Archived from groups: alt.comp.periphs.mainboard.asus (More info?)

I have a friend who just upgraded her computer with
an Asus A7v8X-X motherboard and a fresh install of
Win 98SE. Now she can't get DOS sound out of her
SBLive sound card. She's got all the DOS drivers
installed from the SBLive CD. But I've read the
SBLive's emulation won't work with motherboards
that don't support NMI (non-maskable interrupts).

Does anyone happen to know if the Asus A7v8X-X
supports NMI? There's a product page for it here
http://www.asus.com/products/mb/so [...] erview.htm
but I don't see any mention of NMI or non-maskable interrupts.

One of her favorite old DOS games is Realms of the Haunting,
but it won't work with the DOSBox emulator. So she's been
spending a lot of time trying to get her DOS sound working.
I'm wondering if it's a hopeless cause.

Archived from groups: alt.comp.periphs.mainboard.asus (More info?)

In article <nXuac.5546$NL4.283@newsread3.news.atl.earthlink.net>, spets
<nospam@nospam.com> wrote:

> I have a friend who just upgraded her computer with
> an Asus A7v8X-X motherboard and a fresh install of
> Win 98SE. Now she can't get DOS sound out of her
> SBLive sound card. She's got all the DOS drivers
> installed from the SBLive CD. But I've read the
> SBLive's emulation won't work with motherboards
> that don't support NMI (non-maskable interrupts).
>
> Does anyone happen to know if the Asus A7v8X-X
> supports NMI? There's a product page for it here
> http://www.asus.com/products/mb/so [...] erview.htm
> but I don't see any mention of NMI or non-maskable interrupts.
>
> One of her favorite old DOS games is Realms of the Haunting,
> but it won't work with the DOSBox emulator. So she's been
> spending a lot of time trying to get her DOS sound working.
> I'm wondering if it's a hopeless cause.

Looking in Google, there are a lot of references to NMI,
as mentioned on the Creative site, but as a hardware guy, I
don't understand how this would work with a PCI card. There
is no NMI on there. (But in the PCI standard, I did notice a
reference to SERR# system error signal on a PCI card, needing
to be routed to NMI to raise a system error if something
serious is detected on the PCI card. It is the only place
in the standard where the word "NMI" appears.)

http://www.mameworld.net/emuadvice/sound.html

This post basically identifies that SERR# has to be wired
to the chain of logic leading to NMI.

http://groups.google.com/groups?hl [...] mweb.co.za

I had a look at a reference schematic for an Intel chipset,
and SERR is wired to a pin on the Southbridge. The Southbridge
has register control over what the SERR is connected to. For
example, SERR can be logically connected to the NMI pin that
goes from the Southbridge to the processor. Perhaps this is the
kind of BIOS support that is mentioned - maybe the BIOS needs
to set up the response from SERR to NMI.

As a result of this quick check, I have to think that from a
hardware perspective, this is a "follow the dots" kind of
problem - as long as the HW designer hooks up the pin, the
rest of the solution is in the hands of BIOS writers and
driver writers.

Since no docs are available on Via chipsets, that is as much
analogizing as I can do.

Paul

Archived from groups: alt.comp.periphs.mainboard.asus (More info?)

Paul wrote:
> In article <nXuac.5546$NL4.283@newsread3.news.atl.earthlink.net>, spets
> <nospam@nospam.com> wrote:
>
>
>>I have a friend who just upgraded her computer with
>>an Asus A7v8X-X motherboard and a fresh install of
>>Win 98SE. Now she can't get DOS sound out of her
>>SBLive sound card. She's got all the DOS drivers
>>installed from the SBLive CD. But I've read the
>>SBLive's emulation won't work with motherboards
>>that don't support NMI (non-maskable interrupts).
>>
>>Does anyone happen to know if the Asus A7v8X-X
>>supports NMI? There's a product page for it here
>>http://www.asus.com/products/mb/socketa/a7v8x-x/overview.htm
>>but I don't see any mention of NMI or non-maskable interrupts.
>>
>>One of her favorite old DOS games is Realms of the Haunting,
>>but it won't work with the DOSBox emulator. So she's been
>>spending a lot of time trying to get her DOS sound working.
>>I'm wondering if it's a hopeless cause.
>
>
> Looking in Google, there are a lot of references to NMI,
> as mentioned on the Creative site, but as a hardware guy, I
> don't understand how this would work with a PCI card. There
> is no NMI on there. (But in the PCI standard, I did notice a
> reference to SERR# system error signal on a PCI card, needing
> to be routed to NMI to raise a system error if something
> serious is detected on the PCI card. It is the only place
> in the standard where the word "NMI" appears.)
>
> http://www.mameworld.net/emuadvice/sound.html
>
> This post basically identifies that SERR# has to be wired
> to the chain of logic leading to NMI.
>
> http://groups.google.com/groups?hl [...] mweb.co.za
>
> I had a look at a reference schematic for an Intel chipset,
> and SERR is wired to a pin on the Southbridge. The Southbridge
> has register control over what the SERR is connected to. For
> example, SERR can be logically connected to the NMI pin that
> goes from the Southbridge to the processor. Perhaps this is the
> kind of BIOS support that is mentioned - maybe the BIOS needs
> to set up the response from SERR to NMI.
>
> As a result of this quick check, I have to think that from a
> hardware perspective, this is a "follow the dots" kind of
> problem - as long as the HW designer hooks up the pin, the
> rest of the solution is in the hands of BIOS writers and
> driver writers.
>
> Since no docs are available on Via chipsets, that is as much
> analogizing as I can do.
>
> Paul

Thank you for finding all this out for me. I'll show
my friend the links and see what she thinks. She's
been asking me to explain why the card needs NMI
and the google link you found does just that.

Archived from groups: alt.comp.periphs.mainboard.asus (More info?)

In article <gRDac.6639$yN6.5635@newsread2.news.atl.earthlink.net>, spets
<nospam@nospam.com> wrote:

> Paul wrote:
> > In article <nXuac.5546$NL4.283@newsread3.news.atl.earthlink.net>, spets
> > <nospam@nospam.com> wrote:
> >
> >
> >>I have a friend who just upgraded her computer with
> >>an Asus A7v8X-X motherboard and a fresh install of
> >>Win 98SE. Now she can't get DOS sound out of her
> >>SBLive sound card. She's got all the DOS drivers
> >>installed from the SBLive CD. But I've read the
> >>SBLive's emulation won't work with motherboards
> >>that don't support NMI (non-maskable interrupts).
> >>
> >>Does anyone happen to know if the Asus A7v8X-X
> >>supports NMI? There's a product page for it here
> >>http://www.asus.com/products/mb/socketa/a7v8x-x/overview.htm
> >>but I don't see any mention of NMI or non-maskable interrupts.
> >>
> >>One of her favorite old DOS games is Realms of the Haunting,
> >>but it won't work with the DOSBox emulator. So she's been
> >>spending a lot of time trying to get her DOS sound working.
> >>I'm wondering if it's a hopeless cause.
> >
> >
> > Looking in Google, there are a lot of references to NMI,
> > as mentioned on the Creative site, but as a hardware guy, I
> > don't understand how this would work with a PCI card. There
> > is no NMI on there. (But in the PCI standard, I did notice a
> > reference to SERR# system error signal on a PCI card, needing
> > to be routed to NMI to raise a system error if something
> > serious is detected on the PCI card. It is the only place
> > in the standard where the word "NMI" appears.)
> >
> > http://www.mameworld.net/emuadvice/sound.html
> >
> > This post basically identifies that SERR# has to be wired
> > to the chain of logic leading to NMI.
> >
> >
http://groups.google.com/groups?hl [...] mweb.co.za
> >
> > I had a look at a reference schematic for an Intel chipset,
> > and SERR is wired to a pin on the Southbridge. The Southbridge
> > has register control over what the SERR is connected to. For
> > example, SERR can be logically connected to the NMI pin that
> > goes from the Southbridge to the processor. Perhaps this is the
> > kind of BIOS support that is mentioned - maybe the BIOS needs
> > to set up the response from SERR to NMI.
> >
> > As a result of this quick check, I have to think that from a
> > hardware perspective, this is a "follow the dots" kind of
> > problem - as long as the HW designer hooks up the pin, the
> > rest of the solution is in the hands of BIOS writers and
> > driver writers.
> >
> > Since no docs are available on Via chipsets, that is as much
> > analogizing as I can do.
> >
> > Paul
>
> Thank you for finding all this out for me. I'll show
> my friend the links and see what she thinks. She's
> been asking me to explain why the card needs NMI
> and the google link you found does just that.

Looking in Google, I also find SERR# mentioned in a control file
for WPCREDIT. Maybe SERR# gets coupled into interrupt logic with
that setting, which you could at least query to see whether it is
enabled or not. It could be that the SERR# logic is unreliable
and that is why it isn't enabled by Asus.

http://groups.google.com/groups?hl [...] om&rnum=12

Happy exploring,
Paul