Re: [RFC PATCH 0/4] tests/acceptance: Test U-Boot/Linux from Armbian 20.08 on Orange Pi PC

[Bin Meng]

Hi Philippe,

On Sat, Oct 24, 2020 at 1:56 AM Philippe Mathieu-Daudé <f4bug@amsat.org> wrote:
>
> On 10/23/20 7:42 PM, Bin Meng wrote:
> > Hi Philippe,
> >
> > On Fri, Oct 23, 2020 at 9:18 PM Philippe Mathieu-Daudé <f4bug@amsat.org> 
> > wrote:
> >>
> >> Series meant to help Bin Meng to debug the SD card issue
> >> reported by Michael Roth.
> >
> > Thank you for the patches.
> >
> >>
> >> Philippe Mathieu-Daudé (4):
> >>    Revert "hw/sd: Fix incorrect populated function switch status data
> >>      structure"
> >>    tests/acceptance: Allow running Orange Pi test using cached artifacts
> >>    tests/acceptance: Extract do_test_arm_orangepi_armbian_uboot() method
> >>    tests/acceptance: Test U-Boot/Linux from Armbian 20.08 on Orange Pi PC
> >>
> >>   hw/sd/sd.c                             |  3 +-
> >>   tests/acceptance/boot_linux_console.py | 68 +++++++++++++++++++-------
> >>   2 files changed, 50 insertions(+), 21 deletions(-)
> >
> > With this series, I used:
> >
> > $ ARMBIAN_ARTIFACTS_CACHED=1 AVOCADO_ALLOW_LARGE_STORAGE=1 make 
> > check-acceptance
> >
> > It looks that the failure still exists? Log below:
> >
> > 13-tests_acceptance_boot_linux_console.py_BootLinuxConsole.test_arm_orangepi_bionic_20_08/debug.log:
> >
> > 01:11:27 DEBUG| => boot
> > 01:11:27 DEBUG| unable to select a mode
> > 01:11:27 DEBUG| Device 0: unknown device
> > 01:11:27 DEBUG| BOOTP broadcast 1
> > 01:11:27 DEBUG| DHCP client bound to address 10.0.2.15 (1 ms)
> > 01:11:27 DEBUG| *** Warning: no boot file name; using '0A00020F.img'
> > 01:11:27 DEBUG| Using ethernet@1c30000 device
> > 01:11:27 DEBUG| TFTP from server 10.0.2.2; our IP address is 10.0.2.15
> > 01:11:27 DEBUG| Filename '0A00020F.img'.
> > 01:11:27 DEBUG| Load address: 0x42000000
> > 01:11:27 DEBUG| Loading: *^H
> > 01:11:27 DEBUG| TFTP error: 'Access violation' (2)
> > 01:11:27 DEBUG| Not retrying...
>
> Have you rebuilt qemu-system-arm with the reverted patch included?

Oops, I took it for granted that the `make check-acceptance` will
automatically rebuild the QEMU binary, which is not the case. Should
we enforce the rebuild before testing in Makefiles?

Regards,
Bin