do.
ATI doesn't deserve a penny for leaving their users in such darkness.

Disgusting.

AMD have all my respect and I would like to thank M. Bridgman for his work.
But how can we come to this point ? Releasing for more than 6 months useless drivers for some type of card ?

Hmm, maybe I am bit too hard, there was christmas holydays, they surely didn't have time to work on anything.

[shamooki]

3D works for me with 1680x1050 on my Laptop (T60, Mobility X1300, OpenSUSE 10.3). That's good!

But on the downside, apparently 2D is FAR slower than with 8.40, redrawing windows after moving them takes considerably longer than with the old release. Since I use 3D only occasionally, I'll have to go back to the old driver.

I'll save myself (and you) the rant about ATI not being able to release proper drivers, but instead hope that this time it's because all developers are busy preparing the documentation dump which will eventually enable accelerated 2D and 3D in the open source driver :)

hope springs eternal.

--- Edit:
The 2D slowness problem has gone away after a few reboots. So no rants over ATI devs anymore :) And if they manage to support XRandr properly, I'll praise them in every possible way!

[benwick]

I haven't even installed it yet and I'm already disappointed. I guess that means I once again allowed my sunny optimism to get the better of me. At least they fixed the 1680x1050... right?

[lucky_]

I even can't hope this, since v3200 are yet out of documentation scope.

[benwick]

I haven't even installed it yet and I'm already disappointed. I guess that means I once again allowed my sunny optimism to get the better of me. At least they fixed the 1680x1050... right?

Actually (in response to my own skepticism) I just noticed that the known issues (as of 7.12 actually) finally mentions the blockiness and diagonal tearing in Xv video. So somebody noticed -- now I can retain hope, since they've been so focused on bugfixes. {I don't need new features, I need the old features to work.}

[Extreme Coder]

lol, same deal, same deal..
When I bought my mobo with a built-in ATI, I bought it for the open source drivers ;)

[lowlands]

Just downloaded it and tried to build it with:
$ sh ati-driver-installer-8-01-x86.x86_64.run --buildpkg Fedora/F8

All goes well except that the spec file is missing an installed package:

RPM build errors:
Installed (but unpackaged) file(s) found:
/usr/sbin/amdnotifyui

Anyone have an idea how to fix this?

Update:

Here's a little bit of progress
1) extract the files from ati-driver-installer-8-01-x86.x86_64.run to a directory
$ sh ./ati-driver-installer-8-01-x86.x86_64.run --extract tmp
$ cd tmp

2) go to the Fedora dir in the packages dir
$ cd packages/Fedora

3) open ATI-fglrx.spec-tmpl with an editor, go to the section that starts with %files (line 259) and add %{_sbindir}/amdnotifyui below %{_sbindir}/atieventsd

Now when I go to the root directory (tmp) and try ./ati-installer.sh --buildpkg Fedora/F8 it always returns some help information instead of building the packages.

Anyone know how to make ati-installer.sh build packages?

[givemesugarr]

Anyone got a gentoo ebuild yet? As for maintence release this one is a lot better than 8.43 i think that was 7.11 idk it was the version right after 8.42 which had a few very useless fixes. Least this one has meaningful ones. lol


well, if ati-drivers-8.45 will be added today to portage, then gentoo users would get a lot of new packages since kde4.0.0 also has hit portage official today. as for the ebuild, you can modify it manually from the 8.443 one. but i'm sure that it will hit portage soon (today or tomorrow).

[hotchili]

I am using a Thinkpad T60 with a X1400 under OpenSuSE 10.3 (Kernel 2.6.22-13).
My first attempt to start the X server with my previous xorg configuration was unsuccessful. X11 came up, the fglrx module was loaded but not used and no 3D. After an aticonfig --initial -f everything came up without a problem.

Good:
I am using a 1400x1050 screen. I works!
XV works with mplayer, xine and Skype-Beta with a webcam.
Suspend/Resume works! (even faster than with the older releases)

Strange:
When I start the ccc i get the following errors:
kernel: [fglrx:firegl_free_mutex] *ERROR* mutex id 0x0000000e not found in mutex list
kernel: [fglrx:firegl_free_mutex] *ERROR* mutex id 0x0000000f not found in mutex list

I think this error can be ignored as CCC comes up without a problem. Trying to free something is always OK :-)

Conclusion:
After one hour testing with different applications and some suspend/resume cycles, my ATI depression slowly turns into optimism...

[SledgeHammer_999]

In this system:
Ubuntu 7.10 amd64
x1800xl 256mb

It works just fine. AIGLX also works.

I just hope that they fix the "xV" related issues pretty soon because I hate that my videoplayback is not as good as in WinXP.

[benwick]

I just hope that they fix the "xV" related issues pretty soon because I hate that my videoplayback is not as good as in WinXP.

I agree, and if they fix that I will never have to upgrade the driver again :)

[bridgman]

But on the downside, apparently 2D is FAR slower than with 8.40, redrawing windows after moving them takes considerably longer than with the old release. Since I use 3D only occasionally, I'll have to go back to the old driver.

Are you using Compiz on the new drivers, or are you saying that 2d is slower on Cat 8.01 than on v8.40 with both drivers being used the same way ?

Are you using EXA or XAA acceleration ?

this sucks... i have dual 1440x900 monitors, (which is, i think, a far more common setup than 1680x1050), and was broken with the last release despite no mention in the bug reports. 1440/32 = 22.5, so I guess I'm out of luck. If anyone wants to buy a Radeon 2600 XT HD from me, so that i can buy nvidia, PM me.

Running at 1440x900 on an HD2600 should not be affected by the 64-pixel issue -- AFAIK that is only with the workstation code paths which only kick in on FireGL hardware. This is based on my reading of the release notes though. Have you actually encountered the problem or are you also going from the release notes ?

[Alexander Heß]

I am using a Thinkpad T60 with a X1400 [...] my ATI depression slowly turns into optimism...

Lucky for you. I have a T60p with a V5250 and 3D is corrupted...

So, no optimism for me. AGAIN!

[koolmanoncampus]

Just downloaded it and tried to build it with:
$ sh ati-driver-installer-8-01-x86.x86_64.run --buildpkg Fedora/F8

All goes well except that the spec file is missing an installed package:

RPM build errors:
Installed (but unpackaged) file(s) found:
/usr/sbin/amdnotifyui

Anyone have an idea how to fix this?

Update:

Here's a little bit of progress
1) extract the files from ati-driver-installer-8-01-x86.x86_64.run to a directory
$ sh ./ati-driver-installer-8-01-x86.x86_64.run --extract tmp
$ cd tmp

2) go to the Fedora dir in the packages dir
$ cd packages/Fedora

3) open ATI-fglrx.spec-tmpl with an editor, go to the section that starts with %files (line 259) and add %{_sbindir}/amdnotifyui below %{_sbindir}/atieventsd

Now when I go to the root directory (tmp) and try ./ati-installer.sh --buildpkg Fedora/F8 it always returns some help information instead of building the packages.

Anyone know how to make ati-installer.sh build packages?

Same problem here. It looks like Fedora support is pispoor in this release. Hopefully phorogit will be updated to make a working packaging script for Fedora 8.

Using your trick of adding that to the file list I got the rpms generated.

[LucasG15]

Welps! I recently built a new system with a new ati hd3870 pci-express video card. The 7.12 drivers seemed to work alright except that I couldn't enable 3d-effects which to me is not a big deal but dragging windows around the screen causes slowdowns. CPU usage would spike up and there would be a delay before the windows underneath the one I'm dragging around would get redrawn. If I drag a window around the screen fast enough than it would become corrupted until I stop. After a split second it would get correctly redrawn.
Beyond that, everything else *seems* to be working correctly.

By the way, my method of installation was to create the individual packages from ATI's installer and use 'aticonfig' to generate a new xorg.conf. I'm also running Ubuntu 7.10 gutsy 32bit.

[gmplague]

Are you using Compiz on the new drivers, or are you saying that 2d is slower on Cat 8.01 than on v8.40 with both drivers being used the same way ?

Are you using EXA or XAA acceleration ?



Running at 1440x900 on an HD2600 should not be affected by the 64-pixel issue -- AFAIK that is only with the workstation code paths which only kick in on FireGL hardware. This is based on my reading of the release notes though. Have you actually encountered the problem or are you also going from the release notes ?

Nope, going off release notes.. at work right now, will try when I get home. Thanks for the input.

[shamooki]

Hi brigdman,

thx for tuning in!

Are you using Compiz on the new drivers, or are you saying that 2d is slower on Cat 8.01 than on v8.40 with both drivers being used the same way ?


Yes.
I'm not using Compiz, but KDE/KWin 3.5.7. Specifically, I can observe the slowdown when moving a konsole window that has (pseudo-) transparency enabled (i.e., the screen background image under the konsole is copied with decreased luminance). When I hold my mouse cursor still for a short time, the konsole background is updated, and that's when the entire GUI freezes (with 100% CPU).

This was not happening with an older version of the driver. I think it was 8.40, but unfortunately I can't go back.
After doing rpm --erase for 8.452 and reinstalling 8.40 from the rpm all I get is a black screen :(.
So it seems I'm stuck with slow 2d :mad:

---Edit---
After rebooting twice, the behavior is back to normal - no slowdown anymore. Could this be related to the fact that I have an LCD hooked to my docking station's DVI port.


Are you using EXA or XAA acceleration ?


I don't know. There's no such option in my xorg.conf. How can I find out?

Cheers,

shamooki

[lowlands]

Using your trick of adding that to the file list I got the rpms generated.

That's good news :) Are you also using F8? Which command did you exactly use to build the packages? Did you use ati-installer.sh too?

[Uchikoma]

I've been running on the cutting edge drivers for the past few months now, and so far all is right with the world. (Xv, video playback) with the sticking point being standby/hibernation.

So, I'll give this a try later and see what happens.
On that note, some of the video issues that I've seen users complain about (namely lack of Xv support? and tearing and the like) is now mentioned under known issues.

I must say that even though some of you probably didn't get what you wanted, its nicer to see things being noticed more often now...(at least it seems like it to me). Yes, we know, it's been ages since it popped up, but hey...things seem to be picking up lately.

[Berniyh]

well, if ati-drivers-8.45 will be added today to portage, then gentoo users would get a lot of new packages since kde4.0.0 also has hit portage official today. as for the ebuild, you can modify it manually from the 8.443 one. but i'm sure that it will hit portage soon (today or tomorrow).
Compositing will still not work with it when using kwin. ;) (or :()

Ebuild changes are trivial, one only has to add amdpcsdb.default to the install routine.
Apart from that (and that the =virtual/libstdc++-3.3 dep seems to be unneeded now, which isn't a big deal) it is trivial. ;)

Still this release has exactly the same bugs as the previous one, except maybe the widescreen bug which didn't bother me.

[bridgman]

I'm not using Compiz, but KDE/KWin 3.5.7. Specifically, I can observe the slowdown when moving a konsole window that has (pseudo-) transparency enabled (i.e., the screen background image under the konsole is copied with decreased luminance). When I hold my mouse cursor still for a short time, the konsole background is updated, and that's when the entire GUI freezes (with 100% CPU).

This was not happening with an older version of the driver. I think it was 8.40, but unfortunately I can't go back.

OK. I'm still learning this, but it looks like KWin supports compositing however it is not enabled by default before 4.0. I'm not sure if it is enabled by default in 4.0. Do you happen to know if you have compositing enabled ? If so, that would explain why you are seeing slower 2d on the newer driver ('cause the older driver didn't support compositing through AIGLX).

I don't know. There's no such option in my xorg.conf. How can I find out?

I think your log should indicate which acceleration option is being used. Pretty sure XAA is the default.

[lowlands]

I've been running on the cutting edge drivers for the past few months now, and so far all is right with the world. (Xv, video playback) with the sticking point being standby/hibernation.

So, I'll give this a try later and see what happens.
On that note, some of the video issues that I've seen users complain about (namely lack of Xv support? and tearing and the like) is now mentioned under known issues.

I must say that even though some of you probably didn't get what you wanted, its nicer to see things being noticed more often now...(at least it seems like it to me). Yes, we know, it's been ages since it popped up, but hey...things seem to be picking up lately.

I agree Uchikoma. It's nice to see the known issues list get "refreshed" and that all the issues in the previous release (1680x1050, corruption and suspend) should be fixed. It's just too bad it took another month to get the fixes out. I would love to see (bi-)weekly beta or closed user group releases so these issues can be picked sooner rather than later. Hopefully I'll be able to fix the package build issue so I can actually try the new goodness.

[chrisr]

Lucky for you. I have a T60p with a V5250 and 3D is corrupted...

So, no optimism for me. AGAIN!

But I am wondering: the graphics PCI ID is 0x71D4, correct? What happens if you "hack" the PCI ID to 0x71D5 instead, as was common practice before FireGL support became available anyway? Does 3D work then?

[shamooki]

Thanks for the reply! I just updated my post above; after rebooting twice 2D speed is back to normal. Which is very good! :D

The best problems are those which go away by themselves :)

Do you happen to know if you have compositing enabled ?

No, compositing is not enabled.



I think your log should indicate which acceleration option is being used. Pretty sure XAA is the default.
My Xorg.0.log indicates XAA.

Cheers,

Michael

[NeoBrain]

Well, just installed 8.45 and it can't init AGP. just as 8.44.
Seems like I'll have to wait another month :D

[janl]

I know there was a hotfix which was released for the Windows driver which resolved the AGP issue. Maybe there will be something similar for Linux.

[bridgman]

Another user (Bigon) added their log to Bugzilla #983. Looks like the error from that log is "(EE) fglrx(0): Failed to initialize CMMQS driver."

NeoBrain, you had the "Preinit DAL" error, didn't you ? Is that still what you are seeing, or are you getting the CMMQS message as well ? Which GPU model are you seeing the problem on ?

[chrisbudden]

I still get poor performance when using AIGLX, including the diagonal tearing when scrolling in firefox.

Anyone know of a fix or is it just down to ATI to get it sorted?

[NeoBrain]

No, it was just coming on the first startx after driver installation, on the second try it worked.

My problem was this one:

(**) fglrx(0): ATI GART size: 256 MB
(II) fglrx(0): [pci] find AGP GART
(II) fglrx(0): [agp] Mode=0x1f004e1b bridge: 0x1039/0x0746
(II) fglrx(0): [agp] AGP v1/2 disable mask 0x00000000
(II) fglrx(0): [agp] AGP v3 disable mask 0x00000000
(II) fglrx(0): [agp] enabling AGP with mode=0x1f004f1a
(EE) fglrx(0): [agp] Failed to remmap MC AGP aperture base!
(EE) fglrx(0): cannot init AGP

Which disables the 3d accel.
I have an X1600 (AGP) with 512 MB vram, my system has 512 MB ram, too.
Other users also had the same problem and it helped them to set the AGP aperture size in the BIOS to another vram size.
However, I've tried all possible numbers for that from 32 MB to 256 MB (my BIOS's maximum), but none worked.
Had worked with 8.43 btw, further information is here...:
http://www.phoronix.com/forums/showthread.php?t=7044

I've allready tried many things with givemesuggar, but I came to the decision that I'll better wait until AMD fixes this :rolleyes:

[Compxpert]

Hmm getting poorer 3d performance and what seems like a "frame lag" where it runs fine for a few seconds then freezes for a few and then runs fine for a few more and then freezes and so on. Does this on fgl_glxgears and glxgears. I ran a game also and it would be moving fine then it would appear to skip a bunch of frames and then keep going repeating this the whole time. Anyone else have this issue. I was also able to reproduce this at other resolutions so its not the 64 pixel bug. I just checked dmesg and its flooded with errors related to fglrx. I think I'm going to go back to 7-12 its not like the modeline bug really effected me anyway.

[pjjanak]

So, I'm using Ubuntu 7.10 and suspend obviously wasn't working until Catalyst 7.12. However, widescreen stopped working along with that. So I figured I would just wait until 8.1 for widescreen and suspend (and 3D acceleration, which is really the only plus I care about that ATI proprietary drivers have over the others). However, suspend broke. Goddamn it. It turns off the screen, the harddrive spins up, and then nothing. Dead. I have to do a hard reset. :mad:

Peter

[Uchikoma]

So, I'm using Ubuntu 7.10 and suspend obviously wasn't working until Catalyst 7.12. However, widescreen stopped working along with that. So I figured I would just wait until 8.1 for widescreen and suspend (and 3D acceleration, which is really the only plus I care about that ATI proprietary drivers have over the others). However, suspend broke. Goddamn it. It turns off the screen, the harddrive spins up, and then nothing. Dead. I have to do a hard reset. :mad:

Peter

This after you installed 8.1?

[Nicolas]

Anyone here has problems restarting X ?
If I press Ctrl-Alt-Backspace my PC freezes (only with the background image).

Damn you ATI! Get your act together!!

[Michael]

For the Fedora packaging bug, I'll push updated scripts to Phorogit for it tonight once I get back to my hotel.

[Bigon]

Anyone here has problems restarting X ?
If I press Ctrl-Alt-Backspace my PC freezes (only with the background image).

Damn you ATI! Get your act together!!

Got this too

[Nicolas]

I still get poor performance when using AIGLX, including the diagonal tearing when scrolling in firefox.

Anyone know of a fix or is it just down to ATI to get it sorted?

Scrolling in Firefox with AIGLX enabled is impossible. It's REALLY SLOW depending on the content (eg: flash animation).

So, I've updated my driver to get system crash when restarting X and no improvements at all. Nice. :mad:

[pjjanak]

This after you installed 8.1?

Yessir.

Peter

[sid350]

another crappy release?
they can not call this new version - it is just patch, which fixes the modelines support. I can't see any new fixes.
x1600xt is rotting now under my windows.
and nobody wants to buy my laptop, because of builtin ati videocard! :-E

[Isaac702]

Well I'm dead in the water. I have updated to 8.1 but it crashes when ever I try to configure Xorg through Yast (My old settings don't work either Xorg crashes)

Im not entirely sure what section of the log I should post but others have posted the Backtrace section so here it is.

Backtrace:
0: /usr/sbin/xw(xf86SigHandler+0x6d) [0x492acd]
1: /lib64/libc.so.6 [0x2b337a982bd0]
2: /lib64/libc.so.6 [0x2b337a98663a]
3: /usr/sbin/xw(xf86atoi+0x12) [0x472ea2]
4: /usr/lib64/xorg/modules//drivers/fglrx_drv.so [0x2b337bef69fa]
5: /usr/lib64/xorg/modules//drivers/fglrx_drv.so(atiddxPreInit+0x68d) [0x2b337bef1d8d]
6: /usr/sbin/xw(InitOutput+0x9da) [0x4694ea]
7: /usr/sbin/xw(main+0x295) [0x43a0c5]
8: /lib64/libc.so.6(__libc_start_main+0xf4) [0x2b337a96fb54]
9: /usr/sbin/xw(FontFileCompleteXLFD+0x229) [0x4395e9]

Fatal server error:
Caught signal 11. Server aborting


OS: Opensuse 10.3 64bit
IGP: ATI Radeon X1200 Series - RS690M

I guess im back to testing radeonhd until the next release.

[phlegm]

Things are working great here on my T60. Suspend is working again and compiz is fast. The only thing left for me is the xv playback. Oh and Google earth is all flickery.

Slowly but surely things are getting better

[Alexander Heß]

But I am wondering: the graphics PCI ID is 0x71D4, correct? What happens if you "hack" the PCI ID to 0x71D5 instead, as was common practice before FireGL support became available anyway? Does 3D work then?
Yep, it’s 0x71D4. Where am I supposed to “hack” the ID?

[sok-1]

For me a good driver release:
- Widescreen is working again,
- The memory leak is no more, (I can't use 7.12 on a widescreen)
- distortion on the lower right is no more,
- suspend works

Unfortuantly, compiz feels faster, but not fast enough.

[SledgeHammer_999]

So, I'm using Ubuntu 7.10 and suspend obviously wasn't working until Catalyst 7.12. However, widescreen stopped working along with that. So I figured I would just wait until 8.1 for widescreen and suspend (and 3D acceleration, which is really the only plus I care about that ATI proprietary drivers have over the others). However, suspend broke. Goddamn it. It turns off the screen, the harddrive spins up, and then nothing. Dead. I have to do a hard reset. :mad:

Peter

Well the Release notes state:
Suspending to RAM or DISK on kernel version 2.6.23 or later no longer fails

Well on my up-to-date gutsy "uname -r" gives:
2.6.22-14-generic

So there's your problem!!!

[jackkerouac]

Well, after installing this driver, I cannot even get into the GDM login screen. I just get a blank, black screen with nothing else. ALT-CTRL-F1 or F2 doesn't work. I have to hard-reset.

ATI X1650 Pro, Gutsy 7.10

Here is my Xorg.0.log:


X Window System Version 1.3.0
Release Date: 19 April 2007
X Protocol Version 11, Revision 0, Release 1.3
Build Operating System: Linux Ubuntu (xorg-server 2:1.3.0.0.dfsg-12ubuntu8.1)
Current Operating System: Linux patrick-desktop 2.6.22-14-generic #1 SMP Tue Dec 18 08:02:57 UTC 2007 i686

... blah, blah, blah ...

(II) fglrx(0): detected X.org 7.1.0.0
(II) Loading extension ATIFGLRXDRI
(II) fglrx(0): doing DRIScreenInit
drmOpenDevice: node name is /dev/dri/card0
drmOpenDevice: open result is 7, (OK)
drmOpenDevice: node name is /dev/dri/card0
drmOpenDevice: open result is 7, (OK)
drmOpenByBusid: Searching for BusID PCI:1:0:0
drmOpenDevice: node name is /dev/dri/card0
drmOpenDevice: open result is 7, (OK)
drmOpenByBusid: drmOpenMinor returns 7
drmOpenByBusid: drmGetBusid reports
drmOpenDevice: node name is /dev/dri/card1
drmOpenDevice: open result is -1, (No such device)
drmOpenDevice: open result is -1, (No such device)
drmOpenDevice: Open failed
drmOpenByBusid: drmOpenMinor returns -19
drmOpenDevice: node name is /dev/dri/card2
drmOpenDevice: open result is -1, (No such device)
drmOpenDevice: open result is -1, (No such device)
drmOpenDevice: Open failed
drmOpenByBusid: drmOpenMinor returns -19
drmOpenDevice: node name is /dev/dri/card3
drmOpenDevice: open result is -1, (No such device)
drmOpenDevice: open result is -1, (No such device)
drmOpenDevice: Open failed
drmOpenByBusid: drmOpenMinor returns -19
drmOpenDevice: node name is /dev/dri/card4
drmOpenDevice: open result is -1, (No such device)
drmOpenDevice: open result is -1, (No such device)
drmOpenDevice: Open failed
drmOpenByBusid: drmOpenMinor returns -19
drmOpenDevice: node name is /dev/dri/card5
drmOpenDevice: open result is -1, (No such device)
drmOpenDevice: open result is -1, (No such device)
drmOpenDevice: Open failed
drmOpenByBusid: drmOpenMinor returns -19
drmOpenDevice: node name is /dev/dri/card6
drmOpenDevice: open result is -1, (No such device)
drmOpenDevice: open result is -1, (No such device)
drmOpenDevice: Open failed
drmOpenByBusid: drmOpenMinor returns -19
drmOpenDevice: node name is /dev/dri/card7
drmOpenDevice: open result is -1, (No such device)
drmOpenDevice: open result is -1, (No such device)
drmOpenDevice: Open failed
drmOpenByBusid: drmOpenMinor returns -19
drmOpenDevice: node name is /dev/dri/card8
drmOpenDevice: open result is -1, (No such device)
drmOpenDevice: open result is -1, (No such device)
drmOpenDevice: Open failed
drmOpenByBusid: drmOpenMinor returns -19
drmOpenDevice: node name is /dev/dri/card9
drmOpenDevice: open result is -1, (No such device)
drmOpenDevice: open result is -1, (No such device)
drmOpenDevice: Open failed
drmOpenByBusid: drmOpenMinor returns -19
drmOpenDevice: node name is /dev/dri/card10
drmOpenDevice: open result is -1, (No such device)
drmOpenDevice: open result is -1, (No such device)
drmOpenDevice: Open failed
drmOpenByBusid: drmOpenMinor returns -19
drmOpenDevice: node name is /dev/dri/card11
drmOpenDevice: open result is -1, (No such device)
drmOpenDevice: open result is -1, (No such device)
drmOpenDevice: Open failed
drmOpenByBusid: drmOpenMinor returns -19
drmOpenDevice: node name is /dev/dri/card12
drmOpenDevice: open result is -1, (No such device)
drmOpenDevice: open result is -1, (No such device)
drmOpenDevice: Open failed
drmOpenByBusid: drmOpenMinor returns -19
drmOpenDevice: node name is /dev/dri/card13
drmOpenDevice: open result is -1, (No such device)
drmOpenDevice: open result is -1, (No such device)
drmOpenDevice: Open failed
drmOpenByBusid: drmOpenMinor returns -19
drmOpenDevice: node name is /dev/dri/card14
drmOpenDevice: open result is -1, (No such device)
drmOpenDevice: open result is -1, (No such device)
drmOpenDevice: Open failed
drmOpenByBusid: drmOpenMinor returns -19
drmOpenDevice: node name is /dev/dri/card0
drmOpenDevice: open result is 7, (OK)
drmOpenDevice: node name is /dev/dri/card0
drmOpenDevice: open result is 7, (OK)
drmGetBusid returned ''
(II) fglrx(0): [drm] DRM interface version 1.0
(II) fglrx(0): [drm] created "fglrx" driver at busid "PCI:1:0:0"
(II) fglrx(0): [drm] added 8192 byte SAREA at 0x2000
(II) fglrx(0): [drm] mapped SAREA 0x2000 to 0xb7c2e000
(II) fglrx(0): [drm] framebuffer handle = 0x3000
(II) fglrx(0): [drm] added 1 reserved context for kernel
(II) fglrx(0): DRIScreenInit done
(II) fglrx(0): Kernel Module Version Information:
(II) fglrx(0): Name: fglrx
(II) fglrx(0): Version: 8.37.6
(II) fglrx(0): Date: May 25 2007
(II) fglrx(0): Desc: ATI FireGL DRM kernel module
(II) fglrx(0): Kernel Module version matches driver.
(II) fglrx(0): Kernel Module Build Time Information:
(II) fglrx(0): Build-Kernel UTS_RELEASE: 2.6.22-14-generic
(II) fglrx(0): Build-Kernel MODVERSIONS: no
(II) fglrx(0): Build-Kernel __SMP__: no
(II) fglrx(0): Build-Kernel PAGE_SIZE: 0x1000
(II) fglrx(0): [drm] register handle = 0x00004000
(II) fglrx(0): Interrupt handler installed at IRQ 22.
(II) fglrx(0): Exposed events to the /proc interface
(II) fglrx(0): [drm] ringbuffer size = 0x00100000 bytes
(II) fglrx(0): [drm] texture shared area handle = 0x00008000
(II) fglrx(0): shared FSAAScale=1
(II) fglrx(0): DRI initialization successfull!
(II) fglrx(0): FBADPhys: 0xc0000000 FBMappedSize: 0x00701000
(II) fglrx(0): FBMM initialized for area (0,0)-(1280,1434)
(II) fglrx(0): FBMM auto alloc for area (0,0)-(1280,1024) (front color buffer - assumption)
(II) fglrx(0): Largest offscreen area available: 1280 x 410
(==) fglrx(0): Backing store disabled
(==) fglrx(0): Silken mouse enabled
(II) Loading extension FGLRXEXTENSION
(II) Loading extension ATITVOUT
(**) fglrx(0): DPMS enabled
(II) fglrx(0): GLESX enableFlags = 16
(II) fglrx(0): GLESX is enabled
(II) fglrx(0): Using XFree86 Acceleration Architecture (XAA)
Screen to screen bit blits
Solid filled rectangles
8x8 mono pattern filled rectangles
Solid Lines
Dashed Lines
Offscreen Pixmaps
Setting up tile and stipple cache:
30 128x128 slots
(II) fglrx(0): Acceleration enabled
(WW) fglrx(0): Option "VendorName" is not used
(WW) fglrx(0): Option "ModelName" is not used
(II) fglrx(0): X context handle = 0x1
(II) fglrx(0): [DRI] installation complete
(II) fglrx(0): Direct rendering enabled
(==) fglrx(0): Using hardware cursor
(==) RandR enabled
(II) Initializing built-in extension MIT-SHM
(II) Initializing built-in extension XInputExtension
(II) Initializing built-in extension XTEST
(II) Initializing built-in extension XKEYBOARD
(II) Initializing built-in extension XC-APPGROUP
(II) Initializing built-in extension XAccessControlExtension
(II) Initializing built-in extension SECURITY
(II) Initializing built-in extension XINERAMA
(II) Initializing built-in extension XFIXES
(II) Initializing built-in extension XFree86-Bigfont
(II) Initializing built-in extension RENDER
(II) Initializing built-in extension RANDR
(II) Initializing built-in extension COMPOSITE
(II) Initializing built-in extension DAMAGE
(II) Initializing built-in extension XEVIE
drmOpenDevice: node name is /dev/dri/card0
drmOpenDevice: open result is 8, (OK)
drmOpenByBusid: Searching for BusID PCI:1:0:0
drmOpenDevice: node name is /dev/dri/card0
drmOpenDevice: open result is 8, (OK)
drmOpenByBusid: drmOpenMinor returns 8
drmOpenByBusid: drmGetBusid reports PCI:1:0:0
(EE) AIGLX error: dlsym for __driCreateNewScreen_20050727 failed (/usr/lib/dri/fglrx_dri.so: undefined symbol: __driCreateNewScreen_20050727)
(EE) AIGLX: reverting to software rendering
(II) Loading local sub module "GLcore"
(II) LoadModule: "GLcore"
(II) Loading /usr/lib/xorg/modules/extensions//libGLcore.so
(II) Module GLcore: vendor="X.Org Foundation"
compiled for 1.3.0, module version = 1.0.0
ABI class: X.Org Server Extension, version 0.3
(II) GLX: Initialized MESA-PROXY GL provider for screen 0
(**) Option "CoreKeyboard"
(**) Generic Keyboard: Core Keyboard
(**) Option "Protocol" "standard"
(**) Generic Keyboard: Protocol: standard
(**) Option "AutoRepeat" "500 30"
(**) Option "XkbRules" "xorg"
(**) Generic Keyboard: XkbRules: "xorg"
(**) Option "XkbModel" "pc105"
(**) Generic Keyboard: XkbModel: "pc105"
(**) Option "XkbLayout" "us"
(**) Generic Keyboard: XkbLayout: "us"
(**) Option "CustomKeycodes" "off"
(**) Generic Keyboard: CustomKeycodes disabled
(**) Option "Protocol" "ImPS/2"
(**) Configured Mouse: Device: "/dev/input/mice"
(**) Configured Mouse: Protocol: "ImPS/2"
(**) Option "CorePointer"
(**) Configured Mouse: Core Pointer
(**) Option "Device" "/dev/input/mice"
(**) Option "Emulate3Buttons" "true"
(**) Configured Mouse: Emulate3Buttons, Emulate3Timeout: 50
(**) Option "ZAxisMapping" "4 5"
(**) Configured Mouse: ZAxisMapping: buttons 4 and 5
(**) Configured Mouse: Buttons: 9
(**) Configured Mouse: Sensitivity: 1
(II) XINPUT: Adding extended input device "Configured Mouse" (type: MOUSE)
(II) XINPUT: Adding extended input device "Generic Keyboard" (type: KEYBOARD)
(II) Configured Mouse: ps2EnableDataReporting: succeeded
SetClientVersion: 0 9
(II) fglrx(0): [drm] removed 1 reserved context for kernel
(II) fglrx(0): [drm] unmapping 8192 bytes of SAREA 0x2000 at 0xb7c2e000

Any ideas on why this is happening? 8.40.4 works just fine. It's what I am using now in place of any of the new drivers. Any driver after 8.40.4 gives me the same problems.

Any help would be really appreciated.

[jackkerouac]

And here is my xorg.conf:


# xorg.conf (xorg X Window System server configuration file)
#
# This file was generated by dexconf, the Debian X Configuration tool, using
# values from the debconf database.
#
# Edit this file with caution, and see the xorg.conf manual page.
# (Type "man xorg.conf" at the shell prompt.)
#
# This file is automatically updated on xserver-xorg package upgrades *only*
# if it has not been modified since the last upgrade of the xserver-xorg
# package.
#
# If you have edited this file but would like it to be automatically updated
# again, run the following command:
# sudo dpkg-reconfigure -phigh xserver-xorg

Section "ServerLayout"

# Uncomment if you have a wacom tablet
# InputDevice "stylus" "SendCoreEvents"
# InputDevice "cursor" "SendCoreEvents"
# InputDevice "eraser" "SendCoreEvents"
Identifier "Default Layout"
screen 0 "aticonfig-Screen[0]" 0 0
Inputdevice "Generic Keyboard"
Inputdevice "Configured Mouse"
EndSection

Section "Files"
EndSection

Section "Module"
Load "glx"
EndSection

Section "InputDevice"
Identifier "Generic Keyboard"
Driver "kbd"
Option "CoreKeyboard"
Option "XkbRules" "xorg"
Option "XkbModel" "pc105"
Option "XkbLayout" "us"
EndSection

Section "InputDevice"
Identifier "Configured Mouse"
Driver "mouse"
Option "CorePointer"
Option "Device" "/dev/input/mice"
Option "Protocol" "ImPS/2"
Option "ZAxisMapping" "4 5"
Option "Emulate3Buttons" "true"
EndSection

Section "InputDevice"
Identifier "stylus"
Driver "wacom"
Option "Device" "/dev/input/wacom"
Option "Type" "stylus"
Option "ForceDevice" "ISDV4"# Tablet PC ONLY
EndSection

Section "InputDevice"
Identifier "eraser"
Driver "wacom"
Option "Device" "/dev/input/wacom"
Option "Type" "eraser"
Option "ForceDevice" "ISDV4"# Tablet PC ONLY
EndSection

Section "InputDevice"
Identifier "cursor"
Driver "wacom"
Option "Device" "/dev/input/wacom"
Option "Type" "cursor"
Option "ForceDevice" "ISDV4"# Tablet PC ONLY
EndSection

Section "Monitor"
Identifier "SyncMaster"
Option "DPMS"
EndSection

Section "Monitor"
Identifier "aticonfig-Monitor[0]"
Option "VendorName" "ATI Proprietary Driver"
Option "ModelName" "Generic Autodetecting Monitor"
Option "DPMS" "true"
EndSection

Section "Device"
Identifier "aticonfig-Device[0]"
Driver "fglrx"
Option "VideoOverlay" "on"
Option "OpenGLOverlay" "off"
EndSection

Section "Device"
Identifier "Generic Video Card"
Driver "fglrx"
Option "VideoOverlay" "on"
Option "OpenGLOverlay" "off"
Busid "PCI:1:0:0"
EndSection

Section "Screen"
Identifier "Default Screen"
Device "Generic Video Card"
Monitor "SyncMaster"
Defaultdepth 24
SubSection "Display"
Modes "1280x1024" "1152x864" "1024x768" "832x624" "800x600" "720x400" "640x480"
EndSubSection
EndSection

Section "Screen"
Identifier "aticonfig-Screen[0]"
Device "aticonfig-Device[0]"
Monitor "aticonfig-Monitor[0]"
Defaultdepth 24
SubSection "Display"
Viewport 0 0
Depth 24
EndSubSection
EndSection

Section "Extensions"
Option "Composite" "Disable"
EndSection

Section "ServerFlags"
Option "AIGLX" "off"
EndSection

[enzobelmont]

i don't know if laugh or cry...

another _crappy_ release from _ati_

[chrisg]

8.01 first impressions: (Thinkpad T60, X1400, Kubuntu Gutsy)

- installation was straightforward:

ati-driver-installer-8-01-x86.x86_64.run --buildpkg Ubuntu/gutsy
dpkg -i xorg-driver-fglrx_8.452.1-1_i386.deb fglrx-kernel-source_8.452.1-1_i386.deb fglrx-amdcccle_8.452.1-1_i386.deb
restart X

(I had all the other stuff like compiler tools and restricted modules settings already in place from previous installs)

- everything that worked in 7.12 still works, including suspend/resume
- the screen corruption when dragging icons around on the desktop is gone
- the resolution issues are fixed
- I didn't run benchmarks but my main 3D app googleearth runs about the same as before, certainly not worse
- video playing is essentially the same as before: kind of ok with GL, some flickering with xv. With compositing enabled, video playing is not usable with either GL or xv: too much flicker
- two out of the three issues that remained for me in 7.12 have been resolved: good going ATI!

Overall, for me this is the best fglrx I've used to date. Now make video and compositing work together and I'll be hard pressed to find something else to ask for.

Ah and before I forget, here's my ceterum censeo: ATI please write up a document on what options this driver takes in xorg.conf and what they mean. I know there are some documents out there on the web people put together but it would be really nice to have a complete listing of options from the authoritative source.

[shamooki]

Well I'm dead in the water. I have updated to 8.1 but it crashes when ever I try to configure Xorg through Yast (My old settings don't work either Xorg crashes)
[...]


Try the three R! ;)

I had a similar problem, and it went away after uninstalling the 8.45 rpm, installing the 8.40 rpm, uninstalling that and installing 8.45 again, with reboots inserted in this procedure generously. Feels like windows... :rolleyes:

Having another terminal to ssh into the machine under surgery helps to get around all those black screens on the way. If that's not possible, be sure to boot into runlevel 3 by entering a 3 into the boot options at the prompt.

But in the end it worked for me, and I have a widescreen with 1680x1050 and fast 3D running.

Cheers, shamooki

Lenovo T60
Mobility Radeon X1300
OpenSUSE 10.3 x84_64

[SledgeHammer_999]

Well, after installing this driver, I cannot even get into the GDM login screen. I just get a blank, black screen with nothing else. ALT-CTRL-F1 or F2 doesn't work. I have to hard-reset.

ATI X1650 Pro, Gutsy 7.10

Here is my Xorg.0.log:


Any ideas on why this is happening? 8.40.4 works just fine. It's what I am using now in place of any of the new drivers. Any driver after 8.40.4 gives me the same problems.

Any help would be really appreciated.

From what I see it tries to load an old fglrx(8.37.6). Have you blacklisted the ubuntu's fglrx? info (http://wiki.cchtml.com/index.php/Ubuntu_Gutsy_Installation_Guide)


ati-driver-installer-8.443.1-x86.x86_64.run --buildpkg Ubuntu/gutsy
dpkg -i xorg-driver-fglrx_8.452.1-1_i386.deb fglrx-kernel-source_8.452.1-1_i386.deb fglrx-amdcccle_8.452.1-1_i386.deb
restart X

I think you mean ati-driver-installer-8-01-x86.x86_64.run --buildpkg Ubuntu/gutsy EDIT: You fixed it already!!!

[enzobelmont]

any x200 experiences???

[Bigon]

Any chances to have an hotfix for not working cards? It's the 3rd unusable version in a row for me (memory leak, resolution issue and now not even 3d acceleration)?

[lowlands]

For the Fedora packaging bug, I'll push updated scripts to Phorogit for it tonight once I get back to my hotel.

Thanks Michael, I would appreciate that. Do you know why this does not work:

1) unpack driver into tempdir
2) cd tempdir (and fix missing file in the spec file template)
3) ./ati-installer.sh --buildpkg Fedora/F8

Step 3 always comes back with this error: Unrecognized parameter '' to ati-installer.sh

Am I doing something wrong or is it a bug? Thanks.

[Kano]

You forget to use the "real" version as first option to ati-installer.sh. That's NOT 8-01 but 8.452.1. Example from my script:

bash ati-installer.sh $VER --buildpkg $DISTRIB_ID/$DISTRIB_CODENAME

[sastraxi]

Summary for my Inspiron 6400, Radeon Mobility X1400, Ubuntu/gutsy:
- 1440x900 supported once more.
- suspend not working again (7.12 was fine).
- unable to end an X session; system crashes.
- VLC will not play videos with the OpenGL output plug-in.
- most things noticeably slower.

Too bad, really.

[Kano]

As there are very big compatibity issues especially with lastest Intel G33/P35 chipset (board used: Gigabyte GA-G33-DS3R) I would highly advice ATI to use this for testing, especially RV410 and those cases:

a) switch from ati oss to fglrx
b) switch from fglrx to another version
c) ctrl+alt+backspace - will often result in a state that kdm is still running and new x is therefore not started
d) oss driver with latest xorg + compiz fusion: enable water + enable snow effect + rotate cube
e) add a modeline

Many of those cases force crash the xserver or even the hole system (no ssh login possible anymore).

[slickhenry]

I haven't seen anything mentioned regarding the issue with aiglx under xserver 1.4.

Can anyone enlighten me on wether this is fixed? If not, I'll put upgrading on hold for a while.

[lowlands]

You forget to use the "real" version as first option to ati-installer.sh. That's NOT 8-01 but 8.452.1. Example from my script:

bash ati-installer.sh $VER --buildpkg $DISTRIB_ID/$DISTRIB_CODENAME

Thanks a bunch Kano! That did the trick. Packages have been successfully generated. W00t!

[lucky_]

When does the fglrx comedy will end, when AMD/ATI will have all the spec released ? So anyone could just play with working open-source driver ?
It represents such a waste of resource,

I am pretty sure that the driver's devs at AMD would be very pleased to work for the open-source one instead of continuing to build frustrating / unused (since it won't never be installed by default).
I wouldn't like to be one of them, being flamed at every released isn't motivating. I hope for them that one day they could dev on the open-source side, they would be warmly welcomed, right ?
It is such a pain to see how the industry has build limitations on natural progress of things.
I know the "we don't do what we want" blabla, I only talk about how peoples feel it. And how they can feel to be useful in their job.

What would it be if AMD stopped the release of fglrx for some time, to push the open-source project to the top ?
I guess there are far more hope from every users in the open source driver, so if anyone working on fglrx would work on radeon and radeonhd, who would complain ?
After that they could reuse most of the code to build the closed pieces on top of the open source part.
I know that there it is risky to push the driver if the spec are not out, but I am sure that M. Bridgman wouldn't refuse some help to accomplish his goals.

I know that my post isn't smart nor well thought maybe full of contradiction and not well written (french speaking inside), I only speak about how I feel this rahhhhh it's so egocentric, but who's not frustrated ?

Anyway piece everyone, I feel compassion for devs who certainly work hard but can't express themselves to their full potential.
Hope for better days without fglrx, good night everyone.

[djdoo]

@bridgman:

My friend on a previous post here you ask a fellow which Xorg acceleration type he uses XAA or EXA.

Can we use EXA acceleration with fglrx?? And if yes how? Cause my xorg.0.log always reports about the classic XAA accel...
Should I change something at my xorg.conf?

Also I can confirm Suspend2RAM now works here for my 2.6.24-rc7 kernel on openSuSE 10.3 64-bit system.

Compared to 7.12 AIGLX behaviour is the same, 3D-2D performance the same.

The previous driver enabled XV and Textured features for 32-bit systems for RS690 chipset-X1250 AVIVO IGP! I can confirm that too...

But still no Xv and Textured stuff here for my 64-bit opensuse.. Any ideas when will this be enabled for us?? Cause I was looking forward to 8.1 expecting AVIVO Xv support for 64 bit too... But no luck again :(

Jim

[DarkFoss]

For me the drivers work fine..If anything I see an improvement in fps running UT2K4 over the 7.12's..with them the fps would dip into the yellow and flicker red on occasion.. Solid green with the 8.01.4's
averaged out a nice 92 low 45 high 178..Scrolling in FF is fine, compiz-fusion works when I enable it..

Only downside is using AMD's default installer..Mandriva has yet to fix their --buildpkg script.

[Isaac702]

Try the three R! ;)

I had a similar problem, and it went away after uninstalling the 8.45 rpm, installing the 8.40 rpm, uninstalling that and installing 8.45 again, with reboots inserted in this procedure generously. Feels like windows... :rolleyes:

Thanks for you help. I managed to get it to work in the end.

Although I still have decided to go back to radeonhd as Fglrx 8.1 still renders well anything extremely slowly since the launch of the new code base. ATI fix the regression.

[Kano]

@lucky

As long as you do not have to maintain a distro you whould not wait for ATI. Just buy NVIDIA, much better choice. I would not even consider AMD motherboards anymore - as long as you don't use ATI gfx cards Intel boards are much more stable... I know that does not help to sell AMD/ATI things but the drivers are so bad that if you need more than 2d accelleration or you own such an old card that it works with the standard ati driver you will not be happy. Maybe you can work around some issues but as soon as you visit a friend with other hardware then you will see the difference. ATI cards do not even render everything, they only tested with some games, not more. Thats why tools like gl2benchmark show how incomplete the new codebase is - test 3/4 never worked after 8.40.4 and on 64 bit the tool even segfaults only. I see no way to recommend ATI currently - just using a few gaming benchmarks does not provide the full picture of the driver status.

[legume]

I just hope that they fix the "xV" related issues pretty soon because I hate that my videoplayback is not as good as in WinXP.

FWIW if you haven't tried already, what I do to work around the Xv corruption is use mplayers gl driver.

On my x1600 pro AGP 512meg card

mplayer -vo gl:yuv=4:lscale=1

gives h/w accelerated yuv->RGB and a not too bad looking bicubic scale.

You need working GL of course and I haven't tried with AIGLX.
Older cards may need different options - they are documented in the mplayer man page.

Andy.

PS/OT if any catalyst people see this - can you tell the XP people that directx (dxdiag 3d tests) has been broken on this card since August !! (7-7 last working).

[Melcar]

A "good" driver. Sure, many of the previous faults are still there, but at least they fixed a few things and no apparent regressions have occurred. When they mange to get Xv and OpenGL working with AIGLX and Compiz, then I will really be happy. My R6xx cards have problems generating pixmaps in Compiz, but I hope they fix it in the next few releases.
For me personally, what the devs should really work on is on getting frequency scaling working properly. In some of my cards, the driver will only read the "2D clocks", and since frequency scaling is not yet supported on R6xx chips... well, you get the idea.

[duncan_nz]

I really like this driver release, my max resolution on 7.12 was 1200x1024, despite my monitors native res being 1440x900. With 8.01 version its displaying 1440x900 perfectly.

One thing that ticks me off is that video playback kinda looks pixelated and blocky compared to ati drivers using vlc on my windows partition.

If anyone has options for my xorg.conf to improve video quality, flick me a private message :)

[Moronix]

Hm, yeah, no regressions. I'd just be nice to get usable vsync in opengl. That will probably take quite a while if ever.

[SledgeHammer_999]

...they only tested with some games, not more. Thats why tools like gl2benchmark show how incomplete the new codebase is - test 3/4 never worked after 8.40.4 and on 64 bit the tool even segfaults only...

Hey Kano where do I find gl2benchmark?

[homerhomer]

This driver works better than 7.12, but it still has it issues. I noticed that my computer will hibernate but will not wake properly.

Widescreen works! to tell you the truth, I feel like this was stolen from me and now given back. Thanks :)

So this driver is useable but why the hell is 2d so freaking slow? and, why doesn't my brightness work with 3d games?

I can't help to think that ATI doesn't give a "F" about it's Linux users. Yes, there are working on things and things are getting better, but in all my years of geeking out with the puters I have never came across crap drivers like these. If it's not one thing it's three other.

Please ATI, put out or go home!

I understand memory leaks happen, I can even understand botched software releases. but the crap that end users are experiencing is re-god-damn-diculous.

The whole reason to have ATI on a computer is for the flash. If it wasn't because of this I would have some crap graphics and it would be treated like sounds cards are today.

With that out of the way. The basics must be covered. This includes WORKING! I like to think of ATI as graphics specialist the have some special knowledge on what they are doing. This would mean easy stuff like a working graphics driver is the easy crap.

My definition of WORKING means that it handles basic computer functionality with ease. Really, ATI is suppose to be all about advanced graphics, not "hey it works - sort of"

I will continue to hope that the next driver release is better than the last. But in reality, I can't recommend ATI at all.

As a last note I understand that things don't get fixed overnight. On the other hand video card do. Since my endeavor of issues with ATI/AMD a ton of video cards have been released from ATI and sold. I don't mind the rapid pace of hardware sales but please have your software drivers at the same pace.

Thank you for listening

:rolleyes:

[n3Cre0]

I don't know if it has been mentioned already. But does this version support AIGLX on xorg-server-1.4? I'm still stuck with xorg-server-1.3 :(

[Kano]

@SledgeHammer_999

gl2benchmark 0.1 is in the kanotix repository and also in the original repositry:

deb http://krapfoss-systems.com/files/debian ./
deb-src http://krapfoss-systems.com/files/debian ./

it can be used for etch, feisty and gutsy should work as well (needs openscenegraph 1.2.0).

since lenny you can use a version 0.2 compiled against openscenegraph 2.2, which should be in hardy too, but was broken (don't know of current status), here you can find those packages:

deb http://kanotix.com/files/fix/gl2benchmark.lenny ./
deb-src http://kanotix.com/files/fix/gl2benchmark.lenny ./

[iluva]

Hi. As two other preposters asked, I really would like to know if the 8.01 supports the xorg server 1.4 with this release. If not I'll put the package on hold too.

[ig-88]

i seem to have a few additional problems with 8.01. Console switch is broken, which breaks shutting down, rebooting, and even the possibilty to try suspend. when i switch to a console the display receives no signal and the keybard is unresponsive (also sysrq works). switching bach to X is not possible. the machine reacts on the powerbutton, but is unable to shutdown. it probably hangs somwhere inbetween.
in 7.11 it worked fine, it also works with vesa (also the colors get shifted and radeonhd. i think it also worked in 7.12, which i didn't use much because of the widescreen bug. anyone else with this issue?

and additionally i have the feeling that 2d is realtively slow slow. waiting 100ms before a menu pops up is not what i expect.

and a third thing which i already had in 7.11, was that the small skype icon has a color shift. it turned from apple green to light torquise. this doesn't happen with radeonhd. and here comes the big one: when i first started kde4 all colors were inverted! i the effect disappeard when i enabled compositing in kwin. this may be a qt4 issue.
oh! and i have small glitches in the skype icon with 8.01.

ig-88

PS: i use ubuntu gutsy and a hd2600xt
PPS: i removed all additional options from my xorg.conf. no change.

[hoenny]

For the Fedora packaging bug, I'll push updated scripts to Phorogit for it tonight once I get back to my hotel.

I fixed script. It's working good. I send to your e-mail.

[oyvind]

For once I'm not going to be negative, I'm tired of being negative here ! So, here's a positive list of my own observations with this new release :):



This is the first version of the driver, since the new codebase was introduced, that has worked adequately on my ATI X1400-based laptop. I'm now finally able to ditch fglrx 8.37.6 (The version packaged with Ubuntu Gutsy).


Using 1680x1050 resolution is no longer a problem, phew ! I expected nothing less, though.


Suspend finally works on Ubuntu Gutsy ! This one I'm very happy about.


Finally I can enjoy much better OpenGL performance without any nasty memory leaks.


Dual-screen works better for me, I can now use a "1680x1050+1280x1024"-pairmode without problems, which makes sure that native resolution is used on both my screens and that my (big)desktop isn't actually bigger than what is physically visible. Older fglrx-versions crashed when I tried to use this pairmode (Xorg segfault).


OpenGL-screensavers are finally working (gnome-screensaver), never worked before (they reverted to software-rendering because of problems with finding suitable OpenGL visual).


The new DKMS system is a welcome addition, now I don't have to fiddle with module-assistant during installation, anymore. It worked like a charm.


The driver still has some serious problems. However, many of these problems are now on the "Known issues"-list in the official driver info. That's a good thing! Hope the list will start shrinking again, though.



Yes, there are problems. But I'm not going to repeat them all again, because they are well known. I was going to keep this post on the thumbs-up-side , so I'm sticking to that. This release has rekindled my hopes that real improvements will come further down the road. I'm hoping for more bugfixing, stability and (2D-)performance focus, not new features.

Regards,
oyvind

[oyvind]

I see some people are complaining about reduced 2D-performance. I'm not seeing this on my ATI X1400, Lenovo Z61m. 2D is the same: certainly not the fastest, but not slower than before, either.

Also, if you are experiencing stability problems with console switching, try disabling both OpenGLOverlay and VideoOverlay. If I enable any of these, especially OpenGLOverlay, the driver will often freeze X on logout (which switches to console a short period for me, since I make gdm restart X every time).

[knutbert]

Hi there,

something fundamental seems to have changed in the codebase of 8-1 as opposed to the 7-X releases.

As of 8-1 my Mobility Radeon X1900 is not working properly anymore; I get this very disturbing "failed to initialize ASIC" error

EDIT:
Moved the details of this issue to a new thread (http://phoronix.com/forums/showthread.php?t=7451)

Regards,
knutbert

[makoto42]

Under Gutsy, still cannot Suspend if the fglrx kernel module is loaded. The 3D corruption makes it impossible to use for the whole point of running it. I wish there were some official option to treat my Mobility FireGL pretty much exactly like a X1600 part, since that seemed to work pretty on par with other experiences back when using the PCI ID hack.

[bridgman]

PPS: i removed all additional options from my xorg.conf. no change.

Also, if you are experiencing stability problems with console switching, try disabling both OpenGLOverlay and VideoOverlay. If I enable any of these, especially OpenGLOverlay, the driver will often freeze X on logout (which switches to console a short period for me, since I make gdm restart X every time).

re: conf options and console switching / shutdown issues, we're hearing conflicting reports about what makes the problem go away. Oyvind, are you :

1. REMOVING the lines from conf so that the driver takes defaults for those functions ?

2. DISABLING the functions in conf, possibly by ADDING option lines ?

I am hearing both answers, but it seems unlikely both are true.

[oyvind]

re: conf options and console switching / shutdown issues, we're hearing conflicting reports about what makes the problem go away. Are you :

1. REMOVING the lines from conf so that the driver takes defaults for those functions ?

2. DISABLING the functions in conf, possibly by ADDING option lines ?

I am hearing both answers, but it seems unlikely both are true.

If you are referring to my post about the video overlay options and shutdown freeze, I am explicitly turning both overlay types off, by adding the options to the device section in xorg.conf:

Option "OpenGLOverlay" "off"
Option "VideoOverlay" "off"

Without this, things are not stable. I also make sure the AMD config database is consistent with xorg.conf by editing /etc/ati/amdpcsdb. And btw. it's kind of annyoing when the amdpcsdb is out of sync with xorg.conf because of manual editing ;).

OpenGLOverlay has always been very unstable on my ATI X1400. I remember it used to crash X when the background pixmap was loaded during login, on previous fglrx-releases. Now it causes freezes on logout. I've always disabled it.

If your reply was to something else, then ignore this post :D

[ig-88]

i tried both ways. disabling both using xorg.conf does not help. removing both lines from xorg.conf doesn't either.
i will try 7.11 again after lunch. just to be sure.

[bridgman]

If you are referring to my post about the video overlay options and shutdown freeze, I am explicitly turning both overlay types off, by adding the options to the device section in xorg.conf:

Option "OpenGLOverlay" "off"
Option "VideoOverlay" "off"

That's what I was looking for. Thanks !

[ig-88]

I just checked out 7.12. Switching to a text console works.

[knutbert]

Hi there,

I' getting this disturbing "Failed to initialize ASIC in kernel" error since version 8-01 (previous versions did not show this).

Any idea how to fix this?

Regards,
knutbert

P.S.: I'll be happy to provide more details - in fact I did, but that post somehow never showed up in this thread...

[DirtyHairy]

Observations on my T60 (mobility X1300):

Pro:
- Memory leak is gone
- widescreen works
- OpenGL speed is good
- Textured2D has noticeably improved, some rendering operations still seem to cause minor slowdowns, but overall 2D speed is noticable improved by the option (I don't use compositing or AIGLX for now)
- Suspend to RAM works nicely (as it has always done for me)

Contra:
- XV still is not working correctly
- Strange "diagonal" tearing is still present when vsync is enabled
- Powerplay has developed a tendency to cause deadlocks

Overall, this is one of the most usable releases for me since I aquired this laptop one and a half years ago :) To elaborate a bit (in a hopefully constructive fashion) on the cons:

XV has stopped working correctly somewhere around Summer last year. Since the it is crippled, only two color formats are available, and the only application that plays along nicely with it is mplayer - xine produces garbled output, and vlc can be made working by forcing the color format, but then scaling doesn't work correctly and is blocky. Also, there is a strange sort of "diagonal" tearing (also present with OpenGL and vsync) which looks like a "Z"with the top and bottom lines flipped to the outsides. I still am hoping for a version that fixes this (although I have noticed that xshm and software scaling is suprisingly good).

Much worse is the breakage of powerplay which for the first time causes trouble in this release. If Textured2D is enabled, it causes reproducable deadlocks which affect nearly everything - ssh access isn't possible. However, magic sysrq is working for syncing and rebooting, so at least parts of the kernel are left alive. If I disable Textured2D, it only locks every 4-6 times or so, but of course, that's not really better, so I disabled it for now. Since powerplay makes the difference between ~22 and ~16 watts total power consumption on my notebook, I would consider this a blocker if I wouldn't detest the memory leak even more (not to speak of the widescreen bug, I am confident that pushing this out was meant to be a bad joke).

When these critters get fixed (espeacially the last one), I for the first time will be tempted of assigning any level of "quality" to fglrx (at least on my configuration) :)

If anyone is interested, my configuration is a thinkpad T60 core duo (no "2", so 32 bit) running gentoo stable (xorg-server 1.3).

[linuxhansl]

...

3) open ATI-fglrx.spec-tmpl with an editor, go to the section that starts with %files (line 259) and add %{_sbindir}/amdnotifyui below %{_sbindir}/atieventsd

Now when I go to the root directory (tmp) and try ./ati-installer.sh --buildpkg Fedora/F8 it always returns some help information instead of building the packages.

Anyone know how to make ati-installer.sh build packages?

You need to do:

./ati-installer.sh --arguments --buildpkg Fedora/F8

I have had to do that on and off for the past few releases.

[oyvind]

Hi, just reporting some error messages I'm getting in the kernel log when using OpenGL:


[ 26.284000] [fglrx] Reserve Block - 0 offset = 0X7ffb000 length = 0X5000
[ 26.284000] [fglrx] Reserve Block - 1 offset = 0X0 length = 0X1000000
[ 26.284000] [fglrx] Reserve Block - 2 offset = 0X7fbb000 length = 0X40000
[ 26.456000] [fglrx] interrupt source 10000000 successfully enabled
[ 26.456000] [fglrx] enable ID = 0x00000008
[ 26.456000] [fglrx] Receive enable interrupt message with irqEnableMask: 10000000
[ 29.904000] NET: Registered protocol family 17
[ 44.416000] eth0: no IPv6 routers present
[ 784.388000] input: Virtual ThinkFinger Keyboard as /class/input/input9
[ 1053.648000] input: Virtual ThinkFinger Keyboard as /class/input/input10
[ 1055.720000] [fglrx:firegl_free_mutex] *ERROR* mutex id 0x00000002 not found in mutex list
[ 1056.456000] [fglrx:firegl_free_mutex] *ERROR* mutex id 0x00000002 not found in mutex list
[ 1987.692000] [fglrx:firegl_free_mutex] *ERROR* mutex id 0x00000002 not found in mutex list
[ 1994.728000] [fglrx:firegl_free_mutex] *ERROR* mutex id 0x00000002 not found in mutex list
[ 1996.592000] [fglrx:firegl_free_mutex] *ERROR* mutex id 0x00000002 not found in mutex list
[ 2004.224000] [fglrx:firegl_free_mutex] *ERROR* mutex id 0x00000002 not found in mutex list
[ 2031.516000] input: Virtual ThinkFinger Keyboard as /class/input/input11
[ 2033.824000] [fglrx:firegl_free_mutex] *ERROR* mutex id 0x00000002 not found in mutex list
[ 2033.968000] [fglrx:firegl_free_mutex] *ERROR* mutex id 0x00000002 not found in mutex list
[ 2098.104000] [fglrx:firegl_free_mutex] *ERROR* mutex id 0x0000000a not found in mutex list
[ 2130.840000] [fglrx:firegl_free_mutex] *ERROR* mutex id 0x0000000b not found in mutex list
[ 3025.856000] [fglrx:firegl_free_mutex] *ERROR* mutex id 0x00000002 not found in mutex list
[ 3026.212000] [fglrx:firegl_free_mutex] *ERROR* mutex id 0x00000002 not found in mutex list
[ 3028.060000] [fglrx:firegl_free_mutex] *ERROR* mutex id 0x00000002 not found in mutex list
[ 3028.180000] [fglrx:firegl_free_mutex] *ERROR* mutex id 0x00000002 not found in mutex list
[ 6859.616000] input: Virtual ThinkFinger Keyboard as /class/input/input12
[ 7094.480000] input: Virtual ThinkFinger Keyboard as /class/input/input13
[ 7134.004000] [fglrx:firegl_free_mutex] *ERROR* mutex id 0x00000002 not found in mutex list
[ 7167.560000] [fglrx:firegl_free_mutex] *ERROR* mutex id 0x00000002 not found in mutex list
[ 7208.792000] [fglrx:firegl_free_mutex] *ERROR* mutex id 0x00000010 not found in mutex list
[ 7220.676000] [fglrx:firegl_free_mutex] *ERROR* mutex id 0x00000012 not found in mutex list
[ 7362.628000] [fglrx:firegl_free_mutex] *ERROR* mutex id 0x00000013 not found in mutex list
[ 7472.840000] [fglrx:firegl_free_mutex] *ERROR* mutex id 0x00000014 not found in mutex list
[ 7473.700000] [fglrx:firegl_free_mutex] *ERROR* mutex id 0x00000015 not found in mutex list


The errors pop up when using OpenGL-applications in general. They do not seem to cause any harm, but X hasn't been running for more than 2-3 hours. The Virtual Thinkfinger stuff can be ignored, just my fingerprint reader.