From mboxrd@z Thu Jan 1 00:00:00 1970 From: Eric Auger Subject: Re: [PATCH 1/7] KVM: api: add kvm_irq_routing_extended_msi Date: Thu, 02 Jul 2015 17:22:16 +0200 Message-ID: <55955728.3080506@linaro.org> References: <1435592237-17924-1-git-send-email-eric.auger@linaro.org> <1435592237-17924-2-git-send-email-eric.auger@linaro.org> <011f01d0b498$6a17aeb0$3e470c10$@samsung.com> <55954F6E.5030008@linaro.org> <55955542.6000809@arm.com> Mime-Version: 1.0 Content-Type: text/plain; charset=windows-1252 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: "linux-kernel@vger.kernel.org" , "pbonzini@redhat.com" To: Andre Przywara , Pavel Fedin , "eric.auger@st.com" , "linux-arm-kernel@lists.infradead.org" , Marc Zyngier , "christoffer.dall@linaro.org" , "kvmarm@lists.cs.columbia.edu" , "kvm@vger.kernel.org" Return-path: Received: from mail-wi0-f171.google.com ([209.85.212.171]:38231 "EHLO mail-wi0-f171.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753888AbbGBPWg (ORCPT ); Thu, 2 Jul 2015 11:22:36 -0400 Received: by wibdq8 with SMTP id dq8so76967792wib.1 for ; Thu, 02 Jul 2015 08:22:34 -0700 (PDT) In-Reply-To: <55955542.6000809@arm.com> Sender: kvm-owner@vger.kernel.org List-ID: Hi Andre, On 07/02/2015 05:14 PM, Andre Przywara wrote: > Hi Eric, >=20 > On 02/07/15 15:49, Eric Auger wrote: >> Hi Pavel, >> On 07/02/2015 09:26 AM, Pavel Fedin wrote: >>> Hello! >>> >>>> -----Original Message----- >>>> From: kvm-owner@vger.kernel.org [mailto:kvm-owner@vger.kernel.org]= On Behalf Of Eric Auger >>>> Sent: Monday, June 29, 2015 6:37 PM >>>> To: eric.auger@st.com; eric.auger@linaro.org; linux-arm-kernel@lis= ts.infradead.org; >>>> marc.zyngier@arm.com; christoffer.dall@linaro.org; andre.przywara@= arm.com; >>>> kvmarm@lists.cs.columbia.edu; kvm@vger.kernel.org >>>> Cc: linux-kernel@vger.kernel.org; patches@linaro.org; p.fedin@sams= ung.com; pbonzini@redhat.com >>>> Subject: [PATCH 1/7] KVM: api: add kvm_irq_routing_extended_msi >>>> >>>> On ARM, the MSI msg (address and data) comes along with >>>> out-of-band device ID information. The device ID encodes the devic= e >>>> that composes the MSI msg. Let's create a new routing entry type, >>>> dubbed KVM_IRQ_ROUTING_EXTENDED_MSI and use the __u32 pad space >>>> to convey the device ID. >>>> >>>> Signed-off-by: Eric Auger >>>> >>>> --- >>>> >>>> RFC -> PATCH >>>> - remove kvm_irq_routing_extended_msi and use union instead >>>> --- >>>> Documentation/virtual/kvm/api.txt | 9 ++++++++- >>>> include/uapi/linux/kvm.h | 6 +++++- >>>> 2 files changed, 13 insertions(+), 2 deletions(-) >>>> >>>> diff --git a/Documentation/virtual/kvm/api.txt b/Documentation/vir= tual/kvm/api.txt >>>> index d20fd94..6426ae9 100644 >>>> --- a/Documentation/virtual/kvm/api.txt >>>> +++ b/Documentation/virtual/kvm/api.txt >>>> @@ -1414,7 +1414,10 @@ struct kvm_irq_routing_entry { >>>> __u32 gsi; >>>> __u32 type; >>>> __u32 flags; >>>> - __u32 pad; >>>> + union { >>>> + __u32 pad; >>>> + __u32 devid; >>>> + }; >>>> union { >>>> struct kvm_irq_routing_irqchip irqchip; >>>> struct kvm_irq_routing_msi msi; >>> >>> devid is actually a part of MSI bunch. Shouldn't it be a part of s= truct kvm_irq_routing_msi then? >>> It also has reserved pad. >> Well this makes sense to me to associate the devid to the msi and pu= t >> devid in the pad field of struct kvm_irq_routing_msi. >> >> Andr=E9, Christoffer, would you agree on this change? - I would like= to >> avoid doing/undoing things ;-) - >=20 > Yes, that makes sense to me. TBH I haven't had a closer look at the > patches yet, but clearly devid belongs into struct kvm_irq_routing_ms= i. thanks for your quick reply. OK so let's go with that change. >=20 >>> >>>> @@ -1427,6 +1430,10 @@ struct kvm_irq_routing_entry { >>>> #define KVM_IRQ_ROUTING_IRQCHIP 1 >>>> #define KVM_IRQ_ROUTING_MSI 2 >>>> #define KVM_IRQ_ROUTING_S390_ADAPTER 3 >>>> +#define KVM_IRQ_ROUTING_EXTENDED_MSI 4 >>>> + >>>> +In case of KVM_IRQ_ROUTING_EXTENDED_MSI routing type, devid is us= ed to convey >>>> +the device ID. >>>> >>>> No flags are specified so far, the corresponding field must be se= t to zero. >>> >>> What if we use KVM_MSI_VALID_DEVID flag instead of new KVM_IRQ_ROUT= ING_EXTENDED_MSI definition? I >>> believe this would make an API more consistent and introduce less n= ew definitions. >> do you mean using type =3D=3D KVM_IRQ_ROUTING_MSI and flag =3D=3D >> KVM_MSI_VALID_DEVID? Not sure this is simpler/clearer. s390 paved th= e >> way for new routing entry types. I add a new one here. >=20 > I tend to agree with Pavel's solution. When hacking IRQ routing suppo= rt > into kvmtool I saw that it's nasty being forced to differentiate betw= een > the two MSI routing types. Actually userland should be able to query = the > kernel about what kind of routing it requires. Also there is the issu= e > that we must _not_ set the flag on x86, since that breaks older kerne= ls > (due to that check that Eric removes in 3/7). > So from my point of view the cleanest solution would be to always use > KVM_IRQ_ROUTING_MSI, and add the device ID if the kernel needs it (tr= ue > for ITS guests, false for GICv2M, x86, ...) > I am looking for a clever solution for this now. OK thanks for sharing. I need some more time to study qemu code too. - Eric >=20 > Cheers, > Andre. >=20 >> >> Another solution may be to use new KVM_IRQ_ROUTING_EXTENDED_MSI type= and >> add struct kvm_msi ext_msi in kvm_irq_routing_entry union. It is 8 w= ords >> as well. But most probably this is even uglier. >=20 >> >> Let's see if this thread is heading to a consensus... >> >> Best Regards >> >> Eric >>> >>>> >>>> diff --git a/include/uapi/linux/kvm.h b/include/uapi/linux/kvm.h >>>> index 2a23705..8484681 100644 >>>> --- a/include/uapi/linux/kvm.h >>>> +++ b/include/uapi/linux/kvm.h >>>> @@ -841,12 +841,16 @@ struct kvm_irq_routing_s390_adapter { >>>> #define KVM_IRQ_ROUTING_IRQCHIP 1 >>>> #define KVM_IRQ_ROUTING_MSI 2 >>>> #define KVM_IRQ_ROUTING_S390_ADAPTER 3 >>>> +#define KVM_IRQ_ROUTING_EXTENDED_MSI 4 >>>> >>>> struct kvm_irq_routing_entry { >>>> __u32 gsi; >>>> __u32 type; >>>> __u32 flags; >>>> - __u32 pad; >>>> + union { >>>> + __u32 pad; >>>> + __u32 devid; >>>> + }; >>>> union { >>>> struct kvm_irq_routing_irqchip irqchip; >>>> struct kvm_irq_routing_msi msi; >>>> -- >>>> 1.9.1 >>>> >>>> -- >>>> To unsubscribe from this list: send the line "unsubscribe kvm" in >>>> the body of a message to majordomo@vger.kernel.org >>>> More majordomo info at http://vger.kernel.org/majordomo-info.html >>> >>> Kind regards, >>> Pavel Fedin >>> Expert Engineer >>> Samsung Electronics Research center Russia >>> >>