You are not logged in.
@Vi0l0
Have you tried with a new Steam account?
If you can't sit by a cozy fire with your code in hand enjoying its simplicity and clarity, it needs more work. --Carlos Torres
Offline
Trololol! Seriously? xD
Edit: btw sounds like repairing Windows (via format c: ofc) ;P...
Last edited by Vi0L0 (2013-01-31 16:42:32)
Offline
No man, just messing with you TF2 actually runs fine here, it's just that my CPU overheats so my laptop shuts down
If you can't sit by a cozy fire with your code in hand enjoying its simplicity and clarity, it needs more work. --Carlos Torres
Offline
if it's really translating directx to opengl then in the day when i will run ie. Crysis 3 from Steam on Linux also my CPU will burn (in nanoseconds)...
Edit: just to mention: latelly i bought i7-3770K, edit2: hmm... wondering if i could somehow perform pxp-like switch (got something like Virtu MVP on MB)... hmmm
Edit3: steam is not only buggy, but also hard to debug, even when running `steam -dev` it's not telling a thing about the bug, just like saying "everything is fine, just wait for eternity, maybe i will launch it someday"
Last edited by Vi0L0 (2013-01-31 19:25:10)
Offline
No doubt this has been mentioned in this thread, but is the AMD Catalyst driver for Linux expected to give lower 3D performance than its Windows equivalent? Because I am getting lower performance with it than I do on Windows, by a long way. ATI Radeon HD 5670 with 1 GB video RAM, btw.
I'm not exactly complaining, I'm just curious as to whether others experience this.
Last edited by Archimaredes (2013-02-01 22:44:36)
Offline
No doubt this has been mentioned in this thread, but is the AMD Catalyst driver for Linux expected to give lower 3D performance than its Windows equivalent? Because I am getting lower performance with it than I do on Windows, by a long way. ATI Radeon HD 5670 with 1 GB video RAM, btw.
I'm not exactly complaining, I'm just curious as to whether others experience this.
I guess, you are comparing the Direct3D driver part in Windows with the OpenGL driver part in Arch Linux. If so, then yes, the driver in Windows should be better as expected. If you compare the Windows OpenGL with the Linux OpenGL, then there shouldn't be a significant difference (there was none the last time I checked).
Also which games/programs are you comparing in Windows and Linux, because those Windows games, that use the Direct3D --> OpenGL redirection method (like the Valve games and many Humble Indie Bundle Ports) will be slower in Linux, because the redirection costs more CPU power.
And before you quote the Valve Left4Dead2 is faster in Linux thing: That is more PR than real life scenario. They had such a powerful pc for that test, so that they won't see the bottlenecks.
Offline
Archimaredes wrote:No doubt this has been mentioned in this thread, but is the AMD Catalyst driver for Linux expected to give lower 3D performance than its Windows equivalent? Because I am getting lower performance with it than I do on Windows, by a long way. ATI Radeon HD 5670 with 1 GB video RAM, btw.
I'm not exactly complaining, I'm just curious as to whether others experience this.
I guess, you are comparing the Direct3D driver part in Windows with the OpenGL driver part in Arch Linux. If so, then yes, the driver in Windows should be better as expected. If you compare the Windows OpenGL with the Linux OpenGL, then there shouldn't be a significant difference (there was none the last time I checked).
Also which games/programs are you comparing in Windows and Linux, because those Windows games, that use the Direct3D --> OpenGL redirection method (like the Valve games and many Humble Indie Bundle Ports) will be slower in Linux, because the redirection costs more CPU power.
And before you quote the Valve Left4Dead2 is faster in Linux thing: That is more PR than real life scenario. They had such a powerful pc for that test, so that they won't see the bottlenecks.
Well, I've tested Half-Life, Minecraft and Team Fortress 2. Half-Life and Minecraft aren't exactly intensive so I spotted little difference (but the difference is there), but TF2 struggles.
Offline
but TF2 struggles.
With the 13.2 beta 3 driver?
Offline
I have to disable VariBright every time I reboot. Does anyone else have the same problem?
Also, is there a command line way of disabling it? That would make this bug easier to deal with, at least.
Offline
I have to disable VariBright every time I reboot. Does anyone else have the same problem?
Also, is there a command line way of disabling it? That would make this bug easier to deal with, at least.
Maybe setting PP_VariBrightFeatureEnable variable to PP_VariBrightFeatureEnable=V0 in /etc/ati/amdpcsdb _and_ /etc/ati/amdpcsdb.default will help?
Offline
thebluesgnr wrote:I have to disable VariBright every time I reboot. Does anyone else have the same problem?
Also, is there a command line way of disabling it? That would make this bug easier to deal with, at least.
Maybe setting PP_VariBrightFeatureEnable variable to PP_VariBrightFeatureEnable=V0 in /etc/ati/amdpcsdb _and_ /etc/ati/amdpcsdb.default will help?
Thanks for the advice.
I edited all occurrences in both files but still it seems that VariBright gets reenabled after a reboot.
Offline
Looks like ascarpino deleted both catalyst-utils and lib32-catalyst-utils from AUR and as a reason he wrote that it's duplicate of [community] packages - ok, seems reasonable for me.
Latelly after weirddan455' comment on catalyst@AUR i was thinking about the fact that [community] repo and [catalyst] repo are sharing pkgs in same versions, and actually only -pxp pkgs are in plus for [catalyst] repo. So I was thinking about moving [catalyst] repo step forward and changing it into testing repo with new betas.
There has been more betas then stable releases recently so i don't think it's the bad idea.
For (pxp) users who do like stable versions i will create [catalyst-stable] repo with current stable packages.
What do you think about this?
Vi0L0 wrote:thebluesgnr wrote:I have to disable VariBright every time I reboot. Does anyone else have the same problem?
Also, is there a command line way of disabling it? That would make this bug easier to deal with, at least.
Maybe setting PP_VariBrightFeatureEnable variable to PP_VariBrightFeatureEnable=V0 in /etc/ati/amdpcsdb _and_ /etc/ati/amdpcsdb.default will help?
Thanks for the advice.
I edited all occurrences in both files but still it seems that VariBright gets reenabled after a reboot.
hah, i though it will help. Does it keeping value =V0 after reboot, or is it changing?
Last edited by Vi0L0 (2013-02-06 17:09:02)
Offline
We still don't know what will happen when a newer version of Xorg comes out that Catalyst doesn't support. What will happen in that case, if your repository is removed?
If you can't sit by a cozy fire with your code in hand enjoying its simplicity and clarity, it needs more work. --Carlos Torres
Offline
We still don't know what will happen when a newer version of Xorg comes out that Catalyst doesn't support. What will happen in that case, if your repository is removed?
I'm not planning to remove [catalyst] repos, hope that apocalypsus isn't planning this either
i'm only planning:
- [catalyst] with latest beta
- [catalyst-stable] with latest stable rls
edit:
ofc [catalyst-hd234k] stays untouched
Last edited by Vi0L0 (2013-02-06 16:14:47)
Offline
In that case, I'd say go for it!
If you can't sit by a cozy fire with your code in hand enjoying its simplicity and clarity, it needs more work. --Carlos Torres
Offline
Archimaredes wrote:but TF2 struggles.
With the 13.2 beta 3 driver?
No, I have 13.1 from the official repositories. Should I replace it with catalyst-test from the AUR?
EDIT: Also, does this latest beta also not support xorg-server 1.13? I'm still holding back to 1.12 currently.
Last edited by Archimaredes (2013-02-06 18:51:37)
Offline
ChemBro wrote:Archimaredes wrote:but TF2 struggles.
With the 13.2 beta 3 driver?
No, I have 13.1 from the official repositories. Should I replace it with catalyst-test from the AUR?
EDIT: Also, does this latest beta also not support xorg-server 1.13? I'm still holding back to 1.12 currently.
1.) 13.1 and 13.2 both support xorg-server 1.13.
2.) 13.2 beta 3 has a 300%+ performance boost for TF2. That does not mean, TF2 is now ultra fast, it means, it is now playable. At least for me.
Offline
Archimaredes wrote:ChemBro wrote:With the 13.2 beta 3 driver?
No, I have 13.1 from the official repositories. Should I replace it with catalyst-test from the AUR?
EDIT: Also, does this latest beta also not support xorg-server 1.13? I'm still holding back to 1.12 currently.
1.) 13.1 and 13.2 both support xorg-server 1.13.
2.) 13.2 beta 3 has a 300%+ performance boost for TF2. That does not mean, TF2 is now ultra fast, it means, it is now playable. At least for me.
Great! Just installed catalyst-test, will install new xorg stuff now. Thanks.
Offline
hah, i though it will help. Does it keeping value =V0 after reboot, or is it changing?
Oh, I know. It was good advice.
The amdpcsdb file kept changing after reboot so my change only remained in amdpcdsdb.default. Thanks to the variable you pointed out I did some googling and found the command "sudo aticonfig --set-pcs-u32=MCIL,PP_UserVariBrightEnable,0". Now it seems to have stuck, we'll see.
Thanks again! For the help and for maintaining the repo.
Offline
Hi,
Looks like we (legacy users only ?) will not be able to use desktop effects on KDE 4.10 and higher.
https://bugs.kde.org/show_bug.cgi?id=314602#c6
Last edited by omeringen (2013-02-07 21:18:27)
Offline
Should the 13.1 legacy drivers work with an RS780 (HD3200) on a 3.7.5 kernel and xorg 1.13? I'm asking because I can't seem to get it to work. dmesg shows an error when loading fglrx that says "No supported display adapters were found." I know that seems pretty clear, but I'm seeing conflicting information around the web that seems to indicate that it should work. Can someone please provide me some clarity?
EDIT: From what I've read, it seems catalyst legacy 13.1 still doesn't work with xorg 1.13. Either way, I still get the error message when loading the module.
[ 4.309091] <6>[fglrx] Maximum main memory to use for locked dma buffers: 2620 MBytes.
[ 4.309401] <3>[fglrx:firegl_init_device_list] *ERROR* No supported display adapters were found
[ 4.309405] <3>[fglrx:firegl_init_module] *ERROR* firegl_init_devices failed
EDIT: I checked the aliases from the fglrx module, but I don't see the ven/dev id (1002:9610) for my IGP :-\ Does this mean the legacy 13.1 module won't work at all for me?
Last edited by hanman (2013-02-08 05:09:31)
Offline
Should the 13.1 legacy drivers work with an RS780 (HD3200) on a 3.7.5 kernel and xorg 1.13?
Kernel 3.7 with a patch, I guess. xorg 1.13: No.
Offline
Kernel 3.7 with a patch, I guess. xorg 1.13: No.
I never had any luck with Vi0L0's legacy repository, but the AUR total package worked great! Thanks for the hint that got me going!
Last edited by hanman (2013-02-08 13:56:09)
Offline
Hope it's not out of place here to say thanks for all the work, Vi0L0—and everyone else involved—in keeping my Radeon card up and running all these months. I can't believe I'm actually running xorg 1.13 now without a hitch! Surely that means a new version is due any day now, hah.
github ○
Offline
Hi,
Looks like we (legacy users only ?) will not be able to use desktop effects on KDE 4.10 and higher.
https://bugs.kde.org/show_bug.cgi?id=314602#c6
I did a test and desktop effects are working without any issues. But still there isn't any direct OpenGL rendering support for us.
EDIT: After a few reboots and struggling with KWin, it looks like Blur effect is completely broken with legacy catalyst drivers. After switching to open source drivers, problem is gone. Thank you AMD!!!
Last edited by siavashserver (2013-02-09 08:55:11)
Offline