I'm having trouble getting agpgart support to work with an nforce2
chipset. Is this supported on any kernels? I'm running a Redhat 7.1
system with Redhat's 2.4.9-21 kernel.
--
Carl D. Blake
Director of Engineering
Boeckeler Instruments, Inc.
4650 S. Butterfield Dr.
Tucson, AZ 85714
Phone: 520-745-0001
FAX: 520-745-0004
email: [email protected]
.com
On Mon, 2002-12-23 at 20:50, Carl D. Blake wrote:
> I'm having trouble getting agpgart support to work with an nforce2
> chipset. Is this supported on any kernels? I'm running a Redhat 7.1
> system with Redhat's 2.4.9-21 kernel.
Try to use a newer kernel from Redhat, because that kernel was around
looong before nforce was released. IIRC support for nforce2 was added
around 2.4.19
--
For future reference - don't anybody else try to send patches as vi
scripts, please. Yes, it's manly, but let's face it, so is
bungee-jumping with the cord tied to your testicles.
-- Linus
On Mon, 2002-12-23 at 14:16, Bongani Hlope wrote:
> On Mon, 2002-12-23 at 20:50, Carl D. Blake wrote:
> > I'm having trouble getting agpgart support to work with an nforce2
> > chipset. Is this supported on any kernels? I'm running a Redhat 7.1
> > system with Redhat's 2.4.9-21 kernel.
>
> Try to use a newer kernel from Redhat, because that kernel was around
> looong before nforce was released. IIRC support for nforce2 was added
> around 2.4.19
>
I just upgraded to the 2.4.18 kernel provided by Redhat and it didn't
make any difference. The message I get in dmesg is:
Linux agpgart interface v0.99 (c) Jeff Hartmann
agpgart: Maximum main memory to use for agp memory: 439M
agpgart: unsupported bridge
agpgart: no supported devices found.
You suggested trying 2.4.19, so I downloaded kernel 2.4.20 from
kernel.org and compared its agp code (drivers/char/agp) with the code in
2.4.18. There are a few differences - such as supporting AMD 8151 - but
nothing that indicates improved support for agpgart on the nforce2
chipset. The changelog for 2.4.20 indicated some added support for the
nforce2 chipset, but that seems to be support for the audio and network
portions of the chipset, not agp. I was able to incorporate the audio
changes manually for kernel 2.4.18 by using Nvidia's patches, but I
can't get agp to work.
Any other suggestions? Thanks for your help.
> --
> For future reference - don't anybody else try to send patches as vi
> scripts, please. Yes, it's manly, but let's face it, so is
> bungee-jumping with the cord tied to your testicles.
>
> -- Linus
--
Carl D. Blake
Director of Engineering
Boeckeler Instruments, Inc.
4650 S. Butterfield Dr.
Tucson, AZ 85714
Phone: 520-745-0001
FAX: 520-745-0004
email: [email protected]
.com
On Tue, 2002-12-24 at 00:40, Carl D. Blake wrote:
> On Mon, 2002-12-23 at 14:16, Bongani Hlope wrote:
> > On Mon, 2002-12-23 at 20:50, Carl D. Blake wrote:
> > > I'm having trouble getting agpgart support to work with an nforce2
> > > chipset. Is this supported on any kernels? I'm running a Redhat 7.1
> > > system with Redhat's 2.4.9-21 kernel.
> >
> > Try to use a newer kernel from Redhat, because that kernel was around
> > looong before nforce was released. IIRC support for nforce2 was added
> > around 2.4.19
> >
>
> I just upgraded to the 2.4.18 kernel provided by Redhat and it didn't
> make any difference. The message I get in dmesg is:
>
> Linux agpgart interface v0.99 (c) Jeff Hartmann
> agpgart: Maximum main memory to use for agp memory: 439M
> agpgart: unsupported bridge
> agpgart: no supported devices found.
>
> You suggested trying 2.4.19, so I downloaded kernel 2.4.20 from
> kernel.org and compared its agp code (drivers/char/agp) with the code in
> 2.4.18. There are a few differences - such as supporting AMD 8151 - but
> nothing that indicates improved support for agpgart on the nforce2
> chipset. The changelog for 2.4.20 indicated some added support for the
> nforce2 chipset, but that seems to be support for the audio and network
> portions of the chipset, not agp. I was able to incorporate the audio
> changes manually for kernel 2.4.18 by using Nvidia's patches, but I
> can't get agp to work.
>
> Any other suggestions? Thanks for your help.
Disclaimer at the bottom. ;)
It was once posted on this list that the NForce is similar to some
existing hardware. hence the sound driver uses the Intel i810 audio
driver. I guess if you can search through the archive you might find
that discussion. If you do find it and _are_ clued-up enough, then you
can look at drivers/char/agp/agpgart_be.c approximately line 3900 they
start defining which drivers to use for which bus.
So you need to
1. Find the id for the AGP bus for the NForce and define it in agp.h
(lspci should do the trick IIRC on my PC 00:01.0 is the agp bus)
2. Find out which hardware the bus is similar to, then add the support
for the NForce using the setup for that driver
e.g.
agp.h
add
#ifndef PCI_DEVICE_ID_NVIDIA_NFORCORCE_2 //I'm not sure about the
naming standard
#define PCI_DEVICE_ID_NVIDIA_NFORCORCE_2 0x(Hex value from lspci)
#endif
the in agpgart_be.c
add (for example lets say the bus is similar to Intel's i860 agp bus
then)
#ifdef CONFIG_AGP_INTEL
...
{ PCI_DEVICE_ID_NVIDIA_NFORCORCE_2,
PCI_VENDOR_ID_NVIDIA,
INTEL_I860,
"NVidia",
"NForce2",
intel_860_setup },
....
#endif /* CONFIG_AGP_INTEL */
DISCLAIMER : I do not know anything about the agp code, except for what
I have read when trying to find a solution for you. The examples above
are from what my eyes came across at 1:00 am without prio knowledge to
the agp code. So if you implement these examples you do it at you own
(and maybe your pets) risk. The examples come with.... you should now
how the rest goes ;)
--
For future reference - don't anybody else try to send patches as vi
scripts, please. Yes, it's manly, but let's face it, so is
bungee-jumping with the cord tied to your testicles.
-- Linus
Carl,
Are you using the onboard IDE controller for your hard disk?
If so, please check to see if DMA is enabled or not.
Jeff
----- Original Message -----
From: "Carl D. Blake" <[email protected]>
To: <[email protected]>
Sent: Monday, December 23, 2002 2:40 PM
Subject: Re: nforce2 and agpgart
> On Mon, 2002-12-23 at 14:16, Bongani Hlope wrote:
> > On Mon, 2002-12-23 at 20:50, Carl D. Blake wrote:
> > > I'm having trouble getting agpgart support to work with an nforce2
> > > chipset. Is this supported on any kernels? I'm running a Redhat 7.1
> > > system with Redhat's 2.4.9-21 kernel.
> >
> > Try to use a newer kernel from Redhat, because that kernel was around
> > looong before nforce was released. IIRC support for nforce2 was added
> > around 2.4.19
> >
>
> I just upgraded to the 2.4.18 kernel provided by Redhat and it didn't
> make any difference. The message I get in dmesg is:
>
> Linux agpgart interface v0.99 (c) Jeff Hartmann
> agpgart: Maximum main memory to use for agp memory: 439M
> agpgart: unsupported bridge
> agpgart: no supported devices found.
>
> You suggested trying 2.4.19, so I downloaded kernel 2.4.20 from
> kernel.org and compared its agp code (drivers/char/agp) with the code in
> 2.4.18. There are a few differences - such as supporting AMD 8151 - but
> nothing that indicates improved support for agpgart on the nforce2
> chipset. The changelog for 2.4.20 indicated some added support for the
> nforce2 chipset, but that seems to be support for the audio and network
> portions of the chipset, not agp. I was able to incorporate the audio
> changes manually for kernel 2.4.18 by using Nvidia's patches, but I
> can't get agp to work.
>
> Any other suggestions? Thanks for your help.
> > --
> > For future reference - don't anybody else try to send patches as vi
> > scripts, please. Yes, it's manly, but let's face it, so is
> > bungee-jumping with the cord tied to your testicles.
> >
> > -- Linus
> --
> Carl D. Blake
> Director of Engineering
> Boeckeler Instruments, Inc.
> 4650 S. Butterfield Dr.
> Tucson, AZ 85714
>
> Phone: 520-745-0001
> FAX: 520-745-0004
> email: [email protected]
>
> .com
>
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/
>
On Mon, 2002-12-23 at 16:59, Jeff Nguyen wrote:
> Carl,
>
> Are you using the onboard IDE controller for your hard disk?
> If so, please check to see if DMA is enabled or not.
>
> Jeff
>
I am using the onboard IDE controller and DMA is enabled.
> ----- Original Message -----
> From: "Carl D. Blake" <[email protected]>
> To: <[email protected]>
> Sent: Monday, December 23, 2002 2:40 PM
> Subject: Re: nforce2 and agpgart
>
>
> > On Mon, 2002-12-23 at 14:16, Bongani Hlope wrote:
> > > On Mon, 2002-12-23 at 20:50, Carl D. Blake wrote:
> > > > I'm having trouble getting agpgart support to work with an nforce2
> > > > chipset. Is this supported on any kernels? I'm running a Redhat 7.1
> > > > system with Redhat's 2.4.9-21 kernel.
> > >
> > > Try to use a newer kernel from Redhat, because that kernel was around
> > > looong before nforce was released. IIRC support for nforce2 was added
> > > around 2.4.19
> > >
> >
> > I just upgraded to the 2.4.18 kernel provided by Redhat and it didn't
> > make any difference. The message I get in dmesg is:
> >
> > Linux agpgart interface v0.99 (c) Jeff Hartmann
> > agpgart: Maximum main memory to use for agp memory: 439M
> > agpgart: unsupported bridge
> > agpgart: no supported devices found.
> >
> > You suggested trying 2.4.19, so I downloaded kernel 2.4.20 from
> > kernel.org and compared its agp code (drivers/char/agp) with the code in
> > 2.4.18. There are a few differences - such as supporting AMD 8151 - but
> > nothing that indicates improved support for agpgart on the nforce2
> > chipset. The changelog for 2.4.20 indicated some added support for the
> > nforce2 chipset, but that seems to be support for the audio and network
> > portions of the chipset, not agp. I was able to incorporate the audio
> > changes manually for kernel 2.4.18 by using Nvidia's patches, but I
> > can't get agp to work.
> >
> > Any other suggestions? Thanks for your help.
> > > --
> > > For future reference - don't anybody else try to send patches as vi
> > > scripts, please. Yes, it's manly, but let's face it, so is
> > > bungee-jumping with the cord tied to your testicles.
> > >
> > > -- Linus
> > --
> > Carl D. Blake
> > Director of Engineering
> > Boeckeler Instruments, Inc.
> > 4650 S. Butterfield Dr.
> > Tucson, AZ 85714
> >
> > Phone: 520-745-0001
> > FAX: 520-745-0004
> > email: [email protected]
> >
> > .com
> >
> > -
> > To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> > the body of a message to [email protected]
> > More majordomo info at http://vger.kernel.org/majordomo-info.html
> > Please read the FAQ at http://www.tux.org/lkml/
> >
>
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/
--
Carl D. Blake
Director of Engineering
Boeckeler Instruments, Inc.
4650 S. Butterfield Dr.
Tucson, AZ 85714
Phone: 520-745-0001
FAX: 520-745-0004
email: [email protected]
.com
On Mon, 2002-12-23 at 16:19, Bongani Hlope wrote:
...snip...
>
> It was once posted on this list that the NForce is similar to some
> existing hardware. hence the sound driver uses the Intel i810 audio
> driver. I guess if you can search through the archive you might find
> that discussion. If you do find it and _are_ clued-up enough, then you
> can look at drivers/char/agp/agpgart_be.c approximately line 3900 they
> start defining which drivers to use for which bus.
>
This sounds like a possibility. However, I cannot find any discussion
about what hardware the nforce2 chipset is similar to. I just need to
know if the agp portion of the chipset is similar to something else.
Can anybody point to that information? From what I can determine the
device that is used is the host bridge.
> So you need to
> 1. Find the id for the AGP bus for the NForce and define it in agp.h
> (lspci should do the trick IIRC on my PC 00:01.0 is the agp bus)
> 2. Find out which hardware the bus is similar to, then add the support
> for the NForce using the setup for that driver
> e.g.
>
> agp.h
>
> add
> #ifndef PCI_DEVICE_ID_NVIDIA_NFORCORCE_2 //I'm not sure about the
> naming standard
> #define PCI_DEVICE_ID_NVIDIA_NFORCORCE_2 0x(Hex value from lspci)
> #endif
>
> the in agpgart_be.c
>
> add (for example lets say the bus is similar to Intel's i860 agp bus
> then)
>
> #ifdef CONFIG_AGP_INTEL
> ...
>
> { PCI_DEVICE_ID_NVIDIA_NFORCORCE_2,
> PCI_VENDOR_ID_NVIDIA,
> INTEL_I860,
> "NVidia",
> "NForce2",
> intel_860_setup },
>
> ....
> #endif /* CONFIG_AGP_INTEL */
>
> DISCLAIMER : I do not know anything about the agp code, except for what
> I have read when trying to find a solution for you. The examples above
> are from what my eyes came across at 1:00 am without prio knowledge to
> the agp code. So if you implement these examples you do it at you own
> (and maybe your pets) risk. The examples come with.... you should now
> how the rest goes ;)
>
> --
> For future reference - don't anybody else try to send patches as vi
> scripts, please. Yes, it's manly, but let's face it, so is
> bungee-jumping with the cord tied to your testicles.
>
> -- Linus
--
Carl D. Blake
Director of Engineering
Boeckeler Instruments, Inc.
4650 S. Butterfield Dr.
Tucson, AZ 85714
Phone: 520-745-0001
FAX: 520-745-0004
email: [email protected]
.com
On Thu, 2002-12-26 at 12:52, Jeff Nguyen wrote:
> That's strange because my Red Hat kernel did not enable DMA.
> Did you have to run hdparam to change the IDE setting?
>
Yes, I had to run hdparm to setup the IDE hard drive for maximum
performance.
> Jeff
>
> ----- Original Message -----
> From: "Carl D. Blake" <[email protected]>
> To: "Jeff Nguyen" <[email protected]>
> Sent: Thursday, December 26, 2002 10:51 AM
> Subject: Re: nforce2 and agpgart
>
>
> > On Mon, 2002-12-23 at 16:59, Jeff Nguyen wrote:
> > > Carl,
> > >
> > > Are you using the onboard IDE controller for your hard disk?
> > > If so, please check to see if DMA is enabled or not.
> > >
> > > Jeff
> > >
> >
> > I am using the onboard IDE controller and DMA is enabled.
> >
> > > ----- Original Message -----
> > > From: "Carl D. Blake" <[email protected]>
> > > To: <[email protected]>
> > > Sent: Monday, December 23, 2002 2:40 PM
> > > Subject: Re: nforce2 and agpgart
> > >
> > >
> > > > On Mon, 2002-12-23 at 14:16, Bongani Hlope wrote:
> > > > > On Mon, 2002-12-23 at 20:50, Carl D. Blake wrote:
> > > > > > I'm having trouble getting agpgart support to work with an nforce2
> > > > > > chipset. Is this supported on any kernels? I'm running a Redhat
> 7.1
> > > > > > system with Redhat's 2.4.9-21 kernel.
> > > > >
> > > > > Try to use a newer kernel from Redhat, because that kernel was
> around
> > > > > looong before nforce was released. IIRC support for nforce2 was
> added
> > > > > around 2.4.19
> > > > >
> > > >
> > > > I just upgraded to the 2.4.18 kernel provided by Redhat and it didn't
> > > > make any difference. The message I get in dmesg is:
> > > >
> > > > Linux agpgart interface v0.99 (c) Jeff Hartmann
> > > > agpgart: Maximum main memory to use for agp memory: 439M
> > > > agpgart: unsupported bridge
> > > > agpgart: no supported devices found.
> > > >
> > > > You suggested trying 2.4.19, so I downloaded kernel 2.4.20 from
> > > > kernel.org and compared its agp code (drivers/char/agp) with the code
> in
> > > > 2.4.18. There are a few differences - such as supporting AMD 8151 -
> but
> > > > nothing that indicates improved support for agpgart on the nforce2
> > > > chipset. The changelog for 2.4.20 indicated some added support for
> the
> > > > nforce2 chipset, but that seems to be support for the audio and
> network
> > > > portions of the chipset, not agp. I was able to incorporate the audio
> > > > changes manually for kernel 2.4.18 by using Nvidia's patches, but I
> > > > can't get agp to work.
> > > >
> > > > Any other suggestions? Thanks for your help.
> > > > > --
> > > > > For future reference - don't anybody else try to send patches as vi
> > > > > scripts, please. Yes, it's manly, but let's face it, so is
> > > > > bungee-jumping with the cord tied to your testicles.
> > > > >
> > > > > -- Linus
> > > > --
> > > > Carl D. Blake
> > > > Director of Engineering
> > > > Boeckeler Instruments, Inc.
> > > > 4650 S. Butterfield Dr.
> > > > Tucson, AZ 85714
> > > >
> > > > Phone: 520-745-0001
> > > > FAX: 520-745-0004
> > > > email: [email protected]
> > > >
> > > > .com
> > > >
> > > > -
> > > > To unsubscribe from this list: send the line "unsubscribe
> linux-kernel" in
> > > > the body of a message to [email protected]
> > > > More majordomo info at http://vger.kernel.org/majordomo-info.html
> > > > Please read the FAQ at http://www.tux.org/lkml/
> > > >
> > >
> > > -
> > > To unsubscribe from this list: send the line "unsubscribe linux-kernel"
> in
> > > the body of a message to [email protected]
> > > More majordomo info at http://vger.kernel.org/majordomo-info.html
> > > Please read the FAQ at http://www.tux.org/lkml/
> > --
> > Carl D. Blake
> > Director of Engineering
> > Boeckeler Instruments, Inc.
> > 4650 S. Butterfield Dr.
> > Tucson, AZ 85714
> >
> > Phone: 520-745-0001
> > FAX: 520-745-0004
> > email: [email protected]
> >
> > .com
> >
--
Carl D. Blake
Director of Engineering
Boeckeler Instruments, Inc.
4650 S. Butterfield Dr.
Tucson, AZ 85714
Phone: 520-745-0001
FAX: 520-745-0004
email: [email protected]
.com