qemu-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [PATCH] xen: fix build without pci passthrough


From: Anthony PERARD
Subject: Re: [PATCH] xen: fix build without pci passthrough
Date: Mon, 11 May 2020 14:40:43 +0100

On Mon, May 04, 2020 at 12:14:43PM +0200, Roger Pau Monne wrote:
> has_igd_gfx_passthru is only available when QEMU is built with
> CONFIG_XEN_PCI_PASSTHROUGH, and hence shouldn't be used in common
> code without checking if it's available.
> 
> Fixes: 46472d82322d0 ('xen: convert "-machine igd-passthru" to an accelerator 
> property')
> Signed-off-by: Roger Pau Monné <address@hidden>
> ---
> Cc: Stefano Stabellini <address@hidden>
> Cc: Anthony Perard <address@hidden>
> Cc: Paul Durrant <address@hidden>
> Cc: address@hidden
> ---
>  hw/xen/xen-common.c | 4 ++++
>  hw/xen/xen_pt.h     | 7 +++++++
>  2 files changed, 11 insertions(+)
> 
> diff --git a/hw/xen/xen-common.c b/hw/xen/xen-common.c
> index a15070f7f6..c800862419 100644
> --- a/hw/xen/xen-common.c
> +++ b/hw/xen/xen-common.c
> @@ -127,6 +127,7 @@ static void xen_change_state_handler(void *opaque, int 
> running,
>      }
>  }
>  
> +#ifdef CONFIG_XEN_PCI_PASSTHROUGH
>  static bool xen_get_igd_gfx_passthru(Object *obj, Error **errp)
>  {
>      return has_igd_gfx_passthru;
> @@ -136,6 +137,7 @@ static void xen_set_igd_gfx_passthru(Object *obj, bool 
> value, Error **errp)
>  {
>      has_igd_gfx_passthru = value;
>  }
> +#endif
>  
>  static void xen_setup_post(MachineState *ms, AccelState *accel)
>  {
> @@ -197,11 +199,13 @@ static void xen_accel_class_init(ObjectClass *oc, void 
> *data)
>  
>      compat_props_add(ac->compat_props, compat, G_N_ELEMENTS(compat));
>  
> +#ifdef CONFIG_XEN_PCI_PASSTHROUGH
>      object_class_property_add_bool(oc, "igd-passthru",
>          xen_get_igd_gfx_passthru, xen_set_igd_gfx_passthru,
>          &error_abort);
>      object_class_property_set_description(oc, "igd-passthru",
>          "Set on/off to enable/disable igd passthrou", &error_abort);
> +#endif

It might not be a good idea to have the presence of class property
depending on what is built-in. Instead, I think it would be better to
return an error when QEMU is built without xen_pt and a user is trying
to enable it. That can be done by calling by calling error_setg(errp,
"nop") in xen_set_igd_gfx_passthru().

>  }
>  
>  #define TYPE_XEN_ACCEL ACCEL_CLASS_NAME("xen")
> diff --git a/hw/xen/xen_pt.h b/hw/xen/xen_pt.h
> index 179775db7b..660dd8a008 100644
> --- a/hw/xen/xen_pt.h
> +++ b/hw/xen/xen_pt.h
> @@ -1,6 +1,7 @@
>  #ifndef XEN_PT_H
>  #define XEN_PT_H
>  
> +#include "qemu/osdep.h"

Why do you need osdep?

>  #include "hw/xen/xen_common.h"
>  #include "hw/pci/pci.h"
>  #include "xen-host-pci-device.h"
> @@ -322,7 +323,13 @@ extern void *pci_assign_dev_load_option_rom(PCIDevice 
> *dev,
>                                              unsigned int domain,
>                                              unsigned int bus, unsigned int 
> slot,
>                                              unsigned int function);
> +
> +#ifdef CONFIG_XEN_PCI_PASSTHROUGH
>  extern bool has_igd_gfx_passthru;
> +#else
> +# define has_igd_gfx_passthru false
> +#endif

I don't quite like the use of define here. Could you introduce a
function that return a bool instead? And defining that function in
hw/xen/xen.h like xen_enabled() would be fine I think.

Thanks,

-- 
Anthony PERARD



reply via email to

[Prev in Thread] Current Thread [Next in Thread]