6 Apr
2011
6 Apr
'11
5:04 a.m.
Hey all,
1. I am researching to build a new system here and need to make the
video decision of NVidia or ATI. My goal is to be able to run 3d games
and do my audio work on the same linux installation if possible.
I know in the past I heard bad things about the binary proprietary
drivers and realtime audio(freezing, xruns, etc). Is this still the
case? Does realtime Jack audio have issues with the proprietary
drivers, or with composting window managers?
There are plenty of reviews on linux graphics performance for nvidia,
nouveau, fglrx, and radeon(mesa/gallium3d), but none on xrun stability.
Anyone have experience/recommendations?
2. System76 and ZaReason seem to be the two popular companies in Linux
PC sales. I would like to avoid Ubuntu so that outs System76. Anyone
have good/bad experiences with ZaReason? http://zareason.com
3. Lastly, anyone have any info on multi-booting the Intel MacMini with
Linux for audio? I'd rather not get an Apple, but I need to get my
linux audio fix and my Diablo 3 fix(someday) and am avoiding Windows.
Thanks,
Brian
6 Apr
6 Apr
8:01 a.m.
Brian Bergstrom wrote:
1. I am researching to build a new system here and
need to make the
video decision of NVidia or ATI. My goal is to be able to run 3d games
and do my audio work on the same linux installation if possible.
AFAIK the proprietary nVidia video driver still works a bit better than
the AMD one.
I know in the past I heard bad things about the binary
proprietary
drivers and realtime audio(freezing, xruns, etc). Is this still the
case?
Both nvidia's and fglrx's interrupt handlers are evil. However, you
should be able to minimize their effects on the rest of the system by
using a multi-core CPU (not that you could avoid this nowadays) and
message-signaled interrupts (the nvidia driver needs the module option
"NVreg_EnableMSI=1" to enable them). In the past, there were problems
with MSI on some chipsets, but nowadays it works.
recommendations?
A nVidia card, and a modern chipset (i.e., not nVidia).
composting window managers
^^^^^^^^^^
It seems I haven't kept up with the latest advances in desktop
technology. How does this work? ;-)
Regards,
Clemens
12:54 p.m.
On Wed, Apr 6, 2011 at 2:02 AM, Clemens Ladisch <clemens(a)ladisch.de> wrote:
clemens, two questions:
1) where does this need to be given to the module? is this a kernel
module option or
a X Window option?
2) does nouveau have any way to do this or does it not matter
because of the way it
handles interrupts?
I know in the
past I heard bad things about the binary proprietary
drivers and realtime audio(freezing, xruns, etc). Is this still the
case?
Both nvidia's and fglrx's interrupt handlers are evil. However, you
should be able to minimize their effects on the rest of the system by
using a multi-core CPU (not that you could avoid this nowadays) and
message-signaled interrupts (the nvidia driver needs the module option
"NVreg_EnableMSI=1" to enable them). In the past, there were problems
with MSI on some chipsets, but nowadays it works. 
3:03 p.m.
On Wed, Apr 6, 2011 at 3:54 AM, Paul Davis <paul(a)linuxaudiosystems.com> wrote:
On Wed, Apr 6, 2011 at 2:02 AM, Clemens Ladisch
<clemens(a)ladisch.de> wrote:
clemens, two questions:
1) where does this need to be given to the module? is this a kernel
module option or
a X Window option?
/etc/modules.autoload.d/kernel-2.6
but I haven't been able to prove it helps with my GTX465.
I know in
the past I heard bad things about the binary proprietary
drivers and realtime audio(freezing, xruns, etc). Is this still the
case?
Both nvidia's and fglrx's interrupt handlers are evil. However, you
should be able to minimize their effects on the rest of the system by
using a multi-core CPU (not that you could avoid this nowadays) and
message-signaled interrupts (the nvidia driver needs the module option
"NVreg_EnableMSI=1" to enable them). In the past, there were problems
with MSI on some chipsets, but nowadays it works. 2) does nouveau have any way to do this or does it
not matter
because of the way it
handles interrupts?
If you have it loaded then look at
modinfo nouveau
to see if there's an option to play with.
HTH,
Mark
4:35 p.m.
Mark Knecht wrote:
On Wed, Apr 6, 2011 at 3:54 AM, Paul Davis
<paul(a)linuxaudiosystems.com> wrote:
> > NVreg_EnableMSI=1
>
> 2) does nouveau have any way to do this or does it not matter
> because of the way it handles interrupts?
The noveau driver does this automatically.
Regards,
Clemens
7 Apr
7 Apr
2:26 p.m.
On Wed, Apr 6, 2011 at 10:36 AM, Clemens Ladisch <clemens(a)ladisch.de> wrote:
Mark Knecht wrote:
apparently not:
CPU0 CPU1 CPU2 CPU3 CPU4 CPU5
0: 468 941 1326 11898 6670801
1914669321 IO-APIC-edge timer
1: 0 0 0 0 0
2 IO-APIC-edge i8042
8: 0 0 0 0 0
1 IO-APIC-edge rtc0
9: 0 0 0 0 0
2 IO-APIC-fasteoi acpi
12: 0 0 0 0 0
4 IO-APIC-edge i8042
14: 2 4 5 502 25560
1962065 IO-APIC-edge pata_atiixp
15: 0 0 0 0 0
0 IO-APIC-edge pata_atiixp
16: 0 0 2138 1492054 16
2533 IO-APIC-fasteoi hda_intel
17: 0 1 0 0 0
22 IO-APIC-fasteoi ehci_hcd:usb1, ehci_hcd:usb2, ehci_hcd:usb3
18: 9448 18998 10957057 294899 9
394 IO-APIC-fasteoi ohci_hcd:usb4, ohci_hcd:usb5, ohci_hcd:usb6,
ohci_hcd:usb7
20: 9339 78325838 0 1 75
5373 IO-APIC-fasteoi hdsp
32: 51 0 0 0 6
238 IO-APIC-fasteoi nouveau, hda_intel
44: 0 0 0 0 0
0 IO-APIC-fasteoi ahci
45: 0 0 0 0 0
0 IO-APIC-fasteoi pata_jmicron
46: 0 0 0 0 0
2 IO-APIC-fasteoi firewire_ohci
50: 0 0 0 0 0
0 IO-APIC-fasteoi xhci_hcd:usb8
61: 0 0 0 0 0
0 PCI-MSI-edge ahci
62: 4 2 25 594 100599
7183742 PCI-MSI-edge sky2@pci:0000:04:00.0
On Wed, Apr 6, 2011 at 3:54 AM, Paul Davis
<paul(a)linuxaudiosystems.com> wrote:
> > NVreg_EnableMSI=1
>
> 2) does nouveau have any way to do this or does it not matter
> because of the way it handles interrupts?
The noveau driver does this automatically.
5:16 p.m.
Paul Davis wrote:
On Wed, Apr 6, 2011 at 10:36 AM, Clemens Ladisch
<clemens(a)ladisch.de> wrote:
Sorry, the check was so obvious in the source that I overlooked it. :-#
For nouveau, the module option is msi=1.
Regards,
Clemens
Mark Knecht wrote:
apparently not: On Wed, Apr 6, 2011 at 3:54 AM, Paul Davis
<paul(a)linuxaudiosystems.com> wrote:
> > NVreg_EnableMSI=1
>
> 2) does nouveau have any way to do this or does it not matter
> because of the way it handles interrupts?
The noveau driver does this automatically. 
10:33 a.m.
Clemens Ladisch wrote:
Brian Bergstrom wrote:
[...]
Thanks a lot!!!!
I always had problems with the NVIDIA card sharing the same IRQ with the
firewire port (where I have a Saffire) in my notebook (dual core)...
Using the propietary driver (which works better for graphics), every
time the graphics card was required to use a lot of bandwith, I had a
xruns and the desktop turns unresponsive.... But that makes the trick! I
just added the option for the module and now both graphics and sound
works like a charm, with better distribution of CPUs!!!
Thanks again! :)
Best regards,
Natanael.
I know in the past I heard bad things about the
binary proprietary
drivers and realtime audio(freezing, xruns, etc). Is this still the
case?
Both nvidia's and fglrx's interrupt handlers are evil. However,
you
should be able to minimize their effects on the rest of the system by
using a multi-core CPU (not that you could avoid this nowadays) and
message-signaled interrupts (the nvidia driver needs the module option
"NVreg_EnableMSI=1" to enable them). In the past, there were problems
with MSI on some chipsets, but nowadays it works.
[...]
4980
days inactive
4981
days old
linux-audio-user@lists.linuxaudio.org
7 comments
5 participants
participants (5)
-
Brian Bergstrom -
Clemens Ladisch -
Mark Knecht -
Natanael Olaiz -
Paul Davis