From patchwork Tue Apr 2 08:02:15 2019 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Vitaly Kuznetsov X-Patchwork-Id: 10881223 Return-Path: Received: from mail.wl.linuxfoundation.org (pdx-wl-mail.web.codeaurora.org [172.30.200.125]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id B930717EE for ; Tue, 2 Apr 2019 08:03:58 +0000 (UTC) Received: from mail.wl.linuxfoundation.org (localhost [127.0.0.1]) by mail.wl.linuxfoundation.org (Postfix) with ESMTP id AABFC285AD for ; Tue, 2 Apr 2019 08:03:58 +0000 (UTC) Received: by mail.wl.linuxfoundation.org (Postfix, from userid 486) id 9EFA0288D0; Tue, 2 Apr 2019 08:03:58 +0000 (UTC) X-Spam-Checker-Version: SpamAssassin 3.3.1 (2010-03-16) on pdx-wl-mail.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-2.9 required=2.0 tests=BAYES_00,MAILING_LIST_MULTI autolearn=ham version=3.3.1 Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1 with cipher AES256-SHA (256/256 bits)) (No client certificate requested) by mail.wl.linuxfoundation.org (Postfix) with ESMTPS id 02801287F4 for ; Tue, 2 Apr 2019 08:03:58 +0000 (UTC) Received: from localhost ([127.0.0.1]:55825 helo=lists.gnu.org) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1hBEOj-00074v-AS for patchwork-qemu-devel@patchwork.kernel.org; Tue, 02 Apr 2019 04:03:57 -0400 Received: from eggs.gnu.org ([209.51.188.92]:40219) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1hBENG-0005Sg-LJ for qemu-devel@nongnu.org; Tue, 02 Apr 2019 04:02:32 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1hBENE-0002In-OM for qemu-devel@nongnu.org; Tue, 02 Apr 2019 04:02:26 -0400 Received: from mx1.redhat.com ([209.132.183.28]:31143) by eggs.gnu.org with esmtps (TLS1.0:DHE_RSA_AES_256_CBC_SHA1:32) (Exim 4.71) (envelope-from ) id 1hBENE-0002Hp-Bb for qemu-devel@nongnu.org; Tue, 02 Apr 2019 04:02:24 -0400 Received: from smtp.corp.redhat.com (int-mx08.intmail.prod.int.phx2.redhat.com [10.5.11.23]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mx1.redhat.com (Postfix) with ESMTPS id 7CDB53082E5F; Tue, 2 Apr 2019 08:02:23 +0000 (UTC) Received: from vitty.brq.redhat.com (unknown [10.43.2.155]) by smtp.corp.redhat.com (Postfix) with ESMTPS id 253AB57BF; Tue, 2 Apr 2019 08:02:17 +0000 (UTC) From: Vitaly Kuznetsov To: qemu-devel@nongnu.org Date: Tue, 2 Apr 2019 10:02:15 +0200 Message-Id: <20190402080215.10747-1-vkuznets@redhat.com> MIME-Version: 1.0 X-Scanned-By: MIMEDefang 2.84 on 10.5.11.23 X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-4.5.16 (mx1.redhat.com [10.5.110.46]); Tue, 02 Apr 2019 08:02:23 +0000 (UTC) X-detected-operating-system: by eggs.gnu.org: GNU/Linux 2.2.x-3.x [generic] X-Received-From: 209.132.183.28 Subject: [Qemu-devel] [PATCH v2] ioapic: allow buggy guests mishandling level-triggered interrupts to make progress X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.21 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Cc: Paolo Bonzini , Liran Alon , "Michael S. Tsirkin" Errors-To: qemu-devel-bounces+patchwork-qemu-devel=patchwork.kernel.org@nongnu.org Sender: "Qemu-devel" X-Virus-Scanned: ClamAV using ClamSMTP It was found that Hyper-V 2016 on KVM in some configurations (q35 machine + piix4-usb-uhci) hangs on boot. Root-cause was that one of Hyper-V level-triggered interrupt handler performs EOI before fixing the cause of the interrupt. This results in IOAPIC keep re-raising the level-triggered interrupt after EOI because irq-line remains asserted. Gory details: https://www.spinics.net/lists/kvm/msg184484.html (the whole thread). Turns out we were dealing with similar issues before; in-kernel IOAPIC implementation has commit 184564efae4d ("kvm: ioapic: conditionally delay irq delivery duringeoi broadcast") which describes a very similar issue. Steal the idea from the above mentioned commit for IOAPIC implementation in QEMU. SUCCESSIVE_IRQ_MAX_COUNT, delay and the comment are borrowed as well. Signed-off-by: Vitaly Kuznetsov Reviewed-by: Liran Alon --- Changes since v1: - timer_mod() -> timer_mod_anticipate() [Paolo Bonzini] - Massaged changelog [Liran Alon] - Make implementation look like in-kernel one [Liran Alon] --- hw/intc/ioapic.c | 57 ++++++++++++++++++++++++++++--- hw/intc/trace-events | 1 + include/hw/i386/ioapic_internal.h | 3 ++ 3 files changed, 56 insertions(+), 5 deletions(-) diff --git a/hw/intc/ioapic.c b/hw/intc/ioapic.c index 9d75f84d3b..9fb8dd3450 100644 --- a/hw/intc/ioapic.c +++ b/hw/intc/ioapic.c @@ -139,6 +139,15 @@ static void ioapic_service(IOAPICCommonState *s) } } +#define SUCCESSIVE_IRQ_MAX_COUNT 10000 + +static void ioapic_timer(void *opaque) +{ + IOAPICCommonState *s = opaque; + + ioapic_service(s); +} + static void ioapic_set_irq(void *opaque, int vector, int level) { IOAPICCommonState *s = opaque; @@ -222,13 +231,40 @@ void ioapic_eoi_broadcast(int vector) } for (n = 0; n < IOAPIC_NUM_PINS; n++) { entry = s->ioredtbl[n]; - if ((entry & IOAPIC_LVT_REMOTE_IRR) - && (entry & IOAPIC_VECTOR_MASK) == vector) { - trace_ioapic_clear_remote_irr(n, vector); - s->ioredtbl[n] = entry & ~IOAPIC_LVT_REMOTE_IRR; - if (!(entry & IOAPIC_LVT_MASKED) && (s->irr & (1 << n))) { + + if (((entry & IOAPIC_VECTOR_MASK) != vector) || + !(entry & IOAPIC_LVT_REMOTE_IRR)) { + continue; + } + + trace_ioapic_clear_remote_irr(n, vector); + s->ioredtbl[n] = entry & ~IOAPIC_LVT_REMOTE_IRR; + + if (((entry >> IOAPIC_LVT_TRIGGER_MODE_SHIFT) & 1) != + IOAPIC_TRIGGER_LEVEL) { + continue; + } + + if (!(entry & IOAPIC_LVT_MASKED) && (s->irr & (1 << n))) { + ++s->irq_eoi[vector]; + if (s->irq_eoi[vector] >= SUCCESSIVE_IRQ_MAX_COUNT) { + /* + * Real hardware does not deliver the interrupt immediately + * during eoi broadcast, and this lets a buggy guest make + * slow progress even if it does not correctly handle a + * level-triggered interrupt. Emulate this behavior if we + * detect an interrupt storm. + */ + s->irq_eoi[vector] = 0; + timer_mod_anticipate(s->timer, + qemu_clock_get_ns(QEMU_CLOCK_VIRTUAL) + + NANOSECONDS_PER_SECOND / 100); + trace_ioapic_eoi_delayed_reassert(vector); + } else { ioapic_service(s); } + } else { + s->irq_eoi[vector] = 0; } } } @@ -401,6 +437,8 @@ static void ioapic_realize(DeviceState *dev, Error **errp) memory_region_init_io(&s->io_memory, OBJECT(s), &ioapic_io_ops, s, "ioapic", 0x1000); + s->timer = timer_new_ns(QEMU_CLOCK_VIRTUAL, ioapic_timer, s); + qdev_init_gpio_in(dev, ioapic_set_irq, IOAPIC_NUM_PINS); ioapics[ioapic_no] = s; @@ -408,6 +446,14 @@ static void ioapic_realize(DeviceState *dev, Error **errp) qemu_add_machine_init_done_notifier(&s->machine_done); } +static void ioapic_unrealize(DeviceState *dev, Error **errp) +{ + IOAPICCommonState *s = IOAPIC_COMMON(dev); + + timer_del(s->timer); + timer_free(s->timer); +} + static Property ioapic_properties[] = { DEFINE_PROP_UINT8("version", IOAPICCommonState, version, IOAPIC_VER_DEF), DEFINE_PROP_END_OF_LIST(), @@ -419,6 +465,7 @@ static void ioapic_class_init(ObjectClass *klass, void *data) DeviceClass *dc = DEVICE_CLASS(klass); k->realize = ioapic_realize; + k->unrealize = ioapic_unrealize; /* * If APIC is in kernel, we need to update the kernel cache after * migration, otherwise first 24 gsi routes will be invalid. diff --git a/hw/intc/trace-events b/hw/intc/trace-events index a28bdce925..90c9d07c1a 100644 --- a/hw/intc/trace-events +++ b/hw/intc/trace-events @@ -25,6 +25,7 @@ apic_mem_writel(uint64_t addr, uint32_t val) "0x%"PRIx64" = 0x%08x" ioapic_set_remote_irr(int n) "set remote irr for pin %d" ioapic_clear_remote_irr(int n, int vector) "clear remote irr for pin %d vector %d" ioapic_eoi_broadcast(int vector) "EOI broadcast for vector %d" +ioapic_eoi_delayed_reassert(int vector) "delayed reassert on EOI broadcast for vector %d" ioapic_mem_read(uint8_t addr, uint8_t regsel, uint8_t size, uint32_t val) "ioapic mem read addr 0x%"PRIx8" regsel: 0x%"PRIx8" size 0x%"PRIx8" retval 0x%"PRIx32 ioapic_mem_write(uint8_t addr, uint8_t regsel, uint8_t size, uint32_t val) "ioapic mem write addr 0x%"PRIx8" regsel: 0x%"PRIx8" size 0x%"PRIx8" val 0x%"PRIx32 ioapic_set_irq(int vector, int level) "vector: %d level: %d" diff --git a/include/hw/i386/ioapic_internal.h b/include/hw/i386/ioapic_internal.h index 9848f391bb..70f9fc750a 100644 --- a/include/hw/i386/ioapic_internal.h +++ b/include/hw/i386/ioapic_internal.h @@ -96,6 +96,7 @@ typedef struct IOAPICCommonClass { SysBusDeviceClass parent_class; DeviceRealize realize; + DeviceUnrealize unrealize; void (*pre_save)(IOAPICCommonState *s); void (*post_load)(IOAPICCommonState *s); } IOAPICCommonClass; @@ -111,6 +112,8 @@ struct IOAPICCommonState { uint8_t version; uint64_t irq_count[IOAPIC_NUM_PINS]; int irq_level[IOAPIC_NUM_PINS]; + int irq_eoi[IOAPIC_NUM_PINS]; + QEMUTimer *timer; }; void ioapic_reset_common(DeviceState *dev);