From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from bombadil.infradead.org (bombadil.infradead.org [198.137.202.133]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id 42CACCD98C7 for ; Wed, 10 Jun 2026 00:54:14 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=lists.infradead.org; s=bombadil.20210309; h=Sender:Reply-To:List-Subscribe: List-Help:List-Post:List-Archive:List-Unsubscribe:List-Id:Content-Type:Cc:To: From:Subject:Message-ID:References:Mime-Version:In-Reply-To:Date: Content-Transfer-Encoding:Content-ID:Content-Description:Resent-Date: Resent-From:Resent-Sender:Resent-To:Resent-Cc:Resent-Message-ID:List-Owner; bh=GUYNx7RZ8mYUz0pjgPg+HKHfepMmoInSE1YhyqLiL7U=; b=eCldcOxFlP70liBKg0FzOzhoIw Qxy3bGLVyfgt4YzZN8z00NdmOD3GbYpNwyd7tjTINAbbjWs2qB080RMnCBTPHGTiP7LBtvWJP+nu6 xOOt5PUAhwRLzo83CQVlhcwPo+mJ5IU1sSb6wPSJMoVCtmOrB8mxGB//GskIHJCpdqJKhh6OGiSh+ OqweGHedwk/BrKFYLo43z3Kle4V0xK+0KX6TgxssUXhmZReOnbZB9nkeTGcT3LJpz6Lv34NhrbvsR hLD1+67uUWumeZIXBQS/JxfDeGzDoktOVGziEd6NaVrTLZ7Q3ca8CFkoMDYnbAkoAdkaqLpxm7AFD ZaXBEs1g==; Received: from localhost ([::1] helo=bombadil.infradead.org) by bombadil.infradead.org with esmtp (Exim 4.99.1 #2 (Red Hat Linux)) id 1wX7Cp-00000006Zik-2ZM2; Wed, 10 Jun 2026 00:54:07 +0000 Received: from mail-pg1-x54a.google.com ([2607:f8b0:4864:20::54a]) by bombadil.infradead.org with esmtps (Exim 4.99.1 #2 (Red Hat Linux)) id 1wX7Cg-00000006ZWS-1LNv for linux-arm-kernel@lists.infradead.org; Wed, 10 Jun 2026 00:53:59 +0000 Received: by mail-pg1-x54a.google.com with SMTP id 41be03b00d2f7-c85798977dcso3817084a12.0 for ; Tue, 09 Jun 2026 17:53:57 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20251104; t=1781052837; x=1781657637; darn=lists.infradead.org; h=cc:to:from:subject:message-id:references:mime-version:in-reply-to :date:reply-to:from:to:cc:subject:date:message-id:reply-to; bh=GUYNx7RZ8mYUz0pjgPg+HKHfepMmoInSE1YhyqLiL7U=; b=HJdcDmRfW/FuQtjPQ947ssLF7ylTXNeAEwd4EyrUO9DrPukHYz28vCSR0WgHhQwp9s 0RvbsDplzP+MgfoDj9Gr04RcjtFhyhxnYVOzI/S7muQq2nybQIjBqLnd5uA9v3HX9JDh pl1N1ORyyzYBEePSgK9lFmpp62eC0hxhQXngz7Ei8MsMWNLwQpPErI6r4lL+rgyCNqtM t5psxasPTpPWZBHaThI0teeDF3lQWVcC3t1Uf1v6mgp2iE/IgpU0jJVbB+MuTYMVgyZT oBeeVnmn3zenUmZaBjjTEH7YNydOa9crtduQiPD4pOOlDykkb+saUuavSKjq4Tqs7tHh xqiA== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20251104; t=1781052837; x=1781657637; h=cc:to:from:subject:message-id:references:mime-version:in-reply-to :date:reply-to:x-gm-message-state:from:to:cc:subject:date:message-id :reply-to; bh=GUYNx7RZ8mYUz0pjgPg+HKHfepMmoInSE1YhyqLiL7U=; b=C1gJ1X1dG9/lqPm9Wcy2UiD0/adHA+ZFq2D1GViO1D2rIcdULqNv8dUm+MWg61qxMD oSN/ee8p1zIIqPk88wjZyJ2pfnmD5RkR0rfkfYHwK4cOZrIaKGeJ1nFirCWMOHYxfYts Z9bjRHccsIBgxtdiD4oFPJA9rICoPvZWNMW/vEXkFDLhAi3KxXu5IF9uiXZlih+LiyrF E1aG9Y+2RakW72sioxQTW1XLo5o66O5t+p6FaZxLPc6sFKic0pWXvXDhg+vkjFM1AxJh pXglOnlX6Qm0YgpusgBGFcAUBL9aqZllDmujO6HHFZgYmis6OmGS7qsegemRFdrJbxxy x4sA== X-Forwarded-Encrypted: i=1; AFNElJ/Z+bpxoC/5TDlUToZy1RXhqcgJvOsnrv65cfz8HyYay0KJKvRJKZfpkfnmdCuibQQGy7YlB7kT/7IlqMFDmXy8@lists.infradead.org X-Gm-Message-State: AOJu0YxNiLU3gcVpEWxUgfQRwP/0nlVpmJ801rXuS7y3YTiObMw12eBA AfVjPciDIRDJdokQY66/S1N4lPe+5bIfvu9SDAauZfdFgRro7FjwwTmgoYq5RayYVwMq26ZqC6J lfH/3uQ== X-Received: from pfbhx1.prod.google.com ([2002:a05:6a00:8981:b0:842:5a44:a83c]) (user=seanjc job=prod-delivery.src-stubby-dispatcher) by 2002:a05:6a00:90a7:b0:842:5634:3c1d with SMTP id d2e1a72fcca58-842b0e835d1mr22690620b3a.19.1781052836923; Tue, 09 Jun 2026 17:53:56 -0700 (PDT) Date: Tue, 9 Jun 2026 17:53:32 -0700 In-Reply-To: <20260610005338.2967132-1-seanjc@google.com> Mime-Version: 1.0 References: <20260610005338.2967132-1-seanjc@google.com> X-Mailer: git-send-email 2.54.0.1099.g489fc7bff1-goog Message-ID: <20260610005338.2967132-13-seanjc@google.com> Subject: [PATCH v6 12/18] KVM: selftests: Verify non-postable IRQ remapping in IRQ test From: Sean Christopherson To: Paolo Bonzini , Marc Zyngier , Oliver Upton , Sean Christopherson Cc: Joey Gouly , Steffen Eiden , Suzuki K Poulose , Zenghui Yu , kvm@vger.kernel.org, linux-arm-kernel@lists.infradead.org, kvmarm@lists.linux.dev, linux-kernel@vger.kernel.org, David Matlack , Josh Hilke Content-Type: text/plain; charset="UTF-8" X-CRM114-Version: 20100106-BlameMichelson ( TRE 0.9.0 (BSD) ) MR-646709E3 X-CRM114-CacheID: sfid-20260609_175358_366584_B53BEB2B X-CRM114-Status: GOOD ( 18.17 ) X-BeenThere: linux-arm-kernel@lists.infradead.org X-Mailman-Version: 2.1.34 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Reply-To: Sean Christopherson Sender: "linux-arm-kernel" Errors-To: linux-arm-kernel-bounces+linux-arm-kernel=archiver.kernel.org@lists.infradead.org From: David Matlack Extend the eventfd IRQ test with an '-n' flag to route a subset of device interrupts as NMIs (Non-Maskable Interrupts) into the guest using an alternating pattern of 4 NMIs followed by 4 regular interrupts. While this adds coverage for NMI injection, the primary goal is to validate KVM's handling of non-postable interrupt delivery (AMD and Intel IOMMUs only support posting fixed IRQs targeting a single vCPU). KVM has historically bungled handling transitions between posted and remapped modes. Use NMIs to stress the transitions, because they are a reliable, architectural way to force these code paths. Signed-off-by: David Matlack Co-developed-by: Josh Hilke Signed-off-by: Josh Hilke [sean: add GUEST_RECEIVED_INTERRUPT(), massage changelog] Signed-off-by: Sean Christopherson --- tools/testing/selftests/kvm/irq_test.c | 48 ++++++++++++++++++++------ 1 file changed, 37 insertions(+), 11 deletions(-) diff --git a/tools/testing/selftests/kvm/irq_test.c b/tools/testing/selftests/kvm/irq_test.c index 15791ba0a2f9..3fed76c553dd 100644 --- a/tools/testing/selftests/kvm/irq_test.c +++ b/tools/testing/selftests/kvm/irq_test.c @@ -17,11 +17,17 @@ static u64 timeout_ns = 2ULL * 1000 * 1000 * 1000; static bool guest_ready_for_irqs[KVM_MAX_VCPUS]; static bool guest_received_irq[KVM_MAX_VCPUS]; +static bool guest_received_nmi[KVM_MAX_VCPUS]; static bool irq_affinity; static bool done; #define GUEST_RECEIVED_IRQ(__vcpu) \ SYNC_FROM_GUEST_AND_READ((__vcpu)->vm, guest_received_irq[(__vcpu)->id]) +#define GUEST_RECEIVED_NMI(__vcpu) \ + SYNC_FROM_GUEST_AND_READ((__vcpu)->vm, guest_received_nmi[(__vcpu)->id]) + +#define GUEST_RECEIVED_INTERRUPT(__vcpu, __nmi) \ + ((__nmi) ? GUEST_RECEIVED_NMI(__vcpu) : GUEST_RECEIVED_IRQ(__vcpu)) static u32 guest_get_vcpu_id(void) { @@ -35,6 +41,11 @@ static void guest_irq_handler(struct ex_regs *regs) x2apic_write_reg(APIC_EOI, 0); } +static void guest_nmi_handler(struct ex_regs *regs) +{ + WRITE_ONCE(guest_received_nmi[guest_get_vcpu_id()], true); +} + static void guest_code(void) { x2apic_enable(); @@ -91,7 +102,7 @@ static void trigger_interrupt(struct vfio_pci_device *device, int eventfd) static void kvm_route_msi(struct kvm_vm *vm, u32 gsi, struct kvm_vcpu *vcpu, - u8 vector) + u8 vector, bool use_nmi) { struct { struct kvm_irq_routing header; @@ -102,7 +113,7 @@ static void kvm_route_msi(struct kvm_vm *vm, u32 gsi, struct kvm_vcpu *vcpu, .gsi = gsi, .type = KVM_IRQ_ROUTING_MSI, .u.msi.address_lo = 0xFEE00000 | (vcpu->id << 12), - .u.msi.data = vector, + .u.msi.data = use_nmi ? NMI_VECTOR | (4 << 8) : vector, }, }; @@ -134,13 +145,14 @@ static const char *probe_iommu_type(void) static void help(const char *name) { - printf("Usage: %s [-a] [-d ] [-e] [-h] [-i nr_irqs] [-t iommu_type]\n", name); + printf("Usage: %s [-a] [-d ] [-e] [-h] [-i nr_irqs] [-n] [-t iommu_type]\n", name); printf("\n"); printf("Tests KVM interrupt routing and delivery via irqfd.\n"); printf("-a Affine the device's host IRQ to a random physical CPU\n"); printf("-d Use a VFIO device to send MSI-X interrupts instead of manually signaling the eventfd\n"); printf("-e Set empty GSI routing in-between some interrupts\n"); printf("-i The number of IRQs to generate during the test\n"); + printf("-n Deliver 50 percent of IRQs as non-maskable interrupts\n"); printf("-t Override the IOMMU type to use (vfio_type1_iommu or iommufd)\n"); printf("\n"); exit(KSFT_FAIL); @@ -171,11 +183,12 @@ int main(int argc, char **argv) const char *device_bdf = NULL; const char *iommu_type = NULL; int i, j, c, msix, eventfd; + bool use_nmi = false; struct iommu *iommu; struct kvm_vm *vm; int irq, irq_cpu; - while ((c = getopt(argc, argv, "ad:ehi:t:")) != -1) { + while ((c = getopt(argc, argv, "ad:ehi:nt:")) != -1) { switch (c) { case 'a': irq_affinity = true; @@ -189,6 +202,9 @@ int main(int argc, char **argv) case 'i': nr_irqs = atoi_positive("Number of IRQs", optarg); break; + case 'n': + use_nmi = true; + break; case 't': iommu_type = optarg; break; @@ -202,6 +218,7 @@ int main(int argc, char **argv) vm = vm_create_with_vcpus(nr_vcpus, guest_code, vcpus); vm_install_exception_handler(vm, vector, guest_irq_handler); + vm_install_exception_handler(vm, NMI_VECTOR, guest_nmi_handler); if (device_bdf) { if (!iommu_type) @@ -239,36 +256,45 @@ int main(int argc, char **argv) for (i = 0; i < nr_irqs; i++) { const bool do_set_empty_routing = set_empty_routing && (i & BIT(3)); + const bool do_use_nmi = use_nmi && (i & BIT(2)); struct kvm_vcpu *vcpu = vcpus[i % nr_vcpus]; struct timespec start; if (do_set_empty_routing) kvm_set_empty_gsi_routing(vm); - kvm_route_msi(vm, gsi, vcpu, vector); + kvm_route_msi(vm, gsi, vcpu, vector, do_use_nmi); if (irq_affinity) { irq_cpu = kvm_random_u64(&kvm_rng) % get_nprocs(); proc_irq_set_smp_affinity(irq, irq_cpu); } - for (j = 0; j < nr_vcpus; j++) + for (j = 0; j < nr_vcpus; j++) { TEST_ASSERT(!GUEST_RECEIVED_IRQ(vcpus[j]), "IRQ flag for vCPU %d not clear prior to test", vcpus[j]->id); + TEST_ASSERT(!GUEST_RECEIVED_NMI(vcpus[j]), + "NMI flag for vCPU %d not clear prior to test", + vcpus[j]->id); + } trigger_interrupt(device, eventfd); clock_gettime(CLOCK_MONOTONIC, &start); - while (!GUEST_RECEIVED_IRQ(vcpu) && + while (!GUEST_RECEIVED_INTERRUPT(vcpu, do_use_nmi) && timespec_to_ns(timespec_elapsed(start)) <= timeout_ns) cpu_relax(); - TEST_ASSERT(GUEST_RECEIVED_IRQ(vcpu), - "vCPU %d timed out waiting for IRQ (vector 0x%x) from GSI %d (via CPU %d)\n", - vcpu->id, vector, gsi, irq_cpu); + TEST_ASSERT(GUEST_RECEIVED_INTERRUPT(vcpu, do_use_nmi), + "vCPU %d timed out waiting for %s (vector 0x%x) from GSI %d (via CPU %d)\n", + vcpu->id, do_use_nmi ? "NMI" : "IRQ", + do_use_nmi ? NMI_VECTOR : vector, gsi, irq_cpu); - WRITE_AND_SYNC_TO_GUEST(vm, guest_received_irq[vcpu->id], false); + if (do_use_nmi) + WRITE_AND_SYNC_TO_GUEST(vm, guest_received_nmi[vcpu->id], false); + else + WRITE_AND_SYNC_TO_GUEST(vm, guest_received_irq[vcpu->id], false); } WRITE_AND_SYNC_TO_GUEST(vm, done, true); -- 2.54.0.1099.g489fc7bff1-goog