AMD 8.41.7 Display Driver Released -- The Holy Crap Edition!

[mjordan]

To change your xorg.conf you need to run aticonfig --initial

The driver found in the Ubuntu repository is not 8.41 and thus doesn't support the HD 2600XT.

What does your xorg.conf look like? I've never seen that error on an R600 chip.

This is my xorg.conf after successfully running aticonfig --initial


Section "Monitor"
Identifier "Generic Monitor"
HorizSync 28.0 - 51.0
VertRefresh 43.0 - 60.0
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 "Generic Video Card"
Driver "fglrx"
BusID "PCI:6:0:0"
EndSection

Section "Device"
Identifier "aticonfig-Device[0]"
Driver "fglrx"
EndSection

Section "Screen"
Identifier "Default Screen"
Device "Generic Video Card"
Monitor "Generic Monitor"
DefaultDepth 24
SubSection "Display"
Depth 1
Modes "1024x768" "800x600" "640x480"
EndSubSection
SubSection "Display"
Depth 4
Modes "1024x768" "800x600" "640x480"
EndSubSection
SubSection "Display"
Depth 8
Modes "1024x768" "800x600" "640x480"
EndSubSection
SubSection "Display"
Depth 15
Modes "1024x768" "800x600" "640x480"
EndSubSection
SubSection "Display"
Depth 16
Modes "1024x768" "800x600" "640x480"
EndSubSection
SubSection "Display"
Depth 24
Modes "1024x768" "800x600" "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 "DRI"
Mode 0666
EndSection

Section "Extensions"
Option "Composite" "0"
EndSection


...and still can't run X :S

[Malikith]

Yeah, the driver didn't work for me either, I got the glesx.so issue. I hope the 8.42 driver is the one, if AMD doesn't hurry up, I might have a new nVidia card soon.

[container]

"
Section "Extensions"
Option "Composite" "0"
EndSection
"

Mine looks like this:
"
Section "Extensions"
Option "Composite" "Disable"
EndSection
"

I wonder if that form makes a difference?
edit: I've also got these in the "Device" section:
Option "XAANoOffscreenPixmaps"
Option "UseInternalAGPGART" "yes"
Option "VideoOverlay" "on"
No idea if they matter - they don't look like it.

[mjordan]

"
Section "Extensions"
Option "Composite" "0"
EndSection
"

Mine looks like this:
"
Section "Extensions"
Option "Composite" "Disable"
EndSection
"

I wonder if that form makes a difference?

same error either way :S

thanks though!

[Malikith]

"
Section "Extensions"
Option "Composite" "0"
EndSection
"

Mine looks like this:
"
Section "Extensions"
Option "Composite" "Disable"
EndSection
"

I wonder if that form makes a difference?

I think they're the same thing, he can try this though, add this at the bottom of your xorg.conf:

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

[Alistair]

Roderick Colenbrander has a workaround against current wine that fixes the Visual overlay detection - I've built an overlay on my gentoo for wine 0.9.44 and the patch applies clean and appears to fix the problem of detecting the visual -- unfortunately in this state World of Warcraft is utterly unplayable - the characters and most game features are not visible on the play field -- although terrain and major objects are (i.e. player toons, npc's enemies trees bushes are invisible, but inns, ground, flightpath stands are visible.

I did NOT notice a frame rate improvement in this state ...


*rollback time*

[mjordan]

I think they're the same thing, he can try this though, add this at the bottom of your xorg.conf:

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


didn't make a difference...

[Malikith]

I did some digging around in the Name The Next ATI Driver Contest thread and found a suitable name, I think I found it.

8.41 "life is a one big betatest edition"

[mjordan]

haha, nice :)

I actually am making progress.. no X errors anymore, startx just loads a black screen :S

[container]

haha, nice :)

I actually am making progress.. no X errors anymore, startx just loads a black screen :S

Are you using xorg 7.3? I think ATI says it's somehow not compatible with it, yet.

[rbmorse]

No joy here with my X1950XTX and Ubuntu Gutsy (7.10 - Tribe 5). Install runs to completion, but on reboot the screen refresh is not correct and resists resetting. I either get a skewed presentation and the images are all repeated about four times across the width of the display or Gutsy dumps me into the failsafe settings screen, but that wont let me correct the monitor config nor will it allow selection of the FGLRX driver.

You say there is a problem with Xorg 7.3?

[DarkFoss]

Working here on Mandriva 2007.1 32 bit Mandriva's rpm script is broken but the ati installer worked fine.glxgears went from 3k to over 8k fgl_glxgears went from 1300 to 1800.

Fonts look much crisper, UT2K4 i is smoother faster and looks great... haven't had a chance to do much else tonite.

Nice job ATI.. :) Congrats M Tippet and co !

[mjordan]

Can someone post an xorg.conf that actually works? I'm still stuck at a black screen when I let aticonfig --initial set up my xorg.conf.

I wish it were as simple as envy -g :'(

[container]

Can someone post an xorg.conf that actually works? I'm still stuck at a black screen when I let aticonfig --initial set up my xorg.conf.

I wish it were as simple as envy -g :'(

Here's mine. It might not be the best one, but it works (for me, I assume it depends on your xorg version).


# Xorg configuration created by system-config-display

Section "ServerLayout"
Identifier "single head configuration"
Screen 0 "aticonfig-Screen[0]" 0 0
InputDevice "Keyboard0" "CoreKeyboard"
EndSection

Section "Files"
EndSection

Section "Module"
Load "dbe"
Load "fbdevhw"
Load "record"
Load "freetype"
Load "type1"
Load "dri"
Load "glx"
Load "extmod"
EndSection

Section "InputDevice"
Identifier "Keyboard0"
Driver "kbd"
Option "XkbModel" "pc105"
Option "XkbLayout" "fi"
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 "XAANoOffscreenPixmaps"
Option "UseInternalAGPGART" "yes"
Option "VideoOverlay" "on"
EndSection

Section "Screen"
Identifier "aticonfig-Screen[0]"
Device "aticonfig-Device[0]"
Monitor "aticonfig-Monitor[0]"
DefaultDepth 24
SubSection "Display"
Viewport 0 0
Depth 24
Modes "1024x768" "800x600" "640x480"
EndSubSection
EndSection

Section "DRI"
Group 0
Mode 0666
EndSection

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

[mjordan]

Hmm, well I give up.. I'm running Xorg 7.2 so that's not the issue... just can't get past the black screen..... guess I'll go back to VESA for now.

[rbmorse]

Tried the new driver on Kubuntu 7.04 (Xorg 7.2) with the X1950XTX. Install proceeded to completion, X started, initial impression is quite good.

The KDE control panel lists the loaded driver as ATI Radeon fbdev. There is an entry for ATI Radeon fglrx, but selecting that results in the blinking cursor of indecision when I restart X. Correcting that required the ritual "dpkg-reconfigure xserver-xorg" dance, and that got me back to the ATI Radeon fbdev driver. The ATI Catylist Control Center, OTHO, says driver version 8.41.7. Is fbdev in fact the correct driver?

Checking the configuration with fglrxinfo returned the following error:

fglrxinfo: error while loading shared libraries: libGL.so.1: cannot open shared object file: No such file or directory

In fact, trying any GL related action returned that error. There is a /usr/lib/libGL.so.1.2 on the system and making a symlink with that to the missing file takes care of that.

Fglrxinfo says: display: :0.0 screen: 0
OpenGL vendor string: ATI Technologies Inc.
OpenGL renderer string: Radeon X1950 Series
OpenGL version string: 2.0.6849 Release


The not-a-benchmark glxgears looses about 1000 fps from the 8.37 drivers (from 11500 to 10500), but I haven't restarted X, yet. What's the routine with the gears inside the spinning cube?

[niniendowarrior]

*Sigh*

I haven't even downloaded and already the tomato throwing has started. I'll probably try it soon though the drive to test it isn't as it was.

[Malikith]

Tried the new driver on Kubuntu 7.04 (Xorg 7.2) with the X1950XTX. Install proceeded to completion, X started, initial impression is quite good.

The KDE control panel lists the loaded driver as ATI Radeon fbdev. There is an entry for ATI Radeon fglrx, but selecting that results in the blinking cursor of indecision when I restart X. Correcting that required the ritual "dpkg-reconfigure xserver-xorg" dance, and that got me back to the ATI Radeon fbdev driver. The ATI Catylist Control Center, OTHO, says driver version 8.41.7. Is fbdev in fact the correct driver?

Checking the configuration with fglrxinfo returned the following error:



In fact, trying any GL related action returned that error. There is a /usr/lib/libGL.so.1.2 on the system and making a symlink with that to the missing file takes care of that.

Fglrxinfo says:

The not-a-benchmark glxgears looses about 1000 fps from the 8.37 drivers (from 11500 to 10500), but I haven't restarted X, yet. What's the routine with the gears inside the spinning cube?

Hmmm... Does your xorg.conf actually have the driver listed as fbdev? Thats interesting.

[rbmorse]

xorg.conf says "fglrx" like it should. Yes, this is interesting. The problem could be KDE's though.

A couple of days ago I posted a note indicating that if AMD/ATI was smart they would release this driver as a "beta" version. It sure feels like one.

The other problem I noted is one someone else reported...closing an openGL window leaves an artifact...the window. Switching to term and back clears the error.

[Malikith]

Yeah and if anyone wants to try and help me figure out my issue, which I doubt, I'll probably have to wait until 8.42, check out my Xorg.0.log:

My Xorg log (http://bloodytampon.servegame.com/downloads/Xorg.0.log)

If you have any ideas, feel free to suggest.

Just so you know, 8.40 does work.

[sundown]

Not my cup of tea for my mobility radeon. I guess I'll wait for next month's.

And just for the recored, I bought my laptop without Windows installed on it, some barebones have that option ;)

[Silent Storm]

Oh, so a game gets a few more FPS, and that makes it great? Wow. Obviously doesn't take much to satisfy some.

For others, we think that a few more serious bugs needs to be solved, and I predicted them not to have been, and I was right. Unfortunately.
I'm glad you're happy with ATI's amazing progress though.

Just one question opera: Have you ever developed a complex program that does a lot of processing (or just any program in C or a C like language)? It needn't be a driver, be a hardware-software interface of any kind but a program which is performance critical and huge.

If you've worked in a project like these, I think you'll have that sense of appreciation. Especially in the arena of "extremely programmable hardware".

Since AMD bought ATI, things have changed. They are trying to give a seamless rebirth to a new driver generation without breaking the support. Things would be simpler if they said: "Wait 1 year without any driver and you'll see the miracle" but at least they've "tried" to support new cards and things.

I'm a computer engineer. I'm a fetishist about processors and hardware. I know what takes to program these (taking notes to papers, converting registry values, sometimes debugging assembly crypts). And I use these guys cards since nVidia was not able to make gauraud shading and smoothing in hardware.

ATI is a huge company which has some fair share in customized simulator graphics, mutli-head 2D graphics (Their creepy 9000 chip can do 16 heads if required!) and such. So it was their fault not to support linux but with AMD it has changed.

Yes I have some weird and annoying bugs too. I was using my card with vesa for 8 months but with 8.41 things started to change. I'm not hopeless about ATI. They have opened their spex too. It's equally important as 8.41 drivers.

OTOH, It's important to say that ATI's creepy driver is far better in hotplugging hardware and screen detection than nVidia.

So opera, give the guys some respect and credit. Your lovely nVidia guys have an annoying locking-up bug in their latest drivers and they didn't fixed it for 80 days. c'mon!

Instead pf provocating the board, please shut up. Already nobody listening to you but, I was unable to resist to write this post!.

[Thetargos]

And just for the recored, I bought my laptop without Windows installed on it, some barebones have that option ;)
Where did you get it? I'm currently hunting for a laptop without an OS or preferrably with Linux preloaded. I looked at
Empreror Linux (www.emperorlinux.com), but bringing one of those babies down here is as costly as a MacBook... Pro (quite pricey), and that for a Toucan T61 model, but with customs and other taxes, plus shipping and handling it is virtually twice the price.

[spielc]

Okey i hope to stay as objective as possible in this post (going to be hard because i am biased in some way... actually who the hell isn't biased when it comes to linux and amd/ati?)!

First of all i think almost all of the users do respect your effords Michael. You test every release and well write reviews about it. I don't know if ATI/AMD are so happy about it but probably they weren't with this months review. The only problem you have Michael is to be objective (again i do understand it). In your 8.41 review you said:


...
While the 8.41 release is designed for the R600 series, in our tests with the R300 through R500 series the driver has worked out quite well. There have been some corruption issues and other minor problems but nothing severe.
...


So you did mention that this release is primarily for the r600-series but you said that there were MINOR problems and NOTHING severe on the older cards when you tested the driver. From what i've read in this forums there are BIG problems with this drivers on the older generation cards. Again i do understand that you can't test every single card with an ATI-chip but still when you read the review you get the impression that it the new driver WILL work on older cards and you WILL get the promised huge performance boost. That's why alot of ppl (including myself i have to admit) got so excited about it and that's why we are so disappointed now after we had to remark that it's not as it was promised it would be.

As the only real deal i need the vid-card for in linux (at the moment... in the future it would be cool to have desktop-effects with aiglx/compiz-fusion too...) is watching movies (for playing games i still use good old windows) and i read that xv is still broken (on some cards and i'm one of the lucky ppl who seems to own such a card...) i probably won't try this release (maybe i get bored and tell myself that i can use some kind of adventure and still try it out who knows... ;) )

Once again i do understand that you can't test everything but from a reviewer i do expect some kind of unbiased/objective view of things when he or she reviews something (still i do understand that this is almost impossible as everybody has some kind of personal opinion which affects things... when i would write a ati-linux-driver review it would possibly be alot more skeptical as yours are because of the experiences i made with linux and ati-cards...)

[Malikith]

As the only real deal i need the vid-card for in linux (at the moment... in the future it would be cool to have desktop-effects with aiglx/compiz-fusion too...) is watching movies (for playing games i still use good old windows) and i read that xv is still broken (on some cards and i'm one of the lucky ppl who seems to own such a card...) i probably won't try this release (maybe i get bored and tell myself that i can use some kind of adventure and still try it out who knows... ;) )

Yeah, and maybe you'll be luckier than me, I didn't even get far enough to try it, it simply doesn't work. I hope AMD/ATI can fix that AGP bus issue too (fglrx detecting my agp card as pcie), because from the looks of the xorg log, it still exists in 8.41:

(II) Loading sub module "fglrxdrm"
(II) LoadModule: "fglrxdrm"
(II) Loading /usr/lib/xorg/modules/linux//libfglrxdrm.so
(II) Module fglrxdrm: vendor="FireGL - ATI Technologies Inc."
compiled for 7.1.0, module version = 8.41.7
ABI class: X.Org Server Extension, version 0.3
(--) fglrx(0): VideoRAM: 262144 kByte, Type: DDR3
(II) fglrx(0): PCIE card detected

It should say something like this instead, but i've never in the history of fglrx using this card, have seen it:

(II) Loading sub module "fglrxdrm"
(II) LoadModule: "fglrxdrm"
(II) Loading /usr/lib/xorg/modules/linux//libfglrxdrm.so
(II) Module fglrxdrm: vendor="FireGL - ATI Technologies Inc."
compiled for 7.1.0, module version = 8.41.7
ABI class: X.Org Server Extension, version 0.3
(--) fglrx(0): VideoRAM: 262144 kByte, Type: DDR3
(II) fglrx(0): AGP card detected

[phreadom]

Yeah, and maybe you'll be luckier than me, I didn't even get far enough to try it, it simply doesn't work. I hope AMD/ATI can fix that AGP bus issue too (fglrx detecting my agp card as pcie), because from the looks of the xorg log, it still exists in 8.41:

(II) Loading sub module "fglrxdrm"
(II) LoadModule: "fglrxdrm"
(II) Loading /usr/lib/xorg/modules/linux//libfglrxdrm.so
(II) Module fglrxdrm: vendor="FireGL - ATI Technologies Inc."
compiled for 7.1.0, module version = 8.41.7
ABI class: X.Org Server Extension, version 0.3
(--) fglrx(0): VideoRAM: 262144 kByte, Type: DDR3
(II) fglrx(0): PCIE card detected

It should say something like this instead, but i've never in the history of fglrx using this card, have seen it:

(II) Loading sub module "fglrxdrm"
(II) LoadModule: "fglrxdrm"
(II) Loading /usr/lib/xorg/modules/linux//libfglrxdrm.so
(II) Module fglrxdrm: vendor="FireGL - ATI Technologies Inc."
compiled for 7.1.0, module version = 8.41.7
ABI class: X.Org Server Extension, version 0.3
(--) fglrx(0): VideoRAM: 262144 kByte, Type: DDR3
(II) fglrx(0): AGP card detected

I'm having the same problem... incorrectly detects my AGP card as PCIE. This seems to have the cascade effect of also causing 0x AGP speed (base PCI) and having the driver fail to see my 512mb ram and instead see only 256.

I also have the glesx.so crash on X start, so 3d acceleration etc is completely absent from this release for me.

The AGP/PCIE problem cascade was in the 8.40 drivers as well, but the glesx.so crash is new to 8.41.

[Malikith]

I'm having the same problem... incorrectly detects my AGP card as PCIE. This seems to have the cascade effect of also causing 0x AGP speed (base PCI) and having the driver fail to see my 512mb ram and instead see only 256.

I also have the glesx.so crash on X start, so 3d acceleration etc is completely absent from this release for me.

The AGP/PCIE problem cascade was in the 8.40 drivers as well, but the glesx.so crash is new to 8.41.

Yeah, my video ram amount is actually detected properly (256), just the AGP problem remains, problem with 0x agp, is well, things that already are slow, get alot slower ;). I'm guessing you've got a X1600 pro? I got a x1600xt. I've tried the card in Windows and well, it flies in Windows. Luckily I have a Geforce FX 5500 to hold me off, until a good driver comes, funny thing is, its faster than the x1600xt right now, and its alot more reliable for me. But I would much rather run the x1600xt when the time is right.

[Ralle]

WOOT!
_64 ??

Does this mean what I think it does?

[phreadom]

Yeah, my video ram amount is actually detected properly (256), just the AGP problem remains, problem with 0x agp, is well, things that already are slow, get alot slower ;). I'm guessing you've got a X1600 pro? I got a x1600xt. I've tried the card in Windows and well, it flies in Windows. Luckily I have a Geforce FX 5500 to hold me off, until a good driver comes, funny thing is, its faster than the x1600xt right now, and its alot more reliable for me. But I would much rather run the x1600xt when the time is right.

Yep, x1600Pro 512mb. Unfortunately I don't have a GeForce to fall back on... so I'll just wait it out. I'm having enough fun in Linux aside from that to keep me from going back into Windows unless it's an "emergency". :)

For all the bitching I've done, I really do appreciate the direction AMD/ATI is going... I'm just a little put out after high expectations for the current driver release.

[Malikith]

Yep, x1600Pro 512mb. Unfortunately I don't have a GeForce to fall back on... so I'll just wait it out. I'm having enough fun in Linux aside from that to keep me from going back into Windows unless it's an "emergency". :)

For all the bitching I've done, I really do appreciate the direction AMD/ATI is going... I'm just a little put out after high expectations for the current driver release.

Yeah, hopefully things will go right on the 8.42 driver. I hope the hype that was done on this one, was actually mean't for that one. We're getting AIGLX on that release, which is interesting considering the amount of bugs that still remain.

But hopefully things will go right, AMD knows of the AGP bus issue, just hasn't been fixed yet, perhaps they have the rest of their new codebase going in? I couldn't imagine that being all, theres still alot that needs to be replaced/fixed from what I can see. But this is a start. This is the most they've attempted to do with their linux driver since, well, ever.

[spielc]

Yep, x1600Pro 512mb. Unfortunately I don't have a GeForce to fall back on... so I'll just wait it out. I'm having enough fun in Linux aside from that to keep me from going back into Windows unless it's an "emergency". :)

For all the bitching I've done, I really do appreciate the direction AMD/ATI is going... I'm just a little put out after high expectations for the current driver release.

Same here... I just remarked that i have the same problem as you do (the agp/pci-e and 512 mb issue) so i just stick with 8.40-driver till the 8.42-driver is released with hopefully the fixes included that we need to get a painless linux-experience! :)

[Michael]

So you did mention that this release is primarily for the r600-series but you said that there were MINOR problems and NOTHING severe on the older cards when you tested the driver. From what i've read in this forums there are BIG problems with this drivers on the older generation cards. Again i do understand that you can't test every single card with an ATI-chip but still when you read the review you get the impression that it the new driver WILL work on older cards and you WILL get the promised huge performance boost. That's why alot of ppl (including myself i have to admit) got so excited about it and that's why we are so disappointed now after we had to remark that it's not as it was promised it would be.

As the only real deal i need the vid-card for in linux (at the moment... in the future it would be cool to have desktop-effects with aiglx/compiz-fusion too...) is watching movies (for playing games i still use good old windows) and i read that xv is still broken (on some cards and i'm one of the lucky ppl who seems to own such a card...) i probably won't try this release (maybe i get bored and tell myself that i can use some kind of adventure and still try it out who knows... ;) )

Once again i do understand that you can't test everything but from a reviewer i do expect some kind of unbiased/objective view of things when he or she reviews something (still i do understand that this is almost impossible as everybody has some kind of personal opinion which affects things... when i would write a ati-linux-driver review it would possibly be alot more skeptical as yours are because of the experiences i made with linux and ati-cards...)

There was close to a dozen ATI graphics cards tested with this 8.41 release here internally and I stand by my statement that I had only ever run into minor issues with the 8.41 release driver (back in late July / early August time frame there were known compatibility problems with some Intel motherboard chipsets and a couple other known issues, but those were all corrected by 8.41.5~8.41.6 era). On my main system that I do my daily work from, which I use about ~80 hours a week, had never once had a stability problem or anything wrong with it aside from some corruption on the watermark and mouse. To be honest, I was shocked by some of the problems you guys are having. Those on the ATI beta list also didn't seem to report many problems.

Heck, for those of you with a nasty bug and are committed to seeing the problem resolved, send me a private message with your name, email address, and a brief description of your problem(s) and I can see about getting you on AMD's beta list so you will be able to report your problems directly to AMD and try out the actual beta drivers.

[sundown]

Where did you get it? I'm currently hunting for a laptop without an OS or preferrably with Linux preloaded. I looked at
Empreror Linux (www.emperorlinux.com), but bringing one of those babies down here is as costly as a MacBook... Pro (quite pricey), and that for a Toucan T61 model, but with customs and other taxes, plus shipping and handling it is virtually twice the price.

from http://1toppc.com/Merchant2/
For instance http://1toppc.com/Merchant2/merchant.mv?Screen=PROD&Product_Code=S96S
But in this particular case we're talking about Nvidia now.

[jackkerouac]

There was close to a dozen ATI graphics cards tested with this 8.41 release here internally and I stand by my statement that I had only ever run into minor issues with the 8.41 release driver (back in late July / early August time frame there were known compatibility problems with some Intel motherboard chipsets and a couple other known issues, but those were all corrected by 8.41.5~8.41.6 era). On my main system that I do my daily work from, which I use about ~80 hours a week, had never once had a stability problem or anything wrong with it aside from some corruption on the watermark and mouse. To be honest, I was shocked by some of the problems you guys are having. Those on the ATI beta list also didn't seem to report many problems.


Dude, you hyped up expectations too much and you got burned. That's just the way it is. You should have been more objective and you weren't. You took a credibility hit and that sucks, but maybe it teaches you a lesson not to be so excited by future AMD/ATI driver releases so your readers don't wind up getting so pissed off when the driver doesn't meet the lofty expectations you've set.

Sorry, but that's the way I see it. Yes, AMD/ATI is going in the right direction. Yes, they are opening their drivers, which is great. But this driver doesn't work for me (x1650 Pro), so to me, this driver sucks. It just doesn't live up to the massive hype you gave it. The "Holy Crap Edition"? Maybe as in "Holy Crap, I got screwed by AMD/ATI again!"

[Silent Storm]

Dude, you hyped up expectations too much and you got burned. That's just the way it is. You should have been more objective and you weren't. You took a credibility hit and that sucks, but maybe it teaches you a lesson not to be so excited by future AMD/ATI driver releases so your readers don't wind up getting so pissed off when the driver doesn't meet the lofty expectations you've set.

Sorry, but that's the way I see it. Yes, AMD/ATI is going in the right direction. Yes, they are opening their drivers, which is great. But this driver doesn't work for me (x1650 Pro), so to me, this driver sucks. It just doesn't live up to the massive hype you gave it. The "Holy Crap Edition"? Maybe as in "Holy Crap, I got screwed by AMD/ATI again!"


I'll back michael here because I've a parallel experience when compared to Micahel's. I'm using a X1650XT (1950Pro sibling) and I'm having a good experience with these drivers too. Yes. drivers may have bugs but they didn't hit neither him nor me during the development of this driver. So instead of shouting blindly please try to be a bit cooler, OK?

Edit: Grammar corrections.

[adamk]

Dude, you hyped up expectations too much and you got burned. That's just the way it is. You should have been more objective and you weren't. You took a credibility hit and that sucks, but maybe it teaches you a lesson not to be so excited by future AMD/ATI driver releases so your readers don't wind up getting so pissed off when the driver doesn't meet the lofty expectations you've set.


It certainly met the expectations I had, based on Michael's review. Sorry you've had such a bad time with 8.41, but clearly not everyone is having these problems, so it's not hard to believe that Michael didn't see them either.

Adam

[Malikith]

Dude, you hyped up expectations too much and you got burned. That's just the way it is. You should have been more objective and you weren't. You took a credibility hit and that sucks, but maybe it teaches you a lesson not to be so excited by future AMD/ATI driver releases so your readers don't wind up getting so pissed off when the driver doesn't meet the lofty expectations you've set."

Hey, think of it like this, if you tested the driver, and had no issues with it, or just a few minor ones for the things you do and the things you test, you would be pretty happy with it right? So you would say good things, its natural. So when all the other people with different cards such as AGP x1k cards like myself try it, its a different story.

Sorry, but that's the way I see it. Yes, AMD/ATI is going in the right direction. Yes, they are opening their drivers, which is great. But this driver doesn't work for me (x1650 Pro), so to me, this driver sucks. It just doesn't live up to the massive hype you gave it."

Yes ATI is going in the right direction, I have some confidence in their next driver. Am I frustrated? Sure I am, but I know that in time, this stuff will be a laugh in the past once this all gets fixed.

The "Holy Crap Edition"? Maybe as in "Holy Crap, I got screwed by AMD/ATI again!"

I think with this version of the fglrx drivers, there are two names for it, theres one for the people who had success and the one that well, didn't exactly work out. You sir, win the failure name. Congratulations. I don't know who came up with the holy crap edition name in the first place, but that rocks.

[fexx]

OK,

I've managed to get it running somehow in some situations (Debian/Sid, Mobility X600):

1. The driver doesn't seem to work *without* the fglrx module loaded as opposed to previous drivers
2. On Debian/Sid the link /usr/lib/libGL.so.1 is missing. Do 'cd /usr/lib/; ln -s libGL.so.1.2 libGL.so.1'

Suspend behaviour:

1. s2disk works fine
2. s2ram does work on some kernels

Performance:

Q3A: Framerate drops from 170 to 156 fps (mydemo)
Doom 3: Framerate increases from 30 to 45 fps (standard demo)

Major issue:

Whenever I start anything from KDE's kicker panel the lower right corner gets corrupted (colored stripes) and the cursor drags similar colored stripes behind.

Conclusion:

Performance *has not been* and *is not* the major issue of fglrx drivers, it's unreliability!

Updating the driver has always been some kind of gambling: Will it work at all? Which new issues will arise? Moreover, using an ATI card you get used to being afraid of updating your kernel or xorg version as well because the result is unpredictable (well, usually X is broken afterwards...) In other words, fglrx drivers have caused me way too much trouble, worries and uncertainty. Nothing has changed in this aspect.

What have we learned from the new driver?

1. ATI cards can perform much better the they do with fglrx-drivers < 8.41.7. Well, we knew that before...
2. ATI people *are* able to write better performing drivers on Linux systems. That's really good news.
3. Some hypes are hypes ;-)

If only they had released this as a preview, or public beta. Our expectations would not have been so high...

Regards,

fexx



think I'm going back...

[Trek1701]

Hi fexx,

I'm running Debian Etch with ATI Radeon 9550 AGP from PowerColor and 8.41.7 installed without any issues. About the performance, I haven't noticed any difference.

Here are some basic steps I follow when installing a new ATI driver on Debian and some other commands:
Intall linux-headers:
sudo apt-get install linux-headers-$(uname -r)

Remove older ATI driver:
sudo sh /usr/share/ati/fglrx-uninstall.sh

Intall new ATI driver:
chmod +x ati-driver-installer-8.41.7-x86.x86_64.run
sudo sh ati-driver-installer-8.41.7-x86.x86_64.run

Edit Xorg manually if needed:
sudo gedit /etc/X11/xorg.conf

Edit Catalist manually if needed::
sudo gedit /etc/ati/amdpcsdb

Check if Radeon driver is enable:
lsmod | grep radeon

Check if fglrx driver is enable:
lsmod | grep fglrx

Rebuilding your kernel module dependency:
sudo depmod -a

Disable Radeon driver if enable:
sudo modprobe -r radeon

Enable fglrx driver if disable:
sudo modprobe fglrx

Test 3D hardware acceleration on terminal:
fgl_glxgears


Trek1701

[Kano]

I would never use the standard installer for Debian/Ubuntu. If you want to install it manually, remove all packages with fglrx in the name first, then create a package for Ubuntu, even for Etch, Debian/Edgy works for Xorg 7.1+ which Etch uses. Newer targets are not needed, these just do the same.

[rbmorse]

This is for Malikith in response to #220:

I don't know if this will help any, but mine is pretty similar until you get to this point:

yours:
(II) Initializing built-in extension DAMAGE
(II) Initializing built-in extension XEVIE
(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

Backtrace:
0: /usr/bin/X11/X(xf86SigHandler+0x81) [0x80c8631]
1: [0xffffe420]
2: /usr/lib/xorg/modules//glesx.so [0xb5bb2a24]
3: /usr/lib/xorg/modules//glesx.so [0xb5bac09c]
4: /usr/lib/xorg/modules//glesx.so [0xb5b3f3fb]
5: /usr/lib/xorg/modules//glesx.so [0xb5b51196]
6: /usr/lib/xorg/modules//glesx.so [0xb5b3e3d9]
7: /usr/lib/xorg/modules//glesx.so [0xb5b67c12]
8: /usr/lib/xorg/modules//glesx.so [0xb5b55ebd]
9: /usr/lib/xorg/modules//glesx.so [0xb5b57469]
10: /usr/lib/xorg/modules//glesx.so [0xb5ade6e3]
11: /usr/lib/xorg/modules//glesx.so [0xb5ad3f88]
12: /usr/lib/xorg/modules//glesx.so [0xb5ad4670]
13: /usr/lib/xorg/modules//glesx.so(esutInit+0x74) [0xb5ad3484]
14: /usr/lib/xorg/modules//glesx.so [0xb5acf17b]
15: /usr/lib/xorg/modules//glesx.so(GlesxExtensionInit+0x9b) [0xb5ace3fb]
16: /usr/bin/X11/X(InitExtensions+0xa2) [0x80f6332]
17: /usr/bin/X11/X(main+0x2af) [0x8076c9f]
18: /lib/i686/cmov/libc.so.6(__libc_start_main+0xe0) [0xb7db6050]
19: /usr/bin/X11/X(FontFileCompleteXLFD+0x1ed) [0x80761c1]

Fatal server error:
Caught signal 11. Server aborting

Mine:
(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: dlopen of /usr/lib/dri/fglrx_dri.so failed (/usr/lib/dri/fglrx_dri.so: cannot open shared object file: No such file or directory)
(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 7.2.0, 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" "ExplorerPS/2"
(**) Configured Mouse: Device: "/dev/input/mice"
(**) Configured Mouse: Protocol: "ExplorerPS/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
(**) Option "SendCoreEvents"
(**) stylus: always reports core events
(**) stylus device is /dev/input/wacom
(**) stylus is in absolute mode
(**) stylus: forcing TabletPC ISD V4 protocol
(**) WACOM: suppress value is 2
(**) Option "BaudRate" "9600"
(**) stylus: serial speed 9600
(**) Option "SendCoreEvents"
(**) cursor: always reports core events
(**) cursor device is /dev/input/wacom
(**) cursor is in relative mode
(**) cursor: forcing TabletPC ISD V4 protocol
(**) WACOM: suppress value is 2
(**) Option "BaudRate" "9600"
(**) cursor: serial speed 9600
(**) Option "SendCoreEvents"
(**) eraser: always reports core events
(**) eraser device is /dev/input/wacom
(**) eraser is in absolute mode
(**) eraser: forcing TabletPC ISD V4 protocol
(**) WACOM: suppress value is 2
(**) Option "BaudRate" "9600"
(**) eraser: serial speed 9600
(II) XINPUT: Adding extended input device "eraser" (type: Wacom Eraser)
(II) XINPUT: Adding extended input device "cursor" (type: Wacom Cursor)
(II) XINPUT: Adding extended input device "stylus" (type: Wacom Stylus)
(II) XINPUT: Adding extended input device "Configured Mouse" (type: MOUSE)
(II) XINPUT: Adding extended input device "Generic Keyboard" (type: KEYBOARD)
(**) Option "Device" "/dev/input/wacom"
(EE) xf86OpenSerial: Cannot open device /dev/input/wacom
No such file or directory.
Error opening /dev/input/wacom : No such file or directory
(**) Option "Device" "/dev/input/wacom"
(EE) xf86OpenSerial: Cannot open device /dev/input/wacom
No such file or directory.
Error opening /dev/input/wacom : No such file or directory
(**) Option "Device" "/dev/input/wacom"
(EE) xf86OpenSerial: Cannot open device /dev/input/wacom
No such file or directory.
Error opening /dev/input/wacom : No such file or directory
(II) Configured Mouse: ps2EnableDataReporting: succeeded
Could not init font path element /usr/X11R6/lib/X11/fonts/misc, removing from list!
Could not init font path element /usr/share/fonts/X11/cyrillic, removing from list!
Could not init font path element /usr/X11R6/lib/X11/fonts/Type1, removing from list!
Error in "atiddxMiscUpdateFile" -4


Maybe some knowledgeable person can help sort this out.

[Malikith]

Error in "atiddxMiscUpdateFile" -4

Hmm, at the end of your log, that is something right there I think, however, I don't really know what it is unfortunately. This driver causes some really confusing issues that just don't seem to lead anywhere except in the driver itself. You have a AGP card from the x300-x800 or x1300-1900 I assume?

It seems like yours is getting further than mine, but that atidxMiscUpdateFile thing seems to stop it dead in its tracks. The only thing I can really think of, is that there is really no solution to these issues, and that we're just going to have to wait for the 8.42 driver.

Reason why I say that is because it seems like the R600 support seemed to have screwed support for alot of other cards, especially AGP cards, not that they weren't screwed before.

Also heres another thing:

(EE) AIGLX: reverting to software rendering

[Malikith]

Oh by the way, try adding these I know these probably won't make a difference but put these at the end of your xorg.conf and try starting x, if it fails which i'm guessing it will, let me see your log again. There might be more answers then. Just maybe.

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

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

[rbmorse]

I dunno. Me just stupid user.

I missed that you are using an AGP card. Mine is PCI-e and frankly when I looked at the logs I started from the bottom up to see where the diversion happened. I didn't see anything notable furhter up, but did not really look that closely.

[rbmorse]

Error in "atiddxMiscUpdateFile" -4

Hmm, at the end of your log, that is something right there I think,

I went back and cleaned up all the WACOM associated stuff and that cleared this particular error.

[HighHo]

Just a small update to my previous post regarding gaming and the new drivers with an x800.

Cedega and WoW:
Results in some funny display issues with pixel shaders enabled with the 6.0.3 engine. This can be fixed by either disabling the Pixel Shaders, or as a better solution using the 6.0.1 engine which has no issues with pixel shaders enabled.

UT2004:
No issues plays very well with full graphic options enabled.

Warsow:
0.31 found on the Ubuntu repositories resulted in the game not loading with a message that there is not enough video memory,
downloading the latest version directly from the Warsow website resolves this, and plays without any issues.

[oyvind]

I had heard many positive things about this new release. That's why I just got very disappointed with the 8.41 driver... Many basic problems are still there - in fact, all of the basic problems I had with the previous release of fglrx.

* No visible improvement in 2D rendering speed, and I still need to use the "XaaNoOffscreenPixmaps" "on" Option in xorg.conf, otherwise some web pages will lag/hang when scrolling (digg.com, for instance).

* XVideo/textured video is broken as always, horrible diagonal tearing, and no, setting TexturedVideoSync to On does not make any difference. The scaling quality is truly an abomination, but I guess this is due to the general brokenness of fglrx' XV implementation (on my hardware at least).

* I still get fglrx memory mapping errors in kernel log when starting up a second display (for user switching). This has been a problem for me since I started using fglrx.

* Dual-head / BigDesktop mode is just as quirky as it always has been, with the occasional flash of screen-corruption, which disappears very quickly again (using a Samsung Syncmaster LCD on the VGA output).

I have a Lenovo Z61m with an ATI X1400 in it.

In short, basic things are still broken, and improved OpenGL performance doesn't really give me any value before these issues are fixed. How about making 2D and video acceleration work a bit better ?

OK, so, this was quite negative, but as the disappointment wears off, I'm sure things will improve, at least now that there are specs out. And it is a beta-release, after all, so I really have hopes for 8.42 ..

Cheers,
oyvind

[jackkerouac]

I'll back michael here because I've a parallel experience when compared to Micahel's. I'm using a X1650XT (1950Pro sibling) and I'm having a good experience with these drivers too. Yes. drivers may have bugs but they didn't hit neither him nor me during the development of this driver. So instead of shouting blindly please try to be a bit cooler, OK?

In no way am I 'shouting blindly' at all. I'm posting valid concerns about the information posted here. Information, I might add, that was either exaggerated, incomplete or flat out wrong.

I also run the X1650 and this driver crashes my X when loaded. Perhaps Michael should have tested it a little better/longer before going on about how great the improvements were, that's all I'm saying.

I know that next time Michael goes on and on and on (at least four posts worth) about how great a new driver is going to be, I'm going to be MUCH less likely to take him at face value.

[spielc]

Okey guys i think it's time to stop to accuse each other about who said what, where and when...

However there are some facts that should be stated clearly:

1. In general the driver does seem to show performance improvements
2. XV- and Textured Video-Playback (and thus more or less video playback under linux in general) seem to be still broken in quite alot of cases, so absolutely no improvement there! This is something ATI HAVE to work on FAST, as this is something almost everybody needs/wants.
3. I personally think that the development is going in the right direction but i still think it's not the revolution we were promised to see...

[lenrek]

Well, I took out my old HP laptop that comes with Xpress200M and give the new driver a little spin.

I am using Slack, so just simply issue command line options as '--buildpkg Slackware/All'. Built went smoothly and successfully create two new packages, I do:


$> upgradepkg --install-new fglrx-module-8.41.7-x86-1_kernel_2.6.22.6.tgz
$> upgradepkg --install-new fglrx-x710-8.41.7-x86-1.tgz
$> reboot


After boot up, dmesg output shows:

[fglrx] Maximum main memory to use for locked dma buffers: 1778 MBytes.
[fglrx] GART Table is not in FRAME_BUFFER range
[fglrx] Reserve Block - 0 offset = 0X0 length = 0X40000
[fglrx] Reserve Block - 1 offset = 0X7ff5000 length = 0Xb000
[fglrx] Not implemented, need to call CMM to get the mem info


These seems new, it looks very different from previous driver printout. However, I notice a line stating: GART Table is not in FRAME_BUFFER range - Should I be concern about this?

From Xorg.0.log:

$> grep ATI /var/log/Xorg.0.log
(--) PCI:*(1:5:0) ATI Technologies Inc ATI Radeon XPRESS 200M 5955 (PCIE) rev 0, Mem @ 0xc8000000/27, 0xc0100000/16, I/O @ 0x9000/8
(II) Module fglrx: vendor="FireGL - ATI Technologies Inc."
(II) ATI Proprietary Linux Driver Version Identifier:8.41.7
(II) ATI Proprietary Linux Driver Release Identifier: UNSUPPORTED-8.413.1
(II) ATI Proprietary Linux Driver Build Date: Sep 7 2007 22:35:20
(--) fglrx(0): Chipset: "ATI Radeon Xpress Series" (Chipset = 0x5955)
(--) fglrx(0): board vendor info: third party graphics adapter - NOT original ATI
(II) fglrx(0): VESA VBE OEM: ATI MOBILITY RADEON Xpress 200G Series
(II) fglrx(0): VESA VBE OEM Vendor: ATI Technologies Inc.
(II) Module fglrxdrm: vendor="FireGL - ATI Technologies Inc."
(**) fglrx(0): ATI GART size: 256 MB
(II) Loading extension ATIFGLRXDRI
(II) fglrx(0): Desc: ATI FireGL DRM kernel module
(II) Loading extension ATITVOUT

$> grep fglrx Xorg.0.log | grep Kernel
(**) fglrx(0): Option "KernelModuleParm" "locked-userpages=0"
(**) fglrx(0): KernelModuleParm: "locked-userpages=0"
(II) fglrx(0): Kernel Module Version Information:
(II) fglrx(0): Kernel Module version matches driver.
(II) fglrx(0): Kernel Module Build Time Information:
(II) fglrx(0): Build-Kernel UTS_RELEASE: 2.6.22.6
(II) fglrx(0): Build-Kernel MODVERSIONS: yes
(II) fglrx(0): Build-Kernel __SMP__: yes
(II) fglrx(0): Build-Kernel PAGE_SIZE: 0x1000

$> grep fglrx Xorg.0.log | grep drm
(II) Loading sub module "fglrxdrm"
(II) LoadModule: "fglrxdrm"
(II) Loading /usr/lib/xorg/modules/linux//libfglrxdrm.so
(II) Module fglrxdrm: vendor="FireGL - ATI Technologies Inc."
(II) Loading sub module "fglrxdrm"
(II) LoadModule: "fglrxdrm"
(II) Reloading /usr/lib/xorg/modules/linux//libfglrxdrm.so
(II) fglrx(0): [drm] DRM buffer queue setup: nbufs = 100 bufsize = 65536
(II) fglrx(0): [drm] DRM interface version 1.0
(II) fglrx(0): [drm] created "fglrx" driver at busid "PCI:1:5:0"
(II) fglrx(0): [drm] added 8192 byte SAREA at 0x2000
(II) fglrx(0): [drm] mapped SAREA 0x2000 to 0xb756e000
(II) fglrx(0): [drm] framebuffer handle = 0x3000
(II) fglrx(0): [drm] added 1 reserved context for kernel
(II) fglrx(0): [drm] register handle = 0x00004000

$> grep Compo /var/log/Xorg.0.log
(**) Extension "Composite" is enabled

$> grep AIGLX /var/log/Xorg.0.log
(==) AIGLX enabled
(EE) AIGLX error: dlopen of /usr/lib/xorg/modules/dri/fglrx_dri.so failed (/usr/lib/xorg/modules/dri/fglrx_dri.so: invalid ELF header)
(EE) AIGLX: reverting to software rendering



Well, AIGLX is not working. Is OK, I am expecting this.

Now interesting part.


(8.40)
$> glxgear
3426 frames in 5.0 seconds = 685.044 FPS
3474 frames in 5.0 seconds = 694.785 FPS
3471 frames in 5.0 seconds = 694.120 FPS

(8.41)
$> glxgear
4086 frames in 5.0 seconds = 817.020 FPS
4082 frames in 5.0 seconds = 816.221 FPS
4085 frames in 5.0 seconds = 816.977 FPS


On average, I get about 125 FPS more!


(8.40)
$> fglx_glxgear
...
609 frames in 5.0 seconds = 121.800 FPS
605 frames in 5.0 seconds = 121.000 FPS
621 frames in 5.0 seconds = 124.200 FPS

(8.41)
$> fglx_glxgear
...
723 frames in 5.0 seconds = 144.600 FPS
814 frames in 5.0 seconds = 162.800 FPS
798 frames in 5.0 seconds = 159.600 FPS


On average, I get about 33 FPS more!

This is indeed a great release! :D

EDIT: I haven't got the time to play around the xorg.conf file yet. Maybe I can tweak it more... :D

[rbmorse]

Oh by the way, try adding these I know these probably won't make a difference but put these at the end of your xorg.conf and try starting x, if it fails which i'm guessing it will, let me see your log again. There might be more answers then. Just maybe.

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

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

Adding the serverflags section to Xorg.conf cleared the software rendering error, so as of this moment I have to say the new driver is working very well for me.

[rbmorse]

Also managed to get 8.41.7 working under Ubuntu 7.10 (tribe-5) where it had previously failed. I didn't know that a system reset was required after blacklisting the "fglrx" module in Ubuntu's restricted module manager.

Now works fine, no apparent issues.

[phreadom]

Could people start listing what specific cards they are having successes or failures with?

That way we can get a better idea which specific cards might be causing trouble for the driver.

Thanks!

[SheeEttin]

Figured I might as well post my experience...
I have an ATI Radeon X1600 chip on an ASUS Radeon X1600 Pro board.
When I installed the new driver and rebooted, everything came up fine. Didn't notice any difference, of course. So, I fired up Sauerbraten and turned up all the display settings (shader level 3, water reflections, refractions, and waves, etc.). Worked better than the last driver, until about the third map I loaded, when textures began coming up black. Not only the geometry textures (i.e. world), but the GUI and HUD textures were also black now. (The little screenshot of a map displayed as it is loading was also black.)
I don't remember if I tried UT2004 or not, but I believe I did and it worked fine.

I went back to the 8.40.4 driver.

[phreadom]

Figured I might as well post my experience...
I have an ATI Radeon X1600 chip on an ASUS Radeon X1600 Pro board.
When I installed the new driver and rebooted, everything came up fine. Didn't notice any difference, of course. So, I fired up Sauerbraten and turned up all the display settings (shader level 3, water reflections, refractions, and waves, etc.). Worked better than the last driver, until about the third map I loaded, when textures began coming up black. Not only the geometry textures (i.e. world), but the GUI and HUD textures were also black now. (The little screenshot of a map displayed as it is loading was also black.)
I don't remember if I tried UT2004 or not, but I believe I did and it worked fine.

I went back to the 8.40.4 driver.

PCIE? People seem to be having problems with the AGP versions of the X1600Pro.

[SheeEttin]

PCIE? People seem to be having problems with the AGP versions of the X1600Pro.

Yeah, PCIe.

[ivanovic]

In short:
It looks like all cards that were originally designed for PCIe and use a bridge chip to run on AGP do not work.

My 9800pro AGP runs perfectly fine on my Cheap Asrock Conroe865 board. It works nicely with the new 8.41.7 driver in UT2004 (1600x1200, maximum quality) and in Neverball (1600x1200, max quality) on Gentoo with a vanilla kernel 2.6.22.6, xorg 7.2 (xorg-server 1.3.0).

[Svartalf]

In short:
It looks like all cards that were originally designed for PCIe and use a bridge chip to run on AGP do not work.

My 9800pro AGP runs perfectly fine on my Cheap Asrock Conroe865 board. It works nicely with the new 8.41.7 driver in UT2004 (1600x1200, maximum quality) and in Neverball (1600x1200, max quality) on Gentoo with a vanilla kernel 2.6.22.6, xorg 7.2 (xorg-server 1.3.0).

Well, that's something. It's not QUITE as bad as it seems to be. Still, it's quite disappointing- the driver speed boost is impressive, when you get the drivers to work (I've yet to attempt it on the spare laptop and I'm not in a position to sacrifice any of my other machines yet...). Guess we'll just have to wait and see. AMD's not off of the "DO NOT BUY" list yet...

[jjjj]

Hello,
I'm too using Mobi X1600 and suffering from the same bugs (1. & 2.).
I also noticed that even you can't change the AA settings using the amdcccle if you had previously configured 4XAA the setting is still honered! So I had to manually change from FSAAEnable=Son to FSAAEnable=Soff in /etc/ati/amdpcsdb . With FSAAEnable=Son I experienced crashing of 3D apps.


Hi All,

I've had mostly success with x1600 mobility. Desktop features/bugs/performance are pretty much the same, and OpenGL performance -- as noted -- is significantly better. It seems that the "new driver" is more correctly a "new OpenGL subsystem".

In any case, can anyone help me with the following:
1. I can't change any game display settings from within the game. This has occurred with two Q3 based games. Any resolution or other settings change causes a crash.
2. My brightness in each game is really, really low. I can't seem to change it -- the ingame sliders for gamma don't function. I can set my color gamma using amdcccle on the desktop, which fixes this a bit, but the colors are very wonky in that case.

Thanks,

L.

[gradbert]

I have an RV350 [Mobility Radeon 9600 M10]. The driver seemed to work ok in single screen mode, but in dual screen mode all the open GL programs are messes up. When running a program such as glx gears it looks like it is trying to display everything twice as wide as it should.

[jackkerouac]

Mine's an X1650 Pro AGP and X dies after the install. Yet another AMD/ATI failure.

[mjordan]

Can anyone help me out? I'm at it again today, trying to get this thing to work... I get this when trying to run

sudo sh ./**drivername** --buildpkg Ubuntu/feisty

dh_compress
dh_makeshlibs
dh_installdeb
LD_PRELOAD= dh_shlibdeps
dpkg: /usr/x86_64-linux-gnu/lib64/libc.so.6 not found.
dpkg: /usr/x86_64-linux-gnu/lib64/libstdc++.so.5 not found.
dpkg: /usr/x86_64-linux-gnu/lib64/libpthread.so.0 not found.
dpkg: /usr/x86_64-linux-gnu/lib64/librt.so.1 not found.
dpkg: /usr/x86_64-linux-gnu/lib64/libdl.so.2 not found.
dpkg: /usr/x86_64-linux-gnu/lib64/libm.so.6 not found.
dpkg: /usr/x86_64-linux-gnu/lib64/libgcc_s.so.1 not found.
dpkg: /usr/x86_64-linux-gnu/lib64/libXext.so.6 not found.
dpkg: /usr/x86_64-linux-gnu/lib64/libX11.so.6 not found.
dpkg: /usr/x86_64-linux-gnu/lib/librt.so.1 not found.
dpkg: /usr/x86_64-linux-gnu/lib/libdl.so.2 not found.
dpkg: /usr/x86_64-linux-gnu/lib/libm.so.6 not found.
dpkg: /usr/x86_64-linux-gnu/lib/libpthread.so.0 not found.
dpkg: /usr/x86_64-linux-gnu/lib/libXext.so.6 not found.
dpkg: /usr/x86_64-linux-gnu/lib/libX11.so.6 not found.
dpkg: /usr/x86_64-linux-gnu/lib/libc.so.6 not found.
dpkg: /usr/x86_64-linux-gnu/lib64/libXrandr.so.2 not found.
dpkg: /usr/x86_64-linux-gnu/lib64/libXrender.so.1 not found.
dpkg: /usr/x86_64-linux-gnu/lib64/libfglrx_gamma.so.1 not found.
dpkg: /usr/x86_64-linux-gnu/lib64/libGL.so.1 not found.
dpkg-shlibdeps: failure: dpkg --search gave error exit status 1
dh_shlibdeps: command returned error code 256
make: *** [binary] Error 1
make: Leaving directory `/tmp/fglrx.jI6221'
Removing temporary directory: fglrx-install.Zp6141

[djdoo]

@ mjordan:

You reported earlier that you get a black screen with the new driver when load X...

It is really weird but I get this bug always with fglrx drivers and Xorg 7.2 only with one screen resolution: 1024x768!!!

I saw that your xorg.conf contains as last "mode" the 1024x768
Try add there the mode 1280x1024 and see what happens...
I don't own an ATI card but I take care of a friend's PC which uses an 9550 radeon and I am facing what you describe always with ATI drivers and 7.2 X.

[djdoo]

Now for the topic...

I am an nvidia user who finds nvidia not as good as many folks think...
The fact that their driver can use straight Composite doesn't make it better because it has 2 great blocker bugs for alot of people. And many folks still use XGL on nvidia...

I apreciatte now AMD far more than nvidia(of course we still cannot compare features and goodies overall like SLI) not for the "mysterious" new 8.41 driver but because they gave out the cards' specs and will provide more in the future and that's what counts for me... That's effective community support(at least a very promising start) not just a forum.

Cause here we do not talk about just performance added and bugs added or not solved in the new driver, we are talking about a completely new approach at the community by a really great community "enemy" company.

AMD has really changed the way those guys think about us and was obligated to do that not only cause Linux community grows rapidly but also cause Linux was the only OS supporting 64bit arch, and I think still is.;)

I think the new driver was a bit overboosted here but I blame noone for that cause I can feel every ATI's user agony about the new code and driver.

But guys in my opinion AMD now starts entering the game for good and we should concentrate now at the new heart the driver has which can give such 3D boost stuff and new hope obviously!
And if a parallel good Open driver excists both of the drivers can take advantage of the experience and development of the 2 projects.

So for me the important thing now is the 360 degrees change of the approach to the community by AMD!

Hope that Compiz-Fusion will be an one month wait for our ATI friends! And after that I could also be more helpful than just make comments and thoughts..heh!

Jim

[lenrek]

Hello,
I'm too using Mobi X1600 and suffering from the same bugs (1. & 2.).
I also noticed that even you can't change the AA settings using the amdcccle if you had previously configured 4XAA the setting is still honered! So I had to manually change from FSAAEnable=Son to FSAAEnable=Soff in /etc/ati/amdpcsdb . With FSAAEnable=Son I experienced crashing of 3D apps.

Interesting, this mean, after installation, we should check if this file exist and make the require changes. Do we really need this file? For me, I do not have this file in my system.

My system works fine so far, no screen corruption. Xv works fine. I don't any dual screen or TV output to test, I don't use it anyway...

EDIT: I need to get back to work now, so is off my HP laptop and on my new Dell laptop (is an Intel 965GM). Good luck to the rest of the guys here.

[Malikith]

Now for the topic...

I am an nvidia user who finds nvidia not as good as many folks think...
The fact that their driver can use straight Composite doesn't make it better because it has 2 great blocker bugs for alot of people. And many folks still use XGL on nvidia...

I apreciatte now AMD far more than nvidia(of course we still cannot compare features and goodies overall like SLI) not for the "mysterious" new 8.41 driver but because they gave out the cards' specs and will provide more in the future and that's what counts for me... That's effective community support(at least a very promising start) not just a forum.

Cause here we do not talk about just performance added and bugs added or not solved in the new driver, we are talking about a completely new approach at the community by a really great community "enemy" company.

AMD has really changed the way those guys think about us and was obligated to do that not only cause Linux community grows rapidly but also cause Linux was the only OS supporting 64bit arch, and I think still is.;)

I think the new driver was a bit overboosted here but I blame noone for that cause I can feel every ATI's user agony about the new code and driver.

But guys in my opinion AMD now starts entering the game for good and we should concentrate now at the new heart the driver has which can give such 3D boost stuff and new hope obviously!
And if a parallel good Open driver excists both of the drivers can take advantage of the experience and development of the 2 projects.

So for me the important thing now is the 360 degrees change of the approach to the community by AMD!

Hope that Compiz-Fusion will be an one month wait for our ATI friends! And after that I could also be more helpful than just make comments and thoughts..heh!

Jim

Yeah I agree, however, its not that AMD didn't care, its more like ATI didn't care. Seems like AMD is fixing it all up and cracking the whip on ATI. AMD has always been pretty supportive of Linux, just not ATI.

[Alistair]

Please -- if you have one of these cards, and this issue - add your note (kernel/rel distribution and hardware type) to bug 780 on the ati unofficially bugzilla --

This page -->http://ati.cchtml.com/show_bug.cgi?id=780

If we get more entries there they will have to acknowledge the issue - and we can hope they get it fixed...

More entries -- more likely they chase it

<< edit -> FWIW I have yet to try out my X1650Pro AGP 512Mb on this driver - this I will do on saturday - RL and Work and whatnot being what it is.... I DO have this card working with DRI - - Just NO AGP speed being set card runs in base pci mode ... if X is locking up HARD for you with this card, Increase your agp gart mem size in bios as large as you can, preferably = Memory on card - this resolved the hardlocks for me ..... >>

[Alistair]

Yeah and if anyone wants to try and help me figure out my issue, which I doubt, I'll probably have to wait until 8.42, check out my Xorg.0.log:

My Xorg log (http://bloodytampon.servegame.com/downloads/Xorg.0.log)

If you have any ideas, feel free to suggest.

Just so you know, 8.40 does work.

Looks like you might have composite kicking in there - have you tried the disable for it?

[Malikith]

Looks like you might have composite kicking in there - have you tried the disable for it?

Yep sure have, heres my xorg.conf:

My xorg.conf (http://bloodytampon.servegame.com/downloads/xorg.conf1)

[mjordan]

dh_compress
dh_makeshlibs
dh_installdeb
LD_PRELOAD= dh_shlibdeps
dpkg: /usr/x86_64-linux-gnu/lib64/libc.so.6 not found.
dpkg: /usr/x86_64-linux-gnu/lib64/libstdc++.so.5 not found.
dpkg: /usr/x86_64-linux-gnu/lib64/libpthread.so.0 not found.
dpkg: /usr/x86_64-linux-gnu/lib64/librt.so.1 not found.
dpkg: /usr/x86_64-linux-gnu/lib64/libdl.so.2 not found.
dpkg: /usr/x86_64-linux-gnu/lib64/libm.so.6 not found.
dpkg: /usr/x86_64-linux-gnu/lib64/libgcc_s.so.1 not found.
dpkg: /usr/x86_64-linux-gnu/lib64/libXext.so.6 not found.
dpkg: /usr/x86_64-linux-gnu/lib64/libX11.so.6 not found.
dpkg: /usr/x86_64-linux-gnu/lib/librt.so.1 not found.
dpkg: /usr/x86_64-linux-gnu/lib/libdl.so.2 not found.
dpkg: /usr/x86_64-linux-gnu/lib/libm.so.6 not found.
dpkg: /usr/x86_64-linux-gnu/lib/libpthread.so.0 not found.
dpkg: /usr/x86_64-linux-gnu/lib/libXext.so.6 not found.
dpkg: /usr/x86_64-linux-gnu/lib/libX11.so.6 not found.
dpkg: /usr/x86_64-linux-gnu/lib/libc.so.6 not found.
dpkg: /usr/x86_64-linux-gnu/lib64/libXrandr.so.2 not found.
dpkg: /usr/x86_64-linux-gnu/lib64/libXrender.so.1 not found.
dpkg: /usr/x86_64-linux-gnu/lib64/libfglrx_gamma.so.1 not found.
dpkg: /usr/x86_64-linux-gnu/lib64/libGL.so.1 not found.
dpkg-shlibdeps: failure: dpkg --search gave error exit status 1
dh_shlibdeps: command returned error code 256
make: *** [binary] Error 1
make: Leaving directory `/tmp/fglrx.jI6221'
Removing temporary directory: fglrx-install.Zp6141

@ mjordan:

You reported earlier that you get a black screen with the new driver when load X...

Thanks... I'll try that when I get this compiling.... it wont' recognize my so.6, so.1 etc files... I've tried making links to the directories and even to individual files but no dice. All the files it can't find are actually located in /usr/lib and /usr/lib64, (which are both linked into /usr/x86_64-linux-gnu/ now as well). I can't figure this out...

If noone knows how to fix it can someone upload some pre-built Ubuntu/feisty .deb files?


Thanks

[adamk]

Mine's an X1650 Pro AGP and X dies after the install. Yet another AMD/ATI failure.

You downloaded and installed a driver that specifically wasn't designed for your video card. Sounds like a jackkerouac failure, to me.

Adam

[Kano]

As only Fire* ids are missing - I expect because of incomplete OpenGL functions (interesting post how to patch it) - the mentioned restriction to 2xxx series is an excuse for the the driver status, they only want that you use it with latest cards, but basically it is not mandatory. The wine problems can be solved with updated wine (get git), others usually not. But it definitely shows that the speed is better than before for some games - when the card is running. The "new codebase" can not be the whole driver, too many things have not changed, I think only the real OpenGL part was updated, not the rest - I would have called it updated OpenGL core.

[robotpoetry]

I got this working on Ubuntu Feisty with a Radeon 9600 by:
1. Turning off the ati driver in restricted drivers manager and rebooting.
2. Following these instructions (not using aticonfig): http://wiki.cchtml.com/index.php/Ubuntu_Feisty_Installation_Guide
3. Turning on the ati driver in restricted drivers manager again.

As others have reported, there is a great boost in opengl performance!

However, there are some bugs. Nexuiz segfaults if you try to change the video settings from within the game (change the config file manually instead). Apart from that it works well and is clearly faster.

Neverball hanged after a short while.

Simple apps like fgl_glxgears and the GL screensavers in Gnome work well. Also, some minor visual corruption of the screen can be seen occasionally.

Xv works. I haven't tried TV-out or dual head as I don't use that.

Looking forward to the next driver release.

More importantly, I (like everyone else) am very happy with AMD's new policy of providing specs and helping with a free software driver!

[hobbes]

Please -- if you have one of these cards, and this issue - add your note (kernel/rel distribution and hardware type) to bug 780 on the ati unofficially bugzilla --

This page -->http://ati.cchtml.com/show_bug.cgi?id=780

If we get more entries there they will have to acknowledge the issue - and we can hope they get it fixed...

More entries -- more likely they chase it

<< edit -> FWIW I have yet to try out my X1650Pro AGP 512Mb on this driver - this I will do on saturday - RL and Work and whatnot being what it is.... I DO have this card working with DRI - - Just NO AGP speed being set card runs in base pci mode ... if X is locking up HARD for you with this card, Increase your agp gart mem size in bios as large as you can, preferably = Memory on card - this resolved the hardlocks for me ..... >>

i have just discovered that issue! I was unaware of that until late night today when i tried to play Quake 4 on my Ubuntu box and noticed the poor performance when better quality settings were applied.

I've also added one more entry to the AGP detected as PCIE bug report, let's hope they'll notice.

[ambro814]

As only Fire* ids are missing - I expect because of incomplete OpenGL functions
I must say OpenGL on my Thinkpad T60p (FireGL v5200) works quite well. I can now play GLTron at 'ultra-high' cycle detail without the game crashing and many games in Wine which previously didn't work now work, although some have performance problems. I tried IL-2, Flatout 2 and Guild Wars. IL-2 has some performance problems (previously it crashed while loading almost any level), however Flatout 2 is fully playable after I drop resolution to 1280x1024 (otherwise bad fps). Also Guild Wars works, but only if I add '-dx8 -noshaders', otherwise there is a ton of graphics glitches.

[Raven3x7]

I just installed the driver. Performance is greatly improved in games for my X1800XT although i havent done any specific benches. Nexuiz used to be hardly playable at 1280x1024 and now it runs smoothly at 1680x1050. After running a 3d app a weird white-green artifact appeared on the lower right corner of the screen though it doesnt appear in screenshots. Could this be the watermark issue some ppl had been experiencing previously?

[tsarig]

ATI are concerning all of their attention for their new cards (the HD series) rather than supporting their "older" ones.
they know their drivers suck and expect people to trash their older cards and buy the new ones.

I have a X1950 pro AGP card, which is quite new, and what a sad joke, my previous 9800 pro was outperforming my new beast.

Since I've got this card AGP is forced to X0 instead of X8 because it thinks it's a PCIE card because of the well known AGP to PCIE bridge chip linux issue.

What I want is that ATI WILL PAUSE MAKING DRIVERS and PAUSE THE SUPPORT for their HD cards, focus on the basic issues of the X1000 AGP series, and when all is "well", focus back on the HD's.

I hope that an ATI worker or executive is reading these lines, people PAID A LOT for these cards, they're good cards, please please please MAKE THEM WORK PROPERLY, we all had enough of this, for years. the X1000 series is not here from yesterday, it should be up and running great by now if you had more of your concern for your customers.

And don't tell me "linux is not supported, please install windows", your lastest cat 7.9 drivers for windows does not support my X1950 AGP, it's written in your release notes, "and it's a HD series driver only, please install cat 7.7".

And don't tell me "AGP is out dated, please buy a new PCIE based pc", your X1950 AGP series is less than a year old or so and still in production. and btw there is a new AGP versions of the HD series. will you let them down too?

Dear ATI, if you were a car dealer, you would let the car stand in your showrooms without spare parts and fuel.

And please stop playing around with people with promises like "yeah, the new 8.41 drivers are so great, preformance boost and etc" BECAUSE THEY'RE NOT WORKING.

My next card, for sure would be a green one.

BTW, Nvidia cards are always more expensive than yours,do you cut off the price because you know whom buy your cards will get frustrated??

Thank you for your time, hope this post will reach to ATI's windows some day..

[d2kx]

tsarig,

we had that.

AMD said "don't install it on R500 and under". You didn't hear. AMD said "8.41s focus is R600, 8.42 will focus on older cards". You didn't hear. You are talking about the driver not working. It works.

[Alistair]

i have just discovered that issue! I was unaware of that until late night today when i tried to play Quake 4 on my Ubuntu box and noticed the poor performance when better quality settings were applied.

I've also added one more entry to the AGP detected as PCIE bug report, let's hope they'll notice.


thanks Hobbes!

[Alistair]

ATI are concerning all of their attention for their new cards (the HD series) rather than supporting their "older" ones.
they know their drivers suck and expect people to trash their older cards and buy the new ones.

I have a X1950 pro AGP card, which is quite new, and what a sad joke, my previous 9800 pro was outperforming my new beast.

Sin


Please add your name to bug 780 on the ati unofficial bugzilla which documents the PCIE chips on AGP cards not enabling AGP modes.

http://ati.cchtml.com/show_bug.cgi?id=780

[oliver]

Please add your name to bug 780 on the ati unofficial bugzilla which documents the PCIE chips on AGP cards not enabling AGP modes.

http://ati.cchtml.com/show_bug.cgi?id=780

I keep hearing you say that, but Could you give me some pointers to where I can find more info on it?

[tsarig]

tsarig,

we had that.

AMD said "don't install it on R500 and under". You didn't hear. AMD said "8.41s focus is R600, 8.42 will focus on older cards". You didn't hear. You are talking about the driver not working. It works.

d2kx, let's say you bought a new card, spending xxx $ just to find out it's working at 1/8 of it's power, and the manufacturer just don't care about that, for months..

that's what I've ment to say, I don't care about not getting the latest drivers to work. good for you if you got it working, your're a lucky man.

besides, I'm not the only one who have troubles with the agp/pcie gate.

Originally Posted by Alistair:
Please add your name to bug 780 on the ati unofficial bugzilla which documents the PCIE chips on AGP cards not enabling AGP modes.

http://ati.cchtml.com/show_bug.cgi?id=780

and d2kx, people are posting here to help each other, if you rather lecture people than help them so dont post!

[Alistair]

You downloaded and installed a driver that specifically wasn't designed for your video card. Sounds like a jackkerouac failure, to me.

Adam

And Adam - I'm afraid you actually misunderstand the following alert :

Caution: The ATI Catalyst™ Linux software suite 8.41.7 does not support ATI Workstation products. AMD recommends using the AMD Proprietary Linux software driver version 8.40.4.

This warning is directed at FireGL workstation products which the X1650Pro is not.

Nor does the X1650Pro AGP fit in the

Caution: The AMD Proprietary Linux driver version 8.41.7 is not recommended for the ATI Mobility and Integrated Products. Using this driver on any of the following products may result in stability issues. AMD recommends using the AMD Proprietary Linux software driver version 8.40.4.

Warning -- it is not Mobility nor an Integrated chip


Final point - the last caveat on ATI's release notes:

Caution: The AMD Proprietary Linux driver version 8.41.7 is not recommended for use on ATI Radeon X1900 series of products and lower. Using this driver on any of the following products may result in stability issues. AMD recommends using the AMD Proprietary Linux driver version 8.40.4

Its clear that this is not a "does not support this card" - it is indicating that there are issues - only that.

[Kano]

Btw. everybody can try now the driver even in live mode with ease on Kanotix Thorhammer, just add

fglrx

as option. To use Xorg's detection for the monitor (usually better than the default), use

screen=auto

especially usefull for tft. Of course you can use

nvidia

as option too. You just need more time for fglrx as the dl is 50 mb compared to 10 mb... Wine 0.9.45 is included. To test older fglrx use something like:

fglrx=8.39.4

btw. it autoselects the correct one for any ati card, even dx8 ones. Nothing hurts your hd installs - of course you could install too...

[Alistair]

Mine's an X1650 Pro AGP and X dies after the install. Yet another AMD/ATI failure.

There are known issues with X1000 series cards, and I've found this card is problematic with ***all*** ati drivers - the issue (black screen on X start requiring a restart of X or on occaision a reboot) is resolved when setting the GART memory size in the BIOS = to or greater than the amount of ram on the video card.

I have one of these cards, and have gotten it to run X with DRI enabled, and glxgears and fgl_glxgears will run - however AGP mode 0 prevents any real 3d processing and even substantially slows 2d operations.

[Kano]

Well didn't you know that you have to buy a new board with PCI-E ;)

[Ralle]

When it says _64, does it mean my laptop (32) cannot use it?

[Kano]

That only means that ATI uses a combined package for both, NV uses two different ones.

[Alistair]

Well didn't you know that you have to buy a new board with PCI-E ;)

I'll get right on that -- new motherboard, ram, and I'll break out the soldering kit and put that video chip on a PCIe bus
:p

[wirechief]

I tried the 8.41.7 driver using Kanotix ThorHammer rc6 developed by Kano, it works fine with my Desktops ATI-1300 pro pcie, this card has 512mb of ram.
graphics card.
This is the first time that card has worked in Linux with ATI drivers, congratulations to the ATI-AMD team who finally got a driver to work with my card.

[Alistair]

Ookay folks -- had a few extra minutes this afternoon when I got home from work -- and finally got a chance to chase my issues with X1650Pro AGP card.

1) 8.34.8 fglrx ****WORKS****

my X1650Pro is up and running in AGP mode 8 with DRI with opengl with only some weirdness

I'm running gentoo-sources 2.6.22 - the driver fglrx will not compile with the default ebuild in gentoo tree - copy it to an overlay and add

this line epatch "${FILESDIR}"/8.37.6/fix-ioctl-for-2.6.22.patch

to the ebuild in the stanza that looks like

# This patch is conditional because it contains an api change.
if kernel_is ge 2 6 20; then
epatch packages/Fedora/ati-drivers-2.6.20.patch
fi


then ebuild {full path to the overlay ebuild file} digest
emerge -v ={fully qualified ebuild}

and reboot for good measure

NOW -- caveat on this

DO NOT SWITCH BACK TO CONSOLE after X starts. this locks my machine.

I will now walk the driver tree and see if I can spot the regression.

Please note that I've udpated the bug with the details.

[Alistair]

Okay -- walked through all the drivers in portage for my system with the X1650Pro --did some back to back testing (where the driver initialized things on the X1650Pro)

Keep in mind I'm on Gentoo - so I'm sticking to ebuilds that are available - with a few fixes since the ebuilds typically fit the kernel that was current at the time, I've made minor fixes to get the driver to build with my 2.6.22 kernel.

8.35.5 also enables the card in AGP mode 8

However both 8.35.5 and 8.34 driver provide horrendous performance under
load.
glxgears and fgl_fglrxgears both run slower under these drivers with the
X1650Pro than with the 9550

I have not been able to test on 8.36 {I was unable to get a 64 bit build to
complete}
8.37 enables AGP mode, but the DRI setup never completes and the X setup
does not complete nor do I get a display. (black screen lockup)
8.38 never made portage - if it exists.
8.39 fails to initialize AGP mode.
8.40 fails to initialize AGP mode.
8.41 fails to initialize AGP mode.

[Kano]

You can use Kanotix Thorhammer to test the drivers, much more easy using

install-fglrx-debian.sh -v VER

or even in live mode with

fglrx=v:VER

The script is fully tested with lots of drivers and Xorg 7.1 is old enough to test even drivers for DX8 cards. Of course you can use the hard way too ;)

[xeros]

8.41.7 works not so bad with my Sapphire Radeon 9600 256MB (not Pro) AGP on Kubuntu 7.04.
I have ~2450 FPS on glxgears and ~500 FPS on fgl_glxgears with Dual Head (LCD 1280x1024x24 + CRT 1280x1024x24 + TV 1024x768x24).
The only game that I've tried yet is Nexuiz 2.3.
It works great (from ~40 to ~90 FPS with only disabled Vertex Mapping and World Dynamic Lights - these two options degrade FPS, everything else enabled).
Game segfaults when I apply video changes in menu but no display corruption and other problems.
Xv works well (only ~3% of CPU use when playing movie with MPlayer)

I've managed to change display order with ForcedMonitors option in Device section but the driver doesn't react on this setting.
The drivers uses CRT1 as Display1, TV as Display2 and TMDS1 as Display3 and CRT1+TMDS1 works as clone and TV is the second head. I wanted to use TMDS1 as Display1 (one head) and CRT1 as Display2 with TV as Display3 in clone second head but the driver doesn't look for ForcedMonitors option.

[carpman]

8.41.7 works not so bad with my Sapphire Radeon 9600 256MB (not Pro) AGP on Kubuntu 7.04.

Thats weird. I can't get my Radeon 9800xt to work. I can get X to start, an glxgears works, but anything that uses textures comes back with an x display error. Also, and this is way out there, but my gnome session wouldn't work until I moved .gconf out of the way and let it build a new one. I though this problem was isolated to gnome, but when I went back to 8.40.6 I put my old .gconf back, and it worked. I was getting Connection reset by peer errors in my .xsession with my old .gconf and 8.41.7.

[Thetargos]

So it turns out my girlfriend bought a new laptop and the machine comes equipped with an Xpress 1100 IGP, anyone with a card like this tried the 8.41.7's? Or should we stick with the 8.40 for now?

[oliver]

Ookay folks -- had a few extra minutes this afternoon when I got home from work -- and finally got a chance to chase my issues with X1650Pro AGP card.

1) 8.34.8 fglrx ****WORKS****

my X1650Pro is up and running in AGP mode 8 with DRI with opengl with only some weirdness

I'm running gentoo-sources 2.6.22 - the driver fglrx will not compile with the default ebuild in gentoo tree - copy it to an overlay and add

this line epatch "${FILESDIR}"/8.37.6/fix-ioctl-for-2.6.22.patch

to the ebuild in the stanza that looks like

# This patch is conditional because it contains an api change.
if kernel_is ge 2 6 20; then
epatch packages/Fedora/ati-drivers-2.6.20.patch
fi


then ebuild {full path to the overlay ebuild file} digest
emerge -v ={fully qualified ebuild}

and reboot for good measure

NOW -- caveat on this

DO NOT SWITCH BACK TO CONSOLE after X starts. this locks my machine.

I will now walk the driver tree and see if I can spot the regression.

Please note that I've udpated the bug with the details.

The irony of this all is of course, that my X1950Pro AGP (with PCI-e bridge chip) only works with the 8.34 driver. Anything higher locks up my X session, i can ssh in, but as soon as I kill X my box hardlocks. So the only way i can reboot is to reset eitherway.

Looks like they did 'something' starting with 8.35 if you ask me.

[Svartalf]

So it turns out my girlfriend bought a new laptop and the machine comes equipped with an Xpress 1100 IGP, anyone with a card like this tried the 8.41.7's? Or should we stick with the 8.40 for now?

The brave of heart are free to try the thing- if you don't have some serious Linux expertise under your belt you might want to hold off on anything past 8.40 for fglrx drivers...

[Kano]

When you do it right, switching fglrx drivers is no big deal. Why don't you try the driver yourself?

[Svartalf]

When you do it right, switching fglrx drivers is no big deal. Why don't you try the driver yourself?

Heh... I plan on it- perhaps sometime this week/weekend if time permits to gain intel on the IGP series part behaviors with the codebase. Knowing something of things in this new driver, I'm sure there's some sharp edges on the IGP lineup like there's been with the AGP series of parts. Something that gives them reason to say it's NOT supported for most of the lineup.