From: Eric Auger <eric.auger@redhat.com>
To: eric.auger.pro@gmail.com, eric.auger@redhat.com,
qemu-devel@nongnu.org, qemu-arm@nongnu.org,
peter.maydell@linaro.org, mst@redhat.com, jasowang@redhat.com,
imammedo@redhat.com, peterx@redhat.com,
alex.williamson@redhat.com, clg@redhat.com, philmd@linaro.org,
zhenzhong.duan@intel.com, ddutile@redhat.com
Subject: [PATCH v3 3/5] hw/arm/smmuv3: Move reset to exit phase
Date: Tue, 18 Feb 2025 19:25:33 +0100 [thread overview]
Message-ID: <20250218182737.76722-4-eric.auger@redhat.com> (raw)
In-Reply-To: <20250218182737.76722-1-eric.auger@redhat.com>
Currently the iommu may be reset before the devices
it protects. For example this happens with virtio-scsi-pci.
when system_reset is issued from qmp monitor: spurious
"virtio: zero sized buffers are not allowed" warnings can
be observed. This happens because outstanding DMA requests
are still happening while the SMMU gets reset.
This can also happen with VFIO devices. In that case
spurious DMA translation faults can be observed on host.
Make sure the SMMU is reset in the 'exit' phase after
all DMA capable devices have been reset during the 'enter'
or 'hold' phase.
Signed-off-by: Eric Auger <eric.auger@redhat.com>
Reviewed-by: Zhenzhong Duan <zhenzhong.duan@intel.com>
---
v1 -> v2:
- also migrate smmu_base_reset_hold to smmu_base_reset_exit
which was an oversight of v1
- add a comment emphasizing the importance of doing the reset
in 'exit' phase.
---
hw/arm/smmu-common.c | 9 +++++++--
hw/arm/smmuv3.c | 14 ++++++++++----
hw/arm/trace-events | 1 +
3 files changed, 18 insertions(+), 6 deletions(-)
diff --git a/hw/arm/smmu-common.c b/hw/arm/smmu-common.c
index dd74c2e558..8c1b407b82 100644
--- a/hw/arm/smmu-common.c
+++ b/hw/arm/smmu-common.c
@@ -924,7 +924,12 @@ static void smmu_base_realize(DeviceState *dev, Error **errp)
}
}
-static void smmu_base_reset_hold(Object *obj, ResetType type)
+/*
+ * Make sure the IOMMU is reset in 'exit' phase after
+ * all outstanding DMA requests have been quiesced during
+ * the 'enter' or 'hold' reset phases
+ */
+static void smmu_base_reset_exit(Object *obj, ResetType type)
{
SMMUState *s = ARM_SMMU(obj);
@@ -949,7 +954,7 @@ static void smmu_base_class_init(ObjectClass *klass, void *data)
device_class_set_props(dc, smmu_dev_properties);
device_class_set_parent_realize(dc, smmu_base_realize,
&sbc->parent_realize);
- rc->phases.hold = smmu_base_reset_hold;
+ rc->phases.exit = smmu_base_reset_exit;
}
static const TypeInfo smmu_base_info = {
diff --git a/hw/arm/smmuv3.c b/hw/arm/smmuv3.c
index c0cf5df0f6..b49a59b64c 100644
--- a/hw/arm/smmuv3.c
+++ b/hw/arm/smmuv3.c
@@ -1870,13 +1870,19 @@ static void smmu_init_irq(SMMUv3State *s, SysBusDevice *dev)
}
}
-static void smmu_reset_hold(Object *obj, ResetType type)
+/*
+ * Make sure the IOMMU is reset in 'exit' phase after
+ * all outstanding DMA requests have been quiesced during
+ * the 'enter' or 'hold' reset phases
+ */
+static void smmu_reset_exit(Object *obj, ResetType type)
{
SMMUv3State *s = ARM_SMMUV3(obj);
SMMUv3Class *c = ARM_SMMUV3_GET_CLASS(s);
- if (c->parent_phases.hold) {
- c->parent_phases.hold(obj, type);
+ trace_smmu_reset_exit();
+ if (c->parent_phases.exit) {
+ c->parent_phases.exit(obj, type);
}
smmuv3_init_regs(s);
@@ -1999,7 +2005,7 @@ static void smmuv3_class_init(ObjectClass *klass, void *data)
SMMUv3Class *c = ARM_SMMUV3_CLASS(klass);
dc->vmsd = &vmstate_smmuv3;
- resettable_class_set_parent_phases(rc, NULL, smmu_reset_hold, NULL,
+ resettable_class_set_parent_phases(rc, NULL, NULL, smmu_reset_exit,
&c->parent_phases);
device_class_set_parent_realize(dc, smmu_realize,
&c->parent_realize);
diff --git a/hw/arm/trace-events b/hw/arm/trace-events
index c64ad344bd..7790db780e 100644
--- a/hw/arm/trace-events
+++ b/hw/arm/trace-events
@@ -56,6 +56,7 @@ smmuv3_config_cache_inv(uint32_t sid) "Config cache INV for sid=0x%x"
smmuv3_notify_flag_add(const char *iommu) "ADD SMMUNotifier node for iommu mr=%s"
smmuv3_notify_flag_del(const char *iommu) "DEL SMMUNotifier node for iommu mr=%s"
smmuv3_inv_notifiers_iova(const char *name, int asid, int vmid, uint64_t iova, uint8_t tg, uint64_t num_pages, int stage) "iommu mr=%s asid=%d vmid=%d iova=0x%"PRIx64" tg=%d num_pages=0x%"PRIx64" stage=%d"
+smmu_reset_exit(void) ""
# strongarm.c
strongarm_uart_update_parameters(const char *label, int speed, char parity, int data_bits, int stop_bits) "%s speed=%d parity=%c data=%d stop=%d"
--
2.47.1
next prev parent reply other threads:[~2025-02-18 18:29 UTC|newest]
Thread overview: 7+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-02-18 18:25 [PATCH v3 0/5] Fix vIOMMU reset order Eric Auger
2025-02-18 18:25 ` [PATCH v3 1/5] hw/virtio/virtio-iommu: Migrate to 3-phase reset Eric Auger
2025-02-18 18:25 ` [PATCH v3 2/5] hw/i386/intel-iommu: " Eric Auger
2025-02-18 18:25 ` Eric Auger [this message]
2025-02-18 18:25 ` [PATCH v3 4/5] hw/vfio/common: Add a trace point in vfio_reset_handler Eric Auger
2025-02-18 18:25 ` [PATCH v3 5/5] docs/devel/reset: Document reset expectations for DMA and IOMMU Eric Auger
2025-02-18 22:31 ` [PATCH v3 0/5] Fix vIOMMU reset order Peter Xu
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=20250218182737.76722-4-eric.auger@redhat.com \
--to=eric.auger@redhat.com \
--cc=alex.williamson@redhat.com \
--cc=clg@redhat.com \
--cc=ddutile@redhat.com \
--cc=eric.auger.pro@gmail.com \
--cc=imammedo@redhat.com \
--cc=jasowang@redhat.com \
--cc=mst@redhat.com \
--cc=peter.maydell@linaro.org \
--cc=peterx@redhat.com \
--cc=philmd@linaro.org \
--cc=qemu-arm@nongnu.org \
--cc=qemu-devel@nongnu.org \
--cc=zhenzhong.duan@intel.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).