You are not logged in.

#1 2006-10-14 02:05:06

Alkali
Member
From: United States
Registered: 2006-06-01
Posts: 10

Go ahead and say it, "Please not another ATI issue.&quo

So I'm posting here cause I went to the irc room and got no answer after 4 hours.  I have no dri acceleration unless I follow the advice of putting the modules in the order of:

agpgart intel_agp fglrx

You see when I put it in that order and that order only I believe dri is enabled at next boot.  But when I reboot the computer to see it through I boot to a blank screen.  Not like an out of sync range but it almost wants to load X and it sounds like it's loading X behind all the blankness but it just stays and freezes.  Every time.  No ctrl+alt+del or f1 or backspace will save me. I have to cut the power.  I looked up the advice in forums and on google and I did a cat /proc/mtrr and I don't have any bad lines.  And I've adjusted my xorg.conf countless times to get it to work and I'll post my current conf file below but still to no avail.  I'm using the ati-fglrx package provided and 2d accel works if I don't add intel_agp with the modules in rc.conf.  But as soon as I introduce intel to the module party I'm going where any college freshman goes after hes had too much.  Blacked out.  Anyway, if anyone has this problem just let me know how you got around it or if there is something I missed on google or in the forums just a let me know.  I appreciate any help. My specs will be below and when I can get back into arch again I will post my xorg.conf and or any logfile someone needs to see.  I'm limited to windows right now.

Specs: Dell XPS 3.0 ghz, 1 gig ram, radeon X800 Pro, ati-fglrx 8.29.6, arch 2.16.18, and can post anything else I am forgetting later.

Thanks all.


"If this ends with the both of us dying in a giant explosion killing a Reaper, just remember. I took the kill shot." - Garrus Vakarian

Offline

#2 2006-10-14 09:44:02

iphitus
Forum Fellow
From: Melbourne, Australia
Registered: 2004-10-09
Posts: 4,927

Re: Go ahead and say it, "Please not another ATI issue.&quo

add them to the MODULES=() line in /etc/rc.conf

James

Offline

#3 2006-10-14 14:43:17

Cerebral
Forum Fellow
From: Waterloo, ON, CA
Registered: 2005-04-08
Posts: 3,108
Website

Re: Go ahead and say it, "Please not another ATI issue.&quo

sounds like a silly question, but are you sure intel_agp is your agp driver?

Offline

#4 2006-10-14 15:01:40

iphitus
Forum Fellow
From: Melbourne, Australia
Registered: 2004-10-09
Posts: 4,927

Re: Go ahead and say it, "Please not another ATI issue.&quo

Cerebral wrote:

sounds like a silly question, but are you sure intel_agp is your agp driver?

kernel wouldnt let him load it if it wasnt.

Offline

#5 2006-10-15 23:13:25

Alkali
Member
From: United States
Registered: 2006-06-01
Posts: 10

Re: Go ahead and say it, "Please not another ATI issue.&quo

Sorry it took me so long to respond I was out of town for the day.  I'm almost 100% that intel_agp because my lsmod shows this:

Module                  Size  Used by
ipv6                  237472  12
eth1394                16772  0
snd_seq_oss            29440  0
snd_seq_midi_event      6528  1 snd_seq_oss
snd_seq                47184  4 snd_seq_oss,snd_seq_midi_event
snd_emu10k1           111808  0
snd_rawmidi            19488  1 snd_emu10k1
snd_ac97_codec         87456  1 snd_emu10k1
snd_ac97_bus            2432  1 snd_ac97_codec
snd_pcm_oss            39072  0
snd_pcm                67972  3 snd_emu10k1,snd_ac97_codec,snd_pcm_oss
snd_mixer_oss          14592  1 snd_pcm_oss
snd_seq_device          6924  4 snd_seq_oss,snd_seq,snd_emu10k1,snd_rawmidi
tsdev                   6336  0
rtc                    10292  0
snd_timer              19204  3 snd_seq,snd_emu10k1,snd_pcm
snd_page_alloc          7816  2 snd_emu10k1,snd_pcm
snd_util_mem            3840  1 snd_emu10k1
snd_hwdep               7300  1 snd_emu10k1
snd                    42724  11 snd_seq_oss,snd_seq,snd_emu10k1,snd_rawmidi,snd_ac97_codec,snd_pcm_oss,snd_pcm,snd_mixer_oss,snd_seq_device,snd_timer,snd_hwdep
soundcore               7776  1 snd
usbhid                 47776  0
ppdev                   7556  0
lp                      9604  0
ppp_generic            24084  0
slhc                    6656  1 ppp_generic
e1000                 114496  0
ohci1394               31408  0
ieee1394               83672  2 eth1394,ohci1394
psmouse                35720  0
i2c_i801                7308  0
i2c_core               17280  1 i2c_i801
serio_raw               5636  0
intel_agp              19740  0
i82875p_edac            4996  0
edac_mc                18904  1 i82875p_edac
parport_pc             35684  1
parport                31432  3 ppdev,lp,parport_pc
intel_rng               2688  0
pcspkr                  2816  0
shpchp                 33444  0
pci_hotplug            28616  1 shpchp
uhci_hcd               21004  0
ehci_hcd               29576  0
evdev                   7936  0
fglrx                 393836  0
agpgart                25944  2 intel_agp,fglrx
usbcore               110852  4 usbhid,uhci_hcd,ehci_hcd
reiserfs              236544  2
ide_cd                 35616  0
cdrom                  34464  1 ide_cd
ide_disk               13440  5
generic                 4996  0 [permanent]
ata_piix               10760  0
libata                 85908  1 ata_piix
piix                    9220  0 [permanent]
ide_core              109256  4 ide_cd,ide_disk,generic,piix

Just so it's clear I did make sure I loaded them in order in the /etc/rc.conf
because its only when I force them to load in that order does it go to a black screen.  I'd be more than happy to assist with logfiles but probably the most helpful logfile will be the one I cannot access because it freezes.  But if anyone knows a workaround to get to the logs for the frozen session I'd really be grateful.  It'd give me something new to google  big_smile and I could post the output here.  If anyone knows feel free to let me know.  I'm in town now so I can check and respond to posts anytime.  Thanks for the responses guys.  This seems like a unique problem and I have sent ATI a bug report.  It might be fixed next release it might not.  I'd prefer using having 3D accel cause I'm picky like that.  If tweaking is wrong I don't want to be right.  Thanks guys.


"If this ends with the both of us dying in a giant explosion killing a Reaper, just remember. I took the kill shot." - Garrus Vakarian

Offline

#6 2006-10-16 03:52:49

projektdotnet
Member
Registered: 2006-10-06
Posts: 10

Re: Go ahead and say it, "Please not another ATI issue.&quo

iphitus wrote:
Cerebral wrote:

sounds like a silly question, but are you sure intel_agp is your agp driver?

kernel wouldnt let him load it if it wasnt.

Don't be 100% on that. My kernel let me load via_agp because I have a VIA chipset in most places but I REALLY needed amd64_agp...struck me as odd for sure.

Offline

#7 2006-10-16 04:49:10

Alkali
Member
From: United States
Registered: 2006-06-01
Posts: 10

Re: Go ahead and say it, "Please not another ATI issue.&quo

So do you recommend simply blacklisting intel_agp in the purpose of this experiment or do I change the order to:

agpgart via(ati)_agp fglrx

or leave intel_agp alone?


"If this ends with the both of us dying in a giant explosion killing a Reaper, just remember. I took the kill shot." - Garrus Vakarian

Offline

#8 2006-10-18 06:20:59

Alkali
Member
From: United States
Registered: 2006-06-01
Posts: 10

Re: Go ahead and say it, "Please not another ATI issue.&quo

No bueno  on trying to use ati_agp or via_agp in the modules line of rc.conf instead of intel_agp.  Any other suggestions as to what to try?


"If this ends with the both of us dying in a giant explosion killing a Reaper, just remember. I took the kill shot." - Garrus Vakarian

Offline

#9 2006-10-18 15:59:34

Cerebral
Forum Fellow
From: Waterloo, ON, CA
Registered: 2005-04-08
Posts: 3,108
Website

Re: Go ahead and say it, "Please not another ATI issue.&quo

If you can boot into console and get us /var/log/Xorg.0.log it might help out... otherwise I'm drawing a blank. sad

Offline

#10 2006-10-19 04:35:54

Alkali
Member
From: United States
Registered: 2006-06-01
Posts: 10

Re: Go ahead and say it, "Please not another ATI issue.&quo

Solved. Magically nonetheless.  I fixed the rc.conf back to the intel_agp agpgart fglrx method so the computer would freeze and a log file would be made but by some miracle it booted into kdm.  I was still hesitant to know if fglrx was working but once I added the dri 0666 mode I was 3D accelerated.  I don't know if it was the developers or you guys or maybe even a small miracle but I'm golden now.  Thank you all so much for your time.


"If this ends with the both of us dying in a giant explosion killing a Reaper, just remember. I took the kill shot." - Garrus Vakarian

Offline

#11 2006-10-19 14:32:17

Cerebral
Forum Fellow
From: Waterloo, ON, CA
Registered: 2005-04-08
Posts: 3,108
Website

Re: Go ahead and say it, "Please not another ATI issue.&quo

Congrats - always nice to have something miraculously work. wink

Offline

#12 2006-10-20 12:27:47

Alkali
Member
From: United States
Registered: 2006-06-01
Posts: 10

Re: Go ahead and say it, "Please not another ATI issue.&quo

One quick question.  Why is it that if I end the session with my user dri stops working when I log back in?  Just curious.  Using KDE and I logout to apply setting and I come back and dri is disabled.  If I reboot all is well.  I'm just curious is all.


"If this ends with the both of us dying in a giant explosion killing a Reaper, just remember. I took the kill shot." - Garrus Vakarian

Offline

#13 2006-10-20 13:46:15

Cerebral
Forum Fellow
From: Waterloo, ON, CA
Registered: 2005-04-08
Posts: 3,108
Website

Re: Go ahead and say it, "Please not another ATI issue.&quo

That is odd.  I don't know why it's doing that to you...

Offline

Board footer

Powered by FluxBB