Re: [RFC 0/1] hw/ipmi: add an aspeed IPMI iBT device model

[Titus Rwantare <titusr@google.com>]

On Tue, 28 Sept 2021 at 23:24, Cédric Le Goater <clg@kaod.org> wrote:
>
> Hello Titus,
>
> On 9/29/21 00:39, Titus Rwantare wrote:
> > This patch follows the Handing IPMI for emulating BMC patch set by Hao Wu.
> > Building on top of the work in [PATCH] hw/misc: Add an iBT device model posted
> > by Cédric Le Goater, this iBT model works as a backend to ipmi-host-extern.
>
> Could you please resend with my patch which is still available here :
>
>   https://github.com/legoater/qemu/commit/c6b679690f32534e8992d96752d90d2c4aa48130
>
> or here :
>
>   http://patchwork.ozlabs.org/project/qemu-devel/patch/20210407171637.777743-20-clg@kaod.org/
>
> and in another patch, the modifications you made on top of mine. That
> will help understand where the problem could be.
>
> You can modify patches but you need to keep the original author name,
> commit log, signed-off-by to certify its origin. Then list the changes
> you have made and add your signed-off-by.
>
> The patch you sent seems to be mostly mine and I don't see any of the
> above.
>
> Here is some more info in the kernel documentation :
>
>    https://www.kernel.org/doc/html/v4.12/process/submitting-patches.html#sign-your-work-the-developer-s-certificate-of-origin
>
> I don't think we have the same kind of documentation in QEMU but it's
> the same process.

Thanks, I'm still learning.

>
> > Needed to run:
> > - [PATCH 4/8] hw/ipmi: Refactor IPMI interface, Hao Wu
> > - [PATCH 5/8] hw/ipmi: Take out common from ipmi_bmc_extern.c, Hao Wu
> > - [PATCH 6/8] hw/ipmi: Move handle_command to IPMICoreClass, Hao Wu
> > - [PATCH 7/8] hw/ipmi: Add an IPMI external host device, Hao Wu
>
> I have no idea what these patches do and where they are. They seem to
> modify the core IPMI framework and it is important to understand the
> overall changes. Please include them in the patchset first since they
> are initial requirements.
>

I've pushed everything here: https://github.com/Rwantare/qemu/tree/aspeed-ibt
The other patches are out for review and will likely change before merging

> > Host commandline
> >      -chardev socket,id=ipmichr1,host=localhost,port=9999,reconnect=10 \
> >      -device ipmi-bmc-extern,chardev=ipmichr1,id=bmc0 \
> >      -device isa-ipmi-bt,bmc=bmc0,irq=10 -nodefaults
> >
> > BMC commandline
> >      -chardev socket,id=ipmichr1,host=localhost,port=9999,server=on,wait=off \
> >      -device "ipmi-host-extern,chardev=ipmichr1,responder=/machine/soc/ibt"
>
> Could we have the full command lines also ?
>

Sure, there's not much else there, one of the bmc firmwares I'm testing with
is a quanta-q71l-bmc which should build from OpenBMC.

Host:
qemu-system-x86_64 \
-nographic -m 4G \
-kernel arch/x86/boot/bzImage -append "console=ttyS0 earlyprintk=ttyS0" \
-serial /dev/tty -monitor /dev/pts/9 -initrd u-root.cpio \
-chardev socket,id=ipmichr0,host=localhost,port=9999,reconnect=10 \
-device ipmi-bmc-extern,chardev=ipmichr0,id=bmc0 \
-device isa-ipmi-bt,bmc=bmc0,irq=10 -nodefaults

BMC:
qemu-system-arm \
-machine quanta-q71l-bmc \
-nographic \
-drive file=path/to/openbmc/image,if=mtd,bus=0,unit=0,snapshot=on \
-chardev socket,id=ipmichr1,host=localhost,port=9999,server=on,wait=off \
-device "ipmi-host-extern,chardev=ipmichr1,responder=/machine/soc/ibt"

>
> > Currently, IRQs are not functional, but it is able to connect.
> > The following printout is from the host:
> >
> > [    6.869742] ipmi_si IPI0001:00: IPMI message handler: Found new BMC (man_id:
> >   0x002b79, prod_id: 0x0000, dev_id: 0x00)
> > [   12.393281] ipmi_si IPI0001:00: IPMI bt interface initialized
> >
> > This patch is an RFC because of the missing IRQs and the need for other patches
> > to get merged.
>
> We will need the whole project to start reviewing.
>
> Thanks,
>
> C.

Noted, I think I should be able to send the BT commits from
https://github.com/Rwantare/qemu/tree/aspeed-ibt
after the others are reviewed and merged.
For now, the irq setup is my point of confusion.

Thanks,
-Titus