Re: [Minios-devel] [UNIKRAFT early RFC PATCH 00/11] Enble early SMP support on arm64 kvm plat

["Justin He (Arm Technology China)" <Justin.He@xxxxxxx>]

Hi Julien
Thanks for the comments.
I will record all of your comments about smp support series and address them 
after a while in next version.

--
Cheers,
Justin (Jia He)



> -----Original Message-----
> From: Julien Grall <julien.grall@xxxxxxx>
> Sent: 2019年6月21日 23:33
> To: Justin He (Arm Technology China) <Justin.He@xxxxxxx>; minios-
> devel@xxxxxxxxxxxxxxxxxxxx; Simon Kuenzer <simon.kuenzer@xxxxxxxxx>;
> Sharan.Santhanam@xxxxxxxxx
> Cc: Florian Schmidt <florian.schmidt@xxxxxxxxx>; Felipe Huici
> <felipe.huici@xxxxxxxxx>; yuri.volchkov@xxxxxxxxx; Kaly Xin (Arm
> Technology China) <Kaly.Xin@xxxxxxx>
> Subject: Re: [UNIKRAFT early RFC PATCH 00/11] Enble early SMP support on
> arm64 kvm plat
>
> Hi,
>
> Thank you for sending the series.
>
> On 21/06/2019 07:57, Jia He wrote:
> > This is the early SMP support on arm64 kvm plat which based on previous
> series #1-#5
>
> A reviewer may have no idea what are "series #1 - #5". So I would
> recommend to
> give pointer to each of them. Also, for easing review, it is nice to have a 
> repo
> with everything applied.
>
> But I think basing your work on 6 different series is not really wise. For a
> first, there was a lot of comments on those series. Does it mean this is 
> before
> or after them addressed?
>
> Secondly this is making more difficult to go through the code and potentially
> for you to address comment. May I recommend to focus on the getting series
> #1-#5
> to be merged?
>
> >
> > Most of the codes refer to the implementation in freebsd.
>
> Some general comments here. I know this is an early RFC, but I will point out
> for the next series :).
>
> 1) Please be more verbose in the commit message. This is helpful for
> reviewer
> and future developers to understand what the patch does
> 2) You probably want to make the GIC code SMP safe. Also, the GIC
> CPU ID may be
> different from the MIDR (you don't want to route interrupt to the wrong
> CPU ;)).
> 3) It looks like spinlock is pretty much a NOP today. Is there any plan
> to
> address this?
>
> Cheers,
>
> >
> > Jia He (11):
> >    lib/nolibc: Introduce ndelay and mdelay for busy wait
> >    plat/common/ofw: Move fdt_reg_read_number to header file for static
> >      inline
> >    plat/kvm/smp: Add psci call definitions in psci.h
> >    plat/kvm/smp: Add smp definitions in smp.h
> >    plat/kvm/smp: Add psci call psci_cpu_on
> >    plat/kvm: arm64: Add a parameter to start_mmu
> >    plat/kvm: arm64: Add entry point for APs (secondary cpus)
> >    plat/kvm/arm: Implement smp boot on arm64 kvm plat
> >    plat/kvm/arm: Add simple percpu variable support
> >    lib/ukboot: Enable CONFIG_SMP for arm64
> >    plat/kvm/arm: Enable scheduler on secondary APs
> >
> >   lib/nolibc/exportsyms.uk       |   4 +
> >   lib/nolibc/include/time.h      |   3 +
> >   lib/nolibc/time.c              |  16 ++-
> >   lib/ukboot/Config.uk           |   5 +
> >   lib/ukboot/Makefile.uk         |   1 +
> >   lib/ukboot/boot.c              |  15 +++
> >   plat/common/include/arm/psci.h |  87 ++++++++++++
> >   plat/common/include/smp.h      |  49 +++++++
> >   plat/drivers/include/ofw/fdt.h |  17 ++-
> >   plat/drivers/ofw/fdt.c         |  17 +--
> >   plat/kvm/Makefile.uk           |   1 +
> >   plat/kvm/arm/entry64.S         |  31 +++++
> >   plat/kvm/arm/pagetable64.S     |   7 +-
> >   plat/kvm/arm/setup.c           | 233 +++++++++++++++++++++++++++++++++
> >   14 files changed, 466 insertions(+), 20 deletions(-)
> >   create mode 100644 plat/common/include/arm/psci.h
> >   create mode 100644 plat/common/include/smp.h
> >
>
> --
> Julien Grall
IMPORTANT NOTICE: The contents of this email and any attachments are 
confidential and may also be privileged. If you are not the intended recipient, 
please notify the sender immediately and do not disclose the contents to any 
other person, use it for any purpose, or store or copy the information in any 
medium. Thank you.
_______________________________________________
Minios-devel mailing list
Minios-devel@xxxxxxxxxxxxxxxxxxxx
https://lists.xenproject.org/mailman/listinfo/minios-devel