qca6164 support

[Michal Kazior]

On 6 August 2015 at 18:59, Victor van Poppelen <vvanpo at gmail.com> wrote:
> Still get a firmware crash after 'boot hif start', but different
> behaviour this time.  Here is the dmesg output:
>
> [    6.058379] ath10k_pci 0000:02:00.0: pci probe
> [    6.058933] ath10k_pci 0000:02:00.0: boot pci_mem 0xffffc90002c00000
> [    6.060960] ath10k_pci 0000:02:00.0: pci irq msi-x interrupts 8
> irq_mode 0 reset_mode 0

Can you load ath10k_pci with irq_mode=1 module parameter, please?

Something broke in recent kernel for msi-ranged irqs on ath10k. Enrico
is trying to bisect.

In the meantime forcing shared interrupts is the only workaround.


[...]
> [   67.456339] ath10k_pci 0000:02:00.0: bmi done
> [   67.456368] ath10k_pci 0000:02:00.0: htt tx max num pending tx 1056
> [   67.456396] ath10k_pci 0000:02:00.0: htt rx ring size 2048 fill_level 1023
> [   67.456400] ath10k_pci 0000:02:00.0: boot hif start
> [   67.456427] ath10k_pci 0000:02:00.0: received unsolicited fw crash interrupt
> [   68.456939] ath10k_pci 0000:02:00.0: failed to receive control
> response completion, polling..
> [   68.456948] ath10k_pci 0000:02:00.0: pci hif send complete check
> [   68.457253] ath10k_pci 0000:02:00.0: pci hif send complete check
> [   68.457265] ath10k_pci 0000:02:00.0: pci rx ce pipe 1 len 20
> [   68.457274] ath10k_pci 0000:02:00.0: pci rx: 00000000: 00 00 0c 00
> 00 00 00 00 01 00 02 00 00 07 16 00  ................
> [   68.457282] ath10k_pci 0000:02:00.0: pci rx: 00000010: 01 00 00 00
>                                     ....
> [   68.457299] ath10k_pci 0000:02:00.0: pci hif send complete check
> [   68.457306] ath10k_pci 0000:02:00.0: pci hif send complete check
> [   68.457371] ath10k_pci 0000:02:00.0: pci hif send complete check
> [   68.457484] ath10k_pci 0000:02:00.0: pci hif send complete check
> [   68.457489] ath10k_pci 0000:02:00.0: pci hif send complete check
> [   68.457493] ath10k_pci 0000:02:00.0: pci hif send complete check
> [   68.457602] ath10k_pci 0000:02:00.0: Target ready! transmit
> resources: 2 size:1792

This looks like good news! HTC Control service responded with a
message to host. This means firmware code has finally run. Previously
it did not run at all - instead ROM code crashed before even starting
the uploaded firmware blob.


Michał