From: Alex Williamson <alex.williamson@redhat.com>
To: qemu-devel@nongnu.org
Cc: alex.williamson@redhat.com
Subject: [Qemu-devel] [PULL 3/6] vfio: free dynamically-allocated data in instance_finalize
Date: Tue, 10 Feb 2015 11:37:28 -0700 [thread overview]
Message-ID: <20150210183728.26101.79729.stgit@gimli.home> (raw)
In-Reply-To: <20150210183106.26101.78658.stgit@gimli.home>
From: Paolo Bonzini <pbonzini@redhat.com>
In order to enable out-of-BQL address space lookup, destruction of
devices needs to be split in two phases.
Unrealize is the first phase; once it complete no new accesses will
be started, but there may still be pending memory accesses can still
be completed.
The second part is freeing the device, which only happens once all memory
accesses are complete. At this point the reference count has dropped to
zero, an RCU grace period must have completed (because the RCU-protected
FlatViews hold a reference to the device via memory_region_ref). This is
when instance_finalize is called.
Freeing data belongs in an instance_finalize callback, because the
dynamically allocated memory can still be used after unrealize by the
pending memory accesses.
This starts the process by creating an instance_finalize callback and
freeing most of the dynamically-allocated data in instance_finalize.
Because instance_finalize is also called on error paths or also when
the device is actually not realized, the common code needs some changes
to be ready for this. The error path in vfio_initfn can be simplified too.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
---
hw/vfio/common.c | 5 ++++-
hw/vfio/pci.c | 24 ++++++++++++++----------
2 files changed, 18 insertions(+), 11 deletions(-)
diff --git a/hw/vfio/common.c b/hw/vfio/common.c
index 6e299b3..b230ef1 100644
--- a/hw/vfio/common.c
+++ b/hw/vfio/common.c
@@ -847,7 +847,7 @@ free_group_exit:
void vfio_put_group(VFIOGroup *group)
{
- if (!QLIST_EMPTY(&group->device_list)) {
+ if (!group || !QLIST_EMPTY(&group->device_list)) {
return;
}
@@ -902,6 +902,9 @@ int vfio_get_device(VFIOGroup *group, const char *name,
void vfio_put_base_device(VFIODevice *vbasedev)
{
+ if (!vbasedev->group) {
+ return;
+ }
QLIST_REMOVE(vbasedev, next);
vbasedev->group = NULL;
trace_vfio_put_base_device(vbasedev->fd);
diff --git a/hw/vfio/pci.c b/hw/vfio/pci.c
index 23fe9fa..0271c80 100644
--- a/hw/vfio/pci.c
+++ b/hw/vfio/pci.c
@@ -3248,7 +3248,7 @@ static int vfio_initfn(PCIDevice *pdev)
ret = vfio_populate_device(vdev);
if (ret) {
- goto out_put;
+ return ret;
}
/* Get a copy of config space */
@@ -3258,7 +3258,7 @@ static int vfio_initfn(PCIDevice *pdev)
if (ret < (int)MIN(pci_config_size(&vdev->pdev), vdev->config_size)) {
ret = ret < 0 ? -errno : -EFAULT;
error_report("vfio: Failed to read device config space");
- goto out_put;
+ return ret;
}
/* vfio emulates a lot for us, but some bits need extra love */
@@ -3290,7 +3290,7 @@ static int vfio_initfn(PCIDevice *pdev)
ret = vfio_early_setup_msix(vdev);
if (ret) {
- goto out_put;
+ return ret;
}
vfio_map_bars(vdev);
@@ -3329,17 +3329,24 @@ out_teardown:
pci_device_set_intx_routing_notifier(&vdev->pdev, NULL);
vfio_teardown_msi(vdev);
vfio_unmap_bars(vdev);
-out_put:
+ return ret;
+}
+
+static void vfio_instance_finalize(Object *obj)
+{
+ PCIDevice *pci_dev = PCI_DEVICE(obj);
+ VFIOPCIDevice *vdev = DO_UPCAST(VFIOPCIDevice, pdev, pci_dev);
+ VFIOGroup *group = vdev->vbasedev.group;
+
g_free(vdev->emulated_config_bits);
+ g_free(vdev->rom);
vfio_put_device(vdev);
vfio_put_group(group);
- return ret;
}
static void vfio_exitfn(PCIDevice *pdev)
{
VFIOPCIDevice *vdev = DO_UPCAST(VFIOPCIDevice, pdev, pdev);
- VFIOGroup *group = vdev->vbasedev.group;
vfio_unregister_err_notifier(vdev);
pci_device_set_intx_routing_notifier(&vdev->pdev, NULL);
@@ -3349,10 +3356,6 @@ static void vfio_exitfn(PCIDevice *pdev)
}
vfio_teardown_msi(vdev);
vfio_unmap_bars(vdev);
- g_free(vdev->emulated_config_bits);
- g_free(vdev->rom);
- vfio_put_device(vdev);
- vfio_put_group(group);
}
static void vfio_pci_reset(DeviceState *dev)
@@ -3440,6 +3443,7 @@ static const TypeInfo vfio_pci_dev_info = {
.instance_size = sizeof(VFIOPCIDevice),
.class_init = vfio_pci_dev_class_init,
.instance_init = vfio_instance_init,
+ .instance_finalize = vfio_instance_finalize,
};
static void register_vfio_pci_dev_type(void)
next prev parent reply other threads:[~2015-02-10 18:37 UTC|newest]
Thread overview: 8+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-02-10 18:37 [Qemu-devel] [PULL 0/6] vfio fixes Alex Williamson
2015-02-10 18:37 ` [Qemu-devel] [PULL 1/6] memory: unregister AddressSpace MemoryListener within BQL Alex Williamson
2015-02-10 18:37 ` [Qemu-devel] [PULL 2/6] vfio: cleanup vfio_get_device error path, remove vfio_populate_device callback Alex Williamson
2015-02-10 18:37 ` Alex Williamson [this message]
2015-02-10 18:37 ` [Qemu-devel] [PULL 4/6] vfio: unmap and free BAR data in instance_finalize Alex Williamson
2015-02-10 18:37 ` [Qemu-devel] [PULL 5/6] vfio: Use vfio type1 v2 IOMMU interface Alex Williamson
2015-02-10 18:37 ` [Qemu-devel] [PULL 6/6] vfio: Fix debug message compile error Alex Williamson
2015-02-11 6:03 ` [Qemu-devel] [PULL 0/6] vfio fixes Peter Maydell
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=20150210183728.26101.79729.stgit@gimli.home \
--to=alex.williamson@redhat.com \
--cc=qemu-devel@nongnu.org \
/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).