From mboxrd@z Thu Jan 1 00:00:00 1970 From: Marc Zyngier Date: Wed, 21 Sep 2022 16:59:10 +0000 Subject: Re: [PATCH kernel v2] KVM: PPC: Make KVM_CAP_IRQFD_RESAMPLE support platform dependent Message-Id: <87r104k6bc.wl-maz@kernel.org> List-Id: References: <20220920125143.28009-1-aik@ozlabs.ru> <874jx2kp02.wl-maz@kernel.org> <02490d2a-5e89-f342-f5b3-386406ffb2ea@ozlabs.ru> In-Reply-To: <02490d2a-5e89-f342-f5b3-386406ffb2ea@ozlabs.ru> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: Alexey Kardashevskiy Cc: kvm@vger.kernel.org, linux-arm-kernel@lists.infradead.org, kvmarm@lists.cs.columbia.edu, kvm-riscv@lists.infradead.org, Paolo Bonzini , Anup Patel , kvm-ppc@vger.kernel.org, Nicholas Piggin On Tue, 20 Sep 2022 22:46:21 +0100, Alexey Kardashevskiy wrote: > > > > On 21/09/2022 02:08, Marc Zyngier wr > ote: > > On Tue, 20 Sep 2022 13:51:43 +0100, > > Alexey Kardashevskiy wrote: > >> > >> When introduced, IRQFD resampling worked on POWER8 with XICS. However > >> KVM on POWER9 has never implemented it - the compatibility mode code > >> ("XICS-on-XIVE") misses the kvm_notify_acked_irq() call and the native > >> XIVE mode does not handle INTx in KVM at all. > >> > >> This moved the capability support advertising to platforms and stops > >> advertising it on XIVE, i.e. POWER9 and later. > >> > >> Signed-off-by: Alexey Kardashevskiy > >> Acked-by: Nicholas Piggin > >> [For KVM RISC-V] > >> Acked-by: Anup Patel > >> --- > >> Changes: > >> v2: > >> * removed ifdef for ARM64. > > > > The same argument applies to both x86 and s390, which do select > > HAVE_KVM_IRQFD from the KVM config option. Only power allows this > > option to be selected depending on the underlying interrupt controller > > emulation. > > > > As for riscv and mips, they don't select HAVE_KVM_IRQFD, and this > > isn't a user-selectable option. So why do they get patched at all? > > Before the patch, the capability was advertised on those, with your > proposal it will stop. No, they were never advertised, since none of these architectures select CONFIG_HAVE_KVM_IRQFD, and this option is not selectable by the user. > Which is a change in behavior which some may > say requires a separate patch (like, one per platform). I am > definitely overthinking it though and you are right. Well, either there is no change in behaviour (and therefore I am right), or there is one (and I am wrong). M. -- Without deviation from the norm, progress is not possible.