Radeon RV100 init fails with specific card+mobo - IRQ problem

Radeon RV100 init fails with specific card+mobo - IRQ problem

[Meelis Roos]

I have a test PC with AMD 760MPX chipset (Tyan S2466 Tiger MPX board) 
and Radeon RV100-based AGP card where radeon KMS modesetting does not 
work (does not work with any kernel tried so far (2.6.32, 3.2, 3.7, 
3.8-rc6).

>From dmesg with drm.debug=0xc 

[    3.209617] agpgart-amdk7 0000:00:00.0: AMD 760MP chipset
[    3.325408] agpgart-amdk7 0000:00:00.0: AGP aperture is 512M @ 0xc0000000
[...]
[    3.529106] [drm] Initialized drm 1.1.0 20060810
[    3.797715] [drm] radeon defaulting to kernel modesetting.
[    3.797817] [drm] radeon kernel modesetting enabled.
[    3.799108] [drm] initializing kernel modesetting (RV100 0x1002:0x5159 0x1787:0x0F02).
[    3.799277] [drm] register mmio base: 0xB2000000
[    3.799354] [drm] register mmio size: 65536
[    3.799629] agpgart-amdk7 0000:00:00.0: AGP 2.0 bridge
[    3.799718] agpgart-amdk7 0000:00:00.0: putting AGP V2 device into 4x mode
[    3.799811] radeon 0000:01:05.0: putting AGP V2 device into 4x mode
[    3.799900] radeon 0000:01:05.0: GTT: 512M 0xC0000000 - 0xDFFFFFFF
[    3.799985] radeon 0000:01:05.0: VRAM: 128M 0x00000000E0000000 - 0x00000000E7FFFFFF (64M used)
[    3.800187] [drm] Supports vblank timestamp caching Rev 1 (10.10.2010).
[    3.800268] [drm] Driver supports precise vblank timestamp query.
[    3.800350] radeon 0000:01:05.0: Fatal error during GPU init
[    3.800432] [drm] radeon: finishing device.
[    3.800509] [TTM] Memory type 2 has not been initialized
[    3.800590] [drm] radeon: cp finalized
[    3.800756] radeon: probe of 0000:01:05.0 failed with error -22


01:05.0 VGA compatible controller [0300]: Advanced Micro Devices [AMD] nee ATI RV100 QY [Radeon 7000/VE] [1002:5159] (prog-if 00 [VGA controller])
        Subsystem: Hightech Information System Ltd. Device [1787:0f02]
        Control: I/O+ Mem+ BusMaster- SpecCycle- MemWINV- VGASnoop- ParErr- Stepping+ SERR- FastB2B+ DisINTx-
        Status: Cap+ 66MHz+ UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
        Region 0: Memory at e0000000 (32-bit, prefetchable) [size=128M]
        Region 1: I/O ports at 2000 [size=256]
        Region 2: Memory at b2000000 (32-bit, non-prefetchable) [size=64K]
        [virtual] Expansion ROM at b2020000 [disabled] [size=128K]
        Capabilities: [58] AGP version 2.0
                Status: RQ=48 Iso- ArqSz=0 Cal=0 SBA+ ITACoh- GART64- HTrans- 64bit- FW- AGP3- Rate=x1,x2,x4
                Command: RQ=16 ArqSz=0 Cal=0 SBA+ AGP+ GART64- 64bit- FW- Rate=x4
        Capabilities: [50] Power Management version 2
                Flags: PMEClk- DSI- D1+ D2+ AuxCurrent=0mA PME(D0-,D1-,D2-,D3hot-,D3cold-)
                Status: D0 NoSoftRst- PME-Enable- DSel=0 DScale=0 PME-

Adding some printks reveals it calls
r100_init() -> radeon_irq_kms_init() -> drm_irq_install() -> 
drm_dev_to_irq() and that fails. So no IRQ, no KMS.

lspci does not show Interrupt like on some other PCI devices. 
/proc/interrupts obviously does not contain the irq since we do not 
register any yet.

There is no BIOS option of "Enable IRQ for VGA" or similar (BIOS is 
latest). Could this be the problem?

What can I do to get KMS working on this computer?

-- 
Meelis Roos (mroos@linux.ee)

Re: Radeon RV100 init fails with specific card+mobo - IRQ problem

[Jerome Glisse]

On Fri, Feb 1, 2013 at 7:24 AM, Meelis Roos <mroos@linux.ee> wrote:
> I have a test PC with AMD 760MPX chipset (Tyan S2466 Tiger MPX board)
> and Radeon RV100-based AGP card where radeon KMS modesetting does not
> work (does not work with any kernel tried so far (2.6.32, 3.2, 3.7,
> 3.8-rc6).
>
> >From dmesg with drm.debug=0xc
>
> [    3.209617] agpgart-amdk7 0000:00:00.0: AMD 760MP chipset
> [    3.325408] agpgart-amdk7 0000:00:00.0: AGP aperture is 512M @ 0xc0000000
> [...]
> [    3.529106] [drm] Initialized drm 1.1.0 20060810
> [    3.797715] [drm] radeon defaulting to kernel modesetting.
> [    3.797817] [drm] radeon kernel modesetting enabled.
> [    3.799108] [drm] initializing kernel modesetting (RV100 0x1002:0x5159 0x1787:0x0F02).
> [    3.799277] [drm] register mmio base: 0xB2000000
> [    3.799354] [drm] register mmio size: 65536
> [    3.799629] agpgart-amdk7 0000:00:00.0: AGP 2.0 bridge
> [    3.799718] agpgart-amdk7 0000:00:00.0: putting AGP V2 device into 4x mode
> [    3.799811] radeon 0000:01:05.0: putting AGP V2 device into 4x mode
> [    3.799900] radeon 0000:01:05.0: GTT: 512M 0xC0000000 - 0xDFFFFFFF
> [    3.799985] radeon 0000:01:05.0: VRAM: 128M 0x00000000E0000000 - 0x00000000E7FFFFFF (64M used)
> [    3.800187] [drm] Supports vblank timestamp caching Rev 1 (10.10.2010).
> [    3.800268] [drm] Driver supports precise vblank timestamp query.
> [    3.800350] radeon 0000:01:05.0: Fatal error during GPU init
> [    3.800432] [drm] radeon: finishing device.
> [    3.800509] [TTM] Memory type 2 has not been initialized
> [    3.800590] [drm] radeon: cp finalized
> [    3.800756] radeon: probe of 0000:01:05.0 failed with error -22
>
>
> 01:05.0 VGA compatible controller [0300]: Advanced Micro Devices [AMD] nee ATI RV100 QY [Radeon 7000/VE] [1002:5159] (prog-if 00 [VGA controller])
>         Subsystem: Hightech Information System Ltd. Device [1787:0f02]
>         Control: I/O+ Mem+ BusMaster- SpecCycle- MemWINV- VGASnoop- ParErr- Stepping+ SERR- FastB2B+ DisINTx-
>         Status: Cap+ 66MHz+ UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
>         Region 0: Memory at e0000000 (32-bit, prefetchable) [size=128M]
>         Region 1: I/O ports at 2000 [size=256]
>         Region 2: Memory at b2000000 (32-bit, non-prefetchable) [size=64K]
>         [virtual] Expansion ROM at b2020000 [disabled] [size=128K]
>         Capabilities: [58] AGP version 2.0
>                 Status: RQ=48 Iso- ArqSz=0 Cal=0 SBA+ ITACoh- GART64- HTrans- 64bit- FW- AGP3- Rate=x1,x2,x4
>                 Command: RQ=16 ArqSz=0 Cal=0 SBA+ AGP+ GART64- 64bit- FW- Rate=x4
>         Capabilities: [50] Power Management version 2
>                 Flags: PMEClk- DSI- D1+ D2+ AuxCurrent=0mA PME(D0-,D1-,D2-,D3hot-,D3cold-)
>                 Status: D0 NoSoftRst- PME-Enable- DSel=0 DScale=0 PME-
>
> Adding some printks reveals it calls
> r100_init() -> radeon_irq_kms_init() -> drm_irq_install() ->
> drm_dev_to_irq() and that fails. So no IRQ, no KMS.
>
> lspci does not show Interrupt like on some other PCI devices.
> /proc/interrupts obviously does not contain the irq since we do not
> register any yet.
>
> There is no BIOS option of "Enable IRQ for VGA" or similar (BIOS is
> latest). Could this be the problem?
>
> What can I do to get KMS working on this computer?
>
> --
> Meelis Roos (mroos@linux.ee)

Try booting with radeon.agp_mode=-1

Cheers,
Jerome

Re: Radeon RV100 init fails with specific card+mobo - IRQ problem

[Meelis Roos]

> > Adding some printks reveals it calls
> > r100_init() -> radeon_irq_kms_init() -> drm_irq_install() ->
> > drm_dev_to_irq() and that fails. So no IRQ, no KMS.
> >
> > lspci does not show Interrupt like on some other PCI devices.
> > /proc/interrupts obviously does not contain the irq since we do not
> > register any yet.
> >
> > There is no BIOS option of "Enable IRQ for VGA" or similar (BIOS is
> > latest). Could this be the problem?
> >
> > What can I do to get KMS working on this computer?
> >
> > --
> > Meelis Roos (mroos@linux.ee)
> 
> Try booting with radeon.agp_mode=-1

radeon: `-1' invalid for parameter `agp_mode'

-- 
Meelis Roos (mroos@linux.ee)

Re: Radeon RV100 init fails with specific card+mobo - IRQ problem

[Alex Deucher]

On Fri, Feb 1, 2013 at 10:26 AM, Meelis Roos <mroos@linux.ee> wrote:
>> > Adding some printks reveals it calls
>> > r100_init() -> radeon_irq_kms_init() -> drm_irq_install() ->
>> > drm_dev_to_irq() and that fails. So no IRQ, no KMS.
>> >
>> > lspci does not show Interrupt like on some other PCI devices.
>> > /proc/interrupts obviously does not contain the irq since we do not
>> > register any yet.
>> >
>> > There is no BIOS option of "Enable IRQ for VGA" or similar (BIOS is
>> > latest). Could this be the problem?
>> >
>> > What can I do to get KMS working on this computer?
>> >
>> > --
>> > Meelis Roos (mroos@linux.ee)
>>
>> Try booting with radeon.agp_mode=-1
>
> radeon: `-1' invalid for parameter `agp_mode'

radeon.agpmode=-1

However, I think there are a number of places where we require
interrupts that would need to be worked around (fences, display
hotplug, pageflipping) if not interrupt is available.

Alex

Re: Radeon RV100 init fails with specific card+mobo - IRQ problem

[Meelis Roos]

> >> > Adding some printks reveals it calls
> >> > r100_init() -> radeon_irq_kms_init() -> drm_irq_install() ->
> >> > drm_dev_to_irq() and that fails. So no IRQ, no KMS.
> >> >
> >> > lspci does not show Interrupt like on some other PCI devices.
> >> > /proc/interrupts obviously does not contain the irq since we do not
> >> > register any yet.
> >> >
> >> > There is no BIOS option of "Enable IRQ for VGA" or similar (BIOS is
> >> > latest). Could this be the problem?
> >> >
> >> > What can I do to get KMS working on this computer?
> >> >
> >> > --
> >> > Meelis Roos (mroos@linux.ee)
> >>
> >> Try booting with radeon.agp_mode=-1
> >
> > radeon: `-1' invalid for parameter `agp_mode'
> 
> radeon.agpmode=-1

Still fails in the same place as in the beginning - finding IRQ.
 
> However, I think there are a number of places where we require
> interrupts that would need to be worked around (fences, display
> hotplug, pageflipping) if not interrupt is available.

OK.

-- 
Meelis Roos (mroos@linux.ee)

Re: Radeon RV100 init fails with specific card+mobo - IRQ problem

[Meelis Roos]

> >> > Adding some printks reveals it calls
> >> > r100_init() -> radeon_irq_kms_init() -> drm_irq_install() ->
> >> > drm_dev_to_irq() and that fails. So no IRQ, no KMS.
> >> >
> >> > lspci does not show Interrupt like on some other PCI devices.
> >> > /proc/interrupts obviously does not contain the irq since we do not
> >> > register any yet.
> >> >
> >> > There is no BIOS option of "Enable IRQ for VGA" or similar (BIOS is
> >> > latest). Could this be the problem?
> >> >
> >> > What can I do to get KMS working on this computer?
> >> >
> >> > --
> >> > Meelis Roos (mroos@linux.ee)
> >>
> >> Try booting with radeon.agp_mode=-1
> >
> > radeon: `-1' invalid for parameter `agp_mode'
> 
> radeon.agpmode=-1
> 
> However, I think there are a number of places where we require
> interrupts that would need to be worked around (fences, display
> hotplug, pageflipping) if not interrupt is available.

I did dome more testing with different cards an got a surprise. The 
machine works fine with other AGP cards - one MX440 with nouvea, another 
RV100 QY and a R200. They do have interrupt in lspci -vvv output.

Why does this specific card not have its IRQ routed - card ROM problem?

-- 
Meelis Roos (mroos@linux.ee)

Re: Radeon RV100 init fails with specific card+mobo - IRQ problem

[Alex Deucher]

On Mon, Feb 4, 2013 at 11:32 AM, Meelis Roos <mroos@linux.ee> wrote:
>> >> > Adding some printks reveals it calls
>> >> > r100_init() -> radeon_irq_kms_init() -> drm_irq_install() ->
>> >> > drm_dev_to_irq() and that fails. So no IRQ, no KMS.
>> >> >
>> >> > lspci does not show Interrupt like on some other PCI devices.
>> >> > /proc/interrupts obviously does not contain the irq since we do not
>> >> > register any yet.
>> >> >
>> >> > There is no BIOS option of "Enable IRQ for VGA" or similar (BIOS is
>> >> > latest). Could this be the problem?
>> >> >
>> >> > What can I do to get KMS working on this computer?
>> >> >
>> >> > --
>> >> > Meelis Roos (mroos@linux.ee)
>> >>
>> >> Try booting with radeon.agp_mode=-1
>> >
>> > radeon: `-1' invalid for parameter `agp_mode'
>>
>> radeon.agpmode=-1
>>
>> However, I think there are a number of places where we require
>> interrupts that would need to be worked around (fences, display
>> hotplug, pageflipping) if not interrupt is available.
>
> I did dome more testing with different cards an got a surprise. The
> machine works fine with other AGP cards - one MX440 with nouvea, another
> RV100 QY and a R200. They do have interrupt in lspci -vvv output.
>
> Why does this specific card not have its IRQ routed - card ROM problem?

For some reason the system bios on your box doesn't assign it one.
The video bios isn't really involved.  I'm not sure if there is a way
to force assign an irq in Linux or not.

Alex