[LinuxBIOS] new to LinuxBIOS (help for nova4899r)

[Luis Correia]

Hi!

I've been reading the list for a while and learning how things work around here.

The choice of trying LinuxBIOS is mainly related to a board i've
bought, that has lots of PCI devices and where the factory BIOS simply
doesn't have any type of PCI sharing working properly.

There is no ACPI support.
> From what I've read, making ACPI tables is not easy...

The board is industrial grade, and it can be found here:
http://www.icpamerica.com/products/single_board_computers/5_25_NOVA/NOVA-489...

The version I have has the realtek ethernet chips.

Now, for my questions:

1) if I don't specify a device in config.lb, Linux will still be able
to recognize it
2) about IRQ sharing, what is the best solution
3) is it possible to enable ACPI without having the tables from the factory bios

Below you'll find an 'lspci -vvv' that I managed to produce

Thanks for all the input you may provide.

Luis Correia

lspci -vvv output

00:00.0 Host bridge: Cyrix Corporation PCI Master
    Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr-
Stepping- SERR- FastB2B-
    Status: Cap- 66Mhz- UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort-
<TAbort- <MAbort- >SERR- <PERR-
    Latency: 0

00:0a.0 Ethernet controller: Realtek Semiconductor Co., Ltd.
RTL-8139/8139C/8139C+ (rev 10)
    Subsystem: Realtek Semiconductor Co., Ltd. RT8139
    Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr-
Stepping- SERR- FastB2B-
    Status: Cap+ 66Mhz- UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort-
<TAbort- <MAbort- >SERR- <PERR-
    Latency: 32 (8000ns min, 16000ns max)
    Interrupt: pin A routed to IRQ 9
    Region 0: I/O ports at e000 [size=256]
    Region 1: Memory at d0000000 (32-bit, non-prefetchable) [size=256]
    Capabilities: [50] Power Management version 2
    	Flags: PMEClk- DSI- D1+ D2+ AuxCurrent=375mA PME(D0-,D1+,D2+,D3hot+,D3cold+)
    	Status: D0 PME-Enable- DSel=0 DScale=0 PME-

00:0b.0 Ethernet controller: Realtek Semiconductor Co., Ltd.
RTL-8139/8139C/8139C+ (rev 10)
    Subsystem: Realtek Semiconductor Co., Ltd. RT8139
    Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr-
Stepping- SERR- FastB2B-
    Status: Cap+ 66Mhz- UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort-
<TAbort- <MAbort- >SERR- <PERR-
    Latency: 32 (8000ns min, 16000ns max)
    Interrupt: pin A routed to IRQ 12
    Region 0: I/O ports at e400 [size=256]
    Region 1: Memory at d0001000 (32-bit, non-prefetchable) [size=256]
    Capabilities: [50] Power Management version 2
    	Flags: PMEClk- DSI- D1+ D2+ AuxCurrent=375mA PME(D0-,D1+,D2+,D3hot+,D3cold+)
    	Status: D0 PME-Enable- DSel=0 DScale=0 PME-

00:0c.0 Ethernet controller: Realtek Semiconductor Co., Ltd.
RTL-8139/8139C/8139C+ (rev 10)
    Subsystem: Realtek Semiconductor Co., Ltd. RT8139
    Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr-
Stepping- SERR- FastB2B-
    Status: Cap+ 66Mhz- UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort-
<TAbort- <MAbort- >SERR- <PERR-
    Latency: 32 (8000ns min, 16000ns max)
    Interrupt: pin A routed to IRQ 10
    Region 0: I/O ports at e800 [size=256]
    Region 1: Memory at d0002000 (32-bit, non-prefetchable) [size=256]
    Capabilities: [50] Power Management version 2
    	Flags: PMEClk- DSI- D1+ D2+ AuxCurrent=375mA PME(D0-,D1+,D2+,D3hot+,D3cold+)
    	Status: D0 PME-Enable- DSel=0 DScale=0 PME-

00:12.0 ISA bridge: Cyrix Corporation 5530 Legacy [Kahlua] (rev 30)
    Control: I/O+ Mem+ BusMaster+ SpecCycle+ MemWINV+ VGASnoop- ParErr-
Stepping- SERR- FastB2B-
    Status: Cap- 66Mhz- UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort-
<TAbort- <MAbort- >SERR- <PERR-
    Latency: 64, Cache Line Size 04

00:12.1 Bridge: Cyrix Corporation 5530 SMI [Kahlua]
    Control: I/O- Mem+ BusMaster- SpecCycle- MemWINV- VGASnoop- ParErr-
Stepping- SERR- FastB2B-
    Status: Cap- 66Mhz- UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort-
<TAbort- <MAbort- >SERR- <PERR-
    Region 0: Memory at 40012000 (32-bit, non-prefetchable) [size=256]

00:12.2 IDE interface: Cyrix Corporation 5530 IDE [Kahlua] (prog-if 80 [Master])
    Control: I/O+ Mem- BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr-
Stepping- SERR- FastB2B-
    Status: Cap- 66Mhz- UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort-
<TAbort- <MAbort- >SERR- <PERR-
    Latency: 0
    Region 4: I/O ports at f000 [size=16]

00:12.3 Multimedia audio controller: Cyrix Corporation 5530 Audio [Kahlua]
    Control: I/O- Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr-
Stepping- SERR- FastB2B-
    Status: Cap- 66Mhz- UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort-
<TAbort- <MAbort- >SERR- <PERR-
    Latency: 0
    Region 0: Memory at 40011000 (32-bit, non-prefetchable) [size=128]

00:12.4 VGA compatible controller: Cyrix Corporation 5530 Video
[Kahlua] (prog-if 00 [VGA])
    Subsystem: Cyrix Corporation: Unknown device 0001
    Control: I/O+ Mem+ BusMaster- SpecCycle- MemWINV- VGASnoop- ParErr-
Stepping- SERR- FastB2B-
    Status: Cap- 66Mhz- UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort-
<TAbort- <MAbort- >SERR- <PERR-
    Region 0: Memory at 40800000 (32-bit, non-prefetchable) [size=8M]

00:13.0 USB Controller: Compaq Computer Corporation ZFMicro Chipset
USB (rev 06) (prog-if 10 [OHCI])
    Subsystem: Compaq Computer Corporation ZFMicro Chipset USB
    Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr-
Stepping- SERR- FastB2B-
    Status: Cap- 66Mhz- UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort-
<TAbort- <MAbort- >SERR- <PERR-
    Latency: 32 (20000ns max), Cache Line Size 08
    Interrupt: pin A routed to IRQ 11
    Region 0: Memory at d0006000 (32-bit, non-prefetchable) [size=4K]

[Stefan Reinauer]

* Luis Correia  [060923 10:20]:
> There is no ACPI support.

A spectre is haunting people coping with bios - the spectre of ACPI.

In LinuxBIOS there _is_ ACPI support. But not enabled for all boards,
because it is usually _not required_.
> From what I've read, making ACPI tables is not easy...
>
> The board is industrial grade, and it can be found here:
> http://www.icpamerica.com/products/single_board_computers/5_25_NOVA/NOVA-489...

GX1 is basically supported.
> if I don't specify a device in config.lb, Linux will still be able

Depends :-) Some devices are not easily detectable or need
initialization in LinuxBIOS. Those have to be specified.
> about IRQ sharing, what is the best solution

IRQ routing mainly depends on the wires on the motherboards. It may or
may not be possible physically.

Why do you want to share IRQs. In theory this impacts system performance
quite significantly and is not required with today's APICs.

Does the GX1 have an APIC?
> is it possible to enable ACPI without having the tables from the
>
> factory bios

Yes, definitely. We are not using any factory bios ACPI tables anywhere
LinuxBIOS.

There is no GX1 ACPI tables available in LinuxBIOS yet, so someone would
have to write them. Not too hard, depending on what features you expect
from ACPI.

Stefan

[Luis Correia]

Hi!

On 9/23/06, Stefan Reinauer  wrote:
> Luis Correia  [060923 10:20]:
> > There is no ACPI support.
>
> A spectre is haunting people coping with bios - the spectre of ACPI.
>
> In LinuxBIOS there _is_ ACPI support. But not enabled for all boards,
> because it is usually _not required_.
> > From what I've read, making ACPI tables is not easy...
> >
> > The board is industrial grade, and it can be found here:
> > http://www.icpamerica.com/products/single_board_computers/5_25_NOVA/NOVA-489...
>
> GX1 is basically supported.
> > if I don't specify a device in config.lb, Linux will still be able
>
> Depends :-) Some devices are not easily detectable or need
> initialization in LinuxBIOS. Those have to be specified.
> > about IRQ sharing, what is the best solution
>
> IRQ routing mainly depends on the wires on the motherboards. It may or
> may not be possible physically.

Let's just say that if I have an additional eth card in the PCI slot,
with Redhat 9 if I get all the ethernet cards up and then it panics
complaining about IRQ's.

Therefore I am assuming there are IRQ related issues :)
> Why do you want to share IRQs. In theory this impacts system performance
> quite significantly and is not required with today's APICs.
>
> Does the GX1 have an APIC?

I have no idea, can't you tell that by looking at the lspci output?
> > is it possible to enable ACPI without having the tables from the
> >
> > factory bios
>
> Yes, definitely. We are not using any factory bios ACPI tables anywhere
> LinuxBIOS.
>
> There is no GX1 ACPI tables available in LinuxBIOS yet, so someone would
> have to write them. Not too hard, depending on what features you expect
> from ACPI.

Well, I would just like it to work... This is a perfect hardware
platform for my router project
<pub>Bering uClibc http://www.leaf-project.org/bering-uclibc</pub>
> Stefan

Luis

[yhlu]

why acpi?

You may add some assign_irq in irq_table.c in MB dir. It will set the
IRQ to Device pci conf space. Kernel driver then get irq from that.

YH

On 9/23/06, Luis Correia  wrote:
> Hi!
>
> On 9/23/06, Stefan Reinauer  wrote:
> > Luis Correia  [060923 10:20]:
> > > There is no ACPI support.
> >
> > A spectre is haunting people coping with bios - the spectre of ACPI.
> >
> > In LinuxBIOS there _is_ ACPI support. But not enabled for all boards,
> > because it is usually _not required_.
> > > From what I've read, making ACPI tables is not easy...
> > >
> > > The board is industrial grade, and it can be found here:
> > > http://www.icpamerica.com/products/single_board_computers/5_25_NOVA/NOVA-489...
> >
> > GX1 is basically supported.
> > > if I don't specify a device in config.lb, Linux will still be able
> >
> > Depends :-) Some devices are not easily detectable or need
> > initialization in LinuxBIOS. Those have to be specified.
> > > about IRQ sharing, what is the best solution
> >
> > IRQ routing mainly depends on the wires on the motherboards. It may or
> > may not be possible physically.
>
> Let's just say that if I have an additional eth card in the PCI slot,
> with Redhat 9 if I get all the ethernet cards up and then it panics
> complaining about IRQ's.
>
> Therefore I am assuming there are IRQ related issues :)
> > Why do you want to share IRQs. In theory this impacts system performance
> > quite significantly and is not required with today's APICs.
> >
> > Does the GX1 have an APIC?
>
> I have no idea, can't you tell that by looking at the lspci output?
> > > is it possible to enable ACPI without having the tables from the
> > >
> > > factory bios
> >
> > Yes, definitely. We are not using any factory bios ACPI tables anywhere
> > LinuxBIOS.
> >
> > There is no GX1 ACPI tables available in LinuxBIOS yet, so someone would
> > have to write them. Not too hard, depending on what features you expect
> > from ACPI.
>
> Well, I would just like it to work... This is a perfect hardware
> platform for my router project
> <pub>Bering uClibc http://www.leaf-project.org/bering-uclibc</pub>
> > Stefan
>
> Luis

[Luis Correia]

Hi!

On 9/23/06, yhlu  wrote:
> why acpi?

maybe i'm too used to fix things by turning ACPI on... but wait, that
is with windows...
> You may add some assign_irq in irq_table.c in MB dir. It will set the
> IRQ to Device pci conf space. Kernel driver then get irq from that.

but i must make sure the actual irq is wired to that device or else
things could go wrong
> YH

New subject, how about vga bios, is this GX1 fully supported now? i've
read some comments about it in the ml archives, but couldn't conclude
anything.

And the factory bios also has something called a VSA, it appears to be
some sort of firmware or driver. What is the best descriprion for this
peice of code in the bios?

Well, it is probably best for me to build one bios and then analise
the results...

Luis
> On 9/23/06, Luis Correia  wrote:
> > Hi!
> >
> > On 9/23/06, Stefan Reinauer  wrote:
> > > Luis Correia  [060923 10:20]:
> > > > There is no ACPI support.
> > >
> > > A spectre is haunting people coping with bios - the spectre of ACPI.
> > >
> > > In LinuxBIOS there _is_ ACPI support. But not enabled for all boards,
> > > because it is usually _not required_.
> > > > From what I've read, making ACPI tables is not easy...
> > > >
> > > > The board is industrial grade, and it can be found here:
> > > > http://www.icpamerica.com/products/single_board_computers/5_25_NOVA/NOVA-489...
> > >
> > > GX1 is basically supported.
> > > > if I don't specify a device in config.lb, Linux will still be able
> > >
> > > Depends :-) Some devices are not easily detectable or need
> > > initialization in LinuxBIOS. Those have to be specified.
> > > > about IRQ sharing, what is the best solution
> > >
> > > IRQ routing mainly depends on the wires on the motherboards. It may or
> > > may not be possible physically.
> >
> > Let's just say that if I have an additional eth card in the PCI slot,
> > with Redhat 9 if I get all the ethernet cards up and then it panics
> > complaining about IRQ's.
> >
> > Therefore I am assuming there are IRQ related issues :)
> > > Why do you want to share IRQs. In theory this impacts system performance
> > > quite significantly and is not required with today's APICs.
> > >
> > > Does the GX1 have an APIC?
> >
> > I have no idea, can't you tell that by looking at the lspci output?
> > > > is it possible to enable ACPI without having the tables from the
> > > >
> > > > factory bios
> > >
> > > Yes, definitely. We are not using any factory bios ACPI tables anywhere
> > > LinuxBIOS.
> > >
> > > There is no GX1 ACPI tables available in LinuxBIOS yet, so someone would
> > > have to write them. Not too hard, depending on what features you expect
> > > from ACPI.
> >
> > Well, I would just like it to work... This is a perfect hardware
> > platform for my router project
> > <pub>Bering uClibc http://www.leaf-project.org/bering-uclibc</pub>
> > > Stefan
> >
> > Luis