Re: [PATCH] riscv: sifive_u: Add a "serial" property for board serial number

[Bin Meng]

Hi Palmer,

On Tue, Feb 11, 2020 at 3:55 AM Palmer Dabbelt <address@hidden> wrote:
>
> On Wed, 29 Jan 2020 07:29:11 PST (-0800), Palmer Dabbelt wrote:
> > On Fri, 10 Jan 2020 07:52:05 GMT (+0000), address@hidden wrote:
> >> Hi Palmer,
> >>
> >> On Fri, Nov 22, 2019 at 10:38 AM Palmer Dabbelt
> >> <address@hidden> wrote:
> >>>
> >>> On Thu, 21 Nov 2019 17:10:18 PST (-0800), address@hidden wrote:
> >>> > On Sat, Nov 16, 2019 at 11:08 PM Bin Meng <address@hidden> wrote:
> >>> >>
> >>> >> At present the board serial number is hard-coded to 1, and passed
> >>> >> to OTP model during initialization. Firmware (FSBL, U-Boot) uses
> >>> >> the serial number to generate a unique MAC address for the on-chip
> >>> >> ethernet controller. When multiple QEMU 'sifive_u' instances are
> >>> >> created and connected to the same subnet, they all have the same
> >>> >> MAC address hence it creates a unusable network.
> >>> >>
> >>> >> A new "serial" property is introduced to specify the board serial
> >>> >> number. When not given, the default serial number 1 is used.
> >>> >>
> >>> >> Signed-off-by: Bin Meng <address@hidden>
> >>> >> ---
> >>> >>
> >>> >>  hw/riscv/sifive_u.c         | 21 ++++++++++++++++++++-
> >>> >>  include/hw/riscv/sifive_u.h |  1 +
> >>> >>  2 files changed, 21 insertions(+), 1 deletion(-)
> >>> >>
> >>> >
> >>> > ping?
> >>>
> >>> Sorry, it looks like I dropped this one.  I've put it in the queue for 
> >>> 5.0,
> >>> with a
> >>>
> >>> Reviewed-by: Palmer Dabbelt <address@hidden>
> >>
> >> Has this been applied somewhere?
> >
> > Weird, not sure how I managed to screw this up again.  It's actually on 
> > for-master as
> >
> >     * a828041ba6 - riscv: sifive_u: Add a "serial" property for board 
> > serial number (50 seconds ago) <Bin Meng>
> >
> > with any luck I'll manage to avoid screwing it up a third time.
>
> Ah, OK -- the issue here is that this fails "make check", specifically
>
>     $ make check-qtest-riscv64
>     make[1]: Entering directory '/home/palmerdabbelt/life/riscv/qemu/slirp'
>     make[1]: Nothing to be done for 'all'.
>     make[1]: Leaving directory '/home/palmerdabbelt/life/riscv/qemu/slirp'
>             CHK version_gen.h
>       TEST    check-qtest-riscv64: tests/qtest/cdrom-test
>       TEST    check-qtest-riscv64: tests/qtest/device-introspect-test
>     
> /home/palmerdabbelt/life/riscv/qemu/hw/riscv/sifive_u.c:406:riscv_sifive_u_soc_init:
>  Object 0x55baf3feea00 is not an instance of type sifive_u-machine
>     Broken pipe
>     tests/qtest/libqtest.c:149: kill_qemu() detected QEMU death from signal 6 
> (Aborted)
>     ERROR - too few tests run (expected 6, got 5)
>     make: *** 
> [/home/palmerdabbelt/life/riscv/qemu/tests/Makefile.include:630: 
> check-qtest-riscv64] Error 1
>
> which is probably how it kept getting disappeared -- I just forgot to reply on
> the list.  I'm going to hold it back from the PR I'm staging right now, LMK if
> you have a fix.

OK, I will take a look. I remember I did run "make check" but it did
not report any issue before. Is 'make check-qtest-riscv64' not part of
'make check'?

Regards,
Bin