From: Pierre Morel <pmorel@linux.vnet.ibm.com>
To: Cornelia Huck <cohuck@redhat.com>, qemu-devel@nongnu.org
Cc: thuth@redhat.com, zyimin@linux.vnet.ibm.com, david@redhat.com,
	agraf@suse.de, borntraeger@de.ibm.com
Subject: Re: [Qemu-devel] [PATCH v4 02/10] kvm: remove hard dependency on pci
Date: Mon, 21 Aug 2017 18:02:12 +0200	[thread overview]
Message-ID: <cb66ce20-c407-2462-0e51-8d6751af9f7b@linux.vnet.ibm.com> (raw)
In-Reply-To: <20170821091614.28251-3-cohuck@redhat.com>
On 21/08/2017 11:16, Cornelia Huck wrote:
> The msi routing code in kvm calls some pci functions: provide
> some stubs to enable builds without pci.
> 
> Also, to make this more obvious, guard them via a pci_available boolean
> (which also can be reused in other places).
> 
> Fixes: e1d4fb2de ("kvm-irqchip: x86: add msi route notify fn")
> Fixes: 767a554a0 ("kvm-all: Pass requester ID to MSI routing functions")
> Signed-off-by: Cornelia Huck <cohuck@redhat.com>
> ---
>   accel/kvm/kvm-all.c  |  6 +++---
>   hw/pci/pci-stub.c    | 15 +++++++++++++++
>   hw/pci/pci.c         |  2 ++
>   include/hw/pci/pci.h |  2 ++
>   4 files changed, 22 insertions(+), 3 deletions(-)
> 
> diff --git a/accel/kvm/kvm-all.c b/accel/kvm/kvm-all.c
> index 46ce479dc3..f85553a851 100644
> --- a/accel/kvm/kvm-all.c
> +++ b/accel/kvm/kvm-all.c
> @@ -1248,7 +1248,7 @@ int kvm_irqchip_add_msi_route(KVMState *s, int vector, PCIDevice *dev)
>       int virq;
>       MSIMessage msg = {0, 0};
> 
> -    if (dev) {
> +    if (pci_available && dev) {
>           msg = pci_get_msi_message(dev, vector);
>       }
Hi Conny,
I did not find a case where pci_available is false and dev is true.
but anyway, sure is sure.
> 
> @@ -1271,7 +1271,7 @@ int kvm_irqchip_add_msi_route(KVMState *s, int vector, PCIDevice *dev)
>       kroute.u.msi.address_lo = (uint32_t)msg.address;
>       kroute.u.msi.address_hi = msg.address >> 32;
>       kroute.u.msi.data = le32_to_cpu(msg.data);
> -    if (kvm_msi_devid_required()) {
> +    if (pci_available && kvm_msi_devid_required()) {
>           kroute.flags = KVM_MSI_VALID_DEVID;
>           kroute.u.msi.devid = pci_requester_id(dev);
>       }
> @@ -1309,7 +1309,7 @@ int kvm_irqchip_update_msi_route(KVMState *s, int virq, MSIMessage msg,
>       kroute.u.msi.address_lo = (uint32_t)msg.address;
>       kroute.u.msi.address_hi = msg.address >> 32;
>       kroute.u.msi.data = le32_to_cpu(msg.data);
> -    if (kvm_msi_devid_required()) {
> +    if (pci_available && kvm_msi_devid_required()) {
>           kroute.flags = KVM_MSI_VALID_DEVID;
>           kroute.u.msi.devid = pci_requester_id(dev);
>       }
> diff --git a/hw/pci/pci-stub.c b/hw/pci/pci-stub.c
> index ecad664946..ace43821ca 100644
> --- a/hw/pci/pci-stub.c
> +++ b/hw/pci/pci-stub.c
> @@ -23,10 +23,12 @@
>   #include "monitor/monitor.h"
>   #include "qapi/qmp/qerror.h"
>   #include "hw/pci/pci.h"
> +#include "hw/pci/msi.h"
I think you forgot that...
>   #include "qmp-commands.h"
>   #include "hw/pci/msi.h"
...you already have it included here. Didn't you ?
> 
>   bool msi_nonbroken;
> +bool pci_available;
> 
>   PciInfoList *qmp_query_pci(Error **errp)
>   {
> @@ -38,3 +40,16 @@ void hmp_pcie_aer_inject_error(Monitor *mon, const QDict *qdict)
>   {
>       monitor_printf(mon, "PCI devices not supported\n");
>   }
> +
> +/* kvm-all wants this */
> +MSIMessage pci_get_msi_message(PCIDevice *dev, int vector)
> +{
> +    g_assert(false);
> +    return (MSIMessage){};
> +}
> +
> +uint16_t pci_requester_id(PCIDevice *dev)
> +{
> +    g_assert(false);
> +    return 0;
> +}
> diff --git a/hw/pci/pci.c b/hw/pci/pci.c
> index 258fbe51e2..26f346db2c 100644
> --- a/hw/pci/pci.c
> +++ b/hw/pci/pci.c
> @@ -49,6 +49,8 @@
>   # define PCI_DPRINTF(format, ...)       do { } while (0)
>   #endif
> 
> +bool pci_available = true;
> +
>   static void pcibus_dev_print(Monitor *mon, DeviceState *dev, int indent);
>   static char *pcibus_get_dev_path(DeviceState *dev);
>   static char *pcibus_get_fw_dev_path(DeviceState *dev);
> diff --git a/include/hw/pci/pci.h b/include/hw/pci/pci.h
> index e598b095eb..8bb6449dd7 100644
> --- a/include/hw/pci/pci.h
> +++ b/include/hw/pci/pci.h
> @@ -10,6 +10,8 @@
> 
>   #include "hw/pci/pcie.h"
> 
> +extern bool pci_available;
> +
>   /* PCI bus */
> 
>   #define PCI_DEVFN(slot, func)   ((((slot) & 0x1f) << 3) | ((func) & 0x07))
> 
otherwise LGTM
Regards,
Pierre
-- 
Pierre Morel
Linux/KVM/QEMU in Böblingen - Germany
next prev parent reply	other threads:[~2017-08-21 16:02 UTC|newest]
Thread overview: 44+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-08-21  9:16 [Qemu-devel] [PATCH v4 00/10] zpci detangling Cornelia Huck
2017-08-21  9:16 ` [Qemu-devel] [PATCH v4 01/10] 9pfs: fix dependencies Cornelia Huck
2017-08-21  9:16 ` [Qemu-devel] [PATCH v4 02/10] kvm: remove hard dependency on pci Cornelia Huck
2017-08-21 16:02   ` Pierre Morel [this message]
2017-08-22  9:04     ` Cornelia Huck
2017-08-23 11:05       ` Pierre Morel
2017-08-21  9:16 ` [Qemu-devel] [PATCH v4 03/10] s390x/pci: add stubs Cornelia Huck
2017-08-21  9:16 ` [Qemu-devel] [PATCH v4 04/10] s390x: chsc nt2 events are pci-only Cornelia Huck
2017-08-21 12:24   ` Thomas Huth
2017-08-21  9:16 ` [Qemu-devel] [PATCH v4 05/10] s390x/pci: do not advertise pci on non-pci builds Cornelia Huck
2017-08-21 12:29   ` Thomas Huth
2017-08-21  9:16 ` [Qemu-devel] [PATCH v4 06/10] s390x/ccw: create s390 phb conditionally Cornelia Huck
2017-08-21  9:16 ` [Qemu-devel] [PATCH v4 07/10] s390x/sclp: properly guard pci-specific functions Cornelia Huck
2017-08-21 11:41   ` Halil Pasic
2017-08-21 13:16     ` Cornelia Huck
2017-08-21 13:32       ` Halil Pasic
2017-08-21 13:36         ` Cornelia Huck
2017-08-21 14:58   ` Pierre Morel
2017-08-21 16:24     ` Halil Pasic
2017-08-22  8:39       ` Cornelia Huck
2017-08-22  9:20         ` Halil Pasic
2017-08-22  9:39           ` Cornelia Huck
2017-08-22 12:57             ` Cornelia Huck
2017-08-22 13:00               ` Halil Pasic
2017-08-22 12:58             ` Halil Pasic
2017-08-22 13:24               ` Cornelia Huck
2017-08-22 13:54                 ` Halil Pasic
2017-08-22 14:15                   ` Cornelia Huck
2017-08-22 14:34                     ` Halil Pasic
2017-08-22 15:10                       ` Cornelia Huck
2017-08-22 14:06                 ` Cornelia Huck
2017-08-22 14:27                   ` Halil Pasic
2017-08-22 14:34                     ` Cornelia Huck
2017-08-21  9:16 ` [Qemu-devel] [PATCH v4 08/10] s390x/pci: fence off instructions for non-pci Cornelia Huck
2017-08-23 14:10   ` Halil Pasic
2017-08-23 15:40     ` Cornelia Huck
2017-08-21  9:16 ` [Qemu-devel] [PATCH v4 09/10] s390x/kvm: msi route fixup " Cornelia Huck
2017-08-21 12:00   ` Halil Pasic
2017-08-21 12:13     ` Cornelia Huck
2017-08-21 15:10       ` Halil Pasic
2017-08-21 15:17         ` Thomas Huth
2017-08-21 15:30           ` Halil Pasic
2017-08-23 10:03             ` Cornelia Huck
2017-08-21  9:16 ` [Qemu-devel] [PATCH v4 10/10] s390x: refine pci dependencies Cornelia Huck
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox
  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):
  git send-email \
    --in-reply-to=cb66ce20-c407-2462-0e51-8d6751af9f7b@linux.vnet.ibm.com \
    --to=pmorel@linux.vnet.ibm.com \
    --cc=agraf@suse.de \
    --cc=borntraeger@de.ibm.com \
    --cc=cohuck@redhat.com \
    --cc=david@redhat.com \
    --cc=qemu-devel@nongnu.org \
    --cc=thuth@redhat.com \
    --cc=zyimin@linux.vnet.ibm.com \
    /path/to/YOUR_REPLY
  https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
  Be sure your reply has a Subject: header at the top and a blank line
  before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).