Re: [comedi] MITE transfer error with the NI PCIe card

Hi '=C3=89ric,

On 16/05/2022 15:51, '=C3=89ric Piel' via Comedi: Linux Control and=20
Measurement Device Interface wrote:
> Hi,
>=20
> We have several systems which experience the same error from time to=20
> time. During an acquisition (analog input), the read fails. Then, any=20
> following commands to the board fails with "input/output error" (5). The=
=20
> only way to restore the functionality is to power off the computer, wait=
=20
> a few seconds, and power it on again.
>=20
> On the time of the error, I see such message on the kernel side:
> [ 2843.893823] comedi comedi0: mite: transfer error ffffffff
> [ 2843.893856] comedi comedi0: ni_tio_handle_interrupt: Gi_Gate_Error=20
> detected.
> [ 2843.893861] comedi comedi0: ni_tio_handle_interrupt: Gi_DRQ_Error=20
> detected.
> [ 2843.893867] comedi comedi0: ni_tio_handle_interrupt: Gi_Gate_Error=20
> detected.
> [ 2843.893870] comedi comedi0: ni_tio_handle_interrupt: Gi_DRQ_Error=20
> detected.
> [ 2845.107044] comedi comedi0: timeout loading channel/gain list
> [ 2845.109086] comedi comedi0: timed out waiting for=20
> AO_TMRDACWRs_In_Progress_St to clear
> [ 2845.615575] comedi comedi0: timeout loading channel/gain list
> [ 2845.617689] comedi comedi0: timed out waiting for=20
> AO_TMRDACWRs_In_Progress_St to clear
> [ 2846.123539] comedi comedi0: timeout loading channel/gain list
> [ 2846.125651] comedi comedi0: timed out waiting for=20
> AO_TMRDACWRs_In_Progress_St to clear
> [ 2846.631557] comedi comedi0: timeout loading channel/gain list
> [ 2846.633673] comedi comedi0: timed out waiting for=20
> AO_TMRDACWRs_In_Progress_St to clear
>=20
> Our systems are x86-64 with Ubuntu 18.04 (Linux 5.4) or Ubuntu 20.04=20
> (Linux 5.13). The errors are quite hard to reproduce. On some systems,=20
> it doesn't seem to ever happen, on some others, it's only once in a=20
> month, and few systems experience it almost daily. All the boards are NI=
=20
> PCIe-6251.
>=20
> Any idea of what a "MITE transfer error" is? Is there any way we could=20
> recover from it in an easier way? Could it be a sign of a hardware=20
> issue? Any hint is welcome!

The "MITE transfer error" is a side effect of reading an unexpected=20
value from the MITE status register. Reading ffffffff unexpectedly from=20
a PCI device register is often caused by an unreliable PCI connection,=20
but in this case I'm not so sure because it is affecting several of your=20
systems.

--=20
-=3D( Ian Abbott <[email protected]> || MEV Ltd. is a company  )=3D-
-=3D( registered in England & Wales.  Regd. number: 02862268.  )=3D-
-=3D( Regd. addr.: S11 & 12 Building 67, Europa Business Park, )=3D-
-=3D( Bird Hall Lane, STOCKPORT, SK3 0XA, UK. || www.mev.co.uk )=3D-

--=20
You received this message because you are subscribed to the Google Groups "=
Comedi: Linux Control and Measurement Device Interface" group.
To unsubscribe from this group and stop receiving emails from it, send an e=
mail to [email protected].
To view this discussion on the web visit https://groups.google.com/d/msgid/=
comedi_list/d2a923a2-a1a5-0a94-b52b-eaab157f1435%40mev.co.uk.