Re: [Qemu-devel] [PATCH qemu RFC 4/4] spapr: Implement SLOF-less client_architecture_support

[David Gibson <david@gibson.dropbear.id.au>]

[-- Attachment #1: Type: text/plain, Size: 7328 bytes --]

On Sat, Jul 20, 2019 at 11:28:50AM +1000, Alexey Kardashevskiy wrote:
> QEMU already implements H_CAS called by SLOF. The existing handler
> prepares a diff FDT and SLOF applies it on top of its current tree.
> In SLOF-less setup when the user explicitly selected "bios=no",
> this updates the FDT from the OS, updates it and writes back to the OS.
> The new behavior is advertised to the OS via "/chosen/qemu,h_cas".

I don't love having two different paths here, I'm wondering if we can
unify things at all.

I have wondered at some points if there's anything preventing us just
giving a whole new device tree at CAS, rather than selected updates -
that could simplify several things.

> Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
> ---
>  include/hw/ppc/spapr.h |  5 +++++
>  hw/ppc/spapr.c         | 24 ++++++++++++++++-----
>  hw/ppc/spapr_hcall.c   | 49 +++++++++++++++++++++++++++++++++++++++---
>  3 files changed, 70 insertions(+), 8 deletions(-)
> 
> diff --git a/include/hw/ppc/spapr.h b/include/hw/ppc/spapr.h
> index 7f5d7a70d27e..73cd9cf25b83 100644
> --- a/include/hw/ppc/spapr.h
> +++ b/include/hw/ppc/spapr.h
> @@ -766,9 +766,14 @@ struct SpaprEventLogEntry {
>  
>  void spapr_events_init(SpaprMachineState *sm);
>  void spapr_dt_events(SpaprMachineState *sm, void *fdt);
> +int spapr_h_cas_do_compose_response(SpaprMachineState *spapr, void *fdt,
> +                                    SpaprOptionVector *ov5_updates);
>  int spapr_h_cas_compose_response(SpaprMachineState *sm,
>                                   target_ulong addr, target_ulong size,
>                                   SpaprOptionVector *ov5_updates);
> +#define FDT_MAX_SIZE 0x100000
> +void *spapr_build_fdt(SpaprMachineState *spapr);
> +
>  void close_htab_fd(SpaprMachineState *spapr);
>  void spapr_setup_hpt_and_vrma(SpaprMachineState *spapr);
>  void spapr_free_hpt(SpaprMachineState *spapr);
> diff --git a/hw/ppc/spapr.c b/hw/ppc/spapr.c
> index b097a99951f1..f84895f4a8b4 100644
> --- a/hw/ppc/spapr.c
> +++ b/hw/ppc/spapr.c
> @@ -978,6 +978,19 @@ static bool spapr_hotplugged_dev_before_cas(void)
>      return false;
>  }
>  
> +int spapr_h_cas_do_compose_response(SpaprMachineState *spapr, void *fdt,
> +                                    SpaprOptionVector *ov5_updates)

Not a great function name.

> +{
> +    /* Fixup cpu nodes */
> +    _FDT((spapr_fixup_cpu_dt(fdt, spapr)));
> +
> +    if (spapr_dt_cas_updates(spapr, fdt, ov5_updates)) {
> +        return -1;
> +    }
> +
> +    return 0;
> +}
> +
>  int spapr_h_cas_compose_response(SpaprMachineState *spapr,
>                                   target_ulong addr, target_ulong size,
>                                   SpaprOptionVector *ov5_updates)
> @@ -1009,10 +1022,7 @@ int spapr_h_cas_compose_response(SpaprMachineState *spapr,
>      _FDT((fdt_open_into(fdt_skel, fdt, size)));
>      g_free(fdt_skel);
>  
> -    /* Fixup cpu nodes */
> -    _FDT((spapr_fixup_cpu_dt(fdt, spapr)));
> -
> -    if (spapr_dt_cas_updates(spapr, fdt, ov5_updates)) {
> +    if (spapr_h_cas_do_compose_response(spapr, fdt, ov5_updates)) {
>          return -1;
>      }
>  
> @@ -1232,6 +1242,10 @@ static void spapr_dt_chosen(SpaprMachineState *spapr, void *fdt)
>  
>      /* We always implemented RTAS as hcall, tell guests to call it directly */
>      _FDT(fdt_setprop_cell(fdt, chosen, "qemu,h_rtas", 1));
> +    /* Tell the guest that H_CAS will return the entire FDT now, not the diff */
> +    if (!spapr->bios_enabled) {
> +        _FDT(fdt_setprop_cell(fdt, chosen, "qemu,h_cas", 1));

As with H_RTAS< using qemu,hypertas-functions would be more
appropriate for this.

> +    }
>  
>      spapr_dt_ov5_platform_support(spapr, fdt, chosen);
>  
> @@ -1262,7 +1276,7 @@ static void spapr_dt_hypervisor(SpaprMachineState *spapr, void *fdt)
>      }
>  }
>  
> -static void *spapr_build_fdt(SpaprMachineState *spapr)
> +void *spapr_build_fdt(SpaprMachineState *spapr)
>  {
>      MachineState *machine = MACHINE(spapr);
>      MachineClass *mc = MACHINE_GET_CLASS(machine);
> diff --git a/hw/ppc/spapr_hcall.c b/hw/ppc/spapr_hcall.c
> index b964d94f330b..c5cb06c9d507 100644
> --- a/hw/ppc/spapr_hcall.c
> +++ b/hw/ppc/spapr_hcall.c
> @@ -17,6 +17,7 @@
>  #include "hw/ppc/spapr_ovec.h"
>  #include "mmu-book3s-v3.h"
>  #include "hw/mem/memory-device.h"
> +#include "sysemu/device_tree.h"
>  
>  static bool has_spr(PowerPCCPU *cpu, int spr)
>  {
> @@ -1774,9 +1775,51 @@ static target_ulong h_client_architecture_support(PowerPCCPU *cpu,
>              /* legacy hash or new hash: */
>              spapr_setup_hpt_and_vrma(spapr);
>          }
> -        spapr->cas_reboot =
> -            (spapr_h_cas_compose_response(spapr, args[1], args[2],
> -                                          ov5_updates) != 0);
> +
> +        if (spapr->bios_enabled) {
> +            spapr->cas_reboot =
> +                (spapr_h_cas_compose_response(spapr, args[1], args[2],
> +                                              ov5_updates) != 0);
> +        } else {
> +            int size;
> +            void *fdt, *fdt_skel;
> +            struct fdt_header hdr = { 0 };
> +
> +            cpu_physical_memory_read(args[1], &hdr, sizeof(hdr));
> +            size = fdt32_to_cpu(hdr.totalsize);
> +            if (size > FDT_MAX_SIZE) {
> +                return H_NOT_AVAILABLE;
> +            }
> +
> +            fdt_skel = g_malloc0(size);
> +            cpu_physical_memory_read(args[1], fdt_skel, size);
> +
> +            fdt = g_malloc0(FDT_MAX_SIZE);
> +            fdt_open_into(fdt_skel, fdt, FDT_MAX_SIZE);
> +            g_free(fdt_skel);

fdt_open_into() explicitly permits using the same buffer for both
arguments, so you don't need two allocations here - you can just
allocate FDT_MAX_SIZE, read it in, then fdt_open_into() in place.

You probably should check for errors from fdt_open_into(), though.

> +
> +            if (spapr_h_cas_do_compose_response(spapr, fdt, ov5_updates)) {
> +                g_free(fdt);
> +                return H_NOT_AVAILABLE;
> +            }
> +            fdt_pack(fdt);
> +            if (fdt_totalsize(fdt) > FDT_MAX_SIZE) {

This can't actually happen - you only ever allocated a buffer of size
FDT_MAX_SIZE, so if the DT was going to exceed that you would have hit
FDT_ERR_NOSPACE in an earlier step.

> +                error_report("FDT too big ! 0x%x bytes (max is 0x%x)",
> +                             fdt_totalsize(fdt), FDT_MAX_SIZE);
> +                g_free(fdt);
> +                return H_NOT_AVAILABLE;
> +            }
> +
> +            /* Load the fdt */
> +            cpu_physical_memory_write(args[1], fdt, fdt_totalsize(fdt));
> +
> +            g_free(spapr->fdt_blob);
> +            spapr->fdt_size = fdt_totalsize(fdt);
> +            spapr->fdt_initial_size = spapr->fdt_size;
> +            spapr->fdt_blob = fdt;
> +
> +            qemu_fdt_dumpdtb(fdt, fdt_totalsize(fdt));
> +        }
>      }
>  
>      /*

-- 
David Gibson			| I'll have my music baroque, and my code
david AT gibson.dropbear.id.au	| minimalist, thank you.  NOT _the_ _other_
				| _way_ _around_!
http://www.ozlabs.org/~dgibson

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]